|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 By LINDSEY KIRSCH DARROW A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Lindsey Kirsch Darrow
3 ACKNOWLEDGMENTS I would like to thank my mentor, Dawn Bowers, for her generous time, encouragement, and constant support. I would like to thank M ichael Marsiske for his ready support, and assistance from the beginning to the end of this project. I would also like to thank Michael O kun and his colleagues at the Movement Disorders Center for access to patients and for insightful comments on the stud y. I thank Russell Bauer, Jay Rosenbek lab, especially Laura Zahodne, Utaka Springer, and Ania Mikos for their insightful comments on the study. I thank my husband, Adam Darrow, for his unwavering devotion, love, patien ce, and support for all of my endeavors. Finally, I would like to express appreciation and respect for all of the PD patients who participated in this study, an d whose lives have been altered by mood disorders.
4 TABLE OF CONTENTS page LIST OF TABLES ................................ ................................ ................................ ........................... 6 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ..... 9 CHAPTER 1 STATEMENT OF THE PROBLEM ................................ ................................ ...................... 11 2 BACKGROUND ................................ ................................ ................................ .................... 14 Motor Symptoms in Parkinson Disease ................................ ................................ .................. 14 Depression in PD ................................ ................................ ................................ .................... 15 Defining Apathy ................................ ................................ ................................ ..................... 16 Lack of Accepted Diagnostic Criteria for Apathy ................................ ........................... 18 Apathy Rating Scales in PD ................................ ................................ ............................ 19 Phenomenology of Apathy: Clinical Correlates and Course ................................ ........... 23 Separating Apathy and Depression in PD ................................ ................................ .............. 26 Confirmatory Factor Analysis: Rationale for Selection of Factors ................................ ........ 27 Prior Exploratory Factor Analysis of BDI II and AS ................................ ...................... 29 Potential Contribution of Confirmatory Factory Analysis ................................ .............. 30 Potential Implicatio ns ................................ ................................ ................................ ...... 31 Apathy and Cognitive Impairment ................................ ................................ ......................... 33 Overview of Cognitive Impairment in PD ................................ ................................ ...... 33 Executive Functions: Definition and Review of Findings in PD ................................ .... 35 Cognitive Impairment: Association with Demographic and Disease Variables ............. 39 Mood Disorders in PD: Effect on Cognition. ................................ ................................ .. 40 3 SPECIFIC AIMS OF THE PRESENT STUDY ................................ ................................ ..... 51 4 PARTICIPA NTS AND METHODS ................................ ................................ ...................... 53 Participants ................................ ................................ ................................ ............................. 53 Procedure ................................ ................................ ................................ ................................ 55 Overview of Desi gn ................................ ................................ ................................ ......... 55 Study 1: Examining Apathy and Depression Factors ................................ ...................... 55 Analytic Approach ................................ ................................ ................................ ........... 57 Study 2: Apathy and Cognitive Impairment ................................ ................................ .... 59 Analytic Approach ................................ ................................ ................................ ........... 60 5 RESULTS ................................ ................................ ................................ ............................... 66 Aim 1: Relationship Between Apathy and Depression ................................ .......................... 66
5 Frequency of Apathy and Depression Symptoms ................................ ........................... 66 Differences Between Apathy Depression Groups on Demographics, Disease Variables, and Medication Usage ................................ ................................ ................ 66 Differences Between Apathy Depression Groups on State and Trait Anxiety ............... 69 Factor Structure of Apathy and Depression: Confirmatory Factor Analysis .................. 70 Aim 2: Relationship between Apathy and Cognition ................................ ............................ 74 Cognitive Domain Regression Analyses ................................ ................................ ......... 75 Summary of Regression Results ................................ ................................ ...................... 79 Ap ................................ .......... 79 ................................ ................... 81 Summary of Hier archical Regression Results ................................ .......................... 82 6 DISCUSSION ................................ ................................ ................................ ....................... 112 Prevalence of Apathy and Depression ................................ ................................ .................. 113 Factor Structure of Apathy and Depression in PD ................................ ............................... 114 Apathy and Cognition ................................ ................................ ................................ ........... 119 gnitive Domains ................................ ............................... 119 Apathy and Stroop color word performance: Relationship to anterior cingulate cortex ................................ ................................ ................................ .................. 120 Apathy and speeded v erbal fluency ................................ ................................ ....... 123 Apathy Depression Group Findings ................................ ................................ .............. 125 Comparison to the Current Literature ................................ ................................ ............ 127 Relationship Between Apathy and Anxiety ................................ ................................ .......... 128 Limitations ................................ ................................ ................................ ............................ 129 Conclusions and Directions for Futur e Research ................................ ................................ 131 LIST OF REFERENCES ................................ ................................ ................................ ............. 134 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 145
6 LIST OF TABLES Table page 2 1 ................................ ....................... 48 2 2 Proposed consensus criteria for a syndrome of apathy from the European Psychiatric Association ................................ ................................ ................................ ......................... 49 2 3 Summary of PD studies examining cognition in apathetic versus nonapathetic groups ... 50 4 1 Patient characteristics ................................ ................................ ................................ ......... 62 4 2 Proposed loadings of each item. ................................ ................................ ........................ 63 4 3 Tests categorized into rationally derived cogniti ve domains. DV = dependent variable. ................................ ................................ ................................ .............................. 64 5 1 Demographic, disease variable, and medication average scores, standard deviations, and ranges between no symptom, pure apathy, pure depression, and mixed apathy depression groups ................................ ................................ ................................ ............... 85 5 2 STAI State and Trait average scores, standard deviations, and ranges between no symptom, pure apathy, pure depression, and mixed apathy depression group s ................ 86 5 3 Confirmatory factor analysis loadings and uniquenesses ................................ .................. 90 5 4 Factor correlations ................................ ................................ ................................ ............. 9 1 5 5 Goodness of fit statistics for confirmatory factor analysis of full four factor model and alternative nested models ................................ ................................ ............................ 91 5 6 Descriptive statistics for indi vidual cognitive tests ................................ ........................... 92 5 7 Descriptive statistics for cognitive domains used in hierarchical regressions (z score metric) ................................ ................................ ................................ ................................ 92 5 8 Correlations among cognitive domains (Pearson correlations). ................................ ........ 93 5 9 Multicollinearity statistics for hierarchical multiple regressions ................................ ....... 94 5 10 Hierarchical multiple regression results, showing the relationship between predictors and Executive functioning ................................ ................................ ................................ 96 5 11 Hierarchical multiple regression results, showin g the relationship between predictors and Processing speed ................................ ................................ ................................ ......... 98 5 12 Hierarchical multiple regression results, showing the relationship between predictors and Verbal episodic memory ................................ ................................ ........................... 100
7 5 13 Hierarchical multiple regression results, showing the relationship between predictors and Working Memory ................................ ................................ ................................ ...... 102 5 14 Hierarchical m ultiple regression results, showing the relationship between predictors and Language Functioning ................................ ................................ ............................... 104 5 15 Hierarchical multiple regression results, showing the relationship between predictors and Wisconsin Card Sorting Test, raw number of categories achieved .......................... 106 5 16 Hierarchical multiple regression results, showing the relationship between predictors and Wisconsin Card Sorting T est, raw number of perseverative errors .......................... 108 5 17 Demographic, disease variable differences between apathetic and nonapathetic groups. ................................ ................................ ................................ .............................. 110 5 18 Mood and anxiety differences between apathetic and nonapathetic groups. ................... 111
8 LIST OF FIGURES Figure page 2 1 Overlap between apathy and depression, apathy alone, and depression alone between groups. ................................ ................................ ................................ ................................ 47 5 1 Prevalence of apathy and depression in 161 Parkinson patients ................................ ........ 84 5 2 Mean State Anxiety Scores across Apathy Depression subgroups. ................................ .. 87 5 3 Mean Trait Anxiety scores across Apathy Depression subgroups ................................ .... 88 5 4 Percentages of patients with clinically elevated trait anxiety across Apathy Depression subgroups. ................................ ................................ ................................ ....... 89
9 Abstract of Dissertation Presented to the Graduate School of the University o f Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy APATHY, NEUROCOGNITIVE FUNCTIONING, By Lindsey Kirsch Darrow August 2010 Chair: Dawn Bowers Major: Psychology Apathy is a common neuro the present study were twofold: 1) to test the hypothesis that apath y and depression are separate syndromes in PD, and 2) to determine the effect of apathy on neurocognitive performance in PD patients after controlling for important comorbidities such as dementia, depression, and disease variables. One hundred sixty one non demented PD patients (age = 64.1, 8.4 yrs; UPDRS motor severity = 25.13 8.6) were administered neuropsychological t ests and completed the Apathy Scale and Beck Depression Inventory II Items were proposed to load on to four factors: 1) an apathy factor repres enting loss of motivation, 2) dysphoric mood factor representi ng sadness and negativity, 3) loss of interest an d pleasure factor, and 4) somatic factor representing bodily complaints CFA was used to ex amine the fit of the items to the factors. Hierarchical regression was used to quantify whether apathy uniquely explained variance in specific cognitive domains (e .g. Executive functioning, Processing speed, Verbal episodic memory, Working memory, and Language domains ).
10 There was a good fit for the overall CFA .01 RMSEA was .060 ( p = .16). The four factor model was significantly better than all alternative n ested models at p <.001. Apathy explained incremental variance in Executive functioning, but did not explain signific ant variance in any other cognitive domain. Apathy was negatively related to Executive functioning and this relationship was driven by a significant negative relat ionship with Stroop I nterference performance (p <.01) and a trend for a negative relationsh ip between apathy and semantic fluency (p = .06). Results support the concept that apathy and depressi on are discrete factors T his finding argues for an alternative of the Apathy Scale and Beck Depressio n Inventory II This will help disentangle symptoms related to apathy, depression overlapping symptoms and somatic complaints. F indings also support a relationship b etween apathy and Executive functioning. A pathy is related to certain aspects of Executive functioning, such as c ognitive interference/inhibition and verbal fluency whereas it is not related to other aspects of Executive functioning such as shifting mental set.
11 CHAPTER 1 STATEMENT OF THE PROBLEM Parkinson disease (PD) is one of the most common neurodegenera tive disorders of late life. Over one million Americans suffer from PD, and 50,000 new cases are diagnosed each year (McDonald, Richard, & DeLong, 2003; Nussbaum & Ellis, 2003) The incidence of Parkinson disease increases with age and as the aging population increases the incidence of PD is expected to triple by the year 2050 ( Tanner et al., 2002 ). The disease is characterized neuropathologically by loss of neurons in the substantia nigra pars compacta, dopamine depletion in the basal ganglia, and the presence of Lewy bodies. Patients present with motor symptoms of tremor, bradykinesia (slowness of movement), rigidity, and postural instability. While these motor symptoms are the hallmark features of the disorder neuropsychiatric symptoms are highly prevalent and can be some of the most disturbing, disabling, and complex aspects of PD. One such neuropsychiatric symptom is apathy. Apathy refers to negative/deficit symptoms such as blunted emotions, loss of inte rest, and lack of productivity. Many aspects of apathy are unknown. For example, it is unknown whether apathy is a unique syndrome or a subcomponent of depression in PD. This has important implications both for understanding the neural substrates of moo d disorders (e.g. Do separate neural systems underlie the two types of symptoms?) and for differential diagnosis and treatment (e.g. treatments for depression may not be effective for apathy, and vice versa). Studies to date have sugges ted that depression and apathy are separable (Isella et al., 2002; Kirsch Darrow, Fernandez, Marsiske, Okun, & Bowers, 2006; Pluck & Brown, 2002) However, these studies are limited meth odologically by the use of total scores on clin ical inventories used to assess their presence and severity This is problematic because apathy and depression scales overlap in content. It is possible that a particular symptom endorsement on the B eck D epression I nventory (BDI) might better represent a pathy and is being
12 counted towards depression and vice versa. The primary aim of the present study is to address whether apathy and depression are separable in a way that disentangles the total score confound of overlapping symptomatology. To overcome th e limitation of overlapping symptoms being confirmatory factory analysis will be used to examine individual items of the Beck Depression Inventory II (BDI II) and the Apathy Scale (AS) Based on pre l iminary observat ions, d iscrete factors were proposed to be : 1) an apathy factor representing loss of motivation, 2) a dysphoric mood factor representing sad mood/ negativity, 3) a loss of interest and pleasure factor, representing the overla p between apathy and depression and 4) a somatic factor representing bodily complaints. In this way, an a priori factor structure is hypothesized and can determine how well the items conform to these factors. This will be a strong test of whether apathy and depression are indeed diss ociable in PD. Another unknown factor regarding apathy is its relationship to cognitive functioning. Is apathy related to a particular pattern of cognitive impairment? Several studies have examined this question, but most have confound s (Isella et al., 2002; Pluck & Brown, 2002; Starkstein, Mayberg, Preziosi et al., 1992; Zgaljardic et al., 2007) S ome studies did not screen out patients with dementia. This is problematic because groups of a path etic and nonapathet ic patients appear to have an unequal distribution of dementia cases (Isella et al., 2002; Starkstein, Mayberg, Preziosi et al., 1992) With more demented individuals in the high apathy group, lower cognitive score s could be attributed to deme ntia and not to apathy per se. Second, studies did not control for depression or demographic variables that have been shown to effect cognitive functioning in PD. For example al though Zgaljardic et al. (2007) and Pedersen et al. (2009) scre ened out dementia patients, they did not control for depression (significantly higher in their apathetic group) or disease severity
13 apathetic group at trend for both Zgaljardi c et al. (2007) and Pedersen et al. (2009). Further, the Pedersen (2009) study reported significantly higher UPDRS motor severity in their apathetic group ve rsus their non apathetic group. This makes it difficult to attribute cognitive differences betwee n groups specifically to apathy. The second aim of the present study is to determine whether apathy influences cognitive functioning in a nondemented group of PD patients after controlling for depressio n, and demographic factors (i.e. age, gender educati on ) and disease factors (i.e. duration of PD, severity of PD).
14 CHAPTER 2 BACKGROUND Motor Symptoms in Parkinson Disease The hallmark motor symptoms of Parkinson disease are a resting tremor, bradykinesia, muscular rigidity, and a gait disturbance. Res ting tremor is the most recognizable symptom of PD; however, some patients experience a tremor during activity (postural or action tremor) as well as during rest, and approximately 15% of patients never show tremor during the course of the disease ( Martin et al., 1983 ). Bradykinesia is slowness in execution of movement, while muscular rigidity is tightness and stiffness of the muscles. Tremor, rigidity, and bradykinesia begin unilaterally, but become bilateral as the disorder progresses. Gait is characte rized by stooped posture, shuffling steps, festination (short steps that become quicker and quicker as if the patient were about to run) and propulsion (forward inclination of the body as if the patient were about to fall forward) ( Lieberman, 1995; Tyler, 1992 ). Patients may experience motoric gait and are unable to take any steps forward. Given all of these motor symptoms, PD was once considered solely as a motor disorder. Now, we know that PD is a complex disorder that af fects multiple domains of functioning. For example, PD causes disturbances of mood and motivation including depression, anxiety, and apathy. Of these, apathy is the least understood. As such, the purpose of the present study is further knowledge about t his important non motor symptom of PD. Before turning to specific aims and hypotheses of the present study, relevant background will be presented as follows : 1) an overview of depression in PD, 2) definition of apathy, 3) an overview of apathy in PD and differential diagnostic questions between apathy and depression, and 4) a review of the relationship between mood symptoms and cognitive symptoms in PD.
15 Depression in PD Depression is a common occurrence in PD and has been studied extensively. There have been over 100 published English language studies specifically examining depression in PD. Distinction should be drawn between depression symptoms based on clinical scales versus the diagnosis of Major depressive disorder (MDD) using the Diagnostic and St atistical Manual of Mental Disorders (DSM) IV criteria. Rates found in studies vary, with number of patients meeting DSM criteria being less frequent than those endorsing levels of depression above a particular cut off point on clinical scales (Slaughter, Slaughter, Nichols, Holmes, & Martens, 2001) Even if patients do not meet full diagnostic criteri a, they may still have significant disability from their mood symptoms and benef it from treatment (Judd, Paulus, Wells, & Rapaport, 1996; Lyness et al., 1996) A recent meta analysis by Reijnders et al. (2009) reviewed 36 studies of depression in PD and found the overall prevalence of MDD to b e 19%. This prevalence is much higher than the 2 9% prevalence of MDD found in the general population (Diagnostic and Statistical Manual of Mental Disorders IV, DSM IV). The prevalence of clinically significa nt depression in PD whether or not there was a presence of a DSM diagnosis, was 35% (Reijnders, Ehrt, Lousberg, Aarsland, & Leentjens, 2009) Moreover, depressive symptoms are a primary factor impacting quality of life ( McDonald et al., 2003; Phillips, 1999) In a randomized, multi center study of patients with PD, caregivers, and clinicians, depressive symptoms were cited as the most important factor in patient quality of life ratings more important than medications and disease severity ("Factors impacting on quality of life in Parkinson's disease: results from an international survey," 2002) Unfortunately, depression is difficult to assess in PD. Many of the symptoms of Park inson disease itself overlap with those of depression. Patients often have insomnia, psychomotor slowing, fatigue, and concentration difficulties. Also, the symptom of reduced facial
16 nondepressed individuals with PD. Misdiagnosis of depression can go either way by assigning symptoms to depression that actually represent PD symptoms or by assigning symptoms to PD that actually represent depression. Accurate recognition and diagnosis is critical for appropriate and effective treatment. Failing to treat depression symptoms may cause increased disability whereas inappropriate treatment may lead to unnecessary side effects. Selective serotonin reuptake inhibitors (SSRIs) are the typical first line choice for depression in PD. In addition, a recent clinical trial of antidepressant treatment in 52 PD patients reported that nortriptyline, a dual reuptake inhibitor of serotonin and norepinephrine, was more efficacious than both placebo and the SSRI paroxetine for remitting depressive symptoms (Menza et al., 2009) Selective serotonin reuptake inhibitors (SS RIs) can cause side effects of insomnia, nausea, agitation, and sexual dysfunction. Similarly, the norepinephrine side effect profile includes orthostatic hypotension, constipation, dry mouth, insomnia, and dizziness (Menza et al., 2009) Apathy may not respond to antidepressant agents and thus patients may experience negative side effects without receiving benefit from their mood symptoms. Defi ning Apathy The term first appeared in medical use in the writings of Hugh lings Jackson ( Hughlings Jackson 1931). He distinguished the florid/positive symptoms from the deficit/negative symptoms in schizophrenia. Positive symptoms were seen as an excess of normal function (e.g. hallucinations), whereas negative symptoms (e.g. alogia, apathy, and poverty of speech) were seen as a loss of normal function. Yet, this loss of normal function is seen in disorders beyond schizophrenia. Robert Marin highlighted this fact in the early 1990s. He proposed that apathy could manifest in neurological disorders as both a symptom and a syndrome. His key paper in
17 1991 proposed diagnostic criteria for a syndrome of apathy (Marin, 1991) (see Table 2 1 for criteria ). According to his view, a syndrome of apathy includes a primary lack of motivation Marin defined motivation as a higher order construct that re fers ion intensity and persistence of goal definition by positing that apathy manifests itself in behavioral, cognitive, and emotional domains. At least o ne sy mptom from each domain is required for a diagnosis of apathy. The behavioral domain includes symptoms such as of lack of effort, lack of productivity, and such as loss of interest in new exp The affective domain includes symptoms of flattened affect and lack of response to positive or negative events. The lack of motivation in a syndrome of apathy is primary, and not purely accounted for by intellectual impairment such as dementia, emotional distress, or impaired consciousness (e.g. delirium) Marin emphasized that if apathy did occur in these states (e.g. such as emotional distress such as depression) then apathy would be considered a symptom and not a full syndrome. Levy and Dubois (2006 ) authors defined reduction of voluntary [or goal directed] be They distinguished their definition as being more quantifiable (i.e. observable behavior change) than the psychological term They, like Marin, point ed out that the reduction in voluntary behavior must be below the previo us levels of functioning. They to make sure it is unchanged (i.e. not related to changes in reward contingencies in the
18 environment) and disability status in order to ensure the behavior is not related to the physical inability to perform previous soci al roles, work, and recreation (R. Levy & Dubo is, 2006) Lack of Accepted Diagnostic Criteria for Apathy Although Marin proposed the criteria for a syndrome of apathy, the se criteria have never been fo rmally accepted into the Diagnostic and Statistical Manual of Mental Disorders (DSM). The DSM IV us es the term apat hy only as one of a list of possible symptom s of personality change due to a general medical condition (DSM IV TR). The ICD 10 does not include apathy at all ( ICD 10, World Health Organization). The lack of formally accepted diagnostic cr iteria is problematic because researchers and clinicians do not have a set of standard ized instructions to diagnose patients in a uniform manner. This leads to variability in patient classification and hampers comparisons across studies. It is also diffi cult for treatment oriented clinical trials to appropriate select patients and evaluate outcome (e.g. full remission, partial remission, etc) without clearly ac cepted criteria Towards the goal of creating consensus criteria, Starkstein and colleagues mo dified (Starkstein, 2000; Starkstein & Leentjens, 2008) Starkstein and Leentjens (2008) added a duration requirement that apathy symptoms must occur during most of the day for at least four weeks. They removed the exclusion criteria of 1) intellectual im pairment and 2) emotional distress. Importantly, the authors pointed out that a patient could have both a syndrome of apathy and dementia Patients can also have both a syndrome of apathy and a major depressive d isorder (Starkstein, 2000; Starkstein & Leentjens, 2008) origin al criteria overlooked these dual diagnosis situations. R ecent work from this research group confirmed the common overlap between apat hy and dementia and depression (Starkstein et al., 2009) Out of 164 PD patients assessed, 52 (32%) were diagnosed with apathy based on the A pathy Scale Further, a full 83% of this apathetic g roup had DSM IV based depression diagnoses of major
19 depression, minor depression, or dysthymia. Fifty six percent of the apathetic group had dementia based on DSM IV assessment (Starkstein et al., 2009) Very recently a task force of experts from Europe, Australia, and North America convened in France and developed and published consensus diagnostic criteria for apathy in neuropsychiatric and neurologic al disorders (Robert et al., 2009) Sergio Starkstein was a member of this task force and the proposed criteria are similar to his criteria described above (Starkstein & Leentjens, 2008) The European diagnostic consensus criteria require: loss of, or diminished motivation accompanied by loss of goal directed behavior, cognitive activity, or diminished emotion. See T able 2 2 The duration criteria is that apathy occurs most of the day, every day for at least four weeks. A difference between the European criteria and criteria are the requirement of one symptom in at lea st two of the three domains: behavi or, cognition, and emotion (Robert et al., 2009) P revious criteria required at least one symptoms in all three domains (Starkstein, 2000; Starkstein & Leentjens, 2008) Further, the European criteria structure each domain with two symptoms each one representing self initiated or symptom referring to the pati 2. Apathy Rating Scales in PD R esearchers are currently working towards the goal of uniformly accepted gold standard diagnostic criteria for a syndrome of apathy. In the meanwhile, a number of assess ment scales t hat measure symptoms are used to quantify apathy. These include the Apathy Evaluation Scale (AES), an abbreviated version of the AES known as the Apathy Scale (AS), the Apathy Inventory (AI), the Lille Apathy Rating Scale (LARS), and the Fron t al Systems of Behavior Scale (FrSBe). Single item s are also and the Neuropsychiatric Inventory (item 7).
