1 HEALTH COMORBIDIT I ES AND COGNITION IN NON DISEASE PATIENTS By JACOB D. JONES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTERS OF SCIENCE UNIVERSITY OF FLORIDA 2012
2 2012 Jacob D. Jones
3 TABLE OF CONTENTS page A CKOWLEDGEMENTS ................................ ................................ ................................ .. 5 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 Introduction to the Problem ................................ ................................ ..................... 10 ................................ ...................... 12 Cognition and Comorbidities in the Normal Elderly. ................................ ......... 13 ................................ ....... 15 Statement of the Problem and S pecific Aims ................................ .......................... 17 Specific Aim 1 ................................ ................................ ................................ ... 18 Specific Aim 2 ................................ ................................ ................................ ... 18 2 METHODS ................................ ................................ ................................ .............. 20 Inclusion Criteria ................................ ................................ ................................ ..... 21 Neuropsychology Module of INFORM ................................ .............................. 22 Medi cal Chart Review ................................ ................................ ...................... 23 Exclusion Criteria & Final Sample ................................ ................................ ........... 25 Statistical Analyses ................................ ................................ ................................ 25 Aim 1: Prevalence ................................ ................................ ............................ 25 Aim 2: Cognition ................................ ................................ ............................... 26 3 RESULTS ................................ ................................ ................................ ............... 28 Specific Aim 1: Prevalence of Comorbidities ................................ .......................... 29 Comorbidity Prevalence and Age ................................ ................................ ..... 31 Influence of Demographic and Disease Se verity on Comorbidity Prevalence .. 33 Specific Aim 2: Comorbidities and Cognition ................................ .......................... 37 Hypertension and Hypotension: Influences on C ognition ................................ 41 Pulse Pressure ................................ ................................ ................................ 43 4 DISCUSSION ................................ ................................ ................................ ......... 50 Prevalence: Aim 1 ................................ ................................ ................................ .. 50 Comorbidities and Cognition: Aim 2 ................................ ................................ ........ 53 Unexpected Comorbidity Cognition Relationship: Hypotension ....................... 56
4 Comorbidities and Cognition: Pulse Pressure ................................ .................. 58 Comorbidities and Cognition: Possible Mechanisms ................................ ........ 59 Limitations and Future Directions ................................ ................................ ..... 61 Conclusion ................................ ................................ ................................ ........ 62 APPENDIX A P REVALENCE OF COMORBIDITIES ................................ ................................ .... 65 B P ERCENT OF COMORBITIES REPORTED BY AGE ................................ ............ 67 REFERENCE LIST ................................ ................................ ................................ ........ 68 B IOGRAPHICAL SKETCH ................................ ................................ ............................ 77
5 ACKNOWLEDG E MENTS I thank my mentor and chair Dr. Dawn Bowers, and my committee supervisory committee for their mentoring. I thank everyone at the Center for Movement Disorders and Neurorestoration for their work guidance and support. I thank my family for their encouragement and continuing support throughout my education.
6 LIST OF TABLES Table page 2 1 Cognitive d omains. ................................ ................................ ............................. 23 2 2 List of c omorbidities ................................ ................................ ............................ 25 3 1 Sample c haracteristics ................................ ................................ ....................... 28 3 2 Predictor s of h ypertension ................................ ................................ .................. 33 3 3 Predictors of h ypotension ................................ ................................ ................... 34 3 4 Predictors of c ardiac d isease ................................ ................................ .............. 35 3 5 Predictors of c holesterolemia ................................ ................................ ............. 35 3 6 Predictors of a rthritis ................................ ................................ ........................... 36 3 7 P redictors of p rost ate d isease ................................ ................................ ............ 36 3 8 Cognitive c omposite s cores f or e ach d omain. ................................ .................... 38 3 9 Comorbidities and e xecutive f unction ................................ ................................ 39 3 10 Comorbidities and verbal m emory ................................ ................................ ...... 39 3 11 Comorbidities and processing s peed ................................ ................................ .. 40 3 12 Comorbidities and working m emory ................................ ................................ ... 40 3 13 Comorbidities and l anguage ................................ ................................ ............... 41 3 14 Pulse pressure and executive f un ction ................................ ............................... 44 3 15 Pulse p ressure and verbal m emory ................................ ................................ .... 46 3 16 Pulse pressure and p rocessing s peed ................................ ................................ 47 3 17 Pulse p ressure and working m emory ................................ ................................ 48 3 18 Pulse p ressure and l anguage ................................ ................................ ............. 49
7 LIST OF FIGURES Figure page 1 1 Prevalence of c omorbidities in 1948 PD p atients from NPF QII d atabase ................................ ............................ 17 2 1 Design o verview. ................................ ................................ ............................... 20 3 1 Prevalence of comorbidities in PD s ample (N=341). ................................ .......... 30 3 2 Occurrence of h ypotension and h ypertension in Parkinson s ample ................... 31 3 3 Comorbidi ty prevalence according to age g rouping. ................................ .......... 32 3 4 Executive and v erbal m emory scores a mong b lood p ressure g roups. ............... 42 3 5 Pulse p ressure X UPDRS interaction e xecutive f unction.. ................................ 45 3 6 Pulse pressure X UPDRS i nteraction v erbal m emory.. ................................ ..... 46 3 7 Pulse pressure X UPDRS i nteraction p rocessing s peed.. ................................ 47 4 1 Overlap of white m atter l disease and n ormal a ging.. .......... 61
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Masters of Science HEALTH COMORBIDIT I ES AND COGNITION IN NON DISEASE PATIENTS By Jacob D. Jones May 2012 Chair: Dawn Bowers Major: Psychology Background: Health comorbidities, particularly cardiovascular risk factors, are well known to pose risks for cognitive decline in older adults. To date, little attention has dopaminergic depletion disorder affecting frontal subcortical systems. This study examined the prevalence and contribution of comorbidities on cognitive status in PD patients above and beyond the effects of disease severity. We had two hypotheses. First, comorbidities associated with small vessel disease (i.e., hypertension) would negatively influence executive function, processing speed, and working memory. Second, comorbi dities associated with altered glucose metabolism or acute cardiac events would negatively influence episodic memory based on known associations with hippocampal function. Methods: A cross sectional design was used, including neuropsychological data on 34 1 PD patients obtained from the INFORM database from the Center of Movement Disorders and Neuro Restoration at the University of Florida (UF) Comorbidity data were collected via medical chart review from UF/ShandsClinics. Data were analyzed
9 using a serie s of multiple hierarchical regressions, controlling for PD related disease variables. Results: Overall sample characteristics: 69% male, disease duration 9.7 years, (41. 6%), hypertension (38.1%) and hypotension (30.2%) were the most reported comorbidities. After excluding for possible dementia (Dementia Rating Scores< 5 th percentile) the presence of hypertension significantly contributed to domains of executive function and verbal memory. Conclusions: This study with a large cohort of PD patients provides evidence for a detrimental influence of health comorbidities, particularly hypertension, on cognitive domains that have traditionally been conceptualized as being front ally and/or temporally mediated. Our hypotheses were partially supported and generally in line with findings from normal elderly. Possible mechanisms are discussed.
10 CHAPTER 1 INTRODUCTION Introduction to the Problem e related neurodegenerative disorder, ffecting over half a million people in the United Stated alone annual indirect costs of PD are estimated to be around $20 billion (Huse et al., 2 005). more prevalent in older populations, affecting 1% of the population over 60 years old and reaching prevalence rates of up to 4% in individuals over 80 (Lau & Breteler, 2006). In addition to having an age effect on prevalen ce PD has been shown to be more common among men (Mayeux et al., 1995). symptoms including tremor, rigidity, slowness, stooped posture and balance impairment. However, non motor symptoms are also common giving rise to the view that Parkinson disease is in fact a multisystem disorder involving mood, autonomic, and cognitive symptoms. Cognitive impairment can occur early in the disease course and have a major impact on qualit y of life, survival and the need for nursing home placement (Aarsland et al., 2001; Schrag et al., 2000; Buter et al., 2008). An understanding of the brain mechanisms of cognitive impairment in PD is essential both prognostically and for the development o f new targeted therapies. The most prominent view regarding cognitive symptoms in PD is that they are related to deregulation of the frontal neocortex secondary to reduced dopaminergic input from subcortical systems (Dubois & Pillon, 1997). While PD rel ated neural degeneration is a primary cause of cognitive impairment, it does not account for all of
11 the variance in cognitive decline. Research into additional contributors of cognitive impairment could help in treatment and identification of individuals a t risk for further decline. The overall goal of the present study was to examine other contributory factors that might influence cognitive status in Parkinson patients. Studies among the normal elderly have shown a relationship between cognitive decline and cardiovascular comorbidities such hypertension, high cholesterol and diabetes. The question arises whether similar health comorbidites might also exert a negative influence on cognitive status in patients with PD. Given that a high proportion of PD p atients have risk factors will exert the same type of influence as seen in normal elderly. The specific goals of the present study were to examine the occurrence of he alth comorbidities in large sample of Parkinson patients and determine the extent to which these comorbidities negatively influenced distinct neurocognitive domains (i.e., executive function, processing speed, episodic memory, etc.). To accomplish these g oals, the present study undertook a medical review of Parkinson patients who were part of the Center of Movement Disorders and Neurorestoration (CMDNR) INFORM database and who had undergone a neuropsychological evaluation through the Psychology Clinic at t he University of Florida. We hypothesized that cardiovascular comorbidities, such as hypertension, would exert negative influence on cognitive domains (ie. executive functioning, working memory and processing speed) that are linked to frontal subcortic al integrity. This hypothesis was based on findings from the normal aging literature showing relationships between hypertension, white matter
12 abnormalities in the frontal sub cor tical regions and cognitive performance (Kennedy & Raz, 2009; Hachinski, 1987; Verdelho et al., 2010; Pantoni, 2010) Before discussing this study further, we will provide a brief review of relevant literature including the results from a preliminary study that led to the present project. When originally described in 1817 by James Parkinson, the syndrome which impaired cognitive or mood sequelae. Since that time, numerous studies have shown that cognitive ch anges and decline are indeed common in PD (Levin & Katzen, 2005; Dubois & Pillon, 1997; Zgaljardic et al., 2003; Aarsland & Kurz, 2010). In cross sectional studies, over a quarter of the PD population has disease related dementia whereas longitudinal stu dies suggest that over 80% will develop dementia after 10 15 years following disease onset (Aarsland et al., 2001; Hely et al., 2008). There are two prominent neurotransmitter views regarding cognitive decline in Parkinson disease. One relates to dopa mine depletion in subcortical regions that alters the functional integrity of the frontal neocortex. Cronin Golomb (2010) has argued that dopaminergic degeneration of ne urons in the substantia nigra disrupt multiple basal ganglia loops (Alexander et al., 1986). While deregulation of certain basal ganglia loops results in motor symptoms, circuits with connections to areas in the dorsolateral prefrontal cortex, orbitofronta l cortex and anterior cingulated region may be responsible for impairments in cognitive and behavioral functions. The traditional cognitive phenotype of PD includes cognitive slowing, difficulties with working memory and multi
13 tasking, poor set shifting, and forgetfulness (Zgaljardic et al., 2003, Levin & Katzen, 2005; Zahodne et al., 2011). A second view relates to additional cholinergic deficits that are superimposed on the primary dopaminergic subcortical degeneration (Hilker et al., 2005; Murat, 2003) These cholinergic deficits, thought to arise from disruption of the basal forebrain, impact wide areas of the cortex, particularly posterior regions, and are associated with more severe dementia (rather than mild cognitive changes). Findings by Hilker and colleagues (2005) suggest that development of PD dementia is related to a transition from dopaminergic degeneration in the substantia nigra (where degeneration is first noted in early PD) to further degeneration of cholinergic neurons in the cortex. T he role of Lewy body related degeneration in the development of PD is not thoroughly understood. Lewy bodies are irregular accumulations of protein (mainly alpha synuclein) found in the substantia nigra and may play a role in dopaminergic degeneration (Mer edith et al., 2004). In addition to the substantia nigra, other brain regions such as the locus c o eruleus, hypothalamus and thalamus, may be sites for Lewy body formation (Gibb & Lees, 1988). The presence of Lewy bodies in the cortex of PD patients may be characteristic of dementia (Murat, 2003). Cognition and Comorbidities in the Normal Elderly. For over a decade studies have shown that cardiovascular risk factors (including comorbidities such as hypertension and diabetes) are predictive of cognitive decli ne among elderly individuals (Saxby et al., 2003; Kuo et al., 2004; Kilander et al., 1998; Vicario et al., 2005). A previous longitudinal study examining a wide array of cardiovascular risk factors found only hypertension and diabetes to be related to cogn itive performance among elderly individuals (Knopman et al., 2001). Specifically
