Abstract
Objective:
We sought to determine associations of total, cognitive/affective, and somatic depressive symptoms and antidepressant use with biomarkers of processes implicated in cardiovascular disease in HIV (HIV-CVD).
Methods:
We examined data from 1,546 HIV-positive and 843 HIV-negative veterans. Depressive symptoms were assessed using the Patient Health Questionnaire-9, and past-year antidepressant use was determined from VA pharmacy records. Monocyte (soluble CD14; sCD14), inflammatory (interleukin-6; IL-6), and coagulation (D-dimer) marker levels were determined from previously banked blood specimens. Linear regression models with multiple imputation were run to estimate associations between depression-related factors and CVD-relevant biomarkers.
Results:
Among HIV-positive participants, greater somatic depressive symptoms were associated with higher sCD14 (exp[b]=1.02, 95% CI: 1.00–1.03) and D-dimer (exp[b]=1.06, 95% CI: 1.00–1.11) after adjustment for demographics and potential confounders. Further adjustment for antidepressant use and HIV factors slightly attenuated these relationships. Associations were also detected for antidepressant use, as selective serotonin reuptake inhibitor use was related to lower sCD14 (exp[b]=0.95, 95% CI: 0.91–1.00) and IL-6 (exp[b]=0.86, 95% CI: 0.76–0.96), and tricyclic antidepressant use was related to higher sCD14 (exp[b]=1.07, 95% CI: 1.03–1.12) and IL-6 (exp[b]=1.14, 95% CI: 1.02–1.28). Among HIV-negative participants, total, cognitive/affective, and somatic depressive symptoms were associated with higher IL-6, and tricyclic antidepressant use was related to higher sCD14.
Conclusions:
Our novel findings suggest that (a) monocyte activation and altered coagulation may represent two pathways through which depression increases HIV-CVD risk and that (b) tricyclic antidepressants may elevate and selective serotonin reuptake inhibitors may attenuate HIV-CVD risk by influencing monocyte and inflammatory activation.
Keywords: HIV, depression, somatic symptoms, soluble CD14, interleukin-6, D-dimer
Introduction
Highly-effective antiretroviral therapy (ART) has changed human immunodeficiency virus (HIV) from a life-threatening disease to a chronic disease requiring life-long management. However, with this transformation has come new health concerns. Chief among them is the high morbidity and mortality due to cardiovascular disease in HIV (HIV-CVD) (1). HIV-positive adults, versus HIV-negative adults, have a two-fold greater risk of developing CVD (1), and this elevated risk is not fully explained by established CVD risk factors or HIV factors (2). These observations have stimulated a search for novel risk factors for HIV-CVD that could serve as targets of future CVD prevention efforts.
One such emerging risk factor for HIV-CVD is depression, which consists of cognitive, affective, and somatic symptoms. Depression is common in HIV – it has an estimated prevalence of 40–42%, although rates vary widely across studies (3). Furthermore, consistent with the substantial literature linking depression with future CVD in non-HIV samples (4), our group recently reported that depressive disorders are independently associated with incident acute myocardial infraction (5) and heart failure (6) in people with HIV.
While the mechanisms underlying the depression to HIV-CVD relationship have yet to be elucidated, immune activation, systemic inflammation, and altered coagulation are strong candidates. First, all three processes play prominent roles in current conceptual frameworks of the pathogenesis of HIV-CVD, with HIV-related immune activation and the consequent systemic inflammation and altered coagulation contributing to atherosclerosis and clinical CVD onset (7). One indicator of immune activation is soluble CD14 (sCD14), which is considered a nonspecific marker of monocyte activation (8) and an acute phase protein (9). Second, immune activation, inflammatory, and coagulation markers – including sCD14, interleukin-6 (IL-6), and D-dimer, respectively – have been linked with both subclinical and clinical CVD in people with HIV (10–12). To illustrate, in the Strategies for Management of Antiretroviral Therapy (SMART) study, each standard deviation (SD) increase in IL-6 and D-dimer was associated with a 39% and 40% increased risk of a CVD event, independent of CVD risk factors (13). Third, initial evidence suggests that, in HIV-positive adults, depression may be associated with elevated immune activation (14–19) and inflammatory (19–24) markers. Although these depression-biomarker studies report novel and potentially significant findings, they utilized small samples, did not include an HIV-negative group, had limited adjustment for potential confounders, and/or did not examine depressive symptom clusters or antidepressant medication use.