20 Apathy Evaluation Scale (AES) and Apathy Scale (AS). The AES is an 18 item Likert scale mea suring the behavioral, cognitive, and emotional symptoms of apathy. There are self, clinician, and informant versions. It was developed by Robert Marin (Marin, Biedrzycki, & Firinciogullari, 1991) Marin validated the original scale on approximately 90 subjects aged 55 85 years with a diagno compared their scores to those of normal elderly controls. C onvergent and divergent validity with other scales (e.g. anxiety and depression scales ) was established with the multit rait multimatrix method (Marin et al., 1991) Predictive and external validity were investigated by observing participants in various scenarios such as persistence at video game play and time spent examining novelty gadgets. Self reported apathy scores were negatively correlated with total s cores on video games and the difficulty level at which participants chose to play. Thus, there was a behavioral correlate to the self reported symptoms. The Apathy Scale (AS) is a modified version of the AES. The AES was shortened to 14 items from 18 ite ms and wording was simplified by Starkstein and colleagues (Starkstein et al., 1992). It was validated specifically in PD patients and showed good internal consistency reliability and test retest reliability (Starkstein et al., 1992) The AS has been use d frequently in PD studies since its validation (Isella et al., 2002; Czernecki et al., 2002; Kirsch Darrow et al., 2006; Zahodne et al., 2009; Starkstein et al, 2009). A recent study from our laboratory demonstrated evidence for construct validity of the AS. Twenty eight nondemented PD patients and 19 age They also completed the AS, Lille Apathy Rating Scale (LARS) BDI, and UPDRS motor section. Participants were videota ped while they sat alone with six novel toys/gadgets on a table in front of them. The amount of time spent eng aging with the gadgets was the dependent
21 variable T he apathetic group (classified based on both the AS and LARS) spent significantly less time interacting with the gadgets than the nonapath etic group. Depression, motor severity, and levo dopa equivalent dosage were not related to the amount of time spent engaged with the gadgets Results indicated a strong correlation ic behaviors and the AS, demonstrating evidence for the construct validity of the AS (F erencz et al., 2009; submitted to the American Academy of Neurology). Apathy Inventory (AI). The AI is a 3 item 0 12 point Likert scale assessment of apathy. Adequate i nternal consistency and test retest reliability were established with 60 patients with subsequent studies have used it besides the original authors and it needs to be further validated in a lar ger population of PD patients (Robert et al., 2002; (Leentjens et al., 2008) Lille Apathy Rating Scale (L ARS). The LARS is a 33 item semistructured interview for apathy inal conceptualization It yields a global score and four composite subscores that reflect different dimensions of apathy (i.e. intellectual curiosity, action initiation, emotion, and self awareness ). The total score ranges from 36 to +36, with more positive scores indicating more severe apathy. The LARS was especi ally designed for PD patients, and validated in 159 French PD patients with an d without dementia (Sockeel et al., 2006) It has good internal cons istency, test retest, interrater reliability, and acceptable item total correlation. The authors validated it against clinical judgment of apathy. A recent study from our la boratory validated the English version of this scale in an American population. It used receiver operating characteristic (ROC) analysis to compare the LARS to the AS (Zahodne et al., 2009) A cut off score was identified of 22 (sensitivity = 64%, specificity = 92%, PPV = 88%, NPV = 75%), which was slightly higher than the 16 cut off score in the French population. Recently, the
22 authors have validated a caregiv er version as well, the LARS i (Dujardin, Sockeel, Delliaux, Destee, & Defebvre, 2008) Frontal Sy stems of Behavior Scale (FrSBe) The FrSBe is a 46 item rating scale with self report and family report versions. It measures behavioral traits associated with damage to frontal subcortical circuits. The FrSBe has three subscales, the Apathy Scale (14 i tems), the Disinhibition Scale (15 items), and the Executive Dysfunction Scale (17 items). Patients are asked to rate their pre PD status and their current status on each of the items. The FrSBe has been validated in frontal lobe brain injury and neurode generative disorders, including PD (Grace & Malloy, 2001). Single item assessments: (UPDRS) item 4 and Neuropsychiatric Inventory (NPI) item 7 The UPDRS is the most widely used assessment instru ment in PD and has four sections. Part I includes mood, mentation and behavior. A single item, item 4, assesses motivati on/initiative on a 5 point Likert scale It focuses on withdrawl from behavioral activities and does not capture the emotional dimension of apathy. Mix ed results have been found with regard to the sensitivity and specificity of the UPDRS item 4. Two studies report that a cut off of 2 has adequate sensitivity/specificity when compared to a shorte ne d version of the AS (Starkstein & Merello, 2007) ; Pederson et al., 2007). A study from our laboratory used the full AS for validation purposes and found that a cut of f of 2 had poor sensitivity (52%), while higher cut offs unacceptably lowered the specificity. Consequently, caution was recommend ed for using item 4 as a s creening instrument for apathy (Kirsch Darrow et al., 2009) T he N europsychiatric Inventory (NPI) also ha s a single item that assess es apathy. The NPI is a structured interview given by the clinician to an informan t regarding 10 forms of behavioral
23 disorder that occur with dementia (Cummings, Mega, Gray et al., 1987). These include delusions, hallucinations, agitation/aggression, depression, anxiety, euphoria, apathy, disinhibition, irritability/lability, and aberrant motor behavior. All items assess frequency and severity of the behavior being evaluated. In a small sample (n = 12) of PD patients, the NPI item 7 had good inter rater agreement, otherwise the specific item still needs to be validated in PD patients. Task Force Recommendations for Apathy Scales in PD To evaluate these apathy scales and their use in PD, a task force was commissioned by the Movement Disorders Society in 2008 (Leentjens et al., 2008) T he task force divided the ir findings into three levels: R ecom mended, Suggested, and L isted. They r ecommended the AS They s uggested further stu dy for the AES, LARS and NPI The task force r ecommended using the UPDRS item 4, but only for screening purposes because it is a single item. They did not r ecommend the AI (i.e. they listed it as a scale that was available) The y did not evaluate the F rSBe The authors deemed it outside the scope of the assignment because the FrSBe also evaluates the neuropsychological f eatures of disinhibition and executive dysfunction (Leentjens et al., 2008) The current study used the recommended AS. Phenomenology of Apathy: Clinical Correlates and Course Relationship to demographics and disease variables. P rogress h as been made in identifying the clinical correlates of apathy i n terms of its association with demographic s and disease variables Multiple s tudies have found no relationship between apathy and age or years of education (Starkstein et al., 1992; Pluck & B rown, 2002; Isella et al., 2002; Zgalardic et al., 2007; Dujardin et al., 2007; Pedersen, Alves et al., 2009) O ne exception is a recent population based study in Norway A utho rs reported that their apathetic group of patients with PD was older and had fewer years of education than their non apathetic patients with PD (Pedersen,
24 Larsen, Alves, & Aarsland, 2009) Approximately one quarter of their total patients with PD had a dementia diagnosis, and their apathet ic nonapathetic groups differed with respect to the number of patients with dementia (50% of patients with apathy had dementia and 12% of patients without apathy had dementia). Due to this, it is hard to know whether age and educational differences might have been similar in their non demented apathetic and nonapathetic patients The majority of studies that examined apathy and gender found no relationship between them (Starkstein et al., 1992; Pluck & Brown, 2002; Isella et al., 2002; Zgal j ardic et al., 2007; Dujardin et al., 2007; Peders en, Larsen et al., 2009) However, another population based study in Norway of early drug nai ve nondemented patients with PD found that apathy was significantly associated with male gender. S eventy five percent of patients were male in their apathetic g roup (30 out of 40 patients), and 53% were male in their non apathetic group (72 out of 135 patients (Pedersen, Alves et al., 2009) Gender also signific antly predicted apathy i n the regression models. The authors reported that this finding was unexpected and hypothesized that low testosterone levels may be related to apathy in their male PD patients but testosterone was not measured in the study. Alternately, they also specula ted that female caregivers were more likely to report negative symptoms than male caregivers (Pedersen, Alves et al., 2009) This explanation implies that apathy levels are actu ally similar between genders, but there is a bias of more frequent reporting of apathy by female caregivers It also assumes fem ale caregivers of male patients. There are mixed findings with regard to the relationship between apath y and the severity of disease Several studies do not find a relationship between apathy and disease severity based on the UPDRS III motor subscale (Isella et al., 2002; Dujardin et al., 2007) the
25 Hoe hn Yahr staging of PD severity (Aarsland et al., 1999) or the tremor, akinesia, or rigidity items from the UPDRS III (Starkstei n et al., 1992; Aa rsland et al., 1999). Two other studie s report that higher apathy is associated with greater disease severity based on the UP DRS III motor subscale (Pedersen, A lves et al., 2009) ; (Pedersen, Lars en et al., 2009) Another reports a trend towards higher apathy scores in patients with more severe Hoehn Yahr stages (Zgaljardic et al., 2007) In contrast to the mixed findings regarding disease severity, studies do not find a relationship between apathy and disease duration ( e.g. length of PD ) (Starkstein et al., 1992; Aarsland et al., 199 9; Pluck & Brown, 2002; Isella et a l., 2002; Dujardin et al., 2007; Pedersen, Alves et al., 2009) The fact that apathy is not related to dise ase duration, but is frequently re lated to disease severity may at first seem counterintuitive. However, it is reasonable given the variability of progression of the disease across patients. Some patients can stay relatively healthy with PD for 10 15 or more years, whereas others may dec line more rapidly (Nutt, Hammerstad, & Gancher; Peretz & Cummings, 1988). Therefore, it seems that apathy symptoms may increase with disease severity (at least in some findings) but that duration of the illness is not associated with apathy. Course of apa thy. Apathy occurs in all st ages of PD including early drug naive patients who are recently diagnosed (Pedersen, Alves et al., 2009) However, t he longitudinal course of apath y in terms of factors such as fluctuations in severity across time, length of an episo de, and remission information are relatively unknown In a recently completed study from this lab oratory 139 idiopathic PD patients were assessed at two time points sep arated by 18 to 30 months (mean = 24 months SD = 3.6 months). The AS, BDI, anxiety scores, quality of life scores, and motor scores were completed at these time points. Results indicated that AS and
26 motor scores significantly worsened, while depression scores showed no difference. Worsening apathy correlated with increased motor severity at trend. Worsening apathy was associated with lower baseline AS, worsening anxiety trait scores, motor scores, and quality of life. This study suggests that apathy a nd depression may have a separate course of presentation in moderately severe PD patients, with apathy worsening and depression remaining stable over two years (Zahodne et al., 2009; Submitted to the American Academy of Neurology). Other studies have focu sed on apathy following deep brain stimulation (DBS) surgery for PD. Some studies have found that apathy increases after surgery (Kirsch Darrow et al., 2009, submit ted ). Other information about the longitudinal course in sur gery naive patients is unknown Separating Apathy and Depression in PD T he majority of the literature points to the ability to separate apathy from depression. Specifically in PD, i t has been argued that apathy can occur in the absence of depression and depression can occu r in the a bsence of apathy. Starkstein and colleagues (1992) examined a consecutive series of 50 PD patients for apathy and depression symptoms and found that 12% were at or above the cut off score for clinically significant apathy but did not meet criteria for dep ression (criteria based on DSM III). They also found no difference in apathy levels between patients with and without depression. Many other studies have shown apathy without depression in patients with PD (Levy et al., 1998; Aarsland et al., 1999; Isella et al., 2002; Pluck & Brown, 2002; Kirsch Darrow et al., 2006; Zgal ja rdic et al., 2007; Pedersen et al., 2009) Pluck and Brown (2002) found apathy to be significantly higher in PD than in control osteoarthritis patients. They did not find a difference in depression levels between patients with high level s of apathy and those with low levels of apathy. A study from our labo ratory compared a pa thy and depression in a PD group and a movement disordered control group of dystonia patients (Kirsch Darrow et al., 2006) It was hypothesized that P D patients would have
27 significantly more apathy than patients with dystonia and a large proportion of PD patients would experience apathy in the absence of depression. Eighty PD patients and 20 with dystonia complet ed mood questionnaires that included the Apathy S cale and the Beck Depression Inventory Apathy was significantly more frequent in PD (freq=51%, 41/ 80) than in dystonia (freq=20%, 4/20, p = .012). Moreover, the frequency of apathy in the absence of depression was substantial in PD and nonexiste nt in dystonia (PD = 29%, dystonia = 0%, p <.01) (see Figure 2 1). These data provide support for t he hypothesis that apathy may not be limite d to co occur within depression, but may be its own separate syndrome. However a weakness of current studies i s the manner by which patients are classified as exhibiting 'apathy' or exhibiting 'depression.' To date, this classification has primarily involved using total scores from apathy and depression inventories (ex. Beck Depression Inventory [BDI] and Apathy Scale [AS]). This is problematic because apathy and depression have overlapping symptoms such that the scales overlap in content. The BDI includes item content that is possible that a particular symptom endorsement on the BDI might better represent apathy and is being counted toward depression total score and vice versa. One way to address this limitation is to use a more sophisticated methodological appr oach that of confirmatory factor analysis. Factor analytic techniques can investigate whether items load onto discrete apathy and depression factors. Confirmatory factor analysis in particular will allow specific a priori hypotheses about the underlyin g structure of the data to be tested. Confirmatory Factor Analysis: Rationale for Selection of Factors The present study will take a confirmatory analytic approach to test the hypothesis that Apathy S cale and Beck Depression Inven tory II will load onto 4 factors: 1) an apathy factor representing loss of motivati on, 2) a dysphoric m ood factor representing
28 sadness and negativity, 3) a loss of interest and loss of pleasure factor representing the overlap between apathy and depression and 4) a somatic factor representing bodily complaints. See Table 4 2 for a detailed listing of hypothesized factors and the items that compose them. The rational for these hypothesized factors is based on several ideas about how apathy and depression can be parsed. Sad mood, dysphoria, and tearfulness are commonly found in depressive syndromes. H owever, literature suggests that apathy does not include sad mood, and apathy has (Brown & Pluck, 2000) S ymptoms of thlessness, failure, disappointment, and guilt are hypothesized to be rela ted to depression Apath y does not involve pessimism or negative self and event appraisal. Instead, what distinguish es apathy is loss of motivation. While loss of motivation can b e found n found in depression, it is The rationale for a loss of interest and anhedonia factor is based on the idea that these symptoms are c ommon to both apathy and depression. In fact, a N ational I nstitute of N eurological D isorders and S troke (NINDS) depression in PD workgroup highlighted loss of interest and pleasure a s symptom s that overlaps between depression and primary apathy (Marsh, McDonald, Cummings, & Ravina, 2006) The current study will test this hypot diminished interest or pleasure in all, or almost all, is a criteria for a major depressive disorder (DSM IV TR). Furthermore Pluck and Brown (2002) reported that their apathetic PD patients scored higher in anhedonia (based on the Snaith Hamilton Pleasure Scale) than their nonapathetic PD patients. Lastly, the somatic factor encompasses physical symptoms such as fatigue, changes in appetite and sleep patterns, and changes in sexual drive They are hypothesized to cluster together and to load onto their own
29 sepa rate factor Further, it seems appropriate to separate the physical symptoms because even in t he absence of mood disorder, patients often exper ience these symptoms as part of PD Prior Exploratory Factor Analysis of BDI II and AS The factor structure of the BDI II has been examined in large undergraduate student populations. The BDI II manual reports a two factor solution as the most parsimonious description of the data in an undergraduate sample T he factors corresponded to Cognitive Affective symptom s and Somatic symptoms (Beck, 1996). This factor structure was replicated in a study of over 1,000 Canadian undergraduates (Doz o is, Dobson, & Ahnberg, 1998) Two factors accounted for 46% of the variance and were labeled the Cognitive Affective dimension and the Somatic Vegetative dimension. The Cognitive Affective factor consisted of items such as past failure, worthlessness, self dislike, pessimism, self criticalness, indecisiveness, guilty feelings, suicidality, punishment feelings, and sadness. The Somatic Vegetative factor primarily consisted of the items of changes in sleep, fatigue, loss of energy, irritability, agitation, loss of interest in sex, loss of interest, loss of pleasure, and changes in appetite (Doz o is, Dobson, & Ahnberg, 1998) Consi stent with these exploratory analyses, the present study proposes a separate somatic factor. Notably, the Dozis et al (1998) study includes loss of interest and anhedonia along side the other physical symptoms. In a clinical sample of PD patients with m ore prevalent mood disorders th an a diverse undergraduate population loss of interest/anhedonia are predicted to fall into a separate factor There are no published studies on the structure of the Apathy Scale. The original validation study of the Apat hy Evaluation Scale included an exploratory factor analysis. In a normal elderly adults, three main factors were reported (Marin et al., 1991). These included a General apathy factor (Factor 1) with items involving having initiative, lack of productivity,
30 emotional flatness, lack of effort. Factor 2 included items that dealt with curiosity or novelty seeking, interest in things (in general), learning, new exper iences, and spending time in problems and needing structure for daily activities. In our laboratory, a n exploratory factor analysis (EFA) study of the Apathy S cale was recently completed on 78 non demented PD patients with clinically significant apathy Initial analysis indicated that items 3 (Are you concerned about your condition?) and 13 (Are you neither happy nor sad, just in between?) did not correlate wit h the total apathy score and were excluded from the EFA. Principal axis factoring with Promax rotation was used. Results suggested a three factor solution that resembled Cognitive (items 1, 2, 6 involving learning, interest, and plans/goals for the futur e), Behavioral (items 4, 5, 7, 8 involving effort, seeking out activities, motivation, and energy), and Affective (items 10, 11, 14 involving emotional indifference and lack of concern) domains (Kay et al., submitted to A merican Academy of Neurology 2009 ) Potential Contribution of Confirmatory Factory Analysis Prior studies have investigated the latent structure of the BDI II and the AS. No studies have combined items from both scales in to one analysis. This will al low items to load across scales and c onstructs can be proposed that extend across both scales. Confirmatory factor analysis (CFA) was selected over EFA because it allows a priori hypotheses to be made and tested. CFA will contribute to the PD literature because it will test discrete apathy and depression factors. Individual items can be placed onto specific factors. Additionally, CFA has the benefit of multiple indices to measure the goodness of fit of the hypothesized model to the data, whereas EFA only has one (e.g. square root mean resid ual).
31 Potential Implications If the results support separate apathy and depression factors, this will provide increasing empirical support for the dissociability of these two mood disorders in PD If apathy manifest s a s a separate disorder, this ha s important implications for clinical practice and for the understanding of the neural substrates involved. First of all, apathy can gr eatly a ffect PD recent large scale surv ey of nonmotor symptoms 1,000 Italian PD patients reported th at apathy was one of the symptoms most negatively impacting quality of life (Barone et al., 2009). Further, t reatments specific to apathy in neurological disorders are being investigated. Amphetamines, atypical antipsychotics, dopaminergic agents, and ac etylcholinesterase inhibitors have been examined (van Reekum, Stuss, & Ostrander, 2005) Me thylphenidate (i.e. Ritalin) succ essfully treated apath y in a case study of a man with (Chatterjee & Fahn, 2002) There is preliminary support for the efficacy of some of these pharmacologica l interventions in neurological diseases (Galynker et al., 1997; Kaufer, 1999; Kraus & Maki, 1997; Van Reekum et al., 1995) Moreover, a clinical trial h as very recently been completed a t this center (University of Florida, Movement Disorders Center) investigating repetitive Transcranial Magnetic Stimulation (rTMS) for the treatment of apathy in PD Twenty four PD patients experiencing either mixed apathy /depression (n = 11) or pure apathy (n = 13) underwent apathy, depression, and motor assessment prior, immediately following, one month, and three months after treatment. Treatment was either high frequency rTMS (10Hz) or sham stimulation delivered over t he left dorsolateral prefrontal cortex for 10 days. Both rTMS and sham showed significant improvements in apathy immediately, at one month, and at three months post treatment. There was no difference between the level of improvement derived from rTMS vs. sham, and no differences between the pure apathy and
32 mixed apathy/depression group. This study indicates that brief, daily behavioral interventions are important for apathy, but does not demonstrate a unique effect of rTMS (Fernandez et al., 2009, submit ted to American Academy of Neurology) Furthermore d ifferent neural mechanisms may underlie apathy and depression in PD. Orbito frontal subcortical connections may underlie depression whereas mesial frontal/anterior cingulate cortex ventral tegmental con nections may underlie apathy in PD. Using Positron Emission Tomography (PET), Mayberg and colleagues demonstrated a relationship between depression in PD patients and hypometabolism in the orbital inferior area of the frontal lobe and the c audate, as com pared to nondepressed PD patients (Mayberg, 1994; Mayberg et al., 1990) Apathy may result from dysfunction of the anterior cingulate cortex circuit. This cortico striato pallido thalamo circuit consists of anteri or cingulate cortex ventral striatum ventral pallidum dorsomedial thalamus anterior cingulate cortex. Very few studies have as of yet examined the neuroanatomical basis for apathy in PD. Yet, past research has shown that bilateral lesions in the anteri or cingulate cortex (ACC) leads to a severe form of apathy called akinetic mutism. The patient makes no attempt to act, speak, or initiate activity. Patients can and were not motivated (Damasio & Tranel, 1992 ). Remy and colleagues (2005) examined apathy in twenty Parkinson patients using Positron Emission Tomography with a radioactive ligand that binds selectively to dopamine and norepinephrine r eceptors (Re my, Doder, Lees, Turjanski, & Brooks, 2005) D ecreased binding is a marker for cell loss. Higher apathy on the AES was related to reduced dopamine /norepinephrine binding in the ventral striatum. This is consistent wit h the circuit above. Taken togethe r, findings may indicate that apathy is a separate
33 mood disorder in PD, needing more careful assessment and treatment. One area needing further assessment is the relationship between apathy and cognitive impairment. Apathy and Cognitive Impairment Overvie w of Cognitive Impairment in PD broad spectrum of cognitive impairment, ranging from mild deficits in executive functioning to fully developed dementia. Further, t here is wide variability in cognitive impa irment in non demented patients. I mpairment in executive functioning, language, memory, and visuospatial skills have bee n described in non demented PD patients (Caballol, Marti, & Tolosa, 2007; Dubois & Pillon, 1997; Pillon, Czernecki, & Dubois, 2003; Taylor & Saint Cyr, 1995) Alt hough not pervasive across all individuals with PD memory, psychomotor speed, and executive fun ctioning, can be impaired as early as when the patient is initially diagnosed with P D (Levin & Katzen, 1995, 2005; Levin, Llabre, & Weiner, 1989; Muslimovic, Post, Speelman, & Schmand, 2005) Two domains that are commonly impaired in PD are memory and executive functioning. Memory impairment in PD is characterize d by deficits in free recall, with benefit from cuing and preserved recognit io n. This memory impairment profile is considered secondary to executive dysfunc tion Specifically, i t is thought that PD patients find it diffi cult to organize information for e ncoding and retrieval (Brandt, Shpritz, Munro, Marsh, & Rosenblatt, 2005) S ubcortical dementias ( i.e. also cal led fronto subcortical dementias such as P and Pro gressive Supranuclear Palsy ) are considered to exhibit memory deficits that are sec ondary to inefficient strategies of encoding/retrieval tha t are due to pathology of fronto striatal connections This memory profile can be contrasted with cortical deme prominent memory deficits, rapid
34 forgetting, little benefit from c uing, and impaired recognition (Albert, Feldman, & Willis, 1974; Elias & Treland, 1999; Pillon, Deweer, Agid, & Dubois, 1993) Executive functioning impairment in PD are widely rep orted in the literature, and is common even in patients without otherwi se significant cognitive impairment (Caballol et al., 2007; Muslimovic et al., 2005; Perry & Hodges, 1996; Williams Gray, Foltynie, Brayne, Robbins, & Barker, 2007) The type and severity of executive impairment ca n vary acros s patients. However, deficits in attentional set shifting, planning, concept formation, and inhibition of responses have been consistently described (Owen et al., 1992; Dubois & Pillon, 1997; Weintraub et al., 2005; Muslimovic, Post, Speelman & Schmand, 2005; Williams Gray et al., 2007 ; Caballol et al., 2007 ) A further review of executive functioning will be continued in the next section. Before turning to this a brief overview of PD related dementia is provided next The estimated prevalen ce of dementia in PD is between 25 30% of patients (Aar sland et al., 2005). T he DSM IV diagnostic criteria do not capture all of the characteristics of dementia in PD because of their focus on memory and on impaired activities of daily living. Memory def icits are not always prominent in PD dementia. Also, it can be difficult to determine if impairment in activities of daily living in PD are due to cognitive impairment or due to motor disability. Dementia in PD begins insidiously and is characteriz ed by a slowly progressive cognitive decline. I nitial complaints frequently involve concentration, poor immediate recall, slowed information processing, and word finding problems (Caballol et al., 2007). Factors that are associated with increased risk of demen tia in PD are older age and gre ater severity of motor symptoms (Aarsland, Andersen et al., 2001; Hobson & Meara, 2004) ; Hughes et al., 2000). Further, a longitudinal study by Levy et al. (2002) found support for th e combined effect of age
35 a nd sev erity of motor symptoms. R esu lts indicated that age was an important risk factor for d ementia only when coupled with more severe motor symptoms. Older age without severe motor symptoms was not associated with increased ris k of de mentia (G. Levy et al., 2002) Executive Functions : Definition and Review of Findings in PD The term executive functions is an umbrella ing functions such as planning, shifting mental sets, abstract reasoning, mental flexibility, and problem solving (Burgess, Veitch, de Lacy Costello, & Shallice, 2000; Damasio, 1995; Stuss, Shallice, Alexander, & Pic ton, 1995) Executive functions are the most complex of human behaviors, and are key in our ability to respond and adapt to novel situations (Lezak, 1982 ; Lezak, Howieson, & Loring, 2004 ) Impairment in one or more executive functions can have devastati ng effects on ion/educational performance independ ent functioning at home, and developing a nd maintaining social relationships (Chan, Shum, Toulopoulou, & Chen, 2008; Green, Kern, Braff, & Mintz, 2000) An important aspect of executive functions is how it fractionates into specific no predictive value for anot her type of executive function (Burgess, Alderman, Evans, Emslie, & Wilson, 1998) Over the la st two decades, progress has been made in isolating what areas of the prefrontal cortex are involved in which particular cognitive processes. For example, the dorsolateral prefrontal circuit is involved in planning, goal selecting, sequencing, and working memory whereas the orbitomedial circuit and anterior cingulate circuit are involved in response inhibition (MacDonald, Cohen, Stenger, & Carter, 2000) As it might be imagined, many cases involving executive functioning involve frontal lobe damage. However, importantly, subcortical structures and their connections to prefrontal lobe
36 deficits in PD have been ascribed to dysfunction in the connections between the striatum and prefrontal lobes (Elias & Tre land, 1999; Owen, 2004; Owen et al., 1992; Zgaljardic, Borod, Foldi, & Mattis, 2003; Zgaljardic et al., 2007) Unfortunately, the neuropsychological measurement of executive functions is difficult and sometimes controversial. First, t here is a paradoxic al need to structure a situation where the patient can demonstrate how well they can create structure for themselves (Lezak, 1982). In contrast, during the administration of most cognitive tests, the examiner decides wh at activity the subject is to do co nveys the rules, and leaves relatively little room for discretionary behavior on the part of subject (Lezak, Howieson, & Loring, 2004). To accurately measure executive functions, tests must allow the subject enough room to make decisions, and think of and chose alternatives for their own behavior and cognition (Lezak, Howieson, & Loring, 2004). Ecological validity is another issue when measuring executive functions. Although tests can demonstrate excellent psychometric properties, they may not map on to t he naturalistic tasks found in everyday life (Goldstein, 1 996; Sbordone, 1996). Also, at times, patients with frontal lesions perform equally well as controls on n europsychological tests, but experience multiple difficu lties in everyday life (Shallic e & Burgess, 1991) Thus, tests are not always sensitive to executive functioning difficulties. Daily life may require more complex multi step tasks that require goal setting, prioritization, reliance on prospective memory, and inhibition of inappropria te actions than tests are able to capture (Chan et al., 2008) However, neuropsychological measures of executive functioning are often criticized for tapping a number of executive functions all at the same time, making it difficult to ascertain which specific skill is impaired.
