Citation
Marijuana Use among HIV-Positive and HIV-Negative Men Who Have Sex with Men: Long Term Trends, Predictors of Use and Impact on Cognitive Function

Material Information

Title:
Marijuana Use among HIV-Positive and HIV-Negative Men Who Have Sex with Men: Long Term Trends, Predictors of Use and Impact on Cognitive Function
Creator:
Okafor, Chukwuemeka N
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Epidemiology
Committee Chair:
COOK,ROBERT L
Committee Co-Chair:
CHEN,XINGUANG
Committee Members:
GOODENOW,MAUREEN M
COHEN,RONALD A
PLANKEY,MICHAEL

Subjects

Subjects / Keywords:
cognition
hiv
marijuana
msm
prevalence

Notes

General Note:
Marijuana use is common among HIV positive individuals and although in some individuals its use has been associated with therapeutic benefits, its use is also associated with adverse clinical and behavioral outcomes. In recent years, prevalence of marijuana use in the United States has increased. The objectives of this dissertation are to: (1) examine trends in the prevalence and predictors of marijuana use; (2) identify distinct individual patterns and predictors of these patterns of marijuana use and (3) determine the long-term effect of marijuana use on cognitive performance. In study 1, we performed multivariable Poisson regressions to examine prevalence and predictors of marijuana use over a 29-year period (1984-2013). We found that the prevalence of marijuana use declined over time, but that daily use (among users) increased during the same time period. We also found that HIV positive individuals had higher prevalence of use over time. Study 2 identified four distinct individual patterns of marijuana use over a 29-year period (1984-2013); abstainer/infrequent (65%), decreaser (13%), increaser (12%) and chronic high (10%) use groups. Among the HIV positive individuals, having a detectable viral load was associated with increasing marijuana use in individuals in the increaser group only. In study 3, we did not find any statistically significant association between greater cumulative exposure to marijuana use and performance on three cognitive tests of processing speed and executive function. These three studies increased knowledge on marijuana use among HIV positive individuals.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
8/31/2018

Downloads

This item has the following downloads:


Full Text

PAGE 1

MARIJUANA USE AMONG HIV POSITIVE AND HIV NEGATIVE MEN WHO HAVE SEX WITH MEN: LONG TERM TRENDS, PREDICTORS OF USE AND IMPACT ON COGNITIVE FUNCTION By CHUKWU EMEKA N OKAFOR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2016

PAGE 2

2016 Chukwue meka N Okafor

PAGE 3

To my parents Dr. Raymond Okafor and C ecilia Okafor

PAGE 4

4 ACKNOWLEDGMENTS Firs t and foremost, I give my praises and thanks to God, the Almighty, for favor, blessing and the might that saw me through this research work. I must thank my parents, Dr. Raymond Okafor and Cecilia Okafor for all the love, support, encouragement and your inspiration throughout these years. Frankly, this would not have been possible without you. I must also express my deep appreciation and gratitude to my mentor and dissertation chair, Dr. Robert L Cook for his patienc e, support and guidance this past four years. I have learned so much from you and I feel well poised for a successful scientific career. I am truly fortunate to have had the opportunity to work with him. It will be remiss if I do not thank my committee me mbers: Drs. Michael Plankey, Xinguang Chen, Ronald Cohen and Maureen Goodenow your advice and guidance on this work. I must also thank my siblings, Ebele Obi, Ugo Okafor, Chidinma Ejim and Onyinye Okafor for their words of encouragement that kept me going

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIG URES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW ................................ ..................... 11 The HIV Epidemic ................................ ................................ ................................ ... 11 The Intersection of Drug Use and HIV Infection ................................ ...................... 13 Pharmacodynamics of Marijuana ................................ ................................ ............ 15 Prevalence and Predictors of Marijuana Use in HIV Positive Individuals ................ 17 Trajectories of Marijuana Use ................................ ................................ ................. 20 Clinical Effects of Marijuana Use in HIV Positive Individuals ................................ .. 22 Neurocognitive Effects of Marijuana Use ................................ ................................ 25 Neuropathogenesis of HIV Associated Cognitive Impairments ............................... 29 HIV Associated Neurocognitive Disorders ................................ .............................. 30 Marijuana Use, HIV Infection and Cognitive Function ................................ ............ 34 Significance of Marijuana Use on the Cognitive Functions ................................ ..... 36 2 STATEMENT OF SPECIFIC AIMS AND HYPOTHESIS ................................ ........ 38 Aim 1 Trends in the Prevalence of and Risk Factors for Marijuana Use ................. 38 Aim 2 Trajectories of Marijuana Use and Predictors of Trajectories ....................... 39 Aim 3 Cumulative Exposure to Marijuana and Cognitive Change over 17 years .... 40 3 TRENDS IN THE PREVALENCE OF AND RISK FACTORS FOR MARIJUANA USE ................................ ................................ ................................ ........................ 41 Methods ................................ ................................ ................................ .................. 43 The Multicenter AIDS Cohort Study (MACS) ................................ .................... 43 Participants ................................ ................................ ................................ ....... 44 Measures ................................ ................................ ................................ .......... 45 Covariates ................................ ................................ ................................ ........ 45 Data Analysis ................................ ................................ ................................ .......... 48 Results ................................ ................................ ................................ .................... 49 Sample Characteristics at Baseline ................................ ................................ .. 49 Trends in the Prevalence of Marijuana ................................ ............................. 50 Factors Associated Marijuana Use ................................ ................................ ... 51 Discussion ................................ ................................ ................................ .............. 52

PAGE 6

6 4 TRAJECTORIES OF MARIJUANA USE AND PREDICTORS OF TRAJECTORIES ................................ ................................ ................................ .... 63 Methods ................................ ................................ ................................ .................. 65 The Multicenter AIDS Cohort Study (MACS) ................................ .................... 65 Measures ................................ ................................ ................................ .......... 67 Covariate s ................................ ................................ ................................ ........ 67 Data Analysis ................................ ................................ ................................ .......... 69 Results ................................ ................................ ................................ .................... 71 Sample Characteristics ................................ ................................ ..................... 71 Marijuana Trajectories ................................ ................................ ...................... 72 Time Stable and Time Varying Factors Associated with Trajectories .............. 73 Among All Men ................................ ................................ ........................... 73 Among HIV positive Participants ................................ ................................ 75 Discussion ................................ ................................ ................................ .............. 75 5 CUMULATIVE MARIJUANA USE AND COGNTIVE CHANGE OVER 17 YEARS ................................ ................................ ................................ .................... 89 Methods ................................ ................................ ................................ .................. 92 The Multicenter AIDS Cohort Study (MACS) ................................ .................... 92 Subjects ................................ ................................ ................................ ............ 92 Primary Predictor: Marijuana Exposure ................................ ............................ 9 3 Outcome: Neuropsychological Evaluation ................................ ........................ 94 Covariates ................................ ................................ ................................ ........ 96 Data Analysis ................................ ................................ ................................ .......... 98 Results ................................ ................................ ................................ .................. 100 Sample Characteristics ................................ ................................ ................... 100 Cumul ative Marijuana Exposure and Changes in Cognitive Performance ..... 101 Discussion ................................ ................................ ................................ ............ 102 Conclusions ................................ ................................ ................................ .......... 106 6 DISCUSSION AND CONCLUSIONS ................................ ................................ ... 119 S ummary ................................ ................................ ................................ .............. 119 Study 1 Trend in the Prevalence and Predictors of Marijuana Use ....................... 119 Study 2 Long Term Trajectories of Marijuana use ................................ ................ 122 Study 3 Cumulative Exposures to Marijuana and Cognitive Change .................... 123 Strengths and Limitations ................................ ................................ ..................... 125 Implications for Future Research ................................ ................................ .......... 126 APPENDIX :SELECTIVE ATTRITION ................................ ................................ .......... 128 REFERENCES ................................ ................................ ................................ ............ 129 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 156

PAGE 7

7 LIST OF TABLES Table page 3 1 Characteristics of MACS Participants at Baseline by Enrollment Cohort ............ 56 3 2 Prevalence Ratios of Risk Factors Associated with Current Marijuana Use among All Men ................................ ................................ ................................ .... 57 3 3 Prevalence Ratios of Risk Factors Associated with Current Marijuana Use among HIV positive Men ................................ ................................ .................... 58 3 4 Prevalence Ratios of Risk Factors Associated with Daily Marijuana Use among All Men ................................ ................................ ................................ .... 61 3 5 Prevalence Ratios of Risk Factors Associated with Daily Marijuana Use among HIV positive Men ................................ ................................ .................... 62 4 1 Characteristics of M ACS Men Used in The Trajectories at Baseline .................. 81 4 2 Characteristics of MACS Men at Baseline by Marijuana Trajectories ................. 82 4 3 ................................ ... 83 4 4 Risk Factors for Marijuana Trajectories among HIV .................... 84 4 5 Baseline Characteristics of MACS Men Included and Excluded From the Analysis ................................ ................................ ................................ .............. 85 4 6 Model Fit Statistics for Marijuana Trajectories for All Men ................................ 86 4 7 Average Posterior Probabilities of the Final Four Marijuana Trajectory Solution for All Men ................................ ................................ ............................ 86 4 8 Model Fit Statistics for Trajectories for HIV positive Men ................................ ... 86 4 9 Average Posterior Probabilities of the Final Four Ma rijuana Trajectory Solution for the HIV positive Men ................................ ................................ ....... 86 5 1 Characteristics of MACS Men Included in the Study at Baseline by HIV Status and Marijuana Use ................................ ................................ ................ 110 5 2 Association between cumulative exposure to marijuana use a and changes in cognitive performance among HIV Positive and HIV Negative Participants: MACS 1996 to 2013 ................................ ................................ ......................... 112 5 3 Baseline Characteristics of MACS Men Included and Excluded from Analyses ................................ ................................ ................................ ........... 118

PAGE 8

8 LIST OF FIGURES Figure page 3 1 Annual Prevalence of Current and Daily (Among Current Users) Marijuana Use Among HIV positive and HIV negative MSM in The MACS: Early Cohort .. 59 3 2 Annual Prevalence of Current and Daily (Among Current Users) Marijuana Use Among HIV positive and HIV negative MSM in The MACS: Late Cohort ... 60 4 1 Trajectories of marijuana use among 3,658 HIV positive and HIV negative participants in the Multicenter AIDS Cohort Study (MACS) 1984 2013. ............. 87 4 2 Trajectories of marijuana use among 1,439 HIV+ participants in the Multicenter AIDS Cohort Study (MACS) 1984 2013. ................................ .......... 88 5 1 Flow chart of MACS participants included in the study ................................ ..... 108 5 3 Predicted Z scores on the Trail Making Test Part A by HIV Serostatus ............ 113 5 4 Predicted Z scores on the Trail Making Test Part A by HIV Serostatus ............ 114 5 5 Predicted Z scores on the Symbol Digit Test by HIV Serostatus ...................... 115 5 6 Predicted Z s cores on the Symbol Digit Test HIV positive ............................... 116 5 7 Predicted Z scores on the Symbol Digit Test HIV negative .............................. 117

PAGE 9

9 Abstract of Dissertation Present ed to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MARIJUANA USE AMONG HIV POSITIVE AND HIV NEGATIVE MEN WHO HAVE SEX WITH MEN: LONG TERM TRENDS, PREDICT ORS OF USE AND IMPACT ON COGNITIVE FUNCTION By Chukwue meka N Okafor August 2016 Chair: Robert L Cook Major: Epidemiology Marijuana use is common among HIV positive individuals and although in some individuals its use has been associated with therapeut ic benefits, its use is also associated with adverse clinical and behavioral outcomes. Very little is currently known the prevalence of and factors that contribute to marijuana use in HIV positive individuals. There are also few published reports on indivi dual patterns of marijuana use an d predictors of these pattern s Importantly, data on the effect of long term effect of marijuana use on cognitive function is limited T he objectives of this dissertation are to: (1) examine trends in the prevalence and pre dictors of marijuana use; (2) identify distinct individual patterns and predictors of these patterns of marijuana use and (3) determine the long term effect of marijuana use on cognitive performance. Study 1 showed that the prevalence of marijuana use dec lined, but daily use (among users) increased over time (1984 2013). The results also demonstrated that HIV positive as compared to HIV negative individuals had higher prevalence of marijuana use over time. In s tudy 2 a growth mixture modelling approach wa s used to identif y four distinct individual patterns of marijuana use over a 29 year period (1984

PAGE 10

10 2013); abstainer/infrequent (65%), decreaser (13%), increaser (12%) and chronic high (10%) use groups. Among the HIV positive individuals, having a detectable viral load was associated with increasing marijuana use in individuals in the increaser group only. In study 3, cumulative marijuana exposure was expressed as marijuana use years. In fully adjusted l inear mixed effect regression models there were no sign ificant adverse associations between cumulative exposure to marijuana and 17 year changes in measures of cognitive function including proce ssing speed, executive function and verbal memory The findings from the three studies have advanced our knowledge of marijuana use among HIV positive individuals and its effects on cognitive performance. Our findings also inform areas of future investigations

PAGE 11

11 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW The HIV Epidemic Human immunodeficiency virus (HIV) i s the virus that causes Acquired Immunodeficiency Syndrome (AIDS). HIV causes progressive failure of the immune system (decreased in CD4+ cells but also other aspects of immunity including B cell function and innate inflammation ) with the emergence of lif e threatening opportunistic infections. Studies have estimated that, if left untreated, the median time from initial infection until death is about 10 years (Morgan et al., 2002) Since the epidemic began in 1981, HIV has claimed nearly 675, 000 lives in the United States (CDC, 2015) and 25 million lives worldwide (WHO, 2013) Since the introduction of combination a ntiretroviral therapy (cART ) in the mid1990s, death from HIV/AI DS has drastically declined and the life expectancy of HIV positive individuals in North America is nearly similar to the general uninfected population (Palella et al., 1998; Samji et al., 2013) The Centers for Disease Control and Prevention (CDC) estimates that more than 1.2 million people are living with HIV infection in the Uni ted States, including nearly 156 000 (12.8%) who are unaware of their infection (CDC, 2012) Because of treatment advances, more HIV positive individuals are living longe r and thus the number of HIV positive individuals has increased dramatically. However, the n umber of new infections has remained relatively stable at about 50, 000 new infections per year (CDC, 2015) Current trends indicate that s ome groups are disproportionately represented in the HIV/AIDS epidemic : men who have sex wi th men (MSM ) represent the group with the largest number of new infections and burde n of HIV infection. By race/ethnicity African Americans and Hispanics are the most heavily affected in the United States (CDC,

PAGE 12

12 2015) The greater burden of HIV in these groups are in part due to a number of social and economic challenges including lack of access to care, discrimination, stigma, homophobia, incarceration and poverty (CDC, 2015) Although there is current ly no cure for HIV, advances in treatment, including the availability of cART can help HIV positive individuals reduce the amount of the virus in their body and can also significantly reduce the risk of transmitting HIV to others (Cohe n et al., 2011) In fact, in the current cART era, clinical care for of HIV positive individuals has shifted from an acute infection with high mortality rates to the management of a chronic illness. These developments have changed the approach to HIV prev ention in the United States. The vision of the National HIV/AIDS Strategy is to ensure that everyone with HIV is aware of his or her infection and receives treatment. In addition, impact HIV prevention strategy, aims to achieve the greatest possible reductions in HIV infections by making sure that resources go to regions, populations and prevention strategies where they will have the greatest impact (Gardner, McLees, Steiner, Rio, & B urman, 2011) In July 2013, President Obama established the HIV Care Continuum Initiative -. The HIV Care Continuum or sometimes referred to as the HIV Treatment Cascade is a model that outlines the sequential steps of HIV medical care that HIV positive individuals go through from initial diagnosis to the goal of viral suppression and shows the pr oportion of HIV positive individuals who are engaged at each step. The four major steps of the HIV care continuum include HIV testing and diagnosis, getting and keeping HIV positive individuals in medical care, prescribing ART and helping HIV positive indi viduals achieve viral suppression. According to the most recent CDC data (2015) of

PAGE 13

13 the approximately 1.2 million of HIV positive individuals in the United States, 86%, 40%, 37% and 30% were diagnosed, engaged in medical care, prescribed ART and achieved vi ral suppression respectively (CDC, 2014) Also, the World Health Organization (WHO) now recomm ends that ART should be initiated among all adults with HIV regardless of CD4 cell counts (World Health Organization, 2015) The Intersection o f Drug Use a nd H IV Infection Since the beginning of the HIV epidemic, researchers have linked d rug use an d abuse to HIV/AIDS. It was recognized that HIV was spreading among injection drug users (IDU) (Masur et al., 1981) The sharing of contaminated needles, syringes and injection paraphernalia provide one of the most effective tra nsmission of HIV (Heimer, Myers, Cadman, & Kaplan, 1992) In 2006, one study estimated that approximately 3 million IDUs might be living with HIV worldwide. In the United States, the prevalence of HIV infection among IDUs is estimat ed to be about 16% (Mathers et al., 2008) Alcohol use and other illicit drug use are also common among of HIV positive Study (WIHS) the largest multisite longitudinal cohort study of women living with or at risk for HIV infection in the United States during the period 1994 2006, 50% reported ever using alcohol, including 11% who were either persistent heavy drinkers or increased to heavy drinking over time (Cook et al., 2013 ) Among men living with HIV, prevalence of any alcohol use is even higher: between 46% and 85% reporting any alcohol use in the past six months, with 12.5% reporting consistent hazardous drinking (Mar shall et al., 2014, 2015) Prevalence of smoking among HIV positive individuals has been estimated to be between 25% and 51% (Akhtar Harrell, & Martins, 2014) although recent data indicate that the rates are on a decline

PAGE 14

14 (Akhtar Khaleel et al., 2015; Hessol et al., 2014) In the WIHS, 12% reported using cocaine at their baseline visit In a recent study including over 3,000 HIV positive individuals receiving medical care across four cities in the United States, 9% used amphetamines and 8.5% used crack cocaine in the past 3 months (Mimiaga et al. 2013) Alcohol use particularly binge drinking (defined as 5 or more drinks for men and 4 or more drinks for women at a sitting) onal Institute of Alcohol Abuse NIAAA Publications Distribution Center; 2005.,.) has been associated with high risky sexual behaviors and having a sexual transmitted disease (Cook & Clark, 2005; Hutton, McCaul, Santora, & Erbelding, 2008) Alcohol (Young, Wood, Dong, Kerr, & Hayashi, 2015) methamphetamine, poppers and other stimulant use (Ostrow et al., 2009; Plankey et al., 2007) have been associated with increased risk for HIV seroconversion. Alcohol, smoking and illicit drug use has been linked to decreased ART adh erence (Arnsten et al., 2002; Chander, La u, & Moore, 2006; Newville, Berg, & Gonzalez, 2014; lack of virologic suppression (Baum et al., 2009; Chander et al., cognitive impairment (A. M. Anderson, Higgins, Ownby, & Waldrop Valverde, 2015; Cristiani, Pukay Martin, & Bor nstein, 2004; Gonzalez, Schuster, Vassileva, & Martin, 2011; Meade, Towe, Skalski, & Robertson, 2015; Rothlind et al., 2005) poorer quality of life (Allshouse et al., 2015) and increased HIV disease progression (Baum et al., 2009) Alcohol use has also been associated with worsening of HIV associated neuropathy (Lopez, Becker, Dew, & Caldararo, 2004) and faster liver disease progression and liver associated mortality

PAGE 15

15 (Cooper & Cameron, 2005) In one study, smoking cigarettes was strongly associated with cardiovascular risk and multimorbidities including hypertension, liver disease and diabetes (Hasse et al., 2015) Marijuana use is also increasingly common among HIV positive individuals. Its use has been associated with the relief of a wide range of symptoms including nausea, pain, lack of appetite, depressed mood and sl eep problems many of which HIV positive individuals frequently experience (Corless et al., 2009; Furler, Einarson, Millson, Walmsley, & Bendayan, 2004) However, there is currently limited empirical data to support the efficac y of the clinical benefits of marijuana (Lutge, Gray, & Siegfried, 2013) More importantly, marijuana use has also been associated with a wide range of adverse health effects. Therefo re, it is important to continually monitor prevalence and patterns of marijuana use among HIV positive individuals to identify subgroups more likely to have patterns of harmful or unhealthy use [or potential benefit] and thus help guide targeted preventio n efforts. Pharmacodynamics of Marijuana Marijuana ( the plant) contains about 80 compounds that are unique to the plant including delta 9 tetrahydrocaannabinol (THC) which is the main constituent responsible for the psychoactive effects of marijuana. Fol lowing consumption, THC binds to specific receptors located in brain, which are termed cannabinoid receptors 1 (CB1). CB1 receptors are abundant in the basal ganglia (including the globus pallidus and substantia nigra), cerebellum, and cerebral cortex, wit h moderate concentrations in the amygdala, and hippocampus (Glass, Dragunow, & Faull, 1997; Herkenham et al., 1990) These areas of the brain are involved in a number of cognitive functions including coordinating movement/motor function, executive functions and

PAGE 16

16 learning/memory functions. CB1 receptors within the se brain regions are found mainly on presynaptic neurons which are specialized areas within the axon of the presynaptic cell that contains neurotransmitters involved in the regulation of neurotransmission (Glass et al., 1997; Herkenham et al., 1990) Activation of CB1 receptors by exogenous ligands that bind to it such as THC inhibi ts the synthesis of adenylyl cyclase an enzyme used in the production of cyclic adenosine monophosphate (cAMP) which is an important compound needed to initiate nerve impulses (Devane, Dysarz, Johnson, Melvin, & Howlett, 1988) Reduction in cAMP formation results in ma ny of the processes requiring cAMP to subsequently slow down (Bidaut Russell, Devane, & Howlett, 1990; Devane et al., 1988) In addition, activation of CB1 receptors inhibits calcium channels (Ca 2+ ) responsible for Ca 2+ influx into the neurons. Decreased influx of Ca 2+ in the neurons decreases neurotransmitter release (Pertwee, 1997) In contrast to closing Ca 2+ channels, activ ation of CB1 receptors opens potassium (K + ) channels of the neuron increasing K + conductance (McAllister & Glass, 2002; Pertwee, 1997) Activation of K + conductance produce a shunting effect and decreases the rate of neurotra nsmitters release from the neurons (McAllister & Glass, 2002; Pertwee, 1997) Therefore, THC by binding to and activating CB1 receptors and inhibiting cAMP, decre asing Ca 2+ influx and increasing K + conductance, modulate neurotransmitter release and alter normal communication between nerve cells to produce some of the cognitive effects of marijuana following administration. It is important to note that a second type of cannabinoid receptors 2 (CB2) have also been identified and are present mostly in cells and tissues of the immune system and is proposed to be activated by the binding of other major

PAGE 17

17 cannabinoids including cannabinol (CBN) and cannabidiol (CBD) (Mechoulam & Shvo, 1963; M unro, Thomas, & Abu Shaar, 1993) CB2 expression in immune cells have been proposed to play a role in the anti inflammatory effects and potential therapeutic benefits of marijuana use (Rom & Persidsky, 2013) Prevalence and Predictors of Marijuana Use in HIV Positive Individuals Past studies have reported that between 14% and 56% of HIV positive individuals report past six months marijuana use 2004; Harris et al., 2014; Mimiaga et al., 2013; Prentiss, Power, Balmas, Tzuang, & Israelski, 2004; Skeer et al., 2011; Ware, Rueda, Singer, & Kilby, 2003) Among 503 HIV positive MSM receiving primary care at Fenway Health the largest HIV care center in New England, 33.8% reported marijuana use within the past 3 months (Skeer et al., 2011) Recent data from the WIHS found 14% reporting p ast 6 months marijuana use in 2010 with prevalence of daily use to be 6.1% among all women and 51% among users In a multicenter cohort of HIV positive individuals receiving care in 4 cities in the United States between 2005 and 2010 in the Centers for AIDS Research Network of Integrated Clinical Systems, found that 24% r eported using marijuana in the past 3 months marijuana (Mimiaga et al., 2013) The United States Drug Enforcement Agency (DEA) currently classifies marijuana as an illegal schedule 1 substance. Under this classification, the DEA regards m arijuana as having no acceptable medicinal use with a high potential for addiction with long term use. However, in recent years, state laws have evolved; 23 states have passed laws allowing recreational and/or medicinal marijuana and more states are consid ering passing similar laws. Thus, it is important to understand how the present climate may contribute to increased prevalence of marijuana use. Already, data from the

PAGE 18

18 general United States population 18 years of age and older suggests that prevalence of m arijuana use increased from 4.1% in 2001 2002 to 9.5% in 2012 2013 (Hasin DS, Saha TD, Kerridge BT, & et al 2015) Few studies have assessed trends over time in the prevalence of marijuana use among HIV positive individuals. in a recent study, described the rates of marijuana use in a longitudinal cohort of HIV positive women (n=2776) and found an overall decrease (from 21% to 14%) in past year marijuana use ove r a period of 16 years ( 1994 2010). H o wever, the authors found that daily marijuana use (among past year users) increased from 18% to 51% during the sample time period. Predictors of marijuana use: Sociodemographic factors associated with marijuana use among HIV positive individuals are nearl y similar to that among the general uninfected persons. Younger age, male gender, sexual minority status, racial/ethnic minority status, lower income, lower educational status, internalizing behaviors (e.g. depressive symptoms) and externalizing behaviors (e.g. delinquency) have been associated with marijuana use (David, Catalano, & Miller, 1992) However as some HIV positive individuals use marijuana to relieve HIV related symptoms and side effects of ART (Woolridge et al., 2005) predictors of use may be different that those in the general uninfected population. Accordingly, some studies have found associations between HIV markers of disease progression and marijuana use (Abrams et al., 2003; Milloy et al., 2015) In addition, passage of medical marijuana laws (MMLs) may be associated with increased availability and easier access to marijuana and may c ontribute to increased marijuana use among HIV positive individuals. There is presently much debate on whether passage of these laws increase marijuana use in the

PAGE 19

19 general population as several studies has reported conflicting results (D. M. Anderson, Hansen, & Rees, 2015; Cerd, Wall, Keyes, Galea, & Hasin, 2012; Harper, Strumpf, & Kaufm an, 2012; Hasin DS et al., 2015; Lynne Landsman, Livingston, & Wagenaar, 2013; Wen, Hockenberry, & Cummings, 2015) However, nearly all of the studies assessing MMLs as a risk factor for increased marijuana use have been in adolescent samples. Passage of MMLs may also affect marijuana use in HIV positive individuals as nearly all of these laws list HIV/AIDS as a qualifying condition for medical marijuana There is currently no published repo rt addressing whether passage of medical marijuana law contributes to inc reased marijuana, use among HIV positive individuals. Most of aforementioned studies have been cross sectional and of those utilizing data from longitudinal studies have employed a va riable centered analytical approach with the goal of evaluating trends and predictors of marijuana use represented by population averages rather than individual variability. One d rawback to a variable centered analytical approach is that they are less ge neralizable to individuals with substantial variability in marijuana use from the population average. Thus, this type of analytical approach may not adequately characterize associations between potential risk/protective factors and marijuana use for the ma jority of individuals in a given sample (Dierker, Rose, Tan, & Li, 2010) The advancement of statistical methods now allow the use of a pattern centered (sometimes called person centered, group based or tra jectory analysis) analytical approach to uncover individual heterogeneity in substance use over an extended period of time (Muthn & Shedden, 1999; Nagin, 2005)

PAGE 20

20 Trajectories o f Marijuana Use There is widespread acceptance of the concept m ultiple trajectories or pathways of a behavioral, social or biological process over time (Cicchetti & Rogosch, 1996) With regard to marijuana use, what these suggest is that not all marijuana users follow the same d evelopment al trajectory of use over time i.e. there may be substantial individual variability in marijuana use that differ from the population average. For example, it is possible that some persons who initiate marijuana will progress over time to become persistent heavy/problematic users, whereas others will follow a persistent moderate or light use over time. Furthermore, some may not follow a stable pattern, but have patterns that fluctuate over time, such as those who reduce use with increasing age T hese different trajectories may reflect different etiological pathways toward the progression of marijuana use over time. Therefore, developmental trajectory research aims to uncover these different sub groups of individuals within a population. Uncovering these different developmental trajectories of marijuana use over time may help identify subgroups of individuals with the greatest risk of progressing to harmful/unhealthy patterns of marijuana use In addition, developmental trajectory research can help i dentify modifiable predictors of these different developmental trajectories, which may be useful in the design of target ed intervention programs to prevent or slow the progression to heavy or unhealthy patterns of use. Numerous studies have described tr ajectories and predictors of marijuana use with most studies identifying a range of 3 to 7 distinct trajectories of marijuana use (Brook, Lee, Brown, Finch, & Brook, 2011; Brown, Flory, Lynam, Leukefeld, & Clayton, 2004; Eassey, Gibson, & Krohn, 2014; Ellickson, Martino, & Collins, 2004; Juon Fothergill, Green, Doherty, & Ensminger, 2011; Nelson, Van Ryzin, & Dishion, 2015;

PAGE 21

21 Passarotti, Crane, Hedeker, & Mermelstein, 2015; Schulenberg et al., 2005; Whitesell et al., 2013; Windle & Wiesner, 2004) All of these studies have identified a group w ith no/low marijuana use and a group with persistent heavy use over time Some of these studies report on antecedents or risk factors of the different trajectories of marijuana use trajectories One study by Kuo et al. (2010) found that male late starters were more likely to be aggressive than abstainers (Kuo et al., 2004) Another more recent study found that relative to a group following a trajectory of non escalating marijuana use over time those in the escalating trajectory had higher antisocial behavior (Passarotti et al., 2015) Brook et al. (2011) found that the chronic marijuana use trajector y was associated with criminal behavior (Brook et al., 2011) Also, one study found that among those following a trajectory of persistent heavy user over time had an increased likelihood of a mari juana disorder in adulthood relative to other marijuana use trajectories (Windle & Wiesner, 2004) How ever, almost all these studies have focused on adolescents, youths and young adults i.e. within 12 through 32 years of age and thus may fail to capture patterns of use that occur later in the life course. There is currently no published data on trajectorie s of marijuana use among HIV positive individuals. Given that HIV positive individuals use marijuana for different reasons (i.e. medical versus recreational), different trajectories and predictors of the different trajector ies of marijuana use among HIV po sitive individuals may exist. The importance of a deeper understanding into how trends in the prevalence of marijuana use over time has evolved and the uncovering of different individual patterns of use cannot be overemphasized. While marijuana use has

PAGE 22

22 bee n associated with some therapeutic benefits, its use has also been associated with adverse clinical and neurocognitive effects. Clinical Effects of Marijuana Use in HIV Positive Individuals Marijuana use and antiretroviral therapy (ART) adherence. The intr oduction of antiretroviral therapy (ART) has significantly improved morbidity and mortality associated with HIV infection (R. Detels et al., 2001; HIV CAUSAL Collaboration et al., 2010; Li et al., 1998; Palella et al., 1998) As such, achieving successful treatment and improved long term outcomes of HIV positive individuals requires near perfect adherence to ART. Several studies have found decreased ART adherence among marijuana users (Bonn Miller, Oser, Bucossi, & Trafton, 2012; Corless et al., 2009; Kalichman & Rompa, 2003; Peretti Watel, Spire, Lert, & Obadia, 2006; Tucker, Burnam, Sherbourne, Kung, & Gifford, 2003; Wilson, Doxanakis, & Fairley, 2004) However this finding has not been consistently reported as some studies have found marijuana use to have no significant association with ART adherence (de Jong, Prentiss, McFarland, Machekano, & Israelski, 2005; Lucas, Cheever, Chaisson, & Moore, 2001; Rosen et al., 2013; Slawson et al., 2014) and a nother study reported that marijuana use for the managemen t of nausea (a medical use) is associated with increased ART adherence (de Jong et al., 2005) Further, recent data from the WIHS found l ower adherence to ART among current (past six months) marijuana users compared to non users, but found that ART adherence was not reduced in d aily marijuana users compared to non users Marijuana use, viral load and CD4 counts: Data from preclinical studies suggest that marijuana or other synthetic cannabinoids may have a beneficial effect on reducing HIV viral loads and replication. In vitro studies have shown that

PAGE 23

23 tetrahydrocannabidiol ( THC ) the major co nstituent of marijuana and responsible for its psychoactive effects may reduce HIV viral replication in primary CD4+ T cells (Costantino et al., 2012; Williams et al., 2014) Investigators using male rhesus macaques reported that chronic THC administration initiated prior to and throughout the asymptomatic phase of simian immunodeficienc y virus (SIV) was associated with reduction of plasma and cerebrospinal fluid viral load (Molina et al., 2011) although this protective effect was not seen in a follow up replication study among female rhesus macaques (Amedee et al., 2014) Most of the immunomodulatory effects of marijuana and its constituents are proposed to be mediated via the action of cannabinoids on cannabinoid receptors 2 ( CB2 ) which are predominantly located on pe ripheral immune cells (Rom & Persidsky, 2013) M arijuana use may have an indirect effect on viral load suppression via its potential association with reduced ART adherence. In general, studies on the association between marijuana use and viral load outcomes in HIV positive individuals have been few and conflicting Abrams et al. found no significant short term differences in plasma viral load among 67 HIV positiv e individuals who were randomly assigned either to a 3.95% THC marijuana cigarette, a 2.5mg dronabinol (delta 9 teterahydrocannibol) capsule or a placebo capsule three times daily in a small, short duration (21 days) intervention clinical trial (Abrams et al., 2003) In a more recent study by Ghosn et al.(2014), sex while under the influence of cannabis was significantly associated with increased odds of a detectable HIV 1 RNA in the semen (Ghosn et al., 2014) among a sample of 15 7 HIV positive MSM Two studies have found lower viral load among marijuana users (Milloy et al., 2015; Thames, Mahmood, Burggren, Karimian, & Kuhn, 2015) Thames et al (2015), in a sample of 55

PAGE 24

24 PLWH recruited from HIV clinics in the Los Angeles area, found lower viral load among light and moderate to heavy marijuana users as compared to non users (Thames et al., 2015 ) A fourth study found daily cannabis use to be significantly associated with lower plasma viral load as compared to less than daily use among in a sample of 88 newly infected PLWH in Vancouver, Canada participating in an injection drug use study (Milloy et al., 2015) A more recent study that utilized data from 1,902 HIV positive individuals in the Florida medical monitoring project a CDC sponsored supplemental surveillance project of HIV posi tive individuals in medical care reported no statistically significant association between less than daily or daily marijuana use and decreased odds of viral suppression Specifically, there was no significant difference between adjusted geometric mean v iral loads between both less than daily use (104.7; 95% CI = 30.2, 362.1 copies/mL; p = 0.0684) and daily use (125; 95% CI = 35.4, 446.8 copies/mL; p = 0.1610) as compared to nonuse (81.2; 95% CI: 37.2, 242.8 copies/mL) (Okafor et al., 2016) Marijuana use and quality of life: Studies on the association between marijuana use and quality of life of HIV positive individuals have also been few. Using significant decreased odds of self reported good quality of life among those reporting any marijuana use in the past six months. However, when limiting their analysis to only recent marijuana use, daily marijuana use was significantly associated with higher odds of self reported good quality of life as c ompared to nondaily marijuana users et al., 2012) A more recent cross sectional study of HIV positive individuals in medical