Accordingly, our objective was to determine the independent associations of total depressive symptoms, cognitive/affective and somatic symptom clusters, and antidepressant use (regardless of treatment indication) with monocyte activation (sCD14), inflammatory (IL-6), and coagulation (D-dimer) markers in people living with HIV. While we hypothesized that greater total depressive symptoms would be related to higher and antidepressant use would be related to lower biomarker levels, we did not have hypotheses regarding depressive symptom clusters or antidepressant classes due to the paucity of evidence in HIV. However, a growing literature of non-HIV studies suggests that the somatic symptom cluster is more consistently associated with inflammatory markers than are the other clusters (see (25) for recent review), and a meta-analysis of non-HIV studies indicates that treatment with selective serotonin reuptake inhibitors (SSRIs), but not with other antidepressant classes, may reduce inflammatory markers (26). Moreover, in the lone study examining depressive symptom subtypes and inflammatory markers in HIV, the group with severe/moderate somatic symptoms had elevated IL-6 (22). To achieve our objective, we examined data from the Biomarker Cohort of the Veterans Aging Cohort Study (VACS), which consists of 1,546 HIV-positive and 843 HIV-negative veterans. Inclusion of the HIV-negative group allowed us to assess whether HIV modifies associations between depression-related variables and CVD-relevant biomarkers.
Material and Methods
Participants
Data came from the VACS Biomarker Cohort – a subset of participants from the parent VACS-9 study. VACS-9 is a prospective, multisite, cohort study of HIV-positive veterans and age, sex, race/ethnicity, and clinical site-matched HIV-negative veterans from nine Veterans Affair (VA) medical centers across the U.S. (27). The Biomarker Cohort consists of a subset of VACS-9 participants (n = 2,389) who consented to provide a blood sample at one time point between 2005–2006 (28). The matching procedures were not reapplied to the Biomarker Cohort.
Measures
Depressive Symptoms.
In the parent VACS-9 study, the Patient Health Questionnaire-9 (PHQ-9) (29) was administered multiple times to assess depressive symptoms over the last two weeks. The available studies suggest that the PHQ-9 is a reliable and valid instrument in people with HIV (30). For the present analysis, we used data from the PHQ-9 administration closest to the blood draw date (median [25th-75th percentile] = 31 [0–231] days prior to blood draw). PHQ-9 total scores range from 0–27, with ≥10 being indicative of clinically significant depressive symptoms (31). In addition to the total score (sum of all nine items), we computed the cognitive/affective score as the sum of items 1 (anhedonia), 2 (depressed mood), 6 (low self-esteem), 7 (concentration problems), 8 (psychomotor retardation/agitation), and 9 (suicidal ideation) and the somatic score as the sum of items 3 (sleep disturbance), 4 (fatigue), and 5 (appetite changes). The PHQ-9 has high internal consistency and good sensitivity and specificity for identifying cases of major depressive disorder (29, 31). The cognitive/affective and somatic symptom clusters have been validated across major sociodemographic groups in a recent study utilizing a large, nationally representative sample of U.S. adults (32). For participants who were missing data for one item (n=147), we imputed that value with the mean of the other eight items. The PHQ-9 scores were converted to z-scores prior to analysis so that a 1-unit change corresponded to a 1-SD change.
Antidepressant Use.