37 incorporate more tasks they are likely more ecologically valid. However, when they are more ecologically valid, they do not captur e the fractionation of the executive system Overview of Executive Impairment in PD Despite the difficulties with the measurement of executive functions described above progress has been made in understanding the executive impairment in PD. Studi es hav e compared the performance of PD patients to that of individuals with front al lobe damage. Like patients with frontal lobe damage (Milner, 1964), shifting impairment on the Wisconsin Ca rd Sorting Task (WCST ) (Canavan et al., 1989; Lees & Smith, 1983; Taylor, Saint Cyr, & Lang, 1986) This set shifting difficulty has been found on other tests in addition to the WCST. Downes et al. (1989) found t hat PD patients were impaired on visual discrimination learning when required to shift response sets b etween two stimulus dimensions (Downes et al., 1989) This was replicated and extended by Owen et al. (1991). PD patients and frontal lobe patients were impaired on this task but patients with unilateral temporal lobe damage or amygdala hippoc ampal damage were not impaired (Owen, Roberts, Polkey, Sahakian, & Robbins, 1991) Studies have also found that PD patients are impai red in planning abilities. PD patients were impaired in organizing picture stories on the P icture Arrangement subtest of the WAIS Authors interpreted this as impairment in s equencing and forward planning (Growdon, Corkin, & Rosen, 1990) Further, PD patients were impaired (increased) in terms of the amount of ti me spent thinking about their solution to the Drexel Tower of London planning t est. How ever, accuracy on this test was only impaired in patients with more severe Hoehn and Yahr disease staging (Owen et al., 1992)
38 Weintraub and colleagues (2005) found that increased severity of parkinsoni sm was associated with reduced inhibitory control on the Stroop in terference condition. They administered the Drexel Tower of London test, the Trail Making Test, and the Stroop Color Word Test Golden version, to 46 PD patients. Subscores from the measur es were factor a nalyzed Two factors emerged: 1) a planning factor that encompassed total moves, time violations, and execution time from the Drexel Tower of London test, and 2) a n inhibitory control factor con sisting of errors on Trails B, e rrors on Str oop Color Word task, Interference condition, and rule violations on the Drexel Tower of London test. The inhibitory control factor was negatively correlated with motor slowing, increased parkinsonism, and lower educational level. The planning factor was negatively correlated with apathy as measured by the Apathy Scale (Weintraub et al., 2005) However, this study was limited by a 96% male population (limiting generalizability), a relatively small sample size compared to the number of factors analyzed and the use of pr incipal component (PCA) extraction PCA is controversial and generally discarded in favor of exploratory factor analysis because it does not separate shared and unique variance, and therefore is less helpful in revealing latent variables (Gorsuch, 1990; C ostello & Osborne, 2005). As with most PD symptoms, executive function impairment is thought to be associated with dopaminergic dennervation of the striatum. This, in turn, leads to a cascade of dysfunction including the connections between subcortical an d cortical brain structures. It is thought that connections between the striatum and differential areas of cortex can account for some of the non motor symptoms of PD. For instance, disruptions between basal ganglia and dorsolateral prefrontal cortex is thought to affect working memory and set shifting ability. Disruptions between basal ganglia and anterior cingulate cortex is though to disrupt motivation and response
39 initiation (Zgaljardic, 2003, Lichter, 2000). However, this represents a broad general ization of the effects of complex neurodegenerative processes. In addition to dopamine systems, it is likely that non motor symptoms also involve other systems such as the cholinergic system. Dopaminergic treatment consistently improves motor functioning in PD, but it does not always improve cognition (Cooper et al., 1992; Zgaljardic et al., 2003 ; Gotham, Brown, & Marsden 1998) Cognitive Impairment: Association with Demographic and Disease Variables General demog raphic variables as well as disease specific variables are known to be related to cognitiv e impairment. In normal aging it has been well established that older age and fewer years of education are associated with poorer cognitive functioning (Collie, Shafiq Antonacci, Maruff, Tyler, & Currie, 1999; Scuteri, Palmieri, Lo Noce, & Giampaoli, 2005) In PD, age and education have been found to be associated with executive impairments, particularly on planning tasks (Taylor et al., 1986) In a meta analysis of 25 longitudinal studies of cognition in PD, age was significantly related to decline global cognitive abilities (measured by dementia rating scales such as MMSE) and memory. Lower educational levels were associated with decline in all cognitive domains (e.g. global cognitive abilities, memory, verbal abilities including verbal fluency, reasoning, attention, processing speed, visuo perceptial and visuo spatial abilities) (Muslimovic, Schmand, Speelman & De Hann, 2007). Further, d isease severity has been shown to be an important predictor of cognitive impairment. Studies have d emonstrated that patients at different stages of PD can be differentiated in terms of executive functions (e.g. specifically, planning on the Drexel Tower Test and attentional set shifting on a computerized test of visual stimulus dimensions where subject s had to shift each time to focus on the previously irrelevant stimulus dimension), working memory, and sh ort term spatial memory (Morris et al., 1988; Owen et al., 1992)
40 Mood Diso rders in PD: Effect on Cognition The fact that impairment in executive functions often accompanies PD is clear. W hat seems less wel l established is the effect of mood disorders on cognitive impairme nt in PD. It is notable that even in young, neurologically intact i ndividuals with Majo r D epressi ve D isorder (MDD) studies find that depression has variable effects on cognitio n. S ome studies suggest MDD is related to de creased concentration (one of the dia gnostic criteria for MDD) slowed information processing and impaired working memory ( review ed by Mayberg et al., 2002). O ther studies find no neuropsychological distinctions between moderately depressed individuals and non depressed individuals matched on age and education (Crews et al., 1999). ve Functioni ng in PD. The situation is more complex when exa mining depression in PD because it is a neurological disorder commonly associated with cognitive impairment such as slowed information processing, executive deficits, and memory retrieval problems There ar e mixed findings regarding whether depression exacerbates cognitive impairment in PD. Some studies report no differences in cognitive performance between depressed and non depressed PD patients (Beliauskas et al., 1989; Santamaria et al., 1986; Taylor et al., 1986; Huber et al., 1988 ) Others repor t that depressed PD patients have worse general cognitive functioning on the M ini M ental S tatus E xam (MMSE) and the Dementia Rating Scale (Troster, Stalp et al., 1995; Troster, Paolo et al., 1995; Starkstein et al., 1990 ; Starkstein et al., 1992 ). More recent studies are more methodologically sophisticated because they match depressed PD patients and non depressed PD patients on demographic, disease variables, and gene ral dementia screening measures. These studi es find that depression is associated with greater impairments in executive functioning and working memory (Kuzis, Sabe, Tiberti, Leiguarda, & Starkstein, 1997; Santangelo et al., 2009; Uekermann et al., 2003)
41 An early study by Starkstein and colleagues ( 1992 ) examined approximately 100 PD patients and categorized them into major depression, minor depression, and non depressed based on psychiatric interview with DSM III criteria (Starkstein, Maybe rg, Leiguarda, Preziosi, & Robinson, 1992) Authors re assessed these patients after one year and reported the di fferences between groups T he PD patients with major depression had significantly: 1) greater decline in MMSE scores than the other two grou ps, 2 ) a longer duration of illness, 3) greater declines in activities of daily living over 1 year, and 4 ) greater impairments in mot or scores Authors concluded that major depression in PD is related to faster pr ogression of PD (Starkste in, Mayberg, Leiguarda et al., 1992) Yet the patients with major depression had longer duration of illness (i.e. more years with PD) than the other two groups. This confound s the conclusions because duration of illness could account for the cognitive and ADL declines Other studi es matched disease and demogra phic variables. Troster and colleagues ( 1995 ) carefully matched 44 depressed PD patients 44 nondepressed PD patients and 44 normal elderly controls on disease variables (PD groups only) and de mographic factors and compared neurocognitive profiles (Troster, Stalp, Paolo, Fields, & Koller, 1995) Results indicated that the depr essed PD group but not the non depressed PD group had poorer performance on phonemic fluency, semantic fluency and confrontation naming than healthy controls Comparing depressed PD and non depressed PD groups revealed worse immediate recall and semant ic fluency in the depressed PD group (Troster et al., 1995) However, authors reported that the DRS scores were significantly lower in t he depressed PD group. They then matched the two groups on overall DRS score using a subsample of 15 depressed and 15 non depressed PD patients and found that t he significant differences between PD groups d isappeared Authors discussed the importance o f matching to general cognitive functioning/dementia level.
42 Several studies have match ed all three: demographics, disease variables, and general cogn itive functioning/dementia measure scores These studies find that depressed PD patients perform more poor ly on measures o f executive functioning than non depressed PD patients. Kuzis and colleagues (1997) included four matched gr oups: MDD without PD depressed PD non depressed PD, and no rmal controls. The MDD and depressed PD patients were more severely im paired than the other two groups on verbal auditory attention (i.e. digit span) and phonemic verbal fluency. Depressed PD patients were worse than all three other groups on abstract visual reasoning (i.e. Ravens Progres sive Matrices) and set shifting/ ment al flexibility (i.e. Wisconsin Card Sorting Task). Depressed PD patients completed fewer categories than the three other groups. Uekermann et al. ( 2003 ) also found that reasoning and verbal fluency were imp aired in depressed PD patients relative to normal controls. This study examined patients early in the course of PD. There were two control groups M DD without PD, and healthy controls Depressed and non depressed PD groups were matched on MMSE score s and all groups were matched on demographics and IQ. The depressed PD patients were depress ed based on a BDI score of > 10 (e.g. DSM criteria was not used to diagnose MDD in the PD group) D epressed PD patients, but not non depressed PD patients were impaired compared to normal controls on alternating sem antic verbal fluency (alternating between first names and vegetabl e names), working memory (i.e. Digit S pan), and abstract verbal concept formation (i.e. S imilarities subtest from the Wechsler Abbreviated Scale of Intelligence) (Uekermann et al., 2003) Similarly, i n a recent study using DSM IV MDD diagnost ic crite ria, depressed PD patients s howed more executive impairment than non depressed PD patients (Santangelo et al., 2009) Depressed PD patients performed worse on the Frontal Assessment Battery, semantic
43 fluency, copy of visuo spatial figures, Stroop color reading, and Str oop interference. Authors suggest that copying the visuospatial figures may be related to executive funct ioning because subjects exhibited poor planning in their copies (Santangelo et al., 2009). Apathy on Cognitive Functioning in PD A pathy, too, may affect cog nitive functioning in PD. Six studies examined the relationship between apathy and c ognitive functioning (Isella et al., 2002; Pluck & Brown, 2002; Starkstein, Mayberg, Preziosi et al., 1992; Zgaljardic et al., 2007) See Table 2 3. These s tudies classified PD patients into high and low apathy gr oups and examine d for differences on cognitive measures. The majority of cognitive measures administered were executive functioning tasks. A few other measures of memory, visuospatial skills, and attention w ere interspersed across several s t udies. Apath y was consistently associated with worse performance on semantic and phonemic fluency (Isella et al., 2002; Pedersen, Alves et al., 2009; Pluck & Brown, 2002; Santangelo et al., 2009; Starkstein, Mayberg, Preziosi et al., 1992; Zgaljardic et al., 2007) There were mixed findings regarding cognitive inhibition, as measured by the Stroop Interference task, and set shifting/mental flexibility, as measured by the Wisconsin Card Sorting Task (WCST) Two studies found co gnitive inhibition was worse in the apathetic group compared to the non apathetic group (Pluck & Brown, 2002; Santangelo et al., 2009) and two studies did not find a difference between groups (Pedersen, Alves et al., 2009; Zgaljardic et al., 2007) S imilar ly, there were mixed findings regarding the WCST; one study reported that apathetic patients completed fewer categories than non apathetic patients (Pluck & Brown, 2002), but another found no differences in number of categories completed (Starkstein et al., 1992). The Executive Interview, Frontal Assessment Battery, and C ambridge Examination of Cognition in the Elderly were all more impaired in apathetic versus nonapathetic patients (Isella et al., 2002; Santangelo et al., 2009;
44 Pluck & Brown, 2002, respectively). Generally, the re appears to be support for a negative association between apathy and executive functioning. Further studies are needed to determine whether certain aspects of executive functioning are co nsistently related to apathy. All studies find a relationship between impaired verbal fluency and higher levels of apathy, but other results are mixed. Several memory, visuospatial, and attention tasks were also given. There were mixed findings regard ing memory /new learning, and visuospatial functioning. Paired word learning was worse in apathetic than nonapathetic patien ts (Starkstein et al., 1992). Yet, v erbal story recall (an Italian normed story) word list learning and recall (California Verbal Learning Test II) and visual figure recall (Rey Osterreith figure r ecall) were not different between groups (Isella et al., 2002; Pedersen et al., 2009) Visuospatial skills, specifically visual object perception ( Silhouettes from the Visual Object and S pace Perception battery) and copying simple and complex figures were impaired in apathetic patients (Pedersen et al., 2009; Santangelo et al., 2009). However, visual space perception (Cube Test from the Visual Object and Space Perception battery) and copy ing of a complex figure (Rey Osterreith figure c opy) were not different across groups (Pedersen et al., 2009; Isella et al., 2002). Attention and working memory, as measured by Digit Span and Spatial Span, were not different between apathetic and nonapath etic groups (Starkstein et al., 1992; Isella et al., 2002; Zgal j ardic et al., 2007) This was a consistent finding acr oss studies that measured attention/working memory Although progress is being made in t he understanding of how apathy a ffects cognitio n, there are several confounding factors in the interpretation of the data. First, three out of six studies included patients with probable dementia. The recommended cut off for dementia in PD based on the MMSE is <26 (Dubois et al., 2007) Starkstein et al. (1992), Isella et al. (2002),
45 and Santangelo et al. (2009) included patients below this threshold. This is problematic for the purposes of examining apathy and cognition because of unequal distribution of dem entia cases across apathetic and non apathetic groups. Based on the MMSE means for each group, there are more individuals with probable dementia in the apathetic group for Starkstein et al. (1992) and Isella et al. (2002). Santangelo et al. (2009) collec ted MMSE scores, but did not compare them across their apathetic/anhedonic group classifications. More dementia in apathetic groups confounds conclusions that apathy is related to impaired cognitive functioning because the impairment could be driven by th e overall deficits in general cognitive functioning in the unequal distribution of individuals with dementia. On the same note, depression was often not controlled for across groups. Depressive symptoms were greater in the apathetic groups versus the n onapathetic groups in two studies (Zgal j ardic et al., 2007; Pederson et al., 2009). Differences in depression across apathetic/anhedonic group classifications in Santangelo et al. (2009) is unknown. As with dementia, greater depression in the apathetic g roups given the relationship between depression and cognitive impairment reviewed earlier clouds conclusions that cognitive defic its are related to apathy Lastly, most studies do not measure multiple cognitive domains in their sample. Mostly, studies ad ministered executive tasks plus a sampling of one or two other domains This likely contributed to the inconsistencies in findings regarding memory/new learning and visuospatial tasks. Studies have d ifferent sample characteristics and use different apath y scales and hence different cut off scores to create their groups. Addressing Current Confounds. The present study aim s to address these confounds by controlling for dementia, depression, and assessing a wide range of cogn itive domains within the
46 same sample. It is critical to control for the effects of depression and other demographic and disease variables to determine if apathy has a unique effect on cognition. T he present study is designed to address the unique contribution of apathy to cognitive f unctioning above and beyond that of other variables. This will help determine whether or not apathy is an important factor influencing cognitive performance in PD.
47 Figure 2 1. Overlap between apathy and depression, apathy alone, and depression alone between groups.
48 Table 2 functioning or the standards of his or her age and culture, as evidenced by all three o f the following: BEHAVIOR 1) Diminished goal directed overt behavior as indicated by: Lack of productivity Lack of effort Lack of time spent in activities of interest Behavioral compli ance or dependency on others to structu re activity Diminished socialization or recreation COGNITION 2) Diminished goal directed cognition as indicated by: Lack of interests, lack of interest in learning new things, lack of interest in new experiences Lack of concern abo problems Diminished importance or value attributed to such goal related domains as socialization, recreation, productivity, initiative, perseverance, curiosity EMOTION 3) Diminished emotio nal concomitants of goal directed behavior as indicated by: Unchanging affect Lack of emotional responsivity to positive or negative events Euphoric or flat affect Absence of excitement or emotional intensity
49 Table 2 2. P roposed consensus criteria for a syndrome of a pathy from the European Psychiatric Association For a diagnosis of Apathy, the patient should fulfill criteria A, B, C, and D A. level of functioning and which is not consistent with his age or culture. These changes in motivation may be reported by the patient himself or by the observation of others. B. Presence of at least one symptom in at least two of the three following doma ins for a period of at least four weeks and present most of the time: Domain B1 : Loss of, or diminished, goal directed behavior as evidenced by at least one of the following: 1) Loss of self initiated behavior (for example: starting conversation, doing b asic tasks of day to day living, seeking social activities, communicating choices 2) Loss of environment stimulated behavior (for example: responding to conversation, participating in social activities) Domain B2 : Loss of, or diminished, goal directed co gnitive activity as evidenced by at least one of the following: 1) Loss of spontaneous ideas and curiousity for routine and new events (i.e. challenging tasks, recent news, social opportunities, personal/family and social affairs. 2) Loss of environment s timulated ideas and curiosity for routine and new events (i.e. Domain B3 : Loss of, or diminished emotion as evidenced by at least one of the following: 1) Loss of spontaneous emotion, observed or se lf reported (for example, subjective feeling of weak or absent emotions, or observations by others of a blunted affect) 2) Loss of emotional responsiveness to positive or negative stimuli or events (for example, observer reports of unchanging affect, or o f little emotional reaction to exciting events, personal loss, serious illness, emotional laden news) C. These symotoms (A B) cause clinically significant impairment in personal, social, occupationa., or other important areas of functioning. D. These sym ptoms (A B) are not exclusively explained or due to physical disabilities (e.g. blindness and loss of hearing), to motor disabilities, to diminished level of consciousness or to the direct physiological effect of a substance (e.g. drug of abuse, a medicati on).
50 Table 2 3. Summary of PD studies examining cognition in apathetic versus nonapathetic groups Study N Apathy Measurement Exec. Functioning Tests S igni ficantly Worse in Apathetic vs. Nonapathetic PD P atients Exec. F unctio ning Tests Similar Between G r oups Starkstein et al., 1992 N = 50 Apathy Scale Letter fluency TMT, Part B WCST categories TMT, Part A Isella et al., 2002 N = 30 Apathy Scale Letter fluency Semantic fluency Executive Interview None Pluck & Brown, 2002 N = 45 Apathy Evaluation Scale Letter fluency Semantic fluency WCST categories Stroop Word, Color, and Interference Cambridge Examination of Cog. in the Elderly WCST perseverative errors Zgaljardic et al., 2007 N = 32 Frontal Systems of Behavior Scale Letter fluency Semantic fluency Stroop Word, Color, and Interference Twenty questions from D KEFS Pedersen et al., 2009 N = 175 Neuropsychiatric Interview Semantic fluency Stroop Interference Santangelo et al., 2009 N = 125 Apathy Evaluation Scale Letter fluency Stroop Interference Fro ntal Assessment Battery Planning aspects of copying 2 D spatial figures Semantic fluency Note: WCST = TMT = Trail Making Test; Wisconsin Card Sorting Test; D KEFS = Delis Kaplan Executive Functioning Scale;
51 CHAPTER 3 SPECIFIC AIMS OF THE PRESENT STU DY The specific aims of the study are twofold. Aim 1 is to empirically validate the distinctness of apathy symptoms from depressive symptoms using theory guided confirmatory factor analysis. Although apathy and depression are related constructs and have overlappi ng symptomatology, studies suggest they may occur independently (Isella et al., 2002; M. L. Levy et al., 1998; Pluck & Brown, 2002) A study from our laboratory found that apathy occurred separately from d epression in 29% (23/80) of PD patients (Kirsch Darrow et al., 2006) A weakness of existing studies is the manner by which patients are classified a s exhibiting 'apathy' or exhibiting 'depression.' T his classification has primarily been based on total scores from apathy and depression self report measures for example the Apathy Scale (Starkstein, Mayberg, Preziosi et al., 1992) and the Beck Depression Inventory II ( Beck, 1996 ). This is problematic since apathy and depression have overlapping symptoms (i.e. anhedonia, lack of interest). Thus, symptoms of apathy may be included as depression inventory total score when they actually represent apathy and vice versa. One of the major aims of this study is to addres s this limitation by using a statistical approach that of confirmatory factor analysis. This approach will examine whether items from these two commonly used mood measures will load onto discrete apathy and depression factors. This is proposed to be a st rong er test of whether apathy and depression are indeed separable in PD. Hypothesis 1: Apathy and depression are related, but distinct constructs that are dissociable in Parkinson disease Prediction 1: item level scores on the Apathy Scale an d Beck Depression Inventory II will load onto 4 factors: 1) an apathy factor representing lo ss of motivation, 2) a dysphoric mo od factor representing sadness and negativity, 3) a loss of interest and anhedonia factor representing the overlap between apath y and depression, and 4) a somatic
52 factor representing bodily complaints (e.g. sleep, appetite, fatigue). Theory guided confirmatory factor analysis will be used to examine the fit of the individual items to the proposed factors. Aim 2 is to investigate the relationship between apathy and cognitive fun ctioning controlling for other important variables such as d epression, disease severity, disease duration, and demographics. Previous studies have reported a negative correlation between apathy and executi ve functioning in PD (Isella et al., 2002; Pluck & Brown, 2002; Starkstein, Mayberg, Preziosi et al., 1992; Zgaljardic et al., 2007) However, several of these studies fail to control for dementia and depression an d assess only a limited range of cognitive functions. This study aims to overcome these limitations by examining apathy in relation to the multiple cognitive domains of processing speed, working memory, language, verbal epis odic memory and executive func tioning. This study also controls for dementia and depression. This will help clarify whether apathy in and of itself is related to impaired cognitive functioning in PD. Hypothesis 2 : Apathy in PD is related to impaired frontal lobe function as eviden ced by poor performance on executive functioning tasks. Prediction 2: Apathy will predict poor functioning specifically in the e xecutive functioning domain. O ther cognitive domains will only be predicted by o ther variables such as depression, demographic s, and disease variables
53 CHAPTER 4 PARTICIPANTS AND METHODS Participants Participant s included one hundred sixty one who underwent a clinical neuropsychological evaluation at the University of Florida Neuropsy chology clinic between the dates of 8/2004 and 5/2009, an d agreed to have their data stored for research purposes. Prior to participating in this study, informed consent was obtained according to university and federal guidelines. To be included, PD pati ents had to be between 40 and 9 0 years of age and meet the United Kingdom Brain Bank d iagnostic criteria for idiopathic PD (Hughes, Ben Shlomo, Daniel, & Lees, 1992; Hughes, Daniel, Kilford, & Lees, 1992) These cri teria are based on the presence of at least two of the four cardinal motor signs of PD: 1) bradykinesia (e.g. slowness of initiation of voluntary movement and reduction in speed of voluntary movement ), 2) muscular rigidity, 3) resting tremor, and 4) postur al instability Further, at le ast one of these sign s must be bradykinesia. Patients must have demonstrated a good response to dopaminergic therapy, as defi ned by a marked improvement in p arkinsonian motor signs assessed by the motor subscore of the Unifi Third Edition (UPDRS III ; Fahn & Elton, 1987 ). The UPDRS III is a standard rating tool designed to assess the severity of motor symptoms over the course of the disease. Demonstrating a positive response to levodopa the rapy is required to exclude patients with Parkinson's plus syndromes (e.g., Shy Drager, Multiple Systems Atrophy, Lewy Body disease, Corticobasal Degeneration). Specific exclusion criteria were: 1) co morbid neurological illness (e.g. stroke, traumatic brain injury, brain tumor, comorbid movement disorder), 2) previous neurosurgical treatments such as deep brain stimulation or pallidotomy, 3) evidence of dementia as assessed by the Dementia Rating Scale II ( DRS II; Jurica, Leitten & Mattis, 2001 ) The D RS II covers the
54 domains of attention, initiation/perseveration, visuoconstruction, conceptualization/reaso ning, and memory. There are 144 possible points. Since one of the purposes of this study is to examine the relationship between apathy, depression and cognition in a non demented PD sample, patients scoring < 130 total DRS II points were excluded from the study. Of the 161 participants, 1 11 were men and 50 were women (s ee Table 4 1). Participants ranged in age from 42 to 84 years ( M = 64.1, SD = 8.7) The majority of patients were Caucasian of non Hispanic origin (95%), with a small number of other races and ethnicities (1 patient was African American, 1 patient was Asian American, a nd 6 patients were Hispanic). On average, PD patients had been experiencing parkinsonian symptoms for eight and a half years (i.e. M = 101.8 months, SD = 54.3 months, range 12 251 months). S everity of motor symptoms based on the UPDRS motor section evaluat ed on levodopa medications was an average of 25.13 ( SD = 8.6 range 9 to 47). Approximately one third of the patients were pre surgical candidates for Deep Brain Stimulation (DBS). In the sampl e as a whole, the average apathy, depression, and anxiety symptoms were as follows: The mean AS score was 10.8 (SD = 6.3 range 0 31), and the mean BDI II score was 9.5 (SD = 7.2; range 0 34). Average anxiety levels were based on age and gender relevant manual norms for the State Trait Anxiety Scale (STAI) (STAI manual, Spielberger, 1968). The mean STAI state percentile w as 56 th %ile (SD = 30, range 5 99 th %ile) and the mean STAI trait percentile was 53 rd %ile (SD = 31, 3 99.9 th %ile). Ninety eight percent of PD patients were taking dopaminergic medications (e.g., levodopa, dopamine agonists) at the time of the evalu ation. Three patients (1.9%) were not taking levodopa because they were in the early stages of disease and the movement disorders physician had decided to wait before beginning levodopa treatment. In all but those three patients, leovodpa equivalent dosag es (LEDs) were calculated to quantify the total amount of levodopa patients
55 were taking. The LEDs were created by converting the dopamine agonists (e.g., bromocriptine, pergolide, ropinirole, pramipexole ) into levodopa equivalent doses using the formula d escribed by Hobson and colleages (Hobson et al., 2002). The result s of this calculation w ere then added to the total amount of regular levodopa that the patients were taking. In the current sample of Parkinson patients, levodopa equivalent dosage was an a verage of 813.9 ( SD = 511.4, range 0 2600). Of the tot al group of patients, 30% were prescribed antidepressants (49 out of 161) and 25% were prescribed anxiolytic medication (40 out of 161) Breaking this down further, 11% were taking both antidepressa nts and anxiolytics (18/161), 19% were taking only antidepressants (31/161), and 14 % were taking only anxiolytics (22/16 1 ). Procedure Overview of Design Two studies were completed, both involving the same participants described above. The first study i nvestig ated whether apathy symptoms could be distinguished from depression symptoms using theory guided confirmatory factor analysis. The second study examined the relationship between apathy and cognitive functioning while controlling for depression, dem ographics, and disease variables. Study 1: Examining Apathy and Depression Factors Specific Aim 1: The purpose of Study 1 is to determine whether apathy and depression items can be distinguished using confirmatory factor analysis. Mood Assessment Meas ures. These include d the Beck Depression Inventory II (BDI II) and the Apathy Scale (AS). The BDI II is a 21 item scale that assesses symptoms of depression experienced over the last two weeks ( Beck, 1996 The BDI II is an updated version of the BDI I (Beck, 1978 ). The BDI II retained the original format of the
56 BDI I (21 items with 0 3 Likert scale) and deleted four items (i .e. body image change, work difficulty, weight loss, somatic preoccupation) and replaced them with four items on agitation, worthlessness, loss of energy, and concentration difficulty. Due to these updates, t he BDI II better reflects the DSM IV diagnost ic criteria ( Arbisi 2001 ) Researchers have not yet repeated reliability studies with the BDI II specifically in PD patients however l iterature has shown tha t the BDI I has excellent reliability and validity in PD patients (e.g. internal consistency reliab ility = .88, test retest reliability = .89, criterion validity with DSM IV depression criteria) (Levin, Llabre, & Weiner, 1988; Visser, Leentjens, Marinus, Stiggelbout, & van Hilten, 2006) The Apathy Scale (AS) i s a 14 item scale measuring cognitive, emotional, and behavioral symptoms of apathy (Starkstein, Mayberg, Preziosi et al., 1992) item ver sion developed by Robert Marin (Marin et al., 1991) The original scale was shortened by 4 items, and wording was simplified by Starkstein et al in 1992. The AS was selected because it has previously been used in studies comparing apathy and depression and has shown good psychometric properties in PD (Internal consistency reliability=.76, test retest one week r = .90). Anxiety Assessment Measur e. To assess anxiety, participants were given the State Tr ait Anxiety Inventory (STAI). The STAI is a 40 item, 1 4 Likert scale measuring state and trait anxiety ( Spielberger, 1970 pres Trait anxiety scale test retest reliability ranges from .65 to .86. The STAI correlates highly with other anxiety measures such as the Taylor