14 hypertension and diabetes are risk factors for cognitive symptoms in the areas of memory, attention, executive functioning and processing speed (Saxby et al., 2003). Even amo ng treated hypertensive individuals taking blood pressure medications, executive dysfunction may be present when compared to normotensive individuals (Raz et al., 2003). In addition to a clinical diagnosis of hypertension, subclinical levels of blood press ure, pulse pressure (the difference between systolic and diastolic blood pressure correlating with arterial thickness) and glucose levels have been linked to worse cognitive outcomes (Dahle et al., 2009). Cardiovascular risk factors have been linked to bo th neuropsychological outcomes and brain structure correlates. The effect of hypertension on cognition is thought to be due to subcortical small vessel lesions that can occur in various subcortical areas, particularly the white matter of the frontal and te mporal lobes. In turn, these lesions may impact executive functioning, working memory, processing speed and memory recall (Kennedy & Raz, 2009; Hachinski et al., 1987; Verdelho et al., 2010; Pantoni, 2010). Several studies have shown that the presence o f small vessel lesions, appearing as white matter lesions (WML) on magnetic images, mediates the relationship between memory, executive functioning and hypertension. Greater WML volume in the frontal lobe was significantly related to executive dysfunction regardless of whether or not individuals were on hypertensive medication (Raz et al., 2003). In addition to executive functioning, memory functioning has been associated with temporal lobe and periventricular WML (Smith, et al., 2011). A recent study look ing at different subtypes of small vessel lesions found that hypertension was indicative of both
15 microatheroma (lesions of very small vessels) and lipohyalinosis (lesions of larger small vessels) while diabetes only related to lipohyalinosis (Bezerra et al ., 2012). Diabetes and acute hypoxic events have shown to be related to executive functioning and memory performance (Verdelho et al., 2007; Stewart & Liolitsa, 1999). Furthermore, the detrimental effect of diabetes is independent of small vessel disease, and may be due to additional disruption of glucose metabolism preferentially impacting posterior regions involved in memory functioning (Convit et al., 2003; Leritz et al., 2011). While numerous studies have found hypertension to be related to cognitive impairment, some studies have failed to report a significant relationship between hypertension and cognition (Farmer et al., 1987; Desmond et al., 1993; Launer et al., 1995). A review of previous works on cognition and hypertension came to the conclusion that a majority of longitudinal studies have shown that hypertension does indeed contribute to the development of cognitive impairment (Birns & Karla, 2008). Few studies have examined the relationship betw een cardiovascular comorbidities and cognitive performance in patients with PD. To date, the primary approach taken in the literature has involved determining whether the presence of cardiovascular comorbidities is differentially associated with diagnoses of dementia, mild cognitive impairment or no cognitive impairment (Uc et al., 2009; Haugarvoll et al., 2005; Dalaker et al., 2009). In brief, these studies find no differences in occurrence of cardiovascular comorbidities among demented vs nondemented PD subgroups, or those with mild cognitive impairment In contrast, a recent study in our laboratory was supportive of a relationship between cardiovascular symptoms and simple measures of cognition (Jones et al.,
16 2012). This study incorporated previously col Foundation Quality Improvement Initiative (NPF QII). This sample included data on comorbidity data was limited to six comorbidit ies ( heart/circulation problems diabetes arthritis cancer, respiratory disease other neurological disorders C ognitive data was limited to a verbal fluency task (animal fluency) and a five word immediate and short delay recall In terms of overa ll prevalence, arthritis (47%) and heart/circulatory problems (36%) were the most prevalent comorbidities (Figure 1 1). While heart/circulatory problems were common, they were below what is often reported for prevalence estimates of hypertension ( 60% ) in normal elderly over age 60 years (Wolf Maier et al., 2003 Aronow et al., 2011 ). This lower rate of heart/circulatory problems may be related to a reduction of blood pressure which has shown to be lower among patients on levodopa medication, (McDowell et a l., 1970; Nanhoe Mahabier et al., 2009). In addition to investigating prevalence, the Jones et al. (2012) study also examined the relationship between comorbidities and cognition (animal fluency, delayed memory recall). This was done using a series of hi erarchical regression analyses that controlled for demographic differences (age, gender, living status), PD severity, and medication status. Results showed that heart/circulation problems were associated with worse delayed verbal memory recall performanc e, while presence of diabetes showed a trend relationship with worse semantic performance. While this study provides initial evidence for detrimental impact on vascular related comorbid i ties on cognition in PD, weaknesses are present. This study failed t o differentiate the
17 relationship of hypertension from cardiac disease, as hypertension was lumped into the category of heart circulation problems. Secondly, the effect of hypotension, often observed in PD, was not addressed. Third, the measures of cogniti on were limited and did not broadly capture various neurocognitive domains (i.e., working memory, processing speed, etc.). Finally, there were a variety of other relevant issues with the NPF level (which is strongly correlated with cognition), or whether individuals met criteria for dementia. Figure 1 1 Prevalence of Comorbidities in 1948 PD Patients from NPF QII Database Figure depicts the prevalence of comorbidities among a sample of Statement of the Problem and Specific Aims The overall goal of this study was to examine the prevalence of comorbidities among a sample of non demented Parkinson patients, an d to determine the relationship of these comorbidities to separate cognitive domains. This study attempted to address
18 some of the issues uncovered in our previous NPF study First, this study assessed multiple domains of cognition, rather than global defi cits (ie. dementia vs non dementia). Secondly, this study examined a wide range of comorbidities allowing examination of distinct comorbidities (ie. hypertension, cholesterol and hypotension) from the same organ system. Third, this sample consisted of a la rge cohort of Parkinson patients, screened for possible dementia This study had two specific aims. Specific Aim 1 The first aim was to determine the prevalence of health comorbidities in a University of Florida cohort of P arkinson patients. Based on lit erature in normal elderly, we were particularly interested in the influence of cardiovascular comorbidities such as hypertension, diabetes and acute cardiac event s. Hypothesis 1 : Based on previous studies showing age related increases in comorbidities i n conjunction with the older age of the Parkinson cohort, it was hypothesized that comorbidity prevalence in PD patients would roughly mirror that observed n normal elderly with the exception of hypertension/hypotension. We predicted that hypertension w ould be a common comorbidity in PD but would occur less frequently than in the normal elderly populations (i.e., around 60%). This prediction is based on knowledge about the side effects of dopaminergic medications (lowering of blood pressure and PD auton would be greater in our PD sample due to PD related autonomic dysfunction. Specific Aim 2 The second aim was to examine the unique contribution of comorbidities to specific cognitive domains in P D. Previous research has shown that severity of PD is a
19 main predictor of cognitive functioning; therefore PD related disease variables were statistically controlled for in order to examine the impact that individual comorbidities have on cognitive domain s above and beyond PD related degeneration Hypothesis 1 : The first hypothesis was that t hose comorbidities associated with frontal subcortical disruption (i.e., hypertension) w ould be associated with cognitive performance in the executive functioning, working memory and processing speed domains. These domains are conceptualized as being mediated by frontal lobe processes. Specifically, the occurrence of hypertension would be related to worse scores on measures of executive function, working memory, an d processing speed. Hypothesis 2 : The second hypothesis was that worse verbal memory performance w ould be associated with the presence of acute cardiac events and diabetes, due to potential impact on mesial temporal lobe memory systems (i.e., via altera tions in glucose metabolism or hypoxic events)
20 CHAPTER 2 METHODS Figure 2 1 provides an overview of the study design. In brief, the participants were a convenience sample of Parkinson patients who had undergone neuropsychological evaluation at the Univ ersity of Florida Psychology Clinic. All were being followed in the clinics of the UF Center for Movement Disorders and Neurorestoration. Inclusion criteria are detailed more fully below but include d : a diagnosis of idiopathic PD, available neurocognitiv e test results, and medical workup corresponding to the date of the neuropsychological evaluation (within a minimum of 1 year prior). Figure 2 1. Design Overview. Figure depicts an overview of the study design
21 Inclusio n Criteria Participants were recruited from the INFORM database, a n IRB approved clinical research database Disorders and Neurorestoration (CMDN). To be included in the present study, parti cipants had to meet the following requirements: a) age 18 years or older, b) completion of a neuropsychological evaluation by UF neuropsychologists; and c) meet diagnostic criteria for idiopathic PD by a movement disorders specialty trained neurologist u sing the United Kingdom Brain Bank diagnostic criteria (Hughes et al., 1992). These criteria consist of presence of bradykinesi a and one of the following: muscle rigidity, resting tremor of 4 to 6 Hertz or postural instability. Data pertaining to PD sever ity, duration of symptoms and dopaminergic medication w ere collected from the INFORM database. All measures of disease severity were assessed within a year of neuropsychological testing. These included the Unified Third E dition (UPDRS; Fahn & Elton, 1987), a well validated and standardized rating of PD symptom severity. Motor scores (part III) of the UPDRS were obtained while patients were on medication s by a trained movement disorder neurologist at CMD N. Years with symptoms w ere obtained from the INFORM database as a self reported question assessed by a CMDN neurologist. The amount of dopaminergic medication was indexed by the Levodopa Equivalency Dose (LED). Levodopa equivalency dose is a standardized measurement of the daily amount of anti Parkinsonian medication an individual is receiving. A standard measurement of medication is useful due to variability in dopaminergic calcul ated according to previously published guidelines (Tomlinson et al., 2010).
22 Neuropsychology Module of INFORM Neurocognitive data of PD participants were drawn from the Neuropsychology module of the INFORM database. These data were based on evaluations con ducted by the Neuropsychology Service of the Department of Clinical & Health Psychology between 2004 and 2010. Individuals were included for this study if they had undergone a complete neurocognitive examination. If multiple evaluations occurred, o nly scores from the initial neuropsychological examination were included in analyses. For the present study, five neurocognitive domains were examined : attention/working memory, episodic memory (verbal), language, executive functions, and processing speed ( Table 2 1). The table below depicts the specific measures that rationally based and in line with the approach of Sheline et al. (2006) and similar to that used by Kirsch Darrow (2010). An attempt was made to include more than one measure per domain in order to improve reliability (Jak et al., 2009). However, this was not possible for the language domain. For attention/working memory, one measure was used (Digit Span), t hough both forward and backward span scores were separately extracted. For episodic verbal memory, delayed recall scores were used (instead of learning) due to sensitivity of delayed recall to mesial temporal lobe memory dysfunction (Milner, 1954; Hermann et al., 1992). Normed scores for each neurocognitive measure were obtained from test specific manuals or from the Heaton norms book (Heaton, 2004). All normed scores were converted to z scores in order for all measures to be on a common metric.
23 T able 2 1. Cognitive Domains Domain Measures & Brief Description Working Memory Digit Span : measure of auditory attention and working memory from the WAIS III, with both forward and backwards components, dv = forward span, and backwards span scores Epi sodic Memory Hopkins Verbal L earning T est (HVLT): 12 item word list learning task with 20 minute delay; dv = # items recalled after 20' delay Logical Memory Stories II (WMS III): story memory measure dv = # units recalled after 30 min delay Language Boston Naming Test (BNT): confrontation naming of line drawings of objects/animals; dv = total correctly named Executive Function TrailMaking Test, Part B (TMT B): speeded alternating search of letters and numbers (set shifting), dv = total time St roop Color Word (Golden version): measure of cognitive inhibition of over learned automatic reading response; dv = number items within 45 sec Controlled Oral Word Association Test (COWA): a letter fluency task involving production of words beginning w ith target letter during 60 second trials; dv = number words produced Processing Speed Stroop Single Word Reading (Golden version): timed reading of single words denoting red, green, or blue; dv=number words read in 45" TrailMaking Test, Part A (TMT A ): speeded search of letters displayed on a page; dv = total time Note: individual test references in Lezak (2004). WMS III = Weschler Memory Scale Third edition; dv = dependent variable. For each domain, c omposite scores were calculated by averaging the normed Z scores of all tests in each particular domain. Advantages to using composite scores include better reliability with multiple measures per construct of interest and fewer overall analyses, lowering family wise error rate Medical Chart Revi ew A review of outpatient medical records (UF Shands Clinics) was conducted to record health comorbidities for each patient. Medical charts were reviewed if they occurred the date of or up to a year before the date of the neuropsychological examination The medical comorbidity listing used in the present study was compiled
24 from previous research ( i.e., Charlson Comorbidity Index [ Charlson et al., 1994 ]; Framingham Heart Study [Wilson et al., 1998 ]) and reviewed by physicians from neurology and internal medicine Table 2 2 shows a final list of the medical comorbidities that were coded in the present study. Comorbidities were coded dichotomously as present or not present. A comorbidity was considered if the following information was in the medical record: 1) a note by the physician referring to its presence or 2) the participant was prescribed medication pertaining to a comorbidity. Thus, individuals whose medication status indicated that they had been taking statins were coded as having c holesterolemia. Blood pressure was also recorded from participants medical records Consistent wi th previous studies of cognitive impairment and hypertension, participants having a coded as hypertensive (Farmer et al., 1990; Tzourio et al., 1999). Additional indices of pressure. Pulse pressure has been conceptualized as a surrogate measure of arterial stiffness and was calculated by subtracting diastolic blood pressure from systolic blood pressure (Aronow et al., 2011) The benefit of examining the effect of pulse pressure is that it enables one to use a continuous variable that has shown to be a better predictor of cardiovascular risk, when compared to systolic or diastolic pressure (Blach er et al., 2000; Franklin et al., 1999).
