Abstract
Background:
People with HIV (PWH) are at two times greater risk for major depressive disorder (MDD) than people without HIV (PWoH), which manifests in symptoms across cognitive, somatic, affective, apathy, and anhedonia domains that may differentially impact clinical outcomes. However, few studies have examined whether HIV and its characteristics relate to depressive symptom domains.
Methods:
This secondary, cross-sectional analysis included 3456 participants enrolled in studies at the UCSD HIV Neurobehavioral Research Program and CHARTER sites between 2000 and 2023 (79 % PWH, 78 % male, Age: M = 47.8). Depressive symptom domains were assessed with the Beck Depression Inventory (BDI-II). Multivariable linear regression models evaluated associations between HIV, HIV disease characteristics, and depressive symptom domains while controlling for covariates.
Results:
HIV diagnosis was significantly associated with higher severity across depressive symptom domains (affective: B = 0.51, cognitive: B = 0.44, somatic: B = 1.55, anhedonia: B = 0.52, apathy: B = 0.58, all ps < 0.05) and overall depressive symptoms (BDI-II total: B = 2.51, p < 0.01) while adjusting for covariates. Among PWH, HIV viral suppression was associated with fewer overall depressive symptoms, driven by fewer cognitive and somatic symptoms (ps < 0.01), while higher current CD4+ T-cell count was associated with fewer affective and apathy symptoms (ps < 0.05).
Conclusion:
HIV diagnosis was associated with higher depressive symptoms across all domains. Current HIV disease indicators and duration of HIV disease were associated with select depressive domains to varying degrees, except for the anhedonia domain. These findings highlight the potential importance of examining individual symptom domains as they differentially associate with varying aspects of HIV disease, which may provide insight into specific treatment targets.
Keywords: Depressive domains, Depression, Depressive symptoms, HIV, HIV disease characteristics
1. Introduction
Although advances in antiretroviral therapy (ART) have led to dramatic increases in rates of viral suppression and subsequent life expectancies [1], people with HIV (PWH) still experience disproportionately higher rates of depressive disorders than people without HIV (PwoH; [2]). The prevalence of a clinical diagnosis of major depressive disorder (MDD) has been estimated to be nearly twice as high among PWH compared to PwoH [3] with recent global estimates suggesting that 30–35 % of PWH experience depressive symptoms [4,5]. Depressive disorders are characterized by a broad array of depressive symptoms, such as depressed mood, fatigue, apathy, anhedonia, and somatic disturbances [6]. Despite its multidimensional nature, depression is typically assessed categorically (e.g., depression vs. no depression) rather than dimensionally. To meet criteria for an MDD diagnosis, any 5 out of the 9 possible depressive symptoms in the DSM-5 must be endorsed [6]; thus, a clinical diagnosis often offers little insight into the individual’s specific symptoms. However, researchers have found that depression prognosis and treatment efficacy can vary depending on specific symptomatology and depression subtypes [7,8].
The Beck Depression Inventory-II (BDI-II) is one of the most validated and widely used screening tools for depressive disorders and assesses the severity of 21 different depressive symptoms [9–11]. Although most studies have exclusively examined overall depressive symptom severity (i.e., using the total sum of BDI-II symptom items), research has found that the BDI-II can be divided into three factors or domains of depressive symptoms (i.e., somatic, cognitive, and affective), which has been confirmed via factor analysis and validated across clinical populations, including PWH [12,13]. More recent investigations have additionally identified anhedonia and apathy domains [14–17]. While PWH are known to have a greater risk for depressive disorders, the relationship between HIV and specific depressive symptom domains remains largely unknown. This study aims to fill this gap by examining the associations between HIV diagnosis and specific depressive symptom domains, providing a more nuanced understanding of depressive symptoms in PWH.
Depressive symptom domains warrant additional study among PWH as they have been differentially associated with the clinical outcomes of HIV and other chronic diseases. For example, the somatic factor of the BDI has been found to predict risk for HIV nonsuppression over time [13]. While few studies have examined how depressive symptom domains relate to HIV disease characteristics, previous work has more commonly evaluated their associations with the clinical outcomes of other chronic physical illnesses, such as cardiovascular disease (CVD). A meta-analysis of thirteen prospective studies found that that somatic symptoms (e.g., insomnia, fatigue, loss of libido) were more strongly and consistently associated with increased risk for mortality and cardiovascular events for patients with heart disease than were cognitive/affective symptoms (e.g., pessimism, guilt, self-dislike; [18]). Stewart et al. examined associations between total, cognitive/affective, and somatic depressive symptoms and biomarkers of CVD in PWH and found that somatic symptoms were more strongly associated with increased risk for CVD relative to cognitive/affective symptoms [19]. While prior literature has suggested that the disease indicators of other chronic diseases like CVD are differentially related to individual depression domains, they have rarely been examined in the context of HIV disease outcomes.
The present exploratory analysis thus first aimed to evaluate associations between HIV diagnosis and individual depressive symptom domains, as measured by the BDI-II factors (i.e., cognitive, affective, somatic, anhedonia, and apathy). A secondary aim was to investigate associations between these depressive symptom domains and HIV disease characteristics among PWH (e.g., viral load, CD4 count, ART status, duration of HIV disease).
2. Methods
2.1. Participants
Participants were 3456 individuals (79 % PWH, 78 % male, Age: M = 47.8) enrolled in studies conducted at the UC San Diego HIV Neurobehavioral Research Program (HNRP) or six academic medical centers affiliated with the multi-site CNS HIV Anti-Retroviral Therapy Effects Research (CHARTER) study [1]. For this secondary, cross-sectional analysis, participants were included in the present analysis if they had a known HIV status and had completed the Beck Depression Inventory II (BDI-II) at their most recent study visit between 2000 and 2023. All data used in this analysis were contemporaneously collected during a single study visit, meaning that depressive symptoms, HIV diagnosis, HIV disease characteristics and covariates were assessed at the same time point for each participant.
All participants provided written informed consent to undergo study procedures and for their data to be used in secondary research in their respective studies, which were approved by the UC San Diego Institutional Review Board. The present secondary analysis of existing, de-identified data did not constitute human subjects requiring additional review. While complete inclusion and exclusion criteria varied by the respective studies in which the participants were enrolled, general exclusion criteria across studies included being under the age of 18, having a self-reported diagnosis of schizophrenia, a mood disorder with psychotic features, or neurological or medical conditions that affect the interpretation of neuropsychological testing.