PAGE 25

2 5 care at the University of Colorado clinic assessed the impact of marijuana use on the components of successful aging of HIV positive individuals. Among thei r predominantly male sample (83%) mean age of 51 years, the authors found that quality of life specifically in the social functioning and mental health domains was significantly lower among recent marijuana users (past month) as compared to none recent u sers (Allshouse et al., 2015) Neurocognitive Effects o f Marijuana Use The question of whether marijuana use is a ssociated with adverse and neurocognitive effects of is a subject of intense debate in the scientific community. This section briefly discusses the endocannabinoid system and the mechanisms involved in t data emanating from preclinical studies and in vitro studies. Overall, there have been few studies addressing the association between marijuana use and cognitive functions among HIV positive individuals. This section will also provide a brief summary of the studies conducted to date among HIV uninfected individuals detailing the cognitive domains that marijuana use appears to influence. I will review studies of acute, non acute and residual/long term effects. Studies on Acute effects of marijuana on cogni tive functions: Acute intoxication with marijuana produces alterations in brain functioning in brain regions with CB1 receptors distributions. An early review of the studies among the HIV uninfected published in the 1980s on the acute effects of marijuana concluded that there is evidence of deficits in memory when individuals are intoxicated with marijuana (Ferraro, 1980) Specifically, individuals intoxicated with marijuana demonstrate problems recalling new information presented to them while acutely intoxicated, but do not appear

PAGE 26

26 to have problems recalling information that was presented prior to intoxication. One review ( 2006 ) made similar conclusions (Ranganathan a, 2006) Specifically, the authors concluded that the most consistent deficits associated with acute intoxication of marijuana are in tasks of immediate and delayed recall of information presented while intoxicated. The impairing effects of acute intoxic ation of marijuana appears to be stronger in less experienced users than in experienced users who may have with better tolerance (Crean, Crane, & Mason, 2011) Generally, acute intoxication of marijuana also appears to impair working memory as some studies have observed impairments in working memory following acute administration of vaporized marijuana (Bossong et al., 2012) and smoked marijuana at different doses (Hart et al., 2010) Attention and concentration, abstract reasoning and working memor y also appear to be affected during acute intoxication (Crane, Schuster, Fusar Poli, & Gonzalez, 2013; Crean et al., 2011) Acute intoxication of marijuana appears not to impair decision making or risk taking in occasional or experienced users (J G Ramaekers, Kauert, Theunissen, Toennes, & Moeller, 2009; Weinstein et al., 2008) but may slow decision making skills (Vadhan et al., 2007) Although there are studies with contradicting findings ( 111) Findings to date on the effect of acute intoxication on other cognitive domains including inhibitory control, psychomotor control an d impulsivity are equivocal (Crane et al., 2013) Studies on Non acute effects of marijuana u se on cognitive functions: Among studies examining the non acute effects (i.e. those effects observed after acute intoxication subsides) of marijuana use on cognitive functioning, find similar results as the studies on acute effects Specifically, that asp ects of memory particularly verbal

PAGE 27

27 and episodic memory are consistently impaired in both adult and adolescent marijuana users (Crane et al., 2013; Crean et al., 2011; Gonzalez, 2007) In two studies that examined the neuropsychologica l performance of a longitudinal cohort of adolescents aged 9 to 20 years of age. The studies found that compared to their baselines, only the participants with current heavy marijuana use demonstrated decreases in IQ scores, immediate and delayed memory as well as information processing speed (Fried & Smith, 2001; Fried, Watki nson, & Gray, 2005) The literature on non acute effects of marijuana use on attention and concentration suggests they appear to persist beyond periods of intoxication However, longer duration of abstinence appear to restore cognitive functions in these domains. Following at least 21 days of abstinence from marijuana adolescent marijuana users continued to perform poorly in tasks of attention and concentration than non using controls (Bolla, Brown, Eldreth, Tate, & Cadet, 2002; Hanson et al., 2010; Tapert et al., 2007) and deficits appear worse as total lifetime exposure to marijuana increases (Jacobsen, Mencl, Westerveld, & Pugh, 2004) Similarly, among adult marijuana users with 28 days of abstinence as well as recently abstinent adult marijuana users also demonstrated impairments in tasks of attention and concentration (Bolla et al., 2002; Schol es Balog & Martin Iverson, 2011; Wadsworth, Moss, Simpson, & Smith, 2006) However, following 45 days of abstinence, adolescent marijuana users demonstrated no impairments (Jacobsen et al., 2004) and among adult marijuana users of unkn own last marijuana use also showed no impairments (J. E. Grant, Chamberlain, Schreiber, & Odlaug, 2012) although others have found contradictory findings (Hermann et al., 2007)

PAGE 28

28 With regard to working memory, recently abstinent (after 3 and 13 days) adolescent marijuana users demonstrated worse working memory compared to non using controls (Hanson et al., 2010) also, adolescent marijuana users performed worse on spatial working memory tasks (Harvey, Sellman, Por ter, & Frampton, 2007) even after 28 days of abstinence (Jacobsen et al., 2004) However, several studies report no deficits in working memory following 28 days of abstinence (Hanson et al., 2010; Padula, Schweinsburg, & Tapert, 2007; Schweinsburg et al., 2010) In contrast. among adult marijuana users, many studies report no impairments in working memory among recent abstinent marijuana u sers (Gruber, Sagar, Dahlgren, Racine, & Lukas, 2012; Scholes & Martin Iverson, 2010) In a recent longitudin al study, no impairments in working memory performance was found among recently abstinent marijuana users, former users an controls over a period of 8 years (Tait, Mackinnon, & Christensen, 2011) Therefore, in sum the evidence for the no n acute effects of marijuana use on working memory appears mixed when considering adolescent samples, but appears to demonstrate a consistent pattern of no impairments in adult samples. A dult marijuana users show more consistent deficits in decision makin g and risk taking following 12 hours of abstinence (Whitlow et al., 2004) 15 days (Fernndez Serrano, Prez Garca, Schmidt Ro Valle, & Verdejo Garca, 2010) and even 25 d ays of abstinence (Verdejo Garca, Riva s Prez, Lpez Torrecillas, & Prez Garca, 2006) With regard to inhibition and psychomotor/motor impulsivity, numerous studies report no impairments in psychomotor control (Harvey et al., 2 007) inhibition and motor impulsivity (Tapert et al., 2007) in adolescent marijuana users. Some studies in adult samples report no difference between marijuana users and non using controls (Fontes

PAGE 29

29 et al., 2011; Gonzalez et al., 2012; Gruber, Dahlgren, Sag ar, Gnenc, & Killgore, 2012) Yet, others report impairments in psych omotor performance in adolescent marijuana users following 23 days of abstinence, with evidence of a dose dependent association (MEDINA et al., 2007) Similarly, several studies also find impairments in adult samples (Batti sti et al., 2010; Roberts & Garavan, 2010) Several factors could explain the discrepant findings found in these studies including duration of marijuana use length of abstinence, cumulative exposure to marijuana, whether appropriate non using control gr oups are used and use of cross sectional versus longitudinal designs. In addition, the difficulty levels of the neuropsychological tasks administered may also account for the discrepant findings. For instance, studies using th e Wisconsin Cards Sorting Test find significant impairments (Bolla et al., 2002; H G Pope Jr, Gruber, Hudson, Huestis, & Yurgelun Todd, 2001; Harrison G Pope Jr, Gruber, Hudson, Huestis, & Yurgelun Todd, 2002) whereas studi es using the Stroop Test, a simpler task, find no significant differences between marijuana users and controls Neuropathogenesis o f HIV Associated Cognitive Impairments HIV enters the central nervous system (CNS) early after initial infect ion via a blood brain barrier (BBB) (An, Groves, Gray, & Scaravilli, 1999; Liu et al., 2002) Onc e inside, infection of neighboring cells occurs through direct contact with the infected cell. The other cells in the brain in which the infection spreads to include perivascular macrophages, astrocytes and microglia (Albright, Soldan, & Gonzlez Scarano, 2003) However, others have proposed that mechanisms other than direct infection of brain

PAGE 30

30 cells may mediate the neuropathogenesis of HIV infection, s ince the observation that HIV does not directly infect neurons. One of two of the proposed models for the neuropathogenesis of HIV infection is referred to as the direct model The direct model suggests viral proteins released from infected monocyte derive d cells incite neuronal cell death via their direct interactions with neurons (Gonzlez Scarano & Martn Garca, 2005) The second, indirect model, proposes that neuronal damage is mediate d by a generalized inflammatory processes that is triggered by infected and uninfected non neuronal cells against HIV infection and HIV proteins that are released by infected cells (Gonzlez Scarano & Ma rtn Garca, 2005) In addition, to neuronal damag e, research has suggested that chronic neuroinflammation may also be associated with synaptic disruption and impairment of neurogenesis i.e. the generation and growth of neurons, which lead to the neurol ogical symptoms associated with HIV infection (Gannon, Khan, & Kolson, 2011; Lindl, Marks, Kolson, & Jordan Sciutto, 2010) HIV Associated Neurocognitive Disorder s Early in the HIV epidemic clinicians caring for HIV positive individuals observed neurological symptoms in some of their patients with advanced disease (Snider et al., 1983) Grant et al.(1987) published the first study describing HIV Associated Neurocognitive Disord ers (HAND) in asymptomatic HIV positive individuals (I. Grant et al., 1987) This study demonstrated that HAND is present across all stages of HI V infection. These disorders ranged from subtle neurocognitive impairments t o the profoundly disabling HIV A ssociated dementia (HAD). HAD was more frequently observed among HIV positive individuals in advanced staged of the disease and was associated with increased mortality. Fortunately, the introduction of ART was associated with a drastic decline in the incidence and prevalence of HAD (Maschke et al., 2000) as

PAGE 31

31 more HIV positive individuals were able to suppress their viral loads. However, in the era of cART milder forms of HAND continue to be highly prevalent and can pose difficulties for HIV positive individuals to perform tasks of everyday living including adherence to medications and clinician advice (Robert K Heaton et al., 2004; Thames, Arentoft, Rivera Mindt, & Hinkin, 2013) The classification of HAND was described in 2007 (Antinori et al., 2007) to comprise 3 stages including: (1) Asymptomatic Neuropsychological Impairment (ANI), (2) Minor Neurocognitive Disorder (MND) and (3) HAD. The mildest form of HAND ANI was defined as neuropsychological performance that is one standard deviation below the mean in t wo cognitive domains with no impairment in activities of daily living and no evidence of a preexisting cause other than HIV. MND was defined as in addition to the aforementioned criteria for ANI, impairment in at least one activity of daily living. HAD w as defined as defined as neuropsychological performance that is 2 standard deviations below the mean in 2 cognitive domains with marked impairments in activities of daily living and with no evidence of a preexisting cause other than HIV. Prior to the wides pread use of ART, the overall prevalence of neurocognitive impairment was about 55%; with prevalence of ANI, MND and HAD was 20%, 13% and 20% respectively (R K Heaton et al., 1995) Prevalence of HAND in the era of cART has varied somewhat across studies. In one of t he largest study to assess, the prevalence of HAND in the era of cART was in the CNS HIV Antiretroviral Therapy Effects Research (CHARTER) study. Of the 1,555 HIV positive individuals, 814 (or 52%) had cognitive impairment. Further, the prevalence estimate s for ANI, MND and HAD was 33%, 12% and 2% respectively (R K Heaton et al., 2010) One other study including 200

PAGE 32

32 HIV positive individuals found an overall prevalence of HAND to be about 64%; with 60% ANI; 45 % MND and 0% HAD (Simioni et al., 2010) In a more recent report, using data from 364 HIV positive individuals from the Multicenter AIDS Cohort Study (MACS) found an overall frequency of HAND of 33%. The distribution of HAND categories was as follows: ANI (14%, MND (14%) and HAD (5%). Some researchers have suggested that the afor ementioned approach in the classification of HAND are less stringent and may be overestimating the true prevalence of HAND (Gissln, Price, & Nilsson, 2011; Nightingale et al., 2014) As described above, HAND is defined by one or more standard deviations below normative neuropsychological test performance derived from a control population. Therefore, HAND classification criteria depend substant ially on how the control population reflects the test population. Thus, differences in sociodemographic characteristics between the test population and controls may produce different estimates of HAND. Gissen et al (2011) argue d that based on statistical criteria alone, the approach may be prone to overestimation of HAND (Gissln et al., 2011) This has important clinical implications because it may put HIV positive individuals diagnosed as cognitively impaired under undue anxiety and stress and the use of expensive interventions (Gi ssln et al., 2011) In the classification of HAND, it is important to consider other factors that may predispose to HAND. Hepatitis C infection (HCV) is prevalent among HIV positive individuals and is associated with cognitive impairment independent of H IV. HIV/HCV co infection individuals are nearly two folds more likely to have cognitive impairment as compared to those with monoinfection (Cherner et al., 2005; Clifford, Evans, Yang, & Gulick, 2005) Cerebrovascular disease associated with HIV effects and ART on

PAGE 33

33 endothelial function and cardiovascular disease is another im portant factor that can contribute to cognitive impairment particularly as HIV positive individuals now reach advanced age (Nightingale et al., 2014) Psychoactive drug use including methamphetamine can also predispose to cognitive impairment (Gupta et al., 2011; Rippeth et al., 2004) Immunosuppression before the initiation of ART typically indicated as CD4 nadir (lowest ever CD4 count) has been strongly associated with cognitive impairment (Ellis et al., 2 011; Muoz Moreno et al., 2008) The presence of other psychiatric comorbidities, low educational attainment as well as some genetic factors (e.g. APOE e4) can also contribute to cognitive impairment (Clifford & Ances, 2013) Therefore, it is important that these factors be considered when characterizing cognitive impairment associated with HIV. In addition, cART use has lengthened survi val among HIV positive individuals and thus optimal adherence to and long term use of cART may be associated with increasing survival including cognitive function among HIV positive individuals. The nature of neurocognitive impairment across cognitive doma ins has also evolved in the cART era as compared to the pre CART era. Deficits in speed of information processing, verbal fluency, and motor speed/dexterity characterized neurocognitive impairment in the pre CART era. Whereas, impairments in verbal learnin g/memory, complex attention and executive functions are more common in the cART era (Cysique, Maruff, & Brew, 2004; Robert K Heaton et al., 2011) impairments that overlap with mariju ana related neurocognitive deficits (Crane et al., 2013; Crean et al., 2 Therefore, potential for interactive effects between HIV and marijuana use on these

PAGE 34

34 domains is plausible. In the following section, I will review studies that have investigated the t opic of the association between marijuana use and cognitive function in HIV positive individuals. Marijuana Use, HIV Infection a nd Cognitive Function Christiani et al (2004) (Cristiani et al., 2004) compared HIV positive [Asymptomatic (n=127) and Symptomatic (n=87) and HIV negative individuals (n=74 ) stratified by whether they were frequent (at least once per week in past 12 months) or had minimal or no exposure (less than once per month in past 12 months) to marijuana on a variety of neuropsychological tests. The authors found significant main effec ts for HIV disease stage ( HIV negative vs. Asymptomatic HIV vs. Symptomatic HIV), such that subjects with symptomatic HIV performed worse on all measures of neuropsychological performance. There was a significant main effect for marijuana use; such that ma rijuana users performed worse on a measure of information processing speed and delayed memory. The authors found significant marijuana by HIV interaction effect on a global measure of cognitive impairment; such that symptomatic HIV positive individuals wit h frequent marijuana use had significant higher number of neuropsychological performance that was one standard deviation (SD) below or more the mean of controls. Further analysis revealed that poorer performance on tasks of delayed memory accounted for th e significant interaction effect on this global cognition. T he authors found no significant differences on other measures of cognitive functioning including attention/working memory, executive functions, and motor skills. In contrast, Chang et al (Chang, Cloak, Yakupov, & Ernst, 2006) utilized an extensive battery of neuropsychological test to compare neuropsychological functioning of a sample of HIV positive individuals with regular marijuana use, HIV positive individual s with no or

PAGE 35

35 minimal marijuana use and HIV negative controls The authors found no evidence of an interaction between marijuana and HIV on any test of neuropsychological functioning including tests of executive functions, memory, attention, psychomotor spe ed, and fine and gross motor skills. However, in a more recent study, Gonzalez et al (Gonzalez et al ., 2011) reported additive negative effects of a history marijuana dependence and HIV on complex motor skills among a sample of polysubstance users. In a more recent study, Thames et al. (2015) examined the combined effects of HIV status and marijuana us e on neurocognitive among 84 ( HIV positive =55 and HIV negative =34) participants recruited from HIV clinics in the greater Los Angeles area. The authors grouped p articipants according to their marijuana use status as: light users (2 14 times /week), moder ate to heavy users (18 90 times/week) and nonusers (reported never using marijuana). The HIV positive individuals performed worse in measures of learning/memory [F (1 82) = 15.65 p <.001 16 ] and executive functioning [F ( 1 82) = 3.23 p = .0 3 7 ] as compared to the HIV negative individuals Moderate to heavy marijuana users performed worse on measures of ry [F (2, 82) = than light users and non users. Further, the authors found a significant marijuana by HIV interactive effect in learning and memory such that the HIV positive individuals with moderate to heavy marijuana use performed significantly worse on learning and memory than all other comparisons groups (Thames et al., 2015) The mixed findings observed in these studies are likely a result of their cross sectional designs, participant self selection to participate in the studies, small sample

PAGE 36

36 sizes and lack of adequate control for other type s of substance and other influential factors with potential central nervous system effects such as Hepatitis C infection. In addition, there is evidence that some pre existing neurocognitive deficits and premorbid IQ (Ch eetham et al., 2012; Jackson et al., 2016; Vanyukov et al., 2003) may predate the onset of marijuana and other substance use which would make it problematic to discern whether any neurocognitive deficits observed among marijuana users are direct effects o f marijuana or other antecedent risk factors, especially when cross s ectional study designs are used Significance o f Marijuana Use o n the Cognitive Function s This question is of crucial public health and clinical importance for a number of reasons. Firs t, and as noted earlier, HIV positive individuals increasingly use marijuana both medically and recreationally with the potential for use to be chronic and heavy More so, with the evolving state laws that allow marijuana use for medical purposes for persons with qualified conditions of which HIV positive individuals are eligible, more HIV positive individuals may be considering marijuana us e, despite lack of empirical evidence on its therapeutic effects for its indication of use. Second, it re mains unclear whether marijuana use may affect the neurocognitive functioning of HIV positive individuals As HIV positive individuals are living longe r, it has been proposed that marijuana use to management HIV related symptoms and side effects of ART (e.g. nausea, pain, insomnia mood problems) might be heavy and chronic. As such, whether heavy, chronic marijuana use adverse impacts the neurocognitive f unctioning of persons is not thoroughly understood. Neurocognitive impairment in HIV positive individuals has been associated with a 2.5 times greater risk of poor adherence to ART (Hinkin et al., 2004) increased

PAGE 37

37 dependence in activities of e veryday living (Robert K Heaton et al., 2004) and worse overall quality of life (Parsons, Braaten, Hall, & Robertson, 2006) Therefore, understanding whether long term marijuana use and HIV infection interact negatively on neurocognitive outcomes has important implications for improving treatment outcomes in HIV positive individuals Findings from this study have potential significant implications for neuroAIDS research as it can help determine whether marijuana use impacts on the neurocognitive functioning of HIV positive individuals and if so the magnitude of these deficits which can pave the way for additional research to understand the underlying mechanisms

PAGE 38

38 CHAPTER 2 STATEMENT OF SPECIFIC AIMS AND HYPOTHESIS This dissertation project aims to address thre e specific gaps in the literature regarding marijuana use among HIV positive individuals. Thus, this project consists of three studies with the aims and hypothesis presented below. Aim 1 Trends in the Prevalence of and Risk Factors for Marijuana Use Very l ittle is known about the how trends in the prevalence of marijuana use has evolved over time in HIV positive individuals, especially during a period that changes in legislations toward marijuana use by US states has occurred and the overall population favo rability for medical marijuana use has increased As detailed previously, only one study to date has assessed trends in prevalence of marijuana use over an extended period in HIV positive individuals, but this study was among women 2012) Further, although much work has described predictors of marijuana use in other populations, there are few studies on predictors of marijuana use among HIV positive individuals. Some research has demonstrated that passage of MMLs ma y be associated with increased marijuana use among adolescents (Cerd et al., 2012; Wen et al., 2015) but there are no published studies in HIV posi tive individuals. In this study I will utilize data from a large longitudinal cohort of HIV positive and HIV negative MSM to address the following aims: (1) assess trends in the annual prevalence of current and daily marijuana use over time (1984 2013) (2 ) determine factors associated with current and daily marijuana use over time in th e combined sample and among HIV positive individuals (3) and explore whether passage of MMLs is associate d with increased marijuana use.

PAGE 39

39 Hypothesis: I hypothesize that preva lence of marijuana use will decline over time, but that daily marijuana use (among users) will increase and that HIV related clinical factors such as CD4 counts, HIV viral loads will be positively associated with marijuana use and that passage of MMLs will be associated with increased marijuana use. Aim 2 T rajectories of Marijuana Use and Predictors of Trajectories Trajectory analysis is based on the premise that a given population may not be homogenous in the progression of a behavior (such as marijuana us e) over time, but that individuals within a population are heterogeneous with relation to the behavior. Trajectory analysis aims to identify sub groups of individuals that follow different patterns over time. These different patterns may represent differen t etiologic pathways to and through the behavior. Thus, another goal of trajectory analysis is to identify cofactors of these different trajectories of marijuana use among men who have sex with men The objectives of this aim is to : (1) characterize the lo ngitudinal trajectories of marijuana use in a sample of HIV positive and HIV negative MSM over a period of 29 years, and (2) to identify co factors associated with unique t rajectories of marijuana use as well as those that can change over time that may modi fy the course of the trajectory. Hypothesis: I hypothesize the emergence of between 3 to 5 distinct trajectories of marijuana use over time and that sociodemographic characteristic [younger age, racial minorities, lower educational status, geographic locat ion (e.g. Los Angeles)], HIV positive status, psychosocial (e.g. depressive symptoms) and other substance use will be positively associated with trajectories of marijuana use. I also expect that HIV related clinical characteristics (such as ART use, CD4+ c ounts and detectable viral load) will be positively associated with trajectories of marijuana use.

PAGE 40

40 Aim 3 Cumulative Exposure to Marijuana and Cognitive Change over 17 years In the cART era, the focus of clinical care for HIV positive individuals has shift ed from an acute illness with higher death rates to the management of a chronic illness. HIV positive individuals are on cART for a long period often experience symptoms (e.g. chronic pain, nausea, insomnia and mood problems ) associated with HIV and side e ffects of cART as they age At least a third of HIV positive individuals use marijuana to manage these symptoms and thus marijuana use among HIV positive individuals may be heavy and chronic. One important concern for clinicians caring for HIV positive ind ividuals is the potential that long term and heavy marijuana use may adversely influence the cognitive functions of their patients who are already vulnerable to cognitive effects from HIV itself (R K Heaton et al., 2010; Robert K Heaton et al., 2011; Sacktor et al., 2016) There are presently no published reports on the long term effects of marijuana us e on the cognition function of HIV positive individuals. Therefore, the objective of this aim is to determine the association between cumulative exposure to marijuana and longitudinal changes in cognitive performance in processing speed, executive function s and memory among HIV positive and HIV negative individuals. Hypothesis: We hypothesized that greater cumulative exposure to marijuana use will be associated with worse cognitive performance over time.

PAGE 41

41 CHAPTER 3 TRENDS IN THE PREVALENCE OF AND RISK FAC TORS FOR MARIJUANA USE State laws and attitudes toward marijuana use have continued to evolve: twenty three states and the District of Columbia now allow marijuana for medical and recreational purposes (Cerd et al., 2012; Hasin et al., 2015) Several reports have documented an increase in marijuana use (Hasin DS et al., 2015; Substance Abuse and Mental Health Services Administration, Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings, 2014) as well as daily or near daily use Substance Abuse and Mental Health Services Administration, Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings, 2014) in the general US population since th e mid 2000s. Research among HIV positive individuals in the US suggests that marijuana use is common and higher than the general uninfected population. Rates of current (or past si x months) marijuana use in HIV positive individuals hav e ranged from 23% to 56% 2004; MD, 2001; Mimiaga et al., 2013; Pence et al., 2008; Prentiss et al., 2004; Ware et al., 2003) as compared to 8.5% in the general US population 18 + years of age ( Substance Abuse and Mental Health Services Administration, Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings, 2014) With widespread use of an tiretroviral therapy (ART), HIV positive individuals are living longer and the focus of clinical care has shifted to the management of a chronic diseas e. Observational studies of HIV positive individuals cite therapeutic benefits of marijuana; including re lief of HIV related symptoms as well as side effects of ART (e.g. chronic pain, nausea, insomnia and mood problems) (

PAGE 42

42 Prentiss et al., 2004; Ware et al., 2003; Woolridge et al., 2005) although empirical data on its efficacy and safety of use is limited (Lutge et al., 2013) Imp ort antly, marijuana use in HIV positive individuals has been associated with reduced ART adherence (Bonn Miller et al., 2012; Corless et al., 2009; Peretti Watel et al., 2006) cognitive impairment (Cristiani et al., 2004; Gonzalez et al., 2011) and poorer quali ty of life (Allshouse et al., 2015) Data on long term trends and patterns of marijuana use among HIV positive in dividuals have also been scare. In a recent study that assessed longitudinal patterns of marijuana use in women living with HIV, prevalence of current marijuana use decreased significantly from 21% to 14% over a 16 year period (1994 2010); however, daily u se (among users) increased by more than three fold, increasing from 14.8% in 1994 to 51% in 2010 Past studies of predic tors of marijuana use among HIV positive individuals have found younger age (Bonn Miller et al., 2012; Braitstein et al., 2001) lower educational level (Bonn Miller et al., 2012) alcohol, cigarette and other illicit substances to be positively associated with marijuana use (Allshouse et al., 2015; Bonn Miller et al., 2012; Prentiss et al., 2004) although most of these studies have been cross sectional. Using data fro m a longitudinal cohort of HIV positive women, Kuo et al. (2004) found lower initiation of weekly marijuana use among women with an undetectable viral load and those receiving cART (Kuo et al., 2004) A follow up study in HIV positive women found ma rijuana users to be less likely to be on c ART, but daily marijuana use to be associated with higher CD4 counts In addition, passage of medical marijuana laws (MMLs) may be associated with increased availability and

PAGE 43

43 easier access to marijuana and may contribute to increased use of marijuana. Several studies have showed that passage of MMLs i s associated with increased marijuana use (Cerd et al., 2012; Wen et al., 2015) Other studies either indicate no effect (Hasin et al., 2015; Lynne Landsman et al., 2013) or a decrease in marijuana use following passage of MMLs (Harper et al., 2012) However, nearly all of these studies have focused on adolescent samples. Given that most state MMLs list HIV/AIDS as a qualifying condition for medical use of marijuana (Wilkinson et al., 2016) passage of MMLs may be associated with increased marijuana use among HIV positive individuals. The aim of the present study was to: (1) assess trends i n the annual prevalence of current and daily marijuana use over time (1984 2013) among HIV positive and HIV negative individuals (2) determine factors associated with current and daily marijuana use over time in the combined sample and among HIV positive i ndividuals (3) and explore whether passage of MMLs is associate d with increased marijuana use. Methods The Multicenter AIDS Cohort Study (MACS) The MACS is an ongoing prospective cohort study of the natural and treated history of HIV infection among MSM i n the United States. A total of 6,972 men were enrolled during the project in three waves: 4,954 men in 1984 1985, 668 in 1987 1991, and 1350 in 2001 2003 and at 4 centers located in Baltimore/Washington DC, Chicago, Los Angeles, and Pittsburgh. The study design of the MACS has been described previously (Roger Detels et al., 1992; Dudley et al., 1995; Kaslow et al., 1987) and only the design relevant to the present analyses are described here. The study questionnaires used in the MACS are available at www.statepi.jhsph.edu/macs/forms.html The institutional review boards at the

PAGE 44

44 respective recruitment centers and their community affiliates approved the MACS study protocols and study participant provide inf ormed consent. MACS participants return every 6 months for a physical examination, collection of blood specimens and complete a detailed interview and questionnaires. The interview and questionnaires collect demographic, psychosocial, behavioral, and medic al history data. The questions about recreational drug use, including marijuana, alcohol, poppers, cocaine, crack, heroin, methamphetamine, ecstasy, injection drug use as well as smoking history since last visit were collected using audio computer assiste d self interviewing an approach previously demonstrated to provide more accurate administered questionnaires among MSM (Gribble et al., 2000) Participants The present study uses data from 5,914 [2,742 ( HIV positive ) and 3,172 ( HIV negative ] men who answered questions on marijuana use for at least two or more semi annual visits. For the present analysis, two enrollment periods were defined : the men enrolled prior to 2001 are the early cohorts and those enrolled post 2001 are the late cohorts. The enrollment cohorts were analyzed separately because of difference s in the samples that were recruited: the men in the early cohort were predominantly non Hispanic white, had more years of education, and had fewer symptoms of depression than those in the late cohort (Becker et al., 2014) We included data collected from marijuana use questions from semiannual study visit 1 (data collection starting in April 1, 1984) through visit 59 (data collection ending in September 30, 2013) for the men in the early cohort The period covered for the men in the late cohort included: semiannual visit 40 (data collection starting in October 1, 2003) throu gh visit 59. Visit 40 was

PAGE 45

45 selected as the baseline for the late cohort as this was when the sample size reached its maximum after the expansion of the cohort between 2001 and 2003. For the a nalysis limited to only the HIV positive men, visit 25 (data colle ction starting in April 1, 1996) and visit 40 was selected as the baseline for the men in the early and late coh orts respectively. V isit 25 was selected as the baseline for the HIV positive men in the early cohort because I was interested in the effect o f cART use on rate of marijuana use, which only became available in 1996. Measures Outcome Measure. The following question assessed current marijuana use at Have you used any pot, marijuana or hash since your last visit ? Participants who No users. Among queried with the following question How often did you use pot, marijuana or hash since your last visit e following response optio Covariates Socio demographic Characteristics. calculated from their self reported date of birth. The baseline visit (or index visit) was used to define a three level c ategorical variable for race/ethnicity status (Non Hispanic White, Non Hispanic Black and other), educational attainment (High school diploma or less, some college or college degree, Graduate work or more) and current employment (employed, unemployed). Par ticipants were classified according to the MACS study center and whether they were enrolled prior to or after 2001. Depressive symptoms. The Center for Epidemiologic Studies Depression (CES D) scale, was used to measure clinically significant symptoms of d epression

PAGE 46

46 (Radloff, 1977) The CES D includes components of depressed mood, feelings of worthlessness, sense of hopelessness, sleep disturbance, loss of appetite, and concentration difficulties. Scores on the CES D of 16 or more suggests a clinically significant level of psychologi cal distress (Radloff, 1977 ) Alcohol Use. Using data regarding frequency of drinking and average number of alcoholic drinks since last study visit, alcohol consumption at baseline and at each visit was categorized as low moderate (1 to 2 drinks/day, or 3 to 4 drinks/day no more th an once a month), heavy (3 to 4 drinks/day more than once a month, or 5 or more drinks/day less than once a month or 5 or more drinks/day at least once a month) or no alcohol use. Cigarette Use. Participants were considered current smokers if they respond ed yes to a question asking about any cigarettes smoking since last study visit. Among current smokers, pack years of smoking at initial visit and at each subsequent visit was f cigarettes smoked per day. Stimulant use. Participants were considered to be users of stimulant drugs if they reported the use of any of the following drugs since last study visit: (1) crack cocaine, (2) other forms of cocaine, (3) methamphetamines (or speed, meth or ice), (4) methylenedioxy N methylamphetamine). Clinical factors. The MACS assessed HIV s erostatus using enzyme linked immunosorbent assay with confirmatory Western blot tests on all MAC S participants at

PAGE 47

47 initially HIV negative However, only participants who were HIV positive as at the time of enrollment were included. Standardized flow cytometry was used to quantify CD4+ T lymphocyte subset levels by each MACS site (Giorgi et al., 1990) and categorized as ribonucleic acid ( RNA ) were measured using either the standard reverse transcription polymerase chain reaction assay (Roche Nutley, NJ) or wi th the Roche ultrasensitive assay (Roche Diagnostics ) Standardized viral load measures (across different assays) were used to create a dichotomous variable to denote detectable (> 40 copies / mL) versus undetectable. Hepatitis C infection status was categor ized as hepatitis C virus (HCV) n egative if HCV antibody testing was negative. Participants were classified at each MACS semiannual visit as HCV positive if they were found to be in the process of seroconversion, acute infection, chronic infection, clearin g (between RNA+ and RNA ), or previously HCV positive, but now clear of HCV RNA. In addition to the aforementioned cov ariates described above, I considered that the prevalence of marijuana use among HIV positive participants may be influenced by factors sp ecific to HIV infection such as antiretroviral therapy usage and adherence as has been previously reported Antiretroviral medications used since the last visit were self reported at each semiannual visit and summarized to define Highly active antiretroviral therapy ( HAART ) use (yes/no). HAART was defined according to the U.S. Department of Health and Human Services/Kaiser Panel guidelines es for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV 1 Infected

PAGE 48

48 Data Analysis Using frequencies and percentages, c haracteristics of the sample at the ir b aseline visit stratified by HIV s erostatus and cohort enrollment was described. Y early prevalence of current marijuana use was calculated as the number of participants reporting marijuana use divided by the number of participants seen in the MACS for a given year. Similarly, yearly prevalence of Daily marijuana use was computed as the number of participants reporting daily use divided by the number of current users for each given year. I also plotted both prevalence of current and daily marijuana use o ver the follow up period by c alendar year stratified by HIV s erostatus and cohort enrollment. Poisson regression with robust error variance (Zou, 2004) was used to estimate prevalence rate ratios of current and daily marijuana use in both univariate and multivariable analysis. Separate analyses were conducted for the men in the early and late cohorts and w ithin each enrollment c ohor t; analyses were also conducted separately f or the combined group (i.e. HIV positive a nd HIV negative men) as well as the HIV positive men. The covariates in the multivariable model for the combined group included age, race, educational attainment, emp loyment, study center, depressive symptoms, alcohol, smoking, stimulant drug use, IDU, and hepatitis C status. Furthermore, in order to compare the prevalence rates of the HIV positive to the HIV negative men, the model for the combined group included a va riable to denote HIV s erostatus. The mod els for the HIV positive group examined HIV related clinical factors including: CD4+ cell counts, detectable HIV RNA status and HAART use. Our strategy for constructing the multivariable models was to i nclude covariates that were significant ( .10) in the univariate analyses. In addition, for the

PAGE 49

49 analysis in the men in the early cohort, models were estimated for the period 2002 2013 in order to increase comparisons of the results for the men in the late cohort. The MACS enrolled participants in four states and the District of Columbia (DC). Each of the states for the MACS study centers passed medical marijuana laws (MMLs) at different times. States and years they passed MMLs are as follows: Californi a (1996), Illinois (2013, Maryland (2003), DC (2010) with Pennsylvania not passing a MML in 2013 the end of follow up for this analysis. Using this information, a time varying binary variable for each year (1984 2013) was constructed to indicate the years before a state after the prevalence of marijuana use within states before and after the passage of the law. Only Pennsylvania did not pass a medical marijuana law as of 2013 (the end of the study period) (Pacula, Hunt, & Boustead, 2014) Missing data for predictor variables were imputed using multiple imputation with chained e quations (MICE) (van Buuren, Boshuizen, & Knook, 1999) Five imputed datasets were generated for missing b aseline and time varying cofactor variables which range from 0.2% ( HAART use) to 14.8% (viral load) and the estimates were combined (Toutenburg, 1990 ) Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, North Carolina, USA) and STATA version 11. Results Sample Characteristics at Baseline Table 3 1 displays the baseline characteristics of the 5,914 participants in thi s study stratified by HIV s erostatus and cohort enrollment. The mean age at baseline

PAGE 50

50 ranged from 33 years [standard deviation (SD) =6.7] in the HIV positive men in the early cohort t o 39 years (SD=8.2) in the HIV positive men in the late cohort. The men in the early cohort were predominantly non Hispanic, wh ite (88% in HIV negative and 79% in HIV positive men), whereas the majority of the men in the late cohort were non Hispanic, bla ck (48% in the HIV negative and HIV positive men respectively). At baseline the men in the early cohort wer e more educated (88% in the HIV negative and 84% in the HIV positive men having a college degree or more) than the men in the late cohort (67% in the HIV negative and 58% in the HIV positive men having a college degree or m ore). At baseline, the prevalence of marijuan a use was highest among the HIV positive men in the early cohort (76%) and lowest among the HIV positive men in the late cohort (36%). However, daily marijuana use, among users, was highest among the HIV positiv e men in the late cohort (20%) and lowest among the HIV negative men in the early cohort (9%). Trends in the Prevalence of Marijuana Among the men in the early cohort, the annual prevalence of current marijuana use declined significantly from 80% in 1984 to 33% in 2013 in the HIV positive men and from 58% in 1984 to 22% in 2013 in the HIV negative men (both p for trend <.0001; Figure 3 1 ). The prevalence of daily marijuana use in both the HIV positive and HIV negative men was relatively stable across the p eriod 1984 2013 among all HIV positive (11% to 10%) and HIV negative (5% to 4%) men in the early cohort. In contrast, daily marijuana use among current users increased significantly from 14% in 1984 to 32% in 2013 in the HIV positive men and from 9% in 198 4 to 22% in the HIV negative men (both p for trend <.0001). Among the men in the late cohort, prevalence of current marijuana use declined modestly from 32% in 2002 to 29% in 2013 in the HIV positive

PAGE 51

51 men, and decreased significantly from 37% in 2003 to 26 % in 2013 in the HIV negative men ( p for trend <.0001 ; Figure 3 2 ). The prevalence of daily marijuana use was relatively stable across the period 2002 2013 among all HIV negative (5% to 4%) men in the early cohort, but increased significantly in the HIV positive men (6% to 10%). However, daily marijuana use among current users increased significantly from 17% in 2002 to 37% in 2013 in the HIV positive men and from 16% in 2002 to 34% in 2013 in the HIV negative men (both p for trend <.0001). Factors Associ ated Marijuana Use In the multivariable analysis of data from the combined sample, among the men in the early cohort, HIV positive s erostatus was significantly associated with a 53% higher prevalence of current marijuana use [adjusted prevalence ratio (a PR) =1.53, 95% confidence interval (CI):1.42, 1.64; p <0.001; Table 3 2) and daily marijuana use (aPR=1.70, 95% CI: 1.44, 2.01; p <0.001, supplemental Table 3 4 ). However, in the analysis of the men in the late cohort, there was no statistically significant association between a HIV positive status and prevalence of current (aPR=0.90, 95% CI: 0.79, 1.03; p >0.05) and daily use (aPR=1.31, 95% CI: 0.98, 1.76, p =0.09). Alcohol, smoking and stimulant use were strongly and positively associated with marijuana use i n all models. Among the men in the early cohort, the annual prevalence of marijuana use was 10% higher after passage of a MML (aPR=1.10, CI: 1.07, 1.13; p <0.001; Table 3 2). There was no significant change in annual prevalence of marijuana use associated w ith passing of a MML among the men in the late cohort in the bivariate analysis (aPR=1.09 CI: 0.98, 1.21 ; p > 0.0 5 ; Table 3 2). In the analysis restricted to the HIV positive men, among participants in the late cohort, a CD4+ <500 cells/mm 3 as compared to >500 cells/mm 3 was significantly

PAGE 52

52 associated with increased prevalence of marijuana use (aPR=1.07, 95% CI: 1.01, 1.14; p <0.05; Table 3 3 ). In addition, having a detectable viral load was significantly associated with higher prevalence of marijuana use (aPR =1.10, CI: 1.01, 1.20; p <0.05) among the men in the late cohort. When the analyses for the men in the early and late cohorts were limited to the same time period (i.e. 2002 2013), there were some similarities in co factors for current marijuana use includ ing: non Hispanic blacks having higher prevalence as compared to non Hispanic whites and alcohol, smoking and stimulant use all positively associated with marijuana use (Tables 3 2& 3 3). The risk factors for daily marijuana use were similar to those observ ed for current marijuana use ( Table 3 4 & 3 5 ), although most co factors were not associated with daily use among the men in the late cohort. However, there were some differences including a significant association between positive hepatitis C status and h igher prevalence of daily marijuana use among the men in the early c ohort (Table 3 5). Discussion In this analysis of the MACS cohort, the annual prevalence of current marijuana use decreased over time among all men (1984 2013). However, in contrast, dai ly marijuana use, among those who used marijuana in the previous six months, increased among the HIV positive and HIV negative men in both the early and late cohort enrollment: increasing by more than two folds in nearly all groups. In both the early an d late cohort enrollment, the HIV positive men reported significantly higher prevalence of current and daily marijuana use as compared to the HIV negative men. Higher educational attainment and older age were negatively associated with lower current and d aily marijuana use in nearly all models. Alcohol use, smoking, and stimulant drug use

PAGE 53

53 were positively associated with prevalence of current marijuana use. The prevalence of marijuana use increased after passage of a medical marijuana law in the analysis th at included all men in the early cohort but not for men in the late cohort. Among the HIV positive men, HAART use was not associated with marijuana use, in both enrollment cohorts. However, lower CD4 counts (<500 counts/mm 3 ) and having a detectable viral l oad was associated with increased marijuana use in the men in the late cohort. The contrasting decline in annual prevalence of current marijuana use but increasing prevalence of daily marijuana use among users is consistent with recent data from HIV posit ive al., 2012) where the auth ors found that between 1994 to 2010, there was a significant decrease in prevalence of current marijuana use from 21% to 14%. The most plausible explanation for the declining trend in current marijuana use may be the advancing age of participants in the MA CS cohort. However, among the current users in this study, daily use increased significantly in both the HIV positive and HIV negative men. This increase was observed among the HIV positive and HIV negative men suggesting that other factors that affect bot h groups may explain the increase. One potential explanation may be changes in state marijuana laws allowing for the use of marijuana for medical and recreational purposes. Between 1984 and 2013 the period of this study, three of the four states that hav e MACS sites passed laws legalizing marijuana for medical purposes. In addition, in recent years, attitudes about marijuana use in the US have tempered and there has been an increase in the population acceptance of marijuana use (Gallup, 2 013) The present study found that passage of medical

PAGE 54

54 marijuana law was associated with an increase in daily marijuana use among the men in the late cohort, but not in the men in the early cohort. Among the HIV positive men in the late cohort, having a detectable viral load and lower CD4+ counts (<500 cells/mm 3 ) was positively associated with increased marijuana use. One explanation for this association use may be that the men increased their marijuana use to manage HIV related symptoms. This finding was found only among the participants in the late cohort suggesting that differences in the sample may have accounted for this finding. Participants in the late cohort were predominantly racial/ethnic minority men, had lower education, endorsed more depressiv e symptoms, and had higher proportion of current cigarette smokers, IDUs and positive HCV at baseline (Table 1) all of which have been previously associated wit h poor clinical outcomes in HIV positive individuals (Arnsten et al., 2002; Kleeberger et al., 2001; Rosen et al., 2013; Simoni et al., 2012; Yehia BR, Fleishman JA, Metlay JP, Moore RD, & Gebo KA, 2012) Also, prior studies report no significant or clinically meaningful differences in plasma viral loa d or CD4+ counts among marijuana users (Abrams et al., 2003; Chao et al., 2008; Shoptaw et al., 2012) as compared to nonusers. Y et some have found significantly lower viral loads (Milloy et al., 2015; Thames et al., 2015) and higher CD4 (Thames et al., 2015) counts in marijuana users, although these studies differ methodologically as well as in the samples included. Taken together, these findings underscore the complex relationship between marijuana use and markers of HIV disease stage/progression and therefore warrant further study. There are some li mitations to our study. The study relied on self report of marijuana use and no biological method of measurement to confirm the self reported

PAGE 55

55 data. Although the validity of self reported drug use behaviors using the ACASI has been shown to be accurate (Gribble et al., 2000; Nichols et al., 2014) the possibility that so cial desirability bias may have influenced reporting cannot be ruled out The study used an observational design ; consequently only prevalence of and risk factors for use was assessed, but not whether the associations presented here are causal. Furthermore data w ere from an ongoing longitudinal study with extended follow up, thus attrition due to death or loss to follow up might have influenced the prevalence estimates. This study did not assess prevalence or trends in marijuana use disorder or assess oth er parameters of marijuana use (including route of administration, dose/quantity, or THC potency). In addition, readers should not interpret the findings related to MML and marijuana use as causal. Our study included only four states, two states had insuff icient time windows pre, and post enactment of laws to provided enough information to discern a change in trend. Despite these limitations, the study has notable strengths : in particular data from a large and diverse sample of HIV positive and HIV negative MSM with extensive follow up period were used to assess changes in prevalence of and risk factors for marijuana use. In sum, our study indicates a decline over time in the prevalence of current marijuana in this sample of HIV positive and HIV negative MSM but in contrast, daily use among users has increased in recent years. Given that nearly half of states in the US now have laws allowing medical or recreational marijuana use, the trend found in the study here may continue as more states consider similar legislations (ProCon.org, 2015a)

PAGE 56

56 Table 3 1. Characteristics of MACS Par ticipants at Baseline by Enrollment Cohort Characteristics Early Cohort (N=4,775) Late Cohort (N=1,139) HIV negative (n=2,677) HIV positive (n=2,098) HIV negative (n=495) HIV positive (n=644) n (%) n (%) n (%) n (%) Age ( mean, SD) 34 (8.3) 33 (6.7) 37 (9.7) 39 (8.2) Race Non Hispanic, Whites 2355 (88) 1663 (79) 170 (34) 168 (26) Non Hispanic, Blacks 183 ( 7 ) 262 (13) 235 (48) 309 (48) Other 137 ( 5 ) 173 ( 8 ) 90 (1 8) 167 (26) Education High school diploma or less 305 (12) 328 (16) 166 (34) 272 (42) Some college or college degree 1309 (49) 1178 (57) 221 (45) 275 (43) Graduate work or more 1038 (39) 572 (27) 108 (22) 96 (15) Unemployed 133 ( 5 ) 109 ( 6 ) 137 (30) 139 (23) Study center Pittsburgh 667 (25) 334 (16) 152 (31) 131 (20) Baltimore/Washington DC 732 (27) 456 (22) 134 (27) 160 (25) Chicago 592 (22) 511 (24) 91 (18) 162 (25) Los Angeles 686 (26) 797 (38) 118 (24) 191 (30) Depressive symptoms a 488 (19) 446 (23) 161 (33) 246 (40) Alcohol use None 201 ( 8 ) 128 ( 6 ) 65 (13) 114 (18) Low moderate 1144 (43) 780 (38) 206 (42) 238 (38) Heavy 1295 (49) 1156 (56) 216 (44) 273 (44) Smoking Never 1180 (44) 820 (39) 126 (26) 173 (28) Former 550 (21) 400 (19) 101 (21) 125 (20) Current 936 (35) 864 (41) 260 (53) 327 (52) Stimulant drug use c 1697 (63) 1692 (81) 190 (39) 260 (42) IDU d 76 ( 3 ) 231 (12) 71 (15) 105 (17) Positive Hepatitis C virus antibody 79 ( 3 ) 189 ( 9 ) 68 (14) 122 (19) HAART Use e No 389 (76) 226 (35) Yes 121 (23) 418 (65) CD4 + count (cells/cubic milliliter) > 500 1203 (59) 287 (46) < 500 850 (41) 338 (54 ) HIV viral load Undetectable 71 ( 5 ) 312 (49) Detectable 1256 (95) 325 (51) Marijuana use No 1151 (43) 495 (24) 290 (60) 399 (64) Yes 1523 (57) 1599 (76) 197 (40) 222 (36) Less Often 739 (49) 597 (37) 93 (47) 98 (44) Monthly 274 (18) 284 (18) 21 (11) 28 (13) Weekly 369 (24) 486 (30) 53 (27) 52 (23) Daily 141 ( 9 ) 232 (15) 30 (15) 44 (20) Note. a CESD= Center for E pidemiological Depression Scale; b Binge drinking (i.e. 5 or more drinks per occasion); c Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy; d IDU=Intravenous drug use; e HAART HAART use fo r the early cohort is obtained from MACS visit 25 (data collection starting April, 1996).

PAGE 57

57 Table 3 2. Prevalence Ratios of Risk Factors Associated with Current Marijuana Use among All Men Prevalence Ratios (95% CI) Early Cohort Late Cohort 198 4 2013 2003 2013 2003 2013 Characteristics Univariate Multivariable Multivariable Univariate Multivariable Age, per 10 year increase 0.83 (0.79, 0.87)*** 0.90 (086, 0.94)*** 0.78 (0.63, 0.97)* 0.71 (0.65, 0.78)*** 0.79 (0.73, 0.87)*** Race White, non Hispanic Reference Reference Reference Reference Reference Black, non Hispanic 1.20 (1.06, 1.36)*** 1.09 (0.95, 1.25) 1.32 (1.05, 1.65)* 1.07 (0.90, 1.27) 1.23 (1.01, 1.48)* Other 1.34 (1.18, 1.53)*** 0 .80 (0.59, 1.08) 1.14 (0.93, 1.42) 0.94 (0.73, 1.20) Education HSD or less a Reference Reference Reference Reference Reference Some college/college deg. 0.80 (0.72, 0.89)*** 0.86 (0.78, 0.96)*** 0.95 (0.75, 1.19) 1.00 (0.86, 1.17) 0.98 (0.86, 1.13) Graduate work or more 0.59 (0.53, 0.66)*** 0.76 (0.68, 0.85)*** 0.93 (0.72, 1.21) 0.53 (0.41, 0.70)*** 0.68 (0.52, 0.90)** Employment Employed Reference Reference Reference Reference Unemployed 1.04 (1. 02, 1.07)*** 1.02 (1.00, 1.04)* 1.04 (0.97, 1.12) 1.04 (0.99, 1.10) Study center Pittsburgh Reference Reference Reference Reference Reference Baltimore/Washington DC 1.01 (0.91, 1.14) 1.15 (1.03, 1.29)* 0.90 (0.69, 1.1 7) 0.75 (0.60, 0.94)* 0.93 (0.74 1.15) Chicago 1.32 (0.18, 1.48)*** 1.25 (1.11, 1.41)*** 1.00 (0.79, 1.28) 0.75 (0.61, 0.93)** 0.79 (0.66, 0.95)* Los Angeles 1.14 (1.03, 1.27)*** 1.22 (1.10, 1.35)*** 1.35 (1.06, 1.71)* 1.09 (0.89, 1.33) 1.03 (0 .84, 1.27) Depressive symptoms b CESD < 16 Reference Reference Reference Reference Reference 1.00 (0.99, 1.02) 1.00 (0.99, 1.02) 1.02 (0.98, 1.07) 1.02 (0.97, 1.07) Alcohol use None Reference Reference Reference Reference Reference Low moderate 1.58 (1.52, 1.64)*** 1.44 (1.39, 1 .49)*** 1.44 (1.30, 1.60)*** 1.82 (1.64, 2.03)*** 1.61 (1.47, 1.77)*** Heavy 1.83 (1.76, 1.91)*** 1.57 (1.51, 1.63)*** 1.53 (1.35, 1.74)*** 2.33 (2.05, 2.63)*** 1.93 (1.73, 2.15)*** Smoking Never Reference Reference Reference Re ference Reference Former 1.07 (1.03, 1.11)*** 1.07 (1.03, 1.11)*** 1.67 (1.35, 2.07)*** 1.37 (1.06, 1.77)* Current 1.28 (1.23, 1.34)*** 1.18 (1.14, 1.23)*** 1.96 (1.58, 2.45)*** 1.92 (1.49, 2.49)*** 1.58 (1.22, 2.03)** Stim ulant drug use c No Reference Reference Reference Reference Reference Yes 1.50 (1.46, 1.52)*** 1.40 (1.37, 1.42)*** 1.38 (1.28, 1.48)*** 1.65 (1.52, 1.78)*** 1.51 (1.40, 1.63)*** IDU d No Reference Reference Reference Referenc e Reference Yes 1.33 (1.26, 1.40)*** 1.09 (1.04, 1.14)*** 1.27 (1.02, 1.58)* 1.37 (1.15, 1.64)*** 1.12 (0.94, 1.33) Hepatitis C virus antibody Negative Reference Reference Positive 1.04 (0.98, 1.10) 1.00 (0.88, 1.14) HIV s erostatus HIV negative Reference Reference Reference Reference Reference HIV positive 1.80 (1.67, 1.93)*** 1.53 (1.42, 1.64)*** 1.35 (1.17, 1.55)*** 0.85 (0.73, 0.99)* 0.90 (0.79, 1.03) Medical Marijuana Law 1.17 (1.14, 1.30)*** 1.10 (1.07, 1.13)*** 1.05 (0.90, 1.22) 1.09 (0.98, 1.21) Note. a HSD=High school diploma, b C ESD= Center for Epidemiological Depression Scale; c Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy; d IDU=Intravenous drug use; .10; *P<0.05; **P<0.01; ***p<0.001

PAGE 58

58 Table 3 3. Prevalence Ratios of Risk Factors Associated with Current Marijuana Use among HIV positive Men Prevalence Ratios (95% CI) Early Cohort Late Cohort 1996 2013 2003 2013 2003 2013 Characteristics U nivariate Multivariable Multivariable Univariate Multivariable Age, per 10 year increase 0.78 (0.68, 0.88)*** 0.82 (0.70, 0.97)* 0.72 (0.65, 0.83)*** 0.77 (0.68, 0.88)*** Race White, non Hispanic Reference Reference Reference Reference Reference Black, non Hispanic 1.28 (1.04, 1.59)* 1.38 (1.07, 1.79)* 1.09 (0.85, 1.40) Other 1.15 (0.84, 1.58) 1.06 (0.80, 1.40) 0.78 (0.49, 1.11) 1.19 (0.88, 1.60) 1.14 (0.8 0, 1.63) Education HSD or less a Reference Reference Reference Reference Reference Some college/college deg. 0.88 (0.69, 1.11) 0.83 (0.67, 1.04) 1.11 (0.81, 1.53) 1.88 (0.96, 1.46) Graduate work o r more 0.71 (0.55, 0.93)** 1.15 (0.80, 1.65) 0.51 (0.33, 0.77)** 0.71 (0.45, 1.10) Employment Employed Reference Reference Reference Reference Unemployed 1.08 (1.01, 1.17)* 1.07 (1.00, 1.14)* 1.03 (0.93, 1.14) 1.02 (0.94, 1.12) Study center Pittsburgh Reference Reference Reference Reference Reference Baltimore/Washington DC 0.99 (0.73, 1.32) 1.06 (0.81, 1.39) 0.97 (0.68, 1.38) 0.81 (0.60, 1.11) 0.80 (0.55, 1.14) Chicago 1.20 (0.92, 1.57) 1.03 (0.74, 1.42) 0.66 (0.50, 0.88)** 0.63 (0.49, 0.80)*** Los Angeles 1.40 (1.09, 1.79)** 1.51 (1.18, 1.93)** 1.42 (1.03, 1.95)* 0.99 (0.75, 1.31) 0.71 (0.48, 1.02) Depressive symptoms b C ESD < 16 Reference Reference 1.03 (0.99, 1.08) 1.04 (0.97, 1.10) Alcohol use None Reference Reference Reference Reference Reference Low moderate 1.50 (1.34, 1.67)*** 1.41 (1.27, 1.57)*** 1.46 (1.24, 1.70)*** 1.69 (1.48, 1.9 3)*** 1.50 (1.34, 1.70)*** Heavy 1.67 (1.48, 1.89)*** 1.51 (1.65, 1.71)*** 1.57 (1.32, 1.86)*** 2.18 (1.86, 2.55)*** 1.83 (1.59, 2.12)*** Smoking Never Reference Reference Reference Reference Reference Former 1.33 (1.13, 1.58)*** 1.34 (1.17, 1.54)*** 1.50 (1.17, 1.91)** 1.43 (1.01, 2.01)* 1.35 (1.00, 1.80)* Current 1.52 (1.28, 1.81)*** 1.44 (1.24, 1.68)*** 1.69 (1.33, 2.16)*** 1.99 (1.40, 2.81)*** 1.67 (1.21, 2.26)** Stimulant drug use c No Reference Reference Reference Reference Reference Yes 1.51 (1.41, 1.62)*** 1.45 (1.36, 1.55)*** 1.43 (1.30, 1.58)*** 1.61 (1.45, 1.80)*** 1.48 (1.34, 1.64)*** IDU d No Reference Reference Reference Reference Yes 1.40 (1.10, 1.77)** 1. 19 (0.98, 1.44) 1.27 (1.02, 1.56)* 1.23 (1.02, 1.48)* Hepatitis C virus antibody Negative Reference Reference Positive 0.93 (0.82, 1.04) 1.09 (0.91, 1.29) HAART Use e No Reference Reference Refere nce Reference Yes 1.00 (0.96, 1.05) 1.08 (1.01, 1.18)* 1.03 (0.93, 1.14) CD4 + count > 500 Reference Reference Reference < 500 0.98 (0.93, 1.03) 1.11 (1.05, 1.17)*** 1.07 (1.01, 1.14) HIV viral load Undetectable Reference Reference Reference Reference Detectable 1.01 (0.97, 1.06) 1.13 (1.05, 1.21)*** 1.10 (1.01, 1.20)* Medical Marijuana Law 1.10 (1.00, 1.21)* 1.04 (0.92, 1.16) 1.06 (0.87, 1.28) 1 1.11 (0.91, 1.36) Note. a HSD=High school diploma; b C ESD= Center for Epidemiological Depression Scale; c Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy; d IDU=Intravenous drug use; e HAART =Highly active antiretr oviral therapy;

PAGE 59

59 Figure 3 1. Annual Prevalence of Current and Daily (Among Current Users) Marijuana Use Among HIV positive and HIV negative MSM in The M ACS : Early Cohort 0 10 20 30 40 50 60 70 80 90 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year Prevalence (%) HIV+ Current use HIV+ Daily use (among users) HIVCurrent use HIVDaily use (among users) HIV+ Daily use (among all HIV+) HIVDaily use (among all HIV-)

PAGE 60

60 Figure 3 2. Annual Prevalence of Curre nt and Daily (Among Current Users) Marijuana Use Among HIV positive and HIV negative MSM in The MACS : Late Cohort 0 10 20 30 40 50 60 70 80 90 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year Prevalence (%) HIV+ Current use HIV+ Daily use (among users) HIVCurrent use HIVDaily use (among users) HIV+ Daily use (among all HIV+) HIVDaily use (among all HIV-)

PAGE 61

61 Table 3 4. Prevalence Ratios of Risk Factors Associated with Daily Marijuana Use among All Men Prevalence Ratios (95% CI) Early Cohort Late Cohort 1984 2013 2003 2013 2003 2013 Characteristics Univariate Multivariable Multivariable Univariate Multivariate Age, per 10 year increase 0.80 (0.71, 0.90)*** 1.03 (0.51, 2.08) 0.79 (0.67, 0.94)** -Race White, non Hispanic Reference Reference Reference Black, non Hispanic 1.23 (0.93, 1.61) 1.56 (1.06, 2.29)* 1.23 (0.78, 1.94) Other 1.23 (0.89, 1.69) 0.83 (0.47, 1.46) Education HSD or less a Reference Reference Reference Reference Reference Some college/college deg. 0.62 (0.50, 0.76)*** 0.64 (0.52, 0.80)*** 0.82 (0.47, 1.41) 0.66 (0.49, 0.88)** 0.72 (0.53, 0.97)* Graduate work or more 0.33 (0.26, 0.42)*** 0 .37 (0.29, 0.48)*** 0.72 (0.40, 1.30) 0.34 (0.18, 0.63)*** 0.43 (0.21, 0.86)* Employment Employed Reference Reference Unemployed 1.08 (0.97, 1.20) 1.09 (0.94, 1.25) Study center Pittsburgh Referenc e Reference Reference Reference Reference Baltimore/Washington DC 0.67 (0.51, 0.86)** 0.76 (0.58, 0.99)* 0.71 (0.42, 1.20) 1.06 (0.67, 1.69) 0.78 (0.43, 1.40) Chicago 0.99 (0.78, 1.26) 1.05 (0.82, 1.35) 0.77 (0.43, 1.35) 0.97 (0.62, 1. 52) 0.98 (0.61, 1.57) Los Angeles 0.98 (0.79, 1.21) 1.05 (0.84, 1.31) 1.09 (0.72, 1.67) 1.66 (1.13, 2.42)** 1.04 (0.60, 1.80) Depressive symptoms b CESD < 16 Reference Reference 1.10 (0.96, 1.25) Alcohol use None Reference Reference Reference Reference Low moderate 0.85 (0.74, 0.98)* 0.88 (0.79, 0.99)* 0.87 (0.63, 1.19) 1.08 (0.84, 1.38) Heavy 0.84 (0.72, 0.98)* 0.83 (0.73, 0.95)** 0.84 (0.57, 1.23) 1.05 (0.82, 1.34) Smoking Never Reference Reference Reference Reference Former 1.35 (1.14, 1.61)*** 1.31 (1.09, 1.57)** 0.90 (0.54, 1.52) 1.16 (0.74, 1.82) Current 1.49 (1.25, 1.78)*** 1.37 (1.13, 1.65)** 1.06 (0.62, 1.81) 1.17 (0.74, 1.86) Stimulant drug use c No Reference Reference Reference Reference Yes 1.22 (1.14, 1.30)*** 1.21 (1.14, 1.29)*** 0.97 (0.82, 1.14) 0.99 (0.84, 1.16) IDU d No Reference Re ference Reference Reference Yes 1.39 (1.16, 1.67)*** 1.21 (1.05, 1.40)** 1.57 (1.02, 2.43)* 0.75 (0.44, 1.25) Hepatitis C virus antibody Negative Reference Reference Reference Reference Positive 1.39 (1.14, 1.70)*** 1.23 (1.03, 1.47)* 1.00 (0.70, 1.45) 0.99 (0.74, 1.33) HIV s erostatus HIV negative Reference Reference Reference Reference Reference HIV positive 1.92 (1.63, 2.26)*** 1.70 (1.44, 2.01)*** 1.04 (0.77, 1.41) 1.27 (0.94, 1.72) 1.31 (0.9 Medical Marijuana Law 1.08 (0.94, 1.23) 1.54 (1.22, 1.94)*** 1.50 (1.11, 2.04)** Note. a HSD=High school diploma; b C ESD= Center for Epidemiological Depression Scale; c Includes crack cocaine, methamphetamines (or speed, meth or ice), Ec stasy; d IDU=Intravenous drug use;

PAGE 62

62 Table 3 5. Prevalence Ratios of Risk Factors Associated wit h Daily Marijuana Use among HIV positive Men Prevalence Ratios (95% CI) Early Cohort Late Cohort 1996 2013 2003 201 3 2003 2013 Characteristics Univariate Multivariable Multivariable Univariate Multivariable Age, per 10 year increase 0.80 (0.59, 1.09) 0.90 (0.66, 1.23) 0.91 (0.61, 1.37) 0.82 (0.65, 1.04) 0.90 (0.85, 0.95)*** Race White, n on Hispanic Reference Reference Reference Reference Reference Black, non Hispanic 1.24 (0.82, 1.91) 1.46 (0.90, 2.38) 2.00 (1.16, 3.46)* 0.99 (0.90, 1.09) Other 0.62 (0.25, 1.53) 0.53 (0.20, 1.40) 0.48 (0.11, 2.07) 1.50 (0 .76, 2.98) 1.04 (0.91, 1.19) Education HSD or less a Reference Reference Reference Reference Reference Some college/college deg. 0.78 (0.49, 1.23) 0.81 (0.51, 1.27) 1.10 (0.52, 2.33) 0.60 (0.41, 0.87)** 0.95 0.88, 1.02) Graduate work or more 0.49 (0.29, 0.84)** 0.54 (0.32, 0.93)* 0.86 (0.39, 1.89) 0.35 (0.15, 0.79)* 1.03 (0.89, 1.20) Employment Employed Reference Reference Unemployed 0.99 (0.79, 1.24) 1.05 (0.89, 1.24) Study center Pittsburgh Reference Reference Reference Baltimore/Washington DC 0.75 (0.44, 1.28) 1.48 (0.80, 2.72) 1.05 (0.92, 1.19) Chicago 0.69 (0.40, 1.18) 1.24 (0.66, 2.33) 0.83 (0.75, 0.91)*** Los Angele s 0.90 (0.58, 1.40) 0.85 (0.83, 1.08) Depressive symptoms b CESD < 16 Reference Reference 1.03 (0.91, 1.17) 1.04 (0.89, 1.21) Alcohol use None Reference Reference Low moderate 0.89 (0.71, 1.12) 1.11 (0.84, 1.49) Heavy 0.91 (0.69, 1.19) 1.09 (0.82, 1.47) Smoking Never Reference Reference Former 1.01 (0.55, 1.86) 1.17 (0.63, 2.1 5) Current 1.03 (0.53, 1.98) 1.25 (0.67, 2.31) Stimulant drug use c No Reference Reference Reference Reference Yes 1.02 (0.83, 1.27) 0.96 (0.78, 1.19) IDU d No Refer ence Reference Yes 1.14 (0.80, 1.62) 0.61 (0.29, 1.27) Hepatitis C virus antibody Negative Reference Reference Reference Reference Positive 1.19 (0.78, 1.80) 0.92 (0.56, 1.48) HAART Use e No Reference Reference Reference Yes 0.98 (0.84, 1.14) 0.95 (0.76, 1.18) 0.99 (0.97, 1.01) CD4 + count > 500 Reference Reference Reference < 200 1.22 (0.91, 1.64) 0.91 (0.6 3, 1.31) 200 and < 500 0.98 (0.84, 1.15) 1.02 (0.86, 1.21) 1.00 (0.99, 1.01) HIV viral load Undetectable Reference Reference Reference Detectable 1.06 (0.90, 1.24) 0.91 (0.74, 1.12) 1.01 (0.99, 1 Medical Marijuana Law 1.02 (0.71, 1.48) 1.72 (1.27, 2.33)*** 1.01 (1.00, 1.02)* Note. a HSD=High school diploma; b C ESD= Center for Epidemiological Depression Scale; c Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy ; d IDU=Intravenous drug use; e HAART =Highly active antiretroviral therapy;

PAGE 63

63 CHAPTER 4 TRAJECTO RIES OF MARIJUAN A USE AND PREDICTORS OF TRAJECTORIES Marijuana use is common among persons living with HIV as past studies have reported prevalence rates of current marijuana use between 24% to 56% (Bruce et al., Prentiss et al., 2004) as compared to approximately 7% in the general United States population ( Substance Abuse and Mental Health Services Ad ministration, Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings, NSDUH Series H 46, HHS Publication No. (SMA) 13 4795. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2013. n.d.) Men who have sex with men report higher rates of current and past year marijuana use than their heterosexual counterparts (Cochran, Ackerman, Mays, & Ross, 2004; McCabe, Hughes, Bostwick, West, & Boyd, 2009) Several studies report that HIV positive individuals use marijuana to allevia te stress, anxiety, depression, HIV related symptoms and side effects of antiretroviral therapy (ART) 2004; Ware et al., 2003) In one recent study, among HIV positive individuals who inject drugs and who recentl y seroconverted, heavy cannabis use was associated with lower plasma viral load levels (Milloy et al., 2015) The therapeutic effects of marijuana are proposed to be mediated via the actions of act ive cannabinoid chemicals in marijuana cannabidiol at specific receptor sites: cannabinoid receptors (CB2) located mainly on cells and tissues of the immune system (Borgelt, Franson, Nussbaum, & Wang, 2013; This chapter was previously published Okafor, C. N., Cook, R. L., Chen, X., Surkan, P. J., Becker, J. T., Shoptaw, S ., Plankey, M. W. (2016). Trajectories of Marijuana Use among HIV seropositive and HIV seronegative MSM in the Multicenter AIDS Cohort Study (MACS), 1984 2013. AIDS and Behavior 1 14. http://doi.org/10.1007/s10461 016 1445 3 and is being used with permiss ion

PAGE 64

64 Rom & Persidsky, 2013) In contrast the primary psychoactive cannabinoid in marijuana:THC binds to and activates another receptor site: cannabinoid receptor (CB1) located mainly in areas of the brain (Herkenham et al., 1990) to produce the euphoric and cognitive impairing effects of THC (Fusar Poli et al., 2009) Accordingly, there are concerns that marijuana use may be associated with poorer HIV treatment outcomes. Previous studies have found marijuana use to be associated with decreased cognitive functioning (Cristiani et al., 2004; Gonzalez et al., 2011) as well as reduced ART adherence (Bonn Miller et al., 2012; Corless et al., 2009) which is crucial for PLWH as optimal adherence to ART medications is required for lo ng term viral suppression (Viswanathan et al., 2014) Effective prevention strategies to reduce unhealthy or harmful marijuana use require an in depth understanding of sub groups with different patterns of use. Despite the published evidence that ma rijuana use is common among HIV positive individuals and MSM and the potential adverse health outcomes associated with its use in these population, very little is known about the patterns of marijuana use or how patterns of marijuana use may change over ti me in these population. Developmental research suggests different rather than similar pathways via which individuals initiate and progress to unhealthy or problem substance use (Arnett, 2005; Kellam & Other s, 1982) over the life course. For instance, individuals who start using substances at an early age have increased risk of progressing to problem use and developing use disorders (DeWit, Adlaf, Offord, & Ogborne, 2000; Tarter, Horner, & Ridenour, 2012) Among HIV positive women, depressive symptoms and a positive hepatitis C infection was associated with a pattern of persistent heavy drinking over time

PAGE 65

65 (Cook et al., 2013) Another study found that low i ncome and concurrent substance use predicted consiste nt hazardous drinking among HIV positive MSM (Marshall et al., 2014) Therefore, understanding the natural history of marijuana use and the identification of different trajectories of use over time is important in order for intervention programs to be most effective. For instance, the identification of different patterns of marijuana use over time can help characterize sub groups of individua ls with the greatest risk of progressing to heavy or unhealthy patterns of marijuana use and reveal unique predictors of such patterns which can be used to inform targeted intervention programs. This study is unique in that it follows HIV positive individu als and MSM longitudinally over an extended period of many years to characterize the natural history of their marijuana use. Past studies on substance use patterns in these population have often focused on alcohol or heavy episodic drinking (Marshall et al., 2014, 2015) cigarette smoking (Akhtar Khaleel et al., 2016) or stimulant use (Lim et al., 2010) Therefore, the objectives of th e present study are to characterize the longitudinal trajectories of m arijuana use in a sample of HIV positive and HIV negative MSM over a period of 29 years, and to identify factors associated with unique trajectories of marijuana use and as well as those that can change over time that may modify the course of the trajectory. Methods The Multicenter AIDS Cohort Study ( MACS) The MACS is an ongoing prospective cohort study of the natural and treated history of HIV infection among MSM in the United States. A total of 6,972 men were enrolled during the history of the project in three waves: 4,954 men in 1984 1985, 668 in 1987 1991, and 1350 in 2001 2003 and at four centers located in Baltimore/Washington

PAGE 66

66 DC, Chicago, Los Angeles, and Pittsburgh. The study desig n of the MACS has been described previously (Roger Detels et al., 1992; Dudley et al., 1995; Kaslow et al., 1987) and only the design relevant to the present analysis is described here. The study questionnaires used in the MACS are available at www.statepi.jhsph.edu/macs/forms.html The institutional review boards at the respective recruitment centers and their community affiliates approved the MACS study protocols and all study participants provided inform ed consent. MACS participants return every 6 months for a physical examination, collection of blood specimens and completion of a detailed interview and questionnaires. The interview and questionnaires include demographic, psychosocial, behavioral, and med ical history data. The questions about their recreational drug use, including marijuana, alcohol, poppers, cocaine, crack, heroin, methamphetamine, ecstasy, injection drug use as well as smoking history since their last visit were collected using audio com puter assisted self interviewing (ACASI), an approach previously demonstrated to provide more accurate assessments of administered questionnaires among MSM (Gribble et al., 2000) This analysis included data collected from marijuana standardized use questions from semiannual study visits 1 (April, 1984 September, 1984) through visit 59 (April 2013 September 2013). The study sample included 3,658 participants who had data about marijuana use for at least 25% of their possible study visits during the follow up period. Specifically, the men enrolled in 1984 1985 and 1987 1991 ha d 15 and 13 visits or more, while the men enrolled in 2001 2003 had six or more visits. The median years of follow up was 11.5 years (interquartile range: 9.5 21.0).

PAGE 67

67 Measures Outcome Measure. The MACS protocols assessed marijuana use at each study visit Have you used any pot, marijuana or hash since your last visit No users. queried with the following qu How often did you use Pot, Marijuana or Hash since your last visit Covariates Sociodemographic Characteristics. The baseline visit (or index visit) was used to define a three level categorical variable for race/ethnicity status (Non Hispanic, White, Non Hispanic Black and other), educational attainment (high school diploma or less, some college or college degree, graduate work or more) and employment (employed or unemployed). Participants were classified according to the MACS study center they were enrolled and whether they were enrolled prior to or after 2001. Depressive symptoms. The Center for Epidemiologic Studies Depression (CES D) scale, was used to measure clinically significant symptoms of depression (Radloff, 1977) The CESD includes components of depressed mood, feelings of worthlessness, sense of hopelessness, sleep disturbance, loss of appetite, and concentration difficulties. Scores on the CES D of 16 or more suggests a clinically significant level of psychological distress (Radloff, 1977) Alcohol Use. Using data regarding frequency of drinking and average number of alcoholic drinks alcohol consumption at baseline and at each study visit was nks per week or any binge drinking i.e. 5 or more drinks per occasion)

PAGE 68

68 Alcohol Abuse and Alcoholism. Helping patien low or moderate use (any drinking not meeting criteria for hazardous use) or no alcohol use. Tobacco Use. Participants were classified as nev er, former and current smokers of cigarettes at each study visit. Participants were asked two questions including: whether they ever smoked cigarettes and whether they smoke cigarettes now. ent smokers, pack years of smoking at initial visit and at each subsequent visit was calculated using per day. Stimulant/Recreational Drug Use. Participants were considere d to be users of stimulant drugs if they reported the use of any of the following drugs at baseline and at each study visit: (1) crack cocaine, (2) other forms of cocaine, (3) methamphetamines methylenedioxy N methylamphetamine). Clinical Factors. The MACS protocols assessed HIV serostatus using an enzyme linked immunosorbent assay with confirmatory Western blot tests on all MACS isit and at each study visit for participants who were initially HIV negative However, only participants who were HIV positive at the time of enrollment were included. Standardized flow cytometry was used to quantify CD4+ T

PAGE 69

69 lymphocyte subset levels by eac h MACS site (Giorgi et al., 1990) and categorized 200/mm 3 201 500/mm 3 and > 500/mm3. Levels of plasma HIV RNA were measured using either the standard reverse transcription polymerase chain reaction assay (Roche Nutley, NJ) or with the Roche ultrasensitive assay (Roche Diagnostics. Standardized vir al load measures (across different assays) were used to create a dichotomous variable to denote detectable (> 400 copies per mL) versus undetectable viral load. H CV infection status was categorized as HCV negative if HCV antibody testing was negative. Part icipants were classified at each MACS study visit as HCV positive if they were found to be in the process of seroconversion, acute infection, chronic infection, clearing (between RNA+ and RNA ), or previously HCV positive, but now clear of HCV RNA. In addi tion to above mentioned covariates described above I considered that factors specific to HIV infection might influence trajecto ries of marijuana use among HIV positive participants ART use was dichotomized as use of any ART since last study visit versus no therapy used. Attrition. Two binary variables were constructed and used as covariates to adjust for the impact of attrition: one for participants who dropped out or were lost to follow up (n=1,394) and another for those who died within the follow up pe riod (n=643). Data Analysis Self reported frequency of marijuana across the follow up period identified trajectories using a semi parametric group based mixture method: PROC TRAJ SAS procedure (Jones, Nagin, & Roeder, 2001) frequency of marijuana use over their follow single model consisting of distinct trajectories. The procedure c alculates the probability of each participant belonging to each trajectory group and assigns individuals

PAGE 70

70 into trajectories based on their highest probability of trajectory membership (Jones et al., 2001; Nagin, 2005) Participants were followed from the time of enrollment until either the time of death, lost to follow up or until the end of the study period (MACS visit 59 or September 2013). A sequence of models with 2 5 trajectories, assuming linear, quadratic and cubic shape of the trajectory group curves were fit Several factors were considered in determining model fit and the optimal number of trajectory groups (and trajectory shape) that best represented the heterogeneity of groups within the data: including a priori knowledge from previous research on trajectories of marijuana use (Brook et al., 2011; Brook JS, Zhang C, & Brook DW, 2011; Ellickson et al., 2004; Nelson et al., 2 015; Passarotti et al., 2015; Schulenberg et al., 2005; Windle & Wiesner, 2004) model fit statistics including Bayesian information criterion (BIC) (Raftery, n.d.) Akaike Inf ormation Criterion (AIC; smaller value suggesting better fit) (Akaike, 1974) average posterior probability (entropy) of group membership (a measure of classification quality; greater than 0.7 suggest adequate internal reliability) (Nagin, 2005) significance of the shape of the trajec tory group curves (e.g. linear, quadratic) iterative process, starting with a quadratic specification for the shape of the trajectory group curves and assessing whether an ad ditional group resulted in a better model fit based on the criteria described above. H igher order shapes of the trajectory group curves (e.g. cubic) were then estimated and non significant terms subsequently dropped Models used a zero inflated Poisson dis tribution to account for the large proportion of participants who reported not using marijuana.

PAGE 71

71 C ovariates of inter est were included in the trajectories models, after selecting the model with the optimal numbe r of groups and shape of change. For this analy sis, two types of covariates were considered : time fixed/risk factors of trajectory group membership and time varying covariates. These time fixed/risk factors comprise characteristics established before or at the time of the initial period of trajectories that may serve to predict membership in a given trajectory. Time varying covariates measured during the course of the trajectory provide trajectory group specific estimates of whether these covariates alter the course of the trajectory. One advantage of t he PROC TRAJ software is that it allows for joint estimating of the parameters that describe the shape of the trajectory group curves, adjusted odds ratio (for risk factors for trajectory membership) and the coefficient estimates (for the time varying cova riates). In order to account for potential differences in marijuana use by geographic region/site and MACS enrollment cohort, all models included variables for MACS center and enrollment cohort. Models were estimated for all participants as well as by HIV s erostatus M odels that included all participants was adjusted for sociodemographic characteristics, depressive symptoms, substance use variables, hepatitis C infection status, attrition variables and HIV sero status. In the analysis restricted to HIV posit ive parti cipants, I included other c linical factors relevant to HIV positive status such as ART use, CD4 counts and viral load detectability. All analysis was performed in SAS 9.4 (SAS Institute, Inc., Cary, NC). Results Sample Characteristics The 3,658 p articipants in this study contributed 105, 595 person visits, the median number of visits was 23 (Interquartile Range [IQR]: 19, 42) representing

PAGE 72

72 approximately 11 years (Table 4 1). Among those who were HIV positive (n=1,439 or 39%), the mean age at baseli ne visit was 35 years (STD: 7.7), median number of visits was 23 (IQR= 18, 38), the majority (62%) were Non Hispanic, Whites and 24% were non Hispanic, Blacks (Table 4 1). At baseline, among the HIV positive participants, marijuana use was high (62%: 52% u sed less than daily and 10% daily), 90% used alcohol (25% meeting criteria for hazardous use), 67% reported stimulant/recreational drug use, 44% were current smokers and 29% were classified as having clinically significant depressive symptoms (CES ble 4 1). At baseline, the HIV negative participants in this study reported lower marijuana use (52%; 48% less than daily and 6% daily use), stimulant/recreational drug use (59%), rates of current smoking (35%) and depressi ve symptoms (21%) than the HIV po sitive participants. Both groups were similar with regard to alcohol use (Table 4 1). Marijuana Trajectories reported frequency of marijuana use across the follow up period identified four groups with d istinct trajectori es of marijuana use. A four group solution was selected based on model parsimony, interpretability of trajectories, BIC and AIC values, significance of the polynomial growth terms, average posterior probabilities (which ranged between 0.9 5 0.99) and trajectory g 6 4 9 displays m odel fit information and average posterior probabilities of all models Figure 4 1 displays the trajectories of marijuana use of these four groups, which I Ab stainers or Infrequent use Increasers Chronic high abstained or used infrequently during the follow up period characterized the abstainer or infrequent use group (65% of the entire sample). The decreaser group (1 3% of the

PAGE 73

73 entire sample) consisted of a group of men who reduced their use from nearly weekly use to infrequent use over the follow up period. The increaser group (12% of the entire sample) comprised a group of men who initially decreased their use during the first 10 years of follow up, after which they began to increase their use over time. The chronic high group (10% of the entire sample) represents a group of men who persistently used marijuana nearly daily ove r the follow up period. Figure 4 2 displays trajectories of marijuana use among HIV positive participants: 61% were in the abstainer/infrequent use group 14% were in the decreaser group, 14% in increaser group and 11% in the chronic high group. Table 4 2 displays the baseline characteristics of t he entire sample by the four identified trajectory groups of marijuana use. The median numbers of visits were similar across the different marijuana trajectory groups and on average participants in the abstainer/infrequent use group were older at baseline compared to the other groups. Frequency of marijuana use at baseline varied across the marijuana trajectory groups: as the proportion of daily users were <1% in the abstainer/infrequent users 3% in the decreasers 7% in the increasers and 54% in the chron ic high group. Racial status and detec table HIV viral load (among HIV positive participants) were similar across the marijuana trajectory groups. Time Stable and Time Varying Factors Associated with Trajectories Among All Men A HIV positive status relat ive to being HIV negative was associated with significant increased odds of membership in all groups reporting marijuana use: decreaser [Adjusted odds ratio (AOR) =1.61, 95% confidence interval (CI): 1.27 2.05; p<0.001], increaser (AOR=1.63, 95% CI: 1.26 2.12; p<0.001) and chronic high

PAGE 74

74 (AOR=1.72, 95% CI: 1.32 2.23; p<0.001 ; Table 4 3 ) as compared to being in the abstainer/infrequent use group. The men who reported any marijuana use relative to those who abstained or used infrequently during the follo w up period were significantly more likely to be younger. In addition, non Hispanic blacks, as compared to non Hispanic whites had significantly increased odds of membership in the decreaser (AOR)=1.42, 95% CI: 1.01 1.98; p=0.041) and increaser (AOR=1.43 95% CI: 1.00 2.06; p=0.046 ; Table 4 3 ) groups relative to those in the abstainer/infrequent use group. Those who had completed graduate work or more (at their baseline visit) as compared to those with a high school diploma or less had reduced odds of m embership in decreaser (AOR=0.61, 95% CI: 0.44 0.87; p<0.001 ; Table 4 3 ) and chronic high (AOR=0.61, 95% CI: 0.41 0.89; p<0.001 ; Table 4 3 ) groups relative to the abstainer/infrequent use group. There were several significant associations with MACS stu dy center on membership in the marijuana use trajectories. Those enrolled in the Los Angeles center as compared to nearly all other MACS centers had higher odds of reporting any marijuana use relative to being in the abstainer/infrequent use group. Partici pants who were lost to follow up and those who died during the follow up period had significantly increased odds of membership in only the increaser trajectory group. Table 4 3 also displays the results for the effects of time varying covariates. These e stimates are trajectory group specific: i.e. they indicate whether a covariate measured over time alters the trajectory of marijuana use within that trajectory group only. The estimates for the time varying covariates are growth parameters that describe th e amount of change/deviation from the average long term trajectory of marijuana use within the trajectory. Depressive symptoms were significantly associated with increasing

PAGE 75

75 marijuana use within the abstainer/infrequent (p<0.05) decreaser (p<0.01) and chr onic high groups (p<0.001). Alcohol use (low/moderate and hazardous use), being a current smoker, stimulant/recreational drug use and IDU were each significantly associated with A mong HIV positive Participants HIV positive participants who reported any m arijuana use were significantly younger relative to those who abstained or used infrequently ( Table 4 4 ). Non Hispanic, blacks as compared to non Hispanic, whites, had increased odds of membership in the decreaser group (AOR=1.82, 95% CI: 1.10 2.99; p=0. 021). Table 4 4 displays re sults for the effect of time varying covariates on trajectory levels. Alcohol use over time was associated with increasing marijuana use in all trajectory groups (all ps <0.001); however, hazardous use was not associated with inc reasing marijuana use in the chronic high group (p>0.05). Current cigarette smoking and stimulants/recreational drug use during follow up was significantly associated with increasing marijuana use across all trajectory groups (all ps <0.001). ART use was s ignificantly associated with decreasing marijuana use in the abstainer/infrequent (p=0.047) and increaser groups (p=0.045). Detectable HIV RNA over time was significantly associated with increasing marijuana use in the increaser group (p=0.002). Discussio n This study utilized data from the MACS cohort to assess different patterns of marijuana use and to examine both risk factors and time varying correlates associated with the different trajectories of marijuana use. Our analysis revealed that MSM in the M ACS exhibited four distinct trajectories of marijuana use over time including abstainer/infrequent users, decreasers increasers and chronic high users. Most of the men in this cohort displayed a pattern of abstaining or infrequent use over time (65%)

PAGE 76

76 whil e about 10% who used daily or near daily at their index visit continued this pattern of use over their follow up visits. About a quarter of the men changed their pattern of use over time, either decreasing (~13% of the men) or increasing use (~12% of the m en). Overall, our analysis suggested that these patterns of marijuana use over time were similar for both HIV positive and HIV negative participants. In the analys is among all men, HIV positive status was associated with membership across all three traject ory groups reporti ng any marijuana use. Among HIV positive participants, having a detectable HIV RNA over time was associated with increasing marijuana use only among the men that increased their marijuana use during the follow up period. Self re ported ART use over time in HIV positive men was associated with reducing marijuana use in the abstainer/infrequent and increaser groups. Overall, alcohol, cigarette, stimulant/recreational drug use and IDU over time were associated with increasing marijuana use in nearly all trajectory groups. T his study is unique because there has not been a previous study that has examined trajectories of marijuana use in HIV positive and HIV negative MSM over a long period of follow up. Prior studies that have assessed trajector ies of marijuana use have focused on adolescents transitioning into young adulthood (Brook et a l., 2011; Brook JS et al., 2011; Ellickson et al., 2004; Nelson et al., 2015; Passarotti et al., 2015; Schulenberg et al., 2005; Windle & Wiesner, 2004) or racial/ethnic minorities (Juon et al., 2011; Pahl, Brook, & Koppel, 2011; Whitesell et al., 2013) with a few studies reporting traj ectories of use covering adulthood (Brook, Lee, & Brook, 2014; Brook, Lee, Finch, & Brook, 2014; Pardini, Bechtold, Loeber, & White, 20 15) Direct comparisons of the results from our study with prior research may not be straightforward due to the

PAGE 77

77 different populations studied and age periods covered. However, nearly all studies on trajectories of marijuana use have identified a group tha t abstained or used infrequently, with some identifying a chronic high user group and a few identifying a group that increased (Ellickson et al., 2004; Schulenberg et al., 2005) and decreased (Schulenberg et al., 2005; Windle & Wiesner, 2004) their use. The present study found that a HIV positive status was associated with membership in the decreaser increaser and chroni c high marijuana trajectory group, a finding that suggests that overall HIV positive MSM in the MACS were more likely to use marijuana as compared to HIV negative MSM. This finding is consistent with a number of studies reporting higher r ates of marijuana use among HIV positive individuals as compared to HIV uninfected populations al., 2004; Harris et al., 2014; Mimiaga et al., 2013; Prentiss et al., 2004; Substance Abuse and Mental Health Services Administration, Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings, NSDUH Series H 46, HHS Publication No. (SMA) 13 4795. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2013. n.d.) HIV positive individuals report using marijuana to alleviate symptoms related to HIV infection as well as side effects of ART, althou gh a substantial portion of HIV positive individuals use marijuana recreationally About 16% of the HIV positive men in this study reported decreasi ng their marijuana use over time. This pattern of decreasing substance use over time was recently observed in a study of trajectories of stimulant use among MACS participants (Lim et al., 2010) The authors also found that the men who decreased stimulant drug use reported significant reduction in risky sexual practices over time.

PAGE 78

78 Among the HIV p ositive MSM in this study, having a detectabl e HIV RNA over time was associated with increasing marijuana use among individuals in the increaser group, but not among the men in the decreaser or chronic high groups. Accordingly, the study found that ART use over time was associated with decreasing mar ijuana use in the abstainer/infrequent and increaser groups. It is important to note that the assessment procedures used in this study make it difficult to ascertain that ART use preceded marijuana use. However, these findings provide some reassurance that there may not be an urgent need to intervene; though, there is a need to continue to study the long term effects of marijuana use on ot her health outcomes both in HIV positive and HIV negative individuals. In the data presented here, among the entire samp le as well as HIV positive individuals, younger age was associated with membership in all marijuana trajectory groups and being non Hispanic, black was associated with membership in the decreaser and increaser groups. In addition, alcohol use, cigarette sm oking, stimulants/recreational drug use and depressive symptoms over time served to increase marijuana use within nearly all marijuana trajectory groups. This finding is consistent with previous studies that found significant overlap between several types of drug use and other psychosocial health problems (Halkitis et al., 2014, 2015; Stall et al., 2003) Accordingly, any prevention approaches to mitigate these behaviors should not focus on one of these behaviors or conditions but must consider these co occurring conditions holistically. This study has some limitations, which I highlight in order for some caution to be exercised in the interpretation of our study findin gs. The study was restricted to MACS

PAGE 79

79 participants who had at least 25% or more study visits in order to estimate stable trajectory models. However, at baseline, those included in this study differed from those not included on a number of sociodemographic a nd clinical characteristics as well as use of substances including marijuana (Table 4 5) Therefore, different trajectories of marijuana use may have emerged if these participants had been included in our study. Furthermore, in the MACS, data on substance use (including marijuana use) was obtained via Audio Computer Assisted Interview (ACASI). Although this method has demonstrated good validity in obtaining sensitive information such as drug use in studie s of HIV positive individuals (Macalino, Celentano, Latkin, Strathdee, & Vlahov, 2002) a nd among MSM (Gribble et al., 2000) the data reported here related to substance use may be subject to social desirability bias and most likely an under reporting with a potential underestimation of the true trajectories of marijuana use. Related t o this issue is other biases related to participation in a large ongoing cohort study like the MACS, participant attrition due to dropouts and mortality, which may result in an underestimation of long term estimates of marijuana use. I ndeed, in the present study found that men who increased their marijuana use and those with chronic high use over time were significantly more likely to die or to drop out during follow up as compared to the abstainer/infrequent group. What this suggests is that the attrition in these groups may have precluded us from identifying what their patterns of marijuana use would have been if they had remained in the study. In addition, participants in the MACS represent a highly cooperative cohort of MSM retained in an ongoing cohort study and thus our findings here may not be generalizable to the larger MSM population. Finally, the semi parametric group based modelling used in this study has

PAGE 80

80 been criticized for its tendency to over identify trajectory groups within a population (Bauer & Curran, 2003) Accordingly, Nagin and Tremblay caution that groups extracted from the group based traj ectory models should be thought of as approximations of the more complex underlying reality of individual level trajectories of a behavior (Nagin, 2005; Nagin & Tremblay, 2005) and as such reification of trajectory groups should be done with caution. In sum, this study used data from a large sample, with a long period of follow up and utilized frequency measures of marijuana use to describe the na tural history of marijuana use in HIV positive and HIV negative MSM. Our study revealed different trajectories of use over time: with approximately 1 in 10 of the men em erg ed as chronic heavy users or increased their use over time. Future investigations need to determine whether long term patterns of heavy use are associated with adverse consequences e specially among HIV positive persons.

PAGE 81

81 Table 4 1. Characteristics of M ACS Men Used in The Trajectories at Baseline All men (N=3,658) n (%) HIV negative (n=2,219) n (%) HIV positive (n=1,439) n (%) No of visits, median, IQR 23 (19, 42) 22 (20, 44) 23 (18, 38) Age (mean, SD) 35 (8.2) 35 (8.4) 35 (7.7) Race, n (%) Non Hispanic, Whites 2692 (74) 1797 (81) 895 (62) Non Hispanic, Blacks 620 (17) 278 (13) 342 (24) Other 346 ( 9 ) 144 ( 6 ) 202 (14) Education, n (%) High school diploma or less 621 (17) 289 ( 13) 332 (23) Some college or college degree 1797 (49) 1057 (48) 740 (52) Graduate work or more 1227 (34) 864 (39) 363 (25) Unemployed, n (%) 370 (10) 200 ( 9 ) 170 (12) Study center, n (%) Baltimore/Washington DC 94 2 (26) 622 (28) 320 (22) Chicago 766 (21) 424 (19) 342 (24) Pittsburgh 813 (22) 557 (25) 256 (18) Los Angeles 1137 (31) 616 (28) 521 (36) Study enrollment, n (%) Pre 2001 2726 (75) 1818 (82) 908 (63) Post 2001 932 (25) 401 (18) 531 (37) Depressive symptoms, n (%) CESD < 16 2642 (76) 1669 (79) 973 (71) 844 (24) 449 (21) 395 (29) Alcohol use, n (%) None 312 ( 9 ) 177 ( 8 ) 135 (10) Low/moderate 2441 (68) 1533 (70) 908 (65) Hazardous use a 826 (23) 468 (21) 358 (25) Smoking, n (%) Never 1494 (41) 968 (44) 526 (37) Former 733 (20) 461 (21) 272 (19) Current 1394 (39) 774 (35) 620 (4 4) Cumulative pack years, median (IQR), y 2.2 (0, 18) 1.0 (0, 18) 3.9 (0, 18) Stimulants/recreational substance use, n (%) 2259 (62) 1308 (59) 951 (67) IDU, n (%) 296 ( 8 ) 109 ( 5 ) 187 (14) Positive Hepatitis C virus antibody, n (%) 262 ( 7 ) 104 ( 5 ) 158 (11) Detectable HIV RNA (>40 copies/ml), n (%) --1090 (91) Current CD4 + count (cells per cubic milliliter) < 200 --129 ( 9) --517 (36) > 500 --779 (55) ART use, n (%) --345 (24) History of AIDS, n (%) --557 (39) Marijuana use, n (%) None 1578 (43) 1031 (47) 547 (39) Less than daily 1791 (50) 1054 (48) 736 (52) Daily 258 ( 7 ) 123 ( 6 ) 135 (10) Note a Hazardous alcohol use defined as > 14 drinks per week or Binge Drinking (1.e. 5 or more drinks per occasion); b Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy. IQR=Interquartile Range. SD= Standard Deviation. CESD= Center for Epidemiological Depression Scale. ART=Antiretroviral Therapy; IDU=Intravenous Drug Use

PAGE 82

82 Table 4 2. Characteristics of MACS Men at Baseline by Marijuana Trajectories Characteristics Abstainer (n=2,380) Decreasers( n=483) Increasers (n=425) Chronic high ( n=370) 2/F P value No of visits, median (IQR) 23 (19, 43) 23 (19, 45) 21 (18, 28) 23 (18, 45) 10.5 <.0001 Age, mean (SD) 36 (8.4) 33 (7.8) 33 (7.2) 32 (6.8) 28.3 <.0001 Race, n (%) 8.2 .2173 White, non Hispanic 1769 (74) 334 (69) 313 (79) 2 76 (75) Black, non Hispanic 389 (16) 95 (20) 79 (19) 57 (15) Other 222 (10) 54 (11 ) 33 ( 8 ) 37 (10) Education, n (%) 61.7 <.0001 High school diploma or less 375 (16) 96 (20) 81 (19) 69 (19) Som e college or college degree 1101 (46) 275 (57) 208 (49) 213 (58) Graduate work or more 896 (38) 111 (23) 134 (32) 86 (23) Unemployed, n (%) 225 ( 9 ) 65 (13) 37 ( 9 ) 43 (11) 8.9 .0302 Study center, n (%) 54. 1 <.0001 Baltimore/Washington DC 667 (28) 98 (20) 114 (27) 63 (17) Chicago 495 (21) 108 (22) 89 (21) 74 (20) Pittsburgh 555 (23) 95 (20) 73 (17) 90 (24) Los Angeles 149 (35) 182 (38) 149 (35) 143 (39) Study enrollment, n (%) 7.3 .0611 Post 2001 626 (26) 132 (27) 90 (21) 84 (23) Depressive symptoms, n (%) 20.6 .0001 503 (22) 140 (30) 97 (24) 104 (30) Alcohol use, n (%) 113.1 <.0001 None 261 (11) 17 ( 4 ) 16 ( 4 ) 18 ( 5 ) Low/moderate 1635 (70) 305 (64) 274 (67) 227 (62) Hazardous use 432 (19) 152 (32) 120 (29) 122 (33) Smoking, n (%) 193.5 <.0001 Never 1141 (48) 135 (28) 134 (32) 84 (23) Former 481 (20) 78 (16) 88 (21) 86 (23) Current 734 (31) 267 (56) 194 (47) 199 (54) Cumulative pack years, medi an (IQR), y 0.05 (0, 15) 5.6 (0, 18) 5.2 (0, 22) 9.4 (0.13, 22) 17.9 <.0001 Stimulants/recreational substance use, n (%) 1231 (52) 397 (83) 332 (79) 299 (81) 100.4 <.0001 IDU, n (%) 143 ( 6 ) 72 (16) 40 (10) 41 (12) 52.7 <.0 001 Positive Hepatitis C virus antibody, n (%) 158 ( 7 ) 49 (10) 29 ( 7 ) 26 ( 7 ) 7.4 .0588 HIV positive 817 (34) 231 (48) 207 (49) 184 (50) 71.9 <.0001 236 (29) 48 (21) 29 (14) 32 (17) 27.7 <.0001 619 (76) 172 (74) 162 (78) 137 (74) 1.2 .7417 Current CD4 + 10.8 .0927 < 200 84 (10) 16 ( 7 ) 18 ( 9 ) 11 ( 6 ) 310 (38) 79 (35) 64 (31) 64 (35) > 500 416 (52) 133 (58) 124 (60) 106 (59) 283 (35) 99 (43) 100 (48) 75 (41) 15.7 .0013 Lost to follow up 935 (39) 159 (33) 179 (42) 119 (32) 15.3 .0015 Died 336 (14) 95 (20) 104 (24) 83 (22) 41.1 <.0001 Marijuana use, n (%) 2231.0 <.0001 None 2049 (92) 218 (50) 140 (36) 20 ( 6 ) Less than daily 155 ( 7 ) 199 (46) 207 (54) 132 (40) Da ily 4 (<1) 15 ( 3 ) 39 ( 7 ) 181 (54) Note. a Hazardous alcohol use defined as > 14 drinks per week or Binge Drinking (1.e. 5 or more drinks per occasion); b Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy. IQR= Interquartile Range. SD= Standard Deviation. CESD= Center for Epidemiological Depression Scale. ART=Antiretroviral Therapy. IDU=Intravenous Drug Use; HIV positive participants. Chi square tests (for categorical variables) and ANOVAs (for continuou s variables) was used to examine differences in trajectory groups.

PAGE 83

83 Table 4 3. Risk Factors for Marijuana Trajectories a mong All Men Characteristics Abstainer (n=2,380) Decreasers (n=483) Increasers (n=425) Chronic high (n=370) AOR 95% CI AOR 95% CI AOR 95% CI AOR 95% CI Age, per year increase Reference 0.96 (0.95 0.98)*** 0.96 (0.95 0.98)*** 0.95 (0.94 0.97)*** Race (vs. White, non Hispanic) Black, non Hispanic Reference 1.42 (1.01 1.98)* 1.43 (1.00 2.06)* 1.01 (0.68 1.51) Other Reference 0.79 (0.53 1.18) 0.62 (0.39 0.98)** 0.75 (0.48 1.18) Education (vs. HSD or Less) Some college or college degree Reference 1.04 (0.78 1.39) 0.86 (0.62 1.18) 0.94 (0.68 1 .30) Graduate work or more Reference 0.61 (0.44 0.87)*** 0.85 (0.60 1.21) 0.61 (0.41 0.89)*** Unemployed Reference 1.23 (0.87 1.74) 1.14 (0.77 1.69) 1.17 (0.79 1.74) Study center (vs. Los Angeles) Baltimore/Wa shington DC Reference 0.66 (0.50 0.88)*** 0.74 (0.55 0.99)* 0.53 (0.38 0.75)*** Chicago Reference 0.81 (0.61 1.09) 0.72 (0.52 0.99)* 0.76 (0.55 1.05) Pittsburgh Reference 0.60 (0.44 0.82)*** 0.57 (0.41 0.79 )* 0.78 (0.57 1.07) Study enrollment (vs. Pre 2001) Post 2001 Reference 0.99 (0.74 1.38) 0.81 (0.56 1.17) 0.74 (0.51 1.08) HIV s erostatus (vs. HIV negative ) HIV positive Reference 1.61 (1.27 2.05)*** 1.63 (1. 26 2.12)*** 1.72 (1.32 2.23)*** Lost to follow up Reference 1.01 (0.78 1.28) 1.66 (1.28 2.17)*** 0.82 (0.63 1.08) Died Reference 1.24 (0.90 1.69) 2.22 (1.60 3.07)*** 1.22 (0.88 1.71) Time Varying Covariates Influenc Trajectory specific growth parameters (95% CI) 0.01 (0.00 0.02)* 0.06 (0.03 0.09)** 0.01 ( 0.02 0.05) 0.13 (0.10 0.17)*** Low/moderat e alcohol use 0.08 (0.07 0.10)** 0.69 (0.64 0.73)*** 0.72 (0.67 0.77)*** 0.24 (0.19 0.29)*** Hazardous alcohol use a 0.16 (0.14 0.19)** 0.81 (0.76 0.86)*** 0.83 (0.76 0.89)*** 0.25 (0.20 0.31)*** Current Smoker 0.07 ( 0.05 0.08)* 0.29 (0.26 0.33)*** 0.25 (0.21 0.29)*** 0.21 (0.18 0.24)*** Stimulants/Recreational drug use b 0.24 (0.22 0.25)*** 0.44 (0.41 0.47)*** 0.49 (0.45 0.52)*** 0.23 (0.19 0.26)*** IDU 0.14 (0.08 0.21)*** 0.22 (0.13 0.32)*** 0.46 (0.31 0.61)*** 0.17 (0.04 0.30)*** Positive Hepatitis C virus antibody 0.02 ( 0.04 0.00) 0.00 ( 0.04 0.04) 0.04 ( 0.15 0.09) 0.07 (0.02 0.13)*** Note. Models were estimated simultaneously within the Proc Traj software, a Hazardous alcohol use defined as > 14 drinks per week or Binge Drinking (1.e. 5 or more drinks per occasion); b Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy. IQR=Interquartile Range. SD= Standard Deviation. CES D= Center for Epidemiological Depression Scale. ART=Antiretroviral Therapy. IDU=Intravenous Drug Use. OR=Odds Ratio; AOR=Adju sted Odds ratio; P <0.05; ** P <0.01; *** P <0.001

PAGE 84

84 Table 4 4. Risk Factors for Marijuana Trajec tor ies a mong HIV positive Men Ch aracteristics Abstainer (n=820) Decreasers (n=235) Increasers (n=207) Chronic high (n=177) AOR 95% CI AOR 95% CI AOR 95% CI AOR 95% CI Age, per year increase Reference 0.98 (0.95 1.00)* 0.97 (0.95 0.99)* 0.95 (0.93 0.98)** Race (v s. White, non Hispanic) Black, non Hispanic Reference 1.82 (1.10 2.99)* 1.15 (0.70 1.91) 1.30 (0.76 2.22) Other Reference 1.24 (0.71 2.14) 0.66 (0.36 1.20) 0.70 (0.37 1.30) Education (vs. HSD or Less) Some c ollege or college degree Reference 0.98 (0.64 1.51) 1.11 (0.71 1.74) 1.02 (0.65 1.60) Graduate work or more Reference 0.56 (0.32 0.98)* 1.21 (0.72 2.02) 0.62 (0.35 1.10) Study center (vs. Los Angeles) Baltimore /Washington DC Reference 0.64 (0.39 1.05) 0.72 (0.46 1.12) 0.55 (0.33 0.91)** Chicago Reference 1.00 (0.65 1.55) 0.69 (0.45 1.07) 0.77 (0.48 1.23) Pittsburgh Reference 0.73 (0.43 1.24) 0.54 (0.32 0.90)* 0 .75 (0.45 1.23) Study enrollment (vs. Pre 2001) Post 2001 Reference 0.53 (0.32 0.89)** 0.70 (0.42 1.17) 0.51 (0.29 0.87)* History of AIDS Reference 1.04 (0.64 1.69) 1.16 (0.74 1.82) 0.81 (0.49 1.32) Lost to fol low up Reference 1.36 (0.86 2.14) 1.02 (0.63 1.65) 0.87 (0.54 1.42) Died Reference 1.24 (0.74 2.08) 1.36 (0.84 2.20) 1.07 (0.63 1.82) Time Varying Covariates Influencing Trajectory of Marijuana Use Within Each Trajectory Am ong HIV positive Trajectory specific growth parameters (95% CI) 0.01 ( 0.00 0.04) 0.15 (0.09 0.20)*** 0.06 ( 0.12 0.01)* 0.09 (0.03 0.15)** Low/moderate alcohol use 0.11 (0.09 0.14)*** 0.54 (0.47 0.62)*** 0.78 (0.70 0.85)*** 0.16 (0.08 0.25)*** Hazardous alc ohol use a 0.20 (0.16 0.25)*** 0.72 (0.62 0.82)*** 0.93 (0.83 1.04)*** 0.00 ( 0.10 0.09) Current Smoker 0.07 (0.05 0.10)*** 0.25 (0.20 0.31)*** 0.51 (0.45 0.57)*** 0.38 (0.32 0.43)*** Stimulants/Recreational drug use b 0.25 (0.22 0.27)*** 0.53 (0.47 0.59)*** 0.52 (0.46 0.58)*** 0.34 (0.28 0.39)*** IDU 0.09 (0.01 0.18)* 0.02 ( 0.11 0.16) 0.18 ( 0.34 0.03)* 0.41 (0.20 0.62)** Positive Hepatitis C virus antibody 0.00 ( 0.03 0.03 ) 0.28 (0.20 0.36)*** 0.03 ( 0.11 0.04) 0.25 (0.15 0.35)*** ART Use 0.04 ( 0.07 0.01)* 0.03 ( 0.02 0.09) 0.14 ( 0.21 0.07)* 0.02 ( 0.09 0.05) Detectable HIV RNA 0.00 ( 0.02 0.02) 0.01 ( 0.06 0.04) 0.09 (0 .04 0.14)** 0.00 ( 0.05 0.06) CD4 + count 0.02 ( 0.05 0.00) 0.02 ( 0.04 0.09) 0.05 ( 0.02 0.12) 0.03 ( 0.04 0.12) > 500 0.03 ( 0.06 0.00)* 0.03 ( 0.10 0.04) 0.04 ( 0.12 0.04) 0.26 (0.18 0.34)*** Note. Models were estimated simultaneously w ithin the Proc Traj software a Hazardous alcohol use defined as > 14 drinks per week or Binge Drinking (1.e. 5 or more drinks per occasion); b Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy. IQR=Interquartile Range. SD= Standard D eviation. CESD= Center for Epidemiological Depression Scale. ART= Antiretroviral Therapy. IDU=Intravenous Drug Use. OR=Odds R atio; AOR=Adjusted Odds ratio; P <0.05; ** P <0.01; *** P <0.001

PAGE 85

85 Table 4 5. Baseline Characteristics of MACS Men Included and Excluded From the Analysis Included (n=3,658) Excluded (n=2,680) 2/F P value n (%) n (%) Age (mean, SD) 35 (8.2) 33 (8.1) 51.6 <.0001 Race, n (%) 14.4 .0007 Non Hispanic, Whites 2692 (74) 1855 (69) Non Hi spanic, Blacks 620 (17) 542 (20) Other 346 ( 9 ) 278 (11) Education, n (%) 71.3 <.0001 High school diploma or less 621 (17) 610 (23) Some college or college degree 1797 (49) 1371 (52) Graduate work or more 1227 (34 ) 654 (25) Unemployed, n (%) 370 (10) 334 (13) 8.6 .0033 Study center, n (%) 19.1 .0003 Baltimore/Washington DC 942 (26) 647 (24) Chicago 766 (21) 686 (26) Pittsburgh 813 (22) 571 (21) Los Angeles 1137 (31) 776 (29) Study enrollment, n (%) 130.9 <.0001 Pre 2001 2726 (75) 2312 (86) Post 2001 932 (25) 368 (14) Depressive symptoms, n (%) CESD < 16 2642 (76) 1884 (75) 0.26 .6069 844 (24) 621 (25) Alcohol use, n (%) 25.8 <.0001 None 312 ( 9 ) 238 ( 9 ) Low/moderate 2441 (68) 1630 (62) Hazardous use a 826 (23) 746 (29) Smoking, n (%) 47.0 <.0001 Never 1494 (41) 905 ( 34) Former 733 (20) 502 (19) Current 1394 (39) 1242 (47) Cumulative pack years, median (IQR), y 2.2 (0, 18) 4.8 (0, 22) 22.1 <.0001 Stimulants/recreational substance use, n (%) 2259 (62) 1807 (68) 22.2 <.0001 IDU, n (%) 296 ( 8 ) 242 (10) 2.9 .0873 Positive Hepatitis C virus antibody, n (%) 262 ( 7 ) 253 (10) 11.0 .0009 HIV status Negative 2219 (61) 1176 (44) 175.1 <.0001 Positive 1439 (39) 1504 (56) 10 90 (91) 866 (97) 24.4 <.0001 Current CD4 + 16.8 .0002 < 200 129 ( 9) 172 (12) 517 (36) 612 (41) > 500 779 (55) 702 (47) 345 (24) 106 ( 9 ) 96.2 <.0001 557 (39) 1125 (75) 391.2 <.0001 Marijuana use, n (%) 30.3 <.0001 None 1578 (43) 936 (35) Less than daily 1791 (50) 1492 (56) Daily 258 ( 7 ) 229 ( 9 ) Note. a Hazardous alcoho l use defined as > 14 drinks per week or Binge Drinking (i.e. 5 or more drinks per occasion); b Includes crack cocaine, methamphetamines (or speed, meth or ice), Ecstasy. IQR=Interquartile Range. SD= Standard Deviation. CESD= Center for Epidemio logical Depr ession Scale. ART=Antiretroviral Therapy. IDU=I HIV positive participant; Chi square tests (for categorical variables) and ANOVAs (for continuous variables) were used to examine differences in trajectory groups.

PAGE 86

86 Table 4 6. Mo del Fit Statistics for Marijuana Trajectories for All Men Number of trajectories AIC BIC Prevalence of smallest trajectory 2 133445.5 133470.3 19.4% 3 122307.0 122344.2 12.4% 4 118414.1 118463.7 9.7% 5 114833.9 114895.9 5.4% Final 4 group solu tion 117859.8 117921.8 10.0% Note AIC=Alkaike Information Criterion, BIC=Bayesian Information Criterion Table 4 7. Average Posterior Probabilities of the Final Four Marijuana Trajector y Solution for All Men Marijuana trajectory groups Marijuana tr ajectory group Abstainer Decreaser Increaser Chronic high group Abstainer 0.992 0.017 0.016 0.000 Decreaser 0.005 0.968 0.013 0.013 Increaser 0.002 0.009 0.967 0.003 Chronic high user 0.000 0.004 0.002 0.983 Note Based on the final 4 group solution. P robabilities on the diagonal are average posterior probabilities of trajectory membership among persons assigned to that group, while, the probabilities on the off diagonal are the average posterior probabilities of group membership among persons not assig ned to that group Table 4 8. Model Fit Statistics for Tr ajector ies for HIV positive Men Number of trajectories AIC BIC Prevalence of smallest trajectory 2 53857.24 53878.33 21.5% 3 50408.71 50440.34 12.8% 4 48840.64 48882.81 7.7% 5 47919.24 47971.95 7.6% Final 4 group solution 48835.34 48882.78 10.9% Note AIC=Alkaike Information Criterion, BIC=Bayesian Information Criterion. Table 4 9. Average Posterior Probabilities of the Final Four Marijuana Trajectory Solution for the HIV posi tive Men Marijuana trajectory groups Marijuana trajectory group Abstainer Decreaser Increaser Chronic high group Abstainer 0.991 0.009 0.027 0.000 Decreaser 0.004 0.965 0.014 0.008 Increaser 0.003 0.017 0.956 0.004 Chronic high user 0.000 0.007 0.002 0.986 Note Based on the final 4 group solution. Probabilities on the diagonal are average posterior probabilities of trajectory membership among persons assigned to that group, while, the probabilities on the off diagonal are the average posterior probab ilities of group membership among persons not assigned to that group

PAGE 87

87 Figure 4 1. Trajectories o f marijuana use among 3,658 HIV positive and HIV negative participants in the Multicenter AIDS Cohort Study (MACS) 1984 2013. The solid lines represe nt the predicted probabilities of frequency of marijuana use conditional on membership in one of the four marijuana trajector y groups, while the dotted lines represent the actual frequency of marijuana use given group membership. The y axis represents the conditional probabilities of frequency of marijuana use, while the x axis represents the years of follow up. 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Frequency of Marijuana Use Years of Follow Up Chronic High (9.7%) Chronic High-Actual Increasers (11.8%) Increasers-Actual Decreasers (13.4%) DecreasersActual Abstainers (65.2%) Abstainers-Actual Weekly Monthly Less Often None Daily

PAGE 88

88 Figure 4 2.Trajectories of marijuana use among 1,439 HIV+ participants in the Multicenter AIDS Cohort Study (MACS) 1984 2013. The solid l ines represent the predicted probabilities of frequency of marijuana use conditional on membership in one of the marijuana tr ajectory groups, while the dotted lines represent the actual frequency of marijuana use given group membership. The y axis represen ts the conditional probabilities of frequency of marijuana use, while the x axis represents the years of follow up 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Frequency of Marijuana Use Years of Follow Up Chronic High (10.9%) Chronic High-Actual Increasers (13.6%) Increasers-Actual Decreasers (14.2%) Decreasers-Actual Abstainers (61.3%) Abstainers-Actual None L ess Often Monthly Weekly Daily

PAGE 89

89 CHAPTER 5 CUM ULATIVE MARIJUANA USE AND COGNTI VE CHANGE OVER 17 YEARS Marijuana use is co mmon among HIV positive individuals in the United States, with some studies reporting that 14% to 56% of PLWH to have used marijuana in the past six months ouza et al., 2012; Furler et al., 2004; MD, 2001; Mimiaga et al., 2013; Pence et al., 2008; Prentiss et al., 2004; Ware et al., 2003) In addition, recent data among HIV positive women found that daily use among users increased significantly by more than 2 folds in recent years HIV positive individuals who use marijuan a have touted its use to provide therapeutic benefits for a wide variety of symptoms including relief from pain, nausea, reduced appetite and improved mood many al., 2004) Most of these repor ts have been self report and there is limited empirical data on the efficacy and safety of its use among HIV positive individuals (Lutge et al., 2013) The current climate of increased permissiveness of marijuana use in the general United States population (Gallup, 2013) an d evolving state legislations allowing recreational and medical marijuana use may ha ve important public health a nd clinical consequences. These laws may increase access to and further increase rates of use among PLWH as most medical marijuana laws list HIV/AIDS as a qualifying condition for medical marijuana (Wilkinson et al., 2016) However, marijuana use has been associated with adverse health consequences, includin g neurocognitive impairment which HIV positive individuals are particularly vulnerable to (Robert K Heaton et al., 2011) There is evidence that marijuana use may be associated with impaired cognitive function particularly in domains of learning and memory among HIV uninfected samples

PAGE 90

90 2006) Data on the impact of marijuana use on other cognitive domains including psychomotor speed, executive function and motor skills have be en less conclusive (Crane et al., 2013; Crean et al., 2011; Gonzalez, 2007) HIV positive individuals are vulnerable to cognitive impairments via direct effects of the virus, comorbid conditions associated with HIV (e.g. hepatitis C inf ection), antiretroviral therapy, and other factors (e.g. metabolic, cardiovascular factors and ageing) (Clifford & Ances, 2013; Gannon et al., 2011; Sanmarti et al., 2014) Recent data estimates that 31% of HIV positive individuals experience cognitive impairments in the current HAART era (Sacktor et al., 2016) The most common cognitive disturbances observed among HIV positive individuals in the HAART era are in learning/memory and executive functions (Cysique et al., 20 04; Robert K Heaton et al., 2011) Impairments in these domains overlap with the impairments seen among marijuana users. Thus it is plausible that HIV positive individuals who use marijuana to experience increased burden of these cognitive impairment s Da ta on the effect of marijuana use among HIV positive individuals is scarce. Two studies have found significant adverse marijuana by HIV interactive effect. Cristiani et al. (2004) found symptomatic HIV positive individuals who used marijuana frequently per form worse than controls on measures of delayed memory (Cristiani et al., 2004) In another study, HIV positive individuals with moderate to heavy marijuana use performed significantly worse on learning and memory tasks than other comparison groups (Thames et al., 2015) However, two other studies found no significant marijuana by HIV interaction on tasks of procedural learning or complex motor skills (Gonzalez et al.,

PAGE 91

91 2011) or in other cognitive domains (Chang et al., 2006) although there was evidence of additive negative ef fects on complex motor skills in the study by Gonzalez et al. (2011) (Gonzalez et al., 2011) The d iscrepant findings from the aforementioned studies may be due to a host of factors including their cross sectional designs, small sample sizes, differences in the sociodemographic and clinical profiles of the samples studied, differences in the neuropsycho logical tests used and lack of control of other factors associated with marijuana and cognitive function. In addition, among the studies finding negative effects of marijuana use on cognitive function, it is unclear whether the cognitive impairments are a consequence of marijuana use or its cause. With cross sectional designs it is difficult to discern their temporality. One of the frequently debated questions regarding marijuana use is whether long term heavy use is associated with persistent cognitive imp airments. This question is of particularly importance for HIV positive individuals as marijuana use may be chronic and heavy when using to manage symptoms of a lifelong illness. The MACS is an ongoing prospective cohort study of MSM living with or at ris k for HIV infection. The MACS has continuously collected data on marijuana use since its inception assessed neuropsychological measurements for nearly 27 years, and provides an ideal opportunity to assess the long term impact of marijuana use on cognition The aim of this study was to determine effects of cumulative years of exposure to marijuana use and longitudinal changes in cognitive performance in HIV positive and HIV negative MSM. It was hypothesized that greater cumulative years of exposure to marij uana use will be associated with worse cognitive performance.

PAGE 92

92 Methods The Multicenter AIDS Cohort Study (MACS) The MACS is an ongoing prospective cohort study of the natural and treated history of HIV infection among MSM in the United States. 6,972 men w ere enrolled during the history of the project in three waves: 4,954 men in 1984 1985, 668 in 1987 1991, and 1350 in 2001 2003 and at 4 centers located in Baltimore/Washington DC, Chicago, Los Angeles, and Pittsburgh. The study design of the MACS has been described previously (Roger Detels et al., 1992; Dudley et al., 1995; Kaslow et al., 1987) and only the design relevant to the present analysis are described here. The study questionnaires used in the MACS are available at www.statepi.jhsph.edu/macs/forms.html The institutional review boards at the respective recruitment centers and their community affiliates approved the MACS study protocols and all participants provided informed consent Subjects MACS participants return every 6 months for physical examinations, HIV testing laboratory testing, structured clinical interviews, collection of alcohol, smoking and other substance use and neuropsychological testing. Figure 5 1 displays the flow of participant s included in the current study. Of the larger MACS cohort, 5,712 of the men completed at least one visit wi th cognitive function measurements. O nly those HIV positive participants who had initiated HAART were included in this study Of the 2,463 HIV posit ive men eligible for the analysis, 1,121 who did not initiate HAART were excluded The date of HAART initiation is considered to be the midpoint between the last visit reporting no HAART use (last no HAART ) and the first visit at which HAART

PAGE 93

93 use was report ed (first HAART ). HIV negative participants were selected among those seen on or after MACS semiannual visit 25 (data collection starting April 1, 1996 to September 30, 1996). For both HIV positive and HIV men, only those with at least two o r more visit wi th cognitive function measures were included 407 of the men were further excluded because they reported: (1) history of learning disability (2) stroke (3) nerve disorders/brain neoplasms (4) seizures (5) multiple sclerosis and (6) head injury with loss of consciousness greater than 1 hour Those included in the study (N=1 982 ) and those excluded (n= 407 ) had similar ra tes of current marijuana use (31% versus 29 %; p =.8034) and median cumulative m arijuana use years (median= 0.01 vs. 0.01; p =.4080) at baseline (see Table 5 9 ). However, at baseline, the men who were included in the study were more likely to complete a co llege degree or more (80% vs. 68 %; p <.00 0 1), less likely to endorse having depressive symptoms on the CES D (28 % vs. 38%; p <.0001), more likely to report any alcohol use (8 6 % vs. 77%; p =.0003 ), had lower rates of smoking (7 1% vs. 7 6% -current and former smoking; p =.0 421 ), lower median cumul ative pack years of smoking (2.1 vs. 4.7; p =.00 96 ), less likely to be HCV positive (8 % vs. 12%; p = <.0001 ) and had higher median CD4 nadir cell count (281 vs. 242; p=.0024 ; Table 5 9 ). Primary Predictor: Marijuana Exposure The MACS questionnaires assessed m arijuana use at each MACS semiannual Have you used any pot, marijuana or hash since your last visit No use rs. How often did you use pot, marijuana or hash since your las t visit P

PAGE 94

94 frequency of marijuana use at each study visit was used to compute cumulative exposure to marijuana use while in the MACS study in marijuana use years. Using the frequen cy of marijuana use response, I estimated the average number of days using marijuana since their last visit (typically approximately six months) by assigning weights to each frequency measure. For instance, if an individual used d aily during the last six months, I calculated his average number of days using marijuana as 30.5 days multiplied by 6 months. Thus, it was estimated that this individual used marijuana 182 day s in the past six months. Specifically, the weights assigned to each frequency of use category are as follows: daily use, weight=30.5; weekly use=4.36; monthly use, weight=1; less often, weight=0.33, and none use, weight=0. C umulative exposure to marijuana use was estimated by adding the total number of days using mari juana during follow up until their last study visit and divid ed by 365 (days) Thus, one marijuana use year is equivalent to using marijuana every day for 1 year. I computed the marijuana use year more fully, by carrying forward or backward data on mariju ana use depending on the nearest visit with data. Outcome: Neuropsychological E valuation The MACS began assessment of neuropsychological test performance in 1988 with the goal of characterizing the pattern and severity of neurop sychological deficits among HIV positive individuals. The MACS protocol administered a comprehensive neuropsychological exam to both HIV negative and HIV positive participants. The Trail Making Tests (TMT) (Reitan, 1 992) P art A (TMT A) measures psyc homotor speed, whereas the TMT P art B (TMT B) evaluates both psychomotor speed and executive function. In the TMT A, 25 circles numbered 1 25 are distributed randomly over a sheet of paper. The respondent is required to connect the numbers in

PAGE 95

95 ascending order (i.e. 1 2 3, etc.) by making pencil lines without lifting the pencil from the piece of paper as quickly as possible. In the TMT B, the circles include numbers and letters randomly distributed over the sheet of pap er i ncludes circled letters. Test administrators instruct the participant to draw lines to connect the circled numbers and letters in an alternating numeric and alphabetic sequence (i.e. 1 A, 2 B, 3 C). Scoring is determined as the time in seconds required for completing the connections of the circles. Errors affect the respondents score in that the correction of errors is included in the completi on time of the task. In the TMT A and TMT B higher scores indicate worse performance. The Symbol Digits Modalities Tests (SDMT) (Smith, 1982) requires elements of attention, visuoperceptual processing, working memory and psychomotor speed. In the SDMT, test administrator presents participants a coding ke y on a sheet of paper consisting of nine abstract symbols, each paired with a number. Following a series of black boxes, located below the abstract symbols the participants is required to scan the key and write down the number corresponding to each of the abstract symbol as quickly as possible. The nine symbols are in a random number sequence. Scoring is determined as the number of correct paired abstract symbols and numbers over a period of 90 seconds. In the SDMT, higher s cores indicate better performance The Rey Auditory Verbal Learning Test (RAVLT) evaluates verbal learning and memory. The test administrator reads aloud to the participants 15 nouns (List A) and the participant is required to repeat back as many words as they can remember from the list. free recall. The sixth trial requires the participant to recall (Immediate Recall) words

PAGE 96

96 from List A without further presentation of the words. After a 20 minute delay period, the par ticipant is required to recall words from list A (Delayed Recall). The score for each trial is the number of words recalled. The score from the delayed recall was used in the primary analysis. Covariates Sociodemographic Characteristics. age at each visit was calculated from their self reported date of birth. Participants self reported their race/ethnicity status at their baseline visit and were categorized as non Hispanic, white, n on Hispanic Black and other The baseline visit (or index visit) was used to define a categorical variable for educational attainment (High school diploma or less, some college or college degree, Graduate work or more) and employment (employed, unemployed). Participants were classified according to the MACS study center and whether they were enrolled prior to or after 2001 Depressive symptoms. The Center for Epidemiologic Studies Depression (CES D) scale, was used to measure clinically significant symptoms of depression (Radloff, 1977) Th e CESD includes components of depressed mood, feelin gs of worthlessness, sense of hopelessness, sleep disturbance, loss of appetite, and concentration difficulties. Scores on the CES D of 16 or more suggests a clinically significant level of psychological distress (Radloff, 1977) Alcohol Exposure. The t ypical number of alcoholic drin ks per week consumed by each participant was computed reported average number of drinking days per week by the average number of drinks per drinking day (range: 0 84 drinks/week). A drink was defined as one 12 ounce beer ( ~355 mL), one 4 5 ounce glass of wine (~120 150 mL), or one mixed drink with 1.5 ounces (~44 mL) of

PAGE 97

97 80 proof hard liquor. The few (<1%) reporting 10 or more drinks per drinking day were classified as 12 drinks. Using this data on average number of alcoholi c drinks per week since last study visit, I computed cumulative alcohol consumption by adding up the total number of drinks reported during follow up. Cigarette Exposure. The MACS protocols asked all participants the following The MACS protocols queried for quantity of cigarettes smoked among current smokers with the cumulative exposure to cigarettes was computed and defined in pack years, with 1 pack year of exposure equivalent to 7300 cigarettes (1 year x 365days/year x 1 pack/day x 20 cigarettes/pack) (Akhtar Khaleel et al., 2015) Illicit Drug Exposure. The MACS questionnaires asked participants the frequency of use of poppers (inhaled nitrites) and cocaine (crack or any form of cocaine) using similar response options as in marijuana use. Similar to m ar ijuana use year, I calculated cumulative exposure to poppers and cocaine in use years. Clinical Factors. The MACS protocols assessed HIV s erostatus using enzyme linked immunosorbent assay with confirmatory Western blot tests on all MACS participants at e ts who were initially HIV negative Standardized flow cytometry was used to quantify CD4+ T lymphocyte subset levels by each MACS site (Giorgi et al., 1990) and categorized as 3 201 500/mm 3 and > 500/mm3. Levels of pla sma HIV RNA were measured using either the standard reverse transcription polymerase chain reaction assay (Roche

PAGE 98

98 Nutley, NJ) or with the Roche ultrasensitive assay (Roche Diagnostics. Standardized viral load measures (across different assays) was used to c reate a dichotomous variable to denote detectable (> 200 copies per mL) versus undetectable. HCV infection status was dichotomized as HCV negative if HCV antibody testing was negative. Participants were classified at each MACS semiannual visit as HCV posit ive if they were found to be in the process of seroconversion, acute infection, chronic infection, clearing (between RNA+ and RNA ), or previously HCV positive, but now clear of HCV RNA. A t each visit, the HIV positive men were asked about their ART medic ations and adherence. The MACS protocols classified ART medications as NRTI, PIs, and non NRTI. I calculated c umulative use of each class of ART based on the number of ART medications reported in each classification and weighted for self reported adherence The weights were 1, 0.975, 0.85, 0.375, 0 for adherence levels of 100%, 95 99%, 75 94%. The approach multiplied the number of ART medications at each study visit by the adherence levels and the weighted values then cumulated. Questions to ascertain adher ence to ART was incorporated into the MACS semiannual visit 30 (data collection staring in October 1, 1998 March 31, 1998) and thus adherence prior to October 1998 was assumed to be 100%. Data Analysis Data analysis was performed from October 1, 1996, to September 30, 2013. C haracteristics of the sample at baseline stratified by HIV s erostatus and marijuana use was described using chi square tests for categorical variable and t tests for continuous variables. The exposure variable (marijuana use years) wa s represented as a cumulative continuous exposure measurement over time. To test potential nonlinear associations, the cumulative marijuana exposure variables was modeled as flexible

PAGE 99

99 restric ted cubic splines in adjusted models in order to al low the associations with cognitive measures to take different shapes at lower vs. higher levels of exposure. S eparate mixed effects linear regression models was used to estimate associations between cumulative exposure to marijuana (in use years) and chan ges in the 4 measures of cognitive function over time. Figure 5 2 displays a schema for t he analysis of cumulative marijuana exposure and cognitive function measure over time. The models included random effects for intercepts and slopes to account for within person correlations of cognitive measures over time and to allow for participant specific rates of cognitive change. C ognitive measures were included until death, loss to follow up or end of the follow up Inverse probabilities of attrition weights (IPAW) were calculated to address the propensity for selective attrition (see Appendix for description of the approach for IP AW) Models were fit using maximum likelihood estimation and all models assumed an unstructured variance covaria nce structure. R esidual plots from the mixed models were examined to check for normality of the distributed errors with constant variance. Thes e mixed effects linear regression models estimated slopes or the incremental difference in cognitive performance observed with 1 unit of additional marijuana use year of exposure while adjusting for participant factors. Because the outcome scores were expr essed as Z scores, the model coefficients indicate incremental standardized units associated with additional units of marijuana use year. For example a model coefficient of 0.20 can be interpreted as each additional 1 unit of marijuana exposure is associa ted with 0 .20 lower standardized units in cognitive measures on over time.

PAGE 100

100 The number of years since baseline was used as the time metric. A nalyses were conducted separately by HIV s erostatus and in order to minimize complexities of the models; HIV by cum ulative marijuana exposure interactions was not estimated Model A included a variable adjusting for prac tice effects (number of times participant s w ere exposed to each neuropsychological test ) and baseline cognitive scores age (flexibly modeled via 3 kno t cubic splines), race, education, MACS study center, enrollment cohort and IPAW. Model B included the variables from model A and added potential confounders in the association between marijuana use and cognitive performance including: cumulative alcohol, smoking, stimulant s and poppers exposure, history of IDU, depressive symptoms, HCV infect ion, hypertension and current ma rijuana use (at each visit). M odels for HIV positive participants were additionally adjusted for cumulative exposure to HAART and CD4+ nadir. A ll statistical analyses were performed in SAS version 9.4 (SAS Institute Inc., Cary, North Carolina, USA). Statistical tests for significance were two tailed and P <.05 was considered statistically significant. Results Sample C haracteristics Partic ipants were 1,982 men with 817 ( 41 %) HIV positive an d 1165 ( 59 %) HIV negative (Table 5 1). Mean age of the sample at baseline was 43 years [(standard deviation (SD) = 9.9 ], and majority were non Hispanic whites (61%), and complet ed some college or more (80 %). At baseline, 7 6 % reported any alcohol use; 7 1 % reported ever smoking ( 31 % former smokers and 4 0 % current smokers) 1 6 % reported stimulant and 2 4 % reported popper use. At baseline, prevalence of marijuana use was 5 5%. Compared to the HIV negative part ic ipants in the sample, the HIV positive men were more likely to be racial minorities (non Hispanic blacks and others), more likely to be

PAGE 101

101 enrolled post 2001, reported more depressive symptoms (CES D>16), were more likely to report ever smoking (former and cu rrent smoking), more likely to report stimulant use and to be HCV positive. In contrast, the HIV positive men were significantly less likely to report alcohol use and high blood pressure (Table 5 1). Also, compared to non users, marijuana users in the samp le were significantly younger, less educated, more likely to be racial minorities, reported more depressive symptoms, more likely to report ever smoking cigarettes, alcohol and other illicit drug use ( popper s and stimulant s) Cumulative Marijuana Exposure and Changes in Cognitive Performance In both initial (Model A; Table 5 2) and fully adjusted models cumulative exposure to marijuana (in use years) was not associated with performance on all 3 measures of cognitive performance among HIV positive participan ts; with point estimates of the coefficients indicating better performance with additional exposure to marijuana. For instance, for the Trail Making Tests Parts A and B, each additional 5 years of exposure to marijuana was associated with .16 and .07 highe r standardized units respectively. In contrast each additional 5 years of exposure to marijuana use was associated with a .1 0 lower standardized units on the Symbol Digit test. Among the HIV positive participants, initial models (Model A; Table 5 2) demon strated that each additional 5 years of exposure to marijuana use was associated with .224 and .268 higher standardized units on the Trail Making Tests Part A and Symbol Digits Test. However, in fully adjusted models, the association remained significant with each additional 5 years of exposure to marijuana use was associated with .245 higher standardized units (95% CI, .013 to .477; P =.038 ) in the Symbol Digits Test.

PAGE 102

102 In both the HIV positive and HIV negative men, there was no association between cumulat ive exposure to marijuana and RAVLT performance; although the coefficients demonstrated worse performance with each additional 5 year exposure to marijuana use. Figures 5 3 to 5 7 displays the predicted z scores for each cognitive function measure by cumulative exposure to marijuana in use year by HIV status. Discussion In this 17 year study of cumulative exposure to m arijuana and changes in cognitive function measures, found no significant associations of cumulative exposure to marijuana with processing speed executive functions and verbal memory among the HIV positive men. Among the HIV negative men, the study found significant associations with incr easing exposure to marijuana and improved performance on one measure of processing spe ed, but no significant associations with the other cognitive measures of processing speed as well as executive functions and verbal memory. Reports of marijuana exposure and cognitive function among HIV positive adults have been few and have produced seemingly conflicting results (Chang et al., 2006; Cristiani et al., 2004; Gonzalez et al., 2011; Thames et a l., 2015) Our results are in contrast studies reporting significant association between worse verbal learning and memory and weekly marijuana use for over a period of a year (Cristiani et al., 2004; Thames et al., 2015) Also, another study reported worse executive functions with weekly marijuana use (Thames et al., 2013) whereas, others find no associations on other cognitive ability domains (Chang et al., 2006; Cristiani et al., 2004; Gonzalez et al., 2011) Many of these studies have been limited by their small sample sizes, cross sectional de signs a nd most did not did not evaluate associations of heavy long term marijuana exposure and cognitive functions per se.

PAGE 103

103 Our findings are also in contrast with other research among HIV unin fected marijuana exposure has been associated with worse verbal learning and memory, executive functions (Auer et al., 2016; Fried et al., 2005; Meier et al., 2012; Solowij et al., 2002) For instance, in o ne recent cohort study of 3 385 men and women, s ignificant dose response associations with worse verbal memory. Specifically, the investigators found that for every 5 years of marijuana use verbal memory was 0.13 standardized units lower than those with no exposure. T h e investigators also found no signi ficant associations of cumulative marijuana exposure with executive function or processing speed (Auer et al., 2016) However, t hat study was limited, however, by the availability of cognitive function measurement at only one time point in contrast to the present study that repeated cognitive function measurements for over 17 year s. In another study among 1037 participants in New Zealand followed up for 38 years found that persistent regular cannabis use 4 days per week or more was associated with IQ and neuropsychological decline (including declines in executive functions and processing speed) as compared to those reporting nonregular use who demonstrated no neuropsychological decline (Meier et al., 2012) Comparison between this study and the present study is difficult because marijuana exposure was defined by cannabis dependence following the Diagnostic and Statistical Manual of Mental Diso rders (DSM) Marijuana exposure might affect cognition via the actions of THC on CB1 receptor sites predominantly located in specific brain regions including hippocampus, cerebellum, basal ganglia, substantia nigra and globus pallidus (Herkenham et al., 1990; Westlake, Howlett, Bonner, Matsuda, & Herkenham, 1994) These brain regions

PAGE 104

104 ar e thought to be involved in cognitive functions, including learning/memory, motor functions as well as attention and working memory. Thus, activation of CB1 receptors in these regions may produce cognitive effects associated with marijuana use. Alternative ly, long term exposure to marijuana use may be associated with morphological and functional changes in these brain regions. Some studies have demonstrated smaller grey matter volume in the parahippocampal gyrus (Matochik, Eldreth, Cadet, & Bolla, 2005) reduced hippocampal and amygdala volumes in long term heavy marijuana users (Ycel et al., 2008) Others have found reduced regional blood volumes in cerebellum among long term marijuana use compared to controls (Sneider et al., 2006) This study found consistently worse cognitive function in the verbal memory task in both the H IV positive and HIV negative men, although these associations were not significant. It is possible that the fewer cognitive functions measures for the RAVLT might in part explain the lack of significant associations. T he association between increasing expo sure to marijuana and better performance on the Symbol Digit Test among the HIV negative men was unexpected, although there is a substantial body of literature documenting neuroprotective effects of (Eubanks et al., 2006; Ramrez, Blzquez, Gmez del Pulgar, Guzmn, & de Ceballos, 2005) The mechanism b y which marijuana may be neuroprotective is unclear, but might be explained by the immunomodulatory and anti inflammatory properties of cannabinoid receptor 2 agonists particularly cannabidiol and cannabinol (Rom & Persidsky, 2013; Snchez & Garca Merino, 2012) Whether CB2 expression and these cannabinoids have potential neuroprotective benefits merits further invest igation.

PAGE 105

105 The present study has expanded our knowledge of the associations between cumulative marijuana exposure and cognitive functions in large cohort of HIV positive and HIV negative men for 17 years of follow up with repeated neuropsychological testing has not been examined previously. The relatively wide confidence intervals around the coefficient estimates mig ht indicate the less precise nature of the marijuana exposure variable. Future investigations might collect more precise measures of marijuana th at potentially capture the cannabinoid constituents in the marijuana. Our study has some limitations which are noteworthy when interpre ting our findings. T he cumulative marijuana exposure variable was computed prospectively via self report and there was no laboratory test conducted to validate the self reported marijuana use. Second is that cumulative marijuana exposure variable was cumulated beginning from the time participants enrolled in the MACS and thus this study was not able to account for past expos ures prior to enrollment in the MACS. This is nontrivial as the majority of MACS participants enrolled in the study when they were in their midlife when use of marijuana may be on the decline. This study was also not able to capture age of fi rst exposure to marijuana. Studies have previously demonstrated that e arly initiation of marijuana use or early adolescents use may confer profound effects on cognition via its effect on the developing adolescent brain (Jager & Ramsey, 2008; Lisdahl, Gilbart, Wright, & Shollenbarger, 2013; Lisdahl, Wright, Medina Kirchner, Maple, & Shollenbarger, 2014) T he study did not distinguish whether marijuana use was recreational or medicinal. This is critical given that the CBD in marijuana has been associated with therapeutic effects in cluding neuroprotection. Thus

PAGE 106

106 variability in the THC/CBD ratio of the marijuana consumed wi thin individuals may confer differential effects on cognition. Furthermore, It is possible that repeated cognitive function measurements and test familiarity may have underestimated the true rate of cognitive decline and thus may explain the small effect s in the current study (Atkinson et al., 2010) although the study adjusted for this in all models by including the number of times a participant was exposed to the particular test. In addition, t he average level of education in this study is may be higher than that of the HIV infected population in general. Higher levels of education has been associated with increased cognitive reserve (Satz, 1993) Majority of the participants included in the current study were non Hispanic white MSM. Taken this together, our study may be less generalizable to other HIV populations (e.g. women). Finally, the study did not assess the effects of cumu lative marijuana exposure with other cognitive domains due to the fewer neuropsychological assessments and data available for these domains It is possible that significant effects of marijuana exposure on these other cognitive domains especially in the learning and memory domains will have emerged if more data were available. Conclusions In the sample of HIV positive and HIV negative men; there were no statistically significant associations between cumulative exposure to marijuana for over 17 years and adverse changes in cognitive function measurements in processing speeds, executive functions and verbal memory. The present analysis is among the first to analyze data from repeated cognitive assessments over an extended period of time to examine changes i n performance with cumulative exposure to marijuana. The findings are not definitive evidence that long term heavy marijuana exposure is not associated

PAGE 107

107 with worse cognitive. Additional investigations into whether different strains of marijuana and or THC/ CBD ratio may be associated with differential cognitive effects need clarification.

PAGE 108

108 5,712 men with at least one cognitive function measurement 2,723 HIV negative Men 2,463 HIV positive men 526 Excluded Seroconverts 1646 Excluded 1,121 Did not initiate HAART 285 Had < 2 cognitive measurements 240 History of Learning disorder Stroke Seizures Nerve disorder Multiple Sclerosis Head injury w/LOC GT 1hr 817 HIV p ositive men included i n the analysis 1 558 Excluded 1, 235 Not seen at MACS visit 25 156 Had < 2 cognitive measurements 167 History of Learning disorder Stroke Seizures Nerve disorder Multiple Sclerosis He ad injury w/LOC GT 1hr 1165 HIV n egative men included in the analysis Note MACS indicates Mul ticenter AIDS Cohort Study, Seroconverts indicates HIV negative men who become HIV positive during follow up, HAART indicates Highly Active Antiretroviral Therapy, LOC indicates Loss of Consciousness F igure 5 1 Flow chart of MACS participants included in the study

PAGE 109

109 Figure 5 2 Scheme for Exposure and Outcome Definition 1984 2013 Primary Predictor: 29 Year Cumulative Marijuana Use 1996 Study period for the 17 year outcome: C hanges in C ognitive Performance

PAGE 110

110 Table 5 1 Characteristics o f MACS Men Included in the Study at Baseline by HIV Status and Marijuana Use Characteristics All men (N=1 982 ) HIV negative (n=1165) HIV positive (n= 817 ) Marijuana use Marijuana use Yes (n=397) No (n=768) Yes (n= 477 ) No (n= 339 ) Age, yr. mean (SD) 43 ( 9.9 ) 42 (10.9) 46 (10.2) 40 (8.3 ) 4 2 (8.4 ) Race, n (%) White, non Hispanic 1 201 (61) 252 (63) 553 (72) 234 (4 9 ) 1 61 (48 ) Black, non Hispanic 513 (26) 100 (25) 146 (19) 1 65 (3 5 ) 102 (3 0 ) Other 2 68 (14) 45 (11) 69 ( 9 ) 78 ( 16 ) 76 ( 22 ) Education, n (%) High school diploma or less 3 99 (20) 87 (22) 90 (12) 12 2 ( 26 ) 100 ( 30 ) Some college or college degree 9 88 (50) 202 (51) 374 (49) 264 ( 55 ) 1 47 ( 43 ) Graduate work or more 5 93 (30) 107 (27) 303 (40) 91 ( 19 ) 82 ( 27 ) Study center, n (%) Baltimore/Washington DC 4 9 7 (25) 91 (23) 225 (29) 9 8 (2 1 ) 83 ( 25 ) Chicago 3 84 (1 9) 71 (18) 110 (14) 120 (2 5 ) 83 (2 5 ) Pittsburgh 508 (26) 106 (27) 215 (28) 10 5 (2 2 ) 82 (2 3 ) Los Angeles 5 9 3 (30) 129 (32) 218 (28) 1 54 (3 2 ) 19 ( 27 ) Study enrollment, n (%) Pre 2001 1 065 (5 4 ) 231 (58) 522 (68) 204 ( 43 ) 10 7 ( 32 ) Post 2001 917 (4 6 ) 166 (42) 246 (32) 2 73 ( 57 ) 232 (6 8 ) Depressive symptoms n (%) 13 96 (7 2 ) 293 (75) 579 (77) 300 ( 64 ) 239 ( 74 ) CESD >16 533 (2 8 ) 99 (25) 169 (23) 1 68 ( 36 ) 8 6 ( 26 ) Alcohol use, n (%) None 2 84 (1 4 ) 19 ( 5 ) 138 (18) 37 ( 8 ) 93 ( 27 ) 1 to 3 drinks/wk. 1044 (5 3 ) 200 (50) 427 (56) 238 ( 50 ) 16 6 ( 49 ) 4 to 13 drinks/wk. 4 2 9 (2 2 ) 111 (28) 157 (20) 1 22 ( 2 6) 47 ( 14 ) 13 or more drinks/wk. 2 2 4 (11) 67 (17) 46 ( 6 ) 80 (17) 33 (10) Cumulative drink years, median (IQR) 2.4 (0. 36 10.5 ) 3.5 (0.41, 14.0 ) 2.4 (0.21, 10.9) 2.74 (0. 46 10.0 ) 0. 3 (0.0 1 2. 78 ) Smoking, n (%) Never 5 7 5 ( 29 ) 75 (19) 285 (37) 99 ( 21 ) 130 ( 38 ) Former 614 ( 31 ) 104 (26) 260 (34) 139 (2 9) 6 7 (2 0 ) Current 7 92 (4 0 ) 218 (55) 223 (29) 239 ( 50 ) 14 2 ( 42 ) Cumulative pack years of smok ing, median, (IQR) 2. 1 (0, 18. 4 ) 6.2 (0, 22.5) 0 (0, 12.7) 5.4 (0 2 0.1 ) 0.52 (0, 14.9) Stimulants use, n (%) No 1 664 (8 4 ) 304 (77) 720 (94) 297 ( 62 ) 293 ( 86 ) Yes 317 (1 6 ) 93 (23) 48 ( 6 ) 180 (38) 46 (14) Cumulative cocaine use years, median, (IQR) 0 (0, 0) 0 (0, 0) 0 (0, 0) 0 (0, 0.03 ) 0 (0, 0) Poppers, n (%) No 1 503 (7 6 ) 273 (69) 641 (83) 297 ( 62 ) 293 ( 86 ) Yes 4 78 (2 4 ) 124 (31) 127 (17) 180 ( 38 ) 46 ( 14 )

PAGE 111

111 Table 5 1. Continued Characteristics All men (N=1 982 ) HIV negative (n=1165) HIV positive (n=670) Marijuana use Marijuana use Yes (n=397) No (n=768) Yes (n= 477 ) No (n= 339 ) Cumulative popper use year s, median, (IQR) 0 (0, 0.0 8 ) 0 (0, 0.20) 0 (0, 0.07) 0 .01 (0, 0.15 ) 0 (0, 0) IDU, n (%) No 1 737 (8 8 ) 349 (88) 720 (94) 3 78 ( 79 ) 289 (8 5 ) Yes 2 45 (1 2 ) 48 (12) 48 ( 6 ) 99 ( 21 ) 50 (1 5 ) Hepatitis C virus antibody, n (%) Negative 1 346 (9 2 ) 230 (96) 517 (97) 330 (8 6 ) 1 57 ( 88 ) Positive 113 ( 8 ) 9 ( 4 ) 17 ( 3 ) 52 (1 4 ) 22 ( 12 ) Diabetes, n (%) No 919 ( 89 ) 163 (92) 323 (89) 21 5 ( 85 ) 169 ( 91 ) Yes 113 (1 1 ) 14 ( 8 ) 38 (11) 37 (1 5 ) 1 7 ( 9 ) Hypertension, n (%) No 1 337 (70) 271 (71) 509 (67) 344 (7 4 ) 225 ( 69 ) Yes 5 79 (30) 111 (29) 247 (33) 1 20 (2 6 ) 99 ( 31 ) Undetectable 289 ( 51 ) --193 (51) 85 (49) Detectable 272 ( 49 ) --182 (49) 90 (51) CD4 + 2 81 (1 57 4 12 ) --2 83 (1 63 4 06 ) 2 79 (1 53 4 16 ) Current CD4 + count (cells per cubic milliliter) < 200 34 4 ( 42 ) --144 (30) 200 (59) 238 (2 9 ) --162 ( 34 ) 76 ( 22 ) > 500 235 (2 9 ) --171 ( 36 ) 6 3 ( 19 ) HAART PI HAART 432 (5 5 ) --2 57 (5 5 ) 1 75 (5 4 ) NNRTI HAART (without PI) 307 ( 39 ) --1 82 ( 39 ) 1 24 ( 38 ) Other HAART 53 ( 7 ) --25 ( 5 ) 28 ( 8 ) Current marijuana use None 888 ( 45 ) ----Less often 608 (31 ) 194 (49) --101 (40 ) Monthly 93 ( 5 ) 45 (11) --36 (14) Weekly 1 06 ( 6 ) 95 (24) --52 (21) Daily 286 ( 7 ) 63 (16) --63 (25) Cumulative marijuana use years, median, (IQR) 0.0 1 (0, 0. 35 ) 0.17 (0.02, 1.32) 0 (0, 0.02) 0 .24 (0 .01 1.02 ) 0 (0, 0 ) Note IQR=Interquartile range; SD=Standard deviation; CESD= Center for Epidemiological Depression Scale; wk. = week; IDU=Intravenous Drug Use; RNA copies/ml; CD4 + nadir=Lowest CD4 count while on MACS study; HAART = Highly active antiretroviral therapy; PI=Protease inhibitors; NNRTI=Non nucleoside rev erse transcriptase inhibitors;

PAGE 112

112 Table 5 2. Association between cumulative exposure to marijuana use a and changes in cognitive performance among HIV Positive and HIV Negative Participants: MACS 1996 to 2013 HIV positive ( N=817) HIV negative (N=1165) Variable Model 1 b Model 2 c Model 1 b Model 2 c Trail Making Test Part A Coefficient (95% CI) P Val ue Coefficient (95% CI) P Value Coefficient (95% CI) P Value Coefficient (95% CI) P Value Time, per 1 y r. .008 ( .000 to .017 ) .059 .004 ( .007 to .017) .440 .025 (.020 to .030) <.0001 .028 (.022 to .033) <.0001 Age, per 5 yr. .031 ( .004 to .067 ) .083 .008 ( .048 to .031) .683 .037 ( .065 to .009) .007 .034 ( .063 to .005) .018 Baseline cognitive score per 1 point .719 (.683 to .755) <.0001 .657 (.617 to .696) <.0001 .706 (.675 to .737) <.0001 .702 (.670 to .733) <.0001 For every 5 marijuana use yr. f .225 ( 031 to .483) .085 164 ( .140 to .470) .290 .224 (.009 to .438) .040 .223 ( .012 to .460) .063 Test of nonlinearity .046 .041 Intercept .612 ( .916 to .308) <.0001 .393 ( .785 to .000) .050 .024 ( .229 to .27 7) .852 .070 ( .332 to .191) .598 Log likelihood 19756.9 17110.3 30044.2 28344.3 Trail Making Test Part B Time, per 1 yr. .009 (.000 to .017) .029 .004 ( .006 to .015) .469 .019 (.015 to .024) <.0001 .020 (.015 to .025) <.0001 Age, per 5 yr. .014 ( .016 to .046) .357 .005 ( .041 to .030) .768 .044 ( .069 to .019) .000 .041 ( .067 to .016) .001 Baseline cognitive score per 1 point .742 (.707 to .777) <.0001 .682 (.643 to .720) <.0001 .717 (.687 to .748) <.0001 .716 (.686 to .746) <.0001 For every 5 marijuana use yr. f .156 ( .074 to .387) .184 .074 ( .203 to .351) .600 .143 ( .047 to .333) .140 .160 ( .049 to .369) .160 Test of nonlinearity .058 .138 Intercept .571 ( .844 to .299) <.0001 .660 ( 1.01 to .30 2) .000 .019 ( .246 to .208) .867 .100 ( .335 to .133) .400 Log likelihood 16850.4 14919.4 25329.5 23849.4 Symbol Digit Modality Test Time, per 1 yr. .005 ( .003 to .013) .211 .007 ( .004 to .018) .222 .013 (.008 to .019) <.0001 016 (.010 to .022) <.0001 Age, per 5 yr. .004 ( .030 to .040) .796 .006 ( .044 to .032) .758 .038 ( .068 to .009) .009 .040 ( .070 to 010) .009 Baseline cognitive score per 1 point .636 (.597 to .675) <.0001 .621 (.579 to .664) <.0001 .646 (.6 12 to .680) <.0001 .643 (.609 to .677) <.0001 For every 5 marijuana use yr. f .049 ( .286 to .188) .6848 .109 ( .390 to .172) .448 .268 (.056 to .480) .013 .245 (.013 to .477) .038 Test of nonlinearity .089 .013 Intercept .427 ( .731 to .123 ) .006 .421 ( .809 to .032) .033 .091 ( .180 to .364) .508 Log likelihood 15099.6 13208.6 22835.1 21718.4 Rey Auditory Verbal Learning Test Time, per 1 yr. .005 ( .004 to .016) .272 .003 ( .011 to .019) .629 .014 (.007 to .020) <.0001 .012 (.004 to .019) .001 Age, per 5 yr. .052 ( 107 to .001) .056 .062 ( .125 to .001) .056 .049 ( .089 to .010) .014 .031 ( .072 to .008) .774 Baseline cognitive score per 1 point .654 (.609 to .699) <.0001 .608 (.559 to .657) <.0001 .6 21 (.587 to .655) <.0001 .621 (.586 to .656) <.0001 For every 5 marijuana use yr. f .269 ( .668 to .129) .185 .255 ( .752 to .242) .315 .136 ( .436 to .162) .371 .049 ( .389 to .290) .774 Test of nonlinearity .089 .371 Intercept .067 ( .394 to .529) .775 .087 ( .516 to .691) .777 .017 ( .367 to .333) .9230 Log likelihood 7013.9 6006.6 9278.8 8771.9 Note Time (in years) since baseline was the metric for time and the cognitive z scores was modeled via linear mixed effects models. Models included both fixed and random effects of time a Cumulative exposure to marijuana expressed as marijuana use year, with 1 marijuana use year equivalent to using marijuana every day for 1 year. b The coefficients are for slopes and reflect the increm ental rate of cognitive performance associated a 5 unit additional exposure to marijuana in use year with negative standardized scores indicating worse cognitive performance c Model was adjusted for practice effects (number of times of exposure to cognit ive test ) and baseline cognitive test performance and includes random intercept; d Mode l was adjusted fo r age (flexibly modeled using a 3 knot age spline) race, education, study center enrollment cohort and IPAW; e Models was adjusted for current marijua na use, cumulative exposures to other substances (e.g. alcohol, smoking, cocaine and poppers), history of IDU, depressive symptoms and high blood pressure. Models for the HIV positive men were additionally adjusted for history of AIDS, cumulative years of highly antiretroviral therapy ( HAART ) use and CD4 nadir for the HIV positive participants; f Cumulative marijuana use analyzed as a continuous variable and flexibly modeled via restricted cubic splines

PAGE 113

113 Figure 5 3 Predicted Z scores on the Trail M aking Tes t Part A by HIV Serostatus

PAGE 114

114 Figure 5 4 Predicted Z scores on the Trail Making Test Part A by HIV Serostatus

PAGE 115

115 Figure 5 5 Predicted Z scores on the Symbol Digit Test by HIV Serostatus

PAGE 116

116 Figure 5 6 Predicted Z scores on the Symbol Digit Test HIV positive

PAGE 117

117 Figure 5 7 Predicted Z scores on the Symbol Digit Test HIV negative

PAGE 118

118 Tab le 5 3 Baseline Characteristics of MACS Men Included and Excluded from Anal yses Included (n=1, 982 ) Excluded (n= 407 ) P V alue Age, mean (SD) 43 (9.9) 43 ( 10.0 ) .6883 Race, n (%) .0967 White, non Hispanic 1201 (61) 237 (5 8 ) Black, non Hispanic 513 (26) 1 21 (30) Other 268 (14) 49 (12 ) Educati on, n (%) <.0001 High school diploma or less 399 (20) 1 29 (32 ) Some college or college degree 988 (50) 175 (4 3 ) Graduate work or more 593 (30) 1 03 (2 5 ) Study center, n (%) .1186 Baltimore/Washington DC 497 (25) 1 01 (25) Chicago 384 (19) 84 (2 1 ) Pittsburgh 508 (26) 1 02 (2 5 ) Los Angeles 593 (30) 1 20 (29) Study enrollment, n (%) .8262 Pre 2001 1065 (54) 2 06 (5 1 ) Post 2001 917 (46) 2 01 (4 9 ) Depressive symptoms (CESD >16), n (%) <.0001 1396 (72) 239 (6 2 ) CESD >16 533 (28) 1 46 (3 8 ) Alcohol use, n (%) .0003 None 284 (14) 95 (2 3 ) 1 to 3 drinks/wk. 1044 (53) 109 (49) 4 to 13 drinks/wk. 429 (22) 83 (2 0 ) 13 or more drinks/wk. 224 (11) 30 ( 7 ) Cumulative drink years, median (IQR) 2.4 (0.36, 10.5) 1. 5 (0.1 4 7.99 ) .4651 Smoking, n (%) .0421 Never 575 (29) 98 (2 4 ) Former 614 (31) 1 09 (2 7 ) Current 792 (40) 2 00 (4 9 ) Cumulative pack years of smoking, median, (IQR) 2.1 (0, 18.4) 4. 7 (0, 22.5) .0096 Stimulants use, n (%) .0614 No 1664 (84) 339 (83 ) Yes 317 (16) 68 (1 7 ) Cumulative cocaine use years, median, (IQR) 0 (0, 0) 0 (0, 0) .0439 Poppers, n (%) .4517 No 1503 (76) 326 (80 ) Yes 478 (24) 81 (2 0 ) Cumulative popper use years, median, (IQR) 0 (0, 0.08) 0 (0, 0. 08 ) .2895 IDU, n (%) No 1737 (88) 345 (8 5 ) .0002 Yes 245 (12) 62 (1 5 ) Hepatitis C virus antibody positive, n (%) <.0001 Negative 1346 (92) 225 (8 8 ) Positive 113 ( 8 ) 32 (1 2 ) Diabetes, n (%) .7207 No 919 (89) 159 ( 90 ) Yes 113 (11) 17 (1 0 ) Hypertension, n (% ) .7860 No 1337 (70) 268 ( 68 ) Yes 579 (30) 1 26 (3 2 ) .0262 Undetectable 289 (51) 72 (4 5 ) Detectable 272 (49) 87 ( 55 ) CD4 + nadir (cells/per cubic milliliter), 281 (157, 412) 242 ( 108 3 60 ) <.0001 Current CD4 + .2261 < 200 344 (42) 113 ( 47 ) 238 (29) 71 ( 30 ) > 500 235 (29) 26 ( 23 ) HAART .0035 PI HAART 432 (55) 100 (68 ) NNRTI HAART (without PI) 307 (39) 41 (64) Other HAART 53 ( 7 ) 6 ( 3 ) Current marijuana use 7741 No 888 (45) 238 (62) Yes 608 (31) 149 ( 29 ) Cumulative marijuana use years, median, (IQR) 93 ( 5 ) 0.01 (0, 0.2 3 ) .0223 Trail Making Tests Part A 25 (11.1) 27 (1 1.7 ) .0003 Part B 58 (31.1) 6 7 (4 0.2 ) <.0001 Symbo l Digits Test 52 (12.4) 48 (12.1 ) <.0001 Note IQR=Interquartile range; SD=Standard deviation; CESD= Center for Epidemiological Depression Scale; pies/ml; CD4 + nadir=Lowest CD4 count while on MACS study; HAART = Highly active antiretroviral therapy; PI=Protease inhibitors; NNRTI=Non nucleoside reverse transcriptase inhibitors;

PAGE 119

119 CHAPTER 6 DISCUSSION AND CONCLUSIONS Summary The goal of this dis sertation project was to increase our understanding and expand the current literature on marijuana use among HIV positive individuals by conducted three studies. The three studies utilized from the Multicenter AIDS Cohort Study (MACS) an ongoing prospect ive cohort study of HIV positive and HIV negative MSM, which provided rich data covering almost three decades. The first study aimed to understand how the prevalence of marijuana use has evolved over time and determine d important predictors of use. T he sec ond study utilized a novel statistical approach to identify long term distinct individual patterns of marijuana use, assessed predictors of the distinct individual patterns of use and determined cofactors that may serve to alter the patterns of use over ti me. The final study det ermine d the effects of cumulative exposure to marijuana and changes in cognitive function measures in three cognitive domains including; processing speed, executive function and verbal memory The following sections will summarize th e primary findings from each of these studies, the public health and clinical importance of the findings, strengths and limitations of the studies. I close with considerations for future investigations. Study 1 Trend in the Prevalence and Predictors of Mar ijuana Use The objective of this first study was to assess long term trends in the prevalence of current and daily marijuana use and determine predictors of use for over 29 years (1984 to 2013). I hypothesized that prevalence of current marijuana use will decline over time, but I speculated that daily use would increase. I also expected that racial minorities (non Hispanic blacks), MACS centers in states with medical marijuana laws,

PAGE 120

120 lower educational status, a HIV positive status, other drug use and depres sive symptoms will be associated with marijuana use in the analysis that included all men. Further, it was anticipated that among the HIV positive individuals, clinical factors such as ART use, CD4+ counts and viral load detectability would be associated w ith marijuana use. Study 1 found a decline in current marijuana use and an increase in daily marijuana use (among users) during the study period. This finding is consistent with at least one study that assessed long term trends in marijuana use among HIV p ositive women .The prevalence of marijuana use was approximately 3 0% and 25% among the HIV positive and HIV negative individuals respectively. In contrast, daily use (among users) was 35% and 30% among the HIV positive and HIV negative individuals respectively at the most recent period (i.e. 2013). This study found ver y high rates of marijuana use in the early years of the MACS study. Specifically, the rates from 1984 through 1988 were 80% and 53% respectively for the HIV positive men and 58% and 40% for the HIV negative men. The reason for the high rates of marijuana u se in these early years is not entirely clear. As proposed by the minority stress model, men who have sex with men often experience social pressures/stressors including stigma, discrimination and internalized homophobia after they disclose their sexual min ority status (Meyer, 2003) These stressors may arise from their heterosexual peers and have been proposed as processes that medi ate the relationship between a sexual minority status and elevated rates of substance abuse (McCabe et al. 2009) General attitudes toward lesbian, gay, bisexual and transgender (LGBT) individuals may indicate the levels of acceptance at a population level. One recent study examining national trends in public opinion on LGBT rights in the United

PAGE 121

121 States found that during the mid 1980s to early 1990s Americans reported more negative attitudes toward LGBT individuals than during recent years (e.g. early 200s to 2012) (Flores, 2014) It is possible that the negative attitudes toward gay men during the mid 1980s were associated with higher stressors and subsequ ently higher rates of substance use than in recent times. Further, this study also found that the rates of marijuana use were consistently higher for the HIV positive men compared to the HIV negative men. It is possible that medical use of marijuana among the HIV positive men may partly explain the higher rates. In addition, a HIV positive status may be associated with greater psychosocial problems (e.g. depression and anxiety) which may be linked to elevated marijuana use (and other substances) as coping mechanisms (Green & Feinstein, 2012) These possibilities warrant additional investigation. As expected, alcohol, smoking, stimulant, and other illicit drug use were positively associated with m arijuana use. Non Hispanic black racial status was not consistently associated with marijuana use and depressive symptoms were not associated with marijuana use. Passage of medical marijuana laws did not appear to be consistently associated with marijuana use in all models. Further among HIV positive individuals, there were some tenuous associations between viral load and CD4 counts; such that a detectable viral load and lower CD4 counts, was associated with higher prevalence of marijuana use. These finding s are in contrast with one published longitudinal study among HIV positive women et al., 2012) In analysis t hat included the full sample, found that passage of medical marijuana laws were associated to a modest increase in marijuana use. T his increase was not observed in analysis that

PAGE 122

122 was limited to only the HIV positive individuals This finding, which was unexpected, suggests that the passage of medical marijuana laws do not have a differential positive influence on rates of marijuana use among HIV positive individuals compared to HIV negative individuals. However, given that the c urrent analysis was limited to only four states, this question needs to be re evaluated in another sample that will include HIV positive individuals from more states. Study 2 Long Term Trajectories of Marijuana use In this study, we focused on describing long marijuana use as opposed to describing aggregate data on marijuana use. The principle of describing individual pattern of marijuana use is that there is individual variability in the pattern of a behavior and that this i ndividual variability may explain different etiologic pathways to substance use careers. The MACS data offered a unique opportunity to describe for the first time the natural history of marijuana use among a sample of HIV positive and HIV negative men. An advanced statistical approach was used to identify different patterns (or trajectories) in the sample and to determine predictors of membership in the different trajectories and covariates measured during the course of the trajectory that may serve to alter the course of the trajectory. I speculated that 3 6 distinct trajectories of marijuana use would emerge in this sample based on the number of trajectories observed in prior studies. I further postulated that sociodemographic characteristics, psyc hosocial factors and HIV related clinical factors would predict trajectory membership and alter the trajectories. Our analysis identified four trajectories of marijuana use including groups that abstained or used infrequently, that increased use, decreased use and persistently used heavily over time. The study found similar trajectories of use among HIV positive men. One of the most consistent

PAGE 123

123 predictors of trajectory membership was younger age and a positive HIV status; which was were associated with membe rship in all trajectories that reported marijuana use as compared to the trajectory that abstained. Among the HIV positive individuals, the study found that ART use was associated with increasing the marijuana use among the group that abstained and detect able viral load was associated with increasing marijuana use in the group that increased use. The data collection procedures do not allow to us to determine whether ART use and or having a detectable viral load preceded marijuana use. But this finding sug gests that the association between marijuana use and ART use, adherence and viral suppression needs further clarification. This study is among the first to demonstrate that different trajectories of marijuana use exis t among a sample of HIV positive and HI V negative MSM and that several factors emerged to be associated with the trajectories. It will be important for additional studies to determine whether these different trajectories especially the trajectory of persistent heavy use are associated with adverse health outcomes, particularly among HIV positive individuals. Study 3 Cumulative Exposures to Marijuana and Cognitive Change In this study, it was hypothesized that greater cumulative exposure to marijuana use will be associated with longitudina l changes in cognitive performance. Our study found few significant effects of cumulat ive exposure to marijuana use and adverse changes in cognitive function measures in processing speed, executive functions and verbal memory in both HIV positive and HIV n egative samples. Overall, our finding s were unexpected given that prior studies have demonstrated adverse effects of marijuana use in both HIV positive (Cristiani et al., 2004; Gonzalez et al., 2011; Thames et al., 2015) and HIV negative individuals (Auer et al., 2016; Meier et al., 2012)

PAGE 124

124 There are several potential explanations for these findings. Under some conditions THC and other cannabinoids in marijuana may be neuroprotective (Eubanks et al., 2006; Ramrez et al., 2005) Related to this is that cognitive impairments associated with marijuana tend to dissipate following abstin ence (H G Pope Jr et al., 2001; Harrison G Pope Jr et al., 2002) Furthermore, it is possible that c hronic heavy marijuana use may be associated with the development of tolerance to the impairing eff ects of marijuana on cogniti ve functions (Johannes G. Ramaekers et al., 2011) The study also did not collect historical information about marijuana use prior to enrolling in the MACS specifically and most importantly the age at which they first used marijuana an d their cumulative use prior to enrolling in the MACS. Finally, the preponderance of evidence suggests that the cognitive domains most affected by marijuana use are in learning, memory, attention and concentration and that the impairments are generally of small magnitude (Crane et al., 2013; Gonzalez, 2007; Igor Grant et al., 2003) The MACS protocol selected the cognitive function measures based on the ir sensitivity to detecting HIV associated neurocognitive disorders (Becker et al., 2014) However, cognitive function measures correlate moderately wit h every day cognitive tasks (Chaytor & Schmitter Edgecombe, 2003) Other variab les including informant ratings and behavioral observations variables should be considered along with cognitive function measures Repeated cognitive function measurements as is the case in the MACS may be associated with test familiarity and an underestimation of the true rate of cognitive change. The results from this study are not definitive evidence that long term marijuana use is not associated changes in cognitive functions Our findings need

PAGE 125

125 replic ation in other sociodemographic diverse samples (e.g. women less educated ). The field also needs to underst and whether different strains of marijuana and the THC/CBD ratio of marijuana confer differential cognitive effects. This requires that future studi es consider collecting finer measures of marijuana use than was available for this study. Strengths and Limitations The current project has several strengths. This is the first study to assess and compared long term longitudinal trends in prevalence of an d risk factors for marijuana use among HIV positive and HIV negative individuals. The second study of this project is the also the first to describe and compare long te rm individual level trajectories of marijuana use among HIV positive and HIV negative in dividuals. Study 3 addressed an important research relevant to HIV positive individuals who use marijuana; specifically, whether greater cumulative exposure to marijuana is associated with cognitive changes which up till now have not been addressed. Howe ver, there are some limitations in the studies included in the current project. First is that the studies relied on self reported data for most of the important variables including marijuana use. T hese data may be influenced by recall and social desirabili ty bias although during the later period of the MACS ACASI methods was used to elicit partic ipant response to drug use Related to this issue is the less precise measurement of marijuana used in the studies as the frequency options may not capture other re levant parameters of use. For instance, quantity of use (e.g. number of times/joints smoked per day) has been demonstrated to be a stronger predictor of marijuana related health problems (Walden & Earleywine, 2008; Zeisser et al., 2012) Furthermore, self report measures of marijuana use is weakly associated with more objective measures of

PAGE 126

126 use such as number of j oints per gram, price and subjective potency (van der Pol et al., 2013) In order to advance the field, it is important that more direct biomarkers of marijuana use are utilized. Second is that our results may be influenced by attrition either from death or dropouts, although the s tudies accounted for their effects using different methods (e.g. IPAW and Multiple Imputations) Thirdly is that the studies were among men who have sex with men, were predominantly non Hispanic whites and had greater level of education than a representati ve sample of HIV positive and HIV negative individuals. It is possible that the results may be different among women living with HIV and other race/ethnic groups. Implications for Future Research Study 1 found tenuous associations between passage of medi cal marijuana laws and increased marijuana use. The current study was limited to only four states. Given that more states are considering passing medical marijuana laws in the November ballot an indication that the trend for legislations for medical mariju ana use will continue. There is a need further investigations on whether medical marijuana laws influence increased marijuana use among HIV positive individuals (including women). The second study showed that HIV positive men exhibited distinct individua l patterns of marijuana use over time and several factors were associated with the different patterns. The study identified a group that declined use over time and a group that remained heavy users over time. The group that declined use over time without a ny apparent intervention represents an interesting finding. Future investigations into other factors beyond that available in the current study that predict d eclining use over time may inform the development of preventive interventions by considering these factors in intervention designs. With regard to the group that demonstrated persistent heavy use

PAGE 127

127 over time. There is much to learn about whether this pattern of use is associated with other HIV clinical outcomes such as engagement in medical care, adheren ce to medications and viral suppressions. It will be crucial for the next set of investigations to distinguish between those who use marijuana for medical versus recreational purposes and the THC/CBD contents of the marijuana consumed. The third study di d not find statistically significant associations between cumulative exposure to marijuana use and adverse changes in cognitive function measures over a 17 year period However, much was learned from this study and there remain unanswered questions The wi de confidence intervals around the estimates suggests less precise estimates of marijuana exposure and further underscores the need to utilize more objective measures in future research. There is a need for a need to replicate this study in other samples. A growing amount of research points to sex differences in the impairing effects of marijuana use on cognitive function and thus data is needed on the effect of marijuana use on cognitive function among women with HIV. Other issues that warrant further rese arch include whether different strains and or THC/CBD have differential effects on the brain and cognition function Further if marijuana use is associated with cognitive impairments, does abstinence sufficiently restore cognitive ability

PAGE 128

128 A PPENDIX SELECTIVE ATTRITION Of the 1, 982 participants included in the study at baseline, 1,031 were seen at MACS visit 60 and had data on outcome measures. Therefore, in order to minimize the potential for bias due to selective attrition, i nverse probability of att rition weighting (IPAW) was computed and utilized. In developing the IPAW, a sequence of steps were followed as previously described (Weuve et al., 2012) F irst I constructed models of the probability of continuation in the study i.e. remaining alive and not lost to follow up and from these models, predicted probabilities of remaining in the study for wave of visits contributing to our analysis was computed Next, these probabilities were used to compute analytical weights that are the inverse proportion to the probability of remaining alive and in the study. The intuition behind the weighting approach is t hat observations with characteristics associated with lower probabilities of remaining alive and in the study are assigned higher weights and thus compensating for their underrepresentation of these types of observations in the analysis. C ovariates in the models used to develop these probabilities were selected based on prior research indicating their strong association with attrition as well as those hypothesize d will influence attrition These covariates includ ed : age, race, education, MACS study center, MACS enrollment cohort, alcohol consumption, smoking, marijuana use, stimulant use at the previous visit, cognitive performance at the previous visit, depression at the previous visit and frailty index at the previous visit. The IPAW were in cluded in all m odels estimated.

PAGE 129

129 REFERENCE S Schambelan, M. (2003). Short Term Effects of Cannabinoids in Patients with HIV 1 I nfectionA Randomized, Placebo Controlled Clinical Trial. Annals of Internal Medicine 139 (4), 258 266. http://doi.org/10.7326/0003 4819 139 4 200308190 00008 Akaike, H. (1974). New look at statistical model identification, 19 716 723. Akhtar Khaleel, W. Z Plankey, M. (2016). Long Term Cigarette Smoking Trajectories Among HIV Seropositive and Seronegative MSM in the Multicenter AIDS Cohort Study. AIDS and Behavior 1 9. http://doi.org/10 .1007/s10461 016 1343 8 Albright, A. V., Soldan, S. S., & Gonzlez Scarano, F. (2003). Pathogenesis of human immunodeficiency virus induced neurological disease. Journal of Neurovirology 9 (2), 222 227. http://doi.org/10.1080/13550280390194073 Allshouse, A A. M., MaWhinney, S. S., Jankowski, C. M., Kohrt, W. M., Campbell, T. B., & Erlandson, K. M. M. (2015). The Impact of Marijuana Use on the Successful Aging of HIV Infected Adults. Journal of Acquired Immune Deficiency Syndromes 69 (2), 187 192. http://do i.org/10.1097/QAI.0000000000000562 Amedee, A. M., Nichols, W. A., LeCapitaine, N. J., Stouwe, C. V., Birke, L. L., Lacour, 9 tetrahydrocannabinol administration may not attenuate simian immunodeficiency virus disease pr ogression in female rhesus macaques. AIDS Research and Human Retroviruses 30 (12), 1216 1225. http://doi.org/10.1089/AID.2014.0108 An, S. F., Groves, M., Gray, F., & Scaravilli, F. (1999). Early entry and widespread cellular involvement of HIV 1 DNA in bra ins of HIV 1 positive asymptomatic individuals. Journal of Neuropathology and Experimental Neurology 58 (11), 1156 1162. Anderson, A. M., Higgins, M. K., Ownby, R. L., & Waldrop Valverde, D. (2015). Changes in neurocognition and adherence over six months i n HIV infected individuals with cocaine or heroin dependence. AIDS Care 27 (3), 333 337. http://doi.org/10.1080/09540121.2014.985183 Anderson, D. M., Hansen, B., & Rees, D. I. (2015). Medical Marijuana Laws and Teen Marijuana Use. American Law and Economic s Review ahv002. http://doi.org/10.1093/aler/ahv002 E. (2007). Updated research nosology for HIV associated neurocognitive disorders. Neurology 69 (18), 1789 1799. http://doi.org/10.1212/01.WNL.0000287431.88658.8b

PAGE 130

130 Arnett, J. J. (2005). The Developmental Context of Substance use in Emerging Adulthood. Journal of Drug Issues 35 (2), 235 254. http://doi.org/10.1177/002204260503500202 Arnsten, J. H., Demas, P. A., Grant R. W., Gourevitch, M. N., Farzadegan, H., Howard, A. A., & Schoenbaum, E. E. (2002). Impact of Active Drug Use on Antiretroviral Therapy Adherence and Viral Suppression in HIV infected Drug Users. Journal of General Internal Medicine 17 (5), 377 381. htt p://doi.org/10.1046/j.1525 1497.2002.10644.x Atkinson, T. M., Ryan, J. P., Lent, A., Wallis, A., Schachter, H., & Coder, R. (2010). Three trail making tests for use in neuropsychological assessments with brief intertest intervals. Journal of Clinical and E xperimental Neuropsychology 32 (2), 151 158. http://doi.org/10.1080/13803390902881934 M. J. (2016). Association Between Lifetime Marijuana Use and Cognitive Functio n in Middle Age: The Coronary Artery Risk Development in Young Adults (CARDIA) Study. JAMA Internal Medicine 176 (3), 352 361. http://doi.org/10.1001/jamainternmed.2015.7841 Battisti, R. A., Roodenrys, S., Johnstone, S. J., Pesa, N., Hermens, D. F., & Solo wij, N. (2010). Chronic cannabis users show altered neurophysiological functioning on Stroop task conflict resolution. Psychopharmacology 212 (4), 613 624. http://doi.org/10.1007/s00213 010 1988 3 Bauer, D. J., & Curran, P. J. (2003). Distributional assump tions of growth mixture models: implications for overextraction of latent trajectory classes. Psychological Methods 8 (3), 338 363. http://doi.org/10.1037/1082 989X.8.3.338 Baum, M. K., Rafie, C., Lai, S., Sales, S., Page, B., & Campa, A. (2009). Crack coc aine use accelerates HIV disease progression in a cohort of HIV positive drug users. Journal of Acquired Immune Deficiency Syndromes (1999) 50 (1), 93 99. http://doi.org/10.1097/QAI.0b013e3181900129 Becker, J. T., Kingsley, L. A., Molsberry, S., Reynolds, Selnes, O. A. (2014). Cohort Profile: Recruitment cohorts in the neuropsychological substudy of the Multicenter AIDS Cohort Study. International Journal of Epidemiology dyu092. http://doi.org/10.1093/ije/dyu092 Bidaut Russ ell, M., Devane, W. A., & Howlett, A. C. (1990). Cannabinoid receptors and modulation of cyclic AMP accumulation in the rat brain. Journal of Neurochemistry 55 (1), 21 26. Bolla, K. I., Brown, K., Eldreth, D., Tate, K., & Cadet, J. L. (2002). Dose related neurocognitive effects of marijuana use. Neurology 59 (9), 1337 1343.

PAGE 131

131 Bonn Miller, M. O., Oser, M. L., Bucossi, M. M., & Trafton, J. A. (2012). Cannabis use and HIV antiretroviral therapy adherence and HIV related symptoms. Journal of Behavioral Medicine http://doi.org/10.1007/s10865 012 9458 5 Borgelt, L. M., Franson, K. L., Nussbaum, A. M., & Wang, G. S. (2013). The Pharmacologic and Clinical Effects of Medical Cannabis. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy 33 (2), 195 209. http://doi.org/10.1002/phar.1187 Bossong, M. G., Jager, G., van Hell, H. H., Zuurman, L., Jansma, J. M., Mehta, M. A., tetrahydrocannabinol administration on human encoding and recall memory function: a pharmacologica l FMRI study. Journal of Cognitive Neuroscience 24 (3), 588 599. http://doi.org/10.1162/jocn_a_00156 & Hogg, R. S. (2001). Mary Jane and her patients: sociodemographic and clinical characteristics of HIV positive individuals using medical marijuana and antiretroviral agents. AIDS (London, England) 15 (4), 532 533. Brook, J. S., Lee, J. Y., & Brook, D. W. (2014). Trajectories of Marijuana Use Beginning in Adolescence Pred ict Tobacco Dependence in Adulthood. Substance Abuse 0 (0), 1 8. http://doi.org/10.1080/08897077.2014.964901 Brook, J. S., Lee, J. Y., Brown, E. N., Finch, S. J., & Brook, D. W. (2011). Developmental Trajectories Of Marijuana Use From Adolescence To Adulth ood: Personality And Social Role Outcomes 1,2 Psychological Reports 108 (2), 339 357. http://doi.org/10.2466/10.18.PR0.108.2.339 357 Brook, J. S., Lee, J. Y., Finch, S. J., & Brook, D. W. (2014). Developmental trajectories of marijuana use from adolescenc e to adulthood: Relationship with using weapons including guns. Aggressive Behavior 40 (3), 229 237. http://doi.org/10.1002/ab.21520 Brook JS, Zhang C, & Brook DW. (2011). Developmental trajectories of marijuana use from adolescence to adulthood: Personal predictors. Archives of Pediatrics & Adolescent Medicine 165 (1), 55 60. http://doi.org/10.1001/archpediatrics.2010.248 Brown, T. L., Flory, K., Lynam, D. R., Leukefeld, C., & Clayton, R. R. (2004). Comparing the Developmental Trajectories of Marijuana Use of African American and Caucasian Adolescents: Patterns, Antecedents, and Consequences. Experimental and Clinical Psychopharmacology 12 (1), 47 56. http://doi.org/10.1037/1064 1297.12.1.47 Bruce, D., Harper, G. W., & Fernandez, M. I. (2013). Heavy Marijua na Use Among Gay And Bisexual Male Emerging Adults Living With HIV/AIDS. Journal of HIV/AIDS & Social Services 12 (1), 26 48. http://doi.org/10.1080/15381501.2012.735171

PAGE 132

132 CDC. (2012). Estimated HIV incidence among adults and adolescents in the United Stat es, 2007 2010. HIV Surveillance Supplemental Repor (No. 17) (p. 4). Retrieved from http://www.cdc.gov/hiv/topics/surveillance/ resources/reports/#supplementa CDC. (2014). Vital Signs: HIV Diagnosis, Care, and Treatment Among Persons Living with HIV Unite d States, Retrieved from http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6347a5.htm?s_cid=mm6347a5 _w CDC. (2015). Monitoring selected national HIV prevention and care objectives by using HIV surveillance data United States and 6 dependent areas 2013. HIV S urveillance Supplemental Report (No. 20). Retrieved from : http://www.cdc.gov/hiv/library/reports/surveillance/ Cerd, M., Wall, M., Keyes, K. M., Galea, S., & Hasin, D. (2012). Medical marijuana laws in 50 states: investigating the relationship between st ate legalization of medical marijuana and marijuana use, abuse and dependence. Drug and Alcohol Dependence 120 (1 3), 22 27. http://doi.org/10.1016/j.drugalcdep.2011.06.011 Chander, G., Lau, B., & Moore, R. D. (2006). Hazardous alcohol use: a risk factor f or non adherence and lack of suppression in HIV infection. Journal of Acquired Immune Deficiency Syndromes (1999) 43 (4), 411 417. http://doi.org/10.1097/01.qai.0000243121.44659.a4 Chang, L., Cloak, C., Yakupov, R., & Ernst, T. (2006). Combined and indepen dent effects of chronic marijuana use and HIV on brain metabolites. Journal of Neuroimmune Pharmacology: The Official Journal of the Society on NeuroImmune Pharmacology 1 (1), 65 76. http://doi.org/10.1007/s11481 005 9005 z Chao, C., Jacobson, L. P., Tashk in, D., Martnez Maza, O., Roth, M. D., Margolick, J. in HIV uninfected and HIV infected men. Drug and Alcohol Dependence 94 (1 3), 165 171. http://doi.org/10.1016/j.drugalcdep. 2007.11.010 Chaytor, N., & Schmitter Edgecombe, M. (2003). The ecological validity of neuropsychological tests: a review of the literature on everyday cognitive skills. Neuropsychology Review 13 (4), 181 197. Cheetham, A., Allen, N. B., Whittle, S., Simmon s, J. G., Ycel, M., & Lubman, D. I. (2012). Orbitofrontal Volumes in Early Adolescence Predict Initiation of Cannabis Use: A 4 Year Longitudinal and Prospective Study. Biological Psychiatry 71 (8), 684 692. http://doi.org/10.1016/j.biopsych.2011.10.029

PAGE 133

133 Ch erner, M., Letendre, S., Heaton, R. K., Durelle, J., Marquie Beck, J., Gragg, B., & Grant, I. (2005). Hepatitis C augments cognitive deficits associated with HIV infection and methamphetamine. Neurology 64 (8), 1343 1347. http://doi.org/10.1212/01.WNL.0000 158328.26897.0D Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in developmental psychopathology. Development and Psychopathology 8 (4), 597 600. http://doi.org/10.1017/S0954579400007318 Clifford, D. B., & Ances, B. M. (2013). HIV as sociated neurocognitive disorder. The Lancet Infectious Diseases 13 (11), 976 986. http://doi.org/10.1016/S1473 3099(13)70269 X Clifford, D. B. a, Evans, S. R. b, Yang, Y. b, & Gulick, R. M. c. (2005). The neuropsychological and neurological impact of hepa titis C virus co infection in HIV infected subjects. AIDS Cochran, S. D., Ackerman, D., Mays, V. M., & Ross, M. W. (2004). Prevalence of non medical drug use and dependence among homosexually active men and women in the US population. Addiction 99 (8), 98 9 998. http://doi.org/10.1111/j.1360 0443.2004.00759.x Cohen, M. S., Chen, Y. Q., McCauley, M., Gamble, T., Hosseinipour, M. C., 1 infection with early antiretroviral therapy. The New England Journal of Medicine 365 (6), 493 505. http://doi.org/10.1056/NEJMoa1105243 Cook, R. L., & Clark, D. B. (2005). Is there an association between alcohol consumption and sexually transmitted diseases? A systematic review. Sexually Transmitted Diseases 32 (3) 156 164. H. (2013). Alcohol consumption trajectory patterns in adult women with HIV infection. AIDS and Behavior 17 (5), 1705 1712. http://doi.org/10.1007/s10461 012 0270 6 Cooper, C. L., & Cameron, D. W. (2005). Effect of Alcohol Use and Highly Active Antiretroviral Therapy on Plasma Levels of Hepatitis C Virus (HCV) in Patients Coinfected with HIV and HCV. Clinical Infectious Diseases 41 (Supplement 1), S105 S109. ht tp://doi.org/10.1086/429506 Reynolds, N. (2009). Marijuana effectiveness as an HIV self care strategy. Clinical Nursing Research 18 (2), 172 193. http://doi.org/10.1177/10 54773809334958

PAGE 134

134 Costantino, C. M., Gupta, A., Yewdall, A. W., Dale, B. M., Devi, L. A., & Chen, B. K. (2012). Cannabinoid receptor 2 mediated attenuation of CXCR4 tropic HIV infection in primary CD4+ T cells. PloS One 7 (3), e33961. http://doi.org/10.1371/j ournal.pone.0033961 Crane, N. A., Schuster, R. M., Fusar Poli, P., & Gonzalez, R. (2013). Effects of cannabis on neurocognitive functioning: recent advances, neurodevelopmental influences, and sex differences. Neuropsychology Review 23 (2), 117 137. http:/ /doi.org/10.1007/s11065 012 9222 1 Crean, R. D., Crane, N. A., & Mason, B. J. (2011). An Evidence Based Review of Acute and Long Term Effects of Cannabis Use on Executive Cognitive Functions. Journal of Addiction Medicine March 2011 5 (1), 1 8. http://doi. org/10.1097/ADM.0b013e31820c23fa Cristiani, S. A., Pukay Martin, N. D., & Bornstein, R. A. (2004). Marijuana use and cognitive function in HIV infected people. The Journal of Neuropsychiatry and Clinical Neurosciences 16 (3), 330 335. http://doi.org/10.117 6/appi.neuropsych.16.3.330 Cysique, L. A., Maruff, P., & Brew, B. J. (2004). Prevalence and pattern of neuropsychological impairment in human immunodeficiency virus infected/acquired immunodeficiency syndrome (HIV/AIDS) patients across pre and post highly active antiretroviral therapy eras: A combined study of two cohorts. Journal of NeuroVirology 10 (6), 350 357. http://doi.org/10.1080/13550280490521078 David, J., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol and other d rug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin 112 (1), 64 105. http://doi.org/10.1037/0033 2909.112.1.64 de Jong, B. C., Prentiss, D., McFarland, W., Machekano, R., & Israelski, D. M. ( 2005). Marijuana use and its association with adherence to antiretroviral therapy among HIV infected persons with moderate to severe nausea. Journal of Acquired Immune Deficiency Syndromes (1999) 38 (1), 43 46. Detels, R., Phair, J. P., Saah, A. J., Rinald Vermund, S. (1992). Recent Scientific Contributions to Understanding HIV/AIDS from the Multicenter AIDS Cohort Study. Journal of Epidemiology 2 (2sup), 11 19. http://doi.org/10.2188/jea.2.2sup_11 Detels, R., Tarwater, P., Phair, J. P., Margolick, J., Riddler, S. A., Muoz, A., & Multicenter AIDS Cohort Study. (2001). Effectiveness of potent antiretroviral therapies on the incidence of opportunistic infections before and after AIDS diagnosis. AIDS (London, England) 15 ( 3), 347 355.

PAGE 135

13 5 Devane, W. A., Dysarz, F. A., Johnson, M. R., Melvin, L. S., & Howlett, A. C. (1988). Determination and characterization of a cannabinoid receptor in rat brain. Molecular Pharmacology 34 (5), 605 613. DeWit, D. J., Adlaf, E. M., Offord, D. R., & Ogborne, A. C. (2000). Age at first alcohol use: a risk factor for the development of alcohol disorders. The American Journal of Psychiatry 157 (5), 745 750. http://doi.org/10.1176/appi.ajp.157.5.745 Dierker, L., Rose, J., Tan, X., & Li, R. (2010). Unco vering Multiple Pathways to Substance Use: A comparison of methods for identifying population subgroups. The Journal of Primary Prevention 31 (5 6), 333 348. http://doi.org/10.1007/s10935 010 0224 6 Dudley, J., Jin, S., Hoover, D., Metz, S., Thackeray, R., & Chmiel, J. (1995). The Multicenter AIDS Cohort Study: retention after 9 1/2 years. American Journal of Epidemiology 142 (3), 323 330. Wilson, T. E. (2012). Medicinal an d Recreational Marijuana Use Among HIV 2010. JAIDS Journal of Acquired Immune Deficiency Syndromes 61 (5), 618 626. http://doi.org/10.1097/QAI.0b013e318273ab3a Eassey, J. M., Gibson, C L., & Krohn, M. D. (2014). Using a Group Based Trajectory Approach to Assess Risk and Protective Factors of Marijuana Use. Journal of Drug Issues 22042614542508. http://doi.org/10.1177/0022042614542508 Ellickson, P. L., Martino, S. C., & Collins, R. L. (2004). Marijuana Use From Adolescence to Young Adulthood: Multiple Developmental Trajectories and Their Associated Outcomes. Health Psychology 23 (3), 299 307. http://doi.org/10.1037/0278 6133.23.3.299 Ellis, R. J., Badiee, J., Vaida, F., Letendre, S., He Group. (2011). CD4 nadir is a predictor of HIV neurocognitive impairment in the era of combination antiretroviral therapy. AIDS (London, England) 25 (14), 1747 1751. http://doi.org/10.1097/QAD.0b013e32834a40cd Eubanks, L. M., Rogers, C. J., Beuscher, A. E., Koob, G. F., Olson, A. J., Dickerson, T. J., & Janda, K. D. (2006). A Molecular Link Between the Active Component of Molecular Pharmaceutics 3 (6), 773 777. http://doi.org/ 10.1021/mp060066m Executive Order -HIV Care Continuum Initiative. (2013, July 15). Retrieved May 23, 2016, from https://www.whitehouse.gov/the press office/2013/07/15/executive order hiv care continuum initiative

PAGE 136

136 Fernndez Serrano, M. J., Prez Garca, M ., Schmidt Ro Valle, J., & Verdejo Garca, A. (2010). Neuropsychological consequences of alcohol and drug abuse on different components of executive functions. Journal of Psychopharmacology (Oxford, England) 24 (9), 1317 1332. http://doi.org/10.1177/02698 81109349841 Ferraro, D. P. (1980). Acute effects of marijuana on human memory and cognition. NIDA Research Monograph 31 98 119. Flores, A. (2014). National Trends in Public Opinion on LGBT Rights in the United States Retrieved from http://williamsinstit ute.law.ucla.edu/wp content/uploads/POP natl trends nov 2014.pdf Fontes, M. A., Bolla, K. I., Cunha, P. J., Almeida, P. P., Jungerman, F., Laranjeira, R. tool for detecting execut ive deficits in chronic cannabis users. Journal of Clinical and Experimental Neuropsychology 33 (5), 523 531. http://doi.org/10.1080/13803395.2010.535505 Fried, P. A., & Smith, A. M. (2001). A literature review of the consequences of prenatal marihuana exp osure. An emerging theme of a deficiency in aspects of executive function. Neurotoxicology and Teratology 23 (1), 1 11. Fried, P. A., Watkinson, B., & Gray, R. (2005). Neurocognitive consequences of marihuana -a comparison with pre drug performance. Neurot oxicology and Teratology 27 (2), 231 239. http://doi.org/10.1016/j.ntt.2004.11.003 Furler, M. D., Einarson, T. R., Millson, M., Walmsley, S., & Bendayan, R. (2004). Medicinal and recreational marijuana use by patients infected with HIV. AIDS Patient Care a nd STDs 18 (4), 215 228. http://doi.org/10.1089/108729104323038892 Fusar Poli, P., Crippa, J. A., Bhattacharyya, S., Borgwardt, S. J., Allen, P., Martin tetrahydrocannabinol and cannabidiol on neural activation during emotional processing. Archives of General Psychiatry 66 (1), 95 105. http://doi.org/10.1001/archgenpsychiatry.2008.519 Gallup. (2013). For first time, Americans favor legalizing marijuana. 2013 Retrieved from http://www.webcita tion .org/6QksXxomm. (3 28 2016) Gannon, P., Khan, M. Z., & Kolson, D. L. (2011). Current understanding of HIV associated neurocognitive disorders pathogenesis. Current Opinion in Neurology 24 (3), 275 283. http://doi.org/10.1097/WCO.0b013e32834695fb Gardn er, E. M., McLees, M. P., Steiner, J. F., Rio, C. del, & Burman, W. J. (2011). The Spectrum of Engagement in HIV Care and its Relevance to Test and Treat Strategies for Prevention of HIV Infection. Clinical Infectious Diseases 52 (6), 793 800. http://doi.o rg/10.1093/cid/ciq243

PAGE 137

137 Ghosn, J., Leruez Evarist ANRS EP 49 Study Group. (2014). HIV 1 DNA levels in peripheral blood mononuclear cells and cannabis use are associated with intermittent HIV shedding in semen of men who have sex with men on successful antiretroviral regimens. Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 58 (12), 1763 1770. http://doi.org/10.1093/cid/ciu187 Giorgi, J. V., Park, L. (1990). Quality control in the flow cytometric measurement of T lymphocyte subsets: the multicenter AIDS cohort study experience. The Multicenter AIDS Cohort Study Group. Clin ical Immunology and Immunopathology 55 (2), 173 186. Gissln, M., Price, R. W., & Nilsson, S. (2011). The definition of HIV associated neurocognitive disorders: are we overestimating the real prevalence? BMC Infectious Diseases 11 356. http://doi.org/10. 1186/1471 2334 11 356 Glass, M., Dragunow, M., & Faull, R. L. (1997). Cannabinoid receptors in the human brain: a detailed anatomical and quantitative autoradiographic study in the fetal, neonatal and adult human brain. Neuroscience 77 (2), 299 318. Gonzal ez, R. (2007). Acute and non acute effects of cannabis on brain functioning and neuropsychological performance. Neuropsychology Review 17 (3), 347 361. http://doi.org/10.1007/s11065 007 9036 8 Gonzalez, R., Schuster, R. M., Mermelstein, R. J., Vassileva, J ., Martin, E. M., & Diviak, K. R. (2012). Performance of young adult cannabis users on neurocognitive measures of impulsive behavior and their relationship to symptoms of cannabis use disorders. Journal of Clinical and Experimental Neuropsychology 34 (9), 962 976. http://doi.org/10.1080/13803395.2012.703642 Gonzalez, R., Schuster, R. M., Vassileva, J., & Martin, E. M. (2011). Impact of HIV and a history of marijuana dependence on procedural learning among individuals with a history of substance dependence. Journal of Clinical and Experimental Neuropsychology 33 (7), 735 752. http://doi.org/10.1080/13803395.2011.553584 Gonzlez Scarano, F., & Martn Garca, J. (2005). The neuropathogenesis of AIDS. Nature Reviews. Immunology 5 (1), 69 81. http://doi.org/10.10 38/nri1527 Grant, I., Atkinson, J. H., Hesselink, J. R., Kennedy, C. J., Richman, D. D., Spector, S. A., & McCutchan, J. A. (1987). Evidence for early central nervous system involvement in the acquired immunodeficiency syndrome (AIDS) and other human immun odeficiency virus (HIV) infections. Studies with neuropsychologic testing and magnetic resonance imaging. Annals of Internal Medicine 107 (6), 828 836.

PAGE 138

138 Grant, I., Gonzalez, R., Carey, C. L., Natarajan, L., & Wolfson, T. (2003). Non acute (residual) neuroco gnitive effects of cannabis use: a meta analytic study. Journal of the International Neuropsychological Society: JINS 9 (5), 679 689. http://doi.org/10.1017/S1355617703950016 Grant, J. E., Chamberlain, S. R., Schreiber, L., & Odlaug, B. L. (2012). Neuropsy chological deficits associated with cannabis use in young adults. Drug and Alcohol Dependence 121 (1 2), 159 162. http://doi.org/10.1016/j.drugalcdep.2011.08.015 Green, K. E., & Feinstein, B. A. (2012). Substance Use in Lesbian, Gay, and Bisexual Populatio ns: An Update on Empirical Research and Implications for Treatment. Psychology of Addictive Behaviors 26 (2), 265 278. http://doi.org/10.1037/a0025424 Gribble, J. N., Miller, H. G., Cooley, P. C., Catania, J. A., Pollack, L., & Turner, C. F. (2000). The im pact of T ACASI interviewing on reported drug use among men who have sex with men. Substance Use & Misuse 35 (6 8), 869 890. Gruber, S. A., Dahlgren, M. K., Sagar, K. A., Gnenc, A., & Killgore, W. D. S. (2012). Age of onset of marijuana use impacts inhibi tory processing. Neuroscience Letters 511 (2), 89 94. http://doi.org/10.1016/j.neulet.2012.01.039 Gruber, S. A., Sagar, K. A., Dahlgren, M. K., Racine, M., & Lukas, S. E. (2012). Age of Onset of Marijuana Use and Executive Function. Psychology of Addictive 26 (3), 496 506. http://doi.org/10.1037/a0026269 Everall, I. P. (2011). Dopamine receptor D3 geneti c polymorphism (rs6280TC) is associated with rates of cognitive impairment in methamphetamine dependent men with HIV: preliminary findings. Journal of Neurovirology 17 (3), 239 247. http://doi.org/10.1007/s13365 011 0028 3 Halkitis, P. N., Kapadia, F., Bub K. L., Barton, S., Moreira, A. D., & Stults, C. B. (2015). A Longitudinal Investigation of Syndemic Conditions Among Young Gay, Bisexual, and Other MSM: The P18 Cohort Study. AIDS and Behavior 19 (6), 970 980. http://doi.org/10.1007/s10461 014 0892 y Hal kitis, P. N., Siconolfi, D. E., Stults, C. B., Barton, S., Bub, K., & Kapadia, F. (2014). Modeling substance use in emerging adult gay, bisexual, and other YMSM across time: the P18 cohort study. Drug and Alcohol Dependence 145 209 216. http://doi.org/10 .1016/j.drugalcdep.2014.10.016 Hanson, K. L., Winward, J. L., Schweinsburg, A. D., Medina, K. L., Brown, S. A., & Tapert, S. F. (2010). Longitudinal study of cognition among adolescent marijuana users over three weeks of abstinence. Addictive Behaviors 35 (11), 970 976. http://doi.org/10.1016/j.addbeh.2010.06.012

PAGE 139

139 Harper, S., Strumpf, E. C., & Kaufman, J. S. (2012). Do Medical Marijuana Laws Increase Marijuana Use? Replication Study and Extension. Annals of Epidemiology 22 (3), 207 212. http://doi.org/10.101 6/j.annepidem.2011.12.002 (2014). Patterns and correlates of cannabis use among individuals with HIV/AIDS in Maritime Canada. The Canadian Journal of Infectious Disease s & Medical Microbiology 25 (1), e1 e7. Hart, C. L., Ilan, A. B., Gevins, A., Gunderson, E. W., Role, K., Colley, J., & Foltin, R. W. (2010). Neurophysiological and cognitive effects of smoked marijuana in frequent users. Pharmacology, Biochemistry, and Be havior 96 (3), 333 341. http://doi.org/10.1016/j.pbb.2010.06.003 Harvey, M. A., Sellman, J. D., Porter, R. J., & Frampton, C. M. (2007). The relationship between non acute adolescent cannabis use and cognition. Drug and Alcohol Review 26 (3), 309 319. http ://doi.org/10.1080/09595230701247772 Feng, T. (2015). Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014: results from annual, repeated cross sectional surveys. The Lancet Psychiatry 2 (7), 601 608. http://doi.org/10.1016/S2215 0366(15)00217 5 Hasin DS, Saha TD, Kerridge BT, & et al. (2015). Prevalence of marijuana use disorders in the united states between 2001 2002 and 2012 2013. JAMA Psychia try 1 9. http://doi.org/10.1001/jamapsychiatry.2015.1858 Hasse, B., Tarr, P. E., Marques Marco, Z. (2015). Strong Impact of Smoking on Multimorbidity and Cardiovascular Risk Among Human Immunodeficiency Vi rus Infected Individuals in Comparison With the General Population. Open Forum Infectious Diseases 2 (3), ofv108. http://doi.org/10.1093/ofid/ofv108 CHARTER Group. (20 10). HIV associated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER Study. Neurology 75 (23), 2087 2096. http://doi.org/10.1212/WNL.0b013e318200d727 Heaton, R. K., Franklin, D. R., Ellis, R. J., McCutchan, J. A., Leten dre, S. L., Leblanc, associated neurocognitive disorders before and during the era of combination antiretroviral therapy: differences in rates, nature, and predictors. Journal of Neurovirology 17 (1), 3 16. http://doi.org/10.1 007/s13365 010 0006 1

PAGE 140

140 J. (1995). The HNRC 500 -neuropsychology of HIV infection at different disease stages. HIV Neurobehavioral Research Center. Journal of the I nternational Neuropsychological Society: JINS 1 (3), 231 251. HNRC Group. (2004). The impact of HIV associated neuropsychological impairment on everyday functioning. Jour nal of the International Neuropsychological Society: JINS 10 (3), 317 331. http://doi.org/10.1017/S1355617704102130 Heimer, R., Myers, S. S., Cadman, E. C., & Kaplan, E. H. (1992). Detection by polymerase chain reaction of human immunodeficiency virus type 1 proviral DNA sequences in needles of injecting drug users. The Journal of Infectious Diseases 165 (4), 781 782. Herkenham, M., Lynn, A. B., Little, M. D., Johnson, M. R., Melvin, L. S., Costa, B. R. de, & Rice, K. C. (1990). Cannabinoid receptor localiz ation in brain. Proceedings of the National Academy of Sciences 87 (5), 1932 1936. Hermann, D., Sartorius, A., Welzel, H., Walter, S., Skopp, G., Ende, G., & Mann, K. (2007). Dorsolateral prefrontal cortex N acetylaspartate/total creatine (NAA/tCr) loss in male recreational cannabis users. Biological Psychiatry 61 (11), 1281 1289. http://doi.org/10.1016/j.biopsych.2006.08.027 M. H. (2014). Smoking cessation and recidivism i Human Immunodeficiency Virus Study. American Journal of Preventive Medicine 47 (1), 53 69. http://doi.org/10.1016/j.amepre.2014.02.010 Hile, S. J., Feldman, M. B., Alexy, E. R., & Irvine, M. K. (2016). Recent Tobacco Smoking is As sociated with Poor HIV Medical Outcomes Among HIV Infected Individuals in New York. AIDS and Behavior http://doi.org/10.1007/s10461 015 1273 x Hinkin, C. H., Hardy, D. J., Mason, K. I., Castellon, S. A., Durvasula, R. S., Lam, M. N., & Stefaniak, M. (2004 ). Medication adherence in HIV infected adults: effect of patient age, cognitive status, and substance abuse. AIDS (London, England) 18 (Suppl 1), S19 S25. HIV CAUSAL Collaboration, Ray, M., Logan, R., Sterne, J. A. C., Hernndez Daz, S., Hernn, M. A. (2010). The effect of combined antiretroviral therapy on the overall mortality of HIV infected individuals. AIDS (London, England) 24 (1), 123 137. http://doi.org/10.1097/QAD.0b013e3283324283

PAGE 141

141 Hutton, H. E., McCaul, M. E., Santora, P. B., & E rbelding, E. J. (2008). The Relationship between Recent Alcohol Use and Sexual Behaviors. Alcoholism, Clinical and Experimental Research 32 (11), 2008 2015. http://doi.org/10.1111/j.1530 0277.2008.00788.x Jackson, N. J., Isen, J. D., Khoddam, R., Irons, D. Baker, L. A. (2016). Impact of adolescent marijuana use on intelligence: Results from two longitudinal twin studies. Proceedings of the National Academy of Sciences 113 (5), E500 E508. http://doi.org/10.1073/pnas.1516648113 Jacobsen, L. K., Mencl, W. E., Westerveld, M., & Pugh, K. R. (2004). Impact of cannabis use on brain function in adolescents. Annals of the New York Academy of Sciences 1021 384 390. http://doi.org/10.1196/annals.1308.053 Jager, G., & Ramsey, N. F. (2008 ). Long term consequences of adolescent cannabis exposure on the development of cognition, brain structure and function: an overview of animal and human research. Current Drug Abuse Reviews 1 (2), 114 123. berg, J. (2013). Monitoring the future national results on adolescent drug use: Overview of key findings, 2011. Institute for Social Research, The University of Michigan, Ann Arbor, MI 2012. Jones, B. L., Nagin, D. S., & Roeder, K. (2001). A SAS Procedure Based on Mixture Models for Estimating Developmental Trajectories. Sociological Methods & Research 29 (3), 374 393. http://doi.org/10.1177/0049124101029003005 Juon, H. S., Fothergill, K. E., Green, K. M., Doherty, E. E., & Ensminger, M. E. (2011). Anteced ents and consequences of marijuana use trajectories over the life course in an African American population. Drug and Alcohol Dependence 118 (2 3), 216 223. http://doi.org/10.1016/j.drugalcdep.2011.03.027 Kalichman, S. C., & Rompa, D. (2003). HIV treatment adherence and unprotected sex practices in people receiving antiretroviral therapy. Sexually Transmitted Infections 79 (1), 59 61. http://doi.org/10.1136/sti.79.1.59 Kaslow, R. A., Ostrow, D. G., Detels, R., Phair, J. P., Polk, B. F., & Rinaldo, C. R. (198 7). The Multicenter AIDS Cohort Study: rationale, organization, and selected characteristics of the participants. American Journal of Epidemiology 126 (2), 310 318. Kellam, S. G., & Others, A. (1982). Paths Leading to Teenage Psychiatric Symptoms and Subst ance Use: Developmental Epidemiological Studies in Woodlawn. Retrieved from http://eric.ed.gov/?id=ED215787

PAGE 142

142 Kleeberger, C. A., Phair, J. P., Strathdee, S. A., Detels, R., Kingsley, L., & Jacobson, L. P. (2001). Determinants of Heterogeneous Adherence to HI V Antiretroviral Therapies in the Multicenter AIDS Cohort Study. Journal of Acquired Immune Deficiency Syndromes 26 (1), 82 92. Kuo, W. H., Wilson, T. E., Weber, K. M., Madhava, V., Richardson, J., Delapenha, R., & Jarlais, D. D. (2004). Initiation of Regu lar Marijuana Use Among a Cohort of (WIHS). AIDS Patient Care and STDs 18 (12), 702 713. http://doi.org/10.1089/apc.2004.18.702 Lane, S. D., Cherek, D. R., Tcheremissine, O. V., Li eving, L. M., & Pietras, C. J. (2005). Acute marijuana effects on human risk taking. Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology 30 (4), 800 809. http://doi.org/10.1038/sj.npp.1300620 Li, T. S., Tubiana, R., Katlama, C., Calvez, V., Ait Mohand, H., & Autran, B. (1998). Long lasting recovery in CD4 T cell function and viral load reduction after highly active antiretroviral therapy in advanced HIV 1 disease. Lancet 351 (9117), 1682 1686. Lim, S. H., Ostrow, (2010). Changes in Stimulant Drug Use Over Time in the MACS: Evidence for Resilience Against Stimulant Drug Use Among Men Who Have Sex with Men. AIDS and Behavior 16 (1), 151 158. http://d oi.org/10.1007/s10461 010 9866 x Lindl, K. A., Marks, D. R., Kolson, D. L., & Jordan Sciutto, K. L. (2010). HIV Associated Neurocognitive Disorder: Pathogenesis and Therapeutic Opportunities. Journal of Neuroimmune Pharmacology 5 (3), 294 309. http://doi.o rg/10.1007/s11481 010 9205 z Lisdahl, K. M., Gilbart, E. R., Wright, N. E., & Shollenbarger, S. (2013). Dare to delay? The impacts of adolescent alcohol and marijuana use onset on cognition, brain structure, and function. Frontiers in Psychiatry 4 53. ht tp://doi.org/10.3389/fpsyt.2013.00053 Lisdahl, K. M., Wright, N. E., Medina Kirchner, C., Maple, K. E., & Shollenbarger, S. (2014). Considering Cannabis: The Effects of Regular Cannabis Use on Neurocognition in Adolescents and Young Adults. Current Addicti on Reports 1 (2), 144 156. http://doi.org/10.1007/s40429 014 0019 6 (2002). Human immunodeficiency virus type 1 enters brain microvascular endothelia by macropinocyt osis dependent on lipid rafts and the mitogen activated protein kinase signaling pathway. Journal of Virology 76 (13), 6689 6700.

PAGE 143

143 Lopez, O. L., Becker, J. T., Dew, M. A., & Caldararo, R. (2004). Risk modifiers for peripheral sensory neuropathy in HIV infec tion/AIDS. European Journal of Neurology 11 (2), 97 102. Lucas, G. M., Cheever, L. W., Chaisson, R. E., & Moore, R. D. (2001). Detrimental effects of continued illicit drug use on the treatment of HIV 1 infection. Journal of Acquired Immune Deficiency Synd romes (1999) 27 (3), 251 259. Lutge, E. E., Gray, A., & Siegfried, N. (2013). The medical use of cannabis for reducing morbidity and mortality in patients with HIV/AIDS. The Cochrane Database of Systematic Reviews 4 CD005175. http://doi.org/10.1002/14651 858.CD005175.pub3 Lynne Landsman, S. D., Livingston, M. D., & Wagenaar, A. C. (2013). Effects of state medical marijuana laws on adolescent marijuana use. American Journal of Public Health 103 (8), 1500 1506. http://doi.org/10.2105/AJPH.2012.301117 Lyons, M. J., Bar, J. L., Panizzon, M. S., Toomey, R., Eisen, S., Xian, H., & Tsuang, M. T. (2004). Neuropsychological consequences of regular marijuana use: a twin study. Psychological Medicine 34 (7), 1239 1250. Macalino, G. E., Celentano, D. D., Latkin, C., St rathdee, S. A., & Vlahov, D. (2002). Risk behaviors by audio computer assisted self interviews among HIV seropositive and HIV seronegative injection drug users. AIDS Education and Prevention: Official Publication of the International Society for AIDS Educa tion 14 (5), 367 378. Marshall, B. D. L., Operario, D., Bryant, K. J., Cook, R. L., Edelman, E. J., Gaither, J. Infected Men Who Have Sex With Men: A Cohort Study of United States Veterans1. Drug and Alcohol Dependence http://doi.org/10.1016/j.drugalcdep.2014.12.023 Marshall, B. D. L., Shoveller, J. A., Kahler, C. W., Koblin, B. A., Mayer, K. H., Mimiaga, Have Sex with Men: A Longitudinal, Group Based Analysis. Alcoholism: Clinical and Experimental Research 39 (2), 380 389. http://doi.org/10.1111/acer.12631 Maschke, M., Kastrup, O., Esser, S., Ross, B., Hengge, U., & Hufnagel, A. (2000). Incidence and prevalence of neurologica l disorders associated with HIV since the introduction of highly active antiretroviral therapy (HAART). Journal of Neurology, Neurosurgery & Psychiatry 69 (3), 376 380. http://doi.org/10.1136/jnnp.69.3.376 Masur, H., Michelis, M. A., Greene, J. B., Onorato Cunningham Rundles, S. (1981). An outbreak of community acquired Pneumocystis carinii pneumonia: initial manifestation of cellular immune dysfunction. The New England Journal of Medicine 305 (24), 1431 1438. http://do i.org/10.1056/NEJM198112103052402

PAGE 144

144 Mathers, B. M., Degenhardt, L., Phillips, B., Wiessing, L., Hickman, M., Strathdee, S. A., among people who inject drugs: a systematic review. The Lancet 372 (9651), 1733 1745. http://doi.org/10.1016/S0140 6736(08)61311 2 Matochik, J. A., Eldreth, D. A., Cadet, J. L., & Bolla, K. I. (2005). Altered brain tissue composition in heavy marijuana users. Drug and Alcohol Dependence 77 (1), 23 30. http://do i.org/10.1016/j.drugalcdep.2004.06.011 McAllister, S. D., & Glass, M. (2002). CB1 and CB2 receptor mediated signalling: a focus on endocannabinoids. Prostaglandins, Leukotrienes and Essential Fatty Acids 66 (2 3), 161 171. http://doi.org/10.1054/plef.2001. 0344 McCabe, S. E., Hughes, T. L., Bostwick, W. B., West, B. T., & Boyd, C. J. (2009). Sexual orientation, substance use behaviors and substance dependence in the United States. Addiction 104 (8), 1333 1345. http://doi.org/10.1111/j.1360 0443.2009.02596.x MD, S. S. (2001). Marijuana Use in HIV Positive and AIDS Patients. Journal of Cannabis Therapeutics 1 (3 4), 35 41. http://doi.org/10.1300/J175v01n03_04 Meade, C. S., Towe, S. L., Skalski, L. M., & Robertson, K. R. (2015). Independent effects of HIV infect ion and cocaine dependence on neurocognitive impairment in a community sample living in the southern United States. Drug and Alcohol Dependence 149 128 135. http://doi.org/10.1016/j.drugalcdep.2015.01.034 Mechoulam, R., & Shvo, Y. (1963). Hashish structure of Cannabidiol. Tetrahedron 19 (12), 2073 2078. http://doi.org/10.1016/0040 4020(63)85022 X MEDINA, K. L., HANSON, K. L., SCHWEINSBURG, A. D., COHEN ZION, M., NAGEL, B. J., & TAPERT, S. F. (2007). Neuropsychological functioning in adolescent mar ijuana users: Subtle deficits detectable after a month of abstinence. Journal 13 (5), 807 820. http://doi.org/10.1017/S1355617707071032 Meier, M. H., Caspi, A., Ambler, A., Harrington, H., Houts, R., K T. E. (2012). Persistent cannabis users show neuropsychological decline from childhood to midlife. Proceedings of the National Academy of Sciences 201206820. http://doi.org/10.1073/pnas.1206820109 Meyer, I. H. (2003). Prejudice, Social Stress, and Mental Health in Lesbian, Gay, and Bisexual Populations: Conceptual Issues and Research Evidence. Psychological Bulletin 129 (5), 674 697. http://doi.org/10.1037/0033 2909.129.5.674

PAGE 145

145 Milloy, M. J., Marshall, B., Kerr, T., Richardson, L., (2015). High intensity cannabis use associated with lower plasma human immunodeficiency virus 1 RNA viral load among recently infected people who use injection drugs. Drug and Alcohol Review 34 (2), 135 140. http://doi.o rg/10.1111/dar.12223 Mimiaga, M. J., Reisner, S. L., Grasso, C., Crane, H. M., Safren, S. A., Kitahata, M. M., infected patients engaged in primary care in the United States: findings from the Centers for AIDS Research Network of Integrated Clinical Systems cohort. American Journal of Public Health 103 (8), 1457 1467. http://doi.org/10.2105/AJPH.2012.301162 (2011). Cannabin oid administration attenuates the progression of simian immunodeficiency virus. AIDS Research and Human Retroviruses 27 (6), 585 592. http://doi.org/10.1089/AID.2010.0218 Morgan, D., Mahe, C., Mayanja, B., Okongo, J. M., Lubega, R., & Whitworth, J. A. G. ( 2002). HIV 1 infection in rural Africa: is there a difference in median time to AIDS and survival compared with that in industrialized countries? AIDS (London, England) 16 (4), 597 603. Muoz Moreno, J. A., Fumaz, C. R., Ferrer, M. J., Prats, A., Negredo, E., Garolera, M., HIV infected patients. AIDS Research and Human Retroviruses 24 (10), 1301 1307. http://doi.org/10.1089/aid.2007.0310 Munro, S., Thomas, K. L., & Abu Shaar, M. (1993). Molecular characterization of a peripheral receptor for cannabinoids. Nature 365 (6441), 61 65. http://doi.org/10.1038/365061a0 Muthn, B., & Shedden, K. (1999). Finite mixture modeling with mixture outcomes using the EM algorithm. Biometrics 55 ( 2), 463 469. Nagin, D. S. (2005). Group based Modeling of Development Cambridge, MA: Harvard University Press. Nagin, D. S., & Tremblay, R. E. (2005). Developmental Trajectory Groups: Fact or a Useful Statistical Fiction?*. Criminology 43 (4), 873 904. ht tp://doi.org/10.1111/j.1745 9125.2005.00026.x Nelson, S. E., Van Ryzin, M. J., & Dishion, T. J. (2015). Alcohol, marijuana, and tobacco use trajectories from age 12 to 24 years: Demographic correlates and young adult substance use problems. Development and Psychopathology 27 (Special Issue 01), 253 277. http://doi.org/10.1017/S0954579414000650

PAGE 146

146 Newville, H., Berg, K. M., & Gonzalez, J. S. (2014). The Interaction of Active Substance Use, Depression, and Antiretroviral Adherence in Methadone Maintenance. Inter national Journal of Behavioral Medicine http://doi.org/10.1007/s12529 014 9429 z NIAAA: National Institute of Alcohol Abuse and Alcoholism. Helping patients who drink Center; 20 05. (n.d.). Retrieved March 17, 2015, from http://pubs.niaaa.nih.gov/publications/Practitioner/CliniciansGuide2005/guide.pdf Sleasman, J. W. (2014). Concordance between self reported substance use and toxicology among HIV infected and uninfected at risk youth. Drug and Alcohol Dependence 134 376 382. http://doi.org/10.1016/j.drugalcdep.2013.11.010 Nightingale, S., Winsto, A., Letendre, S., Michael, B. D., McArthur, J. C ., Khoo, S., & Solomon, T. (2014). Controversies in HIV associated neurocognitive disorders. The Lancet. Neurology 13 (11), 1139 1151. http://doi.org/10.1016/S1474 4422(14)70137 1 ordon, J. R., & Safren, S. A. (2015). The unique challenges facing HIV positive patients who smoke cigarettes: HIV viremia, ART adherence, engagement in HIV care, and concurrent substance use. AIDS and Behavior 19 (1), 178 185. http://doi.org/10.1007/s1046 1 014 0762 7 Cook, R. L. (2016). Marijuana use and viral suppression in persons receiving medical care for HIV infection. The American Journal of Drug and Alcohol Abu se 1 8. http://doi.org/10.1080/00952990.2016.1191505 Ostrow, D. G., Plankey, M. W., Cox, C., Li, X., Shoptaw, S., Jacobson, L. P., & Stall, R. C. (2009). Specific Sex Drug Combinations Contribute to the Majority of Recent HIV Seroconversions Among MSM in the MACS. Journal of Acquired Immune Deficiency Syndromes (1999) 51 (3), 349 355. http://doi.org/10.1097/QAI.0b013e3181a24b20 Pacek, L. R., Harrell, P. T., & Martins, S. S. (2014). Cigarette smoking and drug use among a nationally representative sample of HIV positive individuals. The American Journal on Addictions / American Academy of Psychiatrists in Alcoholism and Addictions 23 (6), 582 590. http://doi.org/10.1111/j.1521 0391.2014.12145.x Pacula, R. L., Hunt, P., & Boustead, A. (2014). Words Can Be Dece iving: A Review of Variation Among Legally Effective Medical Marijuana Laws in the United States. Journal of Drug Policy Analysis 7 (1), 1 19. http://doi.org/10.1515/jdpa 2014 0001

PAGE 147

147 Padula, C. B., Schweinsburg, A. D., & Tapert, S. F. (2007). Spatial working memory performance and fMRI activation interaction in abstinent adolescent marijuana users. Psychology of Addictive Behaviors: Journal of the Society of Psychologists in Addictive Behaviors 21 (4), 478 487. http://doi.org/10.1037/0893 164X.21.4.478 Pahl, K., Brook, J. S., & Koppel, J. (2011). Trajectories of Marijuana Use and Psychological Adjustment Among Urban African American and Puerto Rican Women. Psychological Medicine 41 (8), 1775 1783. http://doi.org/10.1017/S0033291710002345 Palella, F. J., Delane y, K. M., Moorman, A. C., Loveless, M. O., Fuhrer, J., Satten, G. with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. The New England Journ al of Medicine 338 (13), 853 860. http://doi.org/10.1056/NEJM199803263381301 Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV 1 Infected Adults and Adolescent. Department of Health and Hu man Services. (2016). Retrieved from https://aidsinfo.nih.gov/contentfiles/lvguidelines/adultandadolescentgl.pdf Pardini, D., Bechtold, J., Loeber, R., & White, H. (2015). Developmental Trajectories of Marijuana Use among Men Examining Linkages with Crimin al Behavior and Psychopathic Features into the Mid 30s. Journal of Research in Crime and Delinquency 52 (6), 797 828. http://doi.org/10.1177/0022427815589816 Parsons, T. D., Braaten, A. J., Hall, C. D., & Robertson, K. R. (2006). Better quality of life wit h neuropsychological improvement on HAART. Health and Quality of Life Outcomes 4 11. http://doi.org/10.1186/1477 7525 4 11 Passarotti, A. M., Crane, N. A., Hedeker, D., & Mermelstein, R. J. (2015). Longitudinal trajectories of marijuana use from adolesce nce to young adulthood. Addictive Behaviors 45 301 308. http://doi.org/10.1016/j.addbeh.2015.02.008 Pence, B. W., Thielman, N. M., Whetten, K., Ostermann, J., Kumar, V., & Mugavero, M. J. (2008). Coping strategies and patterns of alcohol and drug use amo ng HIV infected patients in the United States Southeast. AIDS Patient Care and STDs 22 (11), 869 877. http://doi.org/10.1089/apc.2008.0022 Peretti Watel, P., Spire, B., Lert, F., & Obadia, Y. (2006). Drug use patterns and adherence to treatment among HIV p ositive patients: evidence from a large sample of French outpatients (ANRS EN12 VESPA 2003). Drug and Alcohol Dependence 82 Supplement 1 S71 S79. http://doi.org/10.1016/S0376 8716(06)80012 8 Pertwee, R. G. (1997). Pharmacology of cannabinoid CB1 and CB2 receptors. Pharmacology & Therapeutics 74 (2), 129 180.

PAGE 148

148 Plankey, M. W., Ostrow, D. G., Stall, R., Cox, C., Li, X., Peck, J. A., & Jacobson, L. P. (2007). The relationship between methamphetamine and popper use and risk of HIV seroconversion in the multice nter AIDS cohort study. Journal of Acquired Immune Deficiency Syndromes (1999) 45 (1), 85 92. http://doi.org/10.1097/QAI.0b013e3180417c99 Pope, H. G., Jr, Gruber, A. J., Hudson, J. I., Huestis, M. A., & Yurgelun Todd, D. (2001). Neuropsychological performa nce in long term cannabis users. Archives of General Psychiatry 58 (10), 909 915. Pope, H. G., Jr, Gruber, A. J., Hudson, J. I., Huestis, M. A., & Yurgelun Todd, D. (2002). Cognitive measures in long term cannabis users. Journal of Clinical Pharmacology 4 2 (11 Suppl), 41S 47S. Prentiss, D., Power, R., Balmas, G., Tzuang, G., & Israelski, D. M. (2004). Patterns of marijuana use among patients with HIV/AIDS followed in a public health care setting. Journal of Acquired Immune Deficiency Syndromes (1999) 35 (1) 38 45. ProCon.org. (2015a). 9 States Considering Medical Marijuana Legalization. Retrieved from http://medicalmarijuana.procon.org/view.resource.php?resourceID=002481 Radloff, L. S. (1977). The CES D Scale A Self Report Depression Scale for Research in t he General Population. Applied Psychological Measurement 1 (3), 385 401. http://doi.org/10.1177/014662167700100306 Raftery, A. (n.d.). Bayesian Model Selection in Social Research, 111 63. Ramaekers, J. G., Kauert, G., Theunissen, E. L., Toennes, S. W., & M oeller, M. R. (2009). Neurocognitive performance during acute THC intoxication in heavy and occasional cannabis users. Journal of Psychopharmacology (Oxford, England) 23 (3), 266 277. http://doi.org/10.1177/0269881108092393 Ramaekers, J. G., Theunissen, E. L., de Brouwer, M., Toennes, S. W., Moeller, M. R., & Kauert, G. (2011). Tolerance and cross tolerance to neurocognitive effects of THC and alcohol in heavy cannabis users. Psychopharmacology 214 (2), 391 401. http://doi.org/10.1007/s00213 010 2042 1 Ram rez, B. G., Blzquez, C., Gmez del Pulgar, T., Guzmn, M., & de Ceballos, M. L. neuroprotection mediated by blockade of microglial activation. The Journal of Neuroscience: The Official J ournal of the Society for Neuroscience 25 (8), 1904 1913. http://doi.org/10.1523/JNEUROSCI.4540 04.2005 memory in humans: a review. Psychopharmacology 188 (4), 425 444. http://d oi.org/10.1007/s00213 006 0508 y Reitan, R. M. (1992). Trail Making Test Reitan Neuropsychology Laboratory.

PAGE 149

149 Rippeth, J. D., Heaton, R. K., Carey, C. L., Marcotte, T. D., Moore, D. J., Gonzalez, R., s risk of neuropsychological impairment in HIV infected persons. Journal of the International Neuropsychological Society: JINS 10 (1), 1 14. http://doi.org/10.1017/S1355617704101021 Roberts, G. M. P., & Garavan, H. (2010). Evidence of increased activation underlying cognitive control in ecstasy and cannabis users. NeuroImage 52 (2), 429 435. http://doi.org/10.1016/j.neuroimage.2010.04.192 Rom, S., & Persidsky, Y. (2013). Cannabinoid receptor 2: potential role in immunomodulation and neuroinflammation. Journ al of Neuroimmune Pharmacology: The Official Journal of the Society on NeuroImmune Pharmacology 8 (3), 608 620. http://doi.org/10.1007/s11481 013 9445 9 Liu, H. (2013). Association between use of specific drugs and antiretroviral adherence: findings from MACH 14. AIDS and Behavior 17 (1), 142 147. http://doi.org/10.1007/s10461 011 0124 7 Rothlind, J. C., Greenfield, T. M., Bruce, A. V., Meyerhoff, D. J., Flenniken, D. L. Lindgren, J. A., & Weiner, M. W. (2005). Heavy alcohol consumption in individuals with HIV infection: effects on neuropsychological performance. Journal of the International Neuropsychological Society: JINS 11 (1), 70 83. http://doi.org/10.1017/S13556177 05050095 E. (2016). Prevalence of HIV associated neurocognitive disorders in the Multicenter AIDS Cohort Study. Neurology 86 (4), 334 340. http://doi.org/10.1212/WNL .0000000000002277 North American AIDS Cohort Collaboration on Research and Design (NA ACCORD) of IeDEA. (2013). Closing the Gap: Increases in Life Expectancy among Tre ated HIV Positive Individuals in the United States and Canada. PLoS ONE 8 (12), e81355. http://doi.org/10.1371/journal.pone.0081355 Snchez, A. J., & Garca Merino, A. (2012). Neuroprotective agents: cannabinoids. Clinical Immunology (Orlando, Fla.) 142 (1 ), 57 67. http://doi.org/10.1016/j.clim.2011.02.010 (2014). HIV associated neurocognitive disorders. Journal of Molecular Psychiatry 2 (1), 2. http://doi.org/10.1186/204 9 9256 2 2 Satz, P. (1993). Brain reserve capacity on symptom onset after brain injury: A formulation and review of evidence for threshold theory. Neuropsychology 7 (3), 273 295. http://doi.org/10.1037/0894 4105.7.3.273

PAGE 150

150 Scholes, K. E., & Martin Iverson, M. T. (2010). Cannabis use and neuropsychological performance in healthy individuals and patients with schizophrenia. Psychological Medicine 40 (10), 1635 1646. http://doi.org/10.1017/S0033291709992078 Scholes Balog, K. E., & Martin Iverson, M. T. (2011). Ca nnabis use and sensorimotor gating in patients with schizophrenia and healthy controls. Human Psychopharmacology 26 (6), 373 385. http://doi.org/10.1002/hup.1217 Laetz, V. B. (2005). Trajectories of Marijuana use during the Transition to Adulthood: The Big Picture Based on National Panel Data. Journal of Drug Issues 35 (2), 255 280. http://doi.org/10.1177/002204260503500203 Schweinsburg, A. D., Schweinsburg, B. C., Medin a, K. L., McQueeny, T., Brown, S. A., & Tapert, S. F. (2010). The influence of recency of use on fMRI response during spatial working memory in adolescent marijuana users. Journal of Psychoactive Drugs 42 (3), 401 412. Shoptaw, S., Stall, R., Bordon, J., K Cumulative exposure to stimulants and immune function outcomes among HIV positive and HIV negative men in the Multicenter AIDS Cohort Study. International Journal of STD & AIDS 23 (8), 576 580. http://doi.o rg/10.1258/ijsa.2012.011322 Simioni, S., Cavassini, M., Annoni, J. M., Rimbault Abraham, A., Bourquin, I., Schiffer, long standing suppression of viremia. AIDS (London, England) 24 (9), 1243 1250. http://doi.org/10.1097/QAD.0b013e3283354a7b (2012). Racial/Ethnic Disparities in ART Adherence in the United States: Findings From the MACH14 Study. Journal of Acquired Immune Deficiency Syndromes (1999) 60 (5), 466 472. http://doi.org/10.1097/QAI.0b013e31825db0bd (2011). Patterns of Substance Use Among a Large Ur ban Cohort of HIV Infected Men Who Have Sex With Men in Primary Care. AIDS and Behavior 16 (3), 676 689. http://doi.org/10.1007/s10461 011 9880 7 Slawson, G., Milloy, M. (2014). High Intensity Cannabis Use and Adherence to Antiretroviral Therapy Among People Who Use Illicit Drugs in a Canadian Setting. AIDS and Behavior 1 8. http://doi.org/10.1007/s10461 014 0847 3 Smith, A. (1982). Symbol Digit Modalities Test Western Psychological Service;

PAGE 151

151 Sneider, J. T., Pope, H. G., Silveri, M. M., Simpson, N. S., Gruber, S. A., & Yurgelun Todd, D. A. (2006). Altered regional blood volume in chronic cannabis smokers. Experimental and Clinical Psychopharmacology 14 (4), 422 428. http://doi.org/10.1037/1064 1297.14.4.422 Snider, W. D., Simpson, D. M., Nielsen, S., Gold, J. W., Metroka, C. E., & Posner, J. B. (1983). Neurological complications of acquired immune deficiency syndrome: analysis of 50 patients. Annals of Neurology 14 (4), 403 418. http://doi.org/1 0.1002/ana.410140404 Solowij, N., & Pesa, N. (2010). [Cognitive abnormalities and cannabis use]. Revista brasileira de psiquiatria (So Paulo, Brazil: 1999) 32 Suppl 1 S31 40. Solowij, N., Stephens, R. S., Roffman, R. A., Babor, T., Kadden, R., Miller, M Marijuana Treatment Project Research Group. (2002). Cognitive functioning of long term heavy cannabis users seeking treatment. JAMA 287 (9), 1123 1131. (2003) Association of Co Occurring Psychosocial Health Problems and Increased Vulnerability to HIV/AIDS Among Urban Men Who Have Sex With Men. American Journal of Public Health 93 (6), 939 942. Stall, R., Akhtar Khaleel, W. Z., Cook, R. L., Shoptaw, S., Surkan, P., Beyth, R. J., Plankey, M. (2015). Trends and Predictors of Cigarette Smoking Among HIV Seropositive and Seronegative Men: The Multicenter Aids Cohort Study. AIDS and Behavior http://doi.org/10.1007/s10461 015 1099 6 Substance Abuse and Mental Health Services Administration, Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings, NSDUH Series H 46, HHS Publication No. (SMA) 13 4795. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2013. (n.d. ). Substance Abuse and Mental Health Services Administration, Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings, (2014). (NSDUH Series H 48, HHS). Rockville, MD: Substance Abuse and Mental Health Services Administr ation, 2014. Tait, R. J., Mackinnon, A., & Christensen, H. (2011). Cannabis use and cognitive function: 8 year trajectory in a young adult cohort. Addiction (Abingdon, England) 106 (12), 2195 2203. http://doi.org/10.1111/j.1360 0443.2011.03574.x Tapert, S. F., Schweinsburg, A. D., Drummond, S. P. A., Paulus, M. P., Brown, S. A., Yang, T. T., & Frank, L. R. (2007). Functional MRI of inhibitory processing in abstinent adolescent marijuana users. Psychopharmacology 194 (2), 173 183. http://doi.org/10.1007/s002 13 007 0823 y

PAGE 152

152 Tarter, R. E., Horner, M., & Ridenour, T. (2012). Developmental perspective of substance use disorder etiology. In H. J. Shaffer, D. A. LaPlante, & S. E. Nelson (Eds.), APA addiction syndrome handbook, Vol. 1: Foundations, influences, and exp ressions of addiction (pp. 261 287). Washington, DC, US: American Psychological Association. Thames, A. D., Arentoft, A., Rivera Mindt, M., & Hinkin, C. H. (2013). Functional disability in medication management and driving among individuals with HIV: A 1 y ear follow up study. Journal of Clinical and Experimental Neuropsychology 35 (1), 49 58. http://doi.org/10.1080/13803395.2012.747596 Thames, A. D., Mahmood, Z., Burggren, A. C., Karimian, A., & Kuhn, T. P. (2015). Combined effects of HIV and marijuana use on neurocognitive functioning and immune status. AIDS Care 1 5. http://doi.org/10.1080/09540121.2015.1124983 Toutenburg, H. (1990). Rubin, D.B.: Multiple imputation for nonresponse in surveys. Statistical Papers 31 (1), 180 180. http://doi.org/10.1007/BF0 2924688 Tucker, J. S., Burnam, M. A., Sherbourne, C. D., Kung, F. Y., & Gifford, A. L. (2003). Substance use and mental health correlates of nonadherence to antiretroviral medications in a sample of patients with human immunodeficiency virus infection. The American Journal of Medicine 114 (7), 573 580. http://doi.org/10.1016/S0002 9343(03)00093 7 Vadhan, N. P., Hart, C. L., van Gorp, W. G., Gunderson, E. W., Haney, M., & Foltin, R. W. (2007). Acute effects of smoked marijuana on decision making, as assessed by a modified gambling task, in experienced marijuana users. Journal of Clinical and Experimental Neuropsychology 29 (4), 357 364. http://doi.org/10.1080/13803390600693615 van Buuren, S., Boshuizen, H. C., & Knook, D. L. (1999). Multiple imputation of mis sing blood pressure covariates in survival analysis. Statistics in Medicine 18 (6), 681 694. van der Pol, P., Liebregts, N., de Graaf, R., Korf, D. J., van den Brink, W., & van Laar, M. (2013). Validation of self reported cannabis dose and potency: an ecol ogical study. Addiction 108 (10), 1801 1808. http://doi.org/10.1111/add.12226 Vanyukov, M. M., Tarter, R. E., Kirisci, L., Kirillova, G. P., Maher, B. S., & Clark, D. B. (2003). Liability to substance use disorders: 1. Common mechanisms and manifestations Neuroscience and Biobehavioral Reviews 27 (6), 507 515. Verdejo Garca, A., Rivas Prez, C., Lpez Torrecillas, F., & Prez Garca, M. (2006). Differential impact of severity of drug use on frontal behavioral symptoms. Addictive Behaviors 31 (8), 1373 13 82. http://doi.org/10.1016/j.addbeh.2005.11.003

PAGE 153

153 (2005). Differential effects of MDMA, cocaine, and cannabis use severity on distinctive components of the executive fun ctions in polysubstance users: A multiple regression analysis. Addictive Behaviors 30 (1), 89 101. http://doi.org/10.1016/j.addbeh.2004.04.015 Viswanathan, S., Detels, R., Mehta, S. H., Macatangay, B. J. C., Kirk, G. D., & Jacobson, L. P. (2014). Level of Adherence and HIV RNA Suppression in the Current Era of Highly Active Antiretroviral Therapy (HAART). AIDS and Behavior 19 (4), 601 611. http://doi.org/10.1007/s10461 014 0927 4 Wadsworth, E. J. K., Moss, S. C., Simpson, S. A., & Smith, A. P. (2006). Canna bis use, cognitive performance and mood in a sample of workers. Journal of Psychopharmacology (Oxford, England) 20 (1), 14 23. http://doi.org/10.1177/0269881105056644 Walden, N., & Earleywine, M. (2008). How high: Quantity as a predictor of cannabis relat ed problems. Harm Reduction Journal 5 20. http://doi.org/10.1186/1477 7517 5 20 Ware, M. A., Rueda, S., Singer, J., & Kilby, D. (2003). Cannabis Use by Persons Living with HIV/AIDS: Patterns and Prevalence of Use. Journal of Cannabis Therapeutics 3 (2), 3 15. http://doi.org/10.1300/J175v03n02_02 Sapir, E. (2008). Brain imaging study of the acute effects of Delta9 tetrahydrocannabinol (THC) on attention and motor coordin ation in regular users of marijuana. Psychopharmacology 196 (1), 119 131. http://doi.org/10.1007/s00213 007 0940 7 Wen, H., Hockenberry, J. M., & Cummings, J. R. (2015). The effect of medical marijuana laws on adolescent and adult use of marijuana, alcohol and other substances. Journal of Health Economics 42 64 80. http://doi.org/10.1016/j.jhealeco.2015.03.007 Westlake, T. M., Howlett, A. C., Bonner, T. I., Matsuda, L. A., & Herkenham, M. (1994). Cannabinoid receptor binding and messenger RNA expression in human brain: an in vitro receptor autoradiography and in situ hybridization histochemistry study Neuroscience 63 (3), 637 652. Weuve, J., Tchetgen Tchetgen, E. J., Glymour, M. M., Beck, T. L., Aggarwal, N. T., Wils selective attrition: the example of smoking and cognitive decline. Epidemiology (Cambridge, Mass.) 23 (1), 119 128. http://doi.org/10.1097/EDE.0b013e318230e861

PAGE 154

154 Whitesell, N. R., Asdigian N. L., Kaufman, C. E., Crow, C. B., Shangreau, C., Keane, E. American Indian Adolescents: Patterns and Predictors. Journal of Youth and Adolescence 43 (3), 437 453. http://doi.org/10 .1007/s10964 013 0026 2 Whitlow, C. T., Liguori, A., Livengood, L. B., Hart, S. L., Mussat Whitlow, B. J., term heavy marijuana users make costly decisions on a gambling task. Drug and Alcohol Dependence 76 (1) 107 111. http://doi.org/10.1016/j.drugalcdep.2004.04.009 WHO. (2013). GLOBAL AIDS RESPONSE PROGRESS REPORTING 2013 Retrieved from http://www.who.int/hiv/pub/me/Global_AIDS_Response_Progress_Reporting_20 13_Guidelines_En.pdf Wilkinson, S. T., Yarnell, S., Marijuana Legalization: Impact on Physicians and Public Health. Annual Review of Medicine 67 453 466. http://doi.org/10.1146/annurev med 050214 013454 Williams, J. C., Appelberg, S., Goldberger, B A., Klein, T. W., Sleasman, J. W., & Tetrahydrocannabinol treatment during human monocyte differentiation reduces macrophage susceptibility to HIV 1 infection. Journal of Neuroimmune Pharmacology: The Official Journal of the Society on NeuroImmune Pharmacology 9 (3), 369 379. http://doi.org/10.1007/s11481 014 9527 3 Wilson, K. J., Doxanakis, A., & Fairley, C. K. (2004). Predictors for non adherence to antiretroviral therapy. Sexual Health 1 (4), 251 257. Windle, M., & Wiesner, M. (2004). Trajectories of marijuana use from adolescence to young adulthood: Predictors and outcomes. Development and Psychopathology 16 (4), 1007 1027. http://doi.org/10.1017/S0954579404040118 Woolridge, E., Barton, S., Samuel, J., Osorio, J., Dougherty A., & Holdcroft, A. (2005). Cannabis use in HIV for pain and other medical symptoms. Journal of Pain and Symptom Management 29 (4), 358 367. http://doi.org/10.1016/j.jpainsymman.2004.07.011 World Health Organization. (2015). Guideline on When to Start An tiretroviral Therapy And On Pre Exposure Prophylaxis For HIV Retrieved from http://apps.who.int/iris/bitstream/10665/186275/1/9789241509565_eng.pdf Yehia BR, Fleishman JA, Metlay JP, Moore RD, & Gebo KA. (2012). SUstained viral suppression in hiv infected patients receiving antiretroviral therapy. JAMA 308 (4), 339 342. http://doi.org/10.1001/jama.2012.5927

PAGE 155

155 Young, S., Wood, E., Dong, H., Kerr, T., & Hayashi, K. (2015). Daily alcohol Use as an independent risk factor for HIV seroconversion among people Who inject drugs. Addiction (Abingdon, England) http://doi.org/10.1111/add.13256 Ycel, M., Solowij, N., Respondek, C., Whittle, S., Fornito, A., Pantelis, C., & Lubman, D. I. (2008). Regional brain abnormalities associated with long term heavy cannabis use. Archives of General Psychiatry 65 (6), 694 701. http://doi.org/10.1001/archpsyc.65.6.694 related problems. Addiction Research & Theory 20 (1), 82 92. http://doi.org/10.3109/16066359.2011.569101 Zou, G. (2004). A Modified Poisson Regression Approach to Prospective Studies with Binary Data. American Journal of Epidemiology 159 (7), 702 706. http://doi .org/10.1093/aje/kwh090

PAGE 156

156 BIOGRAPHICAL SKETCH Chukwuemeka Okafor received his Bachelor of Science in Biochemistry from the University of Lagos, Nigeria in 2005. He received his Master of Public Health in Community Health from the University of North Florida, Jacksonville in 2012. He received his PhD in epidemiology from the University of Florida in the summer of 2016. Chukwuemeka Okafor research interests focuses on understanding the clinical and neurobehavioral ef fects of substances of abuse among persons living with or at risk for HIV infection.