Antidepressant use was defined as documentation of a filled prescription for an antidepressant medication up to 365 days before the blood draw date in the VA pharmacy records data. Separate yes/no variables were computed for following antidepressant classes: selective serotonin reuptake inhibitor (SSRI), tricyclic antidepressant (TCA), and miscellaneous other antidepressant. Medications from the following classes were coded as miscellaneous other: monoamine oxidase inhibitor, serotonin-norepinephrine reuptake inhibitor, serotonin antagonist and reuptake inhibitor, norepinephrine reuptake inhibitor, norepinephrine-dopamine reuptake inhibitor, and tetracyclic antidepressant. The treatment indication for the detected antidepressant use is not known. Of note, the indication for SSRIs and certain medications in our miscellaneous category (e.g., serotonin-norepinephrine reuptake inhibitors) is often a depressive disorder, whereas the indication for TCAs and other medications in our miscellaneous category (e.g., trazodone) is often insomnia or pain conditions (33).
Monocyte Activation, Inflammatory, and Coagulation Markers.
As is described elsewhere (28, 34), blood samples were collected at one time point between 2005–2006 using serum separator and EDTA tubes and were shipped to a central repository at the Massachusetts Veterans Epidemiology Research and Information Center. Assays for biomarkers examined here were conducted at the Laboratory for Clinical Biochemistry Research at the University of Vermont and used four controls per sample to assess interassay coefficients of variability (CVs). sCD14 was quantified by an enzyme-linked immunosorbent assay (Quantikine sCD14 Immunoassay, R&D Systems, Minneapolis, MN) with a detectable range of 40–3,200 ng/ml. The interassay CVs ranged from 7.2–8.1%. IL-6 was quantified by a chemiluminescent immunoassay (QuantiGlo IL-6 immunoassay, R&D Systems, Minneapolis, MN) with detectable range of 0.4–10,000 pg/mL. The interassay CVs ranged from 7.7–12.3%. D-dimer was quantified by a STAR automated coagulation analyzer (Diagnostica Stago, Parsippany, NJ) using an immunoturbidometric assay (Liatest D-DI) with a detectable range of 0.01–20 ug/mL. The interassay CVs ranged from 2.8–14.8%. All three biomarkers were natural log transformed before analysis to approximate a normal distribution.
Covariates.
As in our past work (28), covariate data were obtained closest to the blood draw date (see Table 1 for the coding of variables). Demographic factors were age, sex, and race/ethnicity. Several biomedical and behavioral factors were included in our models as covariates. CVD was identified by an International Classification of Diseases, Ninth Revision (ICD-9) code for acute myocardial infarction, unstable angina, cardiovascular revascularization, ischemic stroke, hemorrhagic stroke, heart failure, or cardiomyopathy before the blood draw date. Diabetes was defined by a previously validated metric using glucose values, diabetes medication use, and/or at least one inpatient or two outpatient ICD-9 codes for diabetes (35). Blood pressure was computed as the average of the three routine outpatient measurements obtained closest to the blood draw date. Hypertension was identified by blood pressure ≥140/90 mmHg or documentation of antihypertensive medication. Low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and triglyceride levels were obtained from the VA Corporate Data Warehouse. Statin use was defined by a filled prescription receipt for a 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitor. Hepatitis C infection was identified by a positive hepatitis C virus antibody test or at least one inpatient or two outpatient ICD-9 codes for hepatitis C infection. Estimated glomerular filtration rate (eGFR; a renal function indicator) and hemoglobin were defined by laboratory values extracted from the VA Corporate Data Warehouse. Body mass index and smoking were determined from the VA Health Factors dataset (36). The Alcohol Use Disorders Identification Test (AUDIT-C) (37) and alcohol abuse/dependence ICD-9 codes were used to create a 4-level alcohol use variable: not current, not hazardous (AUDIT-C <4), hazardous or heavy episodic (AUDIT-C ≥4), abuse/dependence (ICD-9 code). Cocaine abuse/dependence was identified by an ICD-9 code for a cocaine use disorder before the blood draw date. We included these substance use variables, as they have been associated with both depression and the biomarkers (28, 38, 39) and, thus, could confound the associations of interest.
Table 1.
