Implications of Apathy for Everyday Functioning Outcomes in Persons Living with HIV Infection

Abstract

Apathy is a relatively common clinical feature of HIV-Associated Neurocognitive Disorders, but little is known about its implications for everyday functioning outcomes. In the present study, we examined the associations between apathy and self-reported instrumental activities of daily living (IADL) and neurocognitive complaints in 75 participants with HIV infection and 52 demographically comparable seronegative comparison subjects. All volunteers completed the apathy subscale of the Frontal Systems Behavioral Scale as part of a comprehensive neuromedical, psychiatric, and neurocognitive research evaluation. When compared with the seronegative comparison participants, the HIV+ group reported significantly higher current levels of apathy, but did not differ in self-report of prior (i.e., pre-seroconversion) apathy. Higher current apathy self-ratings were associated with greater severity of IADL declines and more numerous cognitive complaints in the HIV+ sample, even after adjusting for potential psychiatric (e.g., depression), medical (e.g., hepatitis C co-infection), and neurocognitive predictors. Cognitive complaints, but not IADLs, were also uniquely associated with ratings of executive dysfunction and disinhibition. All told, these findings suggest that apathy may make a unique contribution to important everyday functioning outcomes among persons living with HIV infection. The clinical detection of apathy may help identify HIV-infected individuals at particular risk for functional impairments who may require additional support to maintain independence.

Introduction

HIV infection is commonly accompanied by a wide range of neuropsychiatric complications, including depression (e.g., Atkinson et al., 1988; Ciesla & Roberts, 2001), mania (e.g., Lyketsos, Hanson, Fishman, & Rosenblatt, 1993), and anxiety (Elliott, 1998). Although apathy has long been recognized as a clinical manifestation of HIV-Associated Neurocognitive Disorders (HAND; e.g., Navia, Cho, Petito, & Price, 1986), only in the past decade research has emerged on the prevalence, correlates, and neural underpinnings of this important neuropsychiatric symptom in the setting of HIV infection (Castellon, Hinkin, Wood, & Yarema, 1998; Castellon, Hinkin, & Myers, 2001; Paul, Flanigan et al., 2005; Rabkin et al., 2000). Apathy refers to a cluster of symptoms that include a reduction in self-initiated, goal-directed behavior, and a lack of motoric, emotional, and cognitive motivation (Marin, 1997). Approximately 30–50% HIV-infected persons demonstrate clinically elevated signs of apathy (e.g., Castellon et al., 1998, 2000; Rabkin et al., 2000). The presence and severity of apathy appears to be largely unrelated to HIV disease severity, including CD4 cell counts, HIV RNA viral load in plasma, and AIDS status (e.g., Castellon et al., 1998; Paul, Flanigan et al., 2005; Rabkin et al., 2000), with the possible exception of longer estimated duration of infection (Paul, Flanigan et al., 2005).

Apathy and depression, often characterized by symptoms such as dysphoria and anhedonia, commonly co-occur in HIV infection, which may be expected given the overlap in their clinical presentation (e.g., diminished interest and loss of motivation). In fact, apathy and depression are significantly associated with one another in HIV-infected cohorts, with at least two studies reporting moderately strong, positive correlations between self-reported apathy and depression scores (Castellon et al., 1998; Rabkin et al., 2000). Nevertheless, there is ample evidence of the separability of these related constructs. Excluding items common to both apathy and depression from the Hamilton Rating for Depression and Beck Depression Inventory resulted in an attenuation of their associations (Rabkin et al., 2000). This suggests that while apathy and depression are co-morbid in HIV infection, there may not be a one-to-one correspondence between the two constructs. Three studies also reported that apathy ratings of HIV+ individuals do not co-vary with depressed mood (Paul, Brickman et al., 2005; Paul, Flanigan et al., 2005; Tate et al., 2003), while another study found that elevated levels of apathy may be present in nondepressed HIV-infected persons when compared with seronegatives (Hoare et al., 2010). This is consistent with the dissociation between these two neuropsychiatric syndromes demonstrated by research in Parkinson's disease (PD) and Alzheimer's disease (AD). Findings in these patient populations suggest that these mood disturbances may occur alone or concurrently and may have independent neurophysiological bases (Ott, Noto, & Fogel, 1996; Starkstein et al., 1992). Imaging studies across clinical groups (including HIV-infected persons) show evidence of left prefrontal and limbic system dysregulation being associated with depression, whereas apathy is thought to be driven by medial prefrontal and deep subcortical pathology (Hoare et al., 2010; Paul, Brickman et al., 2005; Starkstein et al., 1992).

Although HIV-associated neural injury is not specific to any one brain region, HAND is reliably associated with damage to frontostriatal circuits, including the anterior cingulate circuit (Jernigan et al., 2011). Lesions to these circuits, which include the ventral striatum and are thought to be involved in emotion regulation and initiation of behavior, are known to produce prominent symptoms of apathy (Cummings, 1993; Tekin & Cummings, 2002). To date, two neuroimaging studies have examined the correspondence between apathy and HIV-associated neural injury. Paul, Brickman et al., (2005) reported a significant association between apathy and lower nucleus accumbens volumes, which was independent of HIV disease severity and depression ratings. Similarly, in a study using diffusion tensor imaging, lower fractional anisotropy in the medial frontal brain regions (e.g., corpus callosum and corona radiata) was observed in the apathetic HIV+ individuals (Hoare et al., 2010). Together, these data provide direct evidence that HIV-associated abnormalities within prefrontostriatal circuits are uniquely associated with higher levels of apathy.