2.2. Measures
Beck Depression Inventory-II:
The Beck Depression Inventory-II (BDI-II; [9]) is a 21-item self-report instrument used to assess the severity of depressive symptoms. Participants rate each item on a 4-point scale ranging from 0 to 3, with higher scores indicating greater severity of symptoms. A total score is calculated by summing the scores of all 21 items. The BDI-II has demonstrated good internal consistency and adequate test-retest reliability for assessing depressive symptoms among PWH [12]. We examined five depressive symptom domains of the BDI-II; affective, cognitive, and somatic [12] as well as anhedonia [20] and apathy [21,22] domains. The affective domain (range 0–12) includes the emotional symptoms of depression, such as crying (e.g., “I cry more than I used to”). The cognitive domain (range 0–27) captures cognitive symptoms related to depression, such as pessimism, guilt, and self-dislike (e.g., “I feel I am a total failure as a person”). The somatic domain (range 0–24) includes physical symptoms associated with depression, such as changes in sleep patterns, appetite, and fatigue (e.g., “I am too tried or fatigued to do most of the things I used to”). The anhedonia and apathy subscales contain items that overlap with the other three subscales. While there is substantial item overlap among the affective, apathy, and anhedonia domains of the BDI-II, we elected to include all three in our analyses given the exploratory nature of our study and the absence of strong empirical evidence favoring one configuration over another in PWH. Specifically, the anhedonia domain symptoms were assessed using items that evaluate loss of interest or pleasure in activities once enjoyed (e.g., “I can’t get any pleasure from things I used to enjoy”; range 0–9). The apathy subscale includes items that assess lack of motivation and loss of interest in activities (e.g., “It’s hard to get interested in anything”; range 0–12). Prior work in PWH have identified these symptoms of reflective of a motivational disturbance component of mood [15,16,21,22]. See Fig. 1 for a full mapping between individual BDI-II items and depressive domains.
Fig. 1.
Mapping between Individual BDI-II Items and Depressive Symptom Domains. Note: References for the depressive symptom domains of the BDI-II are as follows: a) Hobkirk et al., 2015, b) Castellon et al., 2006, c) Marquine et al., 2013, and d) Joiner et al., 2013.
Neurocognitive Impairment (NCI):
Neurocognitive impairment was assessed by detailed neurocognitive evaluation and summarized using the Global Deficit Score (GDS), which measures cognitive functioning across multiple domains relevant to HIV-related NCI. The GDS is calculated by averaging deficit scores across various neuropsychological tests, with higher scores indicating greater impairment [23]. The GDS score is derived by converting the demographically corrected T-scores of each neuropsychological measure into individual deficit scores using a five-point scale (0 = no impairment to 5 = severe impairment) and averaging them [23,24]. The GDS has been used as objective method to determine NCI in PWH, and NCI that is classified in this way is consistent with classifications based upon the Frascati criteria at a cutoff of ≥0.5 [25,26]. The GDS was integrated into the analysis to control for NCI, ensuring that the associations between depressive symptom domains and HIV status were not confounded by neurocognitive deficits.
Mood and Substance Abuse or Dependence Disorders:
Participants’ psychiatric history of mood disorders and substance use disorders, were ascertained using the mood and substance abuse modules of the Composite International Diagnostic Interview (CIDI; [27]). A lifetime diagnosis of any substance use disorder (dependence and/or abuse) was included as a covariate in models examining relationship between an HIV diagnosis and depressive domains, while a current MDD diagnosis was used to descriptively characterize our sample. Given the substantial conceptual and statistical overlap between MDD diagnosis and the depressive symptom domains assessed by the BDI-II, we opted to use MDD status only descriptively rather than modeling it as a covariate. The CIDI is a lay-administered, fully structured instrument commonly used in psychiatric research to capture both lifetime and current psychiatric DSM diagnoses.
Charlson Comorbidity Index:
The Charlson Comorbidity Index (CCI) was used to assess the presence and severity of comorbid medical conditions (e.g., cardiovascular disease, kidney disease, diabetes, cancer). The CCI total is a widely used method for quantifying comorbidities in clinical populations, with higher scores indicating greater comorbidity burden [28].
Antidepressant Use:
Participants self-reported their use of antidepressant medications, which was dichotomized into use of any antidepressant medication or not for the purposes of analysis.
HIV Disease Characteristics:
HIV disease and treatment characteristics, including current CD4 absolute, CD4 nadir, optimal viral suppression, antiretroviral therapy (ART) use (yes/no), and estimated duration of HIV disease were obtained through comprehensive neuromedical evaluations consisting of a structured clinician-administered interview and standard laboratory assays. The absolute CD4 count, representing the number of CD4 T-lymphocytes per cubic millimeter of blood, was used to assess the current state of immune function. The CD4 nadir, which is the lowest recorded CD4 count since HIV diagnosis, was treated as a continuous variable to understand the historical severity of immune system compromise. Levels of HIV RNA in plasma were measured via reverse transcriptase-polymerase chain reaction, with optimal viral suppression defined as being at or below 50 copies/mL. ART status was self-reported by participants, indicating whether they were currently on ART. The estimated duration of HIV disease was calculated from the time between the study visit and the participant’s self-reported first date of HIV diagnosis.
2.3. Statistical analyses
All statistical analyses were performed using R (version 4.4.2). Descriptive statistics (including mean, median, standard deviation, interquartile range) were calculated to describe the demographic and clinical characteristics of the analysis sample. To assess for statistically significant differences in demographic and clinical characteristics between PWH and PwoH in our sample, we conducted Welch’s t-tests for continuous variables and chi-square tests for categorical variables.