Characteristics of Veterans in the VACS Biomarker Cohort by HIV Status and Depression Status
| HIV-Positive (n=1,527) |
HIV-Negative (n=837) |
|||||
|---|---|---|---|---|---|---|
| Depressed (n=348) |
Not Depressed (n=1,179) |
p | Depressed (n=263) |
Not Depressed (n=574) |
p | |
|
Biomarkers | ||||||
| Soluble CD14, median (25th-75th percentile), ng/mL | 1822 (1511–2137) | 1694 (1432–2073) | 0.002 | 1757 (1491–2075) | 1720 (1467–2032) | 0.44 |
| Interleukin-6, median (25th-75th percentile), pg/mL | 2.20 (1.47–3.55) | 2.03 (1.41–3.35) | 0.14 | 2.10 (1.25–3.51) | 1.73 (1.13–2.90) | 0.013 |
| D-dimer, median (25th-75th percentile), μg/mL | 0.29 (0.16–0.53) | 0.26 (0.15–0.48) | 0.091 | 0.32 (0.21–0.53) | 0.29 (0.21–0.53) | 0.80 |
|
Depression Variables | ||||||
| PHQ-9 Total Score (0–27), mean (SD) | 15.9 (5.2) | 2.7 (2.9) | <0.001 | 16.1 (4.9) | 2.9 (2.9) | <0.001 |
| PHQ-9 Cognitive/affective Score (0–18), mean (SD) | 9.6 (4.1) | 1.2 (1.8) | <0.001 | 9.8 (3.7) | 1.4 (1.8) | <0.001 |
| PHQ-9 Somatic Score (0–9), mean (SD) | 6.3 (2.1) | 1.4 (1.6) | <0.001 | 6.4 (2.1) | 1.5 (1.6) | <0.001 |
| Past-Year SSRI Use, n (%) | 180 (52) | 248 (21) | <0.001 | 141 (54) | 121 (21) | <0.001 |
| Past-Year TCA Use, n (%) | 82 (24) | 139 (12) | <0.001 | 69 (26) | 64 (11) | <0.001 |
| Past-Year Other Antidepressant Use, n (%) | 162 (47) | 227 (19) | <0.001 | 138 (52) | 122 (21) | <0.001 |
|
Covariates | ||||||
| Age, mean (SD), years | 51.1 (7.6) | 51.9 (8.4) | <0.001 | 52.0 (7.9) | 54.2 (9.8) | <0.001 |
| Women, n (%) | 10 (3) | 32 (3) | 0.87 | 32 (12) | 49 (9) | 0.10 |
| Race/Ethnicity, n (%) | 0.22 | 0.25 | ||||
| White | 77 (22) | 213 (18) | 56 (21) | 117 (20) | ||
| African American | 224 (64) | 824 (70) | 167 (64) | 395 (69) | ||
| Hispanic | 31 (9) | 101 (9) | 28 (11) | 40 (7) | ||
| Other | 16 (5) | 41 (3) | 12 (5) | 22 (4) | ||
| Cardiovascular Disease, n (%) | 75 (22) | 204 (17) | 0.071 | 72 (27) | 158 (28) | 0.96 |
| Diabetes, n (%) | 67 (19) | 234 (20) | 0.81 | 93 (35) | 154 (27) | 0.012 |
| Hypertension, n (%) | 90 (26) | 272 (23) | 0.29 | 72 (27) | 157 (27) | 0.97 |
| LDL Cholesterol, n (%) | 0.014 | 0.24 | ||||
| <100 mg/dL | 198 (57) | 587 (50) | 138 (52) | 261 (45) | ||
| 100–129 mg/dL | 98 (28) | 330 (28) | 67 (25) | 170 (30) | ||
| 130–159 mg/dL | 29 (8) | 168 (14) | 31 (12) | 79 (14) | ||
| ≥160 mg/dL | 14 (4) | 60 (5) | 15 (6) | 42 (7) | ||
| HDL Cholesterol, n (%) | 0.71 | 0.17 | ||||
| ≥60 mg/dL | 51 (15) | 175 (15) | 50 (19) | 87 (15) | ||
| 40–59 mg/dL | 134 (39) | 475 (40) | 112 (43) | 281 (49) | ||