25 Table 2 2 List of Comorbidities Hypertension Hypotension Hypercholesterolemia High Tryglycerides Anemia Diabetes Thyroid Disease Heart Attack Angina Atrial Fibrillation Arrhythmia Heart Murmur Congestive Heart F ailure Peripheral Vascular disease Pacemaker Surgery Cardiac Stent Surgery Bypass Surgery Valve Surgery HIV/AIDS Stroke Transient Ischemic Event Hemiplegia Osteoarthritis Rheumatoid Arthritis Vasculitis Lupus Asthma Chronic Obstructive Pulmonary Disease (COPD) Pulmonary Embolism Emphysema Gastroesophageal Reflux Disease (GERD) Peptic Ulcer Disease Gallbladder Disease Hepatitis Liver Cirrhosis Kidney Stones Urinary Incontinence Prostate/Benign Prostate Hyperplasi a (BPH) Seizures Head Injury Brain Tumor Peripheral Neuropathy Migraines Glaucoma Cataracts Macular Degeneration Tinnitus Sleep Apnea Restless Leg Syndrome REM Behavior Sleep Disorder (RBSD) Colon Polyps Other List of comorbidities coded for. Exclus ion Criteria & Final Sample Specific exclusion criteria included: history of severe psychiatric disease (i.e., schizophrenia, psychosis), previous brain surgeries for treatment of Parkinson disease (i.e., deep brain stimulation, pallidotomy, fetal cell imp lant), and possible dementia based on impaired scores (<5th %Ile) on the Dementia Rating Scale II (Jurica et al., 2001). This resulted in a final sample size of 341 participants with idiopathic Parkinson disease for the current study. Statistical Analyses Aim 1 : P revalence The prevalence of each comorbidity occurring in at least 10% of the sample was examined for differences in common demographics and PD severity. To examine age related prevalence of comorbidities, individuals were categorized into the fol lowing
26 groups based on age: a function of these age groups. Kruskal Wallis Chi square tests were performed in order to accommodate the categorical nature of both prevalence and age group. A follow up series of logistic regression analyses were conducted among those comorbidities which were shown to differ among age groups, based on the previous set of analyses. The prevalence of the comor bidity served as the outcome/dependent variable. Logistic regression analyses were used due to the dichotomous coding (present vs. not present) of the prevalence of comorbidities. Two blocks of predictors (demographics and PD severity) were set as the inde pendent variable for each regression analysis. The demographic block consisted of age, gender and the highest year of education completed. In order to control for differences in demographic variables, the PD severity block was entered as the second step in our logistic regression analyses. By entering PD severity variables into the second block, this allowed us to examine the unique effect of PD severity on the prevalence of comorbidities, independent of demographic variables. The following variables were e ntered into the PD severity block: duration of symptoms, LED and UPDRS motor score. Aim 2: C ognition A series of hierarchical regression analyses were conducted in order to test whether comorbidities have a unique relationship with specific cognitive doma ins. Five separate regression analyses were conducted for each individual cognitive domain (executive functioning, working memory, processing speed, verbal memory and language). Cognitive composite domain scores were the dependent variable in each
27 regressi on analysis. Previous studies have shown PD severity to be a main predictor of cognitive functioning, thus variables related to PD severity (UPDRS, years with symptoms and LED) were controlled in the first step in order to assess the unique effect of comor bidities. The second step consist ed of comorbidities. For purposes of simplification and to reduce multicollinearity, only those comorbidities present in at least 10% of the sample were included. Follow up analysis for the second aim examine d the impact o f pulse pressure on cognitive domains. Again, a series of hierarchical regressions were conducted to examine the unique effect of vascular functioning on cognition. Variables related to PD severity made up the first step, while pulse pressure and a dichoto mous coding of hypertensive medication status (prescribed/not prescribed hypertensive medications) were included into the second step. Finally a residualized UPDRS by pulse pressure interaction term was calculated and inserted into a final third step. Resi dualizing the interaction term from the main effects helps to reduce multicollinearity and allows one to look at the role that this interaction has on cognition, independent of the main effects of pulse pressure and PD severity. This UPDRS by pulse pressur e allowed examination of vascular functioning as a function of PD severity. All analyses were d least significant difference tests (Saville, 1990; Rothman, 1990)
28 CHAPTER 3 RESULTS Sample characteristics are shown in Table 3 1. In general, the PD participants were in their mid s of age (range from 30 to 90), well educated with about 3 years of college, and a majority were Caucasian (94%). About two thirds of the sample was male, which is typical among the PD population (Van Den Eeden et al., 2003; Haaxma et al., 2007). Duration of PD symptoms ranged from 1 to 33 years, with a mean of 10 years. Severity of motor symptoms was, on average, in the moderate range characterized as having tremor as the pre dominant symptoms (82.7%) with the Approximately 40% were taking antidepressant medications and approximately 25% were taking anti anxiety medications. All participants scored above the 5 th %ile on the Dementia Rating S cale 2, a cognitive screening measure, and as a group had a mean score of 137.3. Table 3 1 Sample Characteristics N=341 PD Patients No DBS 234 Males (69%); 107 Females (31%) Mean Standard Deviation Range Percent Caucasian 94 Age (years) 64. 7 10 30 90 Years of Education 14.9 3 7 20 Years with Symptoms 9.7 6 1 33 UPDRS Motor Score, on medication 26.4 10 7 54 Percent Tremor Predominant 82.7 Levadopa Equivalency Dose 770 540 0 2950 Percent on Antidepressant Medication 38.4 Percent on Anti anxiety Medication 25.5 Percent on Antipsychotic Medication 8.0 Dementia Rating Scale II Total Score 137.3 5 119 144
29 Specific Aim 1: Prevalence of Comorbidities The occurrence of specific comorbidities ranged from a high of 41.6% (cholesterol ) to a low of 0% (HIV/AIDS, vasculitis, lupus, gallbladder disease, hepatitis, liver cirrhosis 1 depicts the occurrence of all comorbidities that were present in at least 10% of the sample. An entire listing of all comorbidit ies and prevalence rates is presented in Appendix A. Comorbidities of similar systems were combined if individual disorders/disease occurred in fewer than 10% of the sample. Three systems (cardiovascular, respiratory and neurologic) of comorbidities wer e computed. An individual was considered as having a cardiovascular comorbidity if any of the following were present: heart attack, angina, atrial fibrilation, arrhythmia, heart murmur, congestive heart failure, peripheral vascular disease, or any cardiac surgery (stent, pacemaker, valve or bypass surgery). The respiratory system comorbidity consisted of COPD, pulmonary embolism, emphysema and asthma. Lastly the following comorbidities were considered in the neurologic system comorbidity: stroke, TIA, seizu res, head injury, brain tumor, and peripheral neuropathy. The three most common comorbidities were high cholesterol, hypertension and orthostatic hypotension, occurring in 41.6%, 38.1% and 30% of our sample respectively. Relative to the rates observed in normal elderly (i.e., approximately 60%), the prevalence of hypertension in the PD sample is substantially less (i.e., 38%) (Wolf Maier et al., 2003, Aronow et al., 2011). Hypotension, however, is approximately twice as common in this PD sample compared t o the normal elderly (Low, 2008; Shin et al., 2004). Lastly, cardiac disease occurred in 26.7% of our sample, and diabetes occurred in 10.3%.
30 Figure 3 1. Prevalence of Comorbidities in PD Sample (N=341). Footnote: GERD = Gastroesophageal reflux disea se ; RBSD = REM behavior sleep disorder; Shown is the percentage of Parkinson patients with various comorbidities based on medical chart review; only comorbidities occurring in at least 10% of the sample are shown. Both hypertension and hypotension oc curred relatively frequently in the PD which is a persistent fall in blood pressure when individuals shift position from sitting to standing. The underlying pathophysi ology in PD usually relates to autonomic dysfunction (i.e., sympathetic cardiac denervation, baroreflex failure) Some individuals with orthostatic hypotension also have co For this reason, the current data were further examined for the co occurrence of hypotension and hypertension, and for the occurrence of each blood pressure condition separately. Figure 3 2 shows that of the 130 individuals with hypertension, more than a third were reported to have co occurring hypote nsion, while roughly half of the hypotensive PDs had concomitant hypertension. Fourteen percent of the total sample
31 of 341 were reported to have both hypertension and orthostatic hypotension. Additionally, hypertension occurred independently in 24% of th e sample, and isolated hypotension occurred in 16% of the sample. Figure 3 2 Occurrence of Hypotension and Hypertension in Parkinson Sample Comorbidity Prevalence and Age Because health problems generally increase with age, the prevalence of comorbidi ties in the PD sample was examined as a function of age. To do so, the PD sample was divided into three age groups (Artinian, 1993): a) old (N = 101, X = 52.8 yrs, SD= 5.6); b) 60 to 69 years old (N =135, X = 64.8 yrs, SD = 2.9); and c) ) (N = 105, X = 75.3 yrs, SD = 4.9). Figure 3 3 depicts the prevalence of comorbidities as a function of these age groupings. As shown, comorbidity prevalence significantly increased with age (Kruskal Wallis analyses) for six of the 12 health conditions: cholesterol, hypertension, hypotension, cardiac, arthritis, and prostrate. In fact, the prevalence of hypertension, choleste rol and cardiac disease were about 20%
32 higher in the oldest age group compared to the other groups. While GERD (p = .065) and respiratory conditions (p = .055) tended to increase with age, the remaining conditions were relatively age stable (i.e., neurolo gic, cancer, RBSD, diabetes). Figure 3 3 Comorbidity Prevalence According to Age Grouping. ** p <0.01 *p<.05, based on Kruskal Wallis nonparametric statistics. Figure depicts prevalence at 3 age groupings for comorbidities that occur at lea st 10% in the sample. Means, standard deviations for each age by comorbidity condition are as follows: Cholesterol by age group (young to old): 27.7%, 37.0%, 61.0%, ( 2 (2) = 25 .3, p <0.001 ); Hypertension : 24.8%, 35.6%, 54.3% ( 2 (2) = 19.6, p <0.001, Hypotension : 23.8%, 28.1%, 39.0% ( 2 (2) = 6.1 p = 0.047 ), Neurologic : 28.7%, 31.1%, 27.6% ( 2 (2) = 0.37, p = ns), Cardiac : 6.9%, 25.2%, 47.6%, ( 2 (2) = 66.2, p < 0.001), Arthritis : 13.9%, 31.1%, 29.5%, ( 2 (2) = 10.3 p = 0.006); GERD 11. 9%, 23.0%, 22.9%, ( 2 (2) = 5.5 p = 0.0 65 ) ; Cancers : 18.8%, 20.0%, 17.1%, ( 2 (2) = 0.31, p = ns); Prostate : 5.0%, 10.4%, 20.0%, ( 2 (2) = 11.6, p < 0.003); RBSD 9.9%, 13.3% 9,5%, ( 2 (2) = 1.08, p = ns); Respiratory : 7.9%, 7.4%, 16.2%, ( 2 (2) = 5.78 p = 0.055); Diabetes : 7.9%, 11.1%, 11.4%, ( 2 (2) = 0.86 p = ns ) ** ** ** ** **
33 Influence of Demographic and Disease Severity on Comorbidity Prevalence To determine whether each comorbidi ty was differentially associated with demographic or disease related factors, we conducted a series of six separate logistic regression analyses. Demographic factors (age, education, sex) were entered into Block 1 and PD severity factors (duration, UPDRS, LED) were entered into Block 2. These factors were regressed, in separate analyses, on the occurrence of each comorbidity: cholesterol, hypertension, low blood pressure, cardiovascular disease, arthritis and prostate disease. For hypertension, the overa ll model was significant ( 2 (6) = 38.353, p<.001; Table 3 2). As shown below, the demographic step was significant ( 2 ( 3 )= 33.216, p<.001), whereas the addition of disease related severity factors was not ( 2 ( 3 ) = (5.136, p=.162). Examination of individual predictors revealed that age was significantly related to hypertension prevalence (odds ratio [OR] = 1.063), with a trend for gender to be associated (OR = .592). Thus, hypertension was associated with older age with a slight preference for male gender. Table 3 2 Predictors of Hypertension HYPERTENSION Chi Square Odd Ratio Sig Final Model 38.353 <0.001 Demographic Block 33.216 <0.001 Age 1.063 <0.001 Gender .592 .058 Education .935 .111 PD Severity Block 5.1 36 .162 Duration .984 .463 LED 1.000 .139 Motor UPDRS 1.014 .262
34 A slightly different pattern occurred for hypotension ( Table 3 3). Again, the overall model was significantly related to hypotension prevalence ( 2 ( 6 ) =35.215, p<.001), as were both demographic ( 2 ( 3 ) =7.280, p =.047) and PD severity factors ( 2 ( 3 ) =27.280, p<.001). However, in contrast to hypertension, higher UPDRS scores (OR=1.052) and shorter duration of symptoms (OR=.899) were associated with hypotension. This suggests th at disease related factors are associated with hypotension in Parkinson dis e ase. None of the demographic variables (i.e., age, gender, education) independently accounted for variance in occurrence of hypotension. Table 3 3. Predictors of Hypotension H YPOTENSION Chi Square Odd Ratio Sig Final Model 35.215 <0.001 Demographic Block 7.280 .047 Age 1.019 .169 Gender 1.414 .230 Education 1.032 .474 PD Severity Block 27.280 <0.001 Duration .089 <.001 LED 1.000 .197 Motor UPDRS 1.052 <.001 For cardiac disease the final model significantly contributed to the presence of cardiac disease ( 2 ( 6 ) =82.448, p<.001), along with demographic and PD severity factors (Table 3 4). Significant individual predicto rs were age, gender, education, and UPDRS motor severity. Thus, cardiac disease was more common among older (OR=1.130) males (OR=.402) with less education (OR=.906) and more severe motor symptoms (OR=1.040).