Given the continuous nature of the cognitive, affective, somatic, anhedonia, and apathy depressive symptom domains, linear regression models were used to model their associations with HIV and HIV disease characteristics. We modeled each depressive symptom domain in separate linear regressions due to significant multicollinearity among domains. As shown in Supplementary Table 1, domains were highly correlated with each other (r = 0.65–0.92, p’s < 0.001). Modeling the depressive domains separately allowed us to detect distinct associations between specific domains, HIV, and HIV disease characteristics. We also modeled overall depressive symptoms (BDI-II total score) to evaluate whether specific domains revealed unique associations that might be obscured when using only a unidimensional approach. To achieve our first aim of evaluating whether HIV was associated with depressive symptom domains, we constructed multivariable linear regression models (Models 1–5) for each domain and overall depressive symptoms (Model 6) that included HIV along with demographic and clinical covariates that were identified a-priori based on their known associations with either HIV and/or depressive symptoms in previous literature. The covariates included in Models 1–6 included: age, sex, lifetime substance abuse and/or dependence, the Charlson Comorbidity Index, neurocognitive impairment (based on the GDS), and current antidepressant use.
To achieve our second aim of evaluating the associations between HIV disease characteristics and depressive symptom domains among PWH, we constructed bivariate (or unadjusted) and multivariable linear regression models for each domain (Models 7–11) and overall depressive symptoms (Model 12) that simultaneously included all HIV disease characteristics (CD4 nadir, CD4 absolute, duration of HIV disease, optimal viral suppression, and current ART use) to allow us to determine whether their associations were independent from one another. To aid in the interpretation of the regression coefficients for the continuous variables CD4 nadir and current CD4, these predictors were scaled to represent 100 cell count increases to aid interpretation. Standardized regression coefficients were also calculated for each model to compare the strength of associations across the depressive symptom domain outcomes while accounting for their varying ranges (e.g., the cognitive domain range is 0–27 and the anhedonia domain only ranges from 0 to 9). A significance level of 0.05 was used across all statistical analyses and comparisons.
For each set of models, the variance inflation factor (VIF) was calculated and examined for each predictor to assess for multicollinearity given the number of variables being included in multivariate models. Multicollinearity is considered present when VIF is higher than 5 [29]. The VIF we observed were below this threshold in our multivariate linear regression models (ranging between 1.03 and 1.71).
3. Results
3.1. Demographic and clinical characteristics of analysis sample
Overall, 10.6 % of participants in our sample currently met criteria for major depressive disorder based on the CIDI (n = 366). Most participants in the sample were PWH (n = 2717, 79 %; PWoH: n = 739, 21 %), male sex (78 %), were 47.8 (SD = 12.2) years old on average, met criteria for DSM-IV lifetime substance abuse or dependence (73 %), were not taking antidepressants (66.6 %) nor were they neurocognitively impaired (60.8 %). Demographic and clinical characteristics of the sample by HIV status are presented in Table 1. For HIV disease characteristics of PWH (see Table 2), the median nadir CD4+ T-cell count was 158 (IQR = 34–292), the median current CD4+ T-cell count was 497 (IQR = 310–711), and the mean estimated duration of disease was 14.4 years (SD = 9.05). The majority of PWH were currently on ART (81.2 %), were virally suppressed (HIV RNA ≤ 50 copies/ml; 63.3 %), and 64.3 % had an AIDS diagnosis.
Table 1.
Demographic and Clinical Characteristics of Sample by HIV status (n = 3456).
| People without HIV (PWoH; n = 739) | People with HIV (PWH; n = 2717) | |
|---|---|---|
|
| ||
| Study Year | ||
| 2000–2007 | 116 (15.7 %) | 1050 (38.6 %) |
| 2008–2013 | 255 (34.5 %) | 643 (23.7 %) |
| 2014–2018 | 170 (23.0 %) | 480 (17.7 %) |
| 2019–2023 | 198 (26.8 %) | 544 (20.0 %) |
| Race / Ethnicity | ||
| Asian | 11 (1.5 %) | 15 (0.6 %) |
| Black | 116 (15.7 %) | 939 (34.6 %) |
| Hispanic or Latino | 163 (22.1 %) | 440 (16.2 %) |
| Other | 30 (4.1 %) | 59 (2.2 %) |
| White | 419 (56.7 %) | 1264 (46.5 %) |
| Age | ||
| Mean (SD) | 47.2 (14.8) | 48.0 (11.4) |
| Sex | ||
| Male | 492 (66.6 %) | 2202 (81.0 %) |
| Female | 247 (33.4 %) | 515 (19.0 %) |
| Substance Abuse or Dependence (Lifetime) | 501 (67.8 %) | 2023 (74.5 %) |
| Charlson Comorbidity Index | ||
| Mean (SD) | 1.39 (1.49) | 5.57 (3.56) |
| Neurocognitive Impairment (GDS) | 208 (28.1 %) | 1147 (42.2 %) |
| Antidepressant Use | 125 (16.9 %) | 1029 (37.9 %) |
| Major Depressive Disorder (Current) | 45 (6.1 %) | 321 (11.8 %) |
| BDI-II Total (Range: 0–63) | ||
| Mean (SD) | 8.34 (9.59) | 12.1 (10.6) |
| BDI-II Affective Domain (Range: 0–12) | ||
| Mean (SD) | 1.51 (2.15) | 2.27 (2.44) |
| BDI-II Cognitive Domain (Range: 0–27) | ||
| Mean (SD) | 3.06 (4.29) | 3.88 (4.72) |
| BDI-II Somatic Domain (Range: 0–24) | ||
| Mean (SD) | 3.77 (4.02) | 5.94 (4.54) |
| BDI-II Anhedonia Domain (Range: 0–9) | ||
| Mean (SD) | 1.36 (1.72) | 2.18 (2.03) |
| BDI-II Apathy Domain (Range: 0–12) | ||
| Mean (SD) | 1.69 (2.19) | 2.61 (2.48) |
Table 2.
HIV Disease Characteristics of People with HIV in Sample (n = 2597).
| CD4 Nadir | |
|---|---|
|
| |
| Median (IQR) | 158 (34–292) |
| Range [Min, Max] | [0, 1550] |
| CD4 Absolute | |
| Median (IQR) | 497 (310–711) |
| Range [Min, Max] | [0, 2710] |
| Duration of HIV Disease (years) | |
| Mean (SD) | 14.4 (9.05) |
| Range [Min, Max] | [0, 40.6] |
| Optimal Viral Suppression (≤50) | 1644 (63.3 %) |
| Current ART | 2109 (81.2 %) |
There was no statistically significant difference in age between PWH and PWoH (t(987.97) = −1.36, p = 0.18), suggesting comparable age distributions. However, PWH had significantly higher Charlson Comorbidity Index scores compared to PWoH, indicating greater overall medical comorbidity burden (t(2897.6) = −47.72, p < 0.001). PWH were more likely to be male (χ2(1) = 69.93, p < 0.001) and to identify as Black or African American, and less likely to identify as White or Hispanic/Latino (χ2(4) = 106.33, p <0.001) compared to PWoH. Clinically, PWH were more likely to meet criteria for current major depressive disorder (χ2(1) = 19.99, p < 0.001), have a lifetime history of substance use disorder (χ2(1) = 12.76, p < 0.001), exhibit neurocognitive impairment (χ2(1) = 47.66, p < 0.001), and report current antidepressant use (χ2(1) = 113.79, p < 0.001).
3.2. Bivariate associations between HIV and depressive symptom domains
Bivariate linear regression models found an HIV diagnosis to be significantly associated with higher scores across all depressive domains, including affective (B = 0.76, β = 0.13, SE = 0.10, p < 0.01), cognitive (B = 0.82, β = 0.07, SE = 0.19, p < 0.01), somatic (B = 2.18, β = 0.19, SE = 0.18, p < 0.01), anhedonia (B = 0.82, β = 0.17, SE = 0.08, p < 0.01), and apathy domains (B = 0.92, β = 0.15, SE = 0.10, p < 0.01), as well as for overall depressive symptoms (B = 3.76, β = 0.15, SE = 0.43, p < 0.01) compared to no HIV diagnosis.
3.3. Independent associations between HIV and depressive symptom domains
Multivariable linear regression models showed that an HIV diagnosis remained significantly associated with higher scores across all subscales, including affective (Model 1: B = 0.51, β = 0.09, SE = 0.11, p < 0.01), cognitive (Model 2: B = 0.44, β = 0.04, SE = 0.22, p < 0.05), somatic (Model 3: B = 1.55, β = 0.14, SE = 0.21, p < 0.01), anhedonia (Model 4: B = 0.52, β = 0.11, SE = 0.09, p < 0.01), and apathy domains (Model 5: B = 0.58, β = 0.10, SE = 0.12, p < 0.01), as well as for overall depressive symptoms (Model 6: B = 2.51, β = 0.10, SE = 0.49, p < 0.01) compared to no HIV diagnosis while controlling for demographic and clinical covariates. See Table 3 for full multivariable linear regression results for each depressive symptom domain (Models 1–5) and for overall depressive symptoms (Model 6).
Table 3.
Multivariable Linear Regression Models between HIV-Status and Depressive Symptom Domains Adjusting for Demographic and non-HIV disease Clinical Covariates (n = 3456).
| Dependent Variables B (SE) |
||||||
|---|---|---|---|---|---|---|
| Model 1: Affective | Model 2: Cognitive | Model 3: Somatic | Model 4: Anhedonia | Model 5: Apathy | Model 6: BDI-II Total | |
|
| ||||||
| HIV Diagnosis | 0.51*** (0.11) | 0.44** (0.22) | 1.55*** (0.21) | 0.52*** (0.09) | 0.58*** (0.12) | 2.51*** (0.49) |
| Age | −0.02*** (0.004) | −0.05*** (0.01) | −0.02*** (0.01) | 0.0002 (0.003) | −0.01** (0.004) | −0.09*** (0.02) |
| Female Sex | 0.11 (0.10) | −0.043 (0.19) | 0.96*** (0.18) | 0.26*** (0.08) | 0.13 (0.10) | 1.03** (0.42) |
| Substance Abuse or Dependence (Lifetime) | 0.56*** (0.09) | 1.33*** (0.17) | 1.11*** (0.17) | 0.41*** (0.07) | 0.57*** (0.09) | 2.99*** (0.39) |
| Charlson Comorbidity Index | −0.01 (0.01) | −0.04 (0.03) | 0.05** (0.03) | 0.02* (0.01) | 0.01 (0.01) | 0.003 (0.06) |
| Neurocognitive Impairment (GDS) | 0.31*** (0.08) | 0.43*** (0.16) | 0.41*** (0.15) | 0.19*** (0.07) | 0.24*** (0.08) | 1.15*** (0.35) |
| Antidepressant Use | 1.07*** (0.09) | 2.10*** (0.16) | 2.06*** (0.16) | 0.95*** (0.07) | 1.19*** (0.09) | 5.22*** (0.37) |
| Intercept | 1.55*** (0.20) | 4.22*** (0.39) | 3.03*** (0.38) | 0.74*** (0.17) | 1.39*** (0.21) | 8.80*** (0.88) |
| Adjusted R2 | 0.08 | 0.09 | 0.11 | 0.09 | 0.09 | 0.11 |
Note: Reference groups were no HIV Diagnosis (or people without HIV), male sex, no life substance abuse/dependence, no neurocognitive impairment, and not taking antidepressants. GDS = global deficit score
p < 0.1
p < 0.05
p < 0.01.
3.4. Independent associations between HIV disease characteristics and depressive symptom domains among PWH
Multivariable linear regression models also found differential associations between HIV disease characteristics and depressive symptom domains among PWH. Higher current CD4 was associated with a significant decrease in affective (Model 7: B = −0.04, β = −0.05, SE = 0.06, p < 0.05) and apathy depressive symptoms (Model 11: B = −0.05, β = −0.06,SE = 0.06, p <0.05). PWH with optimal viral suppression had significantly lower cognitive (Model 8: B = −0.79, β = −0.08, SE = 0.23, p <0.01), somatic (Model 9: B = −0.59, β = −0.06, SE = 0.22, p <0.01), and overall depressive symptoms (Model 12: B = −1.61, β = −0.07,SE = 0.52, p < 0.01) compared to those without, while adjusting for the other HIV disease characteristics. Longer HIV disease duration was also associated with significantly lower affective (Model 7: B = −0.01, β = −0.05, SE = 0.05, p < 0.05), cognitive (Model 8: B = −0.05, β = −0.10, SE = 0.10, p < 0.01), and overall depressive symptoms (Model 12: B = −0.09, β = −0.07, SE = 0.23, p < 0.01). Conversely, we observed no significant associations between nadir CD4+ T-cell count or current ART use with any of the depressive symptom domains or overall depressive symptoms (Models 7–12: all ps > 0.05) while controlling for the other HIV disease characteristics. HIV disease characteristics explained the greatest amount of variance in the model of cognitive depressive symptom domain (Model 8: Adjusted R2 = 0.03) and the least (nonsignificant) in the anhedonia domain (Model 10: Adjusted R2 = 0.002). Table 4 provides the full multivariable linear regression model results between HIV disease characteristics and each depressive symptom domains (Models 7–11), as well as overall depressive symptoms (Model 12).