| <40 mg/dL | 156 (45) | 496 (42) | 92 (35) | 186 (32) | ||
| Triglycerides ≥150 mg/dL, n (%) | 170 (49) | 513 (44) | 0.078 | 91 (35) | 157 (28) | 0.035 |
| Statin Use, n (%) | 79 (23) | 376 (32) | 0.001 | 100 (38) | 257 (45) | 0.067 |
| Hepatitis C Infection, n (%) | 209 (60) | 510 (43) | <0.001 | 98 (37) | 164 (29) | 0.012 |
| eGFR <60 ml/min/1.73 m2, n (%) | 58 (10) | 18 (7) | 0.45 | 92 (8) | 23 (7) | 0.13 |
| Hemoglobin, n (%) | 0.92 | 0.73 | ||||
| ≥14 g/dL | 179 (51) | 602 (51) | 140 (53) | 317 (55) | ||
| 12.0–13.9 g/dL | 130 (37) | 433 (37) | 106 (40) | 217 (38) | ||
| <12 g/dL | 39 (11) | 141 (12) | 16 (6) | 40 (7) | ||
| Body Mass Index ≥30 kg/m2, n (%) | 61 (18) | 185 (16) | 0.44 | 123 (47) | 265 (46) | 0.85 |
| Smoking, n (%) | <0.001 | 0.001 | ||||
| Never | 67 (19) | 300 (25) | 51 (19) | 146 (25) | ||
| Past | 70 (20) | 326 (28) | 63 (24) | 181 (32) | ||
| Current | 211 (61) | 553 (47) | 149 (57) | 245 (43) | ||
| Alcohol Use, n (%) | <0.001 | 0.092 | ||||
| Not Current | 104 (30) | 418 (35) | 99 (38) | 214 (37) | ||
| Not Hazardous | 66 (19) | 284 (24) | 36 (14) | 107 (19) | ||
| Hazardous or Heavy Episodic | 45 (13) | 194 (16) | 33 (13) | 87 (15) | ||
| Abuse/Dependence | 132 (38) | 279 (24) | 95 (36) | 166 (29) | ||
| Cocaine Abuse/Dependence, n (%) | 165 (47) | 386 (33) | <0.001 | 120 (46) | 197 (34) | 0.002 |
| CD4+ T-cell Count, n (%) | 0.002 | --- | --- | --- | ||
| ≥500/mm3 | 100 (29) | 432 (37) | ||||
| 200–499/mm3 | 163 (47) | 539 (46) | ||||
| <200/mm3 | 85 (24) | 204 (17) | ||||
| HIV-1 RNA Level ≥500 copies/mL, n (%) | 150 (43) | 366 (31) | <0.001 | --- | --- | --- |
| Antiretroviral Therapy Use, n (%) | 292 (84) | 1004 (85) | 0.57 | --- | --- | --- |
Note. PHQ-9 total score ≥10: depressed, <10: not depressed. All variables had complete data except (n): IL-6 (2,342), sCD14 (2,354), D-dimer (2,349), PHQ-9 total score (2,364), PHQ-9 cognitive/affective score (2,325), PHQ-9 somatic score (2,296), hypertension (2,385), LDL cholesterol (2,311), HDL cholesterol (2,319), triglycerides (2,349), eGFR (2,380), hemoglobin (2,385), body mass index (2,380), smoking (2,385), alcohol use (2,384), CD4+ T-cell count (1,542), and HIV-1 RNA level (1,542). HIV = human immunodeficiency virus. VACS = Veterans Aging Cohort Study. PHQ-9 = Patient Health Questionnaire-9. SSRI = selective serotonin reuptake inhibitor. TCA = tricyclic antidepressant. LDL = low-density lipoprotein. HDL = high-density lipoprotein. eGFR = estimated glomerular filtration rate. RNA = ribonucleic acid.