35 Table 3 4 Predictors of Cardiac Disease CAR DIAC Chi Square Odd Ratio Sig Final Model 82.448 <0.001 Demographic Block 74.198 <.001 Age 1.130 <.001 Gender .402 .010 Education .906 .045 PD Severity Block 8.250 .041 Duration .966 .213 LED 1.000 .33 8 Motor UPDRS 1.040 .008 As shown in the next two tables, similar patterns were observed for cholesterolemia and arthritis For both, the final model was significant, as was the ibute significantly to the occur rence of either cholesterol or arthritis. For both, increased age was associated with increased occurrence of high cholesterol (OR = 1.061) and increased occurred of arthritis (OR = 1.044) Table 3 5 Predictors of Choleste rolemia CHOLESTEROL Chi Square Odd Ratio Sig Final Model 29.979 <0.001 Demographic Block 26.913 <.001 Age 1.061 <.001 Gender .750 .279 Education .973 .509 PD Severity Block 3.067 .381 Duration .975 .244 LE D 1.000 .453 Motor UPDRS .999 .967
36 Table 3 6 Predictors of Arthritis ARTHRITIS Chi Square Odd Ratio Sig Final Model 16.111 .013 Demographic Block 14.037 .003 Age 1.043 .003 Gender 1.641 .085 Education .968 .4 72 PD Severity Block 2.074 .557 Duration .969 .220 LED 1.000 .306 Motor UPDRS 1.008 .585 Finally, variance in prostate disease prevalence was significantly linked to our final model. Although demographics were significantly relate d to prostate disease prevalence, the PD severity block as a whole did not significantly contribute to our model. Examination of individual predictors revealed that prostate disease was more common among older patients (OR=1.063) with higher UPDRS scores ( OR=1.044). Of note, female gender was not entered into the analyses due to the sex specific nature of this disorder. Table 3 7 Predictors of Prostate Disease PROSTATE Chi Square Odd Ratio Sig Final Model 51.179 <0.001 Demographic Block 46.102 <.001 Age 1.063 .007 Gender NA NA Education 1.061 .340 PD Severity Block 5.077 .166 Duration .986 .692 LED 1.000 .685 Motor UPDRS 1.044 .028
37 Summary of Aim 1 : To summarize the results from Aim 1, we found that cholesterolemia was the most frequently reported comorbidity, followed by hypertension and hypotension. When comparing our sample to previous reports of prevalence among normal elderly cohorts, we observed that hypertension was approximately 20% less common. In contrast, hypotension occurred about twice as frequent in PD relative to that observed in normal elderly (Low, 2008; Shin et al., 2004). Further examination showed that occurrence of hypotension was dependent on PD related severity, whereas hy pertension, cholesterolemia, arthritis and prostate disease were more associated with demographic differences, with age being the primary contributor. Finally, the occurrence of cardiovascular disease was unique in that it was significantly related to both PD severity (i.e., UPDRS) and demographic variables (age, gender, education). Specific Aim 2: Comorbidities and Cognition The purpose of Aim 2 was to determine whether medical comorbities influenced cognitive status, when controlling for the effects of di sease related variables. Only comorbidities present in at least 10% of our sample were considered for this aim. The comorbidities that met this criteria are depicted in Figure 3 1 and included the following: hypertension, hypotension, cholesterolemia, a cute cardiac disease (myocardial infarction or cardiac surgery), diabetes, arthritis, rapid eye movement behavior sleep disorder (RBSD), cancer/tumor, gastroesophageal reflex disease (GERD), and prostate disease (i.e. benign prostate hyperplasia; BPH). Ta ble 3 8 shows the composite scores, represented as z scores, for each of the 5 cognitive domains. The distribution of each domain specific cognitive composite score was examined for normality. Skewness and kurtosis values for all domain composite scores w ere below one. Skewness and kurtosis values < 1 are indicative of good
38 distribution (Jones, 1969). For all regression analyses, collinearity diagnostics indicated that multicollinearity, as measured by tolerance (>.2) and variation inflation factor (<5), w Table 3 8 Cognitive Composite Scores For Each Domain Mean (SD) Range Skewness Kurtosis Executive Functioning .564 (.87) 2.90 to 1.93 .104 .091 Verbal Memory .400 (1.08) 3.78 to 1 .83 378 .232 Processing Speed .728 (.89) 3.20 to 1.90 378 .238 Working Memory .157 (.79) 1.90 to 2.75 .328 .231 Language .206 (1.12) 3.00 to 2.70 .054 .505 Note: all scores are represented as z scores, with a mean of 0 and standard devi ation (SD) of + 1 Results of the series of hierarchical regression analyses for each cognitive domain are shown in the tables below. For each regression, disease severity indicators (UPDRS motor, LED, duration) were entered into block 1, whereas medical comorbidities were entered into block 2. In brief, results of these analyses showed that the overall regression model was significant for 3 of the 5 cognitive domains For executive functioning the final model of PD severity and comorbidities togethe r accounted for a significant amount of variance (F(14,295)=3.38,p<.001; Table 3 9). Disease severity was a significant predictor of executive functioning (F(3,306)=10.247, p<.001) and accounted for 9.1% of the variance. As shown, this effect was due to U PDRS motor scores ( .246). The addition of comorbidities as a whole did not significantly add to the model
39 .127) and hypot comorbidities that significantly predicted executive functioning, though they did so in opposite directions That is, presence of hypertension was associated with reduced executive function scores, whereas the presence of hyp otension was associated with higher executive function scores. Table 3 9 Comorbidities and Executive Function EXECUTIVE FUNCTION Beta Sig Final Model 3.38 0.138 <0.001 PD Severity Block 10.247 0.091 <0.001 Duration 0.15 6 <0.001 LED 0.092 .120 Motor UPDRS 0.246 <0.001 Comorbidity Block 1.460 0.047 .146 Hypertension 0.127 .041 Hypotension 0.134 .019 For delayed verbal memory the overall model accounted for a significant amount of variance (F (14,320)=3.166,p<.001; Table 3 specifically symptom duration .227), were significant indicators of worse verbal memory performance (F(3,331)=9.383, p<.001), explaining 7.8% of variance in verbal memory scores. The addition of comorbidities as a whole 157), however hypertension was significantly .139). Table 3 10 Comorbidities and Verbal Memory DELAYED VERBAL MEMORY Beta Sig Final Model 3.39 0.122 <0.001 PD Severity Block 9.38 3 0.078 <0.001 Duration 0.188 <0.001 LED 0.050 .384 Motor UPDRS 0.227 <0.001 Comorbidity Block 1.433 0.043 .157 Hypertension 0.139 .021
40 For processing speed the final model reached significance (F (14,305) =3.977, p<.001; Table 3 11). Years with .315) were significant variables from the PD severity step (F(3,316)=10.939, p<.001) and indicative of lower processing speeding scores. About 13% of outcome variance was related to indicators of PD severity. Comorbidities as a whole were not significantly individual comorbidities were significantly linked to processing speed performance. Table 3 11 Comorbidities and Processi ng Speed PROCESSING SPEED Beta Sig Final Model 3.977 0.154 <0.001 PD Severity Block 15.647 0.129 <0.001 Duration 0.139 .019 LED 0.086 .619 Motor UPDRS 0.315 <0.001 Comorbidity Block 1.433 0.043 .157 Fo r attention/working memory the final model failed to reach significance (F(14,320)=1.179, p=.290; Table 3 12). Neither the initial PD severity step p=.435) were significant. F urther examination of individual predictors revealed that .132). Table 3 12 Comorbidities and Working Memory WORKING MEMORY Beta Sig Final Model 1.179 0.049 .290 PD Severity Block 1.786 0.016 .150 Duration 0.005 .930 LED 0.015 .800 Motor UPDRS 0.132 .023 Comorbidity Block 1.013 0.033 .435 For the language domain the final model failed to reach significance (F(14,324)=.657, p=.816; Table 3 13). Neither Parkinson disease severity
41 (F(3,335)=.777, p=.508) nor comorbidities added significantly to the model (F(11,324)=.627, p=.806). None of the individual predictors were significant. Table 3 13 Comorbidities and Language L ANGUAGE Beta Sig Final Model 0.657 0.028 .816 PD Severity Block .777 0.007 .508 Duration 0.051 .404 LED 0.007 .908 Motor UPDRS 0.063 .279 Comorbidity Block 1.433 0.043 .806 In summary our final re gression models were significant in accounting for executive functioning, verbal memory and processing speed. The PD severity block contributed to executive functioning, delayed verbal memory and processing speed. Specifically, higher motor UPDRS scores an d longer symptom duration were indicative of worse performance across all three domains. While comorbidities as a whole did not significantly contribute to our model, individual comorbidities did show a relationship to certain cognitive domain composite sc ores. The presence of hypertension was significantly related to lower executive functioning scores and delayed verbal memory performance, and hypotension was related to higher executive functioning composite scores. Hypertension and Hypotension: Influence s on Cognition In light of the unique relationship between hypertension and hypotension in our PD sample, additional analyses were conducted to further examine the influence of hypertension and hypotension on cognitive outcomes. In these analyses, individ uals were grouped according the presence of hypertension and/or hypotension, resulting in 4 groups: a) those with hypertension alone, b) those with hypotension alone, c) those with co occurring hypertension and hypotension, and d) those with neither hyper or
42 hypotension (i.e., normotensive). Two univariate analyses of covariance (ANCOVA) subject variable. The following covariates were used: years with symptoms, LED, and UPDRS motor score. The executive compos ite score was the dependent variable in one ANCOVA, whereas the delayed memory composite was the depende nt variable in the other ANCOVA. Figure 3 4 Executive and Verbal Memory Scores Among Blood Pressure Groups p<.05 from univariate analyses of cov ariance. Figure depicts mean composite z scores for executive function and delayed verbal memory across each of the blood pressure groups. The hypertensive group performed significantly worse than all groups on measures of executive functioning and verbal Figure 3 4 shows the scores of the four blood pressure groups on the executive Group main effect for both executive fu nctioning (F(6,303)=3.568, p=.015) and delayed
43 verbal memory (F(6,328)=2.680, p=.047). For executive function post hoc comparisons (t tests) indicated that PD patients with hypertension alone (X= 0.781 (SD=.769) performed significantly worse than all oth er groups including hypotension alone (X= 0.321 (SD=.769), t= 2.87, p=.004), combined hypertension hypotension (X = 0.437 (SD= .739, t= 2.50, p=.013), and normotensive (X = 0.495, SD=.898, t= 2.21, p=.028). There were no significant differences amo ng the hypotensive, normotensive, and verbal memory post hoc comparisons indicated that PD patients with hypertension (X = 0.633, SD=1.068) performed significantly worse relative to the hypotensive (X = 0.194, SD=1. 098, t= 2.45, p=.015) and the normotensive group (X = 0.301, SD=1.013, t= 2.23, p=.026). The hypertensive group did not differ from the co occurring hypotensive hypertensive group (X= 0.580, SD=1.228, t= 0.63, p=.530). There were no other significan Pulse Pressure To further assess the influence of vascular/blood pressure factors on cognition, a series of hierarchical regression analyses were conducted using pulse pressure (systolic pressure diastolic pressure) as one of the predictor variables. As described earlier, pulse pressure is viewed as a more sensitive index of vascular stress (Aronow et al., 2011). Cognitive domain scores were entered as outcomes for each analysis and were regressed upon PD severity (step 1), pulse pressure and hypertensive medication (step 2) and the pulse pressure X UPDRS interaction (step 3). As an aid for interpretation, the relationship between pulse pressure and cognitive domain scores was plotted as a function of UPDRS scores for those c ognitive domains in which the interaction term was significant. For purposes of demonstration only, the UPDRS motor scores were categorized into quartiles (1 st quartile= 1 to 25%ile; 2 nd quartile= 26 to
44 50%ile; 3 rd quartile= 51 to 75%ile; 4 th quartile= 76 100%ile) when creating scatter plots. However, all regression analyses use raw UPDRS scores (i.e., not quartile groupings). Executive Function: As shown in Table 3 14, the overall model (PD severity, Pulse Pressure, Interaction) was significantly rel ated to executive functioning (F(6,299) = 6.297, p < .001). The initial block (PD severity ) was also significant (F(3,302)=10.125, p<.001) and accounted for 9.1% of executive function variance. .209) and duration of symptoms .195) were significant variables linked to executive functioning scores. The addition of pulse pressure and hypertensive medication did not significantly add to the model PDRS interaction was .113). Visual inspection of executive scores plotted by pulse pressure and UPDRS quartiles revealed that with increasing UPDRS scores (i.e., worsening of PD symptoms), increases in pulse pressure were associated with worse executive functioning scores (Figure 3 5). Table 3 14. Pulse Pressure and Executive Function Executive Functioning Beta Sig Final Model 6.297 0.112 <.001 PD Severity Block 10.125 0.091 <.001 Duration 0.195 <.001 LED 0.094 .175 Motor UPDRS 0.218 <.001 Pulse Pressure Block 1.423 0.009 .243 Pulse Pressure 0.041 .461 Hypertensive Medication 0.083 .134 Interaction Block 4.130 .012 .043 PP X UPDRS 0.113 .043
45 Figure 3 5. Pulse Pressure X UPDRS Interaction Executive Function Figure depicts linear relationship between pulse pressure and executive functioning for each UPDRS quartile. Delayed Verbal Memory: As shown in Table 3 15, res ults of the regression analysis revealed a significant final model for delayed verbal memory. In terms of .2 13) and duration of .127) were significant contributors. The second step in the model (pulse pressure, hypertensive medication) failed to significantly account for a unique amount of verbal memory variance. However, individuals on hypertensiv e medication were .112). The pulse .138). Visual inspection of this relationship with a scatter plot showed a si milar pattern to that observed with executive functioning (Figure 3 6). Namely, individuals with higher pulse
46 pressure levels were more likely to have lower memory scores as their UPDRS scores worsened. Table 3 15 Pulse Pressure and Verbal Memory Verbal Memory Beta Sig Final Model 7.603 .110 <.001 PD Severity Block 9.463 .080 <.001 Duration 0.127 .027 LED 0.072 .213 Motor UPDRS 0.213 <.001 Pulse Pressure Block 2.163 .012 .117 Pulse Pressure 0.030 .575 Hyp ertensive Medication 0.112 .038 Interaction Block 6.676 .018 .010 PP X UPDRS 0.138 .010 Figure 3 6 Pulse Pressure X UPDRS Interaction Verbal Memory Figure depicts linear relationship between pulse pressure and verbal memory fo r each UPDRS quartile. Processing Speed: As shown in Table 3 16, results of the final regression model were significant for processing speed. The first step of PD severity was significant (F(3,312)= 16.004, p<.001) with a R of .133, while the second ste p (F(2,310)=.555,
47 .157) .305) were significant predictors in the first step. Table 3 1 6 Pulse Pressure and Processing Speed Processing Speed Beta Sig F inal Model 6.403 .152 <.001 PD Severity Block 16.004 .133 <.001 Duration 0.157 .006 LED 0.070 .218 Motor UPDRS 0.305 <.001 Pulse Pressure Block .555 .003 .117 Pulse Pressure 0.037 .642 Hypertensive Medicatio n 0.051 .340 Interaction Block 5.785 .016 .017 PP X UPDRS 0.128 .017 Figure 3 7 Pulse Pressure X UPDRS Interaction Processing Speed Figure depicts linear relationship between pulse pressure and processing speed for each UPDRS quartile. The Pulse Pressure X UPDRS term was also significant. Inspection of the scatter plot revealed that processing speed worsened for individuals with more severe disease severity and higher pulse pressures (Figure 3 7).
48 Working Memory : The final regression model did not significantly predict working memory scores, as shown in Table 3 17. Although overall PD severity failed to reach significance in Block 1, higher UPDRS scores were significantly related to worse working memory perfo .116). The addition of the pulse pressure/medication step and the interaction step failed to significantly add to the model. Table 3 17 Pulse Pressure and Working Memory WORKING MEMORY Beta Sig Final Model 1.317 .024 .249 PD Severity Block 1.771 .016 .152 Duration 0.007 .991 LED 0.005 .929 Motor UPDRS 0.116 .039 Pulse Pressure Block 1.265 .008 .284 Pulse Pressure 0.028 .617 Hypertensive Medication 0.087 .123 Interaction Block 0.065 <.001 .794 PP X UPDRS 0.012 .794 Language : Language scores were not significantly related to the final model, as shown in Table 3 18. Neither the first step of PD severity, pulse pressure and hypertensive medication, or the final ste p of pulse pressure X UPDRS significantly explained variance in language scores. Examination of individual predictors revealed that no variables were significantly related to language performance. In summary, three cognitive domains (executive function, p rocessing speed, delayed verbal memory) exhibited a unique relationship with pulse pressure (index of vascular flexibility) and disease severity (UPDRS). Specifically, as disease severity worsened, pulse pressure increases were associated with worse score s on indices of executive function, delayed verbal memory, and processing speed.
49 Table 3 18 Pulse Pressure and Language LANGUAGE Beta Sig Final Model 1.172 .017 .459 PD Severity Block 0.797 .007 .496 Duration 0.050 .399 LED 0.008 .890 Motor UPDRS 0.069 .218 Pulse Pressure Block 1.239 .007 .291 Pulse Pressure 0.090 .106 Hypertensive Medication 0.027 .631 Interaction Block 0.838 .003 .361 PP X UPDRS 0.051 .361
50 CHAPTER 4 DI SCUSSION The results of this study with a cohort of non demented Parkinson patients suggest three major findings. First, vascular related comorbidities, such as hypertension and hypotension were common in our PD cohort, but occurred in different proporti ons compared to normal age peers. Secondly, while PD measures of disease severity (particularly the motor UPDRS scores) were the main contributor to cognitive status across four of five cognitive domains (i.e., executive functioning, verbal memory, process ing speed, working memory), the presence of hypertension independently contributed to worse performance in executive functioning and verbal memory. Furthermore this negative influence of hypertension on executive function seemed modified by overlapping sym ptoms of hypotension. Third, worsening vascular functioning (as indexed by pulse pressure) bec a me more detrimental to aspects of cognitive status as PD severity progressively worsen ed Prevalence: Aim 1 This study found that hypertension occurred in 38.1 % of the PD sample. Previous studies of normal aging have reported estimates of hypertension in the range of 53% to 70% for individuals above 65 years old (Wolf Maier et al., 2003; Aronow et al., 2011). Thus, in our sample of PD patients, hypertension w as about 20% less frequent than estimates in normal elderly. This finding is consistent with our hypothesis that hypertension would be less prevalent than that in normal elderly. Possible explanations for a lower prevalence of hypertension in PD includ e: a) side effects of dopaminergic medications ; and b) PD related autonomic disruption ( McDowell et al., 1970; Nanhoe Mahabier et al., 2009). Scigliano and colleagues (2009)
51 hypertension than those not on levodopa. One explanation for this difference relates to frequent stimulation of peripheral dopamine D2 receptors by levodopa, leading to an inhibition of sympathetic norepinephrine release. This reduction in norepinephrine results in alterations in rennin level and sodium retention, ultimately leading to decreases in vasoconstriction and overall cardiac output. A second explanation for decreased prevalence of hypertension in PD relates to autonomic dysfunction. Autonomic d ysfunction commonly occurs in PD secondary to deposition of Lewy bodies throughout the neuraxis. Autonomic dysfunction also contributes to a reduction norepinephrine (Goldstein, 2003). This reduction in norepinephrine may be explained by degeneration of the locus coeruleus, a primary site for norepinephrine synthesis. The reported prevalence of hypertension in the current Parkinson cohort is within the range of that cited elsewhere in the literature on PD. Previous studies of cardiovascular disease hav e shown that hypertension occurs in 12 to 38% of PD samples (Haugarvoll et al., 2005; Levy et al., 2002). In non demented subsamples the occurrence of hypertension has been reported to be slightly higher, between 18 to 40%. On the other end of the spectru m, orthostatic hypotension occurred in 30% of our sample. This figure is comparable to estimates (20 50% range) from previous PD studies (Oka et al., 2006; Senard et al., 1997). This rate is about twice what would be expected among the normal aging popula tion (Low, 2008; Shin et al., 2004). In PD, orthostatic hypotension has been linked to noradrenergic denervation and baroreflex failure (Goldstein, 2003). The baroreflex is has been described as an automatic mechanism that helps to maintain constant blood pressure by alerting heart rate.
52 Indeed, PD patients with hypotension typically have an abnormal baroreflex response relative to non hypotensive PD patients. Norepinephrine disruption, associated with locus coeruleus degeneration is also associated with baroreflex failure and orthostatic hypotension (Niimi et al., 1999; Goldstein et al., 2002). In terms of other key comorbidities, the occurrence of diabetes (10%), and high cholesterol (41.6%) is comparable to that reported in normal elderly (i.e., diab etes = 10%, Wilson et al., 1986; high cholesterol = 45%, Selvin & Erlinger, 2004). Similarly, in our PD sample, the occurrence of other health conditions (i.e., stroke, peripheral neuropathy, cancer, prostrate problems, respiratory problems, etc.) were a lso similar to estimates described in normal elderly by various sources (Mold et al., 2004; National Heart, Lung and Blood Institute, 2004; American Gastroenterological Association, 2001; Verhamme et al., 2002; Jemal et al., 2007). Thus, the two most comm on neurologic conditions in our PD sample, stroke at 2.3% and peripheral neuropathy at 19.1% grossly approximated the numbers reported for these comorbidities in the elderly (i.e., 4% stroke, 26% peripheral neuropathy; Mold et al., 2004; National Heart, Lu ng and Blood Institute, 2004). Some comorbidities seemed slightly lower in our PD sample (i.e., arthritis), whereas others were slightly higher (i.e., REM Behavior Sleep Disorder, hypotension). The occurrence of arthritis (25.5%) in our PD cohort is sli ghtly lower than a previous prevalence report of arthritis in Americans age 55 65, namely 40% women and 30% men (Reginster, 2002). The high proportion of males in our PD sample may be partially contributory for this difference. Finally, RBSD was reporte d to occur in 11% of our PD sample. This is similar to a previous report (15%) for PD patients (Comella et al., 1998). When comparing the prevalence of RSBD to the health elderly, a previous
53 large epidemiological study in Hong Kong found the prevalence of RBSD to be less than one percent (Chiu et al., 2000). Lastly, comorbidities differed on how they were related to demographic and disease severity data. Hypertension and hypotension were shown to be related to separate predictors. Hypertension was associ ated with increasing age, whereas hypotension was associated with worsening disease (i.e., higher UPDRS scores, longer symptom duration). This reflects the unique role of PD related degeneration on the occurrence of hypotension, while hypertensive is more reflective of separate processes (ie. structural changes in blood vessels ; Aronow et al., 2011). In addition to hypertension, high cholesterol, arthritis, and prostate disease were primarily related to age. Cardiac disease was significantly related to dem ographic variables like age and education as well as to PD severity. The role of PD severity in cardiac disease is not well understood but might potentially relate to a sedentary lifestyle that is common in PD (Nanhoe Mahabier et al., 2009). Comorbidities and Cognition: Aim 2 We found partial support for our hypothesis that vascular comorbidities would contribute to worse performance on cognitive domains that are viewed as being more on cognitive functioning was the presence of hypertension/ hypotension. Other cardiovascular comorbidities (i.e., cardiac, high cholesterol, diabetes) exerted no significant effects after controlling for PD severity. Specifically, the presence of hypert ension was an indicator of worse executive functioning, as indexed by a composite score involving set shifting, cognitive inhibition, and letter fluency. While this e .127), it is independent of differences in PD severity. This relationship between hypertension and
54 executive function is in line with previous studies among the normal elderly (Saxby et al., 2003; Kuo et al., 2004; Kilander et al., 199 8; Vicario et al., 2005). In brief, some studies have found high blood pressure to be associated with worse executive performance among both treated and untreated hypertensive adults. This has not been a universal finding however ( Anson & Para n, 2005; Dur on et al. 2008). Contrary to our predictions, working memory and processing speed scores were not influenced by vascular comorbidities in our cohort of nondemented PD patients. No support was provided for the second hypothesis that verbal memory performan ce would be influenced by cardiac conditions or diabetes. Instead, the occurrence of hypertension was again associated with worse verbal memory. While we did not predict this relationship, hypertension has been shown to be related to worse memory recall performance (Wallace et al., 1985). A previous study by Wu and others (2008) showed that vascular pathology and diabetes are related to memory performance via dissociable hippocampal structures (subiculum vs. dentate gyrus). Because our study did not inv olve neuroimaging and did not specifically examine glucose levels, we cannot speculate as to differential involvement of these hippocampal structures in our participants. In normal elderly, evidence for a detrimental effect of hypertension on executive function has been inconsistent. While numerous studies have found evidence to support such a relationship, others have not (Farmer et al., 1987; Desmond et al., 1993; Launer et al., 1995). A recent review paper by Birns and Karla (2009) concluded that whi le cross sectional studies have provided mixed results for the relationship between high blood pressure and cognition, the majority of longitudinal studies have shown
55 hypertension to be indicative of cognitive impairment. However, this review paper failed to elucidate in their conclusion which cognitive domains were particularly affected by hypertension. In PD, the relationship between vascular comorbidities and cognition has also been controversial and has produced mixed results. There are at least five p rior studies with Parkinson patients that have examined cardiovascular risks in relation to cognition. classification of dementia, rather than subtle variations in cognition. One study was longitudinal (Haugarvoll, et al., 2006) and involved a large cohort of newly diagnosed non demented PD patients (N=130) who were part of a Norwegian longitudinal study. Self reported presence of comorbidities was recorded at baseline. Outcome was vascular risk factors (i.e., hypertension, cardiac disease, smoking) predicted transition to dementia/non dementia when controlling for age and PD related diseas e factors in logistic regression analyses. Four other studies were cross sectional, although two exclusively involved newly diagnosed PD patients (Slawek et al., 2008; Delaker et al., 2009; Lee et al., 2010; Beyer et al., 2006). The major focus of thes the four studies categorized their PD cohort into 3 cognitive subgroups: demented (PDD), mild cognitive impairment (MCI) and unimpaired PD. Th e fourth study used the demented nondemented classification. All studies then examined the occurrence of cardiovascular risk factors among these cognitive subgroups. None of the studies
56 found differential occurrence of vascular comorbidities across any of the dementia/MCI/unimpaired sub groupings. In contrast, the current study found evidence for an influence of hypertension on indices of executive function and verbal memory. What might account for the different results between the current and p revious studies described above? Perhaps most important are differences in approach. First, we excluded individuals who were possibly demented, based on low scores on a dementia screening measure (i.e., DRS 2). Perhaps more importantly, we examined as our outcome, parametric variations across domains of cognition, rather than using a categorical outcome classification such as terms of detecting subtle differences (Hays, 1 994). Second, our definition of hypertension accounted for individuals with high blood pressure (systolic blood pressure >160mmHg). This may help to better capture individuals with untreated/undiagnosed hypertension. Third, differences in findings may be d ue to discrepancies in sample sizes. The previous studies included sample sizes no larger than 163, while our study contained an N of 341. With this large N, we showed that the variance uniquely explained by hypertension was small for both executive functi .13) and .14). Previous studies may have lacked the power to find the small relationship between hypertension and cognition, as found in the current study. Additionally, three of the previously mentioned studies utilize d the same Norwegian cohort (Beyer et al., 2006; Dalaker et al., 2009; Haugarvoll et al., 2005). Unexpected Comorbidity Cognition Relationship: Hypotension One unexpected and unique finding of the present study related to hypotension. We found that the oc currence of hypotension was associated with higher executive
57 function scores. Moreover, when co occurring with hypertension, hypotension appeared to dampen the typical negative influence of hypertension on executive function. This modulation effect was n ot present for verbal memory recall and was specific to executive function. Given this unique observation, two questions arise. Has this phenomenon has been previously described (i.e., has it been reliably observed) a nd, what might be underlying basis f or this relationship? Turning to the literature, there have been mixed findings regarding the effect of hypotension on cognition among normal elderly individuals. A number of studies have found no relationship between hypotension and cognition (Kuo et al., 2004; Rose et al., 2010). The Atherosclerosis Risk in Cognition (ARIC) study of over twelve thousand elderly individuals being followed for 12 years, failed to find an association between presence of orthostatic hypotension and animal fluency scores (Rose et al., 2010). For other cognitive domains, hypotension was associated with worse cognitive outcomes, cardiovascular factors. In PD, only a few studies have examined the rel ationship between orthostatic hypotension and cognition, and this has mainly occurred by comparing demented and non demented PD groups (Peralta et al., 2007; Allcock et al., 2006). Critically in these studies, hypotension was sometimes associated with wor se cognitive scores. While both Peralta et al. (2007) and Allcock et al. (2006) examined postural changes in blood pressure between demented and non demented PDs, only Peralta et al. (2007) found a significant difference on blood pressure changes between the two groups. Additionally, these studies found the presence of hypotension was indicative of worse performance
58 on measures of working memory and verbal fluency. By contrast, the current study found no evidence that hypotension negatively influenced cog nitive scores, and in fact, the opposite was observed. One potentially relevant methodological difference is that our study relied on medical chart review, rather than observation of postural blood pressure changes. The specific mechanism underlying t he positive relationship between hypotension and executive function is unclear, as is the basis for the moderating effects of hypotension on hypertension. For executive function, individuals with co occurring hypertension and hypotension performed on par with individuals without either comorbidity. Again, the specific mechanisms of this effect are unknown, as this phenomenon has not been well documented. However a previous study found that of individuals with hypertension and co occurring hypotension had l ower odds of developing cognitive impairment (Yap et al., 2008). A proposed explanation for this effect is that hypotension results in vasodilation and a reduction of mean arterial pressure. This helps to keep cerebral blood perfusion in the autoregulation range in individuals with hypertension (Novak & Hajjar, 2010). More research is needed to explore this possible mechanism. Comorbidities and Cognition: Pulse Pressure In addition to hypertension, the current study found that pulse pressure (an indicator of vascular health), was related to cognition as the severity of Parkinson disease motor symptoms worsened (i.e., interaction between Pulse Pressure X UPDRS motor scores). This relationship negatively influenced cognitive performance in three domains: exe cutive functioning, verbal memory and processing speed. Importantly, pulse pressure by itself did not directly affect cognitive status (i.e., no main effect).
59 Rather, the influence of poor vascular health (i.e., pulse pressure) exerted increasing worse i nfluence as a function of PD disease severity. Our findings indicate that the relationship between cognitive impairment and vascular comorbidities becomes increasingly apparent in individuals with more severe PD. To our knowledge, no studies have examined the relationship between pulse pressure and cognition in Parkinson patients. Studies among the normal elderly have shown pulse pressure to be related to cognitive domains of memory, executive functioning and processing speed (Waldstein et al., 2008; Dahle et al., 2009). The relationship between pulse pressure and cognitive status is hypothesized to be mediated small vessel lesions and disrupted cerebral perfusion (Qiu et al., 2003; Vicario In PD, small vessel disease (a seen by WML) have shown to have a differential relationship with cognition, depending on severity of PD symptoms. A previous study examining the relationship of white matter severity and cognition failed to find a relationship among newly diagnosed PDs (Dalaker et al., 2009). This is in contrast to previous studies finding significant relationships among more advanced PDs, and may suggest that the relationship between vascular pathology and cognition may be unique for individuals in advanced stages of PD (Beyer et al., 2006; Slawek et al., 2008). Comorbidities and Cognition: Possible Mechanisms Direct i nvestigation of possible mechanisms mediating the relationship between vascular pathology (specifically hypertension) is beyond the scope of this study. Turning to the literature, one proposed mechanism for the relationship between hypertension and cognitive decline relates to cerebral small vessel disease (Verdelho et al., 2010) Cerebral small vessel disease has been described as a group of
60 pathologie s that affect small arteries, arterioles, capillaries and small veins in the brain, and commonly result in lesions of subcortical stuctures (Pantoni, 2010). Recent studies have distinguished between at least two underlying pathophysiologies one involving atherosclerosis of small cerebral vessels and another involving alterations in the lumen The former has been most linked with high lipids and the latter with diabetes (Bezerra et al., 2012). Both are associated with hypertension, smoking, and age. I maging studies of small vessel disease, as indexed by white matter lesions (WML) on MRI, have shown a relationship between WML and motor symptoms in both Parkinson patients and normal elderly (Bohnen & Albin, 2011; see F igure 4 1). In addition to motor sym ptoms, WML have frequently been found to be related to worse performance on frontally mediated cognitive tasks such as set shifting, attention and processing speed (Breteler et al., 1994). Verdelho and colleagues (2010) showed that the relationship between hypertension and cognition was mediated by severity of WML. This was based on statistical findings that hypertension did not exert a significant relation with cognition (memory, executive function, etc.) w hen controlling for size of white matter lesions. The role of small vessel disease in mediating the relationship between hypertension and cognitive impairment in PD has not been directly investigated. However, WML have shown to be indicative of cognitive impairment among PD samples (Beyer et al., 2006; L ee et al., 2010). Lastly, it should be noted the relationship between cognitive impairment, high blood pressure and WML has not been a universal finding (Van De Pol, 2007)
61 Figure 4 1 Aging Diagram shows an overlap of white matter lesions between individuals with PD and normal aging. Nigrostriatal loss in frequent in normal aging, has been related to Parkinsonian symptoms in the absence of PD. From Bohnen & Albin 2011. In addition to e xecutive functioning, the presence of hypertension in the present study was also related to worse scores on measures of delayed verbal memory. Small vessel lesions in periventricular and temporal areas have been associated with memory performance, separat e from executive functioning (Smith et al., 2011). Another possible explanation is that the observed memory impairment is secondary to frontal executive dysfunction (Pillon et al., 1993). Inadequate organization skills (frontally mediated) may lead to diff iculty with retrieval of previously learned words, although recognition skills (temporally mediated) may be intact. Limitations and Future Directions Although this study adds to the literature on cognitive changes in Parkinson disease, it is not without limitations. Findings should be interpreted with caution for several reasons. First, our data consisted of a convenience sample of PD patients who were seen at the University of Florida. As such our findings may not be generalizable to the whole PD popul ation but restricted to those seeking care from a tertiary medical
62 center. Moreover, the sample was even more selective as it consisted of PD patients who were candidates for deep brain stimulation (DBS) or those for whom there was e status (and thus were referred for neuropsychological cognitive status, our sample may have included individuals who may have been more cognitively compromised. Second we relied on accurate documentation of information, but hope and assume that the most notable comorbidities were documented. Third, in a related manner, no di rect measures of comorbidities, other than blood pressure, were collected. Even blood pressure was obtained on only one occasion, rather than multiple readings under different conditions as is typically recommended. Finally, there was no indication of th comorbidities which may potentially play an important and critical role vis a vis impact on brain changes and cognition. Future studies should implement the use of direct measures of comorbidity status. This has specific implica tions for hypotension which has been shown be underreported among PD samples. Additional investigation into possible mechanisms, such as small vessel disease, mediating the effect of comorbidities on cognition is still needed. Studies investigating the lon g term impact of comorbidities on cognition would be beneficial. Future studies should also consider the impact that depression or apathy has on the relationship between comorbidities and cognitive functioning. Conclusion This current study adds to the literature on cognitive changes in Parkinson disease by providing evidence that comorbidities related to cardiovascular health namely
63 hypertension, hypotension, and pulse pressure are associated with cognitive performance. Hypertension was associated wi th worse performance on executive function and delayed verbal memory tasks, whereas hypotension was associated with better performance and even appeared to exert a buffering effect when it co occurred with hypertension. This type of relationship has not b een described before in the PD population and raises a number of questions of regarding the mechanisms underlying these relationships. While hypertension and cardiovascular risk factors may be less common in PD relative to the normal aging population, th ey nevertheless exerted a small but detrimental effect on cognition, particularly executive function and memory. This detrimental effect was present above and beyond the effects of PD neurodegeneration. Though not addressed in the current study, we suspec t that small vessel disease may be occurring in PD patients just as it does in those in older adults without Parkinson disease. However, at the same time, it appears that the influence of vascular factors on cognition may be most detrimental as PD severit y worsens. Taken together, the current findings suggest that cardiovascular comorbidities contribute to the cognitive decline in patients with Parkinson disease, and that this may be even more apparent with worsening disease progression. In light of these findings, consideration should be given to improving vascular health in Parkinson patients, as a means to assist in improving cognitive status Possible interventions could focus around reduction of vascular risk factors and may include strategies such a s maintaining a healthy diet, exercising or better treatment/control vascular related diseases. This can be particularly challenging in a neurodegenerative disorder characterized by overall
64 slowing and motor symptoms (i.e., loss of balance, falls) that fo ster reduced mobility. Due to the high risk of cognitive impairment and dementia in PD, and the frequent occurrence of hypertension in PD and aging populations, even a small benefit to minimize further cognitive decline would be important.
65 APPENDIX A PREVALENCE OF COMORB IDITIES Hypertension 38.1 Hypotension 30.2 Hypercholesterolemia 41.6 High Tryglycerides 5.0 Anemia 0.9 Diabetes 10.3 Thyroid Disease 9.9 Heart Attack 5.6 Angina 7.3 Atrial Fibrillation 7.0 Arrhythmia 2.6 Heart Murmur 2.9 Con gestive Heart Failure 1.8 Peripheral Vascular disease 0.3 Pacemaker Surgery 3.2 Cardiac Stent Surgery 5.9 Bypass Surgery 5.6 Valve Surgery 0.3 HIV/AIDS 0.0 Stroke 2.3 Transient Ischemic Event 2.3 Hemiplegia 0.3 Arthritis 25.5 Vasculitis 0.0 L upus 0.0 Asthma 4.7 COPD 4.3 Pulmonary Embolism 2.1 Emphysema 0.9 GERD 19.6 Peptic Ulcer Disease 2.9 Gallbladder Disease 0.0 Hepatitis 0.0 Liver Cirrhosis 0.0 Colon Polyps 7.6 Cancer 18.8 0.0 Kidney Stones 7.0
66 Urinary Incont inence 9.1 BPH 11.7 Seizures 1.8 Head Injury 3.5 Brain Tumor 0.3 Peripheral Neuropathy 19.1 Migraines 2.9 Glaucoma 2.9 Cataracts 10.6 Macular Degeneration 0.9 Tinnitus 2.3 Sleep Apnea 7.6 Restless Leg Syndrome 3.8 RBSD 11.1
67 APPENDIX B PERCE NT OF COMORBIDITIES REP ORTED BY AGE Cholesterol Hypertension Hypotension Neurologic Cardiac Arthritis GERD Cancer Prostate RBSD Respiratory Diabetes 27.7 24.8 23.8 28.7 6.9 13.9 11.9 18.8 5.0 9.9 7.9 7.9 60 69 37.0 35.6 28.1 31.1 25.2 31.1 23.0 20.0 10.4 13.3 7.4 11.1 61.0 54.3 39.0 27.6 47.6 29.5 22.9 17.1 20.0 9.5 16.2 11.4 Chi Squared 25.3 19.6 6.1 0.4 43.7 10.3 5.5 0.3 11.6 1. 1 5.8 0.9 Sig <.001 <.001 .047 .830 <.001 .006 .065 .854 .003 .581 .055 .651
68 REFERENCE LIST Aarsland, D., Andersen, K., Larsen, J. P., Lolk, A., Nielsen, H., & Kragharensen, P. (2001). Risk of d ementia in Parkinson's d isease. Neurology, 56(6 ), 730 736. Aarsla nd, D., & Kurz, M. (2010). The e pidemiology of dementia associated with Parkinson d isease. Journal of the Neurological Sciences, 289(1 2), 18 22. Alexander, G E, DeLong, M R, & Strick, P L (1986). Parallel organization of functionally s eg regated circuits l inking b asal g anglia and c ortex. Annual Review of Neuroscience, 9(1), 357 381. Allcock, L. M., Kenny, R., Mosimann, U., Tordoff, S., Wesnes, K. A., Hildreth, A. J., et al. (2006). Orthostatic hypotension in Parkinson's d isease: a ssociatio n with c ognitive d ecline? International Journal of Geriatric Psychiatry, 21(8), 778 783. American Gastroenterological Association. The Burden of Chronic Gastrointestinal Diseases Study 2001. Anson, O., & Para n, E. (2005). Hypertension and c ognitive f uncti oning among the elderly: an o verview. American Journal of Therapeutics, 12(4), 359 365. Aronow, W. S., Fleg, J. L., Pepine, C. J., Artinian, N. T., Bakris, G., Brown, A. S., et al. (2011). ACCF/AHA 2011 Expert Consensus Document on Hypertension in the Elde rly. J Am Coll Cardiol, j.jacc.2011.2001.2008. Artinian, NT, Duggan, C, & Miller, P (1993). Age differences in p atient r ecovery p atterns f ollowing c oronary a rtery b ypass s urgery. American Journal of Critical Care, 2(6), 453 461. Beyer, M. K., Aarsland, D., Greve, O. J., & Larsen, J. P. (2006). Visual r ating of White Matter H yperintensities in Parkinson's d isease. Movement Disorders, 21(2), 223 229. Bezerra, D.C., Sharrett, A.R., Matsushita, K., Gottesman, R.F., Shibata, D., Mosley, T.H., et al. (2012). Risk Factors for Lacune Subtypes in the Atherosclerosis Risk in Communities (ARIC) Study. Neurology, 78(2), 102 108. Birns, J., & Kalra, L. (2008). Cognitive Function and Hypertension. J Hum Hypertens, 23(2), 86 96. Blacher, J., Staessen, J. A., Girerd, X., Ga sowski, J., Thijs, L., Liu, L., et al. (2000). Pulse Pressure Not Mean Pressure Determines Cardiovascular Risk in Older Hypertensive Patients. Arch Intern Med, 160(8), 1085 1089. Bohnen, N. I., & Albin, R. L. White matter lesions in Parkinson disease. Nat Rev Neurol, 7(4), 229 236.
69 Breteler, MM, van Amerongen, NM, van Swieten, JC, Claus, JJ, Grobbee, DE, van Gij n, J, et al. (1994). Cognitive c orrelates of v entricular e nlargement and cerebral w hite m atter l esions on Magnetic Resonance Imaging. The Rotterdam Study. Stroke, 25(6), 1109 1115. Buter, T. C., van den Hout, A., Matthews, F. E., Larsen, J. P., Brayne, C., & Aar sland, D. (2008). Dementia and survival in Parkinson d isease. Neurology, 70(13), 1017 1022. Charlson, M., Szatrowski, T. P., Peterson, J., & Gold, J. (1994). Validation of a Combined Comorbidity Index. Journal of Clinical Epidemiology, 47(11), 1245 1251. Chiu, H. F., Wing, Y. K., Lam, L. C., Li, S. W., Lum, C. M., Leung, T., et al. (2000 ). Sleep related Injury in the elderly -an e pidemiological s tudy in Hong Kong. Sleep, 23(4), 513 517. Comella C. L., Nardine T. M., Diederich N. J., & Stebbins G. T. Sleep related violence, injury, and REM sleep behavior disorder in Parkinson's disease. Neurology 1998; 51: 526 9. Convit, Antonio, Wolf, Oliver T., Tarshish, Chaim, & de Leon, Mony J. (2003). Reduced glucose t olerance is a ssociated with p oor memory p erformance and h ippocampal a trophy among n ormal e lderly. Proceedings of the National Academy of Sciences, 100(4), 2019 2022. Cronin Golomb, A. (2010). Pa Neuropsychology Review, 20(2), 191 208. Dahle, C. L. Jacobs, B S ., Raz, N (2009) Aging, vascular risk, and cognition: Blo od glucose, pulse pressure, and cognitive performance in healthy adults. P sychology and Aging, 24(1) 154 162. Dalaker, T. O., Larsen, J. P., Dwyer, M. G., Aarsland, D., Beyer, M. K., Alves, G., et al. (2009). White Matter Hyperintensities do not Impact Co gnitive Function in Patients with Newly Diagnosed Parkinson's Disease. NeuroImage, 47(4), 2083 2089. Dalrymple Alford, J.C., MacAskill, M.R., Nakas, C.T., Livingston, L., Graham, C., Crucian, G.P., et al. (2010). The MoCA. Neurology, 75(19), 1717 1725. de Lau, L. M. L., & Breteler, M. M. B. (2006). Epidemiology of Parkinson's disease. The Lancet Neurology, 5(6), 525 535. Desmond, D. W., Tatemichi, T. K., Paik, M., & Stern, Y. (1993). Risk Factors for Cerebrovascular Disease as Correlates of Cognitive Functi on in a Stroke Free Cohort. Arch Neurol, 50(2), 162 166.
70 Dubois, B., & Pillon, B. (1996). Cognitive d eficits in Parkinson's disease. Journal of Neurology, 244(1), 2 8. Duron E, Hanon O. Vascular risk factors, cognitve decline, and dementia. Vascular Health and Risk Management. 2008; 4(2), 363 381 Fahn S Elton R L Members of the UPDRS Development Committee. Unified Marsden C D Calne D Holstein N 2. Plurham Park, N.J.: Macmillian Healt hcare Information, 1987:153 63. Farmer, M. E., Kittner, S. J., Abbott, R. D., Wolz, M. M., Wolfs, P. A., & White, L. R. (1990). Longitudinally Measured Blood Pressure, Antihypertensive Medication Use, and Cognitive Pe rformance: the Framingham Study. Journal of Clinical Epidemiology, 43(5), 475 480. Farmer, M. E., White, L. R, Abbott, R. D., Kittner, S. J., Kaplan, E., Wolz, M. M., et al. (1987). Blood Pressure and Cognitive Performance. American Journal of Epidemiology 126(6), 1103 1114. Franklin, S. S., Khan, S. A., Wong, N. D., Larson, M. G., & Levy, D. (1999). Is Pulse Pressure Useful in Predicting Risk for Coronary Heart Disease? : The Framingham Heart Study. Circulation, 100(4), 354 360. Gibb, W R, & Lees, A J (19 88). The Relevance of the Lewy Body to the Pathogenesis of Idiopathic Parkinson's Disease. Journal of Neurology, Neurosurgery & Psychiatry, 51(6), 745 752. Goldstein, D. S. (2003). Dysautonomia in Parkinson's Disease: Neurocardiological Abnormalities. The Lancet Neurology, 2(11), 669 676. Goldstein, D. S., Holmes, C. S., Dendi, R., Bruce, S. R., & Li, S. T. (2002). Orthostatic Hypotension from sympathetic denervation in Parkinson's disease. Neurology, 58(8), 1247 1255. Haaxma, C. A., Bloem, B. R., Borm, G. F., Oyen, W. G., Leenders, K. L., Eshuis, S. et al. (2007). Gender differences in Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 78(8), 819 824. Hachinski, V. C., Potter, P., & Merskey, H. (1987). Leuko Araiosis. Arch Neurol, 44(1), 21 23. Haugarvoll, K., Aarsland, D., Wentzel Larsen, T., & Larsen, J. P. (2005). The influence of cerebrovascular risk factors on incident dementia in patients with Parkinson's disease. Acta Neurologica Scandinavica, 112(6), 386 390. Hays W. L. Statistics 5th edition. 1994. Cengage Learning, Clifton Park, NY.
71 Hely, M. A., Reid, W. G. J., Adena, M. A., Halliday, G. M., & Morris, J. L. (2008). The Sydney multicenter study of Parkinson's disease: The inevitability of dementia at 20 years. Movement Disorder s, 23(6), 837 844. Hermann, B. P., Wyler, A. R., Somes, G., Berry, A. D. III, & Dohan, F. C. (1992). Pathological Status of the Mesial Temporal Lobe Predicts Memory Outcome from Left Anterior Temporal Lobectomy. Neurosurgery, 31(4), 652 657. Herrmann, L. L ., Le Masurier, M., & Ebmeier, K. P. (2008). White matter hyperintensities in late life depression: a systematic review. Journal of Neurology, Neurosurgery & Psychiatry, 79(6), 619 624. Hilker, R., Thomas, A. V., Klein, J. C., Weisenbach, S., Kalbe, E., Bu rghaus, L., et al. (2005). Dementia in Parkinson disease. Neurology, 65(11), 1716 1722. Hughes, A. J., Daniel, S. E., Kilford, L., & Lees, A. J. (1992). Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico pathological study of 100 c ases. Journal of Neurology, Neurosurgery & Psychiatry, 55(3), 181 184. Huse, D. M., Schulman, K., Orsini, L., Castelli Haley, J., Kennedy, S., & Lenhart, G. (2005). Burden of illness in Parkinson's disease. Movement Disorders, 20(11), 1449 1454. Jak, A.J ., Bondi, M.W. Delano Wood, L., Wierenga, C. Corey Bloom, J., Salmon D.P. et al. (2009 ). Quantification of five neuropsychological approaches to defining mild cognitive impairment American Journal of Geriatric Psychiatry, 17 (5), pp. 368 375 Jemal, A., Siegel, R., Ward, E., Murray, T., Xu, J., & Thun, M. J. (2007). Cancer Statistics, 2007. CA: A Cancer Journal for Clinicians, 57(1), 43 66. Jones, J., Malaty, I., Okun, M., Bowers, D. (February, 2012) Health Comorbidities, Cognition and Quality of Life i Foundation Quality Initiative. Poster to be presented at the Fortieth Annual Meeting of the International Neuropsychological Society. Montreal, Canada. Jones, T. A. (1969). Skewness and kurtosis as criteria of normality in observed frequency distributions. Journal of Sedimentary Research, 39(4), 1622 1627. Jurica, P. J., Leitten, C. L., & Mattis, S. (2001). DRS 2 : Dementia rating scale 2 : professional manual. Lutz, FL: Psychological Assessment Resources. K ennedy, K. M., & Raz, N. (2009). Aging white matter and cognition: Differential effects of regional variations in diffusion properties on memory, executive functions, and speed. Neuropsychologia, 47(3), 916 927.
72 Kilander, L., Nyman, H., Boberg, M., Hansson L., & Lithe ll, H. (1998). Hypertension Is related to c ognitive impairment : a 20 y ear follow up of 999 m en. Hypertension, 31(3), 780 786. Kirsch Darrow, L. (2010). Apathy, neurocognitive functioning, and Parkinson's disease [electronic resource] / by Lin dsey Kirsch Darrow. Gainesville, Fla: University of Florida. Knopman, D., Boland, L.L., Mosley, T., Howard, G., Liao, D., Szklo, M., et al. (2001). Cardiovascular risk factors and cognitive decline in middle aged adults. Neurology, 56(1), 42 48. Kuo, H. K. Sorond, F., Iloputaife, I., Gagnon, M., Milberg, W., & Lipsitz, L. A. (2004). Effect of Blood Pressure on Cognitive Functions in Elderly Persons. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 59(11), 1191 1194. Launer, L J., Masaki, K., Petrovitch, H., Foley, D., & Havlik, R. J. (1995). The Association Between Midlife Blood Pressure Levels and Late Life Cognitive Function. JAMA: The Journal of the American Medical Association, 274(23), 1846 1851. Leritz, E C., Salat, D H., Williams, V J., Schnyer, D M., Rudolph, J L., Lipsitz, L et al. (2011). Thickness of the human cerebral cortex is associated with metrics of cerebrovascular health in a normative sample of community dwelling older adults. NeuroImage, 54 (4), 2659 2671. Levin, B. E., & Katzen, H. L. (2005). Early cognitive changes and nondementing behavioral abnormalities in Parkinson's disease. Adv Neurol, 96, 84 94. Low, P. (2008). Prevalence of orthostatic hypotension. Clinical Autonomic Research, 18(0), 8 13. M ayeux, R., Marder, K., Cote, L. J., Denaro, J., Hemenegildo, N., Mejia, H., et al. (1995). The Frequency of Idiopathic Parkinson's Disease by Age, Ethnic Group, 142(8), 820 827. McDowell, F. H., & Lee, J. E. (1970). Levodopa, Parkinson's Disease, and Hypotension. Annals of Internal Medicine, 72(5), 751 752. Meredith, G. E., Halliday, G. M., & Totterdell, S. (2004). A critical review of the development and importance of proteinaceo us aggregates in animal models of Parkinson's disease: new insights into Lewy body formation. Parkinsonism; Related Disorders, 10(4), 191 202. Mil ner, B. (1954). Intellectual function of the temporal lobes. Psy chological Bulletin, Vol 51(1), 42 62.
73 Mold, J. W., Vesely, S. K., Keyl, B. A., Schenk, J. B., & Roberts, M. (2004). The prevalence, p redictors, and c onsequences of peripheral s ensory n europa thy in o lder p atients. The Journal of the American Board of Family Practice, 17(5), 309 318. Murat, E. (2003). Dementia associated with Parkinson's disease. The Lancet Neurology, 2(4), 229 237. Nanhoe Mahabier, W., de Laat, K. F., Visser, J. E., Zijlmans, J., de Leeuw, F. & Bloem, B. R. (2009). Parkinson disease and comorbid cerebrovascular disease. Nat Rev Neurol, 5(10), 533 541. National Heart, Lung, and Blood Institute, National Institutes of Health. Morbidity and M ortality: 2004 chart book on cardiovasc ular, lung, and blood diseases. http://www.nhlbi.nih.gov/resources/docs/04a_chtbk.pdf. Accessed January 24, 2012 Niimi, Y., Ieda, T., Hirayama, M., Koike, Y., Sobue, G., Hasegawa, Y., et al. (1999). Clinical and physiological characteristics of autonomic failure with Parkinson's disease. Clinical Autonomic Research, 9(3), 139 144. Novak, V, & Hajjar, I. (2010). The relationship between blood pressure and cognitive function. Nat Rev Cardiol, 7(12), 686 698. Nussbaum, R. L., Ellis, C. (2003). Alzheimer's di sease and Parkinson's disease. The New England journal of medicine, 348(14), 1356. O'brien, R. (2007). A Caution Regarding Rules of Thumb for Variance Inflation Factors. Quality & Quantity, 41(5), 673 690. Oka, H., Mochio, S., Onouchi, K., Morita, M., Yosh ioka, M., & Inoue, K. (2006). Cardiovascular dysautonomia in de novo Parkinson's disease. Journal of the Neurological Sciences, 241(1 2), 59 65. Microvascular Disease in Bra in and Kidney. Hypertension, 46(1), 200 204. Pantoni, L. (2010). Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. The Lancet Neurology, 9(7), 689 701. Parkinson J. Essay on the Shaking Palsy. London, England: Sherwood, Neely, and Jones; 1817. Peralta, C., Stampfer Kountchev, M., Karner, E., Kollensperger, M., Geser, F., Wolf, E., et al. (2007). Orthostatic hypotension and attention in Parkinson's disease with and without dementia. Journal of Neural Tra nsmission, 114(5), 585 588.
74 Pillon, B., Deweer, B., Agid, Y., & Dubois, B. (1993). Explicit m emory in Alzheimer's, Huntington's, and Parkinson's d iseases. Arch Neurol, 50(4), 374 379. Qiu, C, Winblad, B., Viitanen, M., & Fratiglioni, L. (2003). Pulse p ress ure and risk of Alzheimer disease in persons aged 75 years and o lder. Stroke, 34(3), 594 599. Raz, N., Rodrigue, K. M., & Acker, J. D. (2003). Hypertension and the Brain: Vulnerability of the Prefrontal Regions and Executive Functions. Behavioral Neuroscie nce, 117(6), 1169 1180. Reginster, J. Y. (2002). The Prevalence and Burden of Arthritis. Rheumatology, 41(1), 3 6. Rektor, I Goldemund, D Sheardova, K Rektorova, I Michalkova, Z & Dufek, M (2009). Vascular Pathology in Patients with Idiopathic P arkinson's disease. Parkinsonism & Related Disorders, 15(1), 24 29. Rose, K. M., Couper, D., Eigenbrodt, M., Mosley, T., Sharrett, R., & Gottesman, R. (2010). Orthostatic hypotension and cognitive function: the Atherosclerosis Risk in Communities Study. Ne uroepidemiology 34, 1 7. Rothman, K. J. (1990). No Adjustments Are Needed for Multiple Comparisons. Epidemiology, 1(1), 43 46. Saville, D. J. (1990). Multiple Comparison Procedures: The Practical Solution. The American Statistician, 44(2), 174 180. Saxby, B. K., Harrington, F., McKeith, I. G., Wesnes, K., & Ford, G. A. (2003). Effects of Hypertension on Attention, Memory, and Executive Function in Older Adults. Health Psychology, 22(6), 587 591. Schrag, A., Jahanshahi, M., & Quinn, N. (2000). How Does Parki nson's Disease Affect Quality of Life? A Comparison with Quality of Life in the General Population. Movement Disorders, 15(6), 1112 1118. Scigliano, G., Ronchetti, G., Girotti, F., & Musicco, M. (2009). Sympathetic Modulation by Levodopa Reduces Vascular R isk Factors in Parkinson Disease. Parkinsonism & Related Disorders, 15(2), 138 143. Selvin, E., & Erlinger, T. P. (2004). Prevalence of and Risk Factors for Peripheral Arterial Disease in the United States. Circulation, 110(6), 738 743. Senard, J M, Rai, S Lapeyre Mestre, M, Brefel, C, Rascol, O, Rascol, A, et al. (1997). Prevalence of orthostatic hypotension in Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 63(5), 584 589.
75 Sheline, Y. I., Barch, D. M., Garcia, K., Gersing, K., Piepe r, C., Welsh Bohmer, K., et al. (2006). Cognitive Function in Late Life Depression: Relationships to Depression Severity, Cerebrovascular Risk Factors and Processing Speed. Biological Psychiatry, 60(1), 58 65. Shin, C., Abbott, R. D., Lee, H., Kim, J., & K imm, K. (2004). Prevalence and Correlates of Orthostatic Hypotension in Middle aged Men and Women in Korea: the Korean Health and Genome Study. J Hum Hypertens, 18(10), 717 723. Smith, E.E., Salat, D.H., Jeng, J., McCreary, C.R., Fischl, B., Schmahmann, J. D., et al. (2011). Correlations between MRI white matter lesion location and executive function and episodic memory. Neurology, 76(17), 1492 1499. Stewart, R., & Liolitsa, D. (1999). Type 2 Diabetes Mellitus, Cognitive Impairment and Dementia. Diabetic Med icine, 16(2), 93 112. Tomlinson, Claire L., Stowe, R., Patel, S., Rick, C., Gray, R., & Clarke, C. E. (2010). Systematic Review of Levodopa Dose Equivalency Reporting in Parkinson's Disease. Movement Disorders, 25(15), 2649 2653. Tzourio C, Dufouil C, Duci metiere P, Alperovitch A. (1999). Cognitive decline in individuals with high blood pressure: a longitudinal study in the elderly. EVA Study Group. Epidemiology of Vascular Aging. Neurology ; 53:1948 1952. Uc, E. Y., McDermott, M. P., Marder, K. S., Anderson S. W., Litvan, I., Como, P. G., et al. (2009). Incidence of and Risk Factors for Cognitive Impairment in an Early Parkinson Disease Clinical Trial Cohort. Neurology, 73(18), 1469 1477. Van De Pol, L. A., Korf, E. S. C., van Der Flier, W. M., Brashear, H. R., Fox, N. C., Barkhof, F., et al. (2007). Magnetic Resonance Imaging Predictors of Cognition in Mild Cognitive Impairment. Arch Neurol, 64(7), 1023 1028. Van Den Eeden, S. K., Tanner, C. M., Bernstein, A. L., Fross, R. D., Leimpeter, A., Bloch, D. A., e t al. (2003). Incidence of Parkinson's Disease: Variation by Age, Gender, and Race/Ethnicity. American Journal of Epidemiology, 157(11), 1015 1022. Verdelho, A., Madureira, S., Ferro, J. M., Basile, A., Chabriat, H., Erkinjuntti, T., et al. (2007). Differe ntial Impact of Cerebral White Matter Changes, Diabetes, Hypertension and Stroke on Cognitive Performance Among Non disabled Elderly. The LADIS Study. Journal of Neurology, Neurosurgery & Psychiatry, 78(12), 1325 1330. Verdelho, A., Madureira, S., Moleiro, C., Ferro, J.M., Santos, C.O., Erkinjuntti, T., et al. (2010). White Matter Changes and Diabetes Predict Cognitive Decline in the Elderly. Neurology, 75(2), 160 167.
76 Verhamme, K. M. C., Dieleman, J. P., Bleumink, G. S., van der Lei, J., & Sturkenboom, M. C. J. M. (2002). Incidence and p Prevalence of l ower urinary t ract symptoms s uggestive of benign prostatic h yperplasia in primary c European Urology, 42(4), 323 328. Vicario, A., Martinez, C. D., Baretto, D., Casale, A. Diaz, & Nicol osi, L. (2005). Hypertension and Cognitive Decline: Impact on Executive Function. The Journal of Clinical Hypertension, 7(10), 598 604. Wallace, R. B., Lemke, J. H., Morris, M. C., Goodenberger, M., Kohout, F., & Hinrichs, J. V. (1985). Relationship of Fre e recall Memory to Hypertension in the Elderly. The Iowa 65+ Rural Health Study. Journal of Chronic Diseases, 38(6), 475 481.. Wilson, P. W. F., Anderson, K. M., & Kannel, W. B. (1986). Epidemiology of Diabetes Mellitus in the Elderly: The Framingham Study The American Journal of Medicine, 80(5, Supplement 1), 3 9. Wilson, P. W. F., D'Agostino, R. B., Levy, D., Belanger, A. M., Silbershatz, H., & Kannel, W. B. (1998). Prediction of Coronary Heart Disease Using Risk Factor Categories. Circulation, 97(18), 1 837 1847. Wolf Maier, K, Cooper, R., Banegas, J. R., Giampaoli, S, Hense, H., Joffres, M., et al. (2003). Hypertension Prevalence and Blood Pressure Levels in 6 European Countries, Canada, and the United States. JAMA: The Journal of the American Medical As sociation, 289(18), 2363 2369. Wu, W., Brickman, A. M., Luchsinger, J., Ferrazzano, P., Pichiule, P., Yoshita, M., et al. (2008). The Brain in the Age of Old: The Hippocampal Formation is Targeted Differentially by Diseases of Late Life. Annals of Neurolog y, 64(6), 698 706. Yap P. L., Niti M., Yap K. B., Ng T. P. (2008) Orthostatic hypotension, hypotension and cognitive status: early comorbid markers of primary dementia? Dementia and Geriatric Cognitive Disorders, 26: 239 246 Zahodne, L. B., Bowers, D., Pri ce, C. C., Bauer, R. M., Nisenzon, A., Foote, K. D., et al. (2011). The Case for Testing Memory With Both Stories and Word Lists Prior to DBS Surgery for Parkinson's Disease. The Clinical Neuropsychologist, 25(3), 348 358. Zgaljardic, D. (2003). A Review o f the Cognitive and Behavioral Sequelae of Parkinson's Disease: Relationship to Frontostriatal Circuitry. Cognitive and behavioral neurology, 16(4), 193.
77 BIOGRAPHICAL SKETCH Jacob Daniel Jones was born i n 1987 in San Diego California. Jacob grew up mos tly in Camarillo, California where he attended California State University Channel Isla nds (CSUCI) and graduated Suma c um Laude in 2010. Upon graduation, Jacob attended the Clinical and Health Psychology doctoral program at the University of Florida, where his studies focus on cognitive and emotional changes among individuals work toward earning his Ph.D. in Clinical Psychology