57 Man ifest Anxiety Scale and the Institute of Personality and Ability T est (IPAT) Anxiety Scale (r =.80; r = .75 ). ( STAI manual, Spielberger, 1970 ) Analytic Approach First, the prevalence of apathy and depression was examined in this sample of 161 PD patients. Apathy was defined using the recommended cutpoint (Starkstein, Mayberg, Preziosi et al., 1992) D epression was defin ed using the recommended cutpoint from the BDI II manual 13, mild = 14 19, moderate = 20 28, severe 29). Further, Leentjens and colleagues have also recommended using 14 as a cu tpoint for the original BDI I (Leentjens, Verhey, Luijckx, & Troost, 2 000) We also examin ed the frequency of pure apathy II), pure depression BDI apathy and depression symptoms II). Then, f our groups were created based on these classifications. This resulted in the following groups: a) pure ap athy, b) pure depression, c) mixed apathy depression, and d) neither apathy nor depression [no symptoms]. One way analyses of variance (ANOVAs) were used to analyze groups for differences in age, education disease variables (i.e. UPDRS motor score on lev odopa and months with PD) depression, a nxiety, and LED. G ender, antidep ressant usage, and anxiolytic usage were dichotomous variables, therefore chi sq uare tests were used to evaluate significant differences between groups Next t he reliability of each item from the AS and BDI II was examined using item total sed in the analyses. I tems were analyzed using confirmatory factor analysis (CFA). Table 4 2 shows the hypothesized loadings of each item onto the factors C onfirmatory factor analysis was chosen to examine the factor structure because it allows one to propose a priori hypotheses about the underlying structure of the data and then empirically test and evaluate the proposed factor structure
58 Parameters of Confirmatory Factor Analysis. The CFA was conducted with statistical software AMOS 17.0 using maximum likelihood estimation Before factoring, items were examined for univariate normality (e.g. skewness and kurtosis). Since most items were non normally distributed, item parcels were created instead of factoring raw items. The ration al for this is that CFA has an assumption of multivariate normality. Items on p sychopathology scales were skewed towards the lack of psychopath ology (e.g. 0 or 1) and thus present a non normal, positively skewed, and kurtotic distribution. Parceling helps correct for this. Further, parceling creates fewer indicators requires fewer parameters of estimation and thus provides a better fit (Gorsu ch 1983 ; Little, Cunningham, Shahar, & Widaman, 2002) Models were examined for fit based on the following goodness of fit criteria: minimum fit function chi square, root mean square error of approximation (RMSEA), root mean square residual (RMR), normed fit index (NFI), comparative fit index (CFI), incremental fit index (IFI), r elative fit index (RFI), and the Tucker Lewis Index (TLI). Convention al standards were used to determ ine goodness of fit (e.g. ratio of chi square to degrees of freedom 2:1 or less, RMSEA below .05 and nonsignificant, RMR below .05, and NFI, CFI, IFI, RFI TLI above .9. As is conventional, no single fit index is the primary indicator, but the preponderance of evidence must be in support of the fit of the model. Further, a nested model approach was used to test alternatives to the full four factor model. These included a one factor mod el where ume d apathy, dysphoric mood, loss o f interest, and somatic complaints A two f actor model was also tested, where only factors were represented without loss of interest or somatic factors Finally, 2 three factor models (one wi thout loss of interest/pleasure and one without somatic complaints)
59 were tested. The r esultin g chi squared statistics were tested against the hypothesized four factor model. Study 2: Apathy and Cogn itive Impairment Specific Aim 2. The purpose of Study 2 is to examine the effect of apathy on cognitive functioning, controlling for depression, demographics, and disease variables. It was hypothesized that apathy w ould significantly relate to impaire d e xecutive functioning. However, for other cognitive domains, only depression demographic and disease variables would be related to cognitive functioning. Neuropsychological domains and measures: The PD participants completed cognitive and mood testing during a clinical neuropsychological evaluation that lasted approximately 4 5 hours. This evaluation was done solely for clinical purposes and conducted through the University of Florida Psychology Clinic or through the Department of Neurology. Raw sc ores were converted to age, education, and gender based norms per the tes t manual or Heaton based norms (Heaton, Miller, Taylor, & Grant, 2004) Heaton norms were used for the following measures: Trail Making Test Parts A and B, Boston Naming Test, COWA [ F A S ] letter fl uency, Animal fluency Test manual norms were used for the following measures: Digit Symbol from the Weschler Adult Int elligence Scale III (WAIS III), Vocabulary from the Weschler Abbreviate d Scale of Intelligence (WASI), Logical Memory subtests (I,II) from the Wechsler Memory Scale III Hopkins Verb al Learning Test II form I (HVLT II), J udgment of Line Orientation F acial R ecognition T est Wisconsin Card Sorting Test, and the Stroop Color Word Test. Normative scores from these measure s were converted to z scores using a standardized table. This was done for ease of comparison across tests. Of note, all measures were administered to the ful l sample of PD patients (N = 161 ) with one exception the W isconsin Card Sorting Test (W CST ). O nly 68 of the 161 PD participants had data from the WCST due to
60 the fact that this measure was not incorporated into the battery of tests until 2007. Moreover, the WCST was frequently not administered due to time constraints. D ue to the relatively small numb er of individuals receiving this measure data from the WCST was analyzed separately but It is considered as a test of executive functioning because it is thought to measure mental flexibility, ability to develop abstract concepts, and shift mental sets in response to examiner feedback ( Lezak, Howieson, & Loring, 2004 ) Two dependent variables from the WCST were examined: the total numb er of categories achieved, and the total number of perseverative errors P erseverative errors refer to the incorrect repetition of a response when either a) the sub ject continues to sort according to a previously successful principle that is now incorrect, or b) the subject continues to sort according to an initial incorrect guess. Table 4 3 depicts the cognitive tests used in this study and grou ped into rationally derived domains based on the cognitive proce sses they are thought to tap These domains include : 1) E xecutive functioning 2) Processing speed, 3 ) Verbal episodic memory 4) W orking memory, and 5 ) L anguage functioning In assigning tasks to these domains, the approach of Sheline et al (2006) was followed by grouping tasks into rationally derived domains based on the cognitive processes tapped by each task ( Shelin e et al., 2006). Analytic Approach Hierarchical Multiple Regression C ognitive domains composites were created by averag ing the z scores for each tes t within the proposed domain These cognitive domain composite scores were then used a s the outcome varia bles (i.e. dependent variables) in regression analyses Advantages to using composite scores rather than each individual test include better reliability with multiple measures per construct of interest and fewer overa ll analyses, lowering Type I error rat e. Hierarchical regression was used to determine whether apathy has a significant
61 effect on cognitive domains above and beyond the effects of demog raphics, disease variables, depression and anxiety Predictor variables included demographics (age, gender education), disease variables (UPDRS motor score on levodopa, and months with PD), BDI II score, STAI Group type was defined as the apathy depression symptom classification described above (i.e. pure apat hy, pure depression, mixed apathy depression, or no symptoms). This was used to examine whether group type explained significant variance in any cognitive domain. G roup type was a categorical variable with four levels, therefore dummy variables were crea reference group (i.e. group against which the others were compared). Five regressions, one for each cognitive domain were performed with each cognitive domain as the outcome varia ble. If apathy significantly contri buted unique variance to any domain, that domain was examined in terms of individual tests. Predictor variables were entered simultaneously in blocks: age, education, gender (Block 1) UPDRS motor severity (on levodopa), months with PD (Block 2) BDI II score (Block 3) STAI trait anxiety score (Block 4), AS score (Block 5) group type (Block 6).
62 Table 4 1. Patient characteristics Characteristic PD patients (N=161 ) Age 64 .1 (8.7), range 42 84 Men: Women 111 :50 (68.9 % male) Years of Education 15.1 (2.8), range 7 22 On DOPA meds 98 % On Anti depress ants 30 % On An xiolytics 25 % Levodopa Equivalent Dosage 813.0 (511.4), range 0 2600 Disease Subtype 77 % Tremor Predominant 17.4 % Akinetic /Rigid 1.9 % Postural Instability/Gait 3.7% were eit her missing or not given a subtype diagnosis Months Symptoms 101.8 (5 4.3), range 12 251 M otor score (UPDRS, on levodopa) 2 5.5 (8.6), range 9 47 DRS II Total Score 138.8 (3.5), range 130 144 Apathy Scale 10.8 (6.3), range 0 31 Beck Depression In ventory II 9.5 (7.2), range 0 34 State Trait Anxiety Inventory State Percentile 56 th %ile (SD = 30), range 5 99 th %ile State Trait Anxiety Inventory Trait Percentile 53 rd %ile (SD = 31), range 3 99.9 th %ile Note: N = 161. However, four patients were missing UPDRS on scores, (N = 157), 7 patients were missing anxiety scales (N = 157), and 6 patients did not have a subtype diagnosis (N = 155).
63 Table 4 2. Pro posed loadings of each item. Italic = apathy factor, Underlined = dysphoric mood factor It alic & Underlined = overlap factor of loss of interest/pleasure, Bold = somatic factor Apathy Scale Items Beck Depression Inventory II Items Are you interested in learning new things? I feel sad much of the time. Does anything interest you? I feel more d iscouraged about my future than I used to. Are you concerned about your condition? I have failed more than I should have. Do you put much effort into things? I don t enjoy things as much as I used to. Are you always looking for something to do? I feel g uilty over many things I have done or should have done. Do you have plans and goals for the future? I feel I may be punished. Do you have motivation ? I have lost confidence in myself. Do you have the energy for daily activities? I am more critical of my self than I used to be. Does someone have to tell you what to do each day? I have though t s of killing myself, but would not carry them out. Are you indifferent to things? I cry more than I used to. Are you unconcerned with many things? I feel more restl ess or wound up than usual. Do you need a push to get started on things? I am less interested in other people or things than before. Are you neither happy nor sad, just in between? I find it more difficult to make decisions than usual. Would you conside r yourself apathetic? I don t consider myself as worthwhile and useful as I used to. I sleep somewhat more/less than usual. I have less energy than I used to have. I am more irritable than usual. My appetite is somewhat less/greater than usual. I can't concentrate as well as usual I get more tired or fatigued more easily than usual. I am less interested in sex than I used to be.
64 Table 4 3. Tests categorized into rationally derived cognitive domains. DV = dependent variable Cognitive Doma in Test and Description Executive f unctioning Trail Making Test, Part B a test of psychomoto r speed and mental set shifting; Subject must rapidly alternat e between connecting numbers and letters over a page DV = time (in seconds) to completion ; maximu m time Verbal Fluency a test of speeded word production to either a letter of the alphabet (F, A, S) or to a category (animals), DV 1 = total number of words produced to letters F, A, and S over 60 second trials DV 2 = total number of animals pr oduced in a 60 second period Stroop Color Word Test a measure of cognitive inhibition ; Subject must inhibit automa tic response of word reading and instead name the color of the ink the word is printed, DV = number of correct responses over a 45 second t rial Wisconsin Card Sorting Test a measure of mental flexibility, ability to develop abstract concepts, and shift mental sets in response to examiner feedback ; Subject must sort cards according to a principle [color, form, number] that the subject must d educe from examiner feedback, DV 1 = number of categories achieved, DV 2 = number of perseverative errors Processing s peed Trail Making Test, Part A a test of psychomotor speed ; S ubject must rapidly connect numbers in a series spread throughout the pag e, DV = time to completion Digit Symbol, Wechsler Adult Intelligence Test (WA IS III ) a test of psychomotor speed ; S ubject must rapidly write the correctly paired symbol below each number, DV = number of correctly filled i n symbols period Stroop Color Word Test, Reading Trial a test of speed ed single word words [red, green, blue], DV = number of words read aloud during 45 second trial Verbal episodic memory Logical Memory I and II, Wechsle r Memory Scale III (WMS III) a test of re cent memory for detailed stories read aloud to the subject DV1 = total correctly recalled details immediately after presenta t ion DV2 = total of correctly recalled details after a 30 min delay Hopkins Verbal Lea rning Test II (HVLT II) a test of learning and memory of a list of 12 words that can be organized into three categories, DV1 = total number of c orrectly recalled words over the 3 learning tri a 1 DV2= total correctly recalled words after a 20 min delay Working Memory Digit span forward (WAIS III ) a test of auditory attention that involves repeating a number sequence in correct order, DV = highest number of digits repeated forwards Digits Span backward (WAIS III) a test of auditory working memory that involves repeating number sequences in reverse order, DV = highest number of digits repeated backwards
65 Table 4 3. Continued Cognitive Domain Test and Description L anguage Boston Naming Test (BNT) a 60 item test of visual confrontation naming; D V = total number of correctly named items Vocabulary (Weschler Abbreviated Scale of Intelligence ) a measure of verbal knowledge of word definitions, DV = number of correctly defined words with 1 or 2 point answers depending on degree of elaboration
66 CHAPTER 5 RESULTS Aim 1: Relationship Between Apathy and Depression Frequency of Apathy and Depression Symptoms Initial analyses examined the frequency of mood symptoms in the Parkinson sample. Results indicated tha t 54 out of 161 (i.e. 33.5%) patients had clinically signi fic ant Further, 41 out of the 161 patients (i.e. 25.3%) had clinically significant depressive symptoms BDI II score) The relationship between apathy a nd depression was examined by calculating the number of individuals who exhibited apathy ( i.e. II), those who exhibited ( i.e. mixed apathy and depressio n ( i.e. As shown in Figure 5 1, r esults indicated that 17.4% (28/161 patients) had pure apathy 9.3% (15/161 patients) had pure depression and 16% (26/161) had mixed apathy and depression. Differences Between Apathy Depression Group s on Demographics, Disease Variables, and Medication Usage In order to examine differences between groups with pure apathy, pure depression, mixed apathy depression, and no symptoms, four groupings (as defined with the cutpoints described above) were creat ed. This res ulted in the following groups : a) pure apathy (n = 28), b) pure depression (n = 15), c) mixed apathy depression (n = 26 ), and (d) neither apathy nor depression [no symptoms] (n = 92). The groups were then compared in terms of demographic var iables (age, gender, education), disease variables (disease duration, disease severity), and medication usage (levodopa equivalent dosage, antidepressant usage, anxiolyic usage). Separate one way ANOVAs with Bonferroni post hoc comparisons were used to ex amine differences among groups in age, education disease duration (i.e. months with PD), disease severity (UPDRS motor
67 score on levodopa), and amount of levodopa equivalent dosage (i.e. LED). Chi squared tests were used to evaluate differences across gro ups in gender distributio n, antidepressant usage, and anxiolytic usage. Means, standard deviations, ranges F and 2 statistics, and significance values for each analysis are displayed in T able 5 1. Age. One way ANOVA results indicated a significant effe ct of Group on age, F (3, 157 = 3.35, p = .021). Follow up post hoc analyses indicated that the pure apathy group ( M = 68.07 years, SD = 8.4) was significantly older than the mixed apathy depression group ( M = 61.27 years, SD = 10.22, p = .021). No other groups were significa ntly different from each other ( s ee Table 5 1 ) Gender. Chi square analyses indicated that gender distribution was not significantly different between groups. The percentage of males in each group were as follows: 75% of the apathy only group 73% percent of th e depression only group 69% percent of the mixed a pathy depression group and 66% percen t of the no symptom group The gender distribution was not significantly different among groups, 2 (3, N = 161) = .915, p = .822. Educa tion. One way ANOVA results indicated a significant effect of Group on education, F (3, 157 = 3.23, p = .024). Follow up post hoc analyses indicated that the mixed apathy depression group ( M = 13.58 years, SD = 2 5 ) had significantly less education than the no symptom group ( M = 15.48 years, SD = 2 78 p = 013 ). No other groups were significantly different from each other. Disease severity and duration Of note, seven UPDRS on motor scores were missing so the n was 154 One way ANOVA results suggest ed that there was not a significant effect of Group on among groups, F (3, 153 = 1.76, p = .158). One way ANOVA results indicated no significant
68 effect of Group on disease dur ation as measured by number of months with PD among groups, [ F (3, 157 = 2.10, p = .102 ), ( see Table 5 1 motor scores and number of months of PD for each group )] LED. One way ANOVA results indicated a trend for an effect of Group on LED, F (3, 58.6 = 2 .55 p = .058 ). P ost hoc tests revealed that the pure depression group ( M = 1121 SD = 412.1 ) was taking significantly more levodopa than the pure apathy group ( M = 698.58, SD = 418.09 ), p = .047). No other groups were significantly different from each o ther. Antidepressant usage. Chi square analyses indicated that groups differed significantly in terms of percentage of patients taking antidepressants, 2 (3, N = 161) = 8.46, p = .037. The percentage of patients taking antidepressants were as follows: 25% (freq = 23/92) of the no symptom group, 25% (freq = 7/28) of the pure apathy group, 33.3% (freq = 5/15) of the pure depression group, and 53.8% (freq = 14/26) of the mixed apathy depression group. Further chi squared analyses were undertaken to determine where the significant difference was. Results indicated that the mixed apathy depression group (53.8%, 14/26) was taking significantly more antidepre ssants than the no symptom group (25.0%, 23/92), 2 (1, N = 118) = 7.84, p <.01. The mixed apathy depression group was also taking significantly more antidepressants than the pure apathy group, 2 (1, N = 54) = 4.75, p = .030. No other groups were significantly different from each other Anxiolyti c usage. Chi square analyses indicated that groups did not differ significantly in terms of the percentage of patients taking anxiolytics, 2 (3, N = 161) = 3.48, p = .32. The pe rcentage of patients taking anxiolytic s was as follows: 20.7% (freq = 19/92) of the no symptom group, 25% (freq = 7/28) of the pure apathy group, 26.7% (freq = 4/15) of the pure depression group, and 38.5% (freq = 10/26) of the mixed apathy depression group.
69 Differences Between Apathy Depression Groups on State and Trait Anxiety In order to examine the relationship between apathy, depression, and anxiety, the fo ur groups were compared on their state and trait anxiety p ercentile scores from the STAI (see Table 5 2, and Figures 5 2 and 5 3) Two separate one way ANOVAs were used f or state and trait anxiety. Seven of the 161 patients were missing anxiety scores, resulting in a tot al N of 154 for this analysis. icant, the B rown Forsythe F correction was used. Results i ndicated a significant effect of Group on state anxiety, F (3, 74.82 = 19.96, p <.001), and on trait anxiety, F (3, 58.6 = 43.83, p <.001). For state anxiety, follow up post hoc tests revealed that the pure apathy group ( M = 56 th percentile, SD = 30.35) did not significantly differ from either the no symptom group [( M = 47 th percentile, SD = 28.2), p = .4] or the pure depression group [( M = 68 th percentile, SD = 21.2) p = .5) ] However, the pure apathy group had significantly lower state anxiety scores than the mixed apathy depression group [ ( M = 86 th percentile, SD = 14.5), p <.001] These fin dings are depicted in Figure 5 2 Next, trait anxiety scores were compared across groups in post hoc follow up tests. As shown in Figure 5.3, t he pure apathy group h ad significantly higher trait anxiety ( M = 60 th percentile, SD = 24.69) than the no symptom group [( M = 37 th percentile, SD = 25.05), p =.001], but significantly lower trait anxiety than the mixed apathy depression group [( M = 90 th percentile, SD = 14.30), p < .001]. There was no difference between the pure apathy group and the pure depression group [( M = 76.5 th percentile, SD = 25) p = .26]. Finally, the pure depression group did not differ in trait anxiety scores from the mixed apathy depression group ( p = .27). In addition to looking at the mean differences between groups, it is important to look at the relative proportions of clinically significant anxiety in each group. T he Spielberger manual lists a raw score of 40 as a cut score for individuals wit h high anxiety, however in medically ill
70 patients and geriatric patient populations the cut score with the best sensitivity and specificity has found to be higher, ranging from 41 to 55 ( Stark et al., 2002; Stanley et al., 2001; Kvaal et al., 2005) The c orresponding percentile rank s between these two raw scores range between 81 st 100 th percentile. I n th e present study, 93 rd percentile was chosen as the cutpoint for significant anxiety because it represents a mid point between these two percentile ranges from the literature, and is one and a half standard deviations from average C hi squared statistics were used to examine the difference in the occurrence of elevated trait anxiety across groups using rd percentile as the cut point Results indicated that t he proportion of individuals with elevated trait anxiety scores differed across the apathy depression 2 (N = 154, df = 3) = 64.5, p <.001. As shown in Figure 5.4 the p ercentages of patie nts with elevated trait anxiety scores were as follows: 1.1% (1 out of 89) of the no symptom group, 7.4% (2 out of 27) of the pure apathy group, 30.8% (4 out of 13) of the pure depression group, and 64% (16 out of 25) of the mixed apathy depression group Breaking these down into post hoc comparisons, t he pure apathy group (7.4%) did not signific antly differ from the no symptom significance, used to correct for less than 5 per cell, p = .135). The pure apathy group (7.4%) was lower than the pure depression and s ignificantly lower than the mixed apathy depression group (64.0 %), p <.001 Further, the pure depression group was s ignificantly lower than the mixed apathy d epression group at trend (p = .052). T hese results indicate that the prevalence of high anxiety is greatest in the group with mixed apathy and depression. T he pure apathy group has a low prevalence of high anxiety, and it is not significa ntly different f rom the no symptom group Factor Structure of Apathy and Depression : Confirmatory Fa ctor Analysis In addition to using total scores, a primary aim of this study was to examine the individual items on the apathy and depression scales. It was predicted t hat items from the Apathy Scale
71 (AS) and Beck Depression Inventory II (BDI II) would load onto 4 factors: 1) an apathy factor representing lo ss of motivation, 2) a dysphoric mood factor representing dysphoria and negativity, 3) a loss of interest/pleasure factor representing the overlap between apathy and depression, and 4) a somatic factor representing bodily complaints (e.g. sleep, appetite, fatigue). P rior to conducting a confirmatory factor analysis (CFA) on the AS and BDI II items, these items were first screened for internal consistency reliability and for normality This was done by examining the means, standard deviations, and item total correlations for the 21 items of the BDI II and the 14 items of the AS As is conventional for reliability a nalysis, n egatively worded AS items were reverse sco red. each of the total scales, and in terms of wheth er deleting items improved overall alpha. For the AS, most items positively correlat ed with the total apathy score b etween .4 and .7. The one exception was AS item 3. This item (reverse scored) had a negative correlation with the total score ( r = .14). This item states: The negative correlation with the total apathy score i ndicates that as pati ents endorse higher apathy on this item they score low er in overall t otal apathy. Furthermore, internal consistency reliability item total statistics indicated that deletin g this item improve d internal consistency reliability (i.e. Cr onba This indicates that AS item 3 is a psychometrically poor item. Due to its unreliability, AS item 3 w as not used in the confirmatory factor analysis. For the BDI II, all items positively correlated with the total depres sion score (between .36 .68) and alpha was not im proved by deleting any items. The overall alpha was .89. Items were then checked for normality. Results indicated that i ndividual items on both scales tended to be skewed towards the lack o f psychopathol ogy (positively skewed and positively kurtotic) I tem parcels were creat ed by summing items into pairs This is done
72 pseudorandomly by combining items randomly within hypothesized factors. Item parcels improve normality, an assumption for CFA. Because some item parcels were still non norma l, the data we re transformed reduce the positively skew Confirmatory factor analyses. One hundred forty six participants had complete item data for both th e AS and BDI II Fourteen patients had skipped at least one item on either scale, so their item data was incomplete and was not analyzed. The remaining one hundred forty six were constrained to the hy pothesized 4 factor so lution These factors and indicators were : 1) ( 5 item parcel indicators A7_A10, A9_A4, A5_A14, A11_A6, A12_A13 ), 2) D ysphoric mood (6 item parcel indicators, B6_B8, B3_B17, B9_B2, B1_ Loss of Interest and Pleasure (3 item parcel indicators, B4_B12, A1_A2, B13), and 4 ) (4 item parcel indicators, B16_B18, B21_A8, B15_B20, B19) The overall fit o f the model to the data was: 2 (129, N = 146) = 213.3 p <.01 (NFI = .844 CFI = .931 I FI = .932 RFI = .815 TLI = .918 ). The RM SEA was .067 ( p = .04 ), and the RMR was 011 The overall fit was good in terms of the ratio of chi square to degrees of freedom ratio being less than a ratio of 2:1, fit indic es close to 1, and RMR less than .05 However, the RMSEA was significant and greater than .05. M odification indices were examined to see if there was a single parameter that would greatly improve the fit of the model. Interestingly, it was a c orrelated uniqueness (unexplained variance) am ong two indicators (B6_B8 and B 7_B11) that the modification indices provid ed as improving fit the most These items have to do with feelings of punishment, self criticalness, and crying and guilt. This means that there is unexplained var iance in t llowing the unexplained
73 variance of these indicators to correlate improved the model by 17.05 chi squared points. This improve d the fit to: 2 (128 N = 146) = 194 9 p <.01 (NFI = .8 58 CFI = 945 IFI = .9 46 RFI = 830 TLI = .9 34 ). The RMSEA was .06 0 ( p = .16 ) and the RMR was .011. The c hi square to degrees of freedom ratio is slightly lower (1.5:1 versus 1.65:1 before) and fit indices a re closer to 1; additionally RMSEA is nonsignificant, indicating a better fit. Table 5 3 shows these four factors, items standardized loadings and uniquenesses. All the loadings were high, ranging from .59 .87. For the Apathy/Loss of Motivation factor the loadings ranged from .65 to .87. The highest loadings were for items A7_A A4_A For the Dysphoric Mood factor loadings ranged from .59 .87. The highest loading s were items B7_B Loss of Interest and Pleasure factor ranged from .65 .72, with the neither happy nor sad, just in bet Somatic factor ranged from .63 .70, with the Alternative nested m odels. A nested model approach was used t o test alternatives to the four factor model. This was done to determine if the four factor model has a significantly lower 2 (indicating a better fit) than other alternative models. These nested models were identical in structure to the original model except for the number of fac tors (i.e. they included the correlated uniqueness). The results
74 are summarized in Table 5 4 A one f actor model, subsuming dysphoric mood, apathy, loss of was significantly worse than th 2 2 = 194 .9, p <.01). We also tested two factor factor solution without the Somatic factor, and a three factor solution without the Loss of interest/pleasure factors. All of these models were significantly worse than the four facto r model (p <.01). See Table 5 4 for chi square change and significance statistics. Worse (higher) chi square indicates a greater discrepancy between the original and reproduced correlation matrix, and hence a worse fit to the data. Thus the four factor model separating the constructs of apathy, depression, loss of interest/ anhedo nia, and somatic complaints was support ed. This lends support to the hypothesis that apathy an d depression are separable constructs in PD. Further studies will be import ant to disentangle the discrete effects of apathy versus depression. One such area involves examining the separate effects of apathy and depression on cognitive functioning. The next part of this study sought to do this by investigating the unique effect s of apathy on cognition in PD. An important aspect of this aim was to control for demographics (age, gender, education), disease variables (duration, severity), and depression in order to understand the unique effects of apathy on cognitive functioni ng. A im 2: Relationship between Apathy and Cognition The second overarching aim of the present study was to investigate the relationship between apathy and cognitive functioning in PD. It was hypothesized that a pathy is related to impaired frontal lobe functi oning and that deficits would be evidenced on tasks of executive functioning Specifically, it was predicted that apa thy would be related to poor performance on the executive functioning domain, while o ther cognitive domains (e.g. episodic memory, etc)
75 w ould only b e predicted by other variables such as demographics disease variables, and depression First, p equivalents using a standardized table. See Table 5 5 for means and st andard deviations of each cognitive measure Then composite scores were created by averaging the z scores across tasks in each rationally derived domain. These domains consisted of: Executive functioning (Trail Making Test Part B, letter fluency, ani mal fluency, and Stroop Color Word ), Processing speed (Trail Making Test Part A, Digit Symbol, Stroop Color Naming ), Episodic memory (HVLT total, HVLT delay, WMS Logical Memory I, Logical Memory II), Working memory (Digit s pan forwards, Digit s pan backwa rds), and Language processing (Boston Naming Test, Vocabulary). See Table 4 3 for a summary of each test and the associated domain The m eans, standard deviations, and ranges of each domain are presented in Table 5 6 Correlations between cognitive doma ins are presented in Table 5 7 The correlation between each and every cognitive domains were significant ( p <.001). The lowest correlation was between Working memory and Executive functioning ( r = .254) and the highest correlation was between Processing speed and Executive functioning ( r = .659). Cognitive Domain Regression Analyses Next, the cognitive domains were entered as dependent variables in regression analyses Predictors were regressed upon each cognitive domains using hierarchical multiple reg ression. Five hierarchical regressions, one for each cognitive domain, were performed with simultaneous entry of variables in sequential blocks. Demographic variables were entered on Block 1 (age, gender, education). Disease variables were entered on Bl ock 2 (duration of disease, disease severity). Depres sion (BDI II score) was entered on Block 3, and anxiety (STAI trait percentile) was entered on Block 4. Apathy score (AS score) was entered on Block 5. The dummy coded
76 variable of group type (i.e. pur e apathy, pure depression, mixed apathy depression, no symptoms) was entered on Block 6. Multicollinearity statistics of tolerance and Variance Inflation Factors (VIF s ) are presented in Table 5 8 The VIFs for the models range from 1.0 5.0. The VIF ind icates whether the predictor has a strong linear relationship with the other predictors. Values greater than 10 indicate concern that multicollinearity is biasing the models ( Fields, 2005 ). The tolerance statistics range from .20 .99. Values below .20 i ndicate concern that multicollinearity is biasing the models (Fields, 2005). Results of hierarchical multiple regression analyses are provided in Tables 5 9 through 5 15 and include total model R 2 change in R 2 for each set of predictors, b values, standa rd errors, and betas Executive functioning. First, the Executive functioning domain was examined. The final model, with all predictors included (demographics, disease variables, depression, anxiety, a pathy and group type ) explained 13.8 % of th e variance in Executive functioning ( p = .03 1 see Table 5 9 ). Breaking this down by set of predictors and significance of R 2 change statistics the set of demographic variables did not cont ribute significant variance to E xecutive functioning ( p = .6 9). T he set of disease variables did not explain significant variance but was at trend ( R 2 change = 3.5%, p =.07 3 ). However, neither severity of PD ( p = .108), nor duration of PD ( p = .426) had significant s in the final model. A pathy contributed 3.5 % of the variance in Executive functioning ( p =.02 ) Apathy and Executive functioning were inversely related ( = .369, p =.012). Neither anxiety ( p =.145) nor d epression ( p =.273) contribute d significant va riance to Executive functioning Additionally, group type did not explain significant variance ( p = .135).
77 Thus, as predicted, apathy contributed an incremental in crease in variance explained in Executive functioning with increased apathy related to dec rease d performance in the E xecutive functioning domain Next, the remaining cognitive domains were examined. The se included: Processing speed, Verbal e pisodic memory, Working memory, and Language processing. It was predicted that apathy would not add i ncremental variance explained in these cognitive areas. Processing speed. The final model, with all predictors included, significantly explained 21.8 % of the variance in Processing speed ( p <. 0 01, see Table 5 10 ). Breaking this down by predictors, the se t of demogra phic variables accounted for 7.4 % of the variance ( p = .011 ) This variance was completely contributed by gender, with males performing more poorly than females ( = 209 p = .008 ). The set of disease variabl es contributed an additional 5 9 % of variance ( p = .0 0 8 ) S everity of PD was inversely related to Processing speed ( = .175, p =.036 ), while durati on of PD was not related ( p = .2 2 ) D epression explained an additional 2.4% of variance in Processing speed ( p = 048 ) Depression was inversely related to Processing speed ( = .414, p = .014). Neither anxiety ( p = .255) nor apathy ( p = .939) contributed significant variance to Processing speed. Group type explained 5.3% of variance in Processing speed ( p = .027). This was driven by a trend for the no symptom group to score better i n Processing speed than the depression only group ( = .184, p = .065). Verbal episodic memory. The final model explained 2 4.5 % of the variance in V erbal episodic memory ( p <.001, see Table 5 11 ). The set of demograp hic variables explained 16 .4 % of the variance in Verbal episodic memory ( p <.001). Gender signi ficantly predicted Verbal episodic memory, with males performing m ore poorly than females ( = .274 p <.01).
78 Education also significantly predicted list learning memory, with higher education related to better list learning scores ( = .272 p = .001). There was a trend for the set of disease variables to add incremental variance ( R 2 chang e = 2.8%, p =.09). In the final model, disease severity was significantly and inversely related to Verbal episodic memory ( = .163 p = .048). Disease duration was not significantly related ( p = .653). There was also a trend for depression to contribut e variance ( R 2 change = 1.6%, p =.09) was not significant in the final model ( p = .829). Neither anxiety ( p = .47), a pathy ( p = .84), or group type ( p = .11) explained significant variance in Verbal episodic memory Working me mory. The final model did not explain significant variance in W orking memory performance ( p = .204, see Table 5 12 ). However, in earlier models, education contributed significant variance to Working memory. Education was the only predictor that acco unte d for significant incremental variance (5.6%, p = .038 ). Education was positively related to working memory ( =.190, p = .027 ) D isease variables ( p = .37 ) depression ( p = .14 ) anxiety, ( p = .36), apathy ( p = .56 ) and group type ( p = .81) did not e xplain significant variance in W orking memory. Language functioning. The overall model explained 16.1 % of the v ariance in L anguage functioning ( p <.01, see Table 5 13 ). Demographic variables contributed a significant 5 .2% ( p = .049). Education was positively related to L anguage functioning ( = .197, p =.018 ). The re was a trend for the disease variables to contr ibute variance ( R 2 change = 3.8%, p =.054) This was driven by a significant inverse relationship between disease severity and Language functioning ( = .192 p = .027). Disease duration was not significantly related to Language functioning (p = .254).
79 Depression contributed a significant variance of 4.0% to Language functioning ( p = .011 ) However, its weight was insignif icant in the final model ( p = .8 ). Anxiety contributed significant variance at trend ( R 2 change = 2.0%, p =.072) The weight for anxiety was at trend in the final model, with anxiety inversely related to Language functioning ( = .201, p =.073). Apathy ( p = .318) and group type ( p = .86) did not contribute significant variance to Language functioning. Summary of Regression R esults It was predicted that the Executive functioning domain would be related to apathy over and above the effects of demog raphics, disease variables, depression and anxiety This prediction was supported, as increasing apathy was related to impaired E xecutive functioning. Next, it was predicted that the other cognitive domains (Processing speed, Verbal episodic memory, Working memory, Language functioning) would be related to demog raphics, disease variables, depression and anxiety only. The effect of apathy was predic ted to be specific to E xecu tive functioning. This was also supported. A pathy did not contribute incremental variance to Processing speed, Ver bal episodic memory, Working memory, or Language functioning. idual Executive Functioning In order to fully understand the relationship between apathy and E xe cutive functioning the composite domains were broken down into specific tests. For the Executive functioning domain, four hierarchical regressions were compl eted one each for Trail Making Test Part B, letter fluency, animal fluency, and the Stroop color word test. Executive Functioning Tests The same predictors (demographics, disease variables, depression, anxiety, apathy group type ) we re entered in Blocks as above. For the Trail Making Test Part B the final model explained 14.9 % of t he variance ( p <.0 1 9 ). Apathy did not contribute significant variance to the test ( p = .58). Group type also did not contribute significant
80 variance ( p = .53). However, th e set of disease varia bles contributed a significant 6.8% of the variance ( p < 01). Duration of PD was inversely relat ed to performance on the test ( = .197, p =.0 1 9 ), but disease severity was not related ( p = .28 ). Demographics ( p = .10), depressi on ( p = .19 ) and anxiety ( p = .154) did not contribute significant variance to Trail Making Test Part B. Next, speeded verbal fluency measures were examined. For letter fluency, the final model with all the predictors included did not explain signific a nt variance ( p =.59 ). Apathy did not contribute significant variance to letter fluency. In fact, no variable examined contributed significant incremental variance to letter fluency. Group type did not contribute variance ( p = .5). Demographics ( p = .59), disease variables ( p = .46), depression ( p = 6 3 ) and anxiety ( p = .4) did not cont ribute significant variance to letter fluency. For animal fluency the final model explained 11 .9 % of the variance at trend (p =.091 ). There was a trend for a pathy to ex plain variance in animal fluency ( R 2 change = 2.3%, p =.063). Apathy was inversely related to animal fluency ( = .320, p = .032). Group type also explained variance in animal fluency at trend ( R 2 change = 4.2%, p =.097) This was driven by the no symptom group scoring significantly better on animal fluency than the mixed apathy depression group ( = .435, p = .017). Demographics ( p = .725), d isease variables ( p = .152), depression ( p = .152), and anxiety ( p = .470) did not explain significant variance in animal fluency. Stroop Color Word interference condition, was also examined. The final mo del significant ly 12.6 % of the variance at trend ( p =.069 ). Apathy explained a significant 5.5 % o f the variance in Stroop Color Word ( p < .0 1 ), and was inversely related ( = .433, p < .0 1 ). No other variables explained significant incremental variance Demographic variables ( p = .28), disease variables ( p = .33 ), depression ( p = .41), anxiety ( p = .16), and group type ( p = .64) did not contribute significant variance in Stroop Color Word
81 In summary, apathy contributed signif icant variance to Stroop Color Word. Apathy contributed variance at trend to animal fluency. However, apathy did not contribute variance to letter fluency. Apathy also did not contribute significa nt variance to Trail Making Test Part B. Thus, apathy was related to some aspects of execut ive functioning such as cognitive interference and semantic verbal fluency, but was not related to phonemic verbal fluency, or psychomotor speed and switching menta l sets A subsample of the total of Parkinson patients ( 65 of 161 Ss ) was also administered the Wisconsin Card Sort Test (WCST). This test was incorporated into the battery in 2007 (sample collection be gan in 2004). This variable was not used in the domain specific regressions analyses due to the small sample size. However, because a primary aim of this study was to examine the relationship between executive functioning and apathy regression analyses were carried out in this limited sample. Of note, there were no significant differences between the subjects who received the Wis consin versus those who did not Furthermore, demographic, disease, and apathy, depression, and anxiety characteristics of th e samples that were administered the WCST versus not administered the WCST were compared They did not differ on any variable. Using the same methodology of hierarchical linear regressions, the dependent variables of number of categories and number of p erseverative errors (PE) were regressed upon predictors (e.g. demographics, disease variables, depression, apathy). Raw scores were used due to the limited range of the scores once they were normed (e.g most people get a score of >16 th percentile for nu mber of categories ). The descriptive statistics were as follows: number of raw categories achieved ( M = 3.31 SD = 2.1 range 0 6), number of perseverative errors ( M = 24.6 SD = 13.6, range 4 57).
82 Hierarchical r egression results indicated that the fina l model did not explain variance in number of categories achieved ( p = .40, see Table 5 14 ) However, demographics contributed significant increm ental variance in number of categories achieved ( R 2 change = 14 9 %, p =.0 19 ). The variance was completely due to age. Age was inversely related to the number of categories achieved ( = .314, p = .049). No other demographic variable was related. Disease variables ( p = .83), depression ( p = .81), anxiety ( p = .93), apathy ( p = .62 ) and group type ( p = .69) did not explain significant variance in number of categories achieved. Furth er, raw number of perseverative errors was also examined Regress ion results indicated that the final model did not explain significant variance in number of perseverative errors ( p = .14) Similar to categories achieved, significant incremental variance was contributed by demographic variables ( p < .01). Older age was related to more perseverative errors ( = .318, p = .046). More education was related to fewer perseverative errors ( = .363, p <.01). The regression model and parameters are shown in Table 5 15 ; as before, d isease variables (p=.49 ) depressio n ( p = .87) anxiety ( p = .82), apathy ( p = .79 ), and group type ( p = .68) did not explain variance in number of perseverative errors Interestingly, apathy was not related to WCST performance, either with regards to total categories achieved or in regards to number of perseverative errors. This is counter to the a priori prediction Nor was apathy related to the Trail Making Test Part B, another task that involves Summary of Hierarchical Regression Results Aim 2 of the current study was to examine the relationship betwe en apathy and cognitive functioning in Parkinson disease. H ierarc hical regressions were used to quantify whether apathy had a unique and incremental effect of explaining variance in specific cognitive domains of Executive functioning, Processing speed, Ver bal episodic memory, Working memory and
83 Language functioning. Apathy was predicted to be specifically related to Executive functioning and not to other domains. This was supported. Apathy explained incremental variance in Executive functioning. Apathy explained 5.5% of the variance in Executive functioning. Apathy did not explain incremental variance in Processing speed, Verbal episodic memory, Working memory, or Language functioning. The relationship between apathy and Executive functioning was furth er examined by turning to individual test performance. Apathy significantly explained variance in Stroop Color Word performance. Apathy explained variance at trend in animal fluency performance. However, apathy did not explain significant variance in le tter fluency, Trail Making Test Part B, Wisconsin Card Sorting Test number of categories achieved, or the Wisconsin Card Sorting Test number of perseverative errors.
84 Figure 5 1. Prevalence of apathy and depression in 161 Parkinson patients
85 Table 5 1 Demographic, disease variable, and medication average scores, standard deviations, and ranges between no symptom, pure apathy, pure depression, and mixed apathy depression groups Characteristic No symptoms Pure Apathy Pure Depression Both F or 2 value (df = 3, 157) p value N = 92 N = 28 N = 15 N = 26 Age 64.05 (8.38), range 43 80 68.0 (8.4), range 47 84 61.67 (6.76), range 51 77 61.27 (10.2), range 42 84 3.35 .021 Men: Women 61:31 (66.3% male) 21:7 (75% male) 11:4 (73.3% male) 1 8:8 (69.2% male) .915 .822 Y rs of Education 15.48 (2.78), range 12 21 15.32 (2.80), range 11 20 15.33 (3.33), range 12 22 13.58 (2.47), range 7 18 3.23 .024 % on A nti depressants 25.0% 25.0% 33.3% 53.8% 8.45 .037 % on Anxiolytics 20.7% 25.0% 26.7% 38. 5% 3.48 .324 Levodopa Equivalent Dosage 784.4 (525.99), range 0 2600 698.56 (418.09), range 150 1875 1121.72 (412.56), range 434 2100 865.06 (552.39), range 288 2025 2.55 .058 Months of Symptoms 94.8 (50.03), range 12 228 110.75 (65.11), range 24 240 129 .6 (48.27), range 54 251 101.27 (57.31), range 12 241 2.10 .102 M otor score (UPDRS, on levodopa) 24.51 (8.69), range 9 47 28.0 (8.09), range 13 46 22.29 (7.85), range 10 35 24.99 (8.43), range 13 42 1.76 .158 Note: N = 161. However, five patients wer e missing UPDRS on scores, (N = 157 ) Results are presented as means (standard deviations) and ranges.
86 Table 5 2 STAI State and Trait average scores, standard deviations, and ranges between no symptom, pure apathy, pure depression, and mixed apathy depr ession groups Characteristi c No symptoms Pure Apathy Pure Depression Both Browne Forsythe F 2 value p value N = 89 N = 28 N = 13 N = 25 STAI state percentile 46 th %ile (28.20), range 5 97 56 th %ile (30.39), range 11 99 68 th %ile (21.2), r ange 28 97 87 th %ile (14.54), range 39 99 19.96 <.001 STAI trait percentile 37 th %ile (25.1), range 3 96 61 st %ile (24.69), range 6 97 76 th %ile (24.97), range 14 98 91 th %ile (14.3), range 37 100 43.83 <.001 Percentage of patients > 93%i le on STAI trait 1.1% (1/89) 7.4% (2/28) 30.8% (4/13) 64.0% (16/25) 64.5 <.001 Note: Seven patients wer e missing anxiety scores, (N = 154) Results are presented as means (standard deviations) and ranges.
87 Figure 5 2. Mean State Anxiety Scores a cross Apathy Depression sub groups
88 Figure 5 3. Mean Trait A nxiety scores across A pathy D epression subgroups
89 Figure 5 4. Percentages of patients with clinically elevated trait anxiety across Apathy D epression sub groups Indicates significant diff erences, + indicates trends
90 Table 5 3 Confirmatory factor a nalysis loadings and uniquenesses Factor Items Loading Uniqueness Apathy A7_A10 .870 .242 A4 _A 9 .758 .425 A14_A5 .646 .582 A11_A6 .687 .528 A12_A13 .724 .476 Dysphoric B6_B8 .668 .553 Mood B3_B17 .770 .407 B9_B2 .671 .549 B1_B14 .794 .369 B5_B10 .585 .658 B7_B11 .868 .247 Loss of Interest / Pleasure B4_B12 .761 .420 A1_A2 .453 .795 B13 .650 .577 Somatic B19 .701 .508 B15_B20 .640 .591 B21_A8 .671 .550 B16_B18 .630 603
91 Table 5 4. Factor c orrelations Factor Apathy Dysphoric mood Loss of interest/pleasure Somatic complaints Apathy -.526 .801 .735 Dysphoric mood .526 -.895 .584 Loss of interest/pleasure .801 .895 -.870 Somatic complaints .735 .584 .870 Table 5 5 Goodness of f it s tatistics for c onfirmatory f actor a nalysis of f ull f our f actor model and a lternative n ested m odels Model 2 df 2 df p difference Four factor 194.9 128 ---One factor 433.3 134 238.3 6 p <.01 Two factor 278.9 133 84 5 p <.01 Three factor, without somatic 295.2 132 100.2 4 p <.01 Three factor, without loss of interest 212.8 132 17.81 4 p <.01 Note: p values indicate the chi square difference between the alternative models and the four factor model, indicating si gnificantly worse fit for all alternative models.
92 Table 5 6 Descriptive statistics for individual cognitive tests Domains and Cognitive tests Mean (SD) Range N Executive functioning Trail Making Test Part B .706 (1.15) 3.00 2.55 159 Let ter fluency .298 ( 1.07 ) 2.80 2 40 158 Animal fluency .115 ( 1.12 ) 3.00 3 00 158 Stroop Color Word .271(1.05 ) 2.80 3 0 0 155 Processing speed Trail Making Test Part A .579(1.13) 3.00 2.80 161 Digit Symbol .280(.935) 2.35 2.00 158 Stroop Color .785(.77) 2.80 1.70 156 Verbal episodic memory Logical memory I (total) .150(1.04) 2.35 2.35 160 Logical memory II (delay) .354 (1.03) 2.65 2.65 161 HVLT II total .632 (1.15) 3.00 1.70 161 HV LT II delay .768 (1.31) 3.00 1.30 161 Working memory Digit span forwards .208(.89) 1.85 2.35 161 Digit span backwards .183 (.92) 1.70 3.00 161 Language functioning Vocabulary .609 (.79) 1.64 2.33 153 Boston Naming Test .389 (1.01) 2.10 3.00 159 Table 5 7 Descriptiv e statistics for cognitive domains used in hierarchical regressions ( z score metric ) Cognitive domain Mean (SD) Range N Executive functioning .304 (.769 ) 2.40 1.88 15 0 Processing speed .5 1 4 (.740 ) 2.6 0 1. 67 150 Verbal episodic memory .192 ( .882 ) 2.53 1. 90 150 Working memory .20 5 (.781 ) 1.5 5 2. 68 150 Language functioning .527 (.738 ) 1.72 2. 50 150
93 Table 5 8 Correlations among cognitive domains (Pearson correlations) All cogn itive domains were significantly correlated (p <.001). Cognitive domain Exec. functioning Processing speed Verbal episodic memory Working memory Language functioning Exec. function ing 659 534 254 462 Processing speed 659 .437 .271 .317 Verba l episodic memory 534 .437 .289 .431 Working memory 254 .271 .289 .328 Language functioning 462 .317 .431 .328
94 T able 5 9 Multicollinear ity statistics for hierarchical multiple regressions Model Predictor VIF T olerance 1 Age 1.010 .990 Gender 1.019 .981 Education 1.029 .972 2 Age 1.132 .884 Gender 1.039 .963 Education 1.036 .965 Months PD 1.039 .963 UPDRS on 1.168 .856 3 Age 1.164 .859 Gender 1.051 .951 Education 1.065 .939 Months PD 1.051 952 UPDRS on 1.168 .856 BDI II 1.090 .917 4 Age 1.165 .858 Gender 1.051 .951 Education 1.074 .931 Months PD 1.057 .946 UPDRS on 1.168 .856 BDI II 1.905 .525 STAI T 1.807 .553 5 Age 1.210 .826 Gender 1.052 .951 Education 1.092 .916 M onths PD 1.062 .941 UPDRS on 1.188 .842 BDI II 2.494 .401 STAI T 1.851 .540 AS 1.876 .533 6 Age 1.218 .821 Gender 1.057 .946 Education 1.106 .905 Months PD 1.096 .912 UPDRS on 1.210 .826 BDI II 4.897 .204
95 Table 5 9 Continued Model Pr edictor VIF Tolerance STAI T 2.034 .492 AS 3.391 .295 Group: Pure apathy 2.079 .481 Group: Pure depression 1.733 .577 Group: Mixed apathy depression 5.009 .200 Note: VIF = Variance Inflation Factor
96 T able 5 10 H ierarchical multiple regre ssion results showing the relationship between predictors and E xecutive functioning Model Variance explained Predictor B SE B p value 1 Total R 2 = .01 0 (Constant) .023 .553 R 2 = .010 Age .004 .007 .052 .533 p = .686 Gender .146 .137 .088 .291 Education .004 .023 .014 .864 2 Total R 2 = .0 45 (Constant) .237 .568 R 2 = .035 Age .0 01 .008 .013 .884 p = .0 73 + Gender .106 .137 .064 .441 Education .007 .023 .027 .749 UPDRS on .016 .009 .159 .073 + Months PD .001 .001 .093 .265 3 Total R 2 = .0 53 (Constant) .456 .601 R 2 = .0 08 Age .003 .008 .02 9 .744 p = 273 Gender .089 .138 .054 .517 Education .003 .023 .011 .892 UPDRS on .015 .009 .158 075 + Months PD .001 .001 .083 .321 BDI II .010 .009 .093 .273 4 Total R 2 = .068 (Constant) .479 .599 R 2 = 0 14 Age .002 .008 .025 .777 p =.145 Gender .093 .137 .056 .499 Education .006 .023 .022 .792 UPDRS on .015 .009 .158 .073 + Months PD .001 .001 .073 .379 BDI II .001 .012 .014 .903 STAI T .004 .003 .160 .145 5 Total R 2 = .103, (Constant) .460 .590 R 2 = 035 Age .001 .008 .015 .863 p =.020 Gender .100 .135 .061 .460 Education .000 .023 .003 .969 UPDRS on .013 .009 .131 .134 Months PD .001 .001 .087 .292 BDI II .017 .014 .158 .211 STAI T .003 .003 .120 .270 A S .031 .013 .258 .020
97 Table 5 10 Continued Model Variance explained Predictor B SE B p value 6 Total R 2 = .138, (Constant) .390 .747 .549 .390 R 2 = 035 Age .828 .108 .426 .747 p = .135 Gender .696 .215 .012 .549 Educ ation .614 .533 .030 .828 UPDRS on .390 .747 .549 .108 Months PD .828 .108 .426 .426 BDI II .696 .215 .012 .696 STAI T .614 .533 .030 .215 AS .390 .747 .549 .012 Group: Pure apathy vs. No sym .828 .108 .426 .614 Group: Pure depression vs. No sym .696 .215 .012 .533 Group: Mixed apathy depression vs. No sym .614 .533 .030 .030* Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indica tes significant at p <.1
98 Table 5 11 Hierarchical multiple regression results, showing the relationship between predictors and P rocessing speed Model Variance explained Predictor B SE B p value 1 Total R 2 = .0 74 (Constant) .166 .515 R 2 = .074 Age .000 .007 .004 .960 p = 011* Gender .426 .128 .268 .001 Education .005 .021 .019 .813 2 Total R 2 = 133 (Constant) .121 .521 R 2 = .059 Age .004 .007 .051 .540 p = .008* Gender .378 .126 .238 .003 Education .001 .021 .004 .960 UPDRS on .018 .008 .196 .021 Months PD .002 .001 .133 .096 + 3 Total R 2 = 157 (Constant) .481 .547 R 2 = .0 24 Age .002 .007 .023 .782 p = 048* Gender .351 .125 .221 .006 Education .008 .021 .030 .705 UPDRS on .018 .008 .195 .020 Months PD .002 .001 .116 .143 BDI II .017 .008 .160 .048 4 Total R 2 = .164 (Constant) .498 .546 R 2 = 0 08 Age .002 .007 .026 .754 p =. 255 Gender .353 .125 .222 .005 Education .006 .021 .022 .782 UPDRS on .018 .008 .195 .020 Months PD .001 .001 .109 .170 BDI II .008 .011 .081 .446 STAI T .003 .002 .118 .255 5 Total R 2 = .164 (Constant) .497 .548 R 2 = 0 00 Age .002 .007 .027 .748 p =. 939 Gender .354 .126 .222 .006 Education .006 .021 .023 .777 UPDRS on .018 .008 .194 .022 Months PD .001 .001 .109 .170 BDI II .008 .013 .076 .531 STAI T .003 .002 .117 .267 AS .000 .012 .008 939
99 Table 5 11 Continued Model Variance explained Predictor B SE B p value 6 Total R 2 = .218, (Constant) .402 .542 R 2 = 053 Age .003 .007 .040 .627 p = .027 Gender .333 .123 .209 .008 Education .003 .021 .009 .905 UPDRS on .016 .008 .175 .036 Months PD .001 .001 .097 .22 2 BDI II .043 .017 .414 .014 STAI T .002 .003 .106 .325 AS .011 .016 .098 .482 Group: Pure apathy vs. No sym .293 .215 .148 .174 Group: Pure depression vs. No sym .501 .269 .184 .065 + Group: Mixed apathy depression vs. No sym .523 .339 .260 .125 Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
100 Table 5 12 Hierarchical multiple regression results, s howing the relationship between predictors and Verbal episodic memory Model Variance explained Predictor B SE B p value 1 Total R 2 = .164 (Constant) 1.687 .583 R 2 = .164, Age .009 .008 .094 .218 p < .001* Gender .595 .145 .314 <. 00 1* Education .086 .024 .273 .001 2 Total R 2 = .192 (Constant) 1.475 .600 R 2 = .028, Age .013 .008 .130 .106 p = .0 88+ Gender .554 .145 .293 <.001 Education .089 .024 .284 <.001 UPDRS on .016 .009 .143 .0 80 + Months PD .001 .001 .080 .299 3 Total R 2 = .208 (Constant) 1.116 .631 R 2 = .0 16, Age .011 .008 .107 .187 p = .088+ Gender .527 .145 .278 <.001* Education .083 .024 .262 .001* UPDRS on .016 .009 .141 .081+ M onths PD .001 .001 .066 .390 BDI II .016 .010 .133 .088+ 4 Total R 2 = .211 (Constant) 1.104 .632 R 2 = 003 Age .011 .008 .108 .180 p =.468 Gender .529 .145 .279 <.001* Education .084 .024 .267 .001* UPDRS on .016 .009 .141 .081+ Months PD .000 .001 .061 .425 BDI II .010 .013 .084 .413 STAI T .002 .003 .073 .468 5 Total R 2 = .211 (Constant) 1.106 .635 R 2 = 000 Age .011 .008 .112 .177 p =.842 Gender .530 .145 .280 <.001* Education .084 .025 .265 .001* UPDRS on .016 .009 .139 .090+ Months PD .001 .001 .062 .419 BDI II .009 .015 .073 .538 STAI T .002 .003 .070 .493 AS .003 .014 .020 .842
101 Table 5 12 Continued Model Variance explained Predictor B SE B p value 6 Total R 2 = .245 (Constant) 1.097 .635 R 2 = 034 Age .012 .008 .121 140 p = .109 Gender .519 .144 .274 <.001 Education .086 .024 .272 .001 UPDRS on .018 .009 .163 048 Months PD .000 .001 .035 .653 BDI II .004 .020 .035 .829 STAI T .001 .003 .049 .642 AS .018 .019 .132 .333 Group: Pure apathy vs. No sym .066 .252 .028 .793 Group: Pure depression vs. No sym .567 .315 .175 .074+ Group: Mixed apathy depression vs. No sym .208 .397 .087 .602 Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
102 Table 5 13 Hierarchical multiple regression results, sh owing the relationship between predictors and Working Memory Model Variance E xplained Predictor B SE B p value 1 Total R 2 = .056 (Constant) .699 .549 .205 R 2 = .056, Age .003 .007 .031 .697 p = .038* Gender .225 .136 .134 .101 Education .058 .023 .208 .012 2 Total R 2 = .069 (Constant) .817 .570 .154 R 2 = .013, Age .006 .008 .069 .418 p = .371 Gender .207 .138 .123 .135 Education .060 .023 .217 .009 UPDRS on .010 .009 .105 .229 Mon ths PD .001 .001 .076 .356 3 Total R 2 = .083 (Constant) .521 .601 .388 R 2 = .0 14, Age .004 .008 .048 .580 p = .138 Gender .185 .138 .110 .182 Education .055 .023 .197 .019 UPDRS on .010 .009 .104 .232 Months PD .001 .001 .089 .281 BDI II .014 .009 .125 .138 4 Total R 2 = .089, (Constant) .506 .602 .402 R 2 = 006 Age .004 .008 .050 .561 p =.356 Gender .187 .138 .111 .177 Education .057 .023 .203 .016 UPDRS on .010 .009 .104 .23 2 Months PD .001 .001 .095 .252 BDI II .006 .012 .058 .603 STAI T .002 .003 .100 .356 5 Total R 2 = .091, (Constant) .511 .603 .398 R 2 = 002 Age .005 .008 .060 .495 p =.559 Gender .189 .138 .112 .174 Education .0 55 .023 .197 .020 UPDRS on .010 .009 .097 .269 Months PD .001 .001 .091 .271 BDI II .002 .014 .022 .865 STAI T .002 .003 .090 .412 AS .008 .014 .064 .559
103 Table 5 13 Continued Model Variance Explained Predictor B SE B p val ue 6 Total R 2 = .097, (Constant) .577 .615 .350 R 2 = 006 Age .005 .008 .059 .513 p = .813 Gender .191 .140 .114 .173 Education .053 .024 .190 .027 UPDRS on .009 .009 .087 .331 Months PD .001 .001 .087 .308 BDI II .001 .020 .008 .965 STAI T .002 .003 .072 .534 AS .004 .018 .028 .849 Group: Pure apathy vs. No sym .178 .244 .085 .466 Group: Pure depression vs. No sym .032 .305 .011 .916 Group: Mixed apathy depression vs. No sym .265 .384 .125 491 Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
104 Table 5 14 Hierarchical multiple regression results, showing the r elationship between predictors and Language Functioning Model Variance explained Predictor B SE B p value 1 Total R 2 = .052 (Constant) .594 .520 R 2 = .052, Age .006 .007 .068 .403 p = .049* Gender .112 .129 .071 .387 E ducation .055 .022 .209 .012 2 Total R 2 = .090 (Constant) .680 .533 R 2 = .038, Age .011 .007 .136 .111 p = .0 54+ Gender .074 .129 .047 .564 Education .059 .021 .226 .006 UPDRS on .019 .008 .205 .018 Months PD .001 .001 .075 .357 3 Total R 2 = .130 (Constant) .209 .554 R 2 = .0 40, Age .008 .007 .099 .240 p = .011 Gender .039 .127 .024 .761 Education .050 .021 .191 .019 UPDRS on .019 .008 .203 .017 Months PD .001 .001 .097 .22 9 BDI II .022 .008 .210 .011 4 Total R 2 = .150 (Constant) .183 .549 R 2 = 020 Age .009 .007 .104 .215 p =.072+ Gender .043 .126 .027 .734 Education .054 .021 .204 .012 UPDRS on .019 .008 .203 .016 Months PD 001 .001 .108 .177 BDI II .009 .011 .083 .437 STAI T .004 .002 .189 .072 + 5 Total R 2 = .156 (Constant) .190 .549 R 2 = 006 Age .010 .007 .120 .159 p =.318 Gender .046 .126 .029 .718 Education .051 .021 .194 .018 UPDRS on .018 .008 .192 .024 Months PD .001 .001 .102 .201 BDI II .002 .013 .024 .846 STAI T .004 .002 .172 .104 AS .012 .012 .106 .318
105 Table 5 14 Continued Model Variance explained Predictor B SE B p value 6 Total R 2 = .16 1 (Constant) .128 .560 R 2 = 005 Age .010 .007 .117 .176 p = .857 Gender .048 .127 .030 .705 Education .052 .022 .197 .018 UPDRS on .018 .008 .192 .027 Months PD .001 .001 .093 .254 BDI II .004 .018 .042 .807 STAI T .005 .003 .201 .073 + AS .017 .017 .149 .303 Group: Pure apathy vs. No sym .152 .222 .077 .494 Group: Pure depression vs. No sym .180 .278 .066 .519 Group: Mixed apathy depression vs. No sym .144 .350 .072 .681 Note: UPDRS = Unifie d Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
106 Table 5 15 Hierarchical multiple regression results, showing the relationship between pred ictors and Wisconsin Card Sorting Test, raw number of categories achieved Model Variance explained Predictor B SE B p value 1 Total R 2 = .149 (Constant) 5.698 2.094 R 2 = .149, Age .081 .028 .358 .005 p = .019* Gender .45 5 .524 .104 .389 Education .202 .100 .246 .047 2 Total R 2 = .155 (Constant) 5.737 2.152 R 2 = .005, Age .074 .030 .327 .019 p = .830 Gender .408 .539 .094 .452 Education .204 .102 .248 .051 + UPDRS on .021 .035 .07 8 .559 Months PD .000 .005 .008 .948 3 Total R 2 = .156 (Constant) 5.459 2.446 R 2 = .0 01, Age .072 .032 .318 .029 p = .807 Gender .413 .543 .095 .450 Education .206 .104 .251 .051 + UPDRS on .021 .036 .079 .560 Months PD .000 .005 .004 .977 BDI II .010 .042 .032 .807 4 Total R 2 = .156 (Constant) 5.474 2.473 R 2 = 000 Age .072 .033 .319 .031 p =.928 Gender .412 .548 .095 .455 Education .205 .105 .250 .055 + UPDRS on .021 .0 36 .078 .567 Months PD .000 .005 .005 .971 BDI II .007 .054 .022 .894 STAI T .001 .011 .014 .928 5 Total R 2 = .160 (Constant) 5.822 2.584 R 2 = 004 Age .072 .033 .322 .031 p =.617 Gender .461 .560 .106 .414 Educ ation .194 .108 .236 .078 + UPDRS on .020 .036 .074 .593 Months PD .001 .006 .016 .901 BDI II .020 .060 .063 .738 STAI T .002 .011 .033 .839 AS .028 .056 .087 .617
107 Table 5 15 Continued Model Variance explained Predictor B SE B p value 6 Total R 2 = .182 (Constant) 5.249 2.753 R 2 = 022 Age .071 .035 .314 .049 p = .694 Gender .509 .617 .117 .414 Education .190 .114 .231 .102 UPDRS on .020 .037 .074 .601 Months PD .001 .006 .017 .900 B DI II .046 .087 .142 .602 STAI T .005 .012 .072 .690 AS .025 .087 .076 .780 Group: Pure apathy vs. No sym .667 1.204 .105 .582 Group: Pure depression vs. No sym .264 1.231 .037 .831 Group: Mixed apathy depression vs. No sym 1.616 1.374 .291 .245 Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
108 Table 5 16 Hierarchical multiple regression results, showi ng the relationship between predictors and Wisconsin Card Sorting Test, raw number of perseverative errors Model Variance explained Predictor B SE B p value 1 Total R 2 = .211 (Constant) 17.785 13.909 R 2 = .211, Age .547 .176 .375 .003 p = .003* Gender 1.185 3.361 .042 .726 Education 1.944 .674 .343 .005 2 Total R 2 = .231 (Constant) 20.808 14.243 R 2 = .020, Age .475 .194 .326 .018 p = .491 Gender .483 3.456 .017 .889 Education 1.933 .680 .341 .006 UPDRS on .208 .230 .121 .371 Months PD .033 .034 .116 .338 3 Total R 2 = .231 (Constant) 19.582 16.165 R 2 = .0 00, Age .484 .203 .332 .021 p = .869 Gender .445 3.494 .016 .899 Education 1.922 .690 .339 .007 UPDRS on .208 .232 .121 .375 Months PD .032 .035 .113 .360 BDI II .043 .262 .021 .869 4 Total R 2 = .232 (Constant) 19.152 16.414 R 2 = 001 Age .493 .209 .338 .022 p =.819 Gender .462 3.525 .016 .896 Education 1.9 03 .701 .335 .009 UPDRS on .204 .235 .118 .389 Months PD .032 .035 .113 .368 BDI II .093 .342 .045 .786 STAI T .016 .069 .036 .819 5 Total R 2 = .233 (Constant) 17.982 17.119 .298 R 2 = 001 Age .496 .211 .340 .022 p =.789 Gender .680 3.646 .024 .853 Education 1.865 .721 .329 .012 UPDRS on .197 .238 .114 .413 Months PD .030 .036 .107 .402 BDI II .048 .384 .023 .901 STAI T .021 .072 .047 .775 AS .096 .356 .046 .789
109 Table 5 16 Continued Mod el Variance explained Predictor B SE B p value 6 Total R 2 = .256 (Constant) 23.943 18.374 R 2 = 023 Age .464 .226 .318 .046 p = .675 Gender .359 4.019 .013 .929 Education 2.059 .763 .363 .009 UPDRS on .213 .245 .124 .389 Months PD .029 .038 .102 .45 0 BDI II .217 .551 .106 .696 STAI T .009 .078 .020 .908 AS .276 .550 .133 .618 Group: Pure apathy vs. No sym 2.539 7.592 .063 .739 Group: Pure depression vs. No sym 7.890 7.998 .159 .329 Group: Mixed apathy depression vs. No sym 2.9 54 8.542 .084 .731 Note: UPDRS = Unified Parkinson Disease Rating Scale, BDI II = Beck Depression Inventory II, AS = Apathy Scale, indicates significant at p < .05 ; + indicates significant at p <.1
110 Table 5 17 Demograp hic, disease variable differenc es between apathetic and nonapathetic groups. Characteristic Nonapathetic Apathetic p value (N = 107) (N=54 ) Age 63.7 (8.2) range 43 80 64 .8 (9.9) range 42 84 .46 Men: Women 72:35 (67% male) 39:15 (72 % male) .53 Years of Education 15.5 (2.8), range 12 22 14.5 (2.8), range 7 20 .04* % on A nti depressants 26 .2 % (28 out of 107 patients) 38.9 % (21 out of 54 patients) .0 9 % on Anxiolytics 21.5 % (23 out of 107 patients) 31.5% (17 out of 54 patients) 16 Levodopa Equivalent Dosage 831.7 (523.3) 778.7 (489.9) .33 Months of Symptoms 99.7 (51.0) range 12 251 106.2 (61.1) range 12 241 .50 M otor score (UPDRS, on levodopa) 24.2 (8.6), range 9 47 2 6.6 (7.9 ), range 13 4 6 .098 Note: N = 161. However, f ive patients were missing UPDRS on scores, (N = 157 ) Results are presented as means (standard deviations) and ranges.
111 Table 5 18 Mood and anxiety differences between apathetic and nonapathetic groups. Characteristic Nonapathetic Apathetic p value (N = 107) (N= 54 ) Apathy (AS) 7.2 (3.7) 17.9 2 (4.0) --Depression (BDI II) 7.07 (4.86), range 0 23 14.43 (8.60), range 1 34 <.001 State Anxiety (STAI) Percentile 49.05 (28.3), range 5 97 70.34 (28.4), range 11 99 <.001 Trait Anxiety (STAI) Percentile 42.23 (28.17), range 3 98 74.98 (25.24), r ange 6 100 <.001 Note: N = 161. However, six patients were missing STAI State scores, (N = 155), and seven patients were missing STAI Trait scores (N = 154) Results are presented as means (standard deviations) and ranges.
CHAPTER 6 DISCUSSION The present study investigated two hypotheses. The first hypothesis was that apathy and depression, although similar, are separable experiences of mood states that can be dissociated in nondemented patients with Parkinson disease. For this reason, it was pr edicted that individual items from two commonly used depression and apathy self report inventories, the Beck Depression Inventory II and the Apathy Scale, would be dissociable via factor analysis. Specifically, items from these scales were predicted to fa ll onto four main factors: 1) an apathy apathy and depression loss of appetite, insomnia). These factors were based on the concept that depression and apathy differ in terms of their c ore features. Namely, depression is characterized by sadne ss, mood along with loss of motivation. The second hypothesis evaluated in this study involved the relationship between cognition and apathy. Previous studies indi cated a negative association between apathy and executive functioning (Starkstein et al., 1992, Isella et al., 2002, Pluck & Brown, 2002; Zgaljardic et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009). It was hypothesized that apathy would be as sociated with worsened cognitive status, particularly dysfunction of frontal lobe systems. However, all studies until the present one have not controlled for important comorbidities such as dementia, disease variables, and depression. Previous work has a lso focused on executive functioning measures to the relative exclusion of other cognitive domains (language, episodic memory, processing speed, etc). Based on the idea that frontal lobe pathology (i.e. mesial frontal
cortex/anterior cingulate cortex) is putatively involved in apathy, we hypothesized that apathy was related to deteriorated frontal lobe functioning and that this would be evidenced by poor performance in executive functioning tasks. Specifically, apathy scores would predict poor scores in t he executive functioning domain, but not in other cognitive domains. Prevalence of Apathy and Depression The current sample of 161 non demented Parkinson patients was comparable to that of others described in the literature in terms of prevalence of apath y and depression. Apathy was present in approximately one third of the patients in the current Parkinson sample (54 out of 161 patients, 34%). Ten previous studies are known to the authors that have assessed apathy in PD using a self report scale (Starks tein et al., 1992; Pluck & Brown, 2002; Czernecki et al., 2002; Isella et al., 2002; Kirsch Darrow et al., 2006; Dujardin et al., 2007; Zgaljardic et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009; Starkstein et al., 2009). Some of these studie s excluded dementia patients, while some did not. The present results fall into the range repo rted in the nondemented samples of between 23% 44% (Czernecki et al., 2002; Dujardin et al., 2007; Pedersen, Alves et al. 2009; Pluck & Brown, 2002; Zgaljardic et al., 2007) In contrast, other studies did not screen out dementia patients. These studies generally have higher prevalence rates for apathy, ranging from 32% 56% (Starkstein et al., 1992; Isella et al., 2002; Kirsch Darrow et al., 2006; Dujardin et al., 2007; Santangelo et al., 2009; Starkstein et al., 2009). Dujardin and colleagues compared apathy in both a demented group (n = 39) and a nondemented group (n = 120) of Parkinson patients using the Lille Apath y Rating Scale (LARS). The demented group had over twice the prevalence of clinically significant apathy than the nondemented group (56% in the dementia group vs. 24% in the nondemented group) (Dujardin et al., 2007) In addition to PD, a high prevalence of apathy i s also common in other dementia
assess apathy, studies found prevalence rates of 55% 81% (van Reekum et al., 2005) Other subcortical disorders that commonly involve dementia, are frequently associated with apathy. and Corticobasal degenera tion (Aarsland, Ballard, McKeith, Perry, & Larsen, 2001; Litvan, Cummings, & Mega, 1998; Litvan, Paulsen, Mega, & Cummings, 1998; Paulsen, Ready, Hamilton, Mega, & Cummings, 2001) The strong relationship between t he syndromes of apathy and dementia is thought to occur because of similar underlying brain pathology (i.e. rather than an apathy syndrome being the result of im paired cognition ). Both dopaminergic (Brown & Pluck, 2000) and non dopaminergic circuits have been hypothesized as the cause (Dujardin et al., 2007; Cummings & Black, 1998). Importantly, the results from the present study indicated that apathy can occur in the absence of depression. In the current study, 17% of patients had apathy without depress ion, 16% had symptoms of both apathy and depression, and 9% had symptoms of depression only. These findings are consistent with the literature. Several studies broke down their prevalence figures into apathy alone, depression alone, and apathy and depres sion together. Apathy alone prevalence has ranged from 12 29%, depression alone prevalence has ranged from 3 22%, and the prevalence of both apathy and depression has ranged from 12 47% (Starkstein et al., 1992; Isella et al., 2002; K irsch Darrow et al., 2006; Zgalj ardic et al., 2007). The rates from the current study fall within each of these ranges. Factor Structure of Apathy and Depression in PD The majority of studies listed above classify apathy and depression using cutting scores on self report in ventories. However, this problematic because of the use of total scores. Total scores overlook the issue that apathy and depression have overlapping symptoms such that the e. To
overcome this limitation, item level confirmatory factor analysis was used to examine the hypothesis that apathy and depression can be dissociated, but also share some symptomatology. A four factor model was hypothesized that involved: 1) an apathy s of depression, and 4) a somatic factor represent four factor solution fit the data well. Importantly, it fit the data better than all alternative nested models. The four factor model was significantly better at describing the data than a single fac tor model that combined apathy the Beck Depression Inventor y significantly better fit than the two alternative three factor models that omitted either a) the loss of interest and pleasure factor (i.e. included the apathy, depression, and somatic fa ctors) or b) the somatic factor (i.e. included the apathy, depression, and loss of interest and pleasure factor). Taken together, these findings contribute to the growing body of literature suggesting a separation of these two mood states in PD (Levy et a l., 1998; Starkstein et al., 1992 Isella et al., 2002; Kirsch Darrow et al., 2006; Zgal j ardic et al., 2007; Durjardin et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009). The findings argue against the idea that apathy is more accurately classifi ed as a subcomponent of depression. Moreover, this model supports several concepts about the different characteristics of apathy and depression in PD. Depression includes sadness and negative thoughts about the self. One of the item clusters that loa de d most highly on the dysphoric
than u dislike als o loaded highly on the dysphoric mood factor. In contrast, apathy does not involve pessimistic self and event appraisal. Instead, apathetic individuals lack responsiveness to both negative and positive events (Brown & Pluck, 2000). Results lend support for the idea that apathy involves lack of initiation, and lack of effort. In Two additional factors were hypothesized one involving the overlapping symptoms of loss of interest and anhedonia, and one involving somatic symptoms. Both were supported in this study. A National Institute of Neurological Disorders and Stroke (NINDS) depression workgroup highlighted loss of interest as a symptom that overlaps between depressio n and apathy (Marsh et al., 2006 ). This has never b een empirically tested until the present study Furthermore, the depression wor kgroup cautioned that using loss of interest as one of the two major core symptom s of an MDD (i.e. Criteria A1 = sad mood or Criteria A2 = markedly diminished interest or pleasure) could lead to overdiagnosis of MDD because the symptom of loss of interest might be be tter accounted for by a syndrome of apathy (Marsh et al., 2006) The workgroup recommended omitting decreased interest as a core affec tive symptom when diagnosing MDD in PD. They noted similar concerns regarding anhedonia, but decided that it was a core feature of depression as well as involved in apathy; they did not recommend omi tting it from diagnosis of MDD (Marsh et al., 2006) Unfortunately, this solution leaves the symptom of anhedonia in the same scenario as loss of interest. Namely, it could count towards one of the cor e symptoms of depression, or it could
count towards primary apathy. Or, a patient could have both syndromes. This is especially problematic when diagnosing minor depression. Minor depression requires a total of only two symptoms. One of these must be a core de pression symptom (i.e. sad mood or loss of interest/anhedonia); the other can be any of the major depressive disorder symptoms. In PD, minor depression can easily be misdiagnosed because a patient may have loss of interest or anhedonia plus anothe r symptom associated with PD itself (e.g. psychomotor slowing, insomnia, concentration problems). Since the number of required symptoms is so few and they have a high potential to be related to apathy or to PD itself, minor depression diagnoses should be avoided in PD. They remain in the appendix of the DSM IV TR and, as such, have an uncertain nosology and are considered diagnostic research criteria only. Generally, this problem of misdiagnosis using the core feature of loss of interest or anhedonia co uld occur even when the depressive diagnosis has a clear nosology like MDD. The potential to misclassify apathetic patients as depressed using one of these criteria, although less likely because it requires at least 5 de pressive symptoms, is possible. Ap petite, sleep, psychomotor changes, concentration, and fatigue are five symptoms that are all in common between MDD and PD. If these were used purely to diagnose depression, alongside loss of interest or anhedonia, the potential for a false positive diagn osis of MDD is great. Careful evaluation and critical thinking is required for an accurate diagnosis of MDD in PD Given the results of this study, we recommend the following: First, loss of interest/anhedonia symptoms should be evaluated for longitudin al course noting when they onset, fluctuations in severity, be critically evaluated to make sure it is not due to PD itself. Looking at the course of each symptom an d the context of their onset will help with this (e.g. did depressive symptoms begin
roughly at the same time?). For example, did loss of interest begin at the same time as decreases in appetite, middle insomnia, worsened concentration, and fe elings of wo rthlessness? Furthermore, it may be useful to use apathy and depression scales in a modified manner symptoms are added into the total scores, it is hard to deter mine whether they fall into apathy or depressive syndrome clusters. We propose an additional scoring procedure for the scales for use with PD patients. In addition to summing the items to derive total scores, a complimentary method could be used that com the four factors found in this study. This will provide four separate indices of 'pure' apathy and 'pure' depression, overlapping loss of interest/plea sure, and somatic symptoms. Future studi es could use this complimentary scoring m ethod and compare to diagnoses of MDD and apathy based on the consensus crite ria recently proposed by Robert et al. a nd the European task force (Robert et al., 2009) to deter mine if they have diagnostic utility over and above total scores. On a similar note, future studies could use discriminant function analysis to determine the symptoms that best discriminate apathy and depressio n in PD. Groups of patients could be clas sified a priori using the DSM IV SCID for MDD and using the consensus criteria for apathy. Then, the AS and BDI II can be administered to both groups. The items that best discriminate between the two syndromes can then determined and noted for future use of the scales. It is hypothesized that the guilt/worthlessness item, punishment item, suicidal ideation item, and irritability item will discriminate depression from apathy. Items that overlap (e.g. anhedonia, loss of interest), as well as fatigu e and c oncentration items likely will not discriminate between groups.