HIV factors were CD4+ T-cell count, HIV-1 RNA level, and ART use. CD4+ T-cell count and HIV-1 RNA level, measured as part of routine clinical care, were determined from VA Corporate Data Warehouse data obtained closest to the blood draw date (up to 180 days after). ART use was determined from VA pharmacy data for the timeframe 180 days prior to 7 days after the blood draw date and included the following classes: nucleoside reverse-transcriptase inhibitors (NRTI), nonnucleoside reverse-transcriptase inhibitors (NNRTI), and protease inhibitors (PI).
Data Analysis
Participant characteristics were stratified by depression status separately for veterans with and without HIV (see Table 1). Independent samples t tests or Wilcoxon rank-sum tests for continuous variables and chi-square tests for categorical variables were used to assess for differences in participant characteristics by depression status. For these tests, the biomarker variables were examined in their original scale.
Linear regression models were run to estimate the associations of total, cognitive/affective, and somatic depressive symptoms and antidepressant use with monocyte activation, inflammatory, and coagulation markers separately in veterans with and without HIV. Multiple imputations with five imputed datasets were generated using the “mi impute chained” command in Stata. Models were fitted in each imputed dataset and combined using Rubin’s rule to obtain pooled regression coefficients and standard errors.
For each outcome variable (log transformed sCD14, IL-6, and D-dimer), we constructed four models. Model 1 adjusted for demographics (age, sex, and race/ethnicity). Model 2 was our primary model and further adjusted for the following biomedical and behavioral factors that could confound associations between depression-related variables and the biomarkers: CVD, diabetes, hypertension, LDL cholesterol, HDL cholesterol, triglycerides, statin use, hepatitis C infection, renal function, hemoglobin, BMI, smoking, alcohol use, and cocaine abuse/dependence. Model 3 added the three antidepressant use variables (past-year SSRI, TCA, and other antidepressant use) to Model 2. Model 4 added the HIV factors (CD4+ T-cell count, HIV-1 RNA level, and ART use) to Model 3. Separate models were run for each PHQ-9 variable (z-scored total, cognitive/affective, and somatic scores). Models 1–3 were also run in HIV-negative veterans. Finally, the HIV x PHQ-9 total score, HIV x PHQ-9 cognitive/affective score, and HIV x PHQ-9 somatic score interactions were tested separately in Model 2’s that included the HIV main effect and involved entire cohort. Similarly, the HIV x SSRI use, HIV x TCA use, HIV x other antidepressant use interactions were tested simultaneously in a Model 3 that included the HIV main effect and involved entire cohort.
Because our outcome variables were log transformed, we present the exponentiated regression coefficient [exp(b)] and its 95% confidence interval (CI) for each association. For continuous z-scored exposure variables (PHQ-9 variables), we computed the percent change in each biomarker per one-unit (i.e., 1-SD) change in each PHQ-9 variable using the following equation: [exp(b) – 1] x 100. For dichotomous exposure variables (antidepressant use variables), we used the same equation to compute the percent change in each biomarker per one-unit change (i.e., switching from “no” to “yes”) in each antidepressant use variable.
Results
Characteristics of Veterans in the VACS Biomarker Cohort
Table 1 presents descriptive statistics for the biomarkers, depression variables, and covariates stratified by depression status (PHQ-9 total score ≥10: depressed, <10: not depressed,) separately for veterans with and without HIV. Of note, the SDs for the PHQ-9 total, cognitive/affective, and somatic scores were 6.6, 4.3, and 2.7 for the HIV-positive participants and 7.1, 4.7, and 2.9 for the HIV-negative participants.
Among the HIV-positive participants, sCD14 was higher in the depressed group, but significant group differences were not detected for IL-6 or D-dimer. As was expected, the PHQ-9 scores and antidepressant use rates were all higher in the depressed group. Significant differences in covariates were also observed, with the depressed group having higher levels/rates for hepatitis C infection, smoking, alcohol use, cocaine abuse/dependence, and HIV-1 RNA and lower levels/rates for age, LDL cholesterol, statin use, and CD4+ T-cell count.
Among the HIV-negative participants, IL-6 and all depression variables were higher in the depressed group. Significant group differences were not detected for sCD14 or D-dimer. The depressed group had higher levels/rates for diabetes, triglycerides, hepatitis C infection, