Citation
Health Outcomes and Pediatric Asthma: A Longitudinal Study

Material Information

Title:
Health Outcomes and Pediatric Asthma: A Longitudinal Study
Creator:
Li, Zheng
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:
SHENKMAN,ELIZABETH ANN
Committee Co-Chair:
HUANG,I-CHAN
Committee Members:
PRINS,CINDY A
LEITE,WALTER LANA
THOMPSON,LINDSAY ACHESON
Graduation Date:
12/19/2014

Subjects

Subjects / Keywords:
Air pollution ( jstor )
Asthma ( jstor )
Child psychology ( jstor )
Children ( jstor )
Chronic conditions ( jstor )
Demography ( jstor )
Parents ( jstor )
Pediatrics ( jstor )
Pollen ( jstor )
Symptomatology ( jstor )
asthma
children
hrqol

Notes

General Note:
Asthma prevalence and asthma-related healthcare utilization in children continuously increased in the U.S. However, limited studies have used the longitudinal design to examine the natural progression of health outcomes in asthmatic children including symptoms, asthma control status, and health-related quality of life (HRQOL). This dissertation aims to use longitudinal data collected from the NIH Patient-Reported Outcomes Measurement Information System (PROMIS) Pediatric Asthma Study (PAS) to 1) examine trajectories of HRQOL and identify factors associated with variations in initial status and rate of change in HRQOL, 2) investigate the influence of nighttime sleep quality and daytime sleepiness on the association of asthma control status with HRQOL, and 3) investigate effects of ambient air pollution and pollen exposure on asthma control status. This dissertation is a secondary data analysis of the NIH PROMIS PAS. First, latent growth models were applied to test longitudinal associations between asthma control status and HRQOL. Second, multilevel structural equation modeling was used to test the extent to which different domains of sleep problems (nighttime sleep quality and daytime sleepiness) mediate the relationship of asthma control status with HRQOL. Third, linear mixed effect models were performed to examine associations of asthma control status with air pollution and pollen exposure. Aim 1 shows there was a significant improvement in asthma-specific HRQOL over 2 years; however, this improved trend was explained by the change of asthma control and baseline subject's characteristics. Aim 2 shows the change of asthma control status was directly associated with changes of HRQOL at both between-subject and within-subject levels; however, the indirect effect of asthma control status on asthma-specific HRQOL was through the influence of daytime sleepiness at within-subject level, and through the mechanism of nighttime sleep quality and daytime sleepiness at between-subject level. Aim 3 shows increased exposure to PM2.5 and pollen was significantly associated with worse asthma control status over time. Overall, asthma control status was significantly associated with changes in HRQOL among asthmatic children. The problems in nighttime sleep and daytime sleepiness mediate the effect of asthma control on HRQOL. Air pollution and pollen was found to have significant impact on asthma control.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
12/31/2016

Downloads

This item has the following downloads:


Full Text

PAGE 1

HEALTH OUTCOMES AND PEDIATRIC ASTHMA: A LONGITUDINAL STUDY By ZHENG LI 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 2014

PAGE 2

2014 Zheng Li

PAGE 3

To my family: Mo m Dad and my husband Wei

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisors, Dr. I-Chan Huang and Dr. Elizabeth Shenkman for their incredible support and guidance in the past four years throughout my PhD program. Dr. I-Chan Huang has been supportive and provided scientific advice, knowledge and many insightful discussions to me when conducting this dissertation. He is my primary resource for getting my scientific questions answered and I appreciate all his contributions of time, patience, knowledge and ideas on my research during my PhD training. I am also very grateful to Dr. Elizabeth Shenkman for her continuous support and encouragement during my PhD study and research. I would also like to thank my committee members, Dr. Cindy Prins, Dr. Walter Leite, and Dr. Lindsay Thompson for their thoughtful feedback and constructive comments during the dissertation process As my academic advisor, Dr. Cindy Prins has been helpful in providing general advice about academic issues and progress in the past years. Dr. Walter Leite provided constructive suggestions and insightful discussions on methodology issues that helped me address statistical challenges in my dissertation. Dr. Lindsay Thompson provided valuable options on clinical application of patient-reported outcomes among children with asthma. I would like to give special thanks to Dr. I-Chan Huang who allows me to use the dataset collected from his NIH PROMIS Pediatric Asthma Study for my dissertation. Lastly, I would like to thank my parents and husband for their support and love. My parents have been supportive and provided unconditional love and care; I would not have made this far without them. I would especially like to thank my loving, encouraging and patient husband for his faithful support and sacrifices he made for our family.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES ........................................................................................................ 10 LIST OF ABBREVIATIONS ........................................................................................... 11 ABSTRACT ................................................................................................................... 12 CHAPTER 1 INTRODUCTION TO PEDIATRIC ASTHMA AND HEALTH OUTCOMES ............. 14 Epidemiology of Pediatric Asthma .......................................................................... 14 Health Outcomes of Pediatric Asthma .................................................................... 15 Asthma Control and Quality of Life ................................................................... 15 Mediators Influencing the Asthma Control-HRQOL Pathway ........................... 19 Factors Associated with Asthma Outcomes ..................................................... 21 Specific Aims of Dissertation .................................................................................. 24 Specific Aim 1: Examine the Association between Asthma Control and HRQOL ......................................................................................................... 25 Specific Aim 2: Test a Conceptual Framework of Sleep Factors Influencing HRQOL of Asthmatic Children ...................................................................... 26 Specific Aim 3: Investigate the Impact of Ambient Air Pollution and Pollen Exposure on Asthma Control Status ............................................................. 26 Study Design .......................................................................................................... 27 Participants and Data Collection ...................................................................... 27 Measures and Instruments ............................................................................... 29 2 THE RELATIONSHIP BETWEEN ASTHMA CONTROL AND HEALTHRELATED QUALITY OF LIFE AMONG CHILDREN WITH ASTHMA ..................... 37 Introduction ............................................................................................................. 37 Methods .................................................................................................................. 39 Participants and Data Collection ...................................................................... 39 Study Measures ............................................................................................... 40 Statistical Analysis ............................................................................................ 41 Results .................................................................................................................... 44 Participant Characteristics ................................................................................ 44 s Characteristics .............................................................................................. 45 Characteristics .............................................................................................. 45

PAGE 6

6 Longitudinal Associations between Asthma Control and HRQOL .................... 46 Discussion .............................................................................................................. 48 Limitations ............................................................................................................... 52 Conclusion .............................................................................................................. 53 3 LONGITUDINAL ASSOATONS AMONG ASTHMA CONTROL, SLEEP PROBLEMS AND HEALTH-RELATED QUALITY OF LIFE IN CHILDREN WITH ASTHMA ................................................................................................................. 64 Introduction ............................................................................................................. 64 Methods .................................................................................................................. 66 Participants and Data Collection ...................................................................... 66 Study Measures ............................................................................................... 67 Statistical Analysis ............................................................................................ 69 Results .................................................................................................................... 72 Participant Characteristics ................................................................................ 72 Descriptive Analyses for Study Measures ........................................................ 72 Bivariate Associations of Asthma Control with Sleep Problems, HRQOL and Socio Demographics ..................................................................................... 73 Multivariate Associations among Asthma Control, Sleep Problems and HRQOL ......................................................................................................... 73 Path Analysis Results ....................................................................................... 74 Discussion .............................................................................................................. 76 Limitations ............................................................................................................... 80 Conclusion .............................................................................................................. 81 4 THE INFLUENCE OF AMBIT AIR POLLUTION AND POLLEN EXPOSURE ON ASTHMA CONTROL STATUS AMONG ASTHMATIC CHILDREN .................. 88 Introduction ............................................................................................................. 88 Methods .................................................................................................................. 91 Participants and Data Collection ...................................................................... 91 Study Measures ............................................................................................... 91 Statistical Analysis ............................................................................................ 93 Results .................................................................................................................... 95 Study Participant Characteristics ...................................................................... 95 Descriptive Analyses for Study Measures ........................................................ 96 Multivariate Associations among Air Pollution Concentration, Pollen Severity Index and Asthma Control Scores ................................................... 96 Sensitivity Analysis Results .............................................................................. 99 Discussion .............................................................................................................. 99 Limitations ............................................................................................................. 103 Conclusion ............................................................................................................ 104 5 IMPROVING LONG-TERM HEALTH OUTCOMES IN CHILDREN WITH ASTHMA: DISCUSSION OF RESULTS ............................................................... 113

PAGE 7

7 Review of Study Findings ..................................................................................... 113 Limitations ............................................................................................................. 115 Conclusion ............................................................................................................ 116 APPENDIX A CHAPTER 2 INTER-CORRELATION OF ASTHMA CONTROL STATUS AND HRQOL ................................................................................................................. 117 B CHAPTER 3 DETAILED METHODS OF MULTILEVEL STRUCTURAL EQUATION MODELING ....................................................................................... 118 Multilevel Mediation Models .................................................................................. 118 Mplus Sample Code for MSEM ...................................................................... 118 C CHAPTER 4 DETAILED RESULTS FOR AIR POLLUTION AND POLLEN ......... 119 LIST OF REFERENCES ............................................................................................. 123 BIOGRAPHICAL SKETCH .......................................................................................... 139

PAGE 8

8 LIST OF TABLES Table page 1 1 Literature review of longitudinal studies on asthma control and quality of life (QOL) among asthmatics ................................ ................................ ................... 31 2 1 Participant characteristics ................................ ................................ ................... 57 2 2 Bivariate analyses for asthma control associated with socio demographics and asthma specific HRQOL ................................ ................................ .............. 58 2 3 Multivariate linear regression analyses for asthma specific HRQOL associated with asthma control and socio demographic characteristics ............. 59 2 4 Change of asthma specific HRQOL based on unconditional latent growth model (Model 1) and conditional latent growth models (Model 2 &3 ) .................. 60 2 5 The associations of changes in asthma specific HRQOL with asthma control and socio demographics ................................ ................................ ..................... 61 3 1 Study participant characteristics ................................ ................................ ......... 83 3 2 Mean and standard deviation of the asthma control, n ighttime sleep quality, daytime sleepiness and asthma specific HRQOL ................................ ............... 84 3 3 Bivariate association for asthma control with nighttim e sleep quality, daytime sleepiness, asthma specific HRQOL and socio demographics .......................... 85 3 4 Parameter estimate of random intercept models ................................ ................ 86 3 5 Multilevel path analyses for the mediating effects of nighttime sleep quality and daytime sleepiness on the asthma c ontrol HRQOL pathway ....................... 87 4 1 Study participant characteristics ................................ ................................ ....... 106 4 2 Summary statistics of air pollution concentrations, pollen severity index and asthma control scores between September 2010 and May 2012 in Florida ..... 107 4 3 Associations between asthma control summation scores and preceding 7 days average PM 2.5 concentrations ................................ ................................ .. 108 4 4 Associations between asthma control summation scores and preceding 7 days average O 3 concentrations ................................ ................................ ....... 109 4 5 Associations between asthma control summation scores and preceding 7 days average pollen severity index ................................ ................................ ... 110

PAGE 9

9 4 6 Associations between asthma control summation scores and pr eceding 7 days average air pollution and pollen exposure ................................ ................ 111 4 7 Sensitivity analyses for associations between asthma control summation scores and air pollution and pollen exposure ................................ ................... 112 A 1 Inter correlation of asthma control status and asthma specific HRQOL across 4 time points ................................ ................................ ................................ ..... 117 C 1 Number of subjects and observations at different ranges of distance between ................................ .............. 119 C 2 PM 2.5 and O 3 concentrations at EPA air quality monitors and pollen severity index in 8 areas in Florida at 26 weeks ................................ ............................ 120 C 3 2.5 O 3 and pollen at 26 weeks during the study period ................................ ................................ ..................... 121 C 4 Correlations among asthma control summation scores, PM 2.5 O 3 and pollen concentrations at 26 weeks ................................ ................................ .............. 122

PAGE 10

10 LIST OF FIGURES Figure page 1 1 Flow chart demonstrating recruitment of participants in the PROMIS PAS.. ...... 36 2 1 The unconditional latent growth model (LGM) examining trajectories of asthma specific HRQOL across 4 time points ................................ .................... 54 2 2 The conditional latent growth model (LGM) examining the influence of asthma control status on changes of HRQOL ................................ .................... 55 2 3 The final conditional latent growth model (LGM) examining the effect of asthma control and socio demographics on changes of HRQOL over time ....... 56 3 1 Conceptual framework depicting within person and between person relationships using multilevel structural equation modeling ................................ 82 4 1 Locations of study participants and air pollution monitors in Florida, USA ....... 105

PAGE 11

11 LIST OF ABBREVIATIONS ACCI Asthma Control and Communication Instrument AQS Air Quality System EPA Environmental Protection Agency HRQOL Health related quality of life IRT Item response theory LGM Latent growth model NIH National Institutes of Health PAS Pediatric Asthma Study PDSS Pediatric Daytime Sleepiness Scale PROs Patient reported outcomes PROMIS Patient Reported Outcomes Measurement Information System SCHIP Insurance Program SEM Structural equation modeling

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HEALTH OUTCOMES AND PEDIATRIC ASTHMA: A LONGITUDINAL STUDY By Zheng Li December 2014 Chair: Elizabeth Shenkman Major: Epidemiology Asthma prevalence and asthma related healthcare utilization in children continuously increased in the U.S. However, limited studies have used the longitudinal design to examine the natural progression of health outcomes in asthmatic children including symp toms, asthma control status, and health related quality of life (HRQOL). This dissertation aims to use longitudinal data collected from the NIH Patient Reported Outcomes Measurement Information System (PROMIS) Pediatric Asthma Study (PAS) to 1) examine tra jectories of HRQOL and identify factors associated with variations in initial status and rate of change in HRQOL, 2) investigate the influence of nighttime sleep quality and daytime sleepiness on the association of asthma control status with HRQOL, and 3) examine effects of ambient air pollution and pollen exposure on asthma control status. This dissertation is a secondary data analysis of the NIH PROMIS PAS. First, latent growth models were applied to test longitudinal associations between as thma control s tatus and HRQOL. Second, multilevel structural equation modeling was used to test the extent to which different domains of sleep problems (nighttime sleep quality and daytime sleepiness) mediate the relationship of asthma control status with HRQOL.

PAGE 13

13 Third, linear mixed effect models were performed to examine associations of asthma control status with air pollution and pollen exposure. Aim 1 shows there was a significant improvement in asthma specific HRQOL over 2 years; however, this improved trend was expla ined by the change of asthma control status was directly associated with changes of HRQOL at both between subject and within subject levels; however, the indirect effect of as thma control status on asthma specific HRQOL was through the influence of daytime sleepiness at within subject level, and through the mechanism of nighttime sleep quality and daytime sleepiness at between subject level. Aim 3 shows increased exposure to PM 2.5 and pollen was significantly associated with worse asthma control status over time. Overall, asthma control status was significantly associated with changes in HRQOL among asthmatic children. The problems in nighttime sleep and daytime sleepiness medi ate the effect of asthma control on HRQOL. Air pollution and pollen was found to have signif icant impact on asthma control.

PAGE 14

14 CHAPTER 1 INTRODUC TION TO PEDIATRIC ASTHMA AND HEALTH OUTCOMES Epidemiology of Pediatric Asthma Asthma is a common chronic lung disorder characterized by bronchial hyper responsiveness airway inflammation and airflow limitation. When people with asthma are exposed to triggers and irritants, asthma will inflame and narrow the airway which lead s to recurring episodes of symptoms such as wheezing, coughing, chest tightness and short ness of breath. The overall asthma prevalence continuously increased from 7.3% to 8.4% during the period of 2001 2010, which affect ed approximately 25.7 million adults and children in the United States [1, 2] The 2012 National Health Interview Survey (NHIS) showed children and adolescents had higher asthma prevalence than adult population and there were appr oximately 6.8 million (9%) children and adolescents under age 18 years had asthma in 2012 [2] In the past ten years, although asthma related health care visits in primary care settings have declined, asthma related eme rgency department (ED) visit s slightly increased and hospitalization remained the same [1] Children and adults had similar asthma related hospitalization rate s during the period of 2001 2009, however, asthmatic ch ildren had higher rate s of asthma related emergency department (ED) visits and asthma related primary care visits than adults [1] The recent Medical Expenditure Panel Surveys (MEPS) from 2000 through 2009 showed t he expenditure of asthma in adolescents increased 2.5% each year, and it was estimated that total expenditures of asthma was 63 billion in the U .S. in 2009 [3] According to the Asthma and Allergy Foundation of America, asthma is the leading chronic disease that causes nearly 14 million school absenteeism annually [4]

PAGE 15

15 Evidence showed approximately 2.6 million children in Florida had asthma in 2012 [5] The prevalence of pediatric asthma in Florida adolescents has continuously increased from 18% in 2005 to 22% in 20 11 and this trend is consistent with the prevalence in the United States [5] The annual medical expenditure of asthma in Florida was nearly 1.1 billion US dollars [5] Additionally, 25% of asthmatic children had at least one day of school absences due to asthma symptoms [5] Given the fact that important to pay specific attention to asthma related health outcomes such as symptoms, asthma related ED visits, patient reported asthma control and asthma specific HRQOL, identify factors contributing to poor asthma control status and asthma related heal th outcomes, and design effective interventions to help clinicians, health care professionals, researchers and families of asthmatic children to manage asthma symptoms and improve asthma related outcomes. Health Outcomes of Pediatric Asthma Asthma outcomes are considered as the primary or secondary outcomes of many asthma clinical studies, and the Asthma Outcomes Workshop sponsored by several National Institutes of Health (NIH) institutes and the Agency for Healthcare Research and Quality (AHRQ) identified severa l key asthma outcomes, reviewed current measures and discussed how to standardize those outcomes in future studies [6] The Asthma Outcomes Workshop has identified 7 important asthma outcomes including asthma co ntrol measures, quality of life (QOL), symptoms, lung functions, asthma exacerbations, clinical biomarkers, healthcare utilization and cost [6, 7] Asthma C ontrol and Q uality of L ife Several instruments have been developed to assess asthma control status for pediatric population. For example, the Childhood Asthma Control Test (C ACT) [8] the

PAGE 16

16 Asthma Control Questionnaire (ACQ) [9] and the Asthma Control and Communication Instrument (ACCI) [10] are well validated measures to assess key domains of asthma control including symptoms, activity limitations, and use of rescue medicati on for asthma exacerbation. In 2007, the National Asthma Education and Prevention 3) revised the guideline for asthma diagnosis and management [ 11] In contrast to the previous guidelines, the EPR 3 provides a clear definition of asthma control recommends several patient reported measures of asthma control status and underscores the importance of achieving optimal asthma control as the goal of long term asthma management [11] Asthma control is defined as the extent to which manifestations of asthma have been reduced by medical treatment; and well controlled asthma status c an be achieved by reducing two components including current impairments such as frequency of day and night symptoms and functional limitation due to asthma and future risks such as the possibilities of asthma exacerbation, declines in lung function and sid e effects of medication. Asthma quality of life and it has been increasingly used as the important patient reported outcomes (PRO) in clinical research. Several l ongitudinal studies ha ve been performed to evaluate comparative effectiveness of medications on HRQOL as the primary or secondary end points; these studies showed HRQOL was improved in treatment groups for asthmatic children [12 14] The majority of longitudinal studies with PRO such as asthma control and HRQOL largely focused on adults with asthma rather than pediatric population. Table 1 1 summarizes findings from 13 longitudinal studies with a specific

PAGE 17

17 attention on changes of asthma control and/or HRQOL among asthmatics. Those studies either examined the variations in outcomes such as asthma control and HRQOL over time or investigated the associations of changes in asthma control with variati ons in HRQOL. However, there are several limitations on study designs, asthma outcome measures and generalizability in those studies. First although several studies have included asthma control status and HRQOL variables, they merely examined specific risk factors such as panic disorder [15] physical activity [16] self efficacy [16] and health literacy [17] related to changes in association between asthma control and HRQ OL using a longitudinal framework. Second the results derived from asthmatic adults investigating the changes in asthma control related to variations in HRQOL cannot be generaliz ed to asthmatics in the pediatric population [18 20] In addition, several studies focused on clinical indicators without collecting patient reported asthma control status [21, 22] For example, one study [21] suggested changes in forced expiratory volume in 1 second (FEV 1 ) was significantly associated with variations in HRQOL; the other study [22] indicated changes i n HRQOL was significantly associated with variations in individual items assessing sleep disturbance, short acting beta agonist use, activity limitation and healthcare use respectively Three published articles based on a longitudinal study among asthmati cs followed for significantly change over time, additionally, psychological outcomes including anxiety and depression assesse d by the Hospital Anxiety and Depression Scale (HADS) did not

PAGE 18

18 significantly change during the follow up period [23 25] suggesting HRQOL was stable when asthma was well controlled over ti me. However, th ese stud ies did not include subjects with significant changes in asthma control status [26] A recent longitudinal study conducted among asthmatic children found the changes of adherence to daily prescrib ed medication were significantly associated with the changes of asthma symptoms and asthma related health care service utilization including ED visits and hospitalizations [26] The PROMIS PAS data were collected among asthmatic children between 8 and 18 years of age both asthma control and HRQOL were assessed for children, especially, asthma control data were collected prior to pediatric HRQOL data, which provided opportunities to confirm the causal relationship between asthma control and HRQOL. Third very few studies have examined the contribution of asthma control and the rating of HRQ OL [27 29] I examined the contribution of asthma control on HRQOL after taking into account socio demographics, and the extent to which socio demographic factors associat ed with changes in HRQOL using latent growth models (LGM). The advantages of using LGM include the flexibility to handle discrete data and allowing for unequal time intervals on repeated measures and missing data that frequently occur in longitudinal studi es. Additionally, LGM estimates the parameters (e.g., intercept and slope) to represent the linear or non linear trajectories in outcome variables. Application of LGM in this study enables us to model a linear trend for pediatric HRQOL across 4 time points and

PAGE 19

19 examine the impact of changes in asthma control status on HRQOL trajectory after controlling for socio demographic variables. Mediators Influencing the Asthma Control HRQOL Pathway Children with asthma are more likely to report sleep problems than ch ildren without asthma [30, 31] Depending upon different study designs and population characteristics, the percentage of children with asthma experiencing sleep difficulties ranged from 30% t o 40% [32 34] It is especially evident that children with poorly controlled asthma are more likely to experience nighttime awakening and sleep difficulties than children with we ll controlled asthma status [31, 35, 36] The relationship between sleep problems and HRQOL in asthmatic children has been investigated by previous studies. These studies found that frequent nocturnal awakening was associated with more absence from school, more attention problems during classes, greater daytime sleepiness and poor school performance [35, 37, 38] Limited studies have specifically examined the collective influence of poor nighttime sleep quality and frequent daytime sleepiness on asthma control and/or HRQOL. A recent study found that asthmatic children aged 6 15 years who reporte d frequent nocturnal awakening were associated with more activity limitation, school absences and asthma related ED visits, lower quality of life and higher anxiety than those who reported less nocturnal awakening [39] Numerous studies also found that children with poorly controlled asthma status were more likely to report poor nighttime sleep quality, which in turn was associated with low academic performance, absence from school, and poor asthma specific HRQOL [40 43] Our published pilot study using path analytic approach found that poorly controlled asthma status was associated with great daytime sleepiness and lo w asthma specific HRQOL, and daytime sleepiness

PAGE 20

20 significantly mediated the association between asthma control status and asthma specific HRQOL [44] T he complex associations among asthma control sleep problems and HRQOL have not been investigated thoroughly in the literature First, previous studies only focus on a single dimension of sleep problems such as nighttime sleep quality, nighttime sleep time, or daytime sleepiness, and very few studies focus on multiple domains of sleep problems. A recent published study investigated the relationship between objective measures of two sleep domains including sleep quality and duration assessed by actigraphy, and psychological well being measured by the Satisfaction with Life Scal e in adult population [45] The results showed that among several sleep domains including sleep duration, sleep onset latency, awakening time after sleep onset, only variability in sleep duration over 7 days was associ ated with impaired subjective well being. The results also showed subjective sleep quality measured by Pittsburgh Sleep Quality Index mediated the sleep duration and well being pathway. However, the mechanism of multiple sleep factors on health outcome s in a pediatric population is still unclear [45] Given the fact tha t children with asthma experience more sleep difficulties compared to healthy children [33, 34] it is important to investigate the effects of multiple sleep dimensions on asthma control status and asthma related health outcomes such as HRQOL in children with asthma. Second, previous studies have show n the associations either between asthma control status and sleep problems, or between sleep problems and HRQOL, or between asthma control status and HRQOL; no studies have proposed a comprehensive framework to investigate the extent to which multiple domains sleep problems influence the association of a sthma control status and HRQOL

PAGE 21

21 in asthmatic children. In the PROMIS PAS, extensive data on nighttime sleep quality and daytime sleepiness in asthmatic children were collected across 4 time points during 2 year observational period, which provide the opport unity to study the influence of changes in sleep problems on pediatric HRQOL of asthmatic children. Additionally, asthma control and nighttime sleep quality data were collected prior to daytime sleepiness and HRQOL data; therefore, I was able to examine th e causal relationships among inadequately controlled asthma, poor sleep quality, great daytime sleepiness and impaired HRQOL. Because the parent study was a 2 year longitudinal study that collected data across multiple time points, the repeated measures ge nerated multi level data including within subject and between subject levels. In this dissertation, the multilevel structural equation modeling (MSEM) was applied to test the complex associations among asthma control, sleep problems and HRQOL at within sub ject level and between subject level. I specifically examined if changes in nighttime sleep quality and daytime sleepiness would influence the association of asthma control with HRQOL; and if average nighttime sleep quality and daytime sleepiness across 4 repeated measures would significantly mediat e the asthma control HRQOL pathway. Factors Associated with Asthma Outcomes The l iterature suggested that numerous factors were associated with health outcomes of pediatric asthma, and those factors included indi vidual socio demographic characteristics, healthcare delivery system, social support and environmental risk factors [46] Among children with asthma, boys, non Hispanic Black children and children in poor families were m ore likely to be diagnosed with asthma compared to girls, non Hispanic White children and children in families with higher income [1, 2] older age [47 49] low income family [50]

PAGE 22

22 ethnic minorities [51 53] and l ow maternal education [54] were associated with poorly controlled asthma status Evidence showed [27, 28] boys [28] [55] higher family income [27] higher maternal education [27] fewer chronic conditions [29] smoking status [28] were significantly associated with improvement in HRQOL among children with asthma. As a result both graphic characteristics are considered as important factors contributing to prevalence of pediatric asthma and asthma related health outcomes. Environmental risk factors have been proved to influence health outcomes of pediatric respiratory diseases. Prev ious studies suggested outdoor air pollution significantly contribute to poor asthma outcomes including frequent asthma exacerbation, more asthma symptoms and increased asthma related healthcare utilization such as ED visits and hospitalizations [56 58] Air pollution is one of the important environmental factors to be associated with asthma outcomes, and critical air pollutants include CO, SO 2 NO 2 PM 2.5 PM 10 and O 3 [56 59] For the source of air pollutants, CO and NO 2 are primarily produced by motor vehicles. Ground level ozone is produced by interaction between oxides of nitrogen and volatile organic compounds that comes from industry operation and motor vehicles. PM 2.5 and PM 10 are particulate mat ters that are classified by different diameters of solid particles such as dust and smoke or the liquid droplets such as mists and condensing vapor. Particulate matters are mainly produced by industrial facilities and automobiles. SO 2 is primarily produced by industrial facilities such as fuel and coal combustion at power plants. Evidence suggests that these air pollutants contri bute to the adverse health outcomes [59, 60]

PAGE 23

23 Exposure to air pollution can cause adverse effects on respiratory system and increase the susceptibility to respiratory infection; it is specifically the case for O 3 PM 2.5 PM 10 SO 2 and NO 2 that cause irritation and inflammation of the airways and jeopardize lung functioning. In addition to air pollutants, pollen is another environment factor that m ay contribute to allergy and respiratory diseases including asthma. For people with allergy, the immune system overreacts to pollen related to trees and weeds, and their bodies release histamine that causes various asthmatic symptoms such as coughing, wheezing, shortness of breath and chest tigh tness [61] Wang and colleagues reported that higher pollen levels from trees and weeds were associated with higher number of asthma related ED visits and hospitalizations [62] Sparse studies have examined the extent to which exposure of different air pollutants is associated with asthma control status using data collected from individuals with asthma. However the results were still mixed with respect to types of air pollutants and asthma outcomes including asthma control stat us, symptoms and lung function. Maestrelli and colleagues reported increased exposure to PM 10 was marginally associated with poorly controlled asthma status measured by ACT, yet elevated O 3 and SO 2 exposure was significantly associated with decreased lung function measured by exhaled nitric oxide [63] Zora and colleagues found increased exposure to elevated PM 2.5 and O 3 was significantly associated with poor asthma control status measured by ACQ [64] ; however, the observed associations were not statistically significant in part due to the following limitations. First, the data were collected weekly during 3 month observational pe riod significant changes of air pollutants. Second, the size of the study sample was very

PAGE 24

24 small ( n =36) and asthmatic children were recruited from 2 schools in a single city. I focused on the effect of PM 2 5 and O 3 on asthma control status in this dissertation. Because the participants were randomly identified from the entire state of Florida single location, the results would be generalizable to Florida children e nrolled in Medicaid and SCHIP. In contrast to previous studies, the parent study has other advantages including relatively large samples ( n =238 dyads) of asthmatic children and parents and a longer follo w up period of 2 years. There are very few studies focusing on the effect of pollen on asthma control status among pediatric population. A recent study reported approximately 20% of the population in Florida suffered from allergy and there were numerous ki nds of plants that release botanic aeroallergens which contribute to allergies [65] This dissertation use d a longitudinal study with data collected from asthmatic children in Florida to investigate the association of air pollution and pollen levels with asthma control status over 26 week s Third, previous studies have investigated the individual effect of air pollution and pollen exposure on asthma outcomes in different studies or statistical models; very few studi es examined the joint effects of these two important factors. This dissertation focused on the effects of environmental factors on asthma outcomes to examine how air pollution and pollen jointly influence asthma control status and whether pollen had a larg er impact on asthma control status over time than PM 2.5 and O 3 Specific Aims of Dissertation In this dissertation, I proposed and addressed three specific aims on health outcomes of pediatric asthma in a longitudinal framework. The first aim was to examin e the contribution s of asthma control status and participant s socio demographic

PAGE 25

25 characteristics on the changes of asthma specific HRQOL over time; the second aim was to investigate the influence of nighttime sleep quality and daytime sleepiness on the association between asthma control and HRQOL ; the third aim was to investigate the relationships of asthma control with ambient air pollution and pollen exposure Specific Aim 1: Examine the Association between Asthma Control and HRQOL The first aim descri be d trajectories of asthma specific HRQOL in children with asthma followed fo r 2 years and examine d the contribution s of asthma control and baseline socio demographics on the changes of asthma specific HRQOL The firs t aim had two specific objectives. The first objective was to describe a trajectory of asthma specific HRQOL measured by the NIH PROMIS Pediatric Short Form Asthma Impact Scale across 4 measurement occasions during 2 years. The second objective was to inves tigate the effect of asthma control st atus on asthma specific HRQOL by taking into demographic characteristics. It was hypothesized that there would be a significant improvement in asthma specific HRQOL over time It was also hypothesized that indiv iduals with poorly controlled asthma status were more likely to experience impaired asthma specific HRQOL than those with adequately controlled asthma at each measurement occasion, and the rate of changes in asthma specific HRQOL over times was significant ly explained by the variation of asthma control status It was also hypothesized that children characterized with older age, male, non Hispanic Black and Hispanics, greater ed status contributed to lower initial level and slower rate of change in asthma specific HRQOL than children characterized with younger age, female, Caucasian, fewer

PAGE 26

26 partne r status. Specific Aim 2: Test a Conceptual Framework of Sleep Factors Influencing HRQOL of Asthmatic Children The second aim was to investigate the mediating effect s of nighttime sleep quality and daytime sleepiness on the pathway between asthma control and asthma specific HRQOL based on a 2 year observational study Multilevel structure equation modeling (MSEM) was use d to quantify the direct effect of asthma control on asthma specific HRQOL and the indirect effects of asthma control on asthma specific HRQOL through the mechanism of nighttime sleep quality and daytime sleepiness at both between subject and within subject levels. Three pathways that demonstrated the me diating effect s of nighttime sleep quality and daytime sleepiness on asthma control HRQOL pathway were examined. The first pathway investigated the effect of asthma control on asthma specific HRQOL through nighttime sleep quality; the second pathway examin ed the effect of asthma control on asthma specific HRQOL through daytime sleepiness; and the third pathway examined the effect of asthma control on asthma specific HRQOL through the mechanism of both nighttime sleep quality and daytime sleepiness. It was h ypothesized that asthma control status was directly associated with asthma specific HRQOL at both between subject and within subject levels; nighttime sleep quality and daytime sleepiness would significantly mediate the asthma control HRQOL pathway at both between subject and within subject levels. Specific Aim 3: Investigate the Impact of Ambient Air Pollution and Pollen Exposure on Asthma Control Status This third aim was to investigate the relationship s of asthma control status with ambient air pollution and pollen exposure measured across 26 weeks during the 2 year

PAGE 27

27 study period. The third aim had three specific objectives. The first objective was to examine the association s between ambient air pollution including PM 2 5 and O 3 exposure and asthma control status ; the second objective was to examine the effect of pollen on asthma control status among asthmatic children ; the third objective was to investigate the combined effect of air pollution and pollen on asthma control status across 26 measurement occasi ons. It was hypothesized that increased exposure to air pollution including elevated PM 2 5 and O 3 concentration was associated with worse asthma control It was also hypothesized that increased exposure to pollen measured by pollen severity index was assoc iated with worse asthma control Finally, it was hypothesized that increased air pollution and pollen exposure w ere positively associated with worse asthma control among children with asthma ; and pollen exposure had a greater influence on child asthma control status in Florida than PM 2 5 and O 3 Study Design Participants and Data Collection This study is a secondary data analysis which uses data collected from the NIH Patient Reported Outcomes Measurement Information System (PROMIS) Pediatrics Asthma Study (PAS). The NIH PROMIS PAS was a six year U01 project (2010 201 5; PI: Dr. I Chan Huang ) which was designed to pursue two specific aims: 1) to examine the responsiveness of PROMIS Pediatric Short Forms associated with the change of pediatric asthma control status and 2) to establish the clinically minimal important differences (CMIDs). All po tential subjects were identified from the database of Florida Medicaid and SCHIP and 238 dyads of children with asthma and their parents were recruited and agreed to participate in this study at the baseline of first year ( T1 ) The enrollment

PAGE 28

28 criteria for recruiting children and parents included the age between 8 and 17.9 years Florida Medicaid and SCHIP, diagnosis of asthma based on FL Medicaid and SCHIP claim and enrollm ent data (ICD 9 CM: 493.1 (asthma with status asthmaticus) or 493.2 (asthma with acute exacerbation) or other 493.x), at least two asthma related health care visits during the past 12 months and accessible to internet and phone in the past 6 weeks. In the PROMIS PAS, a total of 3,500 children were identified from the Florida Medicaid and SCHIP datebase Our research team was able to contact and speak with meet the study enrollment criteria, 684 subjects refused to participate in the study, 326 subjects verbally agreed to participate in the study via enrollment phone calls, and eventually 238 subjects consented online via the study website. After children and parents were enrolled into this study, research package s w ere sent to the study participants to introduce the study purpose and procedures such as completing weekly report s a research website and measuring lung functioning usin g peak flow meters (Figure 1 1). Data were collected in the second and third years of the project period using a WINDOW approach Asthma control status, health care utilization related to asthma, peak flow values, nighttime sleep quality and quantity, and school functioning were reported weekly (26 weeks in total across 2 years) by parents or primary caregivers through a research website: w eeks 1 13 in the first year and w eeks 14 26 in the second year. PROMIS Pediatric Item Banks and Short Forms were admini stered to assess

PAGE 29

29 pediatric HRQOL through telephone interview s with children at four different time points: the first year baseline (T1), the first year follow up (T2), the second year baseline (T3) and the second year follow up (T4), respectively. T1 and T 3 HRQOL data were collected at Weeks 1 and 14, respectively, and T2 and T4 HRQOL data were collected anytime between Weeks 2 13 and betw een Weeks 15 26, respectively. Using a WINDOW approach, the research team evaluated the change of asthma control status by comparing asthma control status reported in Week 1 to a particular week between Weeks 2 13 of the first year, and asthma control status reported in Week 14 to a particular week between Weeks 15 26 of the second year. If the change status was identified, research coordinators scheduled a telephone interview with children to collect HRQOL data. If asthma control status remained the same during the 13 week window, a telephone interview was scheduled at the end of the observation al period to asthma control status, the data for asthma control and HRQOL would be considered non days (paired data) or 14 days (dyadic data) after reporting asthma control status, the data were considered matched data. Only matched asthma control HRQOL data were included fo r analyses in this dissertation. Measures and Instruments Several well validated instruments were used in the PROMIS PAS. In this study, asthma control status [10] The PROMIS Asthma Impact Scale was used to collect asthma specific HRQOL [66] PAS was assessed using three items measuring difficult y falling asleep and getting up

PAGE 30

30 and sleep disturbance. The Iowa Pediatric Daytime Sleeping Scale (PDSS) was used to assess daytime sleepiness [67] The PROMIS PAS provided rich information on patient reported asthma outcomes to accomplish the three study aims proposed in this dissertation. The longitudinal relationship between asthma control status and HRQOL for aim 1 i s discussed in Chapter 2; the associations among asthma control status, sleep problems and HR QOL for aim 2 are discussed in Chapter 3; the impact of ambient air pollution and pollen exposure on asthma control status is discussed in Chapter 4, and the limitations of this dissertation, directions of future studies and conclusion are discussed in Cha pter 5

PAGE 31

31 Table 1 1. Literature review of longitudinal studies on asthma control and quality of life (QOL) among asthmatics Author Study design Subject Asthma measures QOL measures Statistical analysis Results Limitations Oga, 2005 Observational; 5 year; 6 repeated measures (baseline and annual measures) Adults with stable asthma status (n=87) Pulmonary F unction ; peak flow expiratory flow rate ( PERF ) and airway hyperresponsi veness (AHR) The Hospital Anxiety and Depression Scale (HADS); Asthma Quality of Lif e Questionnaire (AQLQ); and Respiratory Questionnaire (SGRQ) Mixed effects models; outcomes: AQLQ and SGRQ; predictors: FEV1, PEFR Overall AQLQ scores declined significantly but did not reach clinical significant level; SGRQ did not change;FE V1 declined significantly; PEFR and AHR significantly improved No PRO measures for asthma control due to subjects had stable asthma status during follow up; not generalizable to children Hessel ink, 2006 Observational; 2 year; 3 repeated measures (baseline, end of first and second years) Asthma adults (n=380); COPD adults (n=120) FEV1; PEFR The Quality of Life in Respiratory Illness Questionnaire Multilevel analyses; outcomes: changes in FEV1 and HRQOL; predictors (at baseline): PEFR; respiratory sy mptoms complains; ch ronic cough; socio demographics; c orrelation analysis for changes in FEV1 and HRQOL Older age, living in urban, low PEFR, greater weight; less cough was associated with FEV1 decline for asthmatics. Older age, non adherence, more dyspnea lower PEFR was related to HRQOL decrease. Changes in FEV1 were associated with HRQOL No PRO measures for asthma control; not generalizable to children Chen, 2007 Observational; One year; 2 repeated measure (baseline and end of one year) Adults (n=987) The Asthma Therapy Assessment Questionnaire The Mini Asthma Quality of Life Questionnaire (AQLQ) and EuroQoL 5 D (EQ 5D). Multiple linear regression; outcomes: Mini AQLQ and ED 5D scores; predictors: asthma control and socio demographics Poor asthma contro l at baseline was associated with worse QOL at follow up; changes in asthma control associated with AQLQ (asthma specific HRQOL), but not for EQ 5D (health status) Roles of socio demographics on changes in HRQOL were not reported; not generalizable to chil dren

PAGE 32

32 Table 1 1. Continued Author Study design Subject Asthma measures QOL measures Statistical analysis Results Limitations Oga, 2007 Observational; 5 year; 6 repeated measures (baseline and every year) Adults with stable asthma status (n=87) Pulmonary function, PERF and airway hyperresponsi veness The Hospital Anxiety and Depression Scale (HADS); Asthma Quality of Life Questionnaire (AQLQ); and Respiratory Questionnaire (SGRQ) Mixed effects models; outcomes: HADS anxiety and depres sion; predictors: AQLQ, SGRQ, FEV1, PEFR HADS anxiety and depression did not change significantly over time; changes in slope of HADS scores associated with changes in the AQLQ and SGRQ scores; changes in FEV1 and PERF were not associated with changes in H ADS No PRO measures for asthma control due to subjects had stable asthma status during follow up (if exacerbation occurred, treatment was provided, evaluation was postponed till recovery ); not generalizable to children Wood, 2007 Observational; 1 year; 3 repeated measures (baseline, 6 and 12 months) Adults (n=383) The Lara Asthma Symptom Scale (LASS) and FEV1 Asthma Quality of Life Questionnaire (AQLQ) and healthcare use ANOVA and negative binomial regression; outcome: LASS scores; predictors: AQL Q and FEV1 Changes in LASS were associated with changes in FEV1 and QOL. Not generalizable to children King, 2009 Observational; 3 year; 5 repeated measures (every 6 months) Adults (n=213) Sleep disturbance; use of SABA, activity limitation, healthcare use Asthma Quality of Life Questionnaire (AQLQ M) and the SF 36 Multilevel analyses; outcomes: HRQOL; p redictor s : asthma control and socio demographics Activity limitation influenced between subject variation, and sleep disturbance and SABA use impact within subject variation No composite measure for asthma control; not generalizable to children

PAGE 33

33 Table 1 1. Continued Author Study design Subject Asthma measures QOL measures Statistical analysis Results Limitations Oga, 2010 Review Adults with stable asthma status (n=87) Pulmonary function, PERF and airway hyperresponsi veness The Hospital Anxiety and Depression Scale (HADS); Asthma Quality of Life Questionnaire (AQLQ); and Respiratory Questionnaire (SGRQ) Review based on 5 yea r longitudinal study; 6 repeated measures (baseline and every year) For asthmatics, health status, disability and psychological time for people with stable asthma status; but the clinical outcomes such as FEV1, PEFR and AHR chan ged. Not generalizable to children Rohan 2010 A one year tailored problem solving intervention; electronic monitoring for daily medication adherence; PRO measures every 3 months Children (n=92) Electronic devices to record daily fluticasone doses, frequency of asthma symptoms and use of rescue medicine Healthcare utilization Unconditioned latent growth models (LGM); outcome: adherence to medication, symptom rating and healthcare use; Conditional LGM; outcomes: symptom rating, rescue medicine use and healthcare use; predictor: adherence to daily medicine Adherence to daily medicine declined over time, which was associated with increase in healthcare use; changes in adherence was not associated with symptoms and use of rescue medicine NO PRO measures for QOL Guilbe rt, 2011 Online survey; 1 year; 5 repeated measure (baseline and every 3 months) Adults (n=497); c hildren (n=170) The Asthma Control Test (ACT) for adults and the Childhood ACT (C ACT) for children PedsQL3.0 Asthma Module for children and the SF 12 for adults Mixed effects models; outcomes: changes PedsQL3.0 score and SF 12; predictor: ACT (well controlled vs. not well controlled) Adults and children with poorly controlled asthma status tended to have decreased HRQOL Roles of demographics o n changes in HRQOL not reported; asthma control and HRQOL no temporal sequence

PAGE 34

34 Table 1 1. Continued Author Study design Subject Asthma measures QOL measures Statistical analysis Results Limitations Terza no, 2012 Observational; 1 year; 5 repeated measures (baseline and every 3 month) Adults (n=739) The Asthma Control Test (ACT) EuroQoL 5D questionnaire ANOVA analyses, linear regression; outcome: QOL; predictor: asthma control Increase in asthma control was associated with improved HRQOL Not generalizable to children Apter, 2013 Observational; 1 year; 26 weekly repeated measures Adults (n=284) The Asthma Control Questionnaire (ACQ); electronic monitoring for medicine adherence; The Mini Asthma Quality of Life Questionnaire (AQLQ); health literacy: The Short Test of Functional Health Literacy in Adults (S TOFHLA) and the Asthma Numeracy Questionnaire (ANQ) Mixed effects models; outcomes: literacy; predictor: Asthma control and QOL Changes in health literacy was associated with chang es in asthma control and QOL The relationship of asthma control with QOL was not examined; not generalizable to children Favre au, 2014 Observational; 4 years; 2 repeated measures (baseline and follow up) Adults (n=643) The Asthma Control Questionnaire (ACQ); The Asthma Quality of Life (AQLQ) Multivariate general linear models; outcomes: asthma control and QOL at follow up; predictors: panic disorder and the Anxiety Sensitivity Index (ASI) at follow up Panic disorder and ASI at baseline associated with A CQ at follow up, but nor for QOL The relationship of asthma control with QOL was not examined; not generalizable to children

PAGE 35

35 Table 1 1. Continued Author Study design Subject Asthma measures QOL measures Statistical analysis Results Limitations Eliayy an, 2014 Observational; 1 year; 2 repeated measures (baseline and follow up) Adults (n=299) The Asthma Control Test (ACT) The Mini Asthma Quality of Life Questionnaire subscales (MAQLQ); Knowledge, Attitude, and Self efficacy Asthma Questionnaire (KA SE AQ) Path analysis; outcome: asthma control at follow up; predictors: asthma symptoms, emotional status and self efficiency at baseline; mediators: physical activity, healthcare use and medication beliefs during follow up Asthma control at baseline, symp toms, physical activity and self efficacy were directly associated with asthma control at follow up; emotion status at baseline influenced asthma control at follow up through physical activity R elationship s of asthma control with QOL was not examined; not generalizable to children

PAGE 36

36 Figure 1 1. Flow chart demonstrating recruitment of participants in the NIH PROMIS PAS. Adopted from PowerPoint with permission: Huang IC (PI) Thompson L, Knapp C, Shenkman E, DeWalt D. Validation for pediatric patient reported outcomes measurement information system ( PROMIS) in Asthmatic children. 2013 Pediatric PROMIS Annual Meetin g. University of North Carolina at Chapel Hill, Chapel Hill, North Carolin a

PAGE 37

37 CHAPTER 2 THE RELATIONSHIP BETWEEN ASTHMA CONTROL AND H EALTH RELATED QUALITY OF LIFE AMONG CHILDREN WITH ASTHMA Introduction Asthma is one of the most prevalent chronic diseases among children in the United States. There were approximately 7 million (9%) children and adolescents under 18 years had asthma in 2012 [2] Among children with asthm a, boys, non Hispanic Black children and children in poor families were more likely to be diagnosed with asthma compared to girls, non Hispanic White and children in families with higher income [1, 2] [47 49] low income family [50] ethnic minorities [51 53] and low maternal education [54] were associated with poorly [27, 28] boys [28] [55] higher family income [27] higher maternal education [27] fewer chronic conditions [29] smoking [28] were significantly associated with improved HRQOL among children with asthm a. Previous studies show ed that 40 80% of children with asthma experienc ed persistent symptoms such as coughing, wheezing and nocturnal awakening that indicated poorly controlled asthma status [49, 68 72] Numerous studies have linked poorly controlled asthma status to HRQOL impairment based on cross section al study designs [73 76] A systematic review found children with poor asthma control status tend ed to report suboptimal HRQOL including greater activity limitation and more depressive symptoms and fatigue than children with well asthma control status [73] Very few longitudinal studies have been conducted to observe the natural disease progression of asthma in children and adolescents, especially the changes in asthma c ontrol status associated with patient centered outcomes typically health

PAGE 38

38 patient reported outcomes (PROs) and HRQOL. There were several limitations of previous studies on study designs, asthma outcome measures and generalizability. For example, previous st udies mostly examined the changes of asthma control and HRQOL as separate outcomes and the longitudinal relationship of changes in asthma control with variations with HRQOL was rarely investigated. Typically, very few studies have examined the changes of a sthma control related to variations in HRQOL in pediatric asthma. To address the limitations of previous studies, the present study used secondary data collected from the NIH PROMIS PAS to explore the longitudinal association between asthma control and HR QOL. Specifically, the first aim was to examine the trend of asthma specific HRQOL over 2 years; the second aim was to identity important factors associated with changes in asthma specific HRQOL. For the first aim, it was hypothesized there would be an imp rovement in pediatric HRQOL measured by the NIH PROMIS Pediatric Short Form Asthma Impact Scale across 4 repeated measurement occasions during 2 years. It was also hypothesized that asthma control status would influence asthma specific HRQOL at individual time points, in which children with poorly controlled asthma status would be more likely to have impaired asthma specific demographics were hypothesized to relate to the initial status and rate of changes in asthma specific HRQOL. Specifically, children characterized with older age, male, non education and unmarried status would be more likely to have lower initial level and slower rate of change in HRQOL, compared to children characterized with younger age,

PAGE 39

39 female, non education background and married status. Latent growth models (L GMs) were implemented to investigate the changes of asthma specific HRQOL over time and the asthma specific HRQOL over time. Methods Participants and Data Collection This study is a secondary data analysis of the PROMIS PAS which included 238 asthmatic children and their parent s who were followed for 2 years. Participants and the procedure of data collection have been described in Chapter 1. Asthmatic children and thei r primary caregivers were identified from Florida Medicaid and SCHIP. Inclusion criteria for study participants were: age between 8 and 17.9 years for children and 18 years or older for parents SC HIP asthma diagnosis ( ICD 9 CM: 493.1 ( asthma with status asthmaticus) or 493.2 (asthma with acute exacerbation) or other 493.x ) listed in F lorida Medicaid and SCHIP claim and enrollment file, at least two asthma related health care visits during the past year and accessible to internet and phone in the past 6 weeks The Figure 1 1 summarizes the process of patient enrollment and data collection. Of 3,500 eligible children identified from the Fl orida Medicaid and SCHIP claim and enrollment file, 238 dyads of parents and children consented and assented for study participation. In the PROMIS PAS, asthma control status was reported weekly (26 weeks in total across 2 years) by parents or primary care givers through the research website: w eeks 1 13 in the first year and w eeks 14 26 in the second year. Pediatric HRQOL was collected through telephone interview with children at the following time points: the first

PAGE 40

40 year baseline (T1), the first year follow up (T2), the second year baseline (T3) and the second ye ar follow up (T4). Study Measures website. In each year, asthma control status was collected across 13 individual weeks u sing the Asthma Control and Communication Instrumen t (ACCI). The ACCI is a well validated instrument with satisfactory psychometric properties including concurrent validity, discriminant and known group validity [10] The ACCI was developed in 2008 on the basis of the 2007 National Asthma Education Prevention Program (NAEPP) Expert Panel Report 3 (EPR 3) [11] The ACCI is comprised of 11 items ass essing five domains of asthma outcomes including five items measuring asthma control, three items measuring short term asthma related health care, one item measuring direction of asthma symptoms, one item measuring adherence to daily asthma medication and one item measuring asthma concern. O ne open ended question is also used to measure patient and physician communication. The overall asthma control status for individuals is determined by the five items measuring asthma control and each child is assigned t o poorly controlled asthma status if she/he answers all 5 items as adequately controlled; otherwise, each child will be assigned to poorly controlled asthma status if any of 5 questions is responded as poorly controlled. Pediatric HRQOL was self reported b y children twice within the 13 week observational window in each year through telephone interviews. In NIH PROMIS PAS, Pediatric Short Forms were used to collect seven domains of HRQOL including asthma impact, fatigue, depressive symptoms, anxiety, pain, p eer relationship, and physical function mobility. NIH PROMIS Pediatric system created HRQOL item banks using item

PAGE 41

41 response theory, and the Short Forms of individuals item bank s were developed and validated using legacy instruments including the PedsQL [77] The present study focused on a sthma specific HRQOL assessed by the PROM I S Asthma Impact Scale. The response categories of Asthma Impact Scale range from1 to 5, in which 1 indicates Individual items of Asthma Impact Scale were used to estimate the domain score for each child using item response theory methodology, and higher domain scores indicate worse H RQOL. smoking status at home were collected at baseline of the first year. Statistical Analy sis Descriptive analyses were performed to analyze the distribution of socio demographic characteristics among study participants. The percentage or mean and conditions a income and smoking status at home were examined. C orrelation coefficients were estimated to examine the inter correlation of asthma control status with asthma specific HRQOL by i ndividual time points ( Appendix A Table A 1 ) T 2 tests were conducted to examine the bivariate associations of asthma control status with asthma characteristics by individual time points. T values w ere reported for associations of asthma control status with continuous variables including asthma specific HRQOL, 2 values were reported for associations of asthma control status with categorical variables including

PAGE 42

42 home. Linear regression analyses were performed to investigate the multivariate associations among asthma control, asthma specific HR QOL demographic characteristics by individual time points. At each measurement occa sion, a linear regression model w as performed using asthma specific HRQOL as the dependent variable and asthma control as the main independent varia ble by controlling for the influence of education background and smoking status. In each regression analy sis, a dequately controlled asthma status, girls, non Hispanic White, married/living with partners, higher education background (associate degree/some college/ college/ advanced degree) and not smoking at home were treated as reference groups Unconditiona l LGM (Figure 2 1) was conducted to examine the change s in asthma specific HRQOL over time and the conditional LGMs (Figure 2 2 and Figure 2 3 ) were performed to investigate the longitudinal relationship between asthma control and asthma specific HRQOL af baseline of the first year. LGM has been increasingly used in psychosocial and behavioral research especially for analyzing longitudinal outcome data LGM possesses unique advantages over traditiona l models such as generalized estimating equations (GEE) and reported measure analysis of covariance (ANCOVA) [78, 79] because LGM allows for missing data, unequally time spacing, non normal distribution

PAGE 43

43 data, and nonlinear trajectories. LGM can also model discrete measures and take both time invariant and time varying covariates into consideration. In the first step, unconditional LGM (Model 1 ) was specified to describe the specific H RQOL across T1, T2, T3 and T4. The growth pattern of asthma specific HRQOL is represented by two indicators : intercept (I) and slope (S). The intercept (I) is the initial status of asthma specific HRQOL, the slope (S) indicates the rate of change in asthma specific HRQOL over 2 years. For the purpose of model identification, t he factor loadings for the intercept are fixed to 1, and the factor loadings for the slope are assigned base d on the time spacing of T2, T3 and T4 vs. T1 In the present study time was centered at the first measurement occasion (T1) for each participant, whereas the factor loading of the first measurement for all participants is 0, indicating the beginning of o bservations. The average time spacing was 49.86 days between T2 and T1, 358.74 days between T3 and T1 and 411.61 days between T4 and T1. T he time spacing across 4 repeated measures was divided by 30 days to demonstrate the number of months between measurem ent occasions. The final factoring loadings for asthma specific HRQOL measured at T2, T3 and T4 were 1.66, 11.96 and 13.72, respectively. In the second step, two conditional LGMs (Model s 2 and 3) were performed to identify important factors contributing to the changes in asthma specific HRQOL. The first conditional LGM (Model 2) was performed to examine the relationship s of HRQOL age, gender, race/ethnicity, number of chron status and education background. Repeated measures of asthma control status were

PAGE 44

44 treated as the important time varying predictors that were associated with HRQOL at individual time points. Specifically, asthma contr ol status at T1 was assumed to influence HRQOL at T1 and T2; asthma control status at T2 was assumed to influence HRQOL at T2; asthma control status at T3 was assumed to influence HRQOL at T3 and T4; asthma control status at T4 was assumed to influence HRQ OL at T4. Variations in demographic factors. The second conditional LGM (Model 3) was performed by taking into account the influence of socio demographic factors on asthm a control status in individual measurement occasions We used comparative fit index (CFI) and the root mean square error of approximation (RMSEA) to indicate the adequacy of the model fit for both unconditional and conditional LGMs. The values of satisfied model fit indices are: CFI 0.95 and RMSEA 0.06 [80, 81] The LGMs were performed using Mplus 7.3.2 (Muthen and Muthen, Los Angeles, CA), and the rest of the analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC). Results Participant C haracteristics 2 1. For children at baseline of the first year (N=238), the mean age was 12.23 years old (SD: 2.58); 59.92% were boys; 38.40% were n on Hispanic White; the mean number of chronic conditions was 1.53 (SD: 0.83). For parents, the mean age was 40.72 years old (SD: 8.81); the majority of them were married/living with partners (50.63%); most of them had education background of some college, associated degree or college degrees (61.11%), and had family income between $15,000 and $35,000 (44.73%). The majority of the parents smoked at home (82.28%).

PAGE 45

45 Characteristics Poorly controlled asthma status in children was 44.29% at T1, 40.51% at T2, 37.28% at T3 and 40.88% at T4. The mean scores of asthma specific HRQOL in children consistently decreased from T1 to T4: 48.11 (SD: 10.22) at T1, 46.51 (SD: 10.09) at T2, 45.42 (S D: 10.06) at T3 and 44.88 (SD: 10.20) at T4. Table 2 2 shows the results of demographics and asthma specific HRQOL at each measurement occasion. Poorly controlled asthma status was signi ficantly associated with worse asthma specific school and below (p<0.05) and greater number of chronic conditions (p<0.01) were associated with poorly controlled asthma sta tus at T1. Non Hispanic Black (p<0.05) and children with greater number of chronic conditions (p<0.05) were more likely to have poorly controlled asthma status at T3, compared to non Hispanic White and children with fewer number of chronic conditions. Characteristics Table 2 3 shows the associations among asthma control, asthma specific regressi on models. Compar ed to adequately controlled asthma status, p oorly controlled asthma status was significantly associated with poorer asthma specific HRQOL by 5.25 at T1 (p<0.001), 6.24 at T2 (p<0.001), 6.68 at T3 (p<0.001) and 3.83 at T4 (p<0.05), respecti cs. Parents with education background of high school or below was associated with poorer asthma specific compared to parents with

PAGE 46

46 education background o f college or advanced degree B oys were more likely to have better asthma 3.59, p<0.05). Longitudinal Associations between Asthma Control and HRQOL Table 2 4 shows changes of asthma specific HRQOL based on unconditional LGM (Model 1) and conditional LGMs with covariate adjustment (Model s 2 and 3 ) The unconditional LGM showed adequate model fits (X 2 : 7.00, df: 5, p: 0.22, CFI: 0.99, RMSEA: 0.04) suggesting a growth of asthma specific HRQOL over time For the trajectory o f change the mean of the intercepts was 47.80 (p<0.001), and the variance of the intercepts was 69.32 (p<0.001); this indicates individuals possessed different levels of asthma specific HRQOL at the beginning of the study. The mean of the slope was 0.19 (p<0.01) suggesting a significant increase in asthma specific HRQOL among asthmatic children from T1 to T4. T he variance of the slopes was 0.29 (p<0.001) suggesting the rates of increase in asthma specific HRQOL among individuals over time were significant ly different. The correlation between the intercept and slope parameters was 1.84 (p<0.01), which implies that individuals with higher level of asthma specific HRQOL at baseline tended to have slower rates of changes over time during the observational per iod. The two conditional LGMs that added asthma control as a time varying variable baseline socio demographic characteristics into the unconditional LGM were performed to test the extent to which these factors contribute to the trajectori es of asthma specific HRQOL. Specifically, model 2 tested the effect of asthma control at T1 on asthma specific HRQOL at T1 ; asthma control at both T1 and T2 on asthma specific HRQOL at T2 ; asthma control at T3 on asthma specific HRQOL at T3 ; asthma contro l at both T3 and T4 o n asthma specific HRQOL at T4.

PAGE 47

47 characteristics were assumed to contribute to the intercepts and slopes of the asthma specific HRQOL trajectories. The model 2 reveals adequate model fits (CFI: 0.95, RMSEA : 0.03) The mean of the intercept for model 2 was 41.31 (p<0.001) and the variance of the intercepts was 57.75 (p<0.001) suggesting children possessed different levels of asthma specific HRQOL at the beginning after adjusting for their asthma control stat us and socio demographic characteristics (Table 2 4). However, the mean of the slopes was 0.10 (p>0.05) and the variance of the slopes was 0.25 (p<0.001), which suggests the overall asthma specific HRQOL did not change significantly over time after account ing for asthma control as a time varying variable Table 2 5 (Model 2) shows that asthma control status was a significant predictor of impaired asthma p< 0.001), except for the coefficient of asthma control status at T1 related to asthma specific HRQOL at T2 (p>0.05). P a significant effect on the initial level of asthma or below education background was associated with worse asth ma specific HRQOL at baseline. C h significantly contributed to the change of asthma specific HRQOL; boys and Hispanic children were more likely to have greater asth ma 0.43, p<0.05) and non Hispanic Whites 0.64, p<0.01). T he model 3 (Figure 2 3) further tested whether socio demographic variables contribute d to asthma control status at individual time points. The model reve als adequate model fits (CFI: 0.98, RMSEA: 0.03 ) The mean of the intercepts was 41.68 (p<0.001) and the variance of the intercepts was 63.97 (p<0.001), suggesting a

PAGE 48

48 significant variation in asthma specific HRQOL at baseline (Table 2 4). The mean of slopes was 0.31 (p>0.05) and the variance of the slopes was 0.28 (p<0.001), indicating overall asthma specific HRQOL did not change significantly over time after taking all covariates into consideration Similar to model 2, mode 3 ( Table 2 5 ) shows that the effe cts of asthma control on HRQOL at all four time points were all statistically cep t for the effects of asthma control at T1 on asthma educatio n of high school or below reported poorer initial level of asthma specific children reported increasing asthma specific HRQOL over time compared with non Hisp 0. 75, p<0.01). For correlates with asthma control status at individual controlled asthma status at T1 co mpared with college or above education background and fewer number of chronic conditions. Non 0.56, p<0.05) and children with 0.26, p<0.01) were likely to report poorly controlled asthm a status at T3 compared with non Hispanic White and children with fewer number of chronic conditions. Discussion This study tested the relationships between asthma control status and asthma specific HRQOL by using longitudinal data and LGM methodology. Giv en asthma control status was reported prior to reporting of pediatric HRQOL, our results demonstrated a causa l relationship between changes of asthma control status and the variations in asthma specific HRQOL. This study found that although there wa s a

PAGE 49

49 significant improvement in asthma specific HRQOL during the study period, this trend ial level of asthma rates of change in asthma specific HRQOL over time. To the best of our knowledge, this is among limited attempts investigating the longitudinal relationship of asth ma control with asthma specific HRQOL in children using latent growth models with multiple baseline predictors. Results of this study reveal the changes in HRQOL were associated with variations in asthma control status, and it is important for clinicians and health professionals to design specific interventions to improve long term HRQOL through effective management of asthma control status among asthmatic children. Several intervention strategies have been developed for children with asthma to improve the ir HRQOL in the past decades. A systematic review [82] summarized findings from 18 psychosocial interventions attempting to improve HRQOL among asthmatic children and caregivers. This review study found that previous studies primarily focused on asthma education [83, 84] and problem solving skill ap p roach [85, 86] to improve HRQOL through the mechanism of improving asthma knowledge, self management skills, and self efficacy in children alone [85] or in both children and their parents [83, 8 4, 86] Although various education interventions are available [85, 87] numerous studies suggest that a patient centered approach is needed to monitor symptoms using appropriated measures,

PAGE 50

50 involvement in decision making pr ocess [88, 89] Patient centered strategies are especially useful to effectively monitor trajectories of asthma symptoms and design individualized interventions for children with different characteris tics. Given asthma control status has been proven as an important predictor for control and maintain asthma in well controlled status. Rapid learning approaches have been succes sfully used in people with chronic conditions (e.g., chronic pulmonary diseases [90] and cancer survivors [91] ) to monitor long term PRO, assist clinicians to provide timely confidence in self care. This approach could be an important application to address asthma control issue s in children with asthma. In a healthcare delivery system incorporated with rapid learning approaches, PROs are regularly monitored using electronic devices such as tablet computers, and results are transferred and stored in servers to generate symptom information, identify potential problems and guide clinicians to provide individualiz ed treatment plans [92, 93] Additionally, several systematic reviews suggested individualized text messaging via mobile devices could engage participants in learning knowledge of asthma self management skills and improve adherence to daily medication among asthmatic children [94 96] Bidirectional or interactive text messaging can help asthmatics to manage symptoms and increase self efficacy through effective communications between patients and clinicians (e.g., receiving a prompt response from clinicians regarding medication use for asthma relief) [97]

PAGE 51

51 Evidence has revealed that cu ltural background, English proficiency, health literacy and health beliefs were correlated with each other, and these factors will collectively influence healthcare utilization and HRQOL outcomes in asthmatics [98, 99] This is especially the case that Hispanics with low health literacy tended to have suboptimal health beliefs on asthma causes, disease progression and medication use [100, 101] This study showed Hispanic children tended to report increasing asthma specific HRQOL over time compared to non Hispanic White children after taking into account asthma control status. Our results imply Hispanic children may rece ive more benefits from interventions that target improving long term pediatric HRQOL among asthmatics than non Hispanic White. However, a system atic review summarized current asthma interventions and found the majority of them were not culturally or lingui stically sensitive [99] In future studies, use of linguistically sensitive interventions in patient centered care is needed to improve health literacy of Hispanic children with asthma and their parents through e ffective education on their health beliefs. In a clinical setting, implementing routine assessments of asthma control and HRQOL among asthmatic children are critical practice for helping physicians identify poorly controlled asthma and design specific inte long term. The 2007 NAEPP EPR 3 guideline recommends using PRO measures to assess asthma control [11] However, the Asthma Outcomes Workshop in 20 10 sponsored by NIH and Agency for Healthcare Research and Quality has reviewed 17 existing instruments measuring asthma specific HRQOL. Unfortunately, the Workshop concluded that the extant instruments were not satisfied with the high standard for measuri [102] Further revision on

PAGE 52

52 the extant instruments assessing pediatric HRQOL is needed and healthcare professionals and researchers may need to update current asthma guidelines to underscore the importance of monitoring long term HRQOL and provide recommendations of reliable PRO measures of asthma specific HRQOL to clinician and caregivers of children with asthma. Limitations There are several limitations that should b e noted for this study First, most children in this study were from economically disadvantaged families. Therefore, the demographic characteristics measured at baseline of th e first year were included in marital status and family income, which may be related to the asthma specific HRQOL ics need to be assessed with measures of patient reports. The reason of using the parental proxy report of asthma control status i s there are items in the ACCI assessing medication use and we believe parents generally have better knowledge on the types and dosage of asthma medicine than children. Literature showed both parents and children tended to over report adherence to medicatio n compared to objective assessments such as electronic canister measures, although they reported similar use of asthma medication [103, 104] Future studies are needed to examine the discrepancy i report and proxy report.

PAGE 53

53 Conclusion In conclusion, this study found that there was a change in asthma specific HRQOL over time; however, the changes in HRQOL were explained by asthma control status and baseline socio demographics. Low parental education was associated with lower initial level of asthma specific HRQOL and Hispanic children tended to have slower rate of change in it. Innovative strategies including patient centered approaches using text messaging and culturally sensitive approaches to address asthma control problems are important to improve HRQOL in vulnerable asthmatic children.

PAGE 54

54 Figure 2 1. The unconditional latent growth model (LGM) examining the trajectories of asthma specific HRQOL across 4 time points

PAGE 55

55 Figure 2 2. The conditional latent growth model (LGM) examining the influence of asthma control status and socio demographics on changes of HRQOL race/ethnicity, number n background and marital status

PAGE 56

56 Figure 2 3. The final conditional latent growth model (LGM) examining the effect of asthma control and socio demographics on changes of HRQOL over time background and marital status

PAGE 57

57 Table 2 1. Participant characteristics Characteristics Number of subject (%) or mean (SD) 12.23 (2.58) Boys 142 (59.92%) Girls 95 (40.08%) White/non Hispanic 91 (38.40%) Black/non Hispanic 61 (25.74%) Hispanic 64 (27.00%) Other 21 (8.86%) 40.72 (8.81) White/non Hispanic 102 (43.04%) Black/non Hispanic 62 (26.16%) Hispanic 60 (25.32%) Other 13 (5.49%) High school or below 75 (32.05%) Some college / technical/associated degree and college degree 143 (61.11%) Advanced degree 16 (6.84%) Family income,% < $14,999 49 (20.68%) $ 15 ,000 $ 34,999 106 (44.73) $ 35,000 $ 54,999 58 (24.47%) >55,000 24 (10.13%) ,% Never married 42 (17.72%) Married 120 (50.63%) Living with partner in committed relations 10 (4.22%) Separated 10 (4.22%) Divorced 48 (20.25%) Widowed 7 (2.95%) Yes 195 (82.28%) No 42 (17.72%) 1.53 (0.83)

PAGE 58

58 Table 2 2 Bivariate analyses f or asthma control associated with socio demographic s and asthma specific HRQOL by individual time points T1 T2 T3 T4 Poor asthma control (N=97) Adequate asthma control (N=121) t or X 2 value Poor asthma control (N=77) Adequate asthma control (N=114) t or X 2 value Poor asthma control (N=62) Adequate asthma control (N=104) t or X 2 value Poor asthma control (N=65) Adequate asthma control (N=93) t or X 2 value Asthma specific HRQOL 51.37 (10.19) 45.18 (9.08) 4.73 *** 50.24 (8.64) 43.95 (10.15) 4.42 *** 49.57 (10.62) 43.23 (8.97) 4.09 *** 47.35 (9.13) 43.37 (10.60) 2.46 mean (SD) 12.18 (2.44) 12.26 (2.64) 0.23 12.38 (2.56) 12.25 (2.63) 0.34 11.71 (2.39) 12.11 (2.42) 1.03 11.69 (2.28) 12.18 (2.50) 1.26 Boy 59.79 57.02 0.17 59.74 56.14 0.24 56.45 57.28 0.01 56.92 56.99 0.00 Girl 40.21 42.98 40.26 43.86 43.55 42.72 43.08 43.01 race/ethnicity, % White 36.08 40.50 1.67 35.06 38.60 1.29 27.42 38.83 8.48 40.00 31.18 3.15 Black 29.90 22.31 29.87 22.81 40.32 20.39 23.08 34.41 Hispanic 25.77 28.93 25.97 29.82 25.81 28.16 27.69 22.58 Other 8.25 8.26 9.09 8.77 6.45 12.62 9.23 11.83 age in years, mean (SD) 40.23 (8.75) 41.08 (8.82) 0.71 42.38 (10.36) 40.61 (7.65) 1.28 39.58 (9.34) 42.08 (8.44) 1.77 40.28 (8.61) 42.32 (9.27) 1.40 Education, % High school or below 40.63 26.05 5.14 30.26 30.70 0.00 36.07 28.43 1.03 31.25 33.70 0.10 College degree or above 59.38 73.95 69.74 69.30 63.93 71.57 68.75 66.30 status ,% Married 52.58 44.63 1.36 46.75 49.12 0.10 56.45 41.75 3.36 47.69 49.46 0.05 Others 47.42 55.37 53.25 50.88 43.55 58.25 52.31 50.54 Smoking status,% Yes 78.35 85.12 1.68 76.62 86.84 3.36 83.87 81.55 0.14 81.54 84.95 0.32 No 21.65 14.88 23.38 13.16 16.13 18.45 18.46 15.05 Number of chronic conditions, mean (SD) 1.68 (0.80) 1.37 (0.77) 2.90 ** 1.55 (0.85) 1.48 (0.78) 0.53 1.73 (0.91) 1.38 (0.70) 2.58 1.58 (0.86) 1.47 (0.77) 0.85 *p<0.05, **p<0.01, *** p <0.001

PAGE 59

59 Table 2 3 Multivariate linear regression ana lyses for asthma specific HRQOL associated with asthma control and socio demographic characteristics by individual time points T1 (N=214) T2 (N=186) T3 (N=162) T4 (N=156) 95% CI 95% CI 95% CI 95% CI Asthma control (Ref: Adequate control) Poor control 5.25*** (1.3 8 ) [2 53 7 97 ] 6.24*** (1.43) [ 3 42 9.05] 6.68*** (1.67) [ 3.39 9.98) 3.83* (1.65) [ 0.56 7.09] 0.05 (0.28) [ 0.59, 0.50] 0.40 (0.29) [ 0.17, 0.97] 0.40 (0.34) [ 0.27, 1.07] 0.41 (0.36) [ 0.30, 1.13] (Ref: Girls) Boys 1.21 (1.35) [ 3.87, 1.46] 1.26 (1.42) [ 1.55, 4.07] 2.97 (1.55) [ 6.03, 0.09] 3.59* (1.63) [ 6.82, 0.36] White) Black 1.04 (1.74) [ 2.38, 4.46] 0.87 (1.85) [ 2.78, 4.51] 2.20 (2.00) [ 6.16, 1.75] 0.09 (2.05) [ 4.14, 3.96] Hispanics 2.95 (1.67) [ 0.34, 6.24] 2.44 (1.77) [ 1.06, 5.94] 2.44 (2.00) [ 6.39, 1.51] 1.52 (2.15) [ 5.77, 2.72] Others 3.17 (2.58) [ 1.91, 8.25] 4.27 (2.57) [ 0.80, 9.35] 1.20 (2.76) [ 6.66, 4.25] 0.64 (2.87) [ 6.31, 5.03] 0.02 (0.08) [ 0.19, 0.14] 0.08 (0.09) [ 0.26, 0.10] 0.04 (0.10) [ 0.23, 0.16] 0.12 (0.10) [ 0.32, 0.08] Marital status (Ref: Married) Not married 2.14 (1.39) [ 0.61, 4.88] 2.57 (1.46) [ 0.32, 5.46] 0.28 (1.61) [ 2.90, 3.45] 1.15 (1.68) [ 2.18, 4.48] Education (Ref: College or above) High school or below 3.47* (1.47) [0.57, 6.37] 3.80* (1.56) [0.72, 6.88] 0.90 (1.67) [ 2.41, 4.21] 3.12 (1.75) [ 0.34, 6.58] Smoking status (Ref: Not smoking) Smoking at home 1.32 (1.75) [ 4.77, 2.13] 2.88 (1.88) [ 6.59, 0.83] 2.81 (2.03) [ 6.82, 1.21] 1.42 (2.20) [ 5.77, 2.93] Number of chronic conditions 0.54 (0.86) [ 1.15, 2.23] 0.82 (0.87) [ 0.91, 2.55] 0.38 (0.99) [ 2.34, 1.59] 0.02 (1.01) [ 2.03, 1.98] *p<0.05, **p<0.01, *** p <0.001

PAGE 60

60 Table 2 4. Change of asthma specific HRQOL based on unconditional latent growth model (Model 1) and conditional latent growth models with covariate adjustment (Model 2 &3 ) Model 1 Model 2 Model 3 I ntercept (I) S lope (S) Correlation between I & S I ntercept (I) S lope (S) Correlation between I & S I ntercept (I) S lope (S) Correlation between I & S Mean 47.80 *** 0.19 ** 1.84 ** 4 1 31 *** 0 10 1.71 ** 41.68 *** 0.31 2.39 *** Variance 69.32 *** 0.29 *** 57. 75 *** 0.2 5 *** 63.97 *** 0.28 *** Model fit indices X 2 (df) 7.00 (5) 8 3 71 (69) 37.35 (33) P value 0. 22 0.1 1 0. 28 CFI 0.99 0.9 5 0.98 RMSEA 0.04 0.03 0.03 *p<0.05, **p<0.01, *** p <0.001

PAGE 61

61 Table 2 5. The associations of changes in asthma specific HRQOL with asthma control and socio demographics based on conditional latent growth models (Models 2 &3 ) Model 2 Model 3 SE SE Effects to asthma specific HRQOL T1 Asthma control T1 (Ref: Adequate control) Poor control 0.2 3 *** 0.06 0.35*** 0.08 Effects to asthma specific HRQOL T2 Asthma control T1 (Ref: Adequate control) Poor control 0.0 1 0.06 0.16 0.10 Asthma control T2 (Ref: Adequate control) Poor control 0.2 3 *** 0.0 5 0.45*** 0.08 Effects to asthma specific HRQOL T3 Asthma control T3 (Ref: Adequate control) Poor control 0. 31 *** 0.0 6 0.41*** 0.09 Effects to asthma specific HRQOL T4 Asthma control T3 (Ref: Adequate control) Poor control 0.2 6 *** 0.07 0.51*** 0.11 Asthma control T4 (Ref: Adequate control) Poor control 0.1 8 ** 0.05 0.40*** 0.08 Effects to Intercept (I) 0.02 0.08 0.0 1 0.0 8 Boys 0.05 0.16 0.06 0.16 White) Black 0. 20 0.20 0.08 0.20 Hispanic 0.36 0.19 0.33 0.20 Others 0. 49 0.29 0.40 0.25 0.03 0.08 0.03 0.08 Marital status (Ref: Married) Not married 0.2 6 0.16 0.23 0.16 Education (Ref: College or above) High school or below 0.51** 0.16 0.39* 0.17 Number of chronic condition 0.08 0.08 0.02 0.09 Effects to Slope (S) 0.1 3 0.11 0.16 0.11 Boys 0.4 3 0.19 0.37 0.20

PAGE 62

62 Table 2 5. Continued Model 2 Model 3 SE SE Black 0.4 3 0.24 0.47 0.26 Hispanic 0.6 4 ** 0.24 0.75** 0.26 Others 0.6 0 0.34 0.42 0.36 0.04 0.10 0.02 0.11 Marital status (Ref: Married) Not married 0.31 0.20 0.26 0.21 Education (Ref: College or above) High school or below 0.32 0.21 0.28 0.23 Number of chronic condition 0.1 2 0.10 0.16 0.10 Effects to asthma control T1 0.01 0.09 Boys 0.11 0.17 (Ref: White) Black 0.26 0.21 Hispanic 0.11 0.21 Others 0.14 0.32 0.06 0.09 Marital status (Ref: Married) Not married 0.19 0.17 Education (Ref: College or above) High school or below 0.51** 0.18 Number of chronic condition 0.26** 0.08 Effects to asthma control T2 0.04 0.09 Boys 0.17 0.18 Black 0.26 0.24 Hispanic 0.01 0.22 Others 0.09 0.32 age 0.07 0.09 Marital status (Ref: Married) Not married 0.07 0.18

PAGE 63

63 Table 2 5. Continued Model 2 Model 3 SE SE Education (Ref: College or above) High school or below 0.05 0.19 Number of chronic condition 0.05 0.09 Effects to asthma control T3 0.04 0.12 Boys 0.07 0.20 Black 0.56* 0.24 Hispanic 0.23 0.25 Others 0.18 0.37 0.12 0.10 Marital status (Ref: Married) Not married 0.21 0.20 Education (Ref: College or above) High school or below 0.32 0.21 Number of chronic condition 0.26** 0.10 Effects to asthma control T4 0.08 0.11 Boys 0.04 0.20 Black 0.36 0.24 Hispanic 0.02 0.26 Others 0.33 0.37 0.10 0.11 Marital status (Ref: Married) Not married 0.03 0.21 Education (Ref: College or above) High school or below 0.02 0.21 Number of chronic conditions 0.07 0.09 *p<0.05, **p<0.01, *** p <0.001

PAGE 64

64 CHAPTER 3 LONGITUDINAL ASSCTONS AMONG ASTHMA CONTROL, SLEEP PROBLEMS AND HEALTH-RELATED QUALITY OF LIFE IN CHILDREN WITH ASTHMA Introduction Asthma is a common chronic illness which affects approximately 6.9 million children in the United States [2]. Literature suggests children with asthma tend to experience more sleep problems than children without asthma [30, 31]. Sleep problems are prevalent in children with asthma and previous studies showed 30%-40% asthmatic children reported sleep difficulties [32-34]. Numerous studies have linked poorlycontrolled asthma status to sleep problems including frequent nocturnal awakening and sleep disturbance [31, 35, 36]. The association of sleep problems with HRQOL in children with asthma has been investigated by previous research and the findings indicated frequent nighttime awakening was associated with greater daytime sleepiness, more attention problems during classes, poorer academic performance and more missed school days [35, 37, 38]. However, the complex associations among asthma control, different aspects of sleep problems such as sleep disturbance and daytime sleepiness, and HRQOL were still unclear. Our pilot study focused on one aspect of sleep problems and found children with poorly controlled asthma status were more likely to experience impaired asthma-specific HRQOL, and this association was explained by daytime sleepiness as a mediator [44] Previous studies that attempt to understand the mechanism through which sleep factors influence the association between asthma control and HRQOL are limited. First, previous studies only focused on a single dimension of sleep problem such as nighttime sleep disturbance, sleep duration, or daytime sleepiness [44, 45]; very few studies have focused on multiple domains of sleep problems at the same time among children with

PAGE 65

65 asthma. A recent study found sleep duration was associated with subjective well being, and other sleep domains such as sleep onset latency and sleep disturbance were n ot being [45] However, this study was conducted among the general adult population, and it is unclear whether the same findings can be generalized to the pediatric population. It is important to examine multiple sleep domains (e.g., nighttime sleep quality and dayt ime sleepiness) and investigate their roles in mediating the association of asthma control status with HRQOL. Second, previous studies mostly relied on cross sectional study design; therefore, the evidence derived from previous study could inform the associations rather than causal relationships. It is valuable to conduct a longitudinal study that collects asthma control, sleep problems (e.g., nighttime sleep quality an d daytime sleepiness) and HRQOL across multiple measurement occasions to examine if changes in sleep problems would mediate the relationship between asthma control and asthma specific HRQOL. This study aimed to investigate the influence of nighttime sleep quality and daytime sleepiness on the pathway between asthma control and asthma specific HRQOL using the NIH PROMIS Pediatric Asthma Study (PAS) which is a 2 year longitudinal study. Asthma control, nighttime sleep quality, daytime sleepiness and asthma sp ecific HRQOL were measured across 4 time points during 2 year window for individual participants. Repeated measures of variables across multiple time points created two level data, and the mean scores of each variable across 4 repeated measures were used f or results at between subject level and individual scores of each variable across 4 measurement occasions were used for results at within subject level. The main purpose of this study was to investigate the effects of asthma control on

PAGE 66

66 asthma specific HRQO L (defined as direct effects) and the effects of asthma control on asthma specific HRQOL through the mechanism of nighttime sleep quality and daytime sleepiness (defined as indirect effects) at both between subject and within subject levels using multileve l structure equation modeling (MSEM) approach. It was hypothesized that asthma control was directly associated with asthma specific HRQOL at both between subject and within subject levels. Additionally nighttime sleep quality and daytime sleepiness would significantly mediate the direct association of asthma control with HRQOL at both between subject and within subject levels. Methods Participants and Data Collection This study is a secondary data analysis using data collected from the NIH PROMIS PAS. PROM IS PAS was a longitudinal study which collected patient reported asthma outcomes from 238 dyads of asthmatic children and their parents over two years. A detailed flowchart demonstrating data collection procedure was presented in Figure 1 1. In the PROMIS PAS, changes of asthma control status and the corresponding changes in nighttime sleep quality, daytime sleepiness and HRQOL were identified through a pre specified 26 week time window across 2 years. Asthma control status and nighttime sleep quality were reported weekly by parents via the study website, including the weeks 1 13 in the first year and the weeks 14 26 in the second year. Changes in asthma control were identified by comparing the control status in weeks 2 13 to the baseline statu s assessed at week 1 (T1) of the first year, and the control status in weeks 15 26 to the baseline status assessed at week 14 (T3) of the second year. The baseline HRQOL and daytime sleepiness were self reported by children through

PAGE 67

67 telephone interviews at week 1 and week 14. Once the change of asthma control status has been identified, additional HRQOL and daytime sleepiness were self reported by children at any week between the weeks 2 and 13 in the first year (T2) and any week between weeks 15 and 26 in t he second year (T4). Therefore, asthma control and nighttime sleep quality were collected prior to daytime sleepiness and HRQOL. Study Measures The Asthma Control and Communication Instrument (ACCI) was used to assess ACCI has been validated by previous studies and has shown satisfactory psychometric properties including concurrent validity, discriminant and known group validity [10] The ACCI was developed based on the 2007 Natio nal Asthma Education Prevention Program (NAEPP) Expert Panel Report 3 (EPR 3) [11] This instrument is comprised of five items measuring symptoms, use of rescue medicine, occurrence of asthma attack, activity limitation due to asthma and nighttime awakenings to capture the concept of asthma control status. The overall asthma control score of an individual is calculated by summarizing the scores of five individual items with a range from 0 to 19. Higher scores indicate worse asthma control status. assessed by 3 questions: 1) Last week, how difficult was it for your child to settle and fall asleep after bedtime rituals; 2) Last week, how difficult was it for your child to get up in the morning; 3) Last week, how many times did your child wake up during the night? The responses

PAGE 68

68 summarizing the scores of three individual items with a range from 3 to 15. Higher s cores indicate worse nighttime sleep quality. reported twice within the 13 week observational window in each year via telephone interviews (i.e., 4 times across the study periods) alongside HRQOL data. The Io wa Pediatric Daytime Sleeping Scale (PDSS) that is comprised of 8 items was used to measure a single dimension of daytime sleepiness. A previous study has demonstrated that PDSS possess good measurement properties such as internal consistency [67] For each item, five response categories are used with a range from never (score=0) to always (score=4). The higher total scores indicate worse daytime sleepiness. In the NIH PROMIS PAS, the PROMIS Pediatric Short Forms were u sed to collect seven domains of HRQOL including asthma impact, fatigue, depressive symptoms, anxiety, pain, peer relationship, and mobility. This present study only focused on asthma impact domain to capture asthma specific HRQOL. The asthma impact domain is comprised of 8 items selected from a calibrated PROMIS asthma item bank using item response theory methodology; this domain has demonstrated satisfactory psychometric properties by previous studies including unidimensionality and convergent/discriminant validity through comparing asthma impact domain with other legacy instruments such as the SF36 and PedsQL [66, 77] The response categories of Items of individual domain are used to estimate the domain scores for each child, w ith a mean of 50 and SD of 10. Higher domain scores indicate worse HRQOL.

PAGE 69

69 The NIH PROMIS PAS collected baseline characteristics at years 1 and 2 as status, education background, and smoking status at home. Statistical Analys is Descriptive analyses were performed to analyze the distribution of socio demographic characteristics among study participants ( n =238). The percentage or status, numb education background, family income and smoking status at home were examined. To facilitate interpreting the results, the raw scores of asthma control, nighttime sleep quality and d aytime sleepiness domain scores were linearly transformed with a range between 0 (best outcomes) and 100 (worst outcomes). The mean and SD of asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL across 4 measurement occasio ns were reported. Bivariate analyses were conducted to examine the associations of asthma control with nighttime sleep quality, daytime sleepiness, asthma specific HRQOL and correl ation coefficients were reported for the associations of asthma control with nighttime sleep quality, daytime sleepiness, asthma tests was conducted to test the r 2 tests were conducted

PAGE 70

70 respectively. Linea r random intercept models were performed to examine the associations among asthma control, nighttime sleep quality, daytime sleepiness, asthma specific effects that accounts for the clustering effects of repeated HRQOL outcomes of individual subjects. Asthma control status and different variables related to sleep problems were included in 4 different random intercept models to test their associations with asthma specific HRQOL : asthma control in Model 1; nighttime sleep quality in Model 2; daytime sleepiness in Model 3; asthma control, nighttime sleep quality and marital status and education background, were treated as covariates and included in all random intercept models. Girls and non Hispanic White children and parents with unmarried status and education bac kground of college or above were treated as reference groups in the analyses. Multilevel structural equation modeling (MSEM) was applied to investigate the direct effect of asthma control on asthma specific HRQOL and the indirect effect of asthma control o n asthma specific HRQOL through the effects of nighttime sleep quality and daytime sleepiness. Given the inclusion of repeated measures for each participant, MSEM is able to distinguish the difference between within subject (repeated measures) and between subject (persons) variance and model the respective direct and indirect effects at within subject and between subject levels [105, 106] MSEM possesses

PAGE 71

71 several unique characteristics compared to traditional path analytic methods to handle multilevel data and has been applied to several social and behavioral studies [107 109] First, MSEM can incorpora te more than 1 mediator in the path analytic framework. In this study, MSEM was conducted to model the mediating effect of two mediators related to sleep problems (nighttime sleep quality and daytime sleepiness) on the pathway between asthma control and as thma specific HRQOL. Second, MSEM is able to model various relationships among independent variables, mediators and outcome variable at different levels of the data. In this study, asthma control status, nighttime sleep quality, daytime sleepiness, and ast hma related HRQOL were assessed by 4 measurement occasions that can be decomposed by two levels: repeated observation (level 1) and subject (level 2) levels. The intraclass correlations (ICC) of asthma control, nighttime sleep quality, daytime sleepiness a nd asthma specific HRQOL were examined to indicate if there was significant variance between different subjects. Final ly, MSEM is allowed to adjust education background and marital status at baseline, that are assumed to confound the associations between the variables of our study interests. The analytic framework (Figure 3 1) for MSEM was developed based on [106] Using information of within subject and between subject variance, the relationships among the variables of interest were tested using the latent scores rather than the observed scores. To estimate the unbiased confidence intervals (CIs) for indirect effects of asthma control on asthma specific HRQOL through nighttime sleep quality and daytime sleepiness, Monte Carlo si mulation

PAGE 72

72 was performed for all indirect effects. Several studies have reported that Monte Carlo simulation would provide precise CIs when bootstrap method is not feasible under multilevel context including longitudinal study design [105, 110, 111] In this study, Mplus 7.11 (Muthen and Muthen, Los Angeles, CA) was used for multilevel path analyses, and the rest of statistical analyses were performed using SAS V.9.3 (Cary, North Carolina, USA). Results Participant Characteristics Table 3 year (N=238), the mean age was 12.23 years old (SD: 2.58); the mean number of chronic conditions was 1.53 (SD: 0.83). Among study participant, 59.92% were boys; 38.40% were non Hispanic White; 44.24% were overweight. For parents, the mean age was 40.72 years old (SD: 8.81); the majority of them were married/living with partners (50.63%); most had edu cation background of some college, associated degree or college degrees (61.11%), and had family income between $15,000 and $35,000 (44.73%). The majority of the parents smoked at home (82.28%). Descriptive Analyses for Study Measures Table 3 2 shows the r esults of descriptive analyses for asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL across 4 measurement occasions. The number of participants who completed those measurements was 218 at T1, 189 at T2, 166 at T3 and 159 at T4. The mean scores of asthma control measure were 16.88 (SD: 18.66) at T1, 15.58 (SD: 17.78) at T2, 15.30 (SD: 19.02) at T3 and 13.83 (SD: 16.47) T4. The mean scores of nighttime sleep quality measure were 25.13 (SD: 19.35) at T1, 19.46 (SD: 16.97) at T2, 19.63 (SD: 18.70) at T3 and 19.39 (SD:

PAGE 73

73 16.78) T4. The mean scores of daytime sleepiness were 45.27 (SD: 20.87) at T1, 42.11 (SD: 21.44) at T2, 40.40 (SD: 21.02) at T3 and 39.64 (SD: 19.52) T4. The mean scores of asthma specific HRQOL were 48.07 (SD: 1 0.23) at T1, 46.67 (SD: 10.04) at T2, 45.47 (SD: 10.04) at T3 and 44.96 (SD: 10.17) at T4. Bivariate Associations of Asthma Control with Sleep Problems, HRQOL and Socio Demographics Table 3 3 shows the bivariate associations of asthma control with nighttim e sleep quality, daytime sleepiness, asthma demographics. Data of individual measurement occasions were treated as four waves of cross sectional data. In general, asthma control was positively associated with nightti me sleep quality across four time points (p<0.001). Poorer asthma control was associated with greater daytime sleepiness at T1 and T4 (p<0.01). Poorer asthma control was associated with lower asthma specific HRQOL across all time points (p<0.001). Among pa were associated with poorer asthma control at T1 and T3 (p<0.05). Multivariate Associations among Asthma Control, Sleep Problems and HRQOL Table 3 4 shows the associations of asthma sp ecific HRQOL with asthma control, linear random intercept models. Model 1 reveals that poorer asthma control was significantly associated with lower asthma specific HRQO was significantly associated with lower asthma Model 3 shows greater daytime sleepiness was significantl y associated with lower asthma

PAGE 74

74 sleep quality and daytime sleepiness into consideration (Model 4), poorer asthma 19; p<0.001) were significantly associated with lower asthma specific HRQOL; however, the association of nighttime sleep quality with asthma specific HRQOL was no longer l ikely to report poorer asthma control status compared to those with college degree or above in four models (all p<0.05). Path Analysis Results The ICCs for the repeated measurements of asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL were 0.21, 0.48, 0.58, and 0.52, respectively. The ICC of a variable was calculated as the ratio of between group variance to the total variance of that variable. A large ICC indicates a larger between group variation compared with within g roup variation, suggesting homogenous measures (4 repeated measures) within individual participants. The present study uses nighttime sleep quality, daytime sleepiness and asthma specific HRQOL data that need to be addressed using a multilevel analytic framework [112 114] Table 3 5 shows the path coefficients among variables estimated at both wi thin subject and between subject levels. At within subject level, the individual scores of four repeated measurements on asthma control, nighttime sleep quality, daytime sleepiness, and asthma specific HRQOL were used to interpret the relationships among a sthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL. Results reveal that poorer asthma control was associated with poorer nighttime sleep specific

PAGE 75

75 HRQ Additionally, greater daytime sleepiness was significantly associated with lower asthma sign asthma asthma control on asthma specific HRQOL through nighttime sleep quality was 0.01 (p>0.05 ); the magnitude through daytime sleepiness was 0.01 (p<0.05); the magnitude through both nighttime sleep quality and daytime sleepiness was 0.01 (p>0.05). That is, daytime sleepiness rather than nighttime sleep quality significantly explained the relation ship between asthma control and asthma specific HRQOL. At between subject level, the mean scores of four repeated measurements on asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL were used to interpret the relationship s among asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL. Results reveal that poorer asthma control was significantly associated with poorer nighttime sleep quality p<0.05) and lower asthma 0.072, p<0.05), respectively; however, 0.37, p>0.05). Additionally, poorer nighttime sleep quality was significantly associated with but not with asthma 0.26, p>0.05). Greater daytime sleepiness was significantly associated with poorer asthma for asthma control on asthma specific HRQOL throu gh nighttime sleep quality was 0.27 (p>0.05); the magnitude through daytime sleepiness was 0.01 (p>0.05); the magnitude

PAGE 76

76 through both nighttime sleep quality and daytime sleepiness was 0.22 (p<0.05). That is, at the population level, poorer sleep quality and greater daytime sleepiness significantly explain the association between asthma control and asthma specific HRQOL. Discussion This study examined the extent to which nighttime sleep quality and daytime sleepiness contribute to the association between a sthma control and asthma specific HRQOL using longitudinal data collected from a cohort of the PROMIS Pediatric Asthma Study. To our knowledge, this is the first study that uses multilevel SEM methodology to investigate the complex pathway from asthma cont rol to HRQOL through the effect of nighttime sleep quality and daytime sleepiness among children with asthma. We found poor asthma control status was significantly associated with lower asthma specific HRQOL at both between subject and within subject level s. However, the effect of asthma control status on asthma specific HRQOL was through daytime sleepiness at within person level and through both nighttime sleep quality and daytime sleepiness at between person level. Because asthma control and nighttime sle ep quality were reported by parents prior to daytime sleepiness and asthma specific HRQOL assessments completed by children, the identified pathways are deemed as causal relations. The longitudinal design of PROMIS PAS enables us to obtain within subject i nformation for understanding whether the changes of asthma control across four time points were associated with changes in nighttime sleep quality, daytime sleepiness and asthma specific HRQOL, and to obtain between subject (or population) information rega rding whether nighttime sleep quality and daytime sleepiness on average could partially explain the association of asthma control with asthma specific HRQOL.

PAGE 77

77 Previous studies have reported that poor asthma control was associated with poor sleep quality and more episodes of daytime sleepiness, and individuals with poor sleep quality and great sleepiness were likely to report impaired HRQOL [33, 115] control, sleep problems and HRQOL. Consistent with these two studies, our results based on between subject level data revealed that children with poorer asthma control on average reported poorer nighttime sleep quality and lower asthma specific HRQOL than those with better asthma control. Moreover, children with poorer nighttime sleep quality on average reported greater d aytime sleepiness, and children with greater daytime sleepiness reported impaired asthma specific HRQOL. Our findings contribute to the literature by identifying nighttime sleep quality and daytime sleepiness as two important mediators contributing to the association of asthma control with HRQOL. The effect of poor asthma control on asthma specific HRQOL among individuals was through nighttime sleep quality and daytime sleepiness rather than the independent effect of those two variables. Beside the between subject variation, multilevel mediation models provide within subject variation indicating that increasing poor asthma control contributed to greater daytime sleepiness, which in turn to poorer asthma specific HRQOL. An increase in poor asthma control was significantly associated with poorer nighttime sleep quality; however, poor nighttime sleep quality did not significantly lead to either daytime sleepiness or asthma specific HRQOL. Therefore, children in poor asthma control would result in low HRQOL thro ugh the effect of daytime sleepiness, independent of the influence of nighttime sleep problems. Our results reveal a dynamic process of sleep

PAGE 78

78 problems was related to poor asthmatic outcomes, especially the association of worse asthma control with impaired asthma specific HRQOL was partially explained by the mediating effects of changes in daytime sleepiness across multiple measurements. The significant associations of daytime sleepiness with HRQOL that were hold at within person and between subject level s, together with the significant mediating effect of daytime sleepiness on association of asthma control with HRQOL, suggest that than nighttime sleep quality. In this study, nighttime sleep quality measure only captures sleep problems by three items that assess difficulty falling asleep and getting up and sleep disturbance, thus it is unable to untangle various domains in sleep problems. The use of this simple generic measure of sleep quality is not able to sensitively detect the changes of nighttime sleep quality over time. Th is can explain why the mean score of sleep quality and daytime sleepiness mediated the effect of asthma control on HRQOL between subjects but the effec t of changes in subjective sleep quality over time were not associated with either change of daytime sleepiness or asthma specific HRQOL within individuals Several studies have proposed important concepts/domains for assessing self reported sleep problem s. For example, the PROMIS sleep wake function item banks included 27 items to assess sleep disturbance and 16 items to sleep related Sleep Patterns (CRSP) contained 60 item in strument measuring the concepts of sleep pattern, sleep hygiene and sleep disturbances for children [116] Sleep pattern scale at

PAGE 79

79 sleep, parasomnias, and insomnia; sleep hygiene scale captures caffeine and electronics use at bedtime, sleep location and hours of physical activities before bedtim e. This instrument could be applied in future studies to investigate the impact of different aspects of sleep problems on HRQOL among children with asthma. Future studies may also use comprehensive sleep information to represent sleep problems that include objective measures (e.g., polysomnography and actigraphy) to obtain accurate information of nighttime sleep behaviors and patterns [117, 118] Our results underscore the important contributions of nighttime sleep quality and daytime sleepiness on asthma specific HRQOL in children with poor asthma control. In eep history of nighttime sleep quality and daytime sleepiness. A fast screening on sleep history would be valuable for poor asthma control status. Previous studies report ed specific therapies that were effective in improving sleep for asthmatics [119 122] For example, a randomized clinical trial found melatonin could dec rease sleep disturbance among women with asthma [119] Another study reported that melatonin could also significantly reduce sleep latency and number of awakenings for children with chronic insomnia [120] Allergic rhinitis is a common comorbid condition which co occurs with asthma [123 125] and nasal symptoms have significant influenced sleep qu ality; therefore, managing rhinitis symptoms and maintaining nasal patency could be an effective way to improve sleep quality for asthmatics with rhinitis [121, 122] Unfortunately, very few treatm ent

PAGE 80

80 strategies are available for improving excessive daytime sleepiness. A recent review study suggests simulant medicines may be useful to keep normal daytime alertness [126] Additionally, cognitive therapies [127] such as relaxation therapy and stress reduction training have been proved as effective strategies to improve sleep quality [128] and reduce sleep disturbance [129] in people with chronic conditions. Future studies need to test the effect of these cognitive therapies to address sleep problems triggered by poor asthma control status. Limitations There are several limitations that should be n oted when interpreting the results of this study. First, the study results may be not generalizable to general population because the participants were recruited from the Florida Medicaid and SCHIP, and most of the subjects were from families with low econ omic status. Second, multiple factors might play significant roles to confound or mediate the relationship between asthma control and asthma allergic rhinitis may influence the association of ast hma control status with HRQOL. However, due to the nature of secondary data analysis, the detailed history of allergy was not collected in the PROMIS Pediatric Asthma Study. Third, the present study focuses on two dimensions of sleep problems (nighttime sl eep quality and daytime sleepiness) and does not include sleep duration, which is an important aspect of sleep problems and its possible role on the asthma control HRQOL pathway is still unknown. ration at night by using PRO measures of sleep duration and objective measures such as actigraphy.

PAGE 81

81 Conclusion Our study found nighttime sleep quality and daytime sleepiness play crucial roles in the asthm a control HRQOL pathway. Poorly controlled asthma status was associated with impaired HRQOL, and this association was mediated by changes in nighttime sleep quality and daytime sleepiness. Healthcare providers need to monitor asthmatic lity and daytime sleepiness regularly and develop strategies to address sleep problems for mitigating the effects of poor asthma control status on HRQOL

PAGE 82

82 Figure 3 1. Conceptual framework depicting within person and between person relationships using mul tilevel structural equation modeling. *Socio # i ects in the study

PAGE 83

83 Table 3 1. Study p articipant c haracteristics Characteristics Number of subject (%) or mean (SD) 12.23 (2.58) Boys 142 (59.92%) Girls 95 (40.08%) race/ethnicity, % White/non Hispanic 91 (38.40%) Black/non Hispanic 61 (25.74%) Hispanic 64 (27.00%) Other 21 (8.86%) Hyperactivity or attention deficit disorder/ ADD or ADHD 41 (17.30%) Born premature 27 (11.39%) Mental health conditions: depression, anxiety, bipolar disorder and other 8 (3.38%) Epilepsy or other seizure disorders 6 (2.53%) disorder 5 (2.11%) Deaf or hard of hearing 5 (2.11%) High blood pressure/hypertension 4 (1.69%) Diabetes 3 (1.27%) Chronic pain: pain from fibromyalgia, arthritis and other 2 (0.84%) Sickle cell disease 2 (0.84%) 96 (44.24%) 40.72 (8.81) race/ethnicity, % White/non Hispanic 102 (43.04%) Black/non Hispanic 62 (26.16%) Hispanic 60 (25.32%) Other 13 (5.49%) High school or below 75 (32.05%) Some college / technical/associated degree and college degree 143 (61.11%) Advanced degree 16 (6.84%) Family income,% < $14,999 49 (20.68%) $ 15 ,000 $ 34,999 106 (44.73) $ 35,000 $ 54,999 58 (24.47%) >55,000 24 (10.13%) ,% Never married 42 (17.72%) Married 120 (50.63%) Living with partner in committed relations 10 (4.22%) Separated 10 (4.22%) Divorced 48 (20.25%) Widowed 7 (2.95%) Yes 195 (82.28%) No 42 (17.72%)

PAGE 84

84 Table 3 2. Mean and standard deviation of the asthma control, nighttime sleep quality, daytime sleepiness and asthma specific HRQOL scores at 4 measurement occasions T1 (N=218) T2 (N=189) T3 (N=166) T4 (N=158) Asthma control 16.88 (18.66) 15.58 (17.78) 15.30 (19.02) 13.83 (16.47) Nighttime sleep quality 25.13 (19.35) 19.46 (16.97) 19.63 (18.70) 19.39 (16.78) Daytime sleepiness 45.27 (20.87) 42.11 (21.44) 40.40 (21.02) 39.64 (19.52) Asthma specific HRQOL 48.07 (10.23) 46.67 (10.04) 45.47 (10.04) 44.96 (10.17)

PAGE 85

85 Table 3 3 Bivariate association for asthma control with nighttime sleep quality, daytime sleepiness, asthma specific HRQOL and socio demographics by individual time points Asthma control T1 T2 T3 T4 Nighttime sleep quality 0.51*** 0.50*** 0.45*** 0.51*** Daytime sleepiness 0.29*** 0.12 0.03 0.20** Asthma specific HRQOL 0.41*** 0.37*** 0.38*** 0.30*** 0.04 0.06 0.10 0.15 Number of chronic conditions 0.14* 0.14 0.26*** 0.13 0.06 0.00 0.08 0.09 gender, % Girl 16.54 (18.58) 0.23 15.00 (16.92) 0.39 15.35 (17.55) 0.02 13.75 (16.24) 0.05 Boy 17.13 (18.78) 16.02 (18.47) 15.43 (20.18) 13.89 (16.74) White 14.05 (15.97) 1.98 12.96 (13.05) 0.91 12.54 (16.69) 1.44 13.91 (14.90) 0.01 Black 21.34 (21.03) 18.13 (20.52) 20.11 (20.21) 13.62 (17.90) Hispanic 15.75 (18.66) 16.42 (20.88) 14.89 (19.70) 14.10 (15.64) Other 20.00 (20.93) 16.76 (16.20) 13.53 (20.67) 13.53 (20.29) Education, % High school or below 19.79 (18.91) 1.63 12.89 (13.06) 1.52 16.76 (16.27) 0.69 13.53 (14.74) 0.03 College degree or above 15.38 (18.36) 16.56 (19.37) 14.55 (19.96) 13.62 (16.68) ,% Married 14.87 (17.11) 1.66 14.65 (16.59) 0.76 13.62 (18.77) 1.27 12.96 (15.75) 0.68 Others 19.05 (20.05) 16.61 (19.05) 17.37 (19.27) 14.74 (17.26) Smoking status at home,% No 19.23 (18.23) 0.87 23.79 (23.49) 2.32* 15.86 (18.23) 0.15 15.19 (18.25) 0.46 Yes 16.37 (18.76) 13.85 (15.88) 15.29 (19.27) 13.56 (16.16)

PAGE 86

86 Table 3 4 Parameter estimate of random intercept models for asthma specific HRQOL associated with asthma control, nighttime Model 1 Model 2 Model 3 Model 4 95% CI 95% CI 95% CI 95% CI Asthma control 0.17*** (0.02) [0.13, 0.20] 0.14*** (0.02) [0.11, 0.18] Nighttime sleep quality 0.12*** (0.02) [0.08, 0.16] 0.02 (0.02) [ 0.02, 0.06] Daytime sleepiness 0.21*** (0.02) [0.18, 0.24] 0.19*** (0.02) [0.15, 0.22] 0.30 (0.21) [ 0.12, 0.72] 0.34 (0.23) [ 0.11, 0.80] 0.08 (0.20) [ 0.31, 0.47] 0.20 (0.19) [ 0.18, 0.58] (Ref: Girls) Boys 1.43 (1.04) [ 3.47, 0.61] 1.15 (1.11) [ 3.34, 1.04] 0.58 (0.96) [ 2.47, 1.31] 0.67 (0.93) [ 2.49, 1.16] (Ref: White) Black 0.08 (1.34) [ 2.70, 3.55] 0.48 (1.43) [ 2.33, 3.29] 0.37 (1.23) [ 2.05, 2.80] 0.67 (1.19) [ 3.01, 1.67] Hispanics 0.84 (1.30) [ 1.71, 3.39] 1.21 (1.40) [ 1.54, 3.96] 0.82 (1.20) [ 1.55, 3.18] 0.20 (1.17) [ 2.10, 2.49] Others 1.40 (1.92) [ 2.37, 5.18] 1.42 (2.04) [ 2.59, 5.44] 0.45 (1.80) [ 3.98, 3.08] 0.74 (1.70) [ 4.08, 2.61] 0.06 (0.07) [ 0.19, 0.06] 0.10 (0.07) [ 0.24, 0.04] 0.03 (0.06) [ 0.15, 0.09] 0.03 (0.06) [ 0.15, 0.08] Marital status (Ref: Married) Not married 1.30 (1.07) [ 0.80, 3.40] 1.02 (1.15) [ 1.24, 3.28] 0.47 (0.99) [ 1.47, 2.41] 0.35 (0.96) [ 1.54, 2.24] Education (Ref: College or above) High school or below 2.93** (1.12) [0.74, 5.12] 3.13** (1.20) [0.78, 5.48] 2.99** (1.03) [0.95, 5.02] 2.56* (1.00) [0.60, 4.52] Smoking status at home (Ref: Not smoking) Smoking at home 1.92 (1.36) [ 4.59, 0.74] 1.63 (1.46) [ 4.50, 1.24] 0.66 (1.27) [ 3.16, 1.84] 0.09 (1.22) [ 2.49, 2.32] Number of chronic conditions 0.28 (0.65) [ 1.00, 1.56] 0.22 (0.70) [ 1.16, 1.59] 0.64 (0.60) [ 0.54, 1.82] 0.06 (0.58) [ 1.08, 1.21]

PAGE 87

87 Table 3 5 Multilevel path analyses for the mediating effects of nighttime sleep quality and daytime sleepiness on the association between asthma control and asthma specific HRQOL Parameter 95% CI SE Lower Upper Within level (within subject) Direct effect Asthma control > Nighttime sleep quality 0.35 0.04 0.28 0.42 Asthma control > Daytime sleepiness 0.09 0.03 0.04 0.14 Asthma control > Asthma specific HRQOL 0.12 0.02 0.08 0.15 Nighttime sleep quality > Daytime sleepiness 0.09 0.05 0.00 0.19 Nighttime sleep quality > Asthma specific HRQOL 0.03 0.03 0.01 0.08 Daytime sleepiness > Asthma specific HRQOL 0.15 0.03 0.11 0.20 Indirect effect Asthma control > Nighttime sleep quality > Asthma specific HRQOL 0.01 0.01 0.01 0.03 Asthma control > Daytime sleepiness > Asthma specific HRQOL 0.01 0.01 0.00 0.02 Asthma control > Nighttime sleep quality > Daytime sleepiness > Asthma specific HRQOL 0.01 0.00 0.00 0.01 Total indirect effect 0.03 0.01 0.00 0.05 Between level (between subjects) Direct effect Asthma control > Nighttime sleep quality 1.04 0.21 0.69 1.39 Asthma control > Daytime sleepiness 0.37 0.43 1.08 0.34 Asthma control > Asthma specific HRQOL 0.72 0.23 0.35 1.10 Nighttime sleep quality > Daytime sleepiness 0.69 0.24 0.31 1.08 Nighttime sleep quality > Asthma specific HRQOL 0.26 0.14 0.53 0.02 Daytime sleepiness > Asthma specific HRQOL 0.30 0.06 0.21 0.39 Indirect effect Asthma control > Nighttime sleep quality > Asthma specific HRQOL 0.27 0.18 0.61 0.02 Asthma control > Daytime sleepiness > Asthma specific HRQOL 0.11 0.14 0.38 0.14 Asthma control > Nighttime sleep quality > Daytime sleepiness > Asthma specific HRQOL 0.22 0.11 0.06 0.44 Total indirect effect 0.17 0.18 0.59 0.26 *p<0.05 ; Akaike information criterion (AIC): 22867.92; Bayesian information criterion (BIC): 23106.18.

PAGE 88

88 CHAPTER 4 THE INFLUENCE OF AMBIT AIR POLLUTION AND POLLEN EXPOSURE ON ASTHMA CONT ROL STATUS A M ONG ASTHMATIC CHILDREN Introduction Asthma is one of the most prevalent chronic diseases in children. Numerous studies have been conducted to investigate the effect of personal factors on asthma outcomes; however, the impact of environmental factors such as ambient air pollution and pollen on asthma control status has been relatively less explored. Air pollution materials including particulate matter (PM) and O 3 have been associated with poor asthma outcomes [56-59, 130]. Ground level O 3 is produced by chemical reactions between oxides of nitrogen and volatile organic compounds that come from industry operation and motor vehicles. PM can be solid particles such as dust and smoke or the liquid droplets such as mists and condensing vapor. Particles less than 2.5 micrometers (PM 2.5 ) or fine particles are mainly produced by industrial processes such as coal burning, automobile emission, forest fire and agricultural burning. Evidence suggests PM 2.5 and O 3 contribute to the adverse health outcomes in children such as decreased lung function, more asthma symptoms and increased asthma-related healthcare utilization [56, 59, 60, 131] because these air pollutants cause irritation and inflammation of the airways and jeopardize lung functioning [132-134] Besides PM and O 3 other air pollutants such as NO 2 CO and SO 2 are harmful for young children with developing lung functions and immune systems [135]. SO 2 is mainly produced by industry operation while NO 2 and CO 2 are primarily released by motor vehicles. Numerous studies have specifically linked increased exposure to NO 2 in early childhood with higher asthma incidence [136-139]. Previous studies reported

PAGE 89

89 exposure to SO 2 and CO was associated with higher prevalence of asthma and asthma symptoms such as wheeze in children [140, 141] In addition to air pollutants, pollen may contribute to allergy and respiratory diseases including asthma. For individuals with allergy, the immune system overreacts to allergens such as trees, grass and weed pollen and their bodies rele ase histamine that causes various asthma related symptoms such as coughing, wheezing, shortness of breath and chest tightness [61] Several studies have reported that higher pollen levels in grass, trees and weeds w ere associated with more frequent daily asthma symptoms such as wheezing, cough and chest tightness [142] and higher number of asthma related ED visits and hospitalizations [62, 143, 144] Although the relationship between air pollution exposure and pediatric asthma outcomes has been examined by previous studies, these studies contain some limitations to be addressed. First, prev ious studies mostly relied on the aggregate data collected from a single geographical location (e.g., an inner city) or hospital [57, 145 150] as the unit of analyses to examine the relationships of exposure to outdoor air pollution with asthma related health care utilization and pediatric asthma outcomes. As a result, the findings based on s ingle inner city or hospital samples are not generalizable to children with different characteristics who are living in other geographic locations. Additionally, previous studies focus on asthma related health care utilization as health outcome, and neglec t the genuine outcomes such as asthma symptom control, adherence to daily medication, asthma related emergency department (ED) visits, and patient reported outcomes (PROs). These additional outcomes reflect the importance of self management and daily funct ioning related to asthma. Second,

PAGE 90

90 previous studies mostly used the cross sectional framework to examine the association of exposure to air pollutants with asthma outcomes [57, 145 150] Because asthma outcomes are subject to the daily or seasonal change of air pollutant and pollen, the use of cross sectional study design inhibits our ability to investigate the effects of air pollution and pollen on asthmatic outcomes. Third, the collective effects of air pollution and pollen on asthma outcomes were poorly investigated by previous studies and results were still mixed. Several studies found that higher pollen counts rather than air pollution was associated with highest number of asthma related ED visits [62, 144] Other studies suggested ambient air pollution concentrations instead of pollen counts influenced pediatric asthma hospitalization [151, 152] Numerous studies reported increased exposure to both air pollution and pollen were significantly associated with more freque nt asthma related healthcare use [153 156] This study aimed to investigate the relationship between environmental exposure including outdoor air pol lution and pollen and asthma control reported across 26 weeks in a 2 year study period based on a group of asthmatic children in Florida State. The first aim was to examine the associations between air pollution including PM 2.5 and O 3 exposure and asthma c ontrol status in a group of asthma children; the second aim was to examine the effect of pollen on asthma control status; the third aim was to investigate the joint effect of air pollution and pollen on patient reported asthma control across 26 measurement occasions. The following hypotheses were proposed in this study: 1) increased exposure to air pollution including elevated PM 2.5 and O 3 concentration was associated with worse asthma control status; 2) increased exposure to pollen measured by pollen sever ity index was associated with worse asthma control status; 3) increased

PAGE 91

91 air pollution and pollen exposure was both associated with worse asthma control status; Florida State than the air pollution. Methods Participants and Data Collection This study is a secondary data analysis using data collected in the NIH PROMIS PAS. It was a 2 year longitudinal study conducted in Florida State to investigate the sensitivity to change of self reported outcome measures (e.g., HRQOL) corresponding to the change of asthma control status and associated factors. Potential study participants were identified from the Florida Medicaid and SCHIP and 238 dyads of children with asthma and their patie nts have consented and assented to participate in this study. Study Measures In each year, asthma control status was collected across 13 individual weeks using the Asthma Cont rol and Communication Instrument (ACCI). The ACCI has demonstrated satisfactory psychometric properties including concurrent validity, discriminant and known group validity [10] The ACCI is comprised of five items m easuring the concept e an asthma [10] The item scores across five individual items were

PAGE 92

92 summarized to represent the asthma control sum score for individual children with a range from 0 to 19; higher scores indicate worse asthma control status. Daily air quality measurement data of PM 2.5 and O 3 were obtained from the United States Environmental Protection Agency (EPA) Air Quality System (AQS) for 2010 to 2012 in all counties in Florida [157] The EPA AQS data conta in daily reports of air pollutants, dates of measurement, and the locations of specific monitoring sites by latitudes and longitudes. During the study period, 52 and 59 active monitoring sites collected PM 2.5 and O 3 data, respectively (Figure 4 1). The 201 5 digit zip code tabulation area was obtained from U.S. Census Bureau website [158] and ArcMap software (Version 10.2; ESRI, Redlands, California) was used to cal culate the centroid of each zip code where the study participants resided. The centroids of each zip code where the participants resided were linked to the nearest EPA monitoring sites based on the corresponding latitude and longitude. The distance between each centroid of zip codes to each air quality monitoring sites was calculated using ArcMap software and the shortest distance between monitoring sites and the centroid of zip codes where the participants resided was used to assign an EPA monitor to each participant for generating EPA air pollution measures. PM 2.5 and O 3 exposure was calculated as the mean concentrations of the seven days prior to the reporting of asthma control status. Daily pollen data corresponding to the study period was obtained from IMS Health [159] Pollen severity index, a continuous variable ranging from 0 to 12 reflecting the severity of pollen exposure, was used to capture the pollen level. Higher index indicates worse p ollen severity. IMS Health collected daily pollen index in 9 geographic

PAGE 93

93 areas in Florida, including Daytona Beach, Key West, Jacksonville, Orlando, Miami, West Palm Beach, Pensacola, Tallahassee, and Tampa. The study subjects were linked to the IMS Health pollen data based on the zip code where participants resided. Pollen exposure was calculated as the mean pollen severity index of the seven days prior to the reporting of asthma control. Daily temperature and precipitation data in each county in Florida du ring the study period were obtained from the National Climatic Data Center at National Oceanic and Atmospheric Administration (NOAA) [160] In this study, participants consist of tw o cohorts of children and parents. The first cohort was investigated in fall season (from September to December) of 2010 and 2011, and the second cohort was investigated in spring season (from February to May) of 2011 and 2012; therefore, the inclusion of these two cohorts potentially reflects the variation of asthma flare season in Florida. The following variables were collected at baseline of the first year and treated as covariates umber of chronic smoking status at home Statistical Analysis Descriptive analyses were performed to analyze the distribution of socio demographic characteristics among study participants ( n =238). The percentage or education background, family income and smoking status at home were examined. In addition, descriptive analyses were conducted to examine the mean, SD, range and interquartile range (IQR) of air pollution including PM 2.5 and O 3 pollen severity index,

PAGE 94

94 asthma control status scores. IQR is defined as the difference between the third quartile (Q3) and the first quartile (Q1) calculated from descriptive analyses. Bivariate analyses were performed to examine the associations between ambient air pollution an d pollen exposure. ANOVA tests were performed to examine if PM 2.5 and O 3 concentrations were significantly different across EPA monitoring sites by individual weeks. ANOVA tests were also conducted to test if pollen levels were significantly different acro ss 8 areas in Florida by individual weeks (Appendix C). Pearson correlation coefficients were reported for associations among PM 2.5 O 3 and pollen by individual weeks. Linear mixed effect models were performed to examine the associations of weekly asthma control scores with ambient air pollution and pollen levels. The main individual envir onmental factors (independent variables) on asthma control status (dependent variable) using three statistical models. The first set of models only included individual environmental factor such as PM 2.5 O 3 or pollen, respectively, as the main effect. The second set of models included the individual environmental factors as the main effect and specific cohort indicator as a covariate. The third model included the character the joint effects of PM 2.5 O 3 and pollen on asthma control scores. In the mixed effect

PAGE 95

95 models, PM 2.5 O 3 and pollen were modeled as fixed effects and subjects were modeled as random effects. The changes in asthma control sum scores per IQR increase in PM 2.5 O 3 and pollen were reported. We performed sensitivity analyses to examine the robustness o f the joint effects of PM 2.5 O 3 and pollen on asthma control status by taking into account different per week and Carolina, USA). Re sults Study Participant Characteristics 1. For children at baseline of the first year (N=238), the mean age was 12.23 years old (SD: 2.58); 59.92% were boys; 38.40% were non Hispanic White; 44.24% were overweight; the mean number of chronic conditions was 1.53 (SD: 0.83). For parents, the mean age was 40.72 years old (SD: 8.81); the majority of them were married/living with partners (50.63%); the most of them had education background of some college, associated degree or college degrees (61. 11%), and had family income between $15,000 and $35,000 (44.73%). The majority of the parents smoked at home (82.28%). The number of participants and observations

PAGE 96

96 number of monitoring days per week were reported in Table C 1 Appendix C. Descriptive Analyses for Study Measures Table 4 2 shows the mean, SD, range and IQR for PM 2.5 O 3 pollen, temperature, precipitation and asthma control scores by specific cohorts. For coho rt 1 that was investigated in fall season of 2010 and 2011, the mean of PM 2.5 was 7.58 3 (SD: 2.74); the mean of O 3 was 0.04 ppm (SD: 0.01); the mean of pollen severity index was 4.59 (SD: 1.74); the mean of temperature was 27.45 Celsius (SD: 4.01); th e mean of precipitation was 2.95 mm (SD: 4.74); the mean of asthma control scores was 2.64 (SD: 3.66). For cohort 2 that was investigated in spring season of 2011 and 2012, the mean of PM 2.5 3 (SD: 2.51); the mean of O 3 was 0.04 ppm (SD: 0.01) ; the mean of pollen severity index was 8.76 (SD: 1.55); the mean of temperature was 26.90 Celsius (SD: 3.32); the mean of precipitation was 2.23 mm (SD: 3.85); the mean of asthma control scores was 2.57 (SD: 3.46). Correlations among PM 2.5 O 3 and pollen were reported in Table C 3 Appendix C. Additionally, correlations of asthma control sum scores with PM 2.5 O 3 and pollen levels were presented in Table C 4 Appendix C. Multivariate Associations among Air Pollution Concentration, Pollen Severity Index and A sthma Control Scores Table 4 3 shows the associations among PM 2.5 exposure, asthma control scores using linear mixed effect models. Model 1 shows PM 2.5 was positively associate d with control scores per one IQR increase in PM 2.5 Model 2 shows higher PM 2.5 concentration was significantly associated with worse asthma control after adjusting for

PAGE 97

97 2.5 was positively associated with likely to have worse asthma control compared to children with fewer chronic conditions education of control compared to married/living with partner status and education of college degree or above. Table 4 4 shows the associations among O 3 exposure, asthma control scores and subject using linear mixed effect models. Model 1 shows O 3 was negatively associated with 0.02, p>0.05). Model 2 shows O 3 concentration was negatively associated with a 0.02, p>0.05). Model 3 shows O 3 was negatively associated with asthma control scores demographics. In Model 3, childr en with more chronic conditions were more likely to control, where unmarried status

PAGE 98

98 Table 4 5 shows the associations among pollen exposure, asthma control scores using linear mixed effect models. Model 1 shows higher pollen severity index was significantly associated with worse asthma control scor shows higher pollen exposure index was significantly associated with worse asthma severity index was significantly associated with worse asthma control scores after suggesting there was a 0.35 increase in asthma control score per one IQR increase in pollen severity index. Children with more chronic condition s were more likely to have ation of high school or married/living with partner status and college degree education or above. Table 4 6 shows the joint effects of PM 2.5 O 3 and pollen on asthma control sco res across 26 measurement occasions using linear mixed effect models. Model 1 shows higher PM 2.5 association with worse asthma control status, while higher O 3 was marginally associated with 0.16, p<0.1). Model 2 shows higher PM 2.5 associated with worse asthma control after adjusting for the season effect, whereas higher O 3 exposure wa s marginally associated better asthma control scores after

PAGE 99

99 0.16, p<0.1). Model 3 shows higher PM 2.5 control scores after a whereas higher O 3 were more likely to hav e worse asthma control compared to children with less chronic asthma control status compared to parents with married/living with partner status and college degree or above. Sensitivity Analysis Results Table 4 7 shows sensitivity analyses for the joint effects of PM 2.5 O 3 and pollen on asthma control status based on data derived from dif ferent combinations of reveal that higher PM 2.5 exposure was significantly associated with higher asthma control scores (all p<0.01), which suggests increased exposure to P M 2.5 was associated with worse asthma control. O 3 exposure was all negatively associated with asthma km (p< 0.1). Greater pollen exposure was all significantly associated with higher km km (p>0.1). Discussion This study examined the associations of asthma control with ambient air pollution including PM 2.5 O 3 and pollen exposures in a longitudinal framework. Our results reveal

PAGE 100

100 that higher levels of PM 2.5 and pollen are significantly associated with worse asthma control st atus. However, elevated O 3 exposure is marginally associated with better asthma control scores, suggesting increased exposure to O 3 may play a protective role in asthma control. Consistent with findings from previous literature, we found that greater numbe r of chronic conditions was associated with worse asthma control, and parents with high school education or below and unmarried status were more likely to report worse asthma control than parents with education background of college or advanced degree and married status [1, 2] Our results show increased exposure to O 3 was associated with better asthma control status and this association was marginally significant in the main analyses using two po llutants and pollen as predictors (p<0.1) and statistically significant in the sensitivity analyses (p<0.05). Our findings reveal a protective role of elevated O 3 protective effect of O 3 on several asthma outcomes including more asthma symptoms [161] asthma exacerbation [162] more number of pediatric asthma related visits [163, 164] and incidence of other diseases such as pulmonary diseases [165, 166] and cardiovascular diseases [167, 168] Those observed associations mostly occurred during the season of lower temperature that is usually accompanied by relatively low ozone levels compared to the season of higher temperature. Numerous studies have reported that O 3 concentration was much higher in warmer season [169, 170] Both in vivo and in vitro studies assessed the harmful effects of O 3 exposure on human airway cells. To observe the harmful effect of O 3 previous studies exposed study participants or airway cells into a 0.25 1.00 ppm O 3 environment for 2 6 hours [171, 172] In contrast

PAGE 101

101 to the U.S. EPA national ambient air quality standard (NAAQS) f or O 3 (0.075 ppm), the mean of O 3 during the study period is much lower (0.04 ppm). It is possible that the association between elevated O 3 exposure and better asthma control status observed in this study is due to the low O 3 concentration. A recent review study found that exposure to O 3 was associated with asthma incidence for children with frequent outdoor activities and exposure to high level of O 3 [173] Future studies may investigate asthma control and O 3 concentration data during summer among asthmatic children who are physically active in Florida to investigate if the effects of O 3 on asthma control among children remain protective. Another reason that may partially explain the association of elevated O 3 with of O 3 with other air pollutants including NO 2 CO and SO 2 and the effects of those air pollutants on asthma control status. Previous studies have show n O 3 con centrations were negatively correlated with air poll ution such as NO 2, CO and SO 2 [148, 174 178] and NO 2, CO a nd SO 2 may confound the association between O 3 and asthma control status. Due to the limited number of monitors for other air pollution in Florida, we were not able to examine the effects of CO, NO 2 SO 2 and PM 10 possible tha t including multiple air pollutants in the analyses will change the associations between increased O 3 concentrations and better asthma control status. Studies are needed to investigate the influence of exposure to multiple ambient air pollution on asthma c ontrol status in the future. Sensitive analyses have been frequently performed in previous air pollution studies to understand whether analyzing different distances between EPA monitoring

PAGE 102

102 th outcomes [179 181] However, very few studies have evaluated whether the use of different monitoring days per week will generate different results. In our study, although air pollution concentrations were calculated as mean levels in the seven days prior to reporting asthma control status, the actual dates of having asthma attack/exacerbations for an individual corresponding to the dates of air pollution and pollen exposure were still in th e previous week. Therefore, we are not able to identify the common dates that capture air pollution and pollen exposure and the occurrences of asthma attack/ exacerbations. In addition, air pollution data were not collected every day and number of monitori ng days varied each week. As a result, the inclusion of greater number of monitoring days per week would decrease the sample size and loss the statistical power to detect the difference. Using greater than 3 monitoring days per week as the exposure window was perhaps the optimal approach to examine the effects of environmental exposure on asthma control. our sensitivity analyses found greater effects of PM 2.5 and O 3 exposure we re observed location, yet this pattern was not the same for pollen exposure. P ollen severity index was created at 9 areas in Florida; therefore, distances between EPA monitors a nd 2.5 and O 3 rather than pollen Future studies are encouraged to collect pollen data in smaller geographic areas such as counties to test the influence of distance on the relationship of pollen

PAGE 103

103 aggregated by different classes (e.g., tree pollen, grass pollen and weed pollen), which is deemed as a summary measure of pollen. Future studies should distinct diffe rent sources of pollen on asthma control status to benefit children who are allergic to different botanic pollen. Given the harmful effects of ambient air pollution (especially PM 2.5 and O 3 ) and pollen on asthmatic outcomes, an important strategy to impro ve asthma control is to prevent the exposure to these ambient materials. Although the NAEPP EPR 3 have emphasized the control for environmental factors including allergens and irritants, NAEPP guideline did not evaluate the significance of environmental ri sk factors on asthma control status and health outcomes. Physicians and health care agencies need to alert asthmatic children and their parents about the harmful effect of high level of air pollution and pollen on asthmatic outcomes. Parents and school tea chers should keep children away from unnecessary outdoor activities during the days of poor air quality or emission of PM 2.5 and O 3 [173, 182] Governments should motivate the public to use of plug in hybrid electric vehicles to decrease the emission of air pollutants [183] Limitations Several limitations should be considered for this study. First, the study participants were from family with low economic status; our findings may be not generalizable to more representative population. Second, the sources and severity of indoor pollutants (such as radon, PM and carbon monoxide created from antique cooking and heating in household) were not collected in NIH PROMIS PAS. These variables are potential confounding factors to the associations of outdoor air pollutants

PAGE 104

104 and pollen on asthma control status. Previous studies hav e reported children with asthma were more likely to have exposure to environmental tobacco smoke (ETS), tobacco product, than children without asthma [184 186] A recent national survey showed approximately 50% of children with asthma were exposed to ETS and 18% of them had in home exposure to smoking [186] In o ur study, parental smoking status at and this item may not capture the true parental smoking status at home. Third, environmental exposure data were obtained from EPA and IMS, w hich were created at aggregate level rather than individual subject level. The locations of EPA air quality monitoring sites in Florida are determined by the population density, and subjects living in rural area may not have EPA monitoring sites available to measure air pollution exposure. Future studies may collect data in person for participants living in areas without closely monitoring for measuring air quality. Additionally, as discussed, pollen data are aggregated at diffe rent classes of pollen (weed, tree, etc. ) which is deemed as a proxy pollen measure. Conclusion This study found that increased exposures to PM 2.5 and pollen were associated control status, healthca re providers and agencies need to inform the risks of outdoor ambient air pollution and pollen levels for individual asthma children, and develop strategies to prevent asthmatic children from unnecessary exposure to harmful outdoor environmental factors.

PAGE 105

105 Figure 4 1. Locations of study participants and air pollution monitors in Florida, USA

PAGE 106

106 Table 4 1. Study participant characteristics Number of subject (%) or mean (SD) 12.23 (2.58) Boys 142 (59.92%) Girls 95 (40.08%) White/non Hispanic 91 (38.40%) Black/non Hispanic 61 (25.74%) Hispanic 64 (27.00%) Other 21 (8.86%) Hyperactivity or attention deficit disorder/ ADD or ADHD 41 (17.30%) Born premature 27 (11.39%) Mental health conditions: depression, anxiety, bipolar disorder and other 8 (3.38%) Epilepsy or other seizure disorders 6 (2.53%) disorder 5 (2.11%) Deaf or hard of hearing 5 (2.11%) 40.72 (8.81) White/non Hispanic 102 (43.04%) Black/non Hispanic 62 (26.16%) Hispanic 60 (25.32%) Other 13 (5.49%) education background, % High school or below 75 (32.05%) Some college / technical/associated degree and college degree 143 (61.11%) Advanced degree 16 (6.84%) ,% Never married 42 (17.72%) Married 120 (50.63%) Living with partner in committed relations 10 (4.22%) Separated 10 (4.22%) Divorced 48 (20.25%) Widowed 7 (2.95%) Yes 195 (82.28%) No 42 (17.72%)

PAGE 107

107 Table 4 2. Summary statistics of air pollution concentrations, pollen severity index and asthma control scores between September 2010 and May 2012 in Florida Cohort Variable Unit N Mean SD Minimum Maximum IQR Cohort 1 (Fall) PM 2 5 3 1725 7.58 2.74 2.71 25.80 2.93 O 3 ppm 1739 0.04 0.01 0.01 0.06 0.01 Pollen Index score 1789 4.59 1.74 0.49 9.39 2.47 Temperature Celsius 1752 27.45 4.01 11.65 35.96 4.80 Precipitation mm 1789 2.95 4.74 0.00 33.90 3.59 Asthma control Summation score 1786 2.64 3.66 0.00 18.00 4.00 Cohort 2 (Spring) PM 2 5 3 1361 8.73 2.51 2.85 25.47 3.20 O 3 ppm 1388 0.04 0.01 0.01 0.06 0.01 Pollen Index score 1408 8.76 1.55 3.21 11.00 1.96 Temperature Celsius 1382 26.90 3.32 11.43 34.10 3.89 Precipitation mm 1408 2.23 3.85 0.00 35.86 2.64 Asthma control Summation score 1407 2.57 3.46 0.00 18.00 4.00

PAGE 108

108 Table 4 3. Associations between asthma control summation scores and preceding 7 days average PM 2 5 concentrations PM 2 5 locations ( number of subjects= 150 with 2279 observation s ) Model 1 Model 2 Model 3 Beta (SE) 95% CI Beta (SE) 95% CI Beta (SE) 95% CI PM 2 5 0.31*** (0.08) [0.15, 0.46 ] 0.31*** (0.08) [0.15, 0.46] 0.32*** (0.08) [0.16, 0.48 ] Season (Ref: Fall) Spring 0.16 (0.44) [ 1.03, 0.71] 0.23 (0.43) [ 0.61, 1.07 ] 0.06 (0.09) [ 0.24, 0.12 ] (Ref: Girls) Boys 0.47 (0.40) [ 0.32, 1.26 ] (Ref: White) Black 0.78 (0.51) [ 0.23, 1.79 ] Hispanics 0.82 (0.49) [ 0.15, 1.78 ] Others 0.06 (0.76) [ 1.56, 1.44 ] 0.03 (0.03) [ 0.08, 0.03 ] Marital status (Ref: Married) Not married 0.96* (0.41) [0.15, 1.76 ] Education (Ref: College or above) High school or below 0.94* (0.44) [0.07, 1.81 ] Number of chronic conditions 0.80** (0.25) [0.32, 1.29 ] p<0.1; *p<0.5; **p<0.01; ***p<0.001.

PAGE 109

109 Table 4 4. Associations between asthma control summation scores and preceding 7 days average O 3 concentrations 3 number of subjects= 159 with 2728 observation s ) Model 1 Model 2 Model 3 Beta (SE) 95% CI Beta (SE) 95% CI Beta (SE) 95% CI O 3 0.02 (0.07) [ 0.15, 0.11] 0.02 (0.07) [ 0.15, 0.11] 0.02 (0.07) [ 0.15, 0.12] Season (Ref: Fall) Spring 0.01 (0.42) [ 0.83, 0.82] 0.15 (0.40) [ 0.62, 0.93] 0.01 (0.08) [ 0.17, 0.16] (Ref: Girls) Boys 0.43 (0.39) [ 0.33, 1.19] (Ref: White) Black 0.47 (0.49) [ 0.49, 1.42] Hispanics 0.30 (0.47) [ 0.63, 1.22] Others 0.16 (0.76) [ 1.65, 1.33] 0.03 (0.03) [ 0.08, 0.02] Marital status (Ref: Married) Not married 0.80* (0.39) [0.03, 1.57] Education (Ref: College or above) High school or below 1.04* (0.43) [0.20, 1.89] Number of chronic conditions 0.70** (0.25) [0.22, 1.19] *p<0.05, **p<0.01, ***P<0.001

PAGE 110

110 Table 4 5. Associations between asthma control summation scores and preceding 7 days average pollen severity index ( number of subjects =184 with 3193 observation s ) Model 1 Model 2 Model 3 Beta (SE) 95% CI Beta (SE) 95% CI Beta (SE) 95% CI Pollen 0.31* (0.14) [0.03, 0.59] 0.34* (0.15) [0.05, 0.64] 0.35* (0.15) [0.05, 0.65] Season (Ref: Fall) Spring 0.32 (0.40) [ 1.10, 0.46] 0.19 (0.37) [ 0.93, 0.54] 0.01 (0.07) [ 0.16, 0.14] (Ref: Girls) Boys 0.49 (0.35) [ 0.19, 1.17] (Ref: White) Black 0.56 (0.44) [ 0.30, 1.43] Hispanics 0.48 (0.43) [ 0.36, 1.31] Others 0.19 (0.70) [ 1.55, 1.18] 0.03 (0.02) [ 0.07, 0.02] Marital status (Ref: Married) Not married 0.78* (0.36) [0.08, 1.48] Education (Ref: College or above) High school or below 0.84* (0.37) [0.11, 1.57] Number of chronic conditions 0.65** (0.21) [0.23, 1.07] *p<0.05, **p<0.01, *** p <0.001

PAGE 111

111 Table 4 6. Associations between asthma control summation scores and preceding 7 days average air pollution and pollen 2.5 and O 3 locations (number of subjects=141 with 2151 observations) Model 1 Model 2 Model 3 Beta (SE) 95% CI Beta (SE) 95% CI Beta (SE) 95% CI PM 2 5 0.35*** (0.10) [0.16, 0.55] 0.35*** (0.10) [0.16, 0.55] 0.37*** (0.10) [0.17, 0.56] O 3 0.16 (0.09) [ 0.34, 0.02] 0.16 (0.09) [ 0.34, 0.02] 0.16 (0.09) [ 0.34, 0.02] Pollen 0.38* (0.18) [0.02, 0.74] 0.44* (0.19) [0.06, 0.82] 0.44* (0.19) [0.06, 0.83] Season (Ref: Fall) Spring 0.47 (0.49) [ 1.44, 0.50] 0.03 (0.47) [ 0.96, 0.90] 0.05 (0.10) [ 0.24, 0.13] (Ref: Girls) Boys 0.30 (0.42) [ 0.53, 1.13] (Ref: White) Black 0.89 (0.54) [ 0.16, 1.94] Hispanics 0.82 (0.51) [ 0.18, 1.83] Others 0.06 (0.78) [ 1.60, 1.48] 0.03 (0.03) [ 0.08, 0.02] Marital status (Ref: Married) Not married 1.01* (0.42) [0.17, 1.84] Education (Ref: College or above) High school or below 1.05* (0.47) [0.13, 1.97] Number of chronic conditions 0.90*** (0.27) [0.38, 1.43] *p<0.05, **p<0.01, *** p <0.001

PAGE 112

112 Table 4 7. Sensitivity analyses for associations between asthma control summation scores and air pollution and pollen exposure by different number of monitoring days per week and nearest EPA monito rs PM 2 5 O 3 Pollen Beta (SE) 95% CI Beta (SE) 95% CI Beta (SE) 95% CI 0.32** (0.11) [0.11,0.53] 0.20* (0.09) [ 0.39, 0.02] 0.42 (0.22) [ 0.02, 0.86] 0.31*** (0.09) [0.13, 0.49] 0.16 (0.08) [ 0.32, 0.01] 0.50** (0.19) [0.13, 0.86] 0.22** (0.08) [0.07, 0.38] 0.16* (0.08) [ 0.31, 0.01] 0.37* (0.16) [0.05, 0.68] 0.37** (0.12) [0.13, 0.61] 0.21* (0.10) [ 0.41, 0.01] 0.38 (0.24) [ 0.09, 0.84] 0.37*** (0.10) [0.17, 0.56] 0.16 (0.09) [ 0.34, 0.02] 0.44* (0.19) [0.06, 0.83] 0.26** (0.09) [0.08, 0.43] 0.17* (0.08) [ 0.33, 0.01] 0.35* (0.17) [0.02, 0.68] 0.42** (0.13) [0.16, 0.67] 0.23* (0.11) [ 0.44, 0.02] 0.35 (0.25) [ 0.13, 0.84] 0.40*** (0.11) [0.19, 0.62] 0.19* (0.10) [ 0.38, 0.01] 0.45* (0.20) [0.06, 0.85] 0.31** (0.09) [0.12, 0.49] 0.23** (0.09) [ 0.40, 0.06] 0.36* (0.18) [0.01, 0.70] *p<0.05, **p<0.01, *** p <0.001

PAGE 113

113 CHAPTER 5 IMPROVING LONG TERM HEALTH OUTCOMES IN CHILDREN WITH ASTHMA: DISCUSSION OF RESULTS Review of Study Findings This dissertation research is a secondary data analysis which used data collected from the NIH PROMIS PAS to address several research questions related to health outcomes among asthmatic children. Those limitations of previous research included 1) limited evidence is available on the relationship of changes in asthma control s tatus with variations of asthma specific HRQOL over time, 2) a lack of evidence on how the sleep problems influence the association between asthma control and asthma specific HRQOL, and 3) very few studies have been conducted to investigate the joint effec ts of ambient air pollution and pollen exposure on long term asthma control status over time. Three specific aims were proposed in this dissertation to address those limitations. The first aim was to describe a trajectory of asthma specific HRQOL over a 2 year observational period and investigate factors associated with initial status and rate of changes in asthma specific HRQOL; the second aim was to investigate the influence of nighttime sleep quality and daytime sleepiness on the relationship between ast hma control status and HRQOL over a 2 year observational period; and the third aim was to examine the impact of ambient air pollution and pollen exposure on asthma control status across 26 measurement occasions. For aim 1, unconditional LGM was applied to examine the trend of HRQOL across 4 measurement occasions and the conditional LGM was conducted to investigate the factors associated with the intercept and rate of change of HRQOL. The findings suggest that there was a significant improvement in asthma sp ecific HRQOL over 2 years, and this improved trend was explained by asthma control status and

PAGE 114

114 was related to low initial level of asthma specific HRQOL and Hispanic childre n were more likely to have slower rate of change in asthma specific HRQOL. For aim 2, multilevel SEM was performed to investigate the mediating effects of nighttime sleep quality and daytime sleepiness on the pathways between asthma control status and HRQO L across 4 repeated measures. Asthma control status was found to be significantly associated with asthma specific HRQOL at both between subject level and within subject level. However, the association of asthma control status with asthma specific HRQOL was significantly mediated by the daytime sleepiness at within subject level. Both daytime sleepiness and nighttime sleepiness influenced the asthma control HRQOL pathway at between subject (or population) level. For aim 3, mixed effect models were applied to investigate the impact of ambient air pollution and pollen exposure on asthma control status across 26 measurement occasions. The results suggest that increased exposure to PM 2.5 and pollen was significantly associated with poor asthma control status over time. This dissertation provides evidence for researchers, clinicians and policy makers to design specific interventions targeting risk factors of asthma control status and HRQOL to improve long term health outcomes among asthmatic children. Asthma contro l status and HRQOL assessments have been used as the primary and secondary endpoint in clinical studies for asthmatics [13, 14] and the 2007 NAEPP EPR 3 guideline [11] recommends use of PRO tools to assess asthma control status and HRQOL. The results of this dissertation have demonstrated that poor asthma control status was the primary risk factor contributing to the vari ation of HRQOL in the

PAGE 115

115 long run among asthmatic children. In addition, the study results show the important roles of nighttime sleep quality and daytime sleepiness in the pathway between asthma control status and HRQOL. This finding suggests a need of fast screening of sleep problems for clinicians to identify potential sleep problems associated with poorly controlled asthma status and impaired HRQOL among asthmatic children. The increased exposure to ambient air pollution such as PM 2.5 a n d pollen has been linked to poorly controlled asthma status over time in this dissertation. The results provided useful information for health policy makers and agencies to establish local warning systems on a daily basis related to air pollution and high p ollen exposure for asthmatics to avoid the unnecessary exposure. Limitations Se veral limitations in this dissertation work should be noted when interpreting the results. First, the study population was recruited from Florida Medicaid and SCHIP, and the par ticipants were mostly from families with low economic status. Therefore, the study results may not be generalizable to the general population. Second, given the nature of secondary data analysis for this dissertation, several desired variables were not col lected in the NIH PROMIS PAS. For example, objective measures of sleep duration was not collected in the parent study, and the mediating roles of sleep problems does not include sleep quantity and this variable may explain the relationship between asthma c ontrol status and asthma specific HRQOL. Moreover, indoor air quality was not assessed, and indoor air pollutants together with outdoor air pollution and pollen may contribute to poor asthma control status. Third, this dissertation used aggregate level of data obtained the U S EPA AQS and IMS Health. The EPA air quality monitoring sites are established based on

PAGE 116

116 population density. Therefore, the study participants living in areas without close monitoring sites were not included in our study. Additionally, IMS Health collected pollen data in Florida across 9 geographic areas, and the pollen severity index rather than specific type of pollen indexes was used. Future studies need to design feasible method to collect air pollution data for people living in rura l areas without EPA air quality monitoring sites. Conclusion This dissertation research used secondary data collected from the NIH PROMIS PAS to examine the changes of asthma related health outcomes over two years among 238 dyads of asthmatic children and their parents. The results of this dissertation provided rich information to help clinicians and healthcare policy makers understand the natural trajectory of pediatric asthma specific HRQOL over time and factors associated with the changes of asthma contr ol status and HRQOL among asthmatic children. This study generated a foundation for future studies to design specific interventions targeting on those risk factors to maintain well controlled asthma status and improve long term HRQOL of asthmatic children.

PAGE 117

117 APPENDIX A CHAPTER 2 INTER CORRELATION OF ASTHMA CONTROL STATUS AND HRQOL Table A 1. Inter correlation of asthma control status and asthma specific HRQOL across 4 time points T1 T2 T3 T4 Asthma control T1 T2 0.31*** T3 0.39*** 0.09 T4 0.06 0.05 0.22** Asthma specific HRQOL T1 T2 0.64*** T3 0.54*** 0.44*** T4 0.46*** 0.37*** 0.68*** p <0.5; ** p <0.01; *** p<0.001.

PAGE 118

118 APPENDIX B CHAPTER 3 DETAILED METHODS OF MULTILEVEL STRUCTURAL EQUATION MODELING Multilevel Mediation Models The PROMIS PAS was a 2 year longitudinal study which included repeated measures of asthma control status, nighttime sleep quality, daytime sleepiness and HRQOL over 2 years observational period. Multilevel structural equation modeling (MS EM) was applied to explore the influence of sleep problems including nighttime sleep quality and daytime sleepiness on the association between asthma control status and HRQOL at both within subject and between subject level. Mplus Sample Code for MSEM ANAL YSIS: TYPE = twolevel random; MODEL: %WITHIN% SFAsth on quality (bw1); SFAsth on sleepi (bw2); c|SFAsth on ac18; quality on ac18 (aw1); sleepi on ac18 (aw2); sleepi on quality (aw3); %BETWEEN% c quality sleepi SFAsth; c with quality sleepi SFAsth; SFAsth on quality (bb1); SFAsth on sleepi (bb2); SFAsth on ac18 (contextual); quality on ac18 (ab1); sleepi on ac18 (ab2); sleepi on quality (ab3); [c] (wd); quality on cage page chealthn rprela2c craced1 craced2 craced3 rmgrade rcgend; sleepi on cage page cheal thn rprela2c craced1 craced2 craced3 rmgrade rcgend; SFAsth on cage page chealthn rprela2c craced1 craced2 craced3 rmgrade rcgend;

PAGE 119

119 APPENDIX C CHAPTER 4 DETAILED RESULTS FOR AIR POLLUTION AND POLLEN Table C 1. Number of subjects and observations at dif ferent ranges of distance PM 2 5 O 3 Number of subjects Number of observations Number of subjects Number of observations Distance 114 1920 137 2369 155 2629 159 2741 184 3083 183 3123 Monitoring days 184 3083 183 3123 178 2693 183 3109 154 2509 183 3071

PAGE 120

120 Table C 2. PM 2.5 and O 3 concentrations at EPA air quality monitors and pollen severity index in 8 areas in Florida at 26 weeks during the study period PM 2 5 O 3 Pollen F value (P value) F value (P value) F value (P value) Week1 6.08*** 2.14** 15.91*** Week2 1.96** 1.48 9.41*** Week3 3.66*** 1.29 8.83*** Week4 5.64*** 1.54* 6.30*** Week5 2.50*** 1.86** 5.79*** Week6 3.92*** 2.11** 7.18*** Week7 1.53 1.66* 3.99*** Week8 2.57*** 1.51 C 2.19* Week9 4.35*** 1.72* 1.85 Week10 1.59* 1.06 2.20 Week11 2.26** 2.92*** 2.50* Week12 2.93*** 2.52*** 3.68** Week13 3.85*** 1.32 8.91*** Week14 3.89*** 3.73*** 11.43*** Week15 5.21*** 2.35*** 44.58*** Week16 4.96*** 2.46*** 13.94*** Week17 2.29** 1.68* 6.29*** Week18 3.26*** 1.22 3.38** Week19 11.65*** 4.28*** 3.42** Week20 3.39*** 1.92** 3.84*** Week21 8.30*** 2.03** 3.27** Week22 4.73*** 1.41 2.45* Week23 1.58 1.14 1.94 Week24 1.61* 0.97 2.46* Week25 2.21** 1.66* 2.00 Week26 3.97*** 1.37 6.07*** Overall 25.66*** 8.09*** 81.03*** p <0.1; p < 0.5; ** p <0.01; *** p <0.001

PAGE 121

121 Table C 3. coefficients among PM 2.5 O 3 and pollen at 26 weeks during the study period PM 2 5 and O 3 PM 2 5 and Pollen O 3 and Pollen Week1 0.45*** 0.50*** 0.35*** Week2 0.74*** 0.65*** 0.73*** Week3 0.18* 0.57*** 0.12 Week4 0.60*** 0.09 0.28** Week5 0.66*** 0.38*** 0.25** Week6 0.59*** 0.38*** 0.77*** Week7 0.80*** 0.83*** 0.89*** Week8 0.16 0.30*** 0.06 Week9 0.17 0.01 0.46*** Week10 0.40*** 0.54*** 0.74*** Week11 0.65*** 0.60*** 0.72*** Week12 0.03 0.33*** 0.23* Week13 0.19* 0.14 0.77*** Week14 0.72*** 0.22* 0.50*** Week15 0.64*** 0.55*** 0.66*** Week16 0.50*** 0.41*** 0.63*** Week17 0.06 0.50*** 0.26** Week18 0.08 0.09 0.44*** Week19 0.09 0.28** 0.48*** Week20 0.36*** 0.40*** 0.01 Week21 0.16 0.12 0.63*** Week22 0.44*** 0.50*** 0.87*** Week23 0.49*** 0.45*** 0.84*** Week24 0.83*** 0.85*** 0.94*** Week25 0.23* 0.33*** 0.88*** Week26 0.38*** 0.25** 0.79*** p <0.1; p <0.5; ** p <0.01; *** p <0.001

PAGE 122

122 Table C 4. Correlations among asthma control summation scores, PM 2 5 O 3 and pollen concentration s at 26 weeks during the study period ACCI score and PM 2 5 ACCI score and O 3 ACCI score and Pollen Week1 0.05 0.08 0.24** Week2 0.13 0.10 0.13 Week3 0.06 0.01 0.03 Week4 0.17 0.20* 0.08 Week5 0.03 0.04 0.12 Week6 0.09 0.12 0.04 Week7 0.09 0.12 0.09 Week8 0.03 0.05 0.07 Week9 0.12 0.04 0.10 Week10 0.01 0.03 0.04 Week11 0.03 0.03 0.06 Week12 0.06 0.01 0.19* Week13 0.18 0.07 0.18 Week14 0.08 0.11 0.01 Week15 0.06 0.11 0.07 Week16 0.05 0.09 0.11 Week17 0.00 0.01 0.08 Week18 0.02 0.05 0.01 Week19 0.04 0.10 0.09 Week20 0.17 0.10 0.05 Week21 0.10 0.06 0.06 Week22 0.20* 0.02 0.03 Week23 0.01 0.05 0.04 Week24 0.05 0.04 0.07 Week25 0.17 0.14 0.11 Week26 0.06 0.04 0.03 p <0.1; p <0.5; ** p <0.01; *** p <0.001

PAGE 123

123 LIST OF REFERENCE S [1] Akinbami LJ, Moorman JE, Bailey C, Zahran HS, King M, Johnson CA, et al. Trends in asthma prevalence, health care use, and mortality in the united states, 2001 2010 NCHS Data Brief. 2012 May; (94):1 8. [2] Bloom B, Jones LI, Freeman G. Summary health statistics for U.S. children: National health interview survey, 2012. Vita l Health Stat 10. 2013 Dec; (258):1 81. [3] Jang J, Gary Chan KC, Huang H, Sullivan SD. Trends in cost and outcomes amon g adult and pediatric patients with asthma: 2000 2009. Ann Allergy Asthma Immunol. 2013 Dec;111(6):516 22. [4] Asthma and Allergy Foundation of America. Asthma facts and figures. http://www aafa org/display cfm?id=8&s ub=42#_ftn20 [5] Florida department of health, division of community health promotion, bureau of chronic disease prevention, Florida asthma program. Burden of asthma in Florida 2013. [6] Busse WW, Morgan WJ, Taggart V, Togias A. Asthma outcomes workshop : Overview. J Allergy Clin Immunol. 2012 Mar;129(3 Suppl):S1 8. [7] Parulekar AD, Alobaidy A, Hanania NA. Asthma outcomes revisited. Curr Opin Pulm Med. 2013 Jan;19(1):6 12. [8] Liu AH, Zeiger R, Sorkness C, Mahr T, Ostrom N, Burgess S, et al. Development and cross sectional validation of the childhood asthma control test. J Allergy Clin Immunol. 2007 Apr;119(4):817 25. [9] Juniper EF, O'Byrne PM, Guyatt GH, Ferrie PJ, King DR. Development and validation of a questionnaire to measure asthma control. Eur Res pir J. 1999 Oct;14(4):902 7. [10] Patino CM, Okelo SO, Rand CS, Riekert KA, Krishnan JA, Thompson K, et al. The asthma control and communication instrument: A clinical tool developed for ethnically diverse populations. J Allergy Clin Immunol. 2008 Nov;122( 5):936,943.e6. [11] National Asthma Education and Prevention Program. Expert panel report 3 (EPR 3): Guidelines for the diagnosis and management of asthma summary report 2007. J Allergy Clin Immunol. 2007 Nov;120(5 Suppl):S94 138. [12] Horner SD, Brown A. Evaluating the effect of an asthma self management intervention for rural families. J Asthma. 2014 Mar;51(2):168 77.

PAGE 124

124 [13] Kintner EK, Sikorskii A. Randomized clinical trial of a school based academic and counseling program for older school age students. Nu rs Res. 2009 Sep Oct;58(5):321 31. [14] Burkhart PV, Rayens MK, Oakley MG. Effect of peak flow monitoring on child asthma quality of life. J Pediatr Nurs. 2012 Feb;27(1):18 25. [15] Favreau H, Bacon SL, Labrecque M, Lavoie KL. Prospective impact of panic d isorder and panic anxiety on asthma control, health service use, and quality of life in adult patients with asthma over a 4 year follow up. Psychosom Med. 2014 Feb;76(2):147 55. [16] Eilayyan O, Gogovor A, Mayo N, Ernst P, Ahmed S. Predictors of perceived asthma control among patients managed in primary care clinics. Qual Life Res. 2014 May 8. [17] Apter AJ, Wan F, Reisine S, Bender B, Rand C, Bogen DK, et al. The association of health literacy with adherence and outcomes in moderate severe asthma. J Allerg y Clin Immunol. 2013 Aug;132(2):321 7. [18] Terzano C, Cremonesi G, Girbino G, Ingrassia E, Marsico S, Nicolini G, et al. 1 year prospective real life monitoring of asthma control and quality of life in Italy Respir Res. 2012 Dec 6;13:112,9921 13 112. [19 ] Wood PR, Smith B, O'Donnell L, Galbreath AD, Lara M, Forkner E, et al. Quantifying asthma symptoms in adults: The Lara asthma symptom scale. J Allergy Clin Immunol. 2007 Dec;120(6):1368 72. [20] Chen H, Gould MK, Blanc PD, Miller DP, Kamath TV, Lee JH, e t al. Asthma control, severity, and quality of life: Quantifying the effect of uncontrolled disease. J Allergy Clin Immunol. 2007 Aug;120(2):396 402. [21] Hesselink AE, van der Windt DA, Penninx BW, Wijnhoven HA, Twisk JW, Bouter LM, et al. What predicts c hange in pulmonary function and quality of life in asthma or COPD? J Asthma. 2006 Sep;43(7):513 9. [22] King MT, Kenny PM, Marks GB. Measures of asthma control and quality of life: Longitudinal data provide practical insights into their relative usefulness in different research contexts. Qual Life Res. 2009 Apr;18(3):301 12. [23] Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T, Mishima M. Analysis of longitudinal changes in the psychological status of patients with asthma. Respir Med. 2007 Oct;101(10):2133 8. [24] Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T, Koyama H, et al. Longitudinal changes in patient vs. physician based outcome measures did not significantly correlate in asthma. J Clin Epidemiol. 2005 May;58(5):532 9.

PAGE 125

125 [25] Oga T, Tsukino M, Hajiro T, Ikeda A, Koyama H, Mishima M, et al. Multidimensional analyses of long term clinical courses of asthma and chronic obstructive pulmonary disease. Allergol Int. 2010 Sep;59(3):257 65. [26] Rohan J, Drotar D, McNally K, Schluchter M, Riekert K, Vavrek P, et al. Adherence to pediatric asthma treatment in economically disadvantaged A frican A merican children and adolescents: An application of growth curve analysis. J Pediatr Psychol. 2010 May;35(4):394 404. [27] Mansour ME, Kotagal U, Rose B, Ho M, Brewer D, Roy Chaudhury A, et al. Health related quality of life in urban elementary schoolchildren. Pediatrics. 2003 Jun;111(6 Pt 1):1372 81. [28] Petsios KT, Priftis KN, Tsoumakas C, Perperoglou A, Hatziagorou E, Tsanakas JN, et al. Cough affects quality of life i n asthmatic children aged 8 14 more than other asthma symptoms. Allergol Immunopathol (Madr). 2009 Mar Apr;37(2):80 8. [29] Escobar R, Soutullo CA, Hervas A, Gastaminza X, Polavieja P, Gilaberte I. Worse quality of life for children with newly diagnosed at tention deficit/hyperactivity disorder, compared with asthmatic and healthy children. Pediatrics. 2005 Sep;116(3):e364 9. [30] Jensen ME, Gibson PG, Collins CE, Hilton JM, Latham Smith F, Wood LG. Increased sleep latency and reduced sleep duration in children with asthma. Sleep Breath. 2013 Mar;17(1):281 7. [31] Meltzer LJ, Ullrich M, Szefler SJ. Sleep duration, sleep hygiene, and insomnia in adolescents with asthma. The Journal of allergy and clinical immunology. 2014;In Practice. [32] Chugh IM, Khann a P, Shah A. Nocturnal symptoms and sleep disturbances in clinically stable asthmatic children. Asian Pac J Allergy Immunol. 2006 Jun Sep;24(2 3):135 42. [33] Krouse HJ, Yarandi H, McIntosh J, Cowen C, Selim V. Assessing sleep quality and daytime wakefulne ss in asthma using wrist actigraphy. J Asthma. 2008 Jun;45(5):389 95. [34] Stores G, Ellis AJ, Wiggs L, Crawford C, Thomson A. Sleep and psychological disturbance in nocturnal asthma. Arch Dis Child. 1998 May;78(5):413 9. [35] Diette GB, Markson L, Skinner EA, Nguyen TT, Algatt Bergstrom P, Wu AW. Nocturnal asthma in children affects school attendance, school performance, and parents' work attendance. Arch Pediatr Adolesc Med. 2000 Sep;154(9):923 8.

PAGE 126

126 [36] Strunk RC, Sternberg AL, Bacharier LB, Szefler SJ. No cturnal awakening caused by asthma in children with mild to moderate asthma in the childhood asthma management program. J Allergy Clin Immunol. 2002 Sep;110(3):395 403. [37] Dean BB, Calimlim BC, Sacco P, Aguilar D, Maykut R, Tinkelman D. Uncontrolled asth ma among children: Impairment in social functioning and sleep. J Asthma. 2010 Jun;47(5):539 44. [38] van Maanen A, Wijga AH, Gehring U, Postma DS, Smit HA, Oort FJ, et al. Sleep in children with asthma: Results of the PIAMA study. Eur Respir J. 2013 Apr;41 (4):832 7. [39] Daniel LC, Boergers J, Kopel SJ, Koinis Mitchell D. Missed sleep and asthma morbidity in urban children. Ann Allergy Asthma Immunol. 2012 Jul;109(1):41 6. [40] Taras H, Potts Datema W. Childhood asthma and student performance at school. J S ch Health. 2005 Oct;75(8):296 312. [41] Turner S, Eastwood P, Cook A, Jenkins S. Improvements in symptoms and quality of life following exercise training in older adults with moderate/severe persistent asthma. Respiration. 2011;81(4):302 10. [42] Jankowska EA, Wegrzynowska K, Superlak M, Nowakowska K, Lazorczyk M, Biel B, et al. The 12 week progressive quadriceps resistance training improves muscle strength, exercise capacity and quality of life in patients with stable chronic heart failure. Int J Cardiol. 2008 Oct 30;130(1):36 43. [43] Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, Bogels SM. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta analytic review. Sleep Med Rev. 2010 Jun;14(3):179 89. [44] Li Z, Huang IC, Thompson L, Tuli S, Huang SW, Dewalt D, et al. The relationships between asthma control, daytime sleepiness, and quality of life among children with asthma: A path analysis. Sleep Med. 2013 Jul;14(7):641 7. [45] Lemola S, Ledermann T, Friedman EM. Variability of sleep duration is related to subjective sleep quality and subjective well being: An actigraphy study. PLoS One. 2013;8(8):e71292. [46] Lara M, Rosenbaum S, Rachelefsky G, Nicholas W, Morton SC, Emont S, et al. Improving chil dhood asthma outcomes in the united states: A blueprint for policy action. Pediatrics. 2002 May;109(5):919 30. [47] Mosnaim G, Li H, Martin M, Richardson D, Belice PJ, Avery E, et al. Factors associated with levels of adherence to inhaled corticosteroids i n minority adolescents with asthma. Ann Allergy Asthma Immunol. 2014 Feb;112(2):116 20.

PAGE 127

127 [48] Mahut B, Trinquart L, Delclaux C. Influence of age on the risk of severe exacerbation and asthma control in childhood. J Asthma. 2011 Feb;48(1):65 8. [49] Koster E S, Raaijmakers JA, Vijverberg SJ, Koenderman L, Postma DS, Koppelman GH, et al. Limited agreement between current and long term asthma control in children: The PACMAN cohort study. Pediatr Allergy Immunol. 2011 Dec;22(8):776 83. [50] Patel MR, Brown RW, Cl ark NM. Perceived parent financial burden and asthma outcomes in low income, urban children. J Urban Health. 2013 Apr;90(2):329 42. [51] Vasbinder E, Dahhan N, Wolf B, Zoer J, Blankman E, Bosman D, et al. The association of ethnicity with electronically me asured adherence to inhaled corticosteroids in children. Eur J Clin Pharmacol. 2013 Mar;69(3):683 90. [52] Ortega H, Miller DP, Li H. Characterization of asthma exacerbations in primary care using cluster analysis. J Asthma. 2012 Mar;49(2):158 69. [53] Cro cker D, Brown C, Moolenaar R, Moorman J, Bailey C, Mannino D, et al. Racial and ethnic disparities in asthma medication usage and health care utilization: Data from the national asthma survey. Chest. 2009 Oct;136(4):1063 71. [54] Deger L, Plante C, Goudrea u S, Smargiassi A, Perron S, Thivierge RL, et al. Home environmental factors associated with poor asthma control in Montreal children: A population based study. J Asthma. 2010 Jun;47(5):513 20. [55] Indinnimeo L, Chiarotti F, De Vittori V, Baldini L, De Ca stro G, Zicari AM, et al. Risk factors affecting quality of life in a group of Italian children with asthma. Int J Immunopathol Pharmacol. 2014 Apr Jun;27(2):235 44. [56] Sousa SI, Alvim Ferraz MC, Martins FG. Health effects of ozone focusing on childhood asthma: What is now known -a review from an epidemiological point of view. Chemosphere. 2013 Feb;90(7):2051 8. [57] Braback L, Forsberg B. Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: Findings from re cent cohort studies. Environ Health. 2009 Apr 16;8:17,069X 8 17. [58] Trasande L, Thurston GD. The role of air pollution in asthma and other pediatric morbidities. J Allergy Clin Immunol. 2005 Apr;115(4):689 99. [59] Graham LM. All I need is the air that I breath: Outdoor air quality and asthma. Paediatr Respir Rev. 2004;5 Suppl A:S59 64. [60] Goldsmith CA, Kobzik L. Particulate air pollution and asthma: A review of epidemiological and biological studies. Rev Environ Health. 1999 Jul Sep;14(3):121 34.

PAGE 128

128 [61] Gilmour MI, Jaakkola MS, London SJ, Nel AE, Rogers CA. How exposure to environmental tobacco smoke, outdoor air pollutants, and increased pollen burdens influences the incidence of asthma. Environ Health Perspect. 2006 Apr;114(4):627 33. [62] Wang HC, Yous ef E. Air quality and pediatric asthma related emergencies. J Asthma. 2007 Dec;44(10):839 41. [63] Maestrelli P, Canova C, Scapellato ML, Visentin A, Tessari R, Bartolucci GB, et al. Personal exposure to particulate matter is associated with worse health p erception in adult asthma. J Investig Allergol Clin Immunol. 2011;21(2):120 8. [64] Zora JE, Sarnat SE, Raysoni AU, Johnson BA, Li WW, Greenwald R, et al. Associations between urban air pollution and pediatric asthma control in E l P aso, Texas Sci Total En viron. 2013 Mar 15;448:56 65. [65] Phillips JF, Jelks ML, Lockey RF. Important Florida botanical aeroallergens. Allergy Asthma Proc. 2010 Jul Aug;31(4):337 40. [66] Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, et al. The patient reported outcom es measurement information system (PROMIS) developed and tested its first wave of adult self reported health outcome item banks: 2005 2008. J Clin Epidemiol. 2010 Nov;63(11):1179 94. [67] Drake C, Nickel C, Burduvali E, Roth T, Jefferson C, Pietro B. The p ediatric daytime sleepiness scale (PDSS): Sleep habits and school outcomes in middle school children. Sleep. 2003 Jun 15;26(4):455 8. [68] Carroll WD, Wildhaber J, Brand PL. Parent misperception of control in childhood/adolescent asthma: The room to breath e survey. Eur Respir J. 2012 Jan;39(1):90 6. [69] Liu AH, Zeiger RS, Sorkness CA, Ostrom NK, Chipps BE, Rosa K, et al. The childhood asthma control test: Retrospective determination and clinical validation of a cut point to identify children with very poor ly controlled asthma. J Allergy Clin Immunol. 2010 Aug;126(2):267,73, 273.e1. [70] Smith LA, Bokhour B, Hohman KH, Miroshnik I, Kleinman KP, Cohn E, et al. Modifiable risk factors for suboptimal control and controller medication underuse among children wit h asthma. Pediatrics. 2008 Oct;122(4):760 9. [71] Halterman JS, Auinger P, Conn KM, Lynch K, Yoos HL, Szilagyi PG. Inadequate therapy and poor symptom control among children with asthma: Findings from a multistate sample. Ambul Pediatr. 2007 Mar Apr;7(2):1 53 9.

PAGE 129

129 [72] Lozano P, Finkelstein JA, Hecht J, Shulruff R, Weiss KB. Asthma medication use and disease burden in children in a primary care population. Arch Pediatr Adolesc Med. 2003 Jan;157(1):81 8. [73] Everhart RS, Fiese BH. Asthma severity and child qua lity of life in pediatric asthma: A systematic review. Patient Educ Couns. 2009 May;75(2):162 8. [74] Seid M, Limbers CA, Driscoll KA, Opipari Arrigan LA, Gelhard LR, Varni JW. Reliability, validity, and responsiveness of the pediatric quality of life inve ntory (PedsQL) generic core scales and asthma symptoms scale in vulnerable children with asthma. J Asthma. 2010 Mar;47(2):170 7. [75] Juniper EF, Guyatt GH, Feeny DH, Ferrie PJ, Griffith LE, Townsend M. Measuring quality of life in children with asthma. Qu al Life Res. 1996 Feb;5(1):35 46. [76] Gandhi PK, Kenzik KM, Thompson LA, DeWalt DA, Revicki DA, Shenkman EA, et al. Exploring factors influencing asthma control and asthma specific health related quality of life among children. Respir Res. 2013 Feb 23;14: 26,9921 14 26. [77] Thissen D, Varni JW, Stucky BD, Liu Y, Irwin DE, Dewalt DA. Using the PedsQL 3.0 asthma module to obtain scores comparable with those of the PROMIS pediatric asthma impact scale (PAIS). Qual Life Res. 2011 Nov;20(9):1497 505. [78] Curra n PJ, Obeidat K, Losardo D. Twelve frequently asked questions about growth curve modeling. J Cogn Dev. 2010;11(2):121 36. [79] Berlin KS, Parra GR, Williams NA. An introduction to latent variable mixture modeling (part 2): Longitudinal latent class growth analysis and growth mixture models. J Pediatr Psychol. 2013 Dec 4. [80] Yu C. Evaluating cutoff criteria of model fit indices for latent variable models with binary and continuous outcomes. Diss. University of California Los Angeles. 2002. [81] Wu W, West SG, Taylor AB. Evaluating model fit for growth curve models: Integration of fit indices from SEM and MLM frameworks. Psychol Methods. 2009 Sep;14(3):183 201. [82] Clarke SA, Calam R. The effectiveness of psychosocial interventions designed to improve healt h related quality of life (HRQOL) amongst asthmatic children and their families: A systematic review. Qual Life Res. 2012 Jun;21(5):747 64. [83] McGhan SL, Wong E, Sharpe HM, Hessel PA, Mandhane P, Boechler VL, et al. A children's asthma education program: Roaring adventures of puff (RAP), improves quality of life. Can Respir J. 2010 Mar Apr;17(2):67 73.

PAGE 130

130 [84] Cicutto L, Murphy S, Coutts D, O'Rourke J, Lang G, Chapman C, et al. Breaking the access barrier: Evaluating an asthma center's efforts to provide edu cation to children with asthma in schools. Chest. 2005 Oct;128(4):1928 35. [85] Clark NM, Shah S, Dodge JA, Thomas LJ, Andridge RR, Little RJ. An evaluation of asthma interventions for preteen students. J Sch Health. 2010 Feb;80(2):80 7. [86] Seid M, Varni JW, Gidwani P, Gelhard LR, Slymen DJ. Problem solving skills training for vulnerable families of children with persistent asthma: Report of a randomized trial on health related quality of life outcomes. J Pediatr Psychol. 2010 Nov;35(10):1133 43. [87] Bru zzese JM, Evans D, Kattan M. School based asthma programs. J Allergy Clin Immunol. 2009 Aug;124(2):195 200. [88] Ahmed S, Berzon RA, Revicki DA, Lenderking WR, Moinpour CM, Basch E, et al. The use of patient reported outcomes (PRO) within comparative effectiveness research: Implications for clinical practice and health care policy. Med Care. 2012 Dec;50(12):1060 70. [89] Qamar N, Pappalardo AA, Arora VM, Press VG. Patient centered care and its effect on outcomes in the treatment of asthma. Patient Rela t Outcome Meas. 2011 Jul;2:81 109. [90] Kamal AH, Bull J, Stinson C, Blue D, Smith R, Hooper R, et al. Collecting data on quality is feasible in community based palliative care. J Pain Symptom Manage. 2011 Nov;42(5):663 7. [91] Abernethy AP, Herndon JE,2nd Coan A, Staley T, Wheeler JL, Rowe K, et al. Phase 2 pilot study of pathfinders: A psychosocial intervention for cancer patients. Support Care Cancer. 2010 Jul;18(7):893 8. [92] Kamal AH, Miriovsky BJ, Currow DC, Abernethy AP. Improving the management of dyspnea in the community using rapid learning approaches. Chron Respir Dis. 2012 Feb;9(1):51 61. [93] Abernethy AP, Kamal AH, Wheeler JL, Cox C. Management of dyspnea within a rapid learning healthcare model. Curr Opin Support Palliat Care. 2011 Jun;5(2): 101 10. [94] Nickels A, Dimov V. Innovations in technology: Social media and mobile technology in the care of adolescents with asthma. Curr Allergy Asthma Rep. 2012 Dec;12(6):607 12. [95] de Jongh T, Gurol Urganci I, Vodopivec Jamsek V, Car J, Atun R. Mobi le phone messaging for facilitating self management of long term illnesses. Cochrane Database Syst Rev. 2012 Dec 12;12:CD007459.

PAGE 131

131 [96] Li P, Guttmann A. Recent innovations to improve asthma outcomes in vulnerable children. Curr Opin Pediatr. 2009 Dec;21(6): 783 8. [97] Elias P, Rajan NO, McArthur K, Dacso CC. InSpire to promote lung assessment in youth: Evolving the self management paradigms of young people with asthma. Med 2 0. 2013 May 21;2(1):e1 Jun. [98] Wisnivesky JP, Kattan M, Evans D, Leventhal H, Musu meci Szabo TJ, McGinn T, et al. Assessing the relationship between language proficiency and asthma morbidity among inner city asthmatics. Med Care. 2009 Feb;47(2):243 9. [99] Poureslami IM, Rootman I, Balka E, Devarakonda R, Hatch J, Fitzgerald JM. A syste matic review of asthma and health literacy: A cultural ethnic perspective in Canada MedGenMed. 2007 Aug 21;9(3):40. [100] Bearison DJ, Minian N, Granowetter L. Medical management of asthma and folk medicine in a Hispanic community. J Pediatr Psychol. 2002 Jun;27(4):385 92. [101] Koinis Mitchell D, McQuaid EL, Friedman D, Colon A, Soto J, Rivera DV, et al. Latino caregivers' beliefs about asthma: Causes, symptoms, and practices. J Asthma. 2008 Apr;45(3):205 10. [102] Wilson SR, Rand CS, Cabana MD, Foggs MB, Halterman JS, Olson L, et al. Asthma outcomes: Quality of life. J Allergy Clin Immunol. 2012 Mar;129(3 Suppl):S88 123. [103] Bender B, Wamboldt FS, O'Connor SL, Rand C, Szefler S, Milgrom H, et al. Measurement of children's asthma medication adherence by self report mother report, canister weight, and doser CT. Ann Allergy Asthma Immunol. 2000 Nov;85(5):416 21. [104] Bender BG, Bartlett SJ, Rand CS, Turner C, Wamboldt FS, Zhang L. Impact of interview mode on accuracy of child and parent report of adherenc e with asthma controller medication. Pediatrics. 2007 Sep;120(3):e471 7. [105] Preacher KJ, Zhang Z, Zyphur MJ. Alternative methods for assessing mediation in multilevel data: The advantages of multilevel SEM. Structural Equation Modeling. 2011;18(2):161 8 2. [106] Preacher KJ, Zyphur MJ, Zhang Z. A general multilevel SEM framework for assessing multilevel mediation. Psychol Methods. 2010 Sep;15(3):209 33. [107] Cano MA, Lam CY, Chen M, Adams CE, Correa Fernandez V, Stewart DW, et al. Positive smoking outcom e expectancies mediate the association between negative affect and smoking urge among women during a quit attempt. Exp Clin Psychopharmacol. 2014 Aug;22(4):332 40.

PAGE 132

132 [108] Sturgeon JA, Zautra AJ, Arewasikporn A. A multilevel structural equation modeling anal ysis of vulnerabilities and resilience resources influencing affective adaptation to chronic pain. Pain. 2014 Feb;155(2):292 8. [109] Dunkley DM, Ma D, Lee IA, Preacher KJ, Zuroff DC. Advancing complex explanatory conceptualizations of daily negative and p ositive affect: Trigger and maintenance coping action patterns. J Couns Psychol. 2014 Jan;61(1):93 109. [110] Bauer DJ, Preacher KJ, Gil KM. Conceptualizing and testing random indirect effects and moderated mediation in multilevel models: New procedures an d recommendations. Psychol Methods. 2006 Jun;11(2):142 63. [111] Preacher KJ, James SP. Advantages of Monte Carlo confidence intervals for indirect effects. Communication Methods and Measures. 2012;6(2):77 98. [112] Caicedo B, Jones K. Investigating neighb orhood effects on health: Using community survey data for developing neighborhood related constructs. Rev Salud Publica (Bogota). 2014 Feb;16(1):77 89. [113] Chen Q. The impact of ignoring a level of nesting structure in multilevel mixture model A monte ca rlo study. SAGE Open. 2012;2158244012442518. [114] Fenn B, Morris SS, Frost C. Do childhood growth indicators in developing countries cluster? Implications for intervention strategies. Public Health Nutr. 2004 Oct;7(7):829 34. [115] Mastronarde JG, Wise RA Shade DM, Olopade CO, Scharf SM, American Lung Association Asthma Clinical Research Centers. Sleep quality in asthma: Results of a large prospective clinical trial. J Asthma. 2008 Apr;45(3):183 9. [116] Meltzer LJ, Avis KT, Biggs S, Reynolds AC, Crabtree VM, Bevans KB. The children's report of sleep patterns (CRSP): A self report measure of sleep for school aged children. J Clin Sleep Med. 2013 Mar 15;9(3):235 45. [117] Van de Water AT, Holmes A, Hurley DA. Objective measurements of sleep for non laboratory settings as alternatives to polysomnography -a systematic review. J Sleep Res. 2011 Mar;20(1 Pt 2):183 200. [118] Krystal AD, Edinger JD. Measuring sleep quality. Sleep Med. 2008 Sep;9 Suppl 1:S10 7. [119] Campos FL, da Silva Junior FP, de B ruin VM, de Bruin PF. Melatonin improves sleep in asthma: A randomized, double blind, placebo controlled study. Am J Respir Crit Care Med. 2004 Nov 1;170(9):947 51.

PAGE 133

133 [120] Ivanenko A, Crabtree VM, Tauman R, Gozal D. Melatonin in children and adolescents wit h insomnia: A retrospective study. Clin Pediatr (Phila). 2003 Jan Feb;42(1):51 8. [121] Rimmer J, Greenwood A, Bartlett D, Hellgren J. Nasal steroids improve regulation of nasal patency in asthma and mild rhinitis: A randomized cross over trial. Eur Arch Otorhinolaryngol. 2012 Apr;269(4):1133 8. [122] Profita M, Riccobono L, Bonanno A, Chanez P, Gagliardo R, Montalbano AM, et al. Effect of nebulized beclomethasone on airway inflammation and clinical status of children with allergic asthma and rhinitis: A r andomized, double blind, placebo controlled study. Int Arch Allergy Immunol. 2013;161(1):53 64. [123] Boulet LP, Boulay ME. Asthma related comorbidities. Expert Rev Respir Med. 2011 Jun;5(3):377 93. [124] Cazzola M, Segreti A, Calzetta L, Rogliani P. Comor bidities of asthma: Current knowledge and future research needs. Curr Opin Pulm Med. 2013 Jan;19(1):36 41. [125] de Groot EP, Duiverman EJ, Brand PL. Comorbidities of asthma during childhood: Possibly important, yet poorly studied. Eur Respir J. 2010 Sep;3 6(3):671 8. [126] Slater G, Steier J. Excessive daytime sleepiness in sleep disorders. J Thorac Dis. 2012 Dec;4(6):608 16. [127] Berger AM, Parker KP, Young McCaughan S, Mallory GA, Barsevick AM, Beck SL, et al. Sleep wake disturbances in people with cance r and their caregivers: State of the science. Oncol Nurs Forum. 2005 Nov 3;32(6):E98 126. [128] Carlson LE, Speca M, Patel KD, Goodey E. Mindfulness based stress reduction in relation to quality of life, mood, symptoms of stress and levels of cortisol, deh ydroepiandrosterone sulfate (DHEAS) and melatonin in breast and prostate cancer outpatients. Psychoneuroendocrinology. 2004 May;29(4):448 74. [129] Berger AM, VonEssen S, Kuhn BR, Piper BF, Agrawal S, Lynch JC, et al. Adherence, sleep, and fatigue outcomes after adjuvant breast cancer chemotherapy: Results of a feasibility intervention study. Oncol Nurs Forum. 2003 May Jun;30(3):513 22. [130] Babin SM, Burkom HS, Holtry RS, Tabernero NR, Stokes LD, Davies Cole JO, et al. Pediatric patient asthma related eme rgency department visits and admissions in Washington DC, from 2001 2004, and associations with air quality, socio economic status and age group. Environ Health. 2007 Mar 21;6:9. [131] Tzivian L. Outdoor air pollution and asthma in children. J Asthma. 201 1 Jun;48(5):470 81.

PAGE 134

134 [132] Jenerowicz D, Silny W, Danczak Pazdrowska A, Polanska A, Osmola Mankowska A, Olek Hrab K. Environmental factors and allergic diseases. Ann Agric Environ Med. 2012;19(3):475 81. [133] Gilliland FD. Outdoor air pollution, genetic su sceptibility, and asthma management: Opportunities for intervention to reduce the burden of asthma. Pediatrics. 2009 Mar;123 Suppl 3:S168 73. [134] Gielen MH, van der Zee SC, van Wijnen JH, van Steen CJ, Brunekreef B. Acute effects of summer air pollution on respiratory health of asthmatic children. Am J Respir Crit Care Med. 1997 Jun;155(6):2105 8. [135] Bateson TF, Schwartz J. Children's response to air pollutants. J Toxicol Environ Health A. 2008;71(3):238 43. [136] McConnell R, Islam T, Shankardass K, J errett M, Lurmann F, Gilliland F, et al. Childhood incident asthma and traffic related air pollution at home and school. Environ Health Perspect. 2010 Jul;118(7):1021 6. [137] Jerrett M, Shankardass K, Berhane K, Gauderman WJ, Kunzli N, Avol E, et al. Traf fic related air pollution and asthma onset in children: A prospective cohort study with individual exposure measurement. Environ Health Perspect. 2008 Oct;116(10):1433 8. [138] Gehring U, Wijga AH, Brauer M, Fischer P, de Jongste JC, Kerkhof M, et al. Traf fic related air pollution and the development of asthma and allergies during the first 8 years of life. Am J Respir Crit Care Med. 2010 Mar 15;181(6):596 603. [139] Carlsten C, Dybuncio A, Becker A, Chan Yeung M, Brauer M. Traffic related air pollution and incident asthma in a high risk birth cohort. Occup Environ Med. 2011 Apr;68(4):291 5. [140] Penard Morand C, Raherison C, Charpin D, Kopferschmitt C, Lavaud F, Caillaud D, et al. Long term exposure to close proximity air pollution and asthma and allergies in urban children. Eur Respir J. 2010 Jul;36(1):33 40. [141] Arnedo Pena A, Garcia Marcos L, Carvajal Uruena I, Busquets Monge R, Morales Suarez Varela M, Miner Canflanca I, et al. Air pollution and recent symptoms of asthma, allergic rhinitis, and atopic eczema in schoolchildren aged between 6 and 7 years. Arch Bronconeumol. 2009 May;45(5):224 9. [142] DellaValle CT, Triche EW, Leaderer BP, Bell ML. Effects of ambient pollen concentrations on frequency and severity of asthma symptoms among asthmatic child ren. Epidemiology. 2012 Jan;23(1):55 63.

PAGE 135

135 [143] Gleason JA, Bielory L, Fagliano JA. Associations between ozone, PM2.5, and four pollen types on emergency department pediatric asthma events during the warm season in New Jersey : A case crossover study. Enviro n Res. 2014 Jul;132:421 9. [144] Jariwala SP, Kurada S, Moday H, Thanjan A, Bastone L, Khananashvili M, et al. Association between tree pollen counts and asthma ED visits in a high density urban center. J Asthma. 2011 Jun;48(5):442 8. [145] Deger L, Plante C, Jacques L, Goudreau S, Perron S, Hicks J, et al. Active and uncontrolled asthma among children exposed to air stack emissions of sulphur dioxide from petroleum refineries in Montreal Quebec : A cross sectional study. Can Respir J. 2012 Mar Apr;19(2):97 102. [146] Escamilla Nunez MC, Barraza Villarreal A, Hernandez Cadena L, Moreno Macias H, Ramirez Aguilar M, Sienra Monge JJ, et al. Traffic related air pollution and respiratory symptoms among asthmatic children, resident in Mexico C ity: The EVA cohort study. Respir Res. 2008 Nov 16;9:74,9921 9 74. [147] Liu L, Poon R, Chen L, Frescura AM, Montuschi P, Ciabattoni G, et al. Acute effects of air pollution on pulmonary function, airway inflammation, and oxidative stress in asthmatic children. Environ Health Perspect. 2009 Apr;117(4):668 74. [148] Meng YY, Rull RP, Wilhelm M, Lombardi C, Balmes J, Ritz B. Outdoor air pollution and uncontrolled asthma in the S an Joaquin valley, California J Epidemiol Community Health. 2010 Feb;64(2):142 7. [149] Patel MM, Chi llrud SN, Deepti KC, Ross JM, Kinney PL. Traffic related air pollutants and exhaled markers of airway inflammation and oxidative stress in N ew York C ity adolescents. Environ Res. 2013 Feb;121:71 8. [150] Patel MM, Quinn JW, Jung KH, Hoepner L, Diaz D, Perz anowski M, et al. Traffic density and stationary sources of air pollution associated with wheeze, asthma, and immunoglobulin E from birth to age 5 years among N ew York C ity children. Environ Res. 2011 Nov;111(8):1222 9. [151] Magas OK, Gunter JT, Regens JL Ambient air pollution and daily pediatric hospitalizations for asthma. Environ Sci Pollut Res Int. 2007 Jan;14(1):19 23. [152] Ostro B, Lipsett M, Mann J, Braxton Owens H, White M. Air pollution and exacerbation of asthma in A frican A merican children in L os Angeles Epidemiology. 2001 Mar;12(2):200 8. [153] YoussefAgha AH, Jayawardene WP, Lohrmann DK, El Afandi GS. Air pollution indicators predict outbreaks of asthma exacerbations among elementary school children: Integration of daily environmental and sc hool health surveillance systems in Pennsylvania J Environ Monit. 2012 Dec;14(12):3202 10.

PAGE 136

136 [154] Cakmak S, Dales RE, Coates F. Does air pollution increase the effect of aeroallergens on hospitalization for asthma? J Allergy Clin Immunol. 2012 Jan;129(1):2 28 31. [155] Dales RE, Cakmak S, Judek S, Dann T, Coates F, Brook JR, et al. Influence of outdoor aeroallergens on hospitalization for asthma in Canada J Allergy Clin Immunol. 2004 Feb;113(2):303 6. [156] Tosca MA, Ruffoni S, Canonica GW, Ciprandi G. Asth ma exacerbation in children: Relationship among pollens, weather, and air pollution. Allergol Immunopathol (Madr). 2014 Jul Aug;42(4):362 8. [157] United State Environmental Protecti on Agency. Download daily date. http:// w ww.epa.gov/airdata/ad_data_daily.h tml. Accessed October 2, 2014 [158] United States Census Bureau. 2010 TIGER/line shapefile. http:// w ww.census.gov/cgi bin/geo/shapefiles2010/main Accessed October 2, 2014. [159] IMS Health. h ttp:// w ww.imshealth.com/portal/site/imshealth Accessed October 2, 2014. [160] National Climatic Data Center. Climate data online. http:// w ww.ncdc.noaa.gov/cdo web Accessed October 2, 2014. [161] Studnicka M, Hackl E, Pischinger J, Fangmeyer C, Haschke N, Kuhr J, et al. Traffic related NO2 and the prevalence of asthma and respiratory symptoms in seven year olds. Eur Respir J. 1997 Oct;10(10):2275 8. [162] Evans KA, Halterman JS, Hopke PK, Fagnano M, Rich DQ. Increased ultrafine parti cles and carbon monoxide concentrations are associated with asthma exacerbation among urban children. Environ Res. 2014 Feb;129:11 9. [163] Hajat S, Haines A, Goubet SA, Atkinson RW, Anderson HR. Association of air pollution with daily GP consultations for asthma and other lower respiratory conditions in London Thorax. 1999 Jul;54(7):597 605. [164] Anderson HR, Ponce de Leon A, Bland JM, Bower JS, Emberlin J, Strachan DP. Air pollution, pollens, and daily admissions for asthma in London 1987 92. Thorax. 19 98 Oct;53(10):842 8. [165] Atkinson RW, Carey IM, Kent AJ, van Staa TP, Anderson HR, Cook DG. Long term exposure to outdoor air pollution and the incidence of chronic obstructive pulmonary disease in a national English cohort. Occup Environ Med. 2014 Aug 2 0.

PAGE 137

137 [166] Carey IM, Atkinson RW, Kent AJ, van Staa T, Cook DG, Anderson HR. Mortality associations with long term exposure to outdoor air pollution in a national English cohort. Am J Respir Crit Care Med. 2013 Jun 1;187(11):1226 33. [167] Atkinson RW, Carey IM, Kent AJ, van Staa TP, Anderson HR, Cook DG. Long term exposure to outdoor air pollution and incidence of cardiovascular diseases. Epidemiology. 2013 Jan;24(1):44 53. [168] Rich DQ, Kipen HM, Huang W, Wang G, Wang Y, Zhu P, et al. Association between c hanges in air pollution levels during the Beijing Olympics and biomarkers of inflammation and thrombosis in healthy young adults. JAMA. 2012 May 16;307(19):2068 78. [169] Kheirbek I, Wheeler K, Walters S, Kass D, Matte T. PM and ozone health impacts and di sparities in N ew York C ity: Sensitivity to spatial and temporal resolution. Air Qual Atmos Health. 2013 Jun;6(2):473 86. [170] Raun LH, Ensor KB, Persse D. Using community level strategies to reduce asthma attacks triggered by outdoor air pollution: A case crossover analysis. Environ Health. 2014 Jul 11;13:58,069X 13 58. [171] Hatch GE, Duncan KE, Diaz Sanchez D, Schmitt MT, Ghio AJ, Carraway MS, et al. Progress in assessing air pollutant risks from in vitro exposures: Matching ozone dose and effect in huma n airway cells. Toxicol Sci. 2014 Jun 13. [172] Fry RC, Rager JE, Bauer R, Sebastian E, Peden DB, Jaspers I, et al. Air toxics and epigenetic effects: Ozone altered microRNAs in the sputum of human subjects. Am J Physiol Lung Cell Mol Physiol. 2014 Jun 15; 306(12):L1129 37. [173] Guarnieri M, Balmes JR. Outdoor air pollution and asthma. Lancet. 2014 May 3;383(9928):1581 92. [174] Gorai AK, Tuluri F, Tchounwou PB. A GIS based approach for assessing the association between air pollution and asthma in N ew York S tate, USA. Int J Environ Res Public Health. 2014 May 6;11(5):4845 69. [175] Moore K, Neugebauer R, Lurmann F, Hall J, Brajer V, Alcorn S, et al. Ambient ozone concentrations cause increased hospitalizations for asthma in children: An 18 year study in southern California Environ Health Perspect. 2008 Aug;116(8):1063 70. [176 ] O'Connor GT, Neas L, Vaughn B, Kattan M, Mitchell H, Crain EF, et al. Acute respiratory health effects of air pollution on children with asthma in US inner cities. J Allergy Clin Immunol. 2008 May;121(5):1133,1139.e1. [177] Salam MT, Millstein J, Li YF, Lurmann FW, Margolis HG, Gilliland FD. Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter:

PAGE 138

138 Results from the children's health study. Environ Health Perspect. 2005 Nov;113(11):1638 44. [178] Horak F,Jr, Studnicka M, Gartn er C, Spengler JD, Tauber E, Urbanek R, et al. Particulate matter and lung function growth in children: A 3 yr follow up study in Austrian schoolchildren. Eur Respir J. 2002 May;19(5):838 45. [179] Chang HH, Reich BJ, Miranda ML. Time to event analysis of fine particle air pollution and preterm birth: Results from north Carolina 2001 2005. Am J Epidemiol. 2012 Jan 15;175(2):91 8. [180] Janssen NA, Brunekreef B, van Vliet P, Aarts F, Meliefste K, Harssema H, et al. The relationship between air pollution fro m heavy traffic and allergic sensitization, bronchial hyperresponsiveness, and respiratory symptoms in Dutch schoolchildren. Environ Health Perspect. 2003 Sep;111(12):1512 8. [181] Morgenstern V, Zutavern A, Cyrys J, Brockow I, Koletzko S, Kramer U, et al. Atopic diseases, allergic sensitization, and exposure to traffic related air pollution in children. Am J Respir Crit Care Med. 2008 Jun 15;177(12):1331 7. [182] Nadadur SS, Miller CA, Hopke PK, Gordon T, Vedal S, Vandenberg JJ, et al. The complexities of air pollution regulation: The need for an integrated research and regulatory perspective. Toxicol Sci. 2007 Dec;100(2):318 27. [183] Delucchi MA, Yang C, Burke AF, Ogden JM, Kurani K, Kessler J, et al. An assessment of electric vehicles: Technology, infras tructure requirements, greenhouse gas emissions, petroleum use, material use, lifetime cost, consumer acceptance and policy initiatives. Philos Trans A Math Phys Eng Sci. 2013 Dec 2;372(2006):20120325. [184] Quinto KB, Kit BK, Lukacs SL, Akinbami LJ. Envir onmental tobacco smoke exposure in children aged 3 19 years with and without asthma in the United States 1999 2010. NCHS Data Brief. 2013 Aug;(126)(126):1 8. [185] Akinbami LJ, Kit BK, Simon AE. Impact of environmental tobacco smoke on children with asthm a, United States 2003 2010. Acad Pediatr. 2013 Nov Dec;13(6):508 16. [186] Kit BK, Simon AE, Brody DJ, Akinbami LJ. US prevalence and trends in tobacco smoke exposure among children and adolescents with asthma. Pediatrics. 2013 Mar;131(3):407 14.

PAGE 139

139 B IOGRAPHICAL SKETCH Zheng Li received her Bachelor of Medicine from the Capital Medical University in China in 2007. After receiving her bachelor degree, she came to the United States to pursue her master degree. In August of 2008, she accepted a graduate a ssistant position in the School of Community Health Sciences at the University of Nevada, Reno. In May of 2010, she received her Master of Public Health from the University of Nevada, Reno In August of 2010, she was admitted to the Epidemiology PhD progra m in the College of Public Health and Health Professions at the University of Florida and received a Pre Doctoral Fellowship with the Institute for Child Health Policy in the Department of Health Outcomes and Policy at the College of Medicine she began the doctoral program in the Department of Epidemiology and became a research assistant in the Institute of Child Health Policy in the Department of Health Outcomes and Policy Zheng received her PhD in Epidemiology from the University of Florida in December of 20 14.



PAGE 1

OriginalArticleFactorsassociatedwithinjuryinREMsleepbehaviordisorder StuartJ.McCartera ErikK.St.Louisb 1 ChristopherL.Boswellc LucasG.Duefferta NancySlocumba BradleyF.Boeved MichaelH.Silberd EricJ.Olsone TimothyI.Morgenthalere MajaTippmann-Peikertb aMayoClinicCollegeofMedicine,200FirstStreetSouthwest,Rochester,MN55905,USAbMayoCenterforSleepMedicine,DepartmentsofMedicineandNeurology,MayoClinicCollegeofMedicine,200FirstStreetSouthwest,Rochester,MN 55905,USAcDepartmentofFamilyMedicine,MayoClinicCollegeofMedicine,200FirstStreetSouthwest,Rochester,MN55905,USAdMayoCenterforSleepMedicine,DepartmentofNeurology,MayoClinicCollegeofMedicine,20 0FirstStreetSouthwest,Rochester,MN55905,USAeMayoCenterforSleepMedicine,DepartmentsofMedicineandPulmonary&CriticalCareMedicine,MayoClinicCollegeofMedicine,200FirstStreet Southwest,Rochester,MN55905,USA ARTICLEINFO Articlehistory: Received21March2014 Receivedinrevisedform21May2014 Accepted5June2014 Availableonline Keywords: REMsleepbehaviordisorder Injury Falls Synucleinopathy Parkinsonsdisease SubduralhematomaABSTRACT Objective: Asfactorsassociatedwithinjuryinrapideyemovement(REM)sleepbehaviordisorder(RBD) remainlargelyunknown,weaimedtoidentifysuchfactors. Methods: Wesurveyedconsecutiveidiopathic(iRBD)orsymptomaticRBDpatientsseenbetween2008 and2010regardingRBD-relatedinjuries.Associationsbetweeninjuriesandclinicalvariablesweredeterminedwithoddsratios(OR)andmultiplelogisticregressionanalyses.Theprimaryoutcomevariableswereinjuryandinjuryseverity. Results: Fifty-threepatients(40%)responded.Medianagewas69years,and35(73.5%)weremen.Twentyeight(55%)hadiRBD.Twenty-nine(55%)reportedinjury,with37.8%toselfand16.7%tothebedpartner. 11.3%hadmarkedinjuriesrequiringmedicalinterventionorhospitalization,includingtwo(4%)subduralhematomas.iRBDdiagnosis(OR = 6.8, p = 0.016)anddreamrecall(OR = 7.5, p = 0.03)wereassociatedwithinjury;andiRBDdiagnosiswasindependentlyassociatedwithinjuryandinjuryseverityadjusting forage,gender,DEBfrequency,andduration.Falls( p = 0.03)werealsoassociatedwithinjuryseverity. DEBfrequencywasnotassociatedwithinjury,injuryseverity,orfalls. Conclusions: InjuriesappeartobeafrequentcomplicationofRBD,althoughtherelativelylowresponse rateinoursurveycouldhavebiasedresults.iRBDpatientsaremorelikelytosufferinjury…andmore severeinjuries…thansymptomaticRBDpatients.Inaddition,recallofdreamswasalsoassociatedwith injury,anddreamenactmentbehavior(DEB)-relatedfallswereassociatedwithmoresevereinjuries.One inninepatientssufferedinjuryrequiringmedicalintervention.ThefrequencyofDEBdidnotpredictRBDrelatedinjuries,highlightingtheimportanceoftimelyinitiationoftreatmentforRBDinpatientshaving evenrareDEBepisodes.Futureprospectivestudieswillbenecessarytode“nepredictorsofinjuryin RBD. 2014ElsevierB.V.Allrightsreserved. 1.Introduction Rapideyemovement(REM)sleepbehaviordisorder(RBD)isa parasomniacharacterizedbydreamenactmentbehavior(DEB)associatedwiththelossofnormalskeletalmuscleatoniaresulting inabnormal,excessivemotoractivityoftenmirroringdreamcontent duringREMsleep [1] .RBDresultsinmotoractivityrangingfrom simplelimbtwitchestomorecomplexandviolentmovementsthat mayresultininjurytothepatientand/ortheirbedpartner [2…11] Largepopulation-basedstudiesreporttheprevalenceofRBDtobe 0.38…0.5% [2,8,12] .However,probableRBDmayoccurinover6% ofcommunity-dwelling70…89-year-oldindividuals,suggestingthat RBDprevalence,andthereforepossibleresultantinjury,maybeconsiderablyhigherthanpreviouslybelieved,particularlyinvulnerableelderlypatients [3,5] .RBDispredominantlyseeninmenover age60;however,priortoage50,womenandmenareequallylikely todevelopRBD [8,13…16] .Therearetwodiagnosticcategoriesfor RBD,idiopathic(iRBD)orsymptomatic,whichwede“nedasthose patientshavingRBDsymptomsandacomorbidsynucleinopathy *Correspondingauthorat:MayoCenterforSleepMedicine,Departmentsof MedicineandNeurology,MayoClinicCollegeofMedicine,200FirstStreetSouthwest, Rochester,MN55905,USA.Tel.: + 15072667456;fax: + 15072667772. E-mailaddress: stlouis.erik@mayo.edu (E.K.St.Louis).1IndicatesPrincipalInvestigator. http://dx.doi.org/10.1016/j.sleep.2014.06.002 1389-9457/2014ElsevierB.V.Allrightsreserved. SleepMedicine (2014) … ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 Contentslistsavailableat ScienceDirect SleepMedicinejournalhomepage: www.elsevier.com/locate/sleep

PAGE 2

neurodegenerativediseaseincludingParkinsonsdisease(PD),dementiawithLewybodies(DLB),ormultiplesystematrophy(MSA) [2,7,8,10,17,18] .However,becauseapproximately80%ofiRBDpatientsdevelopparkinsonismorcognitivedeclineoverlongitudinalfollow-up,iRBDisconsideredtobeanearlyclinicalmanifestation ofsynucleinopathyneurodegenerativedisorderswithvaryingrates ofdiseaseprogressionpresentingupto50yearspriortothedevelopmentofovertcognitiveormotordecline [2,4,7,8,10,17…19] TheassociationofRBDandinjuryiswidelyacknowledged,occurringin33…96%ofpatientsand/ortheirbedpartners [2,10,20,21] However,verylittleisknownregardingwhatfactorspredisposea RBDpatientortheirbedpartnertoinjury.DEBcanvaryfromharmless(i.e.,singingorfeigningknitting)tomoredangerousbehaviors(i.e.,kicking,punching,or“ringunloadedweapons),whichmay resultinmildinjuriessuchasbruisestomoresevereinjuriessuch aslimbfracturesandsubduralhematomas [2,8,10,21…23] .Thegoals ofRBDtreatmentaretoreduceDEBfrequencyandpreventinjury, withthepresumptionthatminimizingDEBmayreduceinjuryoccurrence [2,10,21,24] .Clonazepamandmelatoninareusedtoreduce thefrequencyandseverityofDEBs [25] .Weaimedtoidentifywhich factorsareassociatedwithinjuryinRBDpatients. 2.Methods 2.1.Ethics TheMayoClinicInstitutionalReviewBoardprovidedhuman subjectsresearchapprovalforthisstudyandoversightofitsstudy activities. 2.2.Subjects Adiagnosisandtext-basedsearchidenti“ed641newlydiagnosedpatientswithRBDatourinstitutionbetween1/1/2000and 12/31/2009.GiventhedicultyindesigningsuitablesurveymeasuresforchildrenwhomaynothavewitnessedsleeptoaccuratelyreportonDEBs,weexcludedpatients < 18yearsofage,resulting in608eligiblesubjects.AllincludedpatientsmetthestandarddiagnosticcriteriaforRBD,includingthepresenceofREMsleep withoutatonia(RSWA)duringpolysomnography,ahistoryofsleeprelatedinjuriousorpotentiallyinjuriousdisruptivebehaviors,and/ orabnormalDEBsduringpolysomnography,absenceofREMrelatedepileptiformactivity,andabsenceofanothersleepdisorder thatbetterexplainedtheirsleepdisturbance [1] .Wemailedasurvey to133patientsseenbetween2008and2010,limitingthestudy sampletotheseyearstoreducerecallbias.Fifty-three(40%)surveys werereturnedandanalyzed.Seventy-eightpatientsdidnotrespond, andtwopatientsdiedbeforereturnofsurveys. 2.3.Surveyinstrument Wedesignedaquestionnaire,whichwascompletedbypatientstogetherwithabedpartnerwhohadobservedtheirRBD,that surveyeddemographics;baselineRBDbehavioralcharacteristicsprior totreatmentincludingfrequency,severity,andtypeofbehaviors, ratingsoflimbmovementintensity,andoccurrenceandqualityof vocalization;andoccurrenceandfrequencyoffallsfrombedand patientorbedpartnerinjuryspeci“callyrelatedtoDEBswithqualitativedescriptionsofanyinjuries.InjuriesunrelatedtoDEBwere notassessed.Injurywasconsideredabinaryvariableanddelineatedbypatientorbedpartnerinjury.Injuryseveritywasde“ned asacategoricalvariablerangingfrommild(nolastingsigns),moderate(bruises),ormarked(injuryrequiringmedicalattentionsuch asalacerationorfracture).Thissurveyinstrumentisshownin Supplemental/OnlineAppendixS1 2.4.Clinicaldata Wereviewedmedicalrecordstocon“rmRBDdiagnostictype aseitheridiopathicorsymptomaticatthetimeofinitialclinicvisit duringthe2008…2010timeperiod,aswellasfollow-upclinicaldata atthetimeofsurveycompletiontoverifypersistenceoftheoriginaldiagnosisatthetimeofsurveyresponses.Wealsoreviewed demographicinformationandmedicationdosagesandabstracted clinicallyimportantcovariatesincludingneurologicandpsychiatrichistory,antidepressantadministration,andpolysomnography data.Patientswithcomorbidneurologicalandsleepdisordersmet thecriteriaforclinicallyprobableDLB,mildcognitiveimpairment (MCI),PD,MSA,obstructivesleepapnea(OSA),andrestlesslegssyndrome(RLS) [1,25] 2.5.Dataanalysis StudydatawerecollectedandmanagedusingREDCapelectronicdatacapturetools [26] .Statisticalcalculationswerecarriedout usingJMPstatisticalsoftware(JMP,Version9.SASInstituteInc.,Cary, NC,USA).Qualitativeandordinaldatawerereportedasabsolute andrelativefrequencies,whilequantitativedatawerereportedas mediansandinterquartilerange(IQR).Wilcoxonrank-sumtestswere usedtocomparecontinuousoutcomes.Categoricalvariableswere comparedwithcontingencytables,chi-squaredtests,andoddsratios (ORs)withdeterminationof95%con“denceintervals(CIs)between injuryandnoninjurygroups.Univariatelogisticregressionanalyseswereperformedtodeterminepotentialassociationsbetween primarydependentvariablesincludinginjury(whethertoself,bed partner,orboth)andinjuryseverity.Variablesapproachingsignificanceusingunivariateanalysiswerethen“tintomultiplelogisticregressionmodels.UnitORswith95%CIswerethendetermined. Thesigni“cancelevelwassetatanalphaof p < 0.05. 3.Results 3.1.Demographicsandclinicaldata KeydemographicandclinicalRBDdataforallsubjectsaresummarizedin Table1 ,includingthegroupofpatientswhoreported injuryandthosewhodidnot Ofthe53respondents,39(73.5%) weremenwithamedianageof69years(IQR = 16.5).Themedian ageofRBDsymptomonsetwas57years(IQR = 16).ThemediandurationofRBDsymptomswas8years(IQR = 13.8).Twenty-eight(55%) ofpatientshadiRBDatthetimeofpolysomnographyandRBDdiagnosis;11ofthese28patientswereseeninfollow-upaftersurvey completionandremainedidiopathic,whiletheremaining17patientshadameandurationbetweenlastclinicalfollow-upandsurvey completionof22.3months.Twenty-“ve(45%)patientswhohad comorbidneurodegenerativediseasewereclassi“edashavingsymptomaticRBDatthetimeofpolysomnographyandRBDdiagnosis. Ten(40%)hadPD,nine(36%)hadDLB/MCI,andsix(24%)hadMSA. Twenty-three(43%)patientshadcomorbiddepressionand25(47%) wereonantidepressantmedications,includingbothselectiveserotoninandnorepinephrinereuptakeinhibitors.Thirty-“ve(66%) patientshadOSA,12(23%)hadRLS,33(61%)had 15periodicleg movementsofsleepperhour,and30(57%)reportedhypersomnolencewithamedianEpworthSleepinessScaleScoreof16 (IQR = 6).PatientsdiagnosedwithcomorbidOSA(apneahypopnea index(AHI) 5)hadamedianAHIof12perhour(IQR = 23.25),and themedianAHIfortheentirecohortwas3(IQR = 11);however,all RBDpatientswithOSAdiagnosesmetInternationalClassi“cation ofSleepDisorders…SecondEdition(ICSD-2)RBDdiagnosticcriteriawithevidenceforpolysomnographicRSWAandhadpersisting dreamenactmentepisodesfollowingsuccessfultreatmentofOSA (i.e.,noneofthesepatientsappearedtohavepseudo-RBDŽ) [27] ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 2 S.J.McCarteretal./SleepMedicine (2014) …

PAGE 3

Demographicdataweresimilarbetweengroupsinwhichinjuryoccurredversusnoinjury,withtheexceptionthatpatientswhoinjured themselvesortheirbedpartnersweremorelikelytohaveiRBD ( p = 0.01),moredreamrecall( p = 0.002),andmoresevereDEBrelatedlimbmovement( p = 0.025)comparedtothosewhosufferednoinjury( Table1 ).Therewerenodifferencesbetween idiopathic( n = 25,89%)andsymptomatic( n = 23,92%)RBDgroups forbedpartneravailabilityforsurveycompletion. 3.2.Injurydata Twenty-nine(55%)patientsreportedinjurypriortotreatment. Self-injuryoccurredmostfrequentlyin20patients(37.8%),while injuriestothebedpartneroccurredinninerespondents(16.7%). Withinthegroupofpatientsreportinginjury,eight(27%)injured boththemselvesandtheirbedpartner.Onlysix(21%)patients(four iRBD,twosymptomaticRBD)whoreportedinjurypresentedtoour sleepcenterwithoccurrenceofDEB-relatedinjurywithin6months priortopresentation,andnoneofthesepatientshadsleeprelatedinjuryasaprimarycomplaintorreasonforreferraltothe sleepcenter.Markedinjuriesrequiringeitheroutpatientmedical interventionorhospitalizationoccurredin11.3%ofallpatients.Of thosesufferinginjury,reportedinjuryseveritywasmildin20.7%, moderate(bruisesorvisibleinjury)in58.6%,andmarkedin20.7%. Thirty-onepatients(58%)reportedfallingoutofbedduringdream enactment.Injuriessuffered,dreamenactment,anddreamcontent variedwidelyamongpatients.Twopatients(4%)sufferedsubduralhematomasafterfallingoutofbed( Fig.1 ).Therangeofqualitativedescriptionsofreportedinjuriesandfrequencyofinjurytypes aresummarizedin Table2 The28patientswithiRBDweresigni“cantlymorelikelythan patientswithsymptomaticRBDtohaveinjuredthemselvesduring DEB(OR = 7.7,CI = 1.9…31.4, p = 0.002),andwhenonlythe11patientswithiRBDdiagnosisseenaftersurveycompletionweresimilarlyanalyzed,thisassociationbecamestronger,althoughCIs widened(OR = 19.0,3.6…138.0, p = 0.003).iRBDdiagnosiswasan independentpredictorofinjurytothepatient( p = 0.008)afteradjustingforage,gender,andfrequencyofDEB.PDwastheonly synucleinopathysubtypeindependentlyassociatedwithalowerfrequencyofreportedRBD-relatedinjuryafteradjustingforageand gender( p = 0.02).iRBDdiagnosis(OR = 6.8,CI = 1.4…40.7, p = 0.016) anddreamrecall(OR = 7.5,CI = 1.2…68.7, p = 0.03)werebothindependentlyassociatedwithinjurytopatientand/orbedpartneradjustingforage,gender,frequencyofDEB,andRBDduration.When 11iRBDpatientswhosediagnosiswasagainveri“edaftersurvey completionweresimilarlyanalyzed,thisassociationagainappearedstronger,albeitwithwiderCIs(OR = 10.7,CI = 2.1…83.2, p = 0.003). Agreaterseverityofpatient/bedpartner-perceivedlimbmovementwasassociatedwithbedpartnerinjury( p = 0.046)afteradjustingforage,sex,andRBDduration.iRBDdiagnosisalsoincreased thelikelihoodofmoresevereinjurytopatientorbedpartnerTable1 DemographicandclinicalvariablesofallRBDpatients,andthegroupsreportinginjuryornoinjurypriorto treatment. All( n = 53)Injury ( n = 29) NoInjury ( n = 24) Demographics,median(IQR) Age69(16.5)67(18.3)71(11) AgeofRBDonset57(16)56(11)57(17.3) RBDduration8(13.8)8.6(10.5)7.7(18.8) Sex, n (%) Male39(73%)21(72%)18(75%) Female14(27%)8(28%)6(25%) IdiopathicRBDPatients n (%)28(53%)22(76%)6(25%) NeurodegenerativeDx, n (%) SynucleinopathyatRBDDx 25(47%)7(24%)18(75%) PD ** 10(19%)2(7%)8(33%) DLB2(4%)1(3%)1(3%) MCI7(13%)2(7%)5(21%) MSA6(11%)2(7%)4(17%) Depression23(43%)14(48%)9(38%) SleepDisorder, n (%) OSA35(66%)18(62%)17(71%) RLS12(23%)7(24%)5(17%) PLMD33(61%)16(55%)17(71%) Medication, n (%) Antidepressant25(47%)13(48%)12(50%) SSRI14(26%)7(24%)7(29%) SNRI/Bupropion11(21%)6(21%)5(21%) Dopaminergic12(23%)5(17%)7(29%) Anticholinergic6(11%)2(7%)4(17%) RBDCharacteristicsSeverityoflimbmovement # 2.52.72.2 Recalleddreams $ n(%)40(75%)27(93%)13(54%) Dreamcontent, n (%) Fighttheme19(36%)13(45%)6(25%) Chasetheme9(17%)6(21%)3(13%) Other8(15%)6(21%)2(8%) FallsfrombedduringDEB, n (%)31(58%)20(69%)11(46%) PD,ParkinsonsDisease;DLB,DementiawithLewyBodies,MCI,MildCognitiveImpairment;MSA,MultipleSystemAtrophy; OSA,ObstructiveSleepApnea;RLS,RestlessLegsSyndrome;PLMD,PeriodicLimbMovementDisorder;SSRI,Selective SerotoninReuptakeInhibitor;SNRI,SelectiveNorepinephrineReuptakeInhibitor. p = 0.01,** p = 0.01,Severityoflimbmovementscored0…3,0 = nomovement,3 = Thrashing/violent moving,#p = 0.028,$p = 0.002. ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 3 S.J.McCarteretal./SleepMedicine (2014) …

PAGE 4

(OR = 10.9,CI = 2.9…54.5, p = 0.0002),andwhenreanalyzedwithonly the11iRBDpatientsseenaftersurveycompletion,thisassociationheld(OR = 12.0,CI = 2.4…78.5, p = 0.002).Inaddition,iRBD ( p = 0.0002)andfallingoutofbedduringDEB( p = 0.03)bothindependentlypredictedmoresevereinjuriesafteradjustingforage, gender,andRBDsymptomduration. ThefrequencyofDEBwasnotassociatedwithinjury,fallingout ofbed,orseverityofreportedinjuries.Comorbiddepression,OSA, antidepressantuse,ordopaminergicdrugusewerenotassociated withinjuryorfrequencyofDEB.Inaddition,dreamtype(“ghting vs.chasing)andvocalization(mumblingvs.screamingorshouting)werenotpredictiveofinjury.However,ayoungerageatRBD symptomonsetwasassociatedwithmorefrequentDEB( p = 0.02) afteradjustingforgender,RBDsymptomduration,andantidepressantuse. 4.Discussion PatientswithiRBDweremorelikelytoinjurethemselvesthan thosewithsymptomaticRBD.Inaddition,patientswhoreported moreseverelimbmovementduringDEB(suchasthrashingand kicking)hadagreaterchanceofinjuringtheirbedpartners.iRBD diagnosisandDEB-relatedfallsfrombedwereassociatedwithmore severeinjuriesinpatientsandtheirbedpartners.Interestingly,DEB frequencywasnotpredictiveofinjuriesorfalls,suggestingthatRBD patientsshouldreceivetreatmenttopreventinjurypotential,regardlessofthereportedDEBsymptomfrequency.Inaddition,DEBrelatedinjurywasnotseenasapresentingcomplaintatoursleep center,suggestingthatpatientsmaynotbeparticularlytroubled byinjurypotential.However,giventhefrequencyofinjurythatwas eventuallyreportedinourpatients,appropriatetreatmentstrategiestopreventfutureinjuryneedstobeconsideredearlyinthe courseofpatientevaluation. Toourknowledge,thisisoneofthe“rstreportsonclinicalfactors associatedwithinjuryinRBDpatients.Theclinicalfeaturesofour respondentsincludingolderage,frequentneurodegenerativedisorders(45%),antidepressantuse(47%),andfrequentinjury(55%) aresimilartopreviouslargeRBDcaseseries [10,19,28] .Injuriessufferedbyourpatientsandtheirbedpartnersrangedfromsimple bumpsandecchymosistobloodyfeetandsubduralhematomas,also similartopreviouslyreportedinjuries [2,10,21,22] OurdatasuggestthatpatientswithsymptomaticRBD,speci“callypatientswithPD,arelesslikelytoinjurethemselvesthanpatientswithiRBD.PatientswithRBDandPDhavebeenshowntobe morelikelytofallfromthebedandinjurethemselvescompared toPDpatientswithoutRBD,butpreviousstudieshavenotcomparedinjuryiniRBDtoPD-RBD [20,29] .Whilewewerenotable toassessthetemporalcourseofRBDfrequencyseverityorinjury potentialinthisretrospectivecohort,iRBDpatientscouldpossibly havemoreinjuriousandsevereDEBsinitially,withDEBsbecomingmilderandlessfrequentastheunderlyingsynucleinopathyprogresses.InvestigatorshavepreviouslyreportedthatRBDsymptoms maydecreaseovertimein26…35%ofpatientswithprogressive neurodegenerativedisorders,withspontaneousremissionofclinicalRBDsymptomsin14…30%ofPD-RBDpatientsperyear [10,30…33] .Otherstudieshavereportedthattheclinicalcoursefor RBDsymptomsmayinitiallyfollowaprogressivelyworseningcourse, followedbyastaticphase,thenasubsequentdecreaseinDEBlater inthedisease [12] .Inaddition,iRBDpatientshavebeenreported tohavemoresevereDEBsandgreaterdreamrecallthanPDand MSApatientswithRBD,althoughdreamrecallorfrequencyand severityofDEBwerenotassociatedwithRBDdiseasestatusinour cohort [34] DecreaseinRBDsymptomswithprogressiveneurodegenerationhasbeensuggestedtobeduetodecreasedpatientfunctionality,moredisruptedsleep,orincreasedmusclerigidityratherthan adecreaseinlesionalpathology,whichcouldpartiallyexplainfewer Fig.1. Subduralhematomaina77-year-oldmanwithidiopathicREMsleepbehaviordisorder.Thepatientdreamthewascatchingapuntwhileplayingfootball anddoveoutofbedstrikinghisheadonthe”oor.Hereportedgaitinstabilityin subsequentweeksandwasfoundtohavelargebilateralfrontalsubduralhematomasrequiringburrholeevacuation.Fiveyearsafterthisevent,hebegandevelopingcognitivedeclineandwassubsequentlydiagnosedwithLewybodydementia. Table2 FrequencyofreportedinjurytypesinRBDpatients. Injurytype, n (%ofinjuredpatients) Mild6(20.7%) Moderate17(58.6%) Marked6(20.7%) Injurydemographics n (%totalpatients) Self20(37.8%) Bedpartner9(17%) Bothselfandbedpartner8(15%) Injuriestopatients( n ,%totalpatients) Subduralhematoma2(4%) Foreheadlaceration2(4%) Fractured/bruisedribs2(4%) Shouldersprain1(2%) Bloodyfeet1(2%) Bruises12(23%) Injuriestobedpartners n (%injuredbedpartners) Bruises6(67%) Openarmwound1(11%) DEBdescriptionsresultinginmarked/moderateinjury Strangledwife Felloutofbedhittingheadon”oorwhilebeingchasedbymenwithgunsŽ Threwdaughteracrossroomaftermistakingherforattackingbear Bitwifesarmwhichrequiredsuturingforrepair Putheadthroughwallresultinginnecksoreness Ranacrossroomintothewall Felloutofbedtryingtocatchapunt Punchedbedpartnerinface ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 4 S.J.McCarteretal./SleepMedicine (2014) …

PAGE 5

injuriesreportedbyoursymptomaticRBDpatients.Interestingly, patientswithPDwereleastlikelytosufferaninjury,possiblydue todegradedmotorfunctioninginPD,renderingDEBlessviolent. Inaddition,ourpatientswithsynucleinopathywereolderthaniRBD patients,therebypotentiallynotbeingasphysicallycapableofinjuriousmovements.Contrarily,RBDpatientshavebeenreportedto displayextraordinarystrengthduringDEBepisodes,soagealone maynotbeindicativeofinjurypotential [22] .OnelargevideopolysomnogramstudyfoundthatDEBmovementsinPD-RBDpatientsweremore”uidthanwakingmovements,suggestingthat corticospinalmotorprojectionsmaybypassŽthebasalganglia, leadingtomorejerkyandviolentmovementsinREMsleepthan duringwakefulnesswhenbasalgangliaprocessingresultsin smoothermovements [22] .Analternativeexplanationforour“nding thatPD-RBDpatientswerelesspronetoinjurythaniRBDpatients mayhavebeenpresentationbiasinourretrospectivestudy,althoughthedearthofsleep-relatedinjuryatpresentationinourstudy cohortdoesnotsuggestthistobelikely.Prospectivestudieswill surelybenecessarytofurtherclarifythereasonswhyPD-RBD patientsmaybelessinjurypronedespitepolysomnographic evidenceformorevigorousREMrelatedthanwakingmotor activity. Wefoundthatyoungeragewasassociatedwithmorefrequent DEBepisodes,potentiallyincreasingopportunitiesforinjury. However,thefrequencyofDEBwasnotassociatedwithinjury,which isconsistentwithonepreviousreportregardingviolentbehaviors duringsleep [35] .Inaddition,previousreportsinpatientswithPD andRBDhaveindicatedthat“ghtingdreammentationandinjury aremorecommoninmenthanwomen [36] .However,nogender differenceswereseenininjuryoccurrenceordreamcontentinour cohort. DEB-relatedfallsfromthebedpredisposedpatientstomore severeinjury.MoreseverelimbmovementsduringDEBwereassociatedwithbedpartnerinjuryandinjuryseverity,intuitivelysensibleaslarger,purposeful,andforcefulmovementssuchaspunching andkickingwouldseemmorelikelytomakecontactwithandpotentiallyinjurethebedpartner. Interestingly,patientswhorecalledtheirdreamcontentwere morelikelytoinjurethemselvesortheirbedpartners.Dreammentationincluding“ghtsorchaseswasreportedby53%ofpatients. Severalpreviousserieshavereportedthataggressivedreamcontent inRBDpatientswasfrequentwhencomparedwithage-andgendermatchedcontrolswithoutanapparentcorrelationwithwaking aggressiveness [21,34,37,38] .Recallbiascouldconfoundthis potentialassociation,aspatientscouldbelikelytorememberdreams inwhichtheyhurtthemselvesortheirbedpartners.However,as suggestedpreviously [39] ,patientswhohavemorevividdreams couldalsobemorelikelytohavecomplexbehaviorsresultingin injury.AsRBDtreatmentdecreasesbothDEBfrequencyandviolent dreamcontent,bothdreamsandbehaviorsinRBDmayhavea commonneuronalgenerator [21,28,38,39] .Ifthisisthecase,then patientswhohavemorevividdreamscouldbemorelikelytohave morecomplexandviolentDEBleadingtoinjury. Inourcohort,58%ofpatientsreportedfallingoutofbedduring DEB,similartoapreviouslyreportedlargeseries [10] .Theriskof fallingshouldbeaddressedwhencounselingRBDpatientsregardingtheimportanceofbedroomsafety.Bedrailsandmattressescan beplacedonthe”oortoprotectagainstinjuryasaresultoffalling outofbedduringDEB.AnovelbedalarmdevicecouldalsopotentiallybeausefuladjunctivetherapytoalertRBDpatientsand/or theirbedpartnerstopossibledangerifthepatientfallsorhasleft thebed [40] Asurprisingandsomewhatcounterintuitive“ndingofourstudy wasthatthefrequencyofDEBwasnotassociatedwithinjuryto eitherthepatientortheirbedpartner,fallingoutofbed,orseverityofinjury,suggestingthatDEBfrequency,severity,andinjurymay actuallybedecoupledinRBDhighlightingtheimportanceofconsideringtreatmentforeachRBDpatient,asinjurymayoccuratany time.ThepracticalimplicationofthisdataisthatRBDpatientsshould betreatedtopreventinjuriouscomplicationsevenwhenDEBsare infrequent.ThetwomaintreatmentoptionsusedforRBDaremelatoninandclonazepam.Melatoninwasrecentlyreportedtobeas effectiveasclonazepamindecreasingDEB,especiallyinneurologicallysensitivepatients,withfewersideeffects [2,7,8,10,21,24,25] Astrengthofourstudyisitsfocusoninjurycharacteristicsina relativelylarge,representative,naturalisticclinicalpracticesample ofRBDpatientswithanacceptableresponseratehavingcomparableclinicalcharacteristicstothosereportedinmostpreviously publishedlargecaseseries [4…6,10,21,24,25] .However,ourstudy hasseveralnotablelimitations.Giventherelativelylowresponse rateof40%overall,concernoverrespondentbiasexists,asitispossiblethatpatientshavingmoresevereorrecentinjuriescouldbe morelikelytohaverespondedtooursurvey,whichcouldfalsely raiseestimatesofthefrequencyandseverityofinjuryinthisseries. AsasurveyofinjuriesoccurringinRBD,itisdiculttodetermine timingoffallsandinjuryandwhetherornottheseoccurredconcomitantlyorasseparateevents.Inaddition,itisverylikelythat wemaynothavebeenabletoidentifyandcontrolforseveralpotentialconfoundingbiasesinselection,referral,sampling,response,andrecall.Presumably,ourcasesmaybemorelikelytohave morefrequentand/orsevereRBDthanthoseevaluatedin community-basedsleepdisordercenters,andthusourconclusionsmaynotbegeneralizabletoallRBDpatients.Furthermore,it couldbearguedthatpatientswithsymptomaticRBDmayhavebeen morepronetocognitiveimpairmentsthatlimitedtheirrecallofpast injuriespriortotreatmentascomparedtoiRBDpatientswithpreservedcognitivefunctioning.Last,wewereunabletoperformneurologicalexaminationatthetimeofsurveyresponseforallsubjects andthereforebasedcategorizationofiRBDorsymptomaticRBDdiagnosisstatusonthelastclinicalfollow-up,possiblyleadingto miscategorizationandunderestimationofpatientswhomayhave convertedfromiRBDtosymptomaticRBDatthetimeofinjury,a particularconcernasrecentevidencesuggeststhatapproximately81%ofpatientswithiRBDeventuallydevelopovertsymptoms andsignsofneurodegenerationwithseriallong-termneurologicalfollow-up [18,41] .Misclassi“cationofiRBDorsymptomaticRBD diagnosismayhaveledtoanoverestimationoftheassociationof iRBDdiagnosistypewithinjuryandinjuryseverity.However,we wereabletoverifyprogression-freestatusandiRBDin11of28subjects,andtheremainingiRBDsubjectshadameandurationbetween thelastclinicalfollow-upandsurveycompletionof22.3months, suggestingthatthemajorityofouriRBDpatientswereaccurately classi“ed.Additionally,wereanalyzedthedataforassociations betweeniRBDandinjuryandinjuryseverityonlyincludingthe11 iRBDpatientswhosediagnoseswereveri“edagainfollowingsurvey completion(thuscon“rmingtheirstatusattimeofsurveycompletion),andfoundsimilar,andevenstronger,associations,althoughgiventhesmallernumberofsubjects,CIestimateswidened asexpected.Futureprospectivestudieswillbenecessarytoaddress theseinherentmethodologicalproblemsinsurvey-basedresearchandtodeterminetheimpactandtimingofinjuriesandpotentialinjury-provokingeventssuchasfallstofurtherdelineatethe relationshipbetweeninjuriesandiRBDandsymptomaticRBDdiagnosistype,DEBfrequencyandseverity,aswellastheimpactof treatmentsoninjurypotential. OurdatashowthatiRBDpatientsappeartobemorelikelyto sufferinjurythansymptomaticRBDpatients,andthatfrequency ofDEBdoesnotappeartopredicteitherinjuryorinjuryseverity, asurprising“ndingwhichsuggeststhattreatmentshouldbeconsideredforallpatientsreportingRBDsymptomsregardlessoffrequencyandseverityofbehaviorstopreventtheirpotentialforinjury. Melatoninmaybeaparticularlyusefultreatmenttoconsiderfor ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 5 S.J.McCarteretal./SleepMedicine (2014) …

PAGE 6

symptomaticRBDpatients,giventhatthesepatientsmaybeless pronetoinjury,butalsoassymptomaticpatientsmaybemoresensitivetowardadversetreatmenteffectssuchascognitivedysfunctionordruginteractionswithothercentrallyactivedopaminergic oranticholinesterasemedications.Inaddition,bedroomsafetymust beaddressedforallpatientswithRBDtopreventinjuriousfallsfrom bed.Weplantoconductfutureprospectivestudiestobetterde“ne predictorsofinjuryinRBD. Disclosures Therewasnooff-labelmedicationuseassociatedwiththisproject. TheprojectdescribedwassupportedbytheNationalCenterfor ResearchResourcesandtheNationalCenterforAdvancingTranslationalSciences,NationalInstitutesofHealth,throughGrantNumber 1UL1RR024150-01. Mr.McCarterreportsnodisclosures. Dr.St.LouisreportsgrantsfromMayoClinicCTSAduringthe courseofthisstudyandothersfromInspire,Inc.outsidethesubmittedwork. Dr.Boswellreportsnodisclosures. Mr.Dueffertreportsnodisclosures. Ms.Slocumbreportsnodisclosures. Dr.BoevehasservedasaninvestigatorforclinicaltrialssponsoredbyCephalon,Inc.,AllonPharmaceuticals,andGEHealthcare.Hereceivesroyaltiesfromthepublicationofabookentitled BehavioralNeurologyofDementia (CambridgeMedicine,2009).He hasreceivedhonorariafromtheAmericanAcademyofNeurology. HeservesontheScienti“cAdvisoryBoardoftheTauConsortium. HereceivesresearchsupportfromtheNationalInstituteonAging (P50AG016574,U01AG006786,RO1AG032306,RO1AG041797) andtheMangurianFoundation. Dr.Silberreportsnodisclosures. Dr.Olsonreportsnodisclosures. Dr.Morgenthalerreportsnodisclosures. Dr.Tippmann-Peikertreportsnodisclosures. Con”ictofinterest TheICMJEUniformDisclosureFormforPotentialCon”ictsofInterestassociatedwiththisarticlecanbeviewedbyclickingonthe followinglink: http://dx.doi.org/10.1016/j.sleep.2014.06.002 Acknowledgments TheprojectdescribedwassupportedbytheNationalCenterfor ResearchResourcesandtheNationalCenterforAdvancingTranslationalSciences,NationalInstitutesofHealth,throughGrantNumber 1UL1RR024150-01.Thecontentissolelytheresponsibilityofthe authorsanddoesnotnecessarilyrepresenttheocialviewsofthe NIH.TheauthorsalsowishtoacknowledgeMs.LoriLynnReinstrom forassistanceinpreparationofthemanuscript. Appendix:Supplementarymaterial Supplementarydatatothisarticlecanbefoundonlineat doi:10.1016/j.endend.2013.05.004 References[1] Theinternationalclassi“cationofsleepdisorders:diagnostic&codingmanual. 2nded.AmericanAcademyofSleepMedicine;2005. [2] BoeveBF.REMsleepbehaviordisorder:updatedreviewofthecorefeatures, theREMsleepbehaviordisorder-neurodegenerativediseaseassociation, evolvingconcepts,controversies,andfuturedirections.An nNYAcadSci 2010;1184:15…54. [3] BoeveBF,MolanoJR,FermanTJ,SmithGE,LinSC,BieniekK,etal.Validation oftheMayoSleepQuestionnairetoscreenforREMsleepbehaviordisorderin anaginganddementiacohort.SleepMed2011;12:445…53. [4] BoeveBF,SilberMH,FermanTJ,LinSC,BenarrochEE,SchmeichelAM,etal. Clinicopathologiccorrelationsin172casesofrapideyemovementsleep behaviordisorderwithorwithoutacoexistingneurologicdisorder.SleepMed 2013. [5] BootBP,BoeveBF,RobertsRO,FermanTJ,GedaYE,PankratzVS,etal.Probable rapideyemovementsleepbehaviordisorderincreasesriskformildcognitive impairmentandParkinsondisease:apopulation-basedstudy.AnnNeurol 2012;71:49…56. [6] FrauscherB,GschliesserV,BrandauerE,MartiI,FurtnerMT,UlmerH,etal. REMsleepbehaviordisorderin703sleep-disorderpatients:theimportance ofelicitingacomprehensivesleephistory.SleepMed2010;11:167…71. [7] IranzoA,SantamariaJ,TolosaE.Theclinicalandpathophysiologicalrelevance ofREMsleepbehaviordisorderinneurodegenerativediseases.SleepMedRev 2009;13:385…401. [8] McCarterSJ,StLouisEK,BoeveBF.REMsleepbehaviordisorderandREMsleep withoutatoniaasanearlymanifestationofdegenerativeneurologicaldisease. CurrNeurolNeurosciRep2012;12:182…92. [9] MontplaisirJ,GagnonJF,FantiniML,PostumaRB,DauvilliersY,DesautelsA, etal.PolysomnographicdiagnosisofidiopathicREMsleepbehaviordisorder. MovDisord2010;25:2044…51. [10] OlsonEJ,BoeveBF,SilberMH.Rapideyemovementsleepbehaviourdisorder: demographic,clinicalandlaboratory“ndingsin93cases.Brain2000;123(Pt 2):331…9. [11] SchenckCH,BundlieSR,EttingerMG,MahowaldMW.Chronicbehavioral disordersofhumanREMsleep:anewcategoryofparasomnia.Sleep 1986;9:293…308. [12] ChiuHF,WingYK,LamLC,LiSW,LumCM,LeungT,etal.Sleep-relatedinjury intheelderly…anepidemiologicalstudyinHongKong.Sleep2000;23:513…17. [13] BodkinCL,SchenckCH.Rapideyemovementsleepbehaviordisorderinwomen: relevancetogeneralandspecialtymedicalpractice.JWomensHealth(Larchmt) 2009;18:1955…63. [14] BonakisA,HowardRS,EbrahimIO,MerrittS,WilliamsA.REMsleepbehaviour disorder(RBD)anditsassociationsinyoungpatients.SleepMed2009;10:641…5. [15] JuYE,Larson-PriorL,DuntleyS.ChangingdemographicsinREMsleepbehavior disorder:possibleeffectofautoimmunityandantidepressants.SleepMed 2011;12:278…83. [16] TemanPT,Tippmann-PeikertM,SilberMH,SlocumbNL,AugerRR.Idiopathic rapid-eye-movementsleepdisorder:associationswithantidepressants, psychiatricdiagnoses,andotherfactors,inrelationtoageofonset.SleepMed 2009;10:60…5. [17] IranzoA,LomenaF,StocknerH,ValldeoriolaF,VilasecaI,SalameroM,etal. Decreasedstriataldopaminetransporteruptakeandsubstantianigra hyperechogenicityasriskmarkersofsynucleinopathyinpatientswithidiopathic rapid-eye-movementsleepbehaviourdisorder:aprospectivestudy[corrected]. LancetNeurol2010;9:1070…7. [18] IranzoA,TolosaE,GelpiE,MolinuevoJL,ValldeoriolaF,SerradellM,etal. Neurodegenerativediseasestatusandpost-mortempathologyinidiopathic rapid-eye-movementsleepbehaviourdisorder:anobservationalcohortstudy. LancetNeurol2013;12:443…53. [19] ClaassenDO,JosephsKA,AhlskogJE,SilberMH,Tippmann-PeikertM,Boeve BF.REMsleepbehaviordisorderprecedingotheraspectsofsynucleinopathies byuptohalfacentury.Neurology2010;75:494…9. [20] ComellaCL,NardineTM,DiederichNJ,StebbinsGT.Sleep-relatedviolence, injury,andREMsleepbehaviordisorderinParkinsonsdisease.Neurology 1998;51:526…9. [21] SchenckCH,MahowaldMW.REMsleepbehaviordisorder:clinical, developmental,andneuroscienceperspectives16yearsafteritsformal identi“cationinSLEEP.Sleep2002;25:120…38. [22] ArnulfI.REMsleepbehaviordisorder:motormanifestationsand pathophysiology.MovDisord2012;27:677…89. [23] SchenckCH,LeeSA,BornemannMA,MahowaldMW.Potentiallylethal behaviorsassociatedwithrapideyemovementsleepbehaviordisorder:review oftheliteratureandforensicimplications.JForensicSci2009;54:1475…84. [24] SchenckCH,MahowaldMW.Polysomnographic,neurologic,psychiatric,and clinicaloutcomereporton70consecutivecaseswithREMsleepbehavior disorder(RBD):sustainedclonazepamecacyin89.5%of57treatedpatients. CleveClinJMed1990;57. [25] McCarterSJ,BoswellCL,StLouisEK,DueffertLG,SlocumbN,BoeveBF,etal. TreatmentoutcomesinREMsleepbehaviordisorder.SleepMed2013;14:237… 42. [26] HarrisPA,TaylorR,ThielkeR,PayneJ,GonzalezN,CondeJG.Researchelectronic datacapture(REDCap)…amet adata-drivenmethodologyandwork”owprocess forprovidingtranslationalresearchinformaticssupport.JBiomedInform 2009;42:377…81. [27] IranzoA,SantamariaJ.Severeobstructivesleepapnea/hypopneamimicking REMsleepbehaviordisorder.Sleep2005;28:203…6. [28] SchenckCH,HurwitzTD,MahowaldMW.Symposium:Normalandabnormal REMsleepregulation:REMsleepbehaviourdisorder:anupdateonaseries of96patientsandareviewoftheworldliterature.JSleepRes1993;2: 224…31. [29] WallaceDM,ShafazandS,CarvalhoDZ,NahabFB,SengunC,RussellA,etal. Sleep-relatedfallingoutofbedinParkinsonsdisease.JClinNeurol.2012;8: 51…7. ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 6 S.J.McCarteretal./SleepMedicine (2014) …

PAGE 7

[30] BoeveBF,SilberMH,FermanTJ,KokmenE,SmithGE,IvnikRJ,etal.REMsleep behaviordisorderanddegenerativedementia:anassociationlikelyre”ecting Lewybodydisease.Neurology1998;51:363…70. [31] BoeveBF,SilberMH,ParisiJE,DicksonDW,FermanTJ,BenarrochEE,etal. SynucleinopathypathologyandREMsleepbehaviordisorderplusdementia orparkinsonism.Neurology2003;61:40…5. [32] GjerstadMD,BoeveB,Wentzel-LarsenT,AarslandD,LarsenJP.Occurrenceand clinicalcorrelatesofREMsleepbehaviourdisorderinpatientswithParkinsons diseaseovertime.JNeurolNeurosurgPsychiatry2008;79:387…91. [33] LavaultS,Leu-SemenescuS,TezenasduMontcelS,CochendeCockV,Vidailhet M,ArnulfI.Doesclinicalrapideyemovementbehaviordisorderpredictworse outcomesinParkinsonsdisease?JNeurol2010;257:1154…9. [34] IranzoA,SantamariaJ,RyeDB,ValldeoriolaF,MartiMJ,MunozE,etal. CharacteristicsofidiopathicREMsleepbehaviordisorderandthatassociated withMSAandPD.Neurology2005;65:247…52. [35] OhayonMM,SchenckCH.Violentbehaviorduringsleep:prevalence, comorbidityandconsequences.SleepMed2010;11:941…6. [36] BjornaraKA,DietrichsE,ToftM.REMsleepbehaviordisorderinParkinsons disease…isthereagenderdifference?ParkinsonismRelatDisord2013;19: 120…2. [37] CochendeCockV.Recentdataonrapideyemovementsleepbehaviordisorder inpatientswithParkinsonsdisease:Analysisofbehaviors,movementsand periodiclegmovements.SleepMed2013;14(8):749…53. [38] FantiniML,CoronaA,ClericiS,Ferini-StrambiL.Aggressivedreamcontent withoutdaytimeaggressivenessinREMsleepbehaviordisorder.Neurology 2005;65:1010…15. [39] SantamariaJ,CarrascoE,KumruH.Relationbetweendreamcontentand movementintensityinREMsleepbehaviordisorder.Sleep2004;27:A289. [40] HowellMJ,ArnesonPA,SchenckCH.AnoveltherapyforREMsleepbehavior disorder(RBD).JClinSleepMed2011;7:639…44A. [41] SchenckCH,BoeveBF,MahowaldMW.Delayedemergenceofaparkinsonian disorderordementiain81%ofoldermeninitiallydiagnosedwithidiopathic rapideyemovementsleepbehaviordisorder:a16-yearupdateonapreviously reportedseries.SleepMed2013;14:744…8. ARTICLEINPRESS Pleasecitethisarticleinpressas:StuartJ.McCarter,etal.,FactorsassociatedwithinjuryinREMsleepbehaviordisorder,SleepMedicine(2014) ,doi: 10.1016/ j.sleep.2014.06.002 7 S.J.McCarteretal./SleepMedicine (2014) …