Table 4.
Multivariable Linear Regression Models between HIV Disease Characteristics and Depressive Symptom Domains among People with HIV (n = 2597).
| Dependent Variables B (SE) |
||||||
|---|---|---|---|---|---|---|
| Model 7: Affective | Model 8: Cognitive | Model 9: Somatic | Model 10: Anhedonia | Model 11: Apathy | Model 12: BDI-II Total | |
|
| ||||||
| CD4 Nadir | 0.05 (0.06) | 0.11* (0.12) | −0.02 (0.12) | 0.01 (0.05) | 0.03 (0.06) | 0.14 (0.27) |
| CD4 Absolute | −0.04** (0.06) | −0.06* (0.12) | −0.05 (0.11) | −0.03 (0.05) | −0.05** (0.06) | −0.15* (0.26) |
| Optimal Viral Suppression (≤ 50 vs. > 50) | −0.23* (0.12) | −0.79*** (0.23) | −0.59*** (0.22) | −0.14 (0.10) | −0.16 (0.12) | −1.61*** (0.52) |
| Duration of HIV Disease (years) | −0.01** (0.05) | −0.05*** (0.10) | −0.02* (0.10) | −0.00 (0.05) | −0.01 (0.05) | −0.09*** (0.23) |
| Current ART (vs. not on ART) | −0.12 (0.15) | −0.20 (0.28) | −0.24 (0.28) | −0.06 (0.12) | −0.13 (0.15) | −0.55 (0.64) |
| Intercept | 2.49*** (0.12) | 4.50*** (0.24) | 6.51*** (0.23) | 2.31*** (0.10) | 2.80*** (0.13) | 13.49*** (0.54) |
| Adjusted R2 | 0.011 | 0.03 | 0.009 | 0.002 | 0.006 | 0.02 |
Notes: ART = antiretroviral therapy. CD4 Nadir and CD4 Absolute were scaled to represent 100 cell count increases to aid interpretation.
p < 0.1
p < 0.05
p < 0.01.
4. Discussion
Despite the increased risk for depressive disorders among PWH and the multidimensional nature of depressive symptoms, little is known about whether HIV diagnosis is differentially associated with depressive symptom domains (affective, cognitive, somatic, anhedonia) and the extent to which these domains are differentially related to HIV disease characteristics. The present study sought to address these gaps to better understand the heterogeneity of depressive symptoms in PWH. While we found that an HIV diagnosis was significantly associated with higher depressive symptoms across all domains, the strength of the association was strongest for the somatic domain (β = 0.14) and weakest for the cognitive domain (β = 0.04) based on standardized regression coefficients accounting for their varying ranges. Our results further suggested the presence of differential associations between HIV disease characteristics across depressive symptom domains except for anhedonia among PWH. Specifically, we found that optimal viral suppression was associated with lower cognitive, somatic, and overall depressive symptoms and that longer duration of HIV disease was associated with lower affective, cognitive, and overall depressive symptoms. In addition, higher current CD4+ T-cell count was uniquely associated with lower affective and apathy depressive symptom domains. There were no independent relationships between CD4 nadir or current ART use and any depressive symptom domains or overall depressive symptoms.
Previous research has similarly found that an HIV diagnosis is associated with increased somatic depressive symptoms, which may be due in part to their overlap with the somatic symptoms of HIV disease and/or its treatment. Kalichman et al. (2000) suggested that the overlap in somatic symptoms between HIV infection and depression may hinder the clinical utility of depression scales [30]. However, more recent research seeking neurobiological mechanisms that may at least partially underlie depression in PWH has suggested that the neuroinflammation elicited by HIV may contribute to the high prevalence of depression among PWH, particularly the somatic depressive symptoms [31]. Increased fatigue, sleep disturbances, loss of appetite, feelings of sadness, and anhedonia have been observed with both depression and inflammation [32]. In addition to somatic symptoms, Kamata et al. (2013) previously reported that PWH had greater apathy symptoms than PWoH, but had not accounted for potential confounders in their analysis [32]. Similar to these prior studies, we also found higher somatic and apathy depressive symptoms to be associated with an HIV diagnosis even after controlling for demographic and clinical covariates.
Few studies to date have examined the individual domains of depressive symptoms and their relationships with HIV disease characteristics for PWH. Perkins et al. previously reported that somatic symptoms of depressive (fatigue and insomnia) were not associated with current CD4 cell count [33], which we also found. Although Kamata et al. previously found apathy symptoms to not be associated with CD4 cell count [33,34], we did find higher CD4 to be associated with lower affective and apathy symptoms which may be due to our larger sample size. In a longitudinal study of youth living with HIV, Kohn et al. (2021) found greater somatic depressive symptoms to be associated with an increased risk for viral nonsuppression. Our results in a broader sample of PWH also suggest that optimal viral suppression is associated with lower somatic symptoms, which may highlight important treatment targets for PWH. We also found that a longer duration of HIV disease was found to be associated with decreases in affective, cognitive, and overall depressive symptoms. It is probable that older age plays a large role in this association, which has been found to be associated with improvements in mental health and may reflect improved coping, emotion regulation, and resiliency [34]. Ultimately, the relatively small amount of variance in depressive symptom domains explained by HIV disease characteristics (Adjusted R2 ranged from 0.03 to 0.002) suggests that effective HIV disease management alone does not address depressive symptoms in PWH.
Our findings regarding differential associations between HIV and its disease characteristics with depressive symptom domains and HIV disease characteristics raise important considerations for clinical practice, specifically the personalization of mental health care for PWH. These findings underscore the limitations of relying solely on total depressive symptoms or categorical diagnoses and instead support the utility of domain- or symptom-specific assessments to better target or match treatments to individual’s unique symptom presentations. For instance, our finding that HIV diagnosis was most strongly associated with somatic depressive symptoms is clinically meaningful given recent meta-analytical evidence showing that somatic complaints are more responsive to pharmacological treatments like antidepressants than to psychotherapy broadly [35,36]. Among psychotherapies, cognitive-behavioral therapy (CBT) has also demonstrated greater efficacy for symptoms such as somatic anxiety and psychomotor agitation compared to other psychotherapies [37], suggesting it also be well suited for PWH given the greater burden of somatic depressive symptoms. These differential effects between treatment modalities may be especially important in HIV care settings, where high rates of polypharmacy and medical comorbidities necessitate precise and efficient intervention [38]. In this context, identifying a patient’s predominant depressive symptom domains may inform clinical decision-making regarding whether pharmacological, psychotherapeutic, or type of psychotherapy are best suited for a patient. Future research should continue evaluating the utility of domain or symptom-specific approaches to personalizing depression treatment for PWH in order to improve treatment outcomes and address the disproportionate burden of depression in this population.
There are several notable limitations to our study that are important to consider when interpreting our findings. First, the cross-sectional nature of our analysis precludes us from being able to establish temporality between HIV, HIV disease characteristics, and the depressive symptoms we examined. Previous research has supported a bidirectional relationship between depression and HIV, such that individuals with depressive disorders are at higher risk for HIV infection and PWH are at higher risk for developing depressive disorders and faster HIV disease progression [39]. Although we controlled for several demographic and non-HIV clinical differences between PWH and PWoH in our models, there remains the possibility of other confounders between the groups given the observational nature of our study and the nature of the PWoH sample, which for many HNRP studies are recruited to serve as a “comparison” group (vs a healthy control sample) to help isolate the effects of HIV in the context of common psychiatric, medical, and substance use comorbidities. Future studies may benefit from expanding these comorbidities to examine their influence across depressive symptom domains. While we identified distinct associations between HIV disease characteristics and several depressive symptom domains, it is important to note the substantial item overlap among the affective, anhedonia, and apathy domains as a notable limitation. Specifically, these domains share several BDI-II items, which likely contributes to the high intercorrelations observed between them and the similar pattern of associations between HIV disease characteristics and affective and apathy domains in particular. Prior research has found apathy and anhedonia to be positively correlated with another potentially due to shared underlying neurobiological mechanisms related to motivation, while also being dissociable [40]. The overlapping measurement content in our domain definitions limits our ability to isolate true domain-specific effects. Future research should consider using validated, domain-specific instruments that minimize item redundancy or adopt symptom-level approaches—such as network analysis [41]—to clarify the distinct contributions of specific symptoms. Lastly, our sample was not drawn from a population-representative sample and so our results may not generalize.
Despite these limitations, there are also notable strengths to our study. We were able to achieve a large sample size by leveraging a center-wide and multi-site database that allowed us to likely have more power to detect smaller associations compared to previous studies. Furthermore, our study is also one of the first to take a multidimensional approach to studying depressive symptoms among PWH by examining individual depressive symptom domains.
5. Conclusions
Our findings represent an important first step towards exploring how multidimensional depressive symptoms across several domains differentially relate to an HIV diagnosis and HIV disease characteristics. The increased prevalence of depression among PWH did not appear to be well explained by any one domain of depressive symptoms in our study, such that HIV showed significant, but relatively small, associations across all depressive symptom domains. Our models also explained a relatively small proportion of the variance in depressive symptom domains, which suggests future studies should also consider investigating other known predictors of depression and how interactions with HIV or its disease characteristics may relate to depressive symptom domains. The possible interactions or relationships between these depressive symptoms domains themselves may also be important to understanding the disproportionate risk for depressive disorders among PWH and represent an important future direction. Longitudinal analyses should also be considered in future studies to help establish causality between HIV, HIV disease characteristics, and depressive symptom domains. It may also be useful to identify cross-sectional and longitudinal depression symptom profiles (vs domains), which may provide even greater insight into the heterogeneity of depression and changes over time in PWoH vs PWH.
Importantly, we identified differential associations between HIV disease characteristics and specific depressive symptom domains, with optimal viral suppression being associated with lower cognitive, somatic, and overall depressive symptoms and higher CD4 cell count being associated with lower apathy and affective depressive symptom domains. These findings further highlight the importance of examining individual symptom domains (opposed to only total depressive symptoms) and potential profiles in future studies as they differentially associate with varying aspects of HIV disease. This nuanced understanding may inform the development of more targeted interventions and clinical practices to improve mental health outcomes for PWH.
Supplementary Material
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jpsychores.2025.112311.
Acknowledgements
The authors extend their sincere gratitude to the research participants, whose time and effort made this analysis possible. We also acknowledge the dedication of the research staff, whose contributions to study coordination, data collection, and participant support were invaluable. Their commitment to rigorous research and scientific integrity greatly enhanced the quality of this work. We also thank the National NeuroHIV Tissue Consortium (NNTC request number: R803) for their assistance with data procurement.
Funding
Maximo Prescott: NIDA (5T32DA031098), Jessica Montoya: NIDA (5K23DA051324).
HNRC: NIMH (P30MH062512), CHARTER: NIH/NIMH/NINDS (N01MH220055), NIH/NIMH/NINDS (HHSN271201000036C), NIH/NIMH/NINDS (HHSN271201000030C), NIMH (R01MH107345), CNTN: NIMH (U24MH100928).
Footnotes
Declaration of competing interest
The authors report no competing conflicts of interest. The authors and affiliated institutions received funding from the National Institute of Health as follows: Maximo Prescott (NIDA 5T32DA031098), Jessica Montoya (NIDA 5K23DA051324), HIV Neurobehavioral Research Center (NIMH, P30MH062512), CHARTER Study (NIH/NIMH/NINDS N01MH220055, NIH/NIMH/NINDS HHSN271201000036C, NIH/NIMH/NINDS HHSN271201000030C, NIMH R01MH107345), California NeuroAIDS Tissue Network (NIMH U24MH100928).
CRediT authorship contribution statement
Maximo Prescott: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Crystal Wang: Writing – review & editing, Writing – original draft, Conceptualization. Miya Gentry: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Donald Franklin: Writing – review & editing, Data curation, Conceptualization. Murray B. Stein: Writing – review & editing, Conceptualization. Joseph Hampton Atkinson: Writing – review & editing, Conceptualization. Ronald Ellis: Writing – review & editing, Conceptualization. Robert K. Heaton: Writing – review & editing, Conceptualization. Jessica Montoya: Writing – review & editing, Methodology, Conceptualization. David Moore: Writing – review & editing, Methodology, Conceptualization. Jennifer Iudicello: Writing – review & editing, Methodology, Formal analysis, Data curation, Conceptualization.
References
- [1].Heaton RK, Clifford DB, Franklin DR Jr., Woods SP, Ake C, Vaida F, Ellis RJ, Letendre SL, Marcotte TD, Atkinson JH, Rivera-Mindt M, Vigil OR, Taylor MJ, Collier AC, Marra CM, Gelman BB, McArthur JC, Morgello S, Simpson DM, McCutchan JA, Abramson I, Gamst A, Fennema-Notestine C, Jernigan TL, Wong J, Grant I, CHARTER group, HIV-associated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER study, Neurology 75 (2010) 2087–2096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Rubin LH, Maki PM, HIV, depression, and cognitive impairment in the era of effective antiretroviral therapy, Curr. HIV/AIDS Rep. 16 (2019) 82–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Ciesla JA, Roberts JE, Meta-analysis of the relationship between HIV infection and risk for depressive disorders, Am. J. Psychiatry 158 (2001) 725–730. [DOI] [PubMed] [Google Scholar]
- [4].Rezaei S, Ahmadi S, Rahmati J, Hosseinifard H, Dehnad A, Aryankhesal A, Shabaninejad H, Ghasemyani S, Alihosseini S, Bragazzi NL, Raoofi S, Kiaee ZM, Ghashghaee A, Global prevalence of depression in HIV/AIDS: a systematic review and meta-analysis, BMJ Support. Palliat. Care 9 (2019) 404–412. [DOI] [PubMed] [Google Scholar]
- [5].Yang Y, Chen B, Zhang H, Huang P, Qian J, Lin L, Zhang L, Cai F, Global prevalence of depressive symptoms among people living with HIV/AIDS: a systematic review and meta-analysis of the past five years, AIDS Care 36 (2024) 153–164. [DOI] [PubMed] [Google Scholar]
- [6].American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, American Psychiatric Association, 2013. [Google Scholar]
- [7].Fava M, Uebelacker LA, Alpert JE, Nierenberg AA, Pava JA, Rosenbaum JF, Major depressive subtypes and treatment response, Biol. Psychiatry 42 (1997) 568–576. [DOI] [PubMed] [Google Scholar]
- [8].Gao S, Calhoun VD, Sui J, Machine learning in major depression: from classification to treatment outcome prediction, CNS Neurosci. Ther. 24 (2018) 1037–1052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Beck AT, Steer RA, Brown GK, BDI-II, Beck Depression Inventory: Manual, Psychological Corporation, 1996. [Google Scholar]
- [10].von Glischinski M, von Brachel R, Hirschfeld G, How depressed is “depressed”? A systematic review and diagnostic meta-analysis of optimal cut points for the Beck depression inventory revised (BDI-II), Qual. Life Res. 28 (2019) 1111–1118. [DOI] [PubMed] [Google Scholar]
- [11].Wang Y-P, Gorenstein C, Psychometric properties of the Beck Depression Inventory-II: a comprehensive review, Rev. Bras. Psiquiatr. 35 (2013) 416–431. [DOI] [PubMed] [Google Scholar]
- [12].Hobkirk AL, The CHARTER Group, Starosta AJ, De Leo JA, Marra CM, Heaton RK, Earleywine M, Psychometric validation of the BDI-II among HIV-positive CHARTER study participants, Psychol. Assess 27 (2015) 457–466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Kohn JN, Loop MS, Kim-Chang JJ, Garvie PA, Sleasman JW, Fischer B, Rendina HJ, Woods SP, Nichols SL, Hong S, Trajectories of depressive symptoms, neurocognitive function, and viral suppression with antiretroviral therapy among youth with HIV over 36 months, J. Acquir. Immune Defic. Syndr. 87 (2021) 851–859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Cogan AB, Persons JB, Kring AM, Using the Beck depression inventory to assess anhedonia: a scale validation study, Assessment 31 (2024) 431–443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Hussain MA, Iudicello JE, Morgan EE, Kamat R, Heaton RK, Grant I, the TMARC Group, Apathy is associated with poorer abstinence self-efficacy in individuals with methamphetamine dependence, Addict. Behav. Rep. 13 (2021) 100331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Kamat R, Morgan E, Marcotte TD, Badiee J, Maich I, Cherner M, de Almeida S, de Pereira AP, Ribeiro CE, Barbosa F, Atkinson JH, Ellis R, HNRP Group, Implications of apathy and depression for everyday functioning in HIV/AIDS in Brazil, J. Affect. Disord. 150 (2013) 1069–1075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Ubukata S, Ueda K, Fujimoto G, Ueno S, Murai T, Oishi N, Extracting apathy from depression syndrome in traumatic brain injury by using a clustering method, J. Neuropsychiatr. Clin. Neurosci 34 (2022) 158–167. [DOI] [PubMed] [Google Scholar]
- [18].de Miranda Azevedo R, Roest AM, Hoen PW, de Jonge P, Cognitive/affective and somatic/affective symptoms of depression in patients with heart disease and their association with cardiovascular prognosis: a meta-analysis, Psychol. Med 44 (2014) 2689–2703. [DOI] [PubMed] [Google Scholar]
- [19].Stewart JC, Polanka BM, So-Armah KA, White JR, Gupta SK, Kundu S, Chang C-CH, Freiberg MS, Associations of Total, cognitive/affective, and somatic depressive symptoms and antidepressant use with cardiovascular disease-relevant biomarkers in HIV: veterans aging cohort study, Psychosom. Med 82 (2020) 461–470. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Joiner TE, Brown JS, Metalsky GI, A test of the tripartite model’s prediction of anhedonia’s specificity to depression: patients with major depression versus patients with schizophrenia, Psychiatry Res. 119 (2003) 243–250. [DOI] [PubMed] [Google Scholar]
- [21].Castellon SA, Hardy DJ, Hinkin CH, Satz P, Stenquist PK, van Gorp WG, Myers HF, Moore L, Components of depression in HIV-1 infection: their differential relationship to neurocognitive performance, J. Clin. Exp. Neuropsychol. 28 (2006) 420–437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Marquine MJ, Iudicello JE, Morgan EE, Brown GG, Letendre SL, Ellis RJ, Deutsch R, Woods SP, Grant I, Heaton RK, Translational methamphetamine AIDS research center (TMARC) group, “frontal systems” behaviors in comorbid human immunodeficiency virus infection and methamphetamine dependency, Psychiatry Res. 215 (2014) 208–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Carey CL, Woods SP, Gonzalez R, Conover E, Marcotte TD, Grant I, Heaton RK, HNRC Group, Predictive validity of global deficit scores in detecting neuropsychological impairment in HIV infection, J. Clin. Exp. Neuropsychol. 26 (2004) 307–319. [DOI] [PubMed] [Google Scholar]
- [24].Revised Comprehensive Norms for an Expanded Halstead-Reitan Battery: Demographically Adjusted Neuropsychological Norms for African American and Caucasian Adults, Professional Manual, Psychological Assessment Resources, 2004. [Google Scholar]
- [25].Blackstone K, Moore DJ, Franklin DR, Clifford DB, Collier AC, Marra CM, Gelman BB, McArthur JC, Morgello S, Simpson DM, Ellis RJ, Atkinson JH, Grant I, Heaton RK, Defining neurocognitive impairment in HIV: deficit scores versus clinical ratings, Clin. Neuropsychol. 26 (2012) 894–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Campbell LM, Fennema-Notestine C, Saloner R, Hussain M, Chen A, Franklin D, Umlauf A, Ellis RJ, Collier AC, Marra CM, Clifford DB, Gelman BB, Sacktor N, Morgello S, McCutchan JA, Letendre S, Grant I, Heaton RK, CHARTER group, use of neuroimaging to inform optimal neurocognitive criteria for detecting HIV-associated brain abnormalities, J. Int. Neuropsychol. Soc 26 (2020) 147–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Composite International Diagnostic Interview (CIDI), American Psychiatric Press, 1993. [Google Scholar]
- [28].Charlson ME, Carrozzino D, Guidi J, Patierno C, Charlson comorbidity index: a critical review of Clinimetric properties, Psychother. Psychosom. 91 (2022) 8–35. [DOI] [PubMed] [Google Scholar]
- [29].Kim JH, Multicollinearity and misleading statistical results, Korean J. Anesthesiol. 72 (2019) 558–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Kalichman SC, Rompa D, Cage M, Distinguishing between overlapping somatic symptoms of depression and HIV disease in people living with HIV-AIDS, J. Nerv. Ment. Dis. 188 (2000) 662–670. [DOI] [PubMed] [Google Scholar]
- [31].Mudra Rakshasa-Loots A, Whalley HC, Vera JH, Cox SR, Neuroinflammation in HIV-associated depression: evidence and future perspectives, Mol. Psychiatry 27 (2022) 3619–3632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Peteet J, Dell ML, Fung WLA, Ethical Considerations at the Intersection of Psychiatry and Religion, Oxford University Press, 2018. [Google Scholar]
- [33].Perkins DO, Leserman J, Stern RA, Baum SF, Liao D, Golden RN, Evans DL, Somatic symptoms and HIV infection: relationship to depressive symptoms and indicators of HIV disease, Am. J. Psychiatry 152 (1995) 1776–1781. [DOI] [PubMed] [Google Scholar]
- [34].Thomas ML, Kaufmann CN, Palmer BW, Depp CA, Martin AS, Glorioso DK, Thompson WK, Jeste DV, Paradoxical trend for improvement in mental health with aging: a community-based study of 1,546 adults aged 21–100 years, J. Clin. Psychiatry 77 (2016) e1019–e1025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Boschloo L, Bekhuis E, Weitz ES, Reijnders M, DeRubeis RJ, Dimidjian S, Dunner DL, Dunlop BW, Hegerl U, Hollon SD, Jarrett RB, Kennedy SH, Miranda J, Mohr DC, Simons AD, Parker G, Petrak F, Herpertz S, Quilty LC, John Rush A, Segal ZV, Vittengl JR, Schoevers RA, Cuijpers P, The symptom-specific efficacy of antidepressant medication vs. cognitive behavioral therapy in the treatment of depression: results from an individual patient data meta-analysis, World Psychiatry 18 (2019) 183–191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Boschloo L, Hieronymus F, Cuijpers P, ICECA work group, clinical response to SSRIs relative to cognitive behavioral therapy in depression: a symptom-specific approach, World Psychiatry 21 (2022) 152–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].O’Driscoll C, Buckman JEJ, Saunders R, Ellard S, Naqvi SA, Singh S, Wheatley J, Pilling S, Symptom-specific effects of counselling for depression compared to cognitive-behavioural therapy, BMJ Ment Health 26 (2023) e300621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Edelman EJ, Gordon KS, Glover J, McNicholl IR, Fiellin DA, Justice AC, The next therapeutic challenge in HIV: polypharmacy, Drugs Aging 30 (2013) 613–628. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Nanni MG, Caruso R, Mitchell AJ, Meggiolaro E, Grassi L, Depression in HIV infected patients: a review, Curr. Psychiatry Rep 17 (2015) 530. [DOI] [PubMed] [Google Scholar]
- [40].Husain M, Roiser JP, Neuroscience of apathy and anhedonia: a transdiagnostic approach, Nat. Rev. Neurosci. 19 (2018) 470–484. [DOI] [PubMed] [Google Scholar]
- [41].Borsboom D, Deserno MK, Rhemtulla M, Epskamp S, Fried EI, McNally RJ, Robinaugh DJ, Perugini M, Dalege J, Costantini G, Isvoranu A-M, Wysocki AC, van Borkulo CD, van Bork R, Waldorp LJ, Network analysis of multivariate data in psychological science, Nat. Rev. Methods Primers 1 (2021), 10.1038/s43586-021-00055-w. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.