Given the overlapping neural circuits involved in apathy and HIV-associated neuropathology and the above-described evidence for their shared variance, one might predict that higher levels of apathy would correspond to greater risk of HIV-associated neurocognitive deficits. However, the literature has yielded mixed findings in that regard. Small, nonsignificant correlations between apathy and a wide range of cognitive abilities were reported in two previous studies (Rabkin et al., 2000; Robinson-Papp et al., 2008). Other studies suggest modest associations between apathy and deficits in episodic memory, cognitive set-switching (Paul, Flanigan et al., 2005), conflict resolution, working memory, and speed of information processing (Castellon et al., 1998, 2000). Thus, the current body of literature indicates that apathy is prevalent in HIV infection and is associated with HIV-related neural injury, but it is likely separable from depression and neurocognitive functioning and may therefore be worthy of clinical consideration.

For example, apathy and loss of motivation may affect the performance on many aspects of instrumental activities and interpersonal functions (Cummings & Mega, 2003). Approximately 35–50% of individuals with HAND experience significant functional impairments (e.g., Heaton et al., 2004). Neurocognitive deficits, psychiatric symptoms, and medical factors (e.g., HIV disease severity) are notable predictors of functional deficits in this population (see Gorman, Foley, Ettenhofer, Hinkin, & van Gorp, 2009 for review). Yet only three studies have evaluated the potential role of apathy in the expression of everyday functioning difficulties in HIV. With regard to antiretroviral (ARV) medication management, apathy has been linked to greater perceived barriers to adherence, poor interpersonal support, lower intentions to adhere, and lower perceived treatment utility (Rabkin et al., 2000). Using Medication Event Monitoring System caps, Barclay, Hinkin and Castellon (2007) reported that higher levels of apathy were associated with ARV nonadherence in younger, but not older HIV+ individuals. In both of these medication-focused studies, the strength and independence of the apathy effects were dampened after controlling for other predictors such as health beliefs, depression, and substance use. Higher levels of apathy in HIV+ persons have also been reported to be associated with poorer health-related quality of life, specifically in the emotional role functioning and mental health domains. Although the magnitude of association was somewhat weaker than that noted for depressive symptoms, apathy remained a significant predictor of lower quality of life after controlling for mood (Tate et al., 2003).

While prior data illustrate that apathy may play at least a modest role in certain critical aspects of the daily lives of persons infected with HIV, its potential unique contribution to other important functional outcomes (e.g., vocational performance, automobile driving, and other instrumental and basic activities of daily living) remains to be studied. In the present study, we therefore aimed to determine the association between apathy and self-reported declines in instrumental activities of daily living and neurocognitive complaints in HIV+ persons. It was specifically hypothesized that elevated levels of current apathy would uniquely predict decline in instrumental activities of daily living and severity of neurocognitive complaints, independent of other known variables associated with these functional outcomes (e.g., depression, neurocognitive deficits).

Methods

Participants

Study participants included 75 HIV+ individuals recruited from HIV treatment centers and community organizations and 52 seronegative comparison subjects who were recruited from community advertisements and word-of-mouth (e.g., friends and family of the seropositive participants). These data were derived retrospectively from a NIDA-funded program project that also recruited individuals who met criteria for methamphetamine dependence within the past year. Given the known influence of chronic methamphetamine use on frontal systems (see Scott et al., 2007 for review), including apathy (Cattie et al., in press) these individuals were excluded from this current study. Individuals with histories of neurological diseases (e.g., seizure disorders, closed head injuries with loss of consciousness greater than 30 min, central nervous systems neoplasms, opportunistic infections) or severe psychiatric illnesses (e.g., schizophrenia) that might affect cognitive functioning were also excluded from the study. The HIV status was determined by enzyme-linked immunosorbent assays and was confirmed by a western blot test. The demographic, clinical, neurocognitive, and psychiatric characteristics of the study sample are displayed in Table 1. Participants were well matched on demographic variables (p > .10), with the exception that the HIV+ group had a larger proportion of Caucasians (p < .01). The seronegative group had lower rates of Major Depressive Disorder, Attention-Deficit/Hyperactivity Disorder (ADHD), and neuropsychological impairment (p < .05).

Table 1.

Sample demographic and clinical characteristics

	HIV− (n = 52)	HIV+ (n = 75)
Demographic variables
Age, years	42.7 (11.6)	45.3 (8.3)
Education, years	13.2 (2.2)	13.1 (1.96)
Sex (%men)	87	89
Ethnicity (% Caucasian)**	44	66
Psychiatric characteristics
Lifetime (current and past) Major depressive disorder (%)*	31	48
ADHD diagnosis (%)*	1	10
Apathy (before), T-score	49.4 (16.6)	53.5 (18.2)
Apathy (after), T-score***	48.3 (18.5)	63.1 (19.75)
Disinhibition (before), T-score*	45.94 (1.84)	51.78 (1.50)
Disinhibition (after), T-score **	45.06 (1.84)	53.13 (1.51)
Executive dysfunction (before), T-score**	50.14 (2.54)	57.69 (2.70)
Executive dysfunction (after), T-score***	48.33 (2.46)	60.75 (2.02)
Correlation between pre-/post-apathy ratings***	–	.50
Correlation between pre-/post-disinhibition ratings***	–	.64
Correlation between pre-/post-executive dysfunction ratings***	–	.64
Substance use history
Lifetime substance use disorder*	15	31
Neuropsychological and functional outcome variables
Proportion NP impaired (%)*	28.8	47.3
IADL complaints, total***	0.5 (1.1)	1.62 (2.1)
PAOFI score, total*	2.32 (5.89)	4.56 (5.04)
HIV disease characteristics
HIV RNA log10 (median, IQR)	–	1.7 [1.7–3.2]
Plasma viral load (% undetectable)	–	65%
Nadir CD4 count (median, IQR)	–	184 [43.5–309.8]
Current CD4 (median, IQR)	–	446.5 [321–607.8]
Proportion on HAART (%)	–	77%
Proportion with AIDS (%)	–	61
Proportion with HCV (%)	–	31

*p < .05.