As an additional next study, larger sample sizes could also be obtained in order to cross validate the factor analysis results. A weakness of the present study is the lack of a large enough sample to cross validate. First, an exploratory factor analysis can be performed. Then, the results of this study can be used in a confirmatory factor analysis. Combining the items into groups of two item pairs (e.g. parcels) could also be considered a weakness as well. Parceling d oes not allow each item to independently load on factors. A stronger item could influence a weaker item in terms of loadings on factors. Parceling was necessary in the present study because of severe non normality of the d ata A sample with a more normal distribution of apathy scores might be able to use individual items Apathy and Cognition In addition to examining the relationship between apathy and depression, a second aim of the present study was to investigate th e relationship between apathy and cognition. Previous studies have reported that increased apathy is associated with decreased executive functioning (Starkstein et al. 1992; Isella et al., 2002; Pluck & Brown, 2002; Zgaljardic et al., 2007; Pedersen et al ., 2009; Santangelo et al., 2009). Based on the idea that frontal lobe pathology is putatively involved in apathy (i.e. mesial frontal cortex/anterior cingulate cortex), it was hypothesized that apathy was related to impaired frontal lobe functioning and that this would be evidenced by poor performance on executive functioning tasks. It was specifically predicted that apathy scores would be significantly related to poor performance in the Executive functioning domain, but not to other cognitive domains (e .g. Processing speed, Verbal episodic memory, Working memory, Language). The hypothesis presented above was supported. Apathy explained significant incremental variance in Executive functioning (3.5%), but not in any other cognitive domain.
For Executive functioning, the influence of apathy was above and beyond that of demographic variables, disease related variables, and depression. Breaking these domains down into the individual component tests, apathy expl ained significant variance in the Stroop Color Word task, Interference condition (5.5%, p <.01). Apathy explained variance in animal fluency at trend level (2.3%, p = .06). However, apathy did not contribute variance to letter fluency or Trail Making Tes t, Part B. The influence of apathy on the Wisconsin Card Sorting Task (WCST) was examined in smaller subgroup of PD patients (n = 68) who were administered this task. Neither the number of categories achieved, nor the number of perseverative errors made were significantly related to apathy. Only age and education were related to WCST performance. Based on these findings, it appears that apathy is related to certain aspects of executive functioning such as speeded semantic verbal fluency [as measured by animal fluency] and cognitive inhibition [as measured by Stroop Color Word test, Interference condition]. Apathy was not related to other aspects of executive functioning such as shifting mental sets [as measured by Trail Making Test, Part B and the WCST ] or phonemic fluency [as measured by F,A,S letter fluency]. Interestingly, the types of executive measures that seem most affected by apathy are those that are known to be associated with the mesial frontal lobe/anterior cingulate cortex (e.g. cognitive inhibition). This is in contrast to the ones that apathy was not associated with (e.g. shifting mental set to changing demands and phonemic fluency) which are more associated with the dorsolateral prefrontal cortex. Apathy and Stroop color word perform ance: Relationship to anterior cingulate c ortex Apathy explained significant incremental variance in performance on the Stroop Color Word test. Out of all the tests that were evaluated, apathy contributed the most variance (5.5%) to the Interference con dition of the Str oop Color Word test. N o other predictors aside from apathy explained significant variance on the test (i.e. not demographics, disease variables, or
depression). The Stroop test has a long history (Stroop, 1935), and has become one of mos t that it took longer for subjects to name the color of the ink that color words were written in when the ink color and color word were incongruent (e.g. the word red written in blue ink), than it took for the subjects just to name the color of colored squares (Stroop, 1935 ). T oday, t he classic Golden version of the Stroop test involves three conditions: the Stroop Word condition, the Stroop Color condition, a nd the Stroop Color Word condition. In the Stroop Word condition, participants read words that denote color names (i.e., red, green, blue). In the Stroop Color (i.e. red, green, blue). Finally, in the Stroop Color Word condition, the cognitive interference condition, participants name the color of the ink of words that denote colors (i.e., word red printed in green ink). The Stroop effect is based on the findi ng that it takes longer to read the name of a word multiple interpretations (Lezak, Howieson, & Loring, 2004). Some researchers attribute the slowing to a defect on mulus due to a second stimulus (Bush et al., 1998; Dyer, 1973) Others have interpreted it to be a hibit the pre potent response of simply reading the word instead of stating the color of the ink the word is printed in (Zajano & Gorman, 1986). Although these may occur in healthy adults, and present as a decrease in reaction time, it often occurs to exag gerated extent in persons with frontal lobe injury (Lezak, Howieson, & Loring, 2004).
performance on the Stroop Color Word Interference condition (p =.001). As apathy inc reased, performance on Stroop Color Word condition worsened. To see if this finding was specific to the cognitive interference condition, post hoc hierarchical regressions were performed with the same predictors and with the Stroop Word reading condition and the Stroop Color naming condition as dependent variables. Apathy did not explain significant variance in either (Word reading, p = .28; Color naming, p = .09). The relationship between Stroop Color Word task, interference condition performance and a pathy is particularly intriguing because both are thought to involve the anterior cingulate cortex (ACC). Multiple studies using fMRI in healthy adults have reported a relationship between the Stroop interferenc e condition and ACC activation (Bench et al., 1993; Carter, Mintun, & Cohen, 1995; Pardo, Pardo, Janer, & Raichle, 1990) Studies find that the ACC is significantly more active during interference than non interference trials (Bush et al., 1998). Further, l esion studies by Stuss and colleagues (2001) have indicated that that bilateral superior medial frontal damage is associated with increased errors and slowing on the interference trial of the Stroop, whereas posterior lesions were n ot related (Stuss, Floden, Alexander, Levine, & Katz, 2001) The neurobiological substrates of apathy are unknown, but are hypothesized to involve the ACC circuit (Brown & Pluck, 2000; Isella et al., 2002). Specifically, apathy is thought to involve the striato thalamo cortical circuit originating in the ventral tegmental area (VTA) and ending in the ACC (VTA ventral striatum ventral pallidum medialdorsal thalamus ACC). These limbic structures are involved in motivation and drive, and are important in translating motivation i nto action (Davidson & Irwin, 1999; Groenewegen, Wr ight, & Beijer,
1996; Mogenson, Jones, & Yim, 1980) Lesions in the region of the ACC and supplemental motor area produce a syndrome of extremely severe apathy, called akinetic mutism. In this syndrome, patients make no effort to communicate or initiate activities: they are conten t to lie silent and motionless (Kant, Duffy, & Pivovarnik, 1998 ; Damasio & Van Hoesen, 1983; T ranel, 1992). In PD, only one study exists that examines the neuroanatomy of apathy. Remy and colleagues used Positron Emission Tomography (PET) imaging and found that apathy was inversely correlated with dopamine and norepinephrine b inding in the ventral striatum (Remy et al., 2005) The ventral striatum is a key structure in the circuit described above. Dys function of the ACC circuit, perhaps neurochemically through loss of dopamine and neuropathologically through Lewy bodies, may be underlie both apathy and Stroop Color Word performance in PD. Future studies should continue to address the relationship betw een apathy, cognitive interference/Stroop, and the ACC. Future studies could use Functional Magnetic Resona nce Imaging (fMRI) during performance on a Stroop task as a noninvasive alternative to PET to investigate whether apathetic PD patients have less ac tivation of the ACC and related circuitry than do non apathetic PD patients. Apathy and speeded verbal f luency In addition to Stroop Color Word, apathy explained variance at trend level for animal (semantic) fluency. In contrast, there was no relationship between apathy and phonemic fluency. An association between apathy and verbal fluency has been found in all studies that have assessed verbal fluency in the context of apathy (Starkstein et al., 1992; Pluck & Brown, 2002; Isella et al., 2002; Zgaljardic et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009). Namely, increased apathy is associated with decreased performance on verbal fluency tasks. It is unknown why the present study found that apathy was related to animal fluency at trend (p = .0 6) yet found no relationship with phonemic fluency. Phonemic fluency is often considered a more
difficult task for PD patients because it requires organization and switching between categories (e.g. executive based) versus semantic fluency which relies on memory and retrieval from semantic storage. Further studies are needed to replicate that apathy has a specific effect on semantic fluency, but not on phonemic fluency. In contrast to these findings, most of the previous studies compared scores between a pathetic and nonapathetic PD groups and found both phonemic and animal fluency to be significantly worse in the apathetic group (Pluck & Brown, 2002; Isella et al., 2002; Zgaljardic et al., 2007). Starkstein and colleagues found their apathetic group to b e lower on phonemic fluency than their nonapathetic group, but di d not assess semantic fluency (Starkstein et al., 1992) Speeded fluency tests require an individual to rapidly produce words, either that begin with a letter of the alphabet (i.e. F, A, and S for letter fluency), or are in a particular category (i.e. animals for animal fluency). To do well on fluency tests, the individual must successfully perform several tasks: a) self initiate responses to a single prompt, b) organize their verbal output into meaningful clusters of related words (i.e. phonologically or semantically), and c) switch between clusters (Troyer, Moscovitch, & Winocur, 1997) Generally, it is well known that frontal lob e lesions, particularly of the left dominant hemisphere, induce impairments on verbal fluency tasks (Baldo & Shimamura, 1998; Milner, 1975) Stuss and colleagues (1998) administered the phonemic [F,A,S] and animal fluency tests to seventy four brain damaged patients and a comparison group of healthy age matched control s (Stuss et al., 1998) They found that patients with superior medial frontal lesions (including the ACC), either alone or with formances differed produced significantly fewer words in the first fifteen seconds of the task than controls. Stuss et
al. (1998) proposed that the ACC may play a role in response initiation in this task (Stuss et al., 1998) In the current study, the mechanism of impaired verbal fluency is unknown. To further understand this relationship, it will be important for future studies to examine performance in more fine grained ways. Some possibilities include examining t he relationship between apathy and aspects of fluency performance including: 1) semantic versus phonemic fluency using different prompts other than animal and F,A,S to see if the specific relationship to semantic fluency versus phonemic fluency is replica ted, 2) measuring cluster size, 3) measuring the number of switches between categories, and 3) scoring the test in 15 second intervals to determine if apathy is associated with problems initiating output to the prompt. Apathy Depression Group Findings Four apathy depression groupings were created that included PD patients with: a) pure apathy, b) pure depression, c) mixed apathy depression, and d) neither apathy nor depression. The main differences between groups were that the pure apathy group was old er than the mixed apathy depression group. No other groups differed in age. Most studies have not found a difference in age between apathetic and nonapathetic groups (Pluck & Brown 2002; Isella et al., 2002; Zg a l jardic et al., 2007; Starkstein et al., 1 992; Santangelo et al., 2009). In contrast, Pedersen and colleagues (2009) also found their apathetic group to be significantly older than their nonapathetic group. Additionally, the mixed apathy depression group had fewer years of education than the grou p with neither apathy nor depression. Although other studies have examined education and have not found a difference between groups (Pluck & Brown 2002; Isella et al., 2002; Zg a l jardic et al., 2007; Starkstein et al., 1992; Pedersen et al., 2009; Santang elo et al., 2009), this is one of the largest samples size to date (N = 161). Most studies sample sizes ranged between N = 30 50
(Pluck & Brown, 2002; Isella et al., 2002; Zgaljardic et al., 2007; Starkstein et al., 1992). As such, the current study may have included a larger range (i.e., 7 to 22 years) of education than previous studies. Although the precise relationship with apathy is unknown, the number of years of education may be a proxy for other variables such as general intellectual functioning o r cognitive reserve. In terms of anxiety, groups differed in terms of state and trait anxiety as measured by the STAI. The mixed gr oup had significantly higher state anxiety than all other groups. There were no other differences in state anxiety betwee n groups. In terms of trait anxiety, percentiles above 93 rd percentile (1.5 SD) were considered and the percentage of patients meeting this criteria was calculated. Results indicated dramatic differences between groups. The results were as follows: 1% for the no symptom group, 7% for the pure apathy group, 31% for the pure depression group, and 64% of the mixed apathy depression group. Turning to the literature regarding anxiety, Pluck & Brown (2002) administered the Hospital Anxiety and Depression Sca le and found no difference in anxiety levels between patients with high versus low apathy. Starkstein and colleagues (1992) administered the Hamilton Rating Scale for Anxiety, and found higher anxiety in their depressed Parkinson group than in their Parki nson group without apathy or depression (Starkstein et al., 1992). However, they did not report comparisons between their apathy only group, or their mixed apathy depression group and their no symptom group. Finally, Aarsland et al. (1999) factor analyze d the symptoms from the Neuropsychiatric Inventory (NPI) from a combined group of demented and nondemented PD patients (Aarsland et al., 1999) Results indicated that two factors emerged one involving delusions, hallucinations, agitation, and irritability and another involving apathy and anxie ty (Aarsland et al., 1999) Certainly, a relationship between apathy and anxiety seems cou nterintuitive given that apathy is thought to
involve disinterest and disengagement rather than preoccupation and worry. A detailed discussion regarding the relationship between apathy and anxiety is provided in the next section. Comparison to the Current Literature The results from this and previous studies support a relationship between apathy and executive functioning. What is less clear is the exact nature and pattern of the specific executive deficits that are related to apathy. Most studies, includ ing the present one (though at trend), find a relationship between fluency and apathy, with apathetic patients performing more poorly on verbal fluency tasks than non apathetic patients (Starkstein et al., 1992; Pluck & Brown, 2002; Isella et al., 2002, Z galjardic et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009). Findings are mixed with regards to the Stroop Color Word Test. Like the present study, all studies used the Golden version of the Stroop. Pluck and Brown (2002) and Santangelo et a l. (2009) found worse Stroop interfer ence performance in apathetic versus nonapathetic patients, whereas Zgaljardic et al. (2007) and Pedersen et al. (2009) did not. In line with these general findings, the present study found that apathy was associated with worse performance on the Stroop interference condition and a trend for worse performance on animal fluency. In contrast, the present study did not find a relationship between apathy and the Wisconsin Card Sorting Test (WCST) in terms of the number o f categories achieved or the number of perseverative errors. Previous findings for this variable, like the Stroop Color Word Test, have also been mixed. Starkstein et al. (1992) did not find a difference between apathetic and nonapathetic groups in terms of WCST categories achieved (Starkstein, Mayberg, Preziosi et al., 1992) This was true both when authors scored the measure continuously (i.e. 0 6 categories achieved), or sc ored it in a dichotomous manner (i.e. with patients achieving 6 categories or fewer than 6 categories). Another study found that apathetic patients achieved fewer categories
than nonapathetic patients (Pluck & Brown, 2002). However, the number of perseve rative errors was not different between the apathetic and nonapathetic groups. Relationship Between Apathy and Anxiety A somewhat unexpected finding was the relationship between apathy and anxiety. This fi nding seems counterintuitive because a pathy is tho ught to involve disinterest and disengagement rather than anxiety and worry. Consistent across all analyses was the finding of high anxiety in the mixed apathy depression group The mixed apathy depression group had significantly higher state anxiety tha n all other groups Similarly, the mixed apathy depression group had significantly higher trait anxiety t han the pure apathy group and the no symptom group. The mixed group and the pure depression group were not significantly different on trait anxiety. When examining clinically elevated trait anxiety, the groups differed dramatically from one another. Over half of the patients in the mixed apathy depression group had clinically elevated anxiety and one third of the patients in the pure depression group had clinically elevated anxiety. In contrast, only 7% of the pure apathy group had clinically elevated anxiety and 1% of the no symptom group had clinically elevated anxiety. Thus, the group with mixed apathy and depression was the highest in trait anxi ety. Additionally, the percentage of elevated trait anxiety in the pure depression group was over three times as high as that of the pure apathy group. These findings suggest that pure apathy is rarely associated with clinically elevated trait anxiety, w hereas the mixed apathy depression group is associated with high trait anxiety. This suggests the existence of a highly comorbid subgroup with a mixture of high apathy, depression, and trait anxiety symptoms. Interestingly, the groups were not different in terms of duration PD or severity of PD motor symptoms. It is well established that anxiety and depression co occur in PD (Nuti et al., 2004 ; Menzaet al., 1993; Henderson et al., 1992). A high degree of overlap can often be found
between patients with depression and anxiety. For example, Pontone and colleagues administered the Structured Clinical Interview for the DSM IV (SCID DSM IV) to one hundred twenty seven non demented PD patients (Pontone et al., 2009) They found that of the individuals with a current anxiety disorder diagnosis, 65% also had a curr ent depressive disorder diagnosis. Further, similar to the overlapping symptom content between depression and apathy measures, there is overlapping symptom content between depression and anxiety measures. For example, the STAI trait scale has patients r indicated that the mixed apathy depression group had significantly higher depression severity (mean BDI II score of 22) than the pure depression group (mean BDI II score of 1 6). Thus, the addition of apathy may only appear to be related to increased clinical elevations of anxiety. Instead, it may be that the more severe depression is driving the relationship with anxiety. In any case, this subgroup with high apathy, depress ion, and anxiety is clearly experiencing high levels of distress, and endorsing all measures of negative symptoms. In future studies, it would be important to determine whether this group of patients might be either perceiving more functional disability f rom PD or experiencing more role loss (e.g. recently had to quit work due to symptoms, unable to be head of household anymore due to cognitive changes, etc) even though their motor symptoms are not significantly worse. Limitations The current study has several limitations. First, when collecting data from patient medical charts, only total scores f or the State Trait Anxiety Inventory (STAI) were gathered.
Item level data was obtained for the Beck Depression Inventory II and Apathy Scale, but not for t he STAI Having STAI items would have allowed further investigation into the nature of the relationship between apathy and anxiety. Specifically, apathy could be examined separately for the group of items relating to depressive content versus the group o f items specific to anxiety. The ratio of the total number of items from all three measures to the sample size of the current study would not have allowed for a confirmatory factor analysis. However, collection of a larger sample could address separate a pathy, depression, and anxiety factors in Parkinson disease. Next, this study did not use psychiatric interviews and DSM IV diagnoses in addition to the BDI II. This would have allowed for apathy prevalence to be assessed within the context of Dysthymia or Major Depressive Disorder (MDD). However, research has shown that patients with significant depression symptoms, even without meeting full criteria for a MDD, still experience significant disability from their mood symptoms and benefit from treatment (Judd, Paulus, Wells & Rapaport, 1996; Lyness et al., 1996). Additionally, DSM IV diagnoses would have helped discriminate the type of anxiety patients were experiencing. Some anxiety disorders are thought to overlap more with depression in PD (i.e. spec ific phobia and anxiety NOS) than others (i.e. agoraphobia and social phobia) (Pontone et al., 2009). Diagnoses would have allowed us to examine disorder based overlap. Finally, this study created composite scores by combing tests together rationally base d on the cognitive processes they were thought to tap. This is a common approach in studies that admi nister a large number of tests (Schinka, Vanderploeg, Rogish, & Ordorica, 2002; Sheline et al., 2006) Yet, rati derived domains. In fact, this was the case when we conducted an exploratory factor analysis of the cognitive tests in our study. For example, fluency loaded onto its own factor (i.e. letter
fluency and animal fluency) and Trail Making Test Part B loaded with Trail Making Test Part A and Digit Symbol. On the other hand, the individual measures that were predicted a priori to be related to apathy were so, regardless of their loadings in the exploratory factor analysis. Conclusions and Directions for Future Research Apathy occurred in one third of the Parkinson patients in this sample. As such, it appears to have a prominent place in the neuropsychiatric profile of nondemented Parkinson patients. Findings from the present study provide support for the hypothesis that apathy and depression are similar, but separable experiences of mood states. Support was found in terms of important discriminating characteristics of apathy and depressio n. Practically, this finding argues for an entory II into apathy, dysphoric mood, loss of interest/pleasure, and somatic complaints indices. This may help disentangle sy mptoms related to apathy, depression, and somatic aspects of PD. This study found an unexpected association between apathy and anxiety. Namely, the apathetic group had significantly more anxiety than the nonapathetic group. However, a dditiona l findings suggested that pure apathy was rarely associated with clinically elevated anxiety whereas mixed apathy depression was highly associated with clinically elevated anxiety. The mixed apathy depression group was twice as high in anxiety symptoms as the pure depression group. At first it seemed as if the addition of apathy was related to the anxiety elevations. However, it turned out that the mixed group was significantly higher in overal l depression symptoms than the pure depression group. Thus, the depres sion vs. the apathy is most likely related to the heightened anxiety symptoms. It also appeared that the total score from the STAI has a similar problem of overlapping features with the BDI II as the AS does (i.e. STAI measure has statements regarding not feeling happy, feeling like a failure, lack of self con fidence, etc).
Future research using a larger sample size can delineate discrete apathy, depression, and anxiety constructs in PD and test these ideas with confirmatory factor analysis. Further, the hierarchical regression results pointed to a relationship between apathy and executive functioning. This is consistent with the small body of literature that has previously examined the association (Starkstein et al., 1992; Pluck & Brown, 2002; Isella et al., 2002; Zgaljardic et al., 2007; Pedersen et al., 2009; Santangelo et al., 2009). What remains to be determined is the specific patterns of executive functioning impairment that are associated with apathy. This study found that some aspects were relat ed to apathy (e.g. cognitive interference, semantic fluency), while others were not (e.g. set shifting, phonemic fluency). The findings from previous studies all point to an association with fluency, but are mixed in regards to cognitive interference and set shifting. Additionally, links between apathy and types of executive functioning impairment are interesting from a theoretical perspective of common underlying neural substrates. Future studies may develop and test more fine grained hypotheses related to this. For example, test batteries could be specifically designed to differentially tap the anterior cingulate cortex/mesial frontal lobe functioning (initiation, cognitive interference, etc) versus dorsolateral prefrontal cortex functioning (set shift ing, complex problem solving, working memory) versus orbitofrontal functioning (perseveration, impulse control, etc). Apathy would be hypothesized to relate most to anterior cingulate cortex/mesial lobe functioning. Further support for the notion that a pathy and depression are separable in Parkinson disease has broad implications for the field of movement disorders. It suggests that clinicians should screen for both conditions in order to appropriately triage and treat patients. Selective serotonin reu ptake inhibitors (SSRIs) and dual serotonin and norepinephrine reuptake inhibitors
improve apathy. Further, in other neurological disorders, treatments for ap athy are being examined in pharmacological areas such as: amphetamines (e.g. methylphenidate), atypical antipsychotics, dopaminergic agents, and acetylcholinesterase inhibitors. These may hold promise for the treatment of apathy in PD. In small (n) studi es, preliminary support has been found for two dopaminergic agents, bromocriptine and amantadine, in treating apathy in traumatic brain injury and poststroke patients (van Reekum, Stuss, & Ostrander, 2005). At the same time that pharmacological research i s getting underway, non pharmacological interventions engaging the patient slowly back into activities and interests could also be investigated. This study adds to the current literature by emphasizing the dissociability of apathy and depression symptom s and highlighting the prominent place apathy has in the neuropsychiatric profile of PD patients. Further, it provides a carefully controlled examination of apathy and cognition in nondemente d PD patients. Results indicate a clear and specific relationship between apathy and executive functioning in PD. As the concept of apathy in PD is delved into further, differential pathology, clinical correlates, and effective treatments will be discove red.
LIST OF REFERENCES Aarsland, D., Andersen, K., Larsen, J. P., Lolk, A., Nielsen, H., & Kragh Sorensen, P. (2001). Risk of dementia in Parkinson's disease: a community based, prospective study. Neurology, 56 (6), 730 736. Aars land, D., Ballard, C., McKeith, I., Perry, R. H., & Larsen, J. P. (2001). Comparison of extrapyramidal signs in dementia with Lewy bodies and Parkinson's disease. J Neuropsychiatry Clin Neurosci, 13 (3), 374 379. Aarsland, D., Larsen, J. P., Lim, N. G., Jan vin, C., Karlsen, K., Tandberg, E., et al. (1999). Range of neuropsychiatric disturbances in patients with Parkinson's disease. J Neurol Neurosurg Psychiatry, 67 (4), 492 496. Albert, M. L., Feldman, R. G., & Willis, A. L. (1974). The 'subcortical dementia' of progressive supranuclear palsy. J Neurol Neurosurg Psychiatry, 37 (2), 121 130. Arbisi, P.A. (2001). Review of the test Beck Depression Inventory II. In B. S. Plake & J.C. Impara (Eds,), The fourteenth mental measurements yearbook (pp. 150 154). Lincoln NE: Buros Institute of Mental Measurements Baldo, J. V., & Shimamura, A. P. (1998). Letter and category fluency in patients with frontal lobe lesions. Neuropsychology, 12 (2), 259 267. Beck, A.T., Steer, R., & Brown, G. (1996). Beck Depression Inventory II San Antonio, TX: The Psychological Corporation. San Antonio, TX. Beck, A.T. (1978). Beck Depression Inventory. San Antonio, TX: The Psychological Corporation. Bench, C. J., Frith, C. D., Grasby, P. M., Friston, K. J., Paulesu, E., Frackowiak, R. S., et al. (1993). Investigations of the functional anatomy of attention using the Stroop test. Neuropsychologia, 31 (9), 907 922. Brandt, J., Shpritz, B., Munro, C. A., Marsh, L., & Rosenblatt, A. (2005). Differential impairment of spatial location memory in Hun tington's disease. J Neurol Neurosurg Psychiatry, 76 (11), 1516 1519. Brown, R. G., & Pluck, G. (2000). Negative symptoms: the 'pathology' of motivation and goal directed behaviour. Trends Neurosci, 23 (9), 412 417. Burgess, P. W., Alderman, N., Evans, J., E mslie, H., & Wilson, B. A. (1998). The ecological validity of tests of executive function. J Int Neuropsychol Soc, 4 (6), 547 558. Burgess, P. W., Veitch, E., de Lacy Costello, A., & Shallice, T. (2000). The cognitive and neuroanatomical correlates of multi tasking. Neuropsychologia, 38 (6), 848 863.