**p < .01.

***p < .001.

Procedure

Participants completed the 14-item apathy subscale of the self-report version of the Frontal Systems Behavioral Scale (FrSBe; Grace & Malloy, 2001). Ratings are on a Likert-type scale that ranges from 1 (“almost never”) to 5 (“almost always”) for each question. Higher ratings indicate more abnormal behavior (e.g., neglecting personal hygiene, leaving things unfinished). Following the manual guidelines, the raw scores are converted into demographically (i.e., age, education, and gender) adjusted T scores. A T-score cut point of 65 or higher is recommended to capture greater degrees of symptomatology and clinical significance. The scale allows for retrospective ratings prior to the injury or illness (before) and for ratings following the injury or illness (after), creating a baseline measure with which to compare subsequent ratings (Grace & Malloy, 2001). Participants were asked to provide retrospective ratings prior to the diagnosis of HIV infection (before) and current ratings (i.e., after seroconversion). Seronegative participants were asked to rate their behavior before age 20 and at present. The FrSBe T-scores (before and after) for the disinhibition and executive dysfunction subscales scores were also used in the analyses. In the HIV+ group, the pre- and post-self-ratings for each subscale were strongly correlated, but not collinear by traditional definitions (see Table 1).

All participants completed a version of the Lawton and Brody (1969) Activities of Daily Living scale, which was modified to enable the detection of temporal changes in ADL dependence at a cross-sectional visit by requiring subjects to rate their “current” and “best” prior level of functioning on numerous daily tasks (e.g., grocery shopping, managing finances, housekeeping; see Heaton et al., 2004). Consistent with prior research (e.g., Woods et al., 2008), the primary dependent variable of interest was the total number of declines reported in current versus past functioning on all of the IADL items [possible range = 0 (no decline) − 13 (increased dependence in all activities)].

Participants also completed the Patient's Assessment of Own Functioning Inventory (PAOFI; Chelune, Heaton, & Lehman, 1986), which is a 41-item questionnaire in which they reported the frequency of difficulties with memory, language and communication, use of hands, sensory-perception, higher level cognitive and intellectual functions, work and recreation. The total number of cognitive complaints (possible range of 0–33) was used as the dependent variable.

Trained interviewers administered the Composite International Diagnostic Interview (CIDI) to the participants. The CIDI is a computer-assisted interview that provides a cross-cultural assessment of alcohol, drug, and mental disorders using DSM–IV criteria (Wittchen et al., 1991). The CIDI was used to obtain diagnoses of current and lifetime major depressive disorder and of psychoactive substance abuse and dependence disorders.

Participants were administered a standardized neuropsychological battery that included tests of executive functions, learning, memory, speed of information processing, verbal fluency, motor skills, and working memory (see Table 2 for tests in each domain). Raw scores were converted to demographically adjusted T-scores (e.g., Heaton et al., 2004), which were converted to deficit scores (range = 0 [T > 39] to 5 [T < 20]), which were then averaged to generate the Global Deficit Score (GDS; Carey et al., 2004). A GDS of ≥0.5 is the standard cutpoint indicative of global neurocognitive impairment. T-scores for the seven domains assessed in both study groups are presented in Table 2.

Table 2.

Neurocognitive functioning in the HIV seronegative and seropositive groups

Domain	Tests	Average domain T-score		Cohen's d	p value
		HIV− (n = 52)	HIV+ (n = 75)
Executive functions	Category Test, WCST-64, TMT Part B, Stroop Color and Word Test	50.70 (8.07)	46.30 (11.44)	0.45	.02
Verbal fluency	FAS Letter Fluency, Category Fluency-Animals	50.29 (7.94)	48.56 (8.61)	0.21	.25
Attention/working memory	WAIS–III Letter–Number Sequencing, Paced Auditory Serial Addition Task	49.95 (8.63)	48.15 (10.06)	0.19	.30
Motor skills	Grooved Pegboard Test	48.61 (9.90)	46.25 (10.53)	0.04	.20
Speed of information processing	WAIS–III Digit Symbol and Symbol Search, TMT Part A	52.69 (9.21)	49.83 (9.06)	0.31	.08
Learning	HVLT-R Total, BVMT–R Total	47.32 (8.29)	46.16 (7.94)	0.14	.43
Recall	HVLT–R Delay Recall, BVMT–R Delay Recall	46.48 (8.41)	45.97 (8.85)	0.06	.74

Notes: WCST-64= Wisconsin Card Sorting Test-64; WAIS-III= Wechsler Adult Intelligence Scale–III; TMT= Trail Making Test; HVLT-R= Hopkins Verbal Learning Test–Revised; BVMT-R= Brief Visuospatial Memory Test–Revised.

Data Analysis

A repeated-measures ANOVA was used to test for an interaction between HIV serostatus (between-subjects) and pre- and post-HIV infection apathy ratings (within-subjects). The prevalence of clinically elevated apathy in the HIV+ and seronegative groups was also examined with a chi-square test. Linear multiple regression analyses were then conducted to determine whether any of the demographic (i.e., ethnicity), psychiatric (i.e., ADHD and lifetime (i.e., current and prior) history of MDD diagnosis), cognitive, and substance use (i.e., lifetime history of any substance use disorder) factors on which the HIV seropositives and seronegatives differed confounded the HIV effect on the current apathy ratings. Next, within the HIV+ group, a repeated-measures approach was used to examine the association between functional complaints and pre- and post-apathy ratings. This approach allowed us to determine whether the association between post-ratings and functional outcomes was statistically larger than that with the pre-ratings (i.e., by virtue of the interaction term and planned follow-up correlations). The alternative analysis in which IADL and PAOFI complaints would serve as the criterion variables and pre-/post-apathy as the predictors would have addressed the contribution of post-apathy ratings in predicting functional outcomes above and beyond that of pre-ratings, which was not our primary question of interest. Moreover, the latter approach would be plagued by increased risk of Type II error due to the strong intercorrelations between pre- and post-ratings, which are better handled by grouping them in a within-subjects fashion. Finally, follow-up multivariable analyses (e.g., Katz, 2006; Peters, 2008) were conducted in which each functional outcome was predicted uniquely by current apathy ratings when considered alongside possible confounding variables (i.e., global neurocognitive impairment, lifetime MDD diagnosis, lifetime history of any substance use disorder, AIDS status, current CD4 count, and HCV serostatus). Results were unchanged when the current MDD was included in the models instead of lifetime MDD. Similar multivariable models using current ratings of disinhibition and executive dysfunction were also analyzed to address the issue of the specificity of the association between apathy and functional outcomes.

Results

HIV Serostatus and Apathy

Repeated-measures ANOVA was conducted with HIV serostatus as the between-subject variable and “before” and “after” apathy ratings as the within-subject variable. There was a main effect of HIV status, F(1, 124) = 9.71, p < .01, with greater levels of apathy reported by HIV+ participants. There was also a main effect of pre- and post-apathy ratings, F(1, 124) = 9.33, p < .01, such that apathy levels increased over time. These effects were tempered by an HIV serostatus by time interaction, F(1, 124) = 13.83, p < .001. Planned post-hoc analyses showed that HIV+ individuals did not differ significantly from their seronegative counterparts for apathy ratings “before,” t(124) = 1.31, p > .1, d = 0.20. Significant group differences were noted on the apathy “after” ratings, t(125) = 4.28, p < .0001, d = 0.76 with HIV+ participants reporting higher levels of apathy (see Table 1).

Multivariable regression with HIV serostatus and confounders (i.e., ADHD diagnosis, lifetime history of any substance use disorder, lifetime diagnosis of MDD, global cognitive impairment, and ethnicity) showed that HIV infection (β = 0.31, t(117) = 3.80, p < .001) and lifetime diagnosis of MDD (β = 0.41, t(117) = 5.02, p < .001) were uniquely associated with higher ratings of current apathy. The full model accounted for a significant amount of variance in the criterion, adjusted R² = .23, p < .0001.

Using the standard FrSBe cutpoint of T > 65, 42% of the HIV+ individuals were classified as apathetic at the current time-point, whereas 19% of the seronegative group fell in this range (χ² = 7.74, p = .005). Odds ratio revealed that after adjusting for depression and other covariates (i.e., ADHD diagnosis, lifetime history of substance use disorder, global cognitive impairment, and ethnicity), HIV+ individuals were approximately three times more likely to endorse clinically elevated levels of current apathy when compared with seronegative subjects (odds ratio = 3.1, 95% confidence interval = 1.35, 7.12).

Apathy and IADLs

Next, in the HIV+ group, to capture the association between current IADL complaints and the change in pre- and post-ratings of apathy, we conducted a repeated-measures model with pre- and post-apathy ratings as the within-subject variable and total number of IADL complaints as the predictor. This analytic approach provided a rigorous method to determine the relative strength of associations between ADLs and apathy. A trend-level main effect of IADL complaints was detected (F(1, 73) = 3.5, p = .07), along with a significant main effect of pre- and post-apathy ratings, F(1, 73) = 6.21, p = .01. An interaction was observed, F(1, 73) = 6.0, p = .02, such that the number of IADL complaints were significantly correlated with apathy “after” ratings (r = .30, p < .01), but not apathy “before” ratings (r = .07, p > .1). A follow-up multiple regression analysis was conducted with IADL complaints as the criterion and apathy rating as a predictor, which allowed us to determine whether post/current apathy was uniquely associated with ADL when considered alongside possible confounding variables (i.e., global neurocognitive impairment, lifetime MDD diagnosis, lifetime history of any substance use disorder, AIDS status, current CD4 count, and HCV serostatus). As shown in Table 3, the model was significant (adjusted R² = .15, F(7, 63) = 2.70, p = .02) and revealed that current apathy rating (β = 0.27, p = .04) and lifetime MDD diagnosis (β = 0.27, p = .03) were the only significant, independent predictors of IADL complaints (all other p > .10).

Table 3.

Clinical predictors of decline in activities of daily living in the HIV+ sample (n = 75)

Variable	Model	B	B 95% CI	Parameter (β)	p-value
Adjusted R2	0.15
F	2.70				.016
Apathy T-score		0.03	0.001, 0.053	0.27	.04
LT-MDD		−0.56	−1.06, −0.06	−0.27	.03
LT-SUD		0.11	−0.47, 0.69	0.05	.71
Global cognitive impairment		0.18	−0.30, 0.65	0.09	.46
AIDS status		−0.43	−0.95, 0.08	−0.02	.10
Current CD4 count		3.51e−5	−0.002, 0.002	0.004	.97
HCV status		−0.36	−0.93, 0.21	−0.16	.22

Notes: LT-MDD= Lifetime Major Depressive Disorder diagnosis; LT-SUD= Lifetime Substance Use Disorder diagnosis; HCV= Hepatitis-C virus.

As shown in Table 4, the highest rates of dependence were reported in the domains of housekeeping, social activities, and comprehension of verbal and written material. A majority of individuals also reported a decline in the work status. We also examined whether apathy was associated with declines in individual domains of everyday functioning. Nominal regression analyses were conducted with the dependence status for each domain on the IADL questionnaire as the dependent variable, and apathy as well as psychiatric (i.e., lifetime MDD and substance use disorder diagnoses), cognitive, and disease characteristics (e.g., AIDS status, current CD4 count, and HCV serostatus) entered as predictors. Three significant models were obtained and apathy was a significant independent predictor of dependence in social activities (χ² = 7.24, p = .007), shopping (χ² = 0.09, p = .01), and comprehension of written/verbal material (χ² = 7.24, p = .007).

Table 4.

Frequency of self-reported declines from “best” to “now” in activities of daily living and number of complaints in domains of cognitive functioning in 75 individuals with HIV infection

Activities of Daily Living	% declined
Work	63
Housekeeping (cleaning)	23
Social Activities	18
Comprehension of reading/TV materials	18
Home repairs	15
Grocery shopping	12
Laundry	12
Cooking	11
Shopping	9
Financial management	8
Telephone use	5
Bathing	5
Medication management	4
Transportation	3
Dressing	1
Cognitive Complaints	Median [IQR]
Memory	1 [0, 4]
Language	1 [0, 3]
Motor	0 [0, 0]
Sensory	0 [0, 0]
General cognition	0 [0, 1]

Apathy and Cognitive Complaints

Similar analyses were conducted to examine the relationship between apathy ratings and total cognitive complaints on the PAOFI. A repeated-measures model was conducted with pre- and post-apathy ratings as the within-subject variable and total number of PAOFI complaints as the predictor. A significant main effect of PAOFI complaints was detected (F(1, 67) = 18.16, p < .001), but there was no main effect for pre- and post-apathy ratings, F(1, 67) = 2.15, p > .1. A significant interaction was observed, F(1, 67) = 13.45, p < .001; apathy “before” ratings were correlated with a number of complaints on the PAOFI at the trend level (r = .24, p = .05). Apathy “after” ratings were significantly associated with cognitive complaints (r = .57, p < .0001), and the magnitude of this correlation was twice that for pre-infection apathy and current cognitive complaints.

A multivariable regression analysis was conducted with the total number of PAOFI complaints as the criterion and current apathy level as well as the confounding variables (i.e., global neurocognitive impairment, lifetime MDD diagnosis, lifetime history of any substance use disorder, AIDS status, current CD4 count, and HCV serostatus) entered as predictors. As shown in Table 5, the overall model was significant, adjusted R² = .40, F(7, 57) = 7.09, p < .0001. Current apathy rating (β = 0.55, p < .001) and global neurocognitive impairment (β = 0.24, p = .02) were the only significant predictors of total cognitive complaints, while the HCV status (β = 0.20, p < .07) was associated at the trend level.

Table 5.

Predictors of cognitive complaints in the HIV+ sample

Variable	Model	B	B 95% CI	Parameter (β)	p-value
Adjusted R2	0.40
F	7.09				<.0001
Apathy T-score		0.13	0.08, 0.18	0.54	<.0001
LT-MDD		−0.63	−1.6, 0.38	−0.13	.22
LT-SUD		0.15	−1.02, 1.32	0.03	.80
Global cognitive impairment		1.17	0.17, 2.17	0.24	.03
AIDS status		−0.22	−1.30, 0.84	−0.04	.67
Current CD4 count		0.001	−0.003, 0.005	0.07	.52
HCV status		1.08	−0.11, 2.27	0.20	.07

Notes: LT-MDD, Lifetime Major Depressive Disorder diagnosis; LT-SUD, Lifetime Substance Use Disorder diagnosis; HCV, Hepatitis-C virus.

Finally, multivariable models were analyzed examining the association of apathy and individual domains (i.e., memory, cognition, sensory, motor, and language; see Table 4) on the PAOFI after adjusting for covariates (i.e., lifetime MDD and substance use disorder diagnoses, AIDS status, HCV status, and current CD4). Significant models were only obtained for the domains of memory (adjusted R² = .36, F(7,62) = 6.69, p < .0001) and general cognitive complaints (adjusted R² = .27, F(7, 62) = 4.51, p = .001). Apathy was an independent, significant predictor of difficulties in memory (β = 0.41, p < .001) and cognition (β = 0.45, p < .001) such that higher ratings of apathy were associated with greater complaints in both domains.

Post-hoc Multivariable Analyses Examining the Association of FrSBe Subscales of Disinhibition and Executive Dysfunction and Functional Outcomes

To test the specificity of the effect of apathy on functional outcomes, we examined the association of disinhibition and executive dysfunction, two neurobehavioral sequelae of HIV infection that are also measured by the FrSBe, with cognitive and IADL complaints. Post-hoc multivariable analyses were conducted with the current ratings of disinhibition and executive dysfunction being considered alongside other possible predictors (i.e., global neurocognitive impairment, lifetime MDD diagnosis, lifetime history of any substance use disorder, AIDS status, current CD4 count, and HCV serostatus) of the two functional outcomes of interest. Current ratings of disinhibition (β = 0.34, p = .01) as well as executive dysfunction (β = 0.63, p < .0001) were independent predictors of cognitive complaints (see Table 6). However, as noted in Table 7, neither rating was associated with IADL complaints.

Table 6.

Models examining disinhibition and executive dysfunction as predictors of cognitive complaints in the HIV+ sample

Variable	Model	B	B 95% CI	Parameter (β)	p-value
Disinhibition
Adjusted R2	0.22
F	3.69				<.01
Disinhibition T-score		0.14	0.03, 0.24	0.34	.01
LT-MDD		−0.58	−1.80, 0.63	−0.12	.34
LT-SUD		0.18	−1.15, 1.52	0.03	.78
Global cognitive impairment		1.51	0.37, 2.64	0.31	.01
AIDS status		0.40	−0.76, 1.57	0.08	.49
Current CD4 count		2.01e−4	−0.004, 0.005	0.01	.93
HCV status		0.55	−0.81, 1.91	0.10	.41
Executive dysfunction
Adjusted R2	0.45
F	8.33				<.0001
Executive dysfunction T-score		0.16	0.11, 0.22	0.63	<.0001
LT-MDD		−0.15	−1.17, 0.87	−0.03	.77
LT-SUD		−0.06	−1.19, 1.07	−0.01	.91
Global cognitive impairment		1.05	0.08, 2.02	0.21	.03
AIDS status		−0.31	−1.35, 0.72	−0.06	.54
Current CD4 count		−0.2.3e−4	−0.004, 0.003	−0.01	.89
HCV status		0.58	−0.56, 1.72	0.10	.31

Notes: LT-MDD= Lifetime Major Depressive Disorder diagnosis; LT-SUD= Lifetime Substance Use Disorder diagnosis; HCV= Hepatitis-C virus.

Table 7.

Models examining disinhibition and executive dysfunction as predictors of IADL decline in the HIV+ sample

Variable	Model	B	B 95% CI	Parameter (β)	p-value
Disinhibition
Adjusted R2	0.12
F	2.36				.03
Disinhibition T-score		0.03	−0.009, .079	0.21	.11
LT-MDD		−0.51	−1.05, 0.03	−24	.07
LT-SUD		0.10	−0.49, 0.69	0.04	.73
Global cognitive impairment		0.25	−0.22, 0.73	0.12	.29
AIDS status		−0.29	−0.79, 0.19	−0.14	.24
Current CD4 count		−3.3e−4	−0.002, 0.002	−0.04	.75
HCV status		−0.45	−1.03, 0.13	−0.19	.12
Executive dysfunction
Adjusted R2	0.13
F	2.55				.02
Executive dysfunction T-score		0.03	−0.002, 0.059	0.26	.06
LT-MDD		−0.49	−1.02, 0.04	−24	.07
LT-SUD		0.07	−0.51, 0.66	0.03	.81
Global cognitive impairment		0.16	−1.01, 0.14	0.08	.52
AIDS status		−0.40	−0.92, 0.11	−0.19	.12
Current CD4 count		−0.0003	−0.002, 0.001	−0.03	.75
HCV status		−0.43	−1.01, 0.14	−0.19	.06

Notes: LT-MDD= Lifetime Major Depressive Disorder diagnosis; LT-SUD= Lifetime Substance Use Disorder diagnosis; HCV= Hepatitis-C virus.

Discussion

The prevalence and clinical correlates of apathy in HIV infection have received increased attention in the past decade. Commensurate with prior research (e.g., Castellon et al., 1998; Hoare et al., 2010), findings from this investigation revealed a significant elevation in self-reported apathy ratings among persons living with HIV infection when compared with seronegative subjects. In fact, HIV infection was associated with a significantly higher prevalence of apathy, with 42% of the HIV+ individuals in our sample reporting clinically significant levels of apathy (vs. 19% in the seronegatives). This prevalence estimate is consistent with the 30–50% rates previously reported in HIV+ persons (e.g., Castellon et al., 1998, 2000; Rabkin et al., 2000). Extending the literature, we observed that self-reported symptoms of apathy prior to HIV infection were comparable with those endorsed by seronegatives, but rose significantly in the period after seroconversion. Importantly, the association between HIV serostatus and current apathy was independent of various confounding factors such as lifetime history of any substance use disorder, psychiatric comorbidities, cognitive impairment, and ethnicity. The elevation in apathy post-HIV infection coupled with the higher prevalence of apathy in the HIV+ group appears to be consistent with the underlying frontostriatal pathophysiology (e.g., involvement of the ventral striatum and anterior cingulate) common to HIV infection and apathy (Paul, Brickman et al., 2005).

This high prevalence of apathy in HIV+ persons warrants attention to this neuropsychiatric symptom for its role in functional outcomes. Consistent with our expectations, an interaction between the number of IADL complaints and apathy ratings prior to and following HIV infection was observed. A strong association was noted for IADL complaints and current apathy levels, such that individuals with higher ratings of apathy reported more IADL difficulties. These findings are consistent with the previous functional outcome studies (i.e., Barclay et al., 2007; Rabkin et al., 2000; Tate et al., 2003), but extend them by showing strong and independent associations between apathy and everyday functions (e.g., social activities, shopping, and comprehension of written/verbal material). Such findings are also commensurate with research in other clinical populations (e.g., AD and HCV) suggests that apathy plays an important role in functional difficulties (Freels et al., 1992; Posada et al., 2010).

Self-reported cognitive complaints provide another lens on daily functioning that might be relevant to apathy. This is particularly true in an HIV+ population, as cognitive complaints are commonly endorsed (e.g., Carter, Rourke, Murji, Shore, & Rourke, 2003) and are correlated with higher levels of functional impairments in tasks such as financial management, medication management, cooking, and employment (e.g., Heaton et al., 2004). In this study, using the total number of cognitive complaints endorsed on the PAOFI, a similar association with apathy as that seen for IADL complaints was noted. HIV+ individuals who endorsed higher ratings of apathy were more likely to report having noticed declines in various cognitive domains relevant to their everyday functioning, especially memory. This is consistent with a study of individuals with mild cognitive impairment (MCI) that demonstrated a positive relationship between subjective cognitive complaints and apathy (Robert et al., 2006). This association may reflect the overlap between cognitive complaints and the cognitive aspects of apathy, such as a lack of goal-directed thought content, diminished motivation with regards to executive functions, and decreased verbal fluency. Fatigue, a strong predictor of cognitive complaints (Woods et al., 2007), may accompany apathy and this may also contribute to the observed association.

Its independence from confounding factors is critical to the clinical and scientific value of apathy as a predictor of functional outcomes. HIV infection is associated with highly prevalent comorbidities that are strong predictors of everyday functioning. These include higher rates of lifetime substance use disorders (Bing et al., 2001), major depressive disorder (Atkinson et al., 1988, 2008; Bing et al., 2001), global neurocognitive impairment (Heaton et al., 1995; Hinkin et al., 2002; Martin et al., 2001), and comorbid HCV infection (e.g., Goulet, Fultz, McGinnis, & Justice, 2005). Functional impairment in HIV disease is also impacted by CD4 cell counts (e.g., Blalock, McDaniel & Farber, 2002) and AIDS status (e.g., Basso & Bornstein, 2000; Heaton et al., 1995). Although there were strong univariate effects of neurocognitive impairment and MDD on ADL complaints, apathy accounted for unique variance beyond these critical factors. In fact, the predictive value of neurocognitive impairment for IADLs (but not cognitive complaints) was diminished by the inclusion of apathy in the model. These findings suggest that relative to other predictors, apathy captures unique aspects of functional impairments and may therefore warrant clinical consideration.

The association between apathy and functional outcomes is consistent with the relationship noted between apathy and lower health-related quality of life. Two studies reported that elevated levels of apathy have a significant, negative impact on quality-of-life ratings (Rabkin et al., 2000; Tate et al., 2003), and this relationship was independent of depressed mood (Tate et al., 2003). Taken together, these findings are consistent with the Wilson and Cleary (1995) model of health-related quality of life, which posits that psychological symptoms may result in functional impairment, thereby impacting an individual's quality of life. Interestingly, a different pattern of association has been reported in ARV adherence studies. Two studies reported that while apathy was associated with adherence behaviors and intentions, it was not an independent predictor of the same after accounting for variables such as depressed mood (Rabkin et al., 2000) and perceptions of self as well as medication utility (Barclay et al., 2007). Whether the discrepancy in these findings of everyday functions is accounted for by the sample characteristics or the methods of assessing depression (i.e., self-report instruments in the studies by Rabkin and colleagues (2000) and Barclay and colleagues (2007) vs. structured clinical interviews in the present study) as well as functional outcomes is unclear. In all, the findings for quality-of-life variables, IADL and cognitive complaints as well as reports of the association between apathy and adherence suggest that apathy may affect specific aspects of everyday functioning more so than others.

As with all cross-sectional clinical investigations, limitations of this study must be acknowledged. First, we used self-report measures of apathy and everyday functioning outcomes, whose accuracy may be affected by bias, depression, and/or mild anosognosia. Apathy ratings were obtained via a self-report questionnaire (i.e., FrSBe) as opposed to a clinical interview (e.g., Neuropsychiatric Inventory; Cummings et al., 1994) and were retrospectively provided for the period prior to HIV infection. This may be problematic in the context of episodic memory deficits commonly observed in HIV+ persons (e.g., Gongvatana, Woods, Taylor, Vigil, & Grant, 2007) and tempers the strength of conclusions that may be drawn based on significant interactions we found for functional outcomes and the pre- and post-HIV infection apathy ratings in a cross-sectional study. The nature of the retrospective ratings raises the concern that a participant's post self-ratings may be primed by the pre-rating, as illustrated by significant correlations between the two scales (see Table 1). Arguing against this interpretation, however, is the finding that post-ratings and not pre-ratings were related to the outcomes of interest. Functional outcomes were also obtained using self-report instruments. Laboratory-based measures of daily functioning (e.g., Direct Assessment of Functional Status; Lowenstein & Bates, 1992) may have provided greater sensitivity to functional declines when compared with those obtained through self-report (Blackstone et al., 2012). Relatedly, that self-reported neurobehavioral symptoms (e.g., current apathy ratings) are predictive of self-reported functional difficulties deserves some consideration. As demonstrated by the post-hoc analyses, current self-reported ratings of apathy, but not disinhibition and executive dysfunction, were independently associated with IADL deficits. However, all three FrSBe subscales were significant predictors of cognitive complaints, which suggests variable specificity of apathy as a predictor of self-rated functional impairments. It is possible that the maintenance of ADL independence, a generally more severe expression of functional decline than the subjective experience of cognitive declines, is more sensitive to the behavioral aspects of apathy.

Another limitation relates to the sample's demographic (i.e., a well-educated, predominately male and Caucasian cohort) and well-managed HIV disease characteristics (e.g., only 10% were immunosuppressed), which may restrict the extrapolation of our findings to more diverse and symptomatic HIV+ groups. The cross-sectional study design is another limitation, as it does not allow us to determine how well apathy predicts functional decline at a later time-point, a question that bears considerable clinical relevance for this population. The quasi-experimental, retrospective study design also did not permit the examination of whether the apathy-associated reduced behavioral repertoire led to the declines in functional abilities or, alternatively, increasing functional difficulties prompted the individuals to behaviorally withdraw to minimize the experience of failures.

In the context of prior studies, these findings suggest that apathy symptoms are more prevalent in HIV+ individuals compared with seronegatives, are more prominent following HIV infection, and adversely impact everyday functioning. Domainwise analyses demonstrated that apathy was independently associated with declines in daily activities such as social activities, shopping, and comprehension of verbal and written information, as well as with an increase in memory and general cognitive complaints. This suggests that specific functional activities may be most sensitive to higher levels of apathy. That elevated levels of apathy on the FrSBe appear to be associated with increased cognitive and functional difficulties in HIV+ individuals has implications for clinicians attempting to ascertain an individual's ability to function and live independently. Given that apathy is associated with a number of adverse outcomes, including functional impairment, preliminary studies across non-HIV patient groups have investigated pharmacological interventions, which suggest promising but methodologically limited evidence for the possible efficacy of dopaminergic agents and amphetamines in the treatment of apathy (van Reekum, Stuss, & Ostrander, 2005). Preliminary studies examining behavioral interventions for apathy in patients with dementia suggest that mentally stimulating activities and interactive social settings may be effective (see Roth, Flashman, & McAllister, 2007 for review). Further investigation of pharmacological and behavioral interventions for HIV-infected cohorts is warranted as such therapies may have a positive impact on functional outcomes and quality of life. Future studies examining apathy in HIV would benefit from a prospective study design, multimodal assessments of daily functioning and apathy symptoms, as well as an exploration of other HIV disease markers of CNS involvement to tease apart the specific behavioral changes associated with apathy and how they manifest in everyday functioning declines.

Funding

The HIV Neurobehavioral Research Center (HNRC) is supported by Center award MH 62512 from NIMH. This study was also supported by P01-DA12065 from NIDA.

Conflict of Interest

None declared.

Appendix

The San Diego HIV Neurobehavioral Research Center [HNRC] group is affiliated with the University of California, San Diego, the Naval Hospital, San Diego, and the Veterans Affairs San Diego Healthcare System, and includes: Director: Igor Grant, M.D.; Co-Directors: J. Hampton Atkinson, M.D., Ronald J. Ellis, M.D., Ph.D., and J. Allen McCutchan, M.D.; Center Manager: Thomas D. Marcotte, Ph.D.; Jennifer Marquie-Beck, M.P.H.; Melanie Sherman; Neuromedical Component: Ronald J. Ellis, M.D., Ph.D. (P.I.), J. Allen McCutchan, M.D., Scott Letendre, M.D., Edmund Capparelli, Pharm.D., Rachel Schrier, Ph.D., Terry Alexander, R.N., Debra Rosario, M.P.H., Shannon LeBlanc; Neurobehavioral Component: Robert K. Heaton, Ph.D. (P.I.), Steven Paul Woods, Psy.D., Mariana Cherner, Ph.D., David J. Moore, Ph.D., Matthew Dawson; Neuroimaging Component: Terry Jernigan, Ph.D. (P.I.), Christine Fennema-Notestine, Ph.D., Sarah L. Archibald, M.A., John Hesselink, M.D., Jacopo Annese, Ph.D., Michael J. Taylor, Ph.D.; Neurobiology Component: Eliezer Masliah, M.D. (P.I.), Cristian Achim, M.D., Ph.D., Ian Everall, FRCPsych., FRCPath., Ph.D. (Consultant); Neurovirology Component: Douglas Richman, M.D., (P.I.), David M. Smith, M.D.; International Component: J. Allen McCutchan, M.D., (P.I.); Developmental Component: Cristian Achim, M.D., Ph.D.; (P.I.), Stuart Lipton, M.D., Ph.D.; Participant Accrual and Retention Unit: J. Hampton Atkinson, M.D. (P.I.), Rodney von Jaeger, M.P.H.; Data Management Unit: Anthony C. Gamst, Ph.D. (P.I.), Clint Cushman (Data Systems Manager); Statistics Unit: Ian Abramson, Ph.D. (P.I.), Florin Vaida, Ph.D., Reena Deutsch, Ph.D., Anya Umlauf, M.S., Tanya Wolfson, M.A.