Bush, G., Whalen, P. J., Rosen, B. R., Jenike, M. A., McInerney, S. C., & Rauch, S. L. (1998). The counting Stroop: an interference task specialized for functional neuroimaging -validation study with functional MR I. Hum Brain Mapp, 6 (4), 270 282. Caballol, N., Marti, M. J., & Tolosa, E. (2007). Cognitive dysfunction and dementia in Parkinson disease. Mov Disord, 22 (S17), S358 S366. Canavan, A. G., Passingham, R. E., Marsden, C. D., Quinn, N., Wyke, M., & Polkey, C. E. (1989). Sequence ability in parkinsonians, patients with frontal lobe lesions and patients who have undergone unilateral temporal lobectomies. Neuropsychologia, 27 (6), 787 798. Carter, C. S., Mintun, M., & Cohen, J. D. (1995). Interference and facilitation effects during selective attention: an H215O PET study of Stroop task performance. Neuroimage, 2 (4), 264 272. Chan, R. C., Shum, D., Toulopoulou, T., & Chen, E. Y. (2008). Asse ssment of executive functions: review of instruments and identification of critical issues. Arch Clin Neuropsychol, 23 (2), 201 216. Chatterjee, A., & Fahn, S. (2002). Methylphenidate treats apathy in Parkinson's disease. J Neuropsychiatry Clin Neurosci, 14 (4), 461 462. Collie, A., Shafiq Antonacci, R., Maruff, P., Tyler, P., & Currie, J. (1999). Norms and the effects of demographic variables on a neuropsychological battery for use in healthy ageing Australian populations. Aust N Z J Psychiatry, 33 (4), 568 5 75. Cooper, J. A., Sagar, H. J., Doherty, S. M., Jordan, N., Tidswell, P., & Sullivan, E. V. (1992). Different effects of dopaminergic and anticholinergic therapies on cognitive and motor function in Parkinson's disease. A follow up study of untreated pati ents. Brain, 115 ( Pt 6) 1701 1725. Czernecki, V., Pillon, B., Houeto, J. L., Pochon, J. B., Levy, R., & Dubois, B. (2002). Motivation, reward, and Parkinson's disease: influence of dopatherapy. Neuropsychologia, 40 (13), 2257 2267. Damasio, A. R. (1995). On some functions of the human prefrontal cortex. Ann N Y Acad Sci, 769 241 251. Damasio, A.R., & Van Hoesen, G.W. (1983). Emotional disturbances associated with focal lesions of the limbic frontal lobe. In K.M. Heilman & P. Satz (Eds.), Neuropsychology o f Human Emotion (pp. 85 110). New York: Guilford. Davidson, R. J., & Irwin, W. (1999). The functional neuroanatomy of emotion and affective style. Trends Cogn Sci, 3 (1), 11 21. Downes, J. J., Roberts, A. C., Sahakian, B. J., Evenden, J. L., Morris, R. G., & Robbins, T. W. (1989). Impaired extra dimensional shift performance in medicated and unmedicated Parkinson's disease: evidence for a specific attentional dysfunction. Neuropsychologia, 27 (11 12), 1329 1343.
Dozis, D. J. A., Dobson, K.S., Ahnberg, J. L (1 998). A Psychometric Evaluation of the Beck Depression Inventory II. Psychological Assessment, 10 (2) 83 89. Dubois, B., Burn, D., Goetz, C., Aarsland, D., Brown, R. G., Broe, G. A., et al. (2007). Diagnostic procedures for Parkinson's disease dementia: r ecommendations from the movement disorder society task force. Mov Disord, 22 (16), 2314 2324. Dubois, B., & Pillon, B. (1997). Cognitive deficits in Parkinson's disease. J Neurol, 244 (1), 2 8. Dujardin, K., Sockeel, P., Delliaux, M., Destee, A., & Defebvre, L. (2008). The Lille Apathy Rating Scale: validation of a caregiver based version. Mov Disord, 23 (6), 845 849. Dujardin, K., Sockeel, P., Devos, D., Delliaux, M., Krystkowiak, P., Destee, A., et al. (2007). Characteristics of apathy in Parkinson's disease Mov Disord, 22 (6), 778 784. Dyer, F. N. (1973). Interference and facilitation for color naming with separate bilateral presentations of the word and color. J Exp Psychol, 99 (3), 314 317. Elias, J. W., & Treland, J. E. (1999). Executive function in Parkin son's disease and subcortical disorders. Semin Clin Neuropsychiatry, 4 (1), 34 40. Factors impacting on quality of life in Parkinson's disease: results from an international survey. (2002). Mov Disord, 17 (1), 60 67. Fahn, S., Elton R. L., & Members of the U PDRS Development Committee (1987). Unified (Eds.), Recent Developments in Parkinsons Disease: Vol. 2. (pp.153 163). New Jersey: Macmillan Health Care Information. Fie ld, A. (2005). Chapter 5: Regression. Discovering Statistics Using SPSS, Second edition (pp. 174 175). Thousand Oaks, CA: SAGE Publications Inc. Fernandez, H. H., Bowers, D., Triggs, W., Perlstein, W., Hass, C., Zahodne, L., Sozda, C., Rodriguez, R.L., Ma laty, I.A., Wu, S.S. (2009). Repetitive transcranial magnetic stimulation (rTMS) for the treatment of apathy in Parkinson's disease: results from a double blind, sham controlled, randomized, controlled trial. Submitted to American Academy of Neurology meet ing 2009. Galynker, I., Ieronimo, C., Miner, C., Rosenblum, J., Vilkas, N., & Rosenthal, R. (1997). Methylphenidate treatment of negative symptoms in patients with dementia. J Neuropsychiatry Clin Neurosci, 9 (2), 231 239. Green, M. F., Kern, R. S., Braff, D. L., & Mintz, J. (2000). Neurocognitive deficits and functional outcome in schizophrenia: are we measuring the "right stuff"? Schizophr Bull, 26 (1), 119 136. Groenewegen, H. J., Wright, C. I., & Beijer, A. V. (1996). The nucleus accumbens: gateway for li mbic structures to reach the motor system? Prog Brain Res, 107 485 511.
Growdon, J. H., Corkin, S., & Rosen, T. J. (1990). Distinctive aspects of cognitive dysfunction in Parkinson's disease. Adv Neurol, 53 365 376. Heaton, R. K., Miller, S. W., Taylor, M. J., Grant, I. (2004). Revised comprehensive norms for an expanded Halstead Reitan battery: demographically adjusted neuropsychological norms for African American and Caucasian adults. Lutz, Florida: Psychological Assessment Resources, Inc. Hobson, P. & Meara, J. (2004). Risk and incidence of dementia in a cohort of older subjects with Parkinson's disease in the United Kingdom. Mov Disord, 19 (9), 1043 1049. Hughes, A. J., Ben Shlomo, Y., Daniel, S. E., & Lees, A. J. (1992). What features improve the a ccuracy of clinical diagnosis in Parkinson's disease: a clinicopathologic study. Neurology, 42 (6), 1142 1146. Hughes, A. J., Daniel, S. E., Kilford, L., & Lees, A. J. (1992). Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico patho logical study of 100 cases. J Neurol Neurosurg Psychiatry, 55 (3), 181 184. Hughlings Jackson, J. (1931). Selected Writings. Philadelphia, PA: Hodder & Stoughton. Isella, V., Melzi, P., Grimaldi, M., Iurlaro, S., Piolti, R., Ferrarese, C., et al. (2002). C linical, neuropsychological, and morphometric correlates of apathy in Parkinson's disease. Mov Disord, 17 (2), 366 371. Judd, L. L., Paulus, M. P., Wells, K. B., & Rapaport, M. H. (1996). Socioeconomic burden of subsyndromal depressive symptoms and major de pression in a sample of the general population. Am J Psychiatry, 153 (11), 1411 1417. Jurica, P.J., Leitten C.J. & Mattis, S. (2001). Dementia Rating Scale 2 : Professional manual Psychological Assessment Resources, Lutz, FL. Kant, R., Duffy, J. D., & Pivov arnik, A. (1998). Prevalence of apathy following head injury. Brain Inj, 12 (1), 87 92. Kaufer, D. I. (1999). Cholinergic therapy for neuropsychiatric symptoms in neurologic disorders. Curr Psychiatry Rep, 1 (1), 78 84. Kay, D., Kirsch Darrow, L., Zahodne, L ., Okun, M.S., Fernandez, H.H., Bowers, D. (2009) Dimensions of apathy in Parkinson's disease: Apathy Scale revisited. Submitted to American Academy of Neurology meeting 2009. Kellison, I., Kirsch Darrow, L., Lopez, M., Crucian, G., Do, T., Fernandez, H., Okun, M., & Bowers, D. (August 2006). Apathy and Executive Function in Non demen Disease. P resented at the American Psychological Association 2006 Convention. Kirsch Darrow, L., Fernandez, H. H., Marsiske, M., Okun, M. S., & Bowers, D. (20 06). Dissociating apathy and depression in Parkinson disease. Neurology, 67 (1), 33 38.
Kirsch Darrow, L., Zahodne, L. B., Hass, C., Mikos, A., Okun, M. S., Fernandez, H. H., et al. (2009). How cautious should we be when assessing apathy with the Unified Pa rkinson's Disease Rating Scale? Mov Disord, 24 (5), 684 688. Kraus, M. F., & Maki, P. M. (1997). Effect of amantadine hydrochloride on symptoms of frontal lobe dysfunction in brain injury: case studies and review. J Neuropsychiatry Clin Neurosci, 9 (2), 222 230. Kuzis, G., Sabe, L., Tiberti, C., Leiguarda, R., & Starkstein, S. E. (1997). Cognitive functions in major depression and Parkinson disease. Arch Neurol, 54 (8), 982 986. Leentjens, A. F., Dujardin, K., Marsh, L., Martinez Martin, P., Richard, I. H., St arkstein, S. E., et al. (2008). Apathy and anhedonia rating scales in Parkinson's disease: critique and recommendations. Mov Disord, 23 (14), 2004 2014. Leentjens, A. F., Verhey, F. R., Luijckx, G. J., & Troost, J. (2000). The validity of the Beck Depressio n Inventory as a screening and diagnostic instrument for depression in patients with Parkinson's disease. Mov Disord, 15 (6), 1221 1224. Lees, A. J., & Smith, E. (1983). Cognitive deficits in the early stages of Parkinson's disease. Brain, 106 (Pt 2) 257 2 70. Levin, B. E., & Katzen, H. L. (2005). Early cognitive changes and nondementing behavioral abnormalities in Parkinson's disease. Adv Neurol, 96 84 94. Levin, B. E., Llabre, M. M., & Weiner, W. J. (1988). Parkinson's disease and depression: psychometric properties of the Beck Depression Inventory. J Neurol Neurosurg Psychiatry, 51 (11), 1401 1404. Levin, B. E., Llabre, M. M., & Weiner, W. J. (1989). Cognitive impairments associated with early Parkinson's disease. Neurology, 39 (4), 557 561. Levy, G., Schup f, N., Tang, M. X., Cote, L. J., Louis, E. D., Mejia, H., et al. (2002). Combined effect of age and severity on the risk of dementia in Parkinson's disease. Ann Neurol, 51 (6), 722 729. Levy, M. L., Cummings, J. L., Fairbanks, L. A., Masterman, D., Miller, B. L., Craig, A. H., et al. (1998). Apathy is not depression. J Neuropsychiatry Clin Neurosci, 10 (3), 314 319. Levy, R., & Dubois, B. (2006). Apathy and the functional anatomy of the prefrontal cortex basal ganglia circuits. Cereb Cortex, 16 (7), 916 928. L ezak, M.D. (1982). The problem of assessing executive functions. International Journal of Psychology, 17 281 297. Lezak, M.D., Howieson, D.B., & Loring, D.W. (2004). Chapter 15: Concept formation and reasoning. In Neuropsychological Assessment, Fourth ed ition (pp.586 591). New York, NY: Oxford University Press, Inc.
Lichter, D.G. (2000). Movement disorders and frontal subcortical circuits. In: Lichter, D.J., Cummings, J.L. Eds. Frontal subcortical Circuits in Psychiatric and Neurological Disorders. New York, NY: The Guilford Press, 260 313. Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling, 9 (2), 151 173. Litvan, I., Cummings, J. L., & Mega, M. (1998). Neuropsychiatric features of corticobasal degeneration. J Neurol Neurosurg Psychiatry, 65 (5), 717 721. Litvan, I., Paulsen, J. S., Mega, M. S., & Cummings, J. L. (1998). Neuropsychiatric assessment of patients with hyperkinetic and hypo kinetic movement disorders. Arch Neurol, 55 (10), 1313 1319. Lyness, J. M., Bruce, M. L., Koenig, H. G., Parmelee, P. A., Schulz, R., Lawton, M. P., et al. (1996). Depression and medical illness in late life: report of a symposium. J Am Geriatr Soc, 44 (2), 198 203. MacDonald, A. W., 3rd, Cohen, J. D., Stenger, V. A., & Carter, C. S. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science, 288 (5472), 1835 1838. Marin, R. S. (1991). Apathy: a neu ropsychiatric syndrome. J Neuropsychiatry Clin Neurosci, 3 (3), 243 254. Marin, R. S., Biedrzycki, R. C., & Firinciogullari, S. (1991). Reliability and validity of the Apathy Evaluation Scale. Psychiatry Res, 38 (2), 143 162. Marsh, L., McDonald, W. M., Cumm ings, J., & Ravina, B. (2006). Provisional diagnostic criteria for depression in Parkinson's disease: report of an NINDS/NIMH Work Group. Mov Disord, 21 (2), 148 158. Mayberg, H. S. (1994). Frontal lobe dysfunction in secondary depression. J Neuropsychiatry Clin Neurosci, 6 (4), 428 442. M ayberg, H.S., Keightley, M., Mahurin, R.K., & Brannon, S.K. (2002). Neuropsychiatric aspects of mood and affective disorders. In S.C. Yudofsky & R.E. Hales (Eds). Textbook of neuropsychiatry and clinical neurosciences Washi ngton, D.C.: American Psychiatric Publishing. Mayberg, H. S., Starkstein, S. E., Sadzot, B., Preziosi, T., Andrezejewski, P. L., Dannals, R. F., et al. (1990). Selective hypometabolism in the inferior frontal lobe in depressed patients with Parkinson's dis ease. Ann Neurol, 28 (1), 57 64. McDonald, W. M., Richard, I. H., & DeLong, M. R. (2003). Prevalence, etiology, and treatment of depression in Parkinson's disease. Biol Psychiatry, 54 (3), 363 375.
Menza, M., Dobkin, R. D., Marin, H., Mark, M. H., Gara, M., Buyske, S., et al. (2009). A controlled trial of antidepressants in patients with Parkinson disease and depression. Neurology, 72 (10), 886 892. Milner, B. (1975). Psychological aspects of focal epilepsy and its neurosurgical management. Adv Neurol, 8 299 321. Mogenson, G. J., Jones, D. L., & Yim, C. Y. (1980). From motivation to action: functional interface between the limbic system and the motor system. Prog Neurobiol, 14 (2 3), 69 97. Morris, R. G., Downes, J. J., Sahakian, B. J., Evenden, J. L., Heald, A ., & Robbins, T. W. (1988). Planning and spatial working memory in Parkinson's disease. J Neurol Neurosurg Psychiatry, 51 (6), 757 766. Muslimovic, D., Post, B., Speelman, J. D., & Schmand, B. (2005). Cognitive profile of patients with newly diagnosed Parki nson disease. Neurology, 65 (8), 1239 1245. Nussbaum, R. L., & Ellis, C. E. (2003). Alzheimer's disease and Parkinson's disease. N Engl J Med, 348 (14), 1356 1364. Nuti, A., Ceravolo, R., Piccinni, A., Dell'Agnello, G., Bellini, G., Gambaccini, G., et al. (2 004). Psychiatric comorbidity in a population of Parkinson's disease patients. Eur J Neurol, 11 (5), 315 320. Owen, A. M. (2004). Cognitive dysfunction in Parkinson's disease: the role of frontostriatal circuitry. Neuroscientist, 10 (6), 525 537. Owen, A. M. James, M., Leigh, P. N., Summers, B. A., Marsden, C. D., Quinn, N. P., et al. (1992). Fronto striatal cognitive deficits at different stages of Parkinson's disease. Brain, 115 ( Pt 6) 1727 1751. Owen, A. M., Roberts, A. C., Polkey, C. E., Sahakian, B. J ., & Robbins, T. W. (1991). Extra dimensional versus intra dimensional set shifting performance following frontal lobe excisions, temporal lobe excisions or amygdalo hippocampectomy in man. Neuropsychologia, 29 (10), 993 1006. Pardo, J. V., Pardo, P. J., Ja ner, K. W., & Raichle, M. E. (1990). The anterior cingulate cortex mediates processing selection in the Stroop attentional conflict paradigm. Proc Natl Acad Sci U S A, 87 (1), 256 259. Paulsen, J. S., Ready, R. E., Hamilton, J. M., Mega, M. S., & Cummings, J. L. (2001). Neuropsychiatric aspects of Huntington's disease. J Neurol Neurosurg Psychiatry, 71 (3), 310 314. Pedersen, K. F., Alves, G., Bronnick, K., Aarsland, D., Tysnes, O. B., & Larsen, J. P. (2009). Apathy in drug naive patients with incident Parkin son's disease: the Norwegian ParkWest study. J Neurol
Pedersen, K. F., Larsen, J. P., Alves, G., & Aarsland, D. (2009). Prevalence and clinical correlates of apathy in Parkinson's disease: a community based study. Parkinsonism Relat Disord, 15 (4), 295 299 Perry, R. J., & Hodges, J. R. (1996). Spectrum of memory dysfunction in degenerative disease. Curr Opin Neurol, 9 (4), 281 285. Phillips, P. (1999). Keeping depression at bay helps patients with Parkinson disease. JAMA, 282 (12), 1118 1119. Pillon, B., Cze rnecki, V., & Dubois, B. (2003). Dopamine and cognitive function. Curr Opin Neurol, 16 Suppl 2 S17 22. Pillon, B., Deweer, B., Agid, Y., & Dubois, B. (1993). Explicit memory in Alzheimer's, Huntington's, and Parkinson's diseases. Arch Neurol, 50 (4), 374 3 79. Pluck, G. C., & Brown, R. G. (2002). Apathy in Parkinson's disease. J Neurol Neurosurg Psychiatry, 73 (6), 636 642. Pontone, G. M., Williams, J. R., Anderson, K. E., Chase, G., Goldstein, S. A., Grill, S., et al. (2009). Prevalence of anxiety disorders and anxiety subtypes in patients with Parkinson's disease. Mov Disord, 24 (9), 1333 1338. Reijnders, J. S., Ehrt, U., Lousberg, R., Aarsland, D., & Leentjens, A. F. (2009). The association between motor subtypes and psychopathology in Parkinson's disease. P arkinsonism Relat Disord, 15 (5), 379 382. Remy, P., Doder, M., Lees, A., Turjanski, N., & Brooks, D. (2005). Depression in Parkinson's disease: loss of dopamine and noradrenaline innervation in the limbic system. Brain, 128 (Pt 6), 1314 1322. Robert, P., On yike, C. U., Leentjens, A. F., Dujardin, K., Aalten, P., Starkstein, S., et al. (2009). Proposed diagnostic criteria for apathy in Alzheimer's disease and other neuropsychiatric disorders. Eur Psychiatry, 24 (2), 98 104. Santangelo, G., Vitale, C., Trojano, L., Longo, K., Cozzolino, A., Grossi, D., et al. (2009). Relationship between depression and cognitive dysfunctions in Parkinson's disease without dementia. J Neurol, 256 (4), 632 638. Schinka, J. A., Vanderploeg, R. D., Rogish, M., & Ordorica, P. I. (2002 ). Effects of alcohol and cigarette use on cognition in middle aged adults. J Int Neuropsychol Soc, 8 (5), 683 690. Schwab, R.S.. & England, A.C. (1969). Projection technique for evaluating surgery in dson (Eds.), Third symposium on (pp. 152 157). Edinburg: Livingstone. Scuteri, A., Palmieri, L., Lo Noce, C., & Giampaoli, S. (2005). Age related changes in cognitive domains. A population based study. Aging Clin Exp Res, 17 ( 5), 367 373.
Shallice, T., & Burgess, P. W. (1991). Deficits in strategy application following frontal lobe damage in man. Brain, 114 ( Pt 2) 727 741. Sheline, Y. I., Barch, D. M., Garcia, K., Gersing, K., Pieper, C., Welsh Bohmer, K., et al. (2006). Cogn itive function in late life depression: relationships to depression severity, cerebrovascular risk factors and processing speed. Biol Psychiatry, 60 (1), 58 65. Slaughter, J. R., Slaughter, K. A., Nichols, D., Holmes, S. E., & Martens, M. P. (2001). Prevale nce, clinical manifestations, etiology, and treatment of depression in Parkinson's disease. J Neuropsychiatry Clin Neurosci, 13 (2), 187 196. Sockeel, P., Dujardin, K., Devos, D., Deneve, C., Destee, A., & Defebvre, L. (2006). The Lille apathy rating scale (LARS), a new instrument for detecting and quantifying apathy: validation in Parkinson's disease. J Neurol Neurosurg Psychiatry, 77 (5), 579 584. Spielberger, C. D., Gorsuch, R. L., & Luschene, R. (1970). T est Manual for the State Trait Anxiety Inventory Palo Alto, CA: Consulting Psychologists Press. Starkstein, S. E., & Leentjens, A. F. (2008). The nosological position of apathy in clinical practice. J Neurol Neurosurg Psychiatry, 79 (10), 1088 1092. Starkstein, S. E., Mayberg, H. S., Leiguarda, R., Prezio si, T. J., & Robinson, R. G. (1992). A prospective longitudinal study of depression, cognitive decline, and physical impairments in patients with Parkinson's disease. J Neurol Neurosurg Psychiatry, 55 (5), 377 382. Starkstein, S. E., Mayberg, H. S., Prezios i, T. J., Andrezejewski, P., Leiguarda, R., & Robinson, R. G. (1992). Reliability, validity, and clinical correlates of apathy in Parkinson's disease. J Neuropsychiatry Clin Neurosci, 4 (2), 134 139. Starkstein, S. E., & Merello, M. (2007). The Unified Park inson's Disease Rating Scale: validation study of the mentation, behavior, and mood section. Mov Disord, 22 (15), 2156 2161. Starkstein, S. E., Merello, M., Jorge, R., Brockman, S., Bruce, D., & Power, B. (2009). The syndromal validity and nosological posit ion of apathy in Parkinson's disease. Mov Disord, 24 (8), 1211 1216. Stroop, J.R. (1935).Studies of interference in serial verbal reactions. J Exp Psychol 18, 643 662. Stuss, D. T., Alexander, M. P., Hamer, L., Palumbo, C., Dempster, R., Binns, M., et al. ( 1998). The effects of focal anterior and posterior brain lesions on verbal fluency. J Int Neuropsychol Soc, 4 (3), 265 278. Stuss, D. T., Floden, D., Alexander, M. P., Levine, B., & Katz, D. (2001). Stroop performance in focal lesion patients: dissociation of processes and frontal lobe lesion location. Neuropsychologia, 39 (8), 771 786. Stuss, D. T., Shallice, T., Alexander, M. P., & Picton, T. W. (1995). A multidisciplinary approach to anterior attentional functions. Ann N Y Acad Sci, 769 191 211.
Tanner, C Jankovic, J.J., & Tolosa, E. (Eds.), 4 th edition. (pp. 90 103). Philadelphia, PA: Lippincott Williams & Wilkins. Taylor, A. E., & Sai nt Cyr, J. A. (1995). The neuropsychology of Parkinson's disease. Brain Cogn, 28 (3), 281 296. Taylor, A. E., Saint Cyr, J. A., & Lang, A. E. (1986). Frontal lobe dysfunction in Parkinson's disease. The cortical focus of neostriatal outflow. Brain, 109 ( Pt 5) 845 883. Troster, A. I., Stalp, L. D., Paolo, A. M., Fields, J. A., & Koller, W. C. (1995). Neuropsychological impairment in Parkinson's disease with and without depression. Arch Neurol, 52 (12), 1164 1169. Troyer, A. K., Moscovitch, M., & Winocur, G. (1997). Clustering and switching as two components of verbal fluency: evidence from younger and older healthy adults. Neuropsychology, 11 (1), 138 146. Uekermann, J., Daum, I., Peters, S., Wiebel, B., Przuntek, H., & Muller, T. (2003). Depressed mood and ex ecutive dysfunction in early Parkinson's disease. Acta Neurol Scand, 107 (5), 341 348. v an Reekum, R., Bayley, M., Garner, S., Burke, I. M., Fawcett, S., Hart, A., et al. (1995). N of 1 study: amantadine for the amotivational syndrome in a patient with trau matic brain injury. Brain Inj, 9 (1), 49 53. van Reekum, R., Stuss, D. T., & Ostrander, L. (2005). Apathy: why care? J Neuropsychiatry Clin Neurosci, 17 (1), 7 19. Visser, M., Leentjens, A. F., Marinus, J., Stiggelbout, A. M., & van Hilten, J. J. (2006). Rel iability and validity of the Beck depression inventory in patients with Parkinson's disease. Mov Disord, 21 (5), 668 672. Weintraub, D., Moberg, P. J., Culbertson, W. C., Duda, J. E., Katz, I. R., & Stern, M. B. (2005). Dimensions of executive function in P arkinson's disease. Dement Geriatr Cogn Disord, 20 (2 3), 140 144. Williams Gray, C. H., Foltynie, T., Brayne, C. E., Robbins, T. W., & Barker, R. A. (2007). Evolution of cognitive dysfunction in an incident Parkinson's disease cohort. Brain, 130 (Pt 7), 178 7 1798. Zahodne, L. B., Kirsch Darrow, L., Fernandez, H. H., Okun, M. S., & Bowers, D. (2009). Divergent trajectories of apathy and depression over two years in medically managed idiopathic Parkinson's disease. Submitted to American Academy of Neurology meeting 2009.
Zahodne, L. B., Young, S., Kirsch Darrow, L., Nisenzon, A., Fernandez, H. H., Okun, M. S., et al. (2009). Examination of the Lille Apathy Rating Scale in Parkinson disease. Mov Disord, 24 (5), 677 683. Zajano, M.J., & Gorman, A. (1986). Stroo p interference as a functionof percentage of congruent items. Perceptual and Motor Skills 63 1087 1096. Zgaljardic, D. J., Borod, J. C., Foldi, N. S., & Mattis, P. (2003). A review of the cognitive and behavioral sequelae of Parkinson's disease: relation ship to frontostriatal circuitry. Cogn Behav Neurol, 16 (4), 193 210. Zgaljardic, D. J., Borod, J. C., Foldi, N. S., Rocco, M., Mattis, P. J., Gordon, M. F., et al. (2007). Relationship between self reported apathy and executive dysfunction in nondemented p atients with Parkinson disease. Cogn Behav Neurol, 20 (3), 184 192.
145 BIOGRAPHICAL SKETCH Lindsey Kirsch Darrow was born in Atlanta, Georgia and received her B.S. in n euroscience from Furman University. During college, she obtained research experience at the Centers for Disease Control and Prevention in Atlanta, GA. Currently, she is a doctoral candidate in Clinical & Health Psychology at the University of Florida, where she is specializing in neuropsychology. She has accepted a position as a neuropsych ology intern at the University of Alabama/Vetera ns Affairs Medical Center Consortium Site in Birmingham, AL (September 2009 thesis, originally presented at the Amer ican Academy of Neurolog y meeting in Miami, FL in 2005, under the mentorship of Dr. Dawn Bowers. This project was funded by a National Institutes of Health N ational Research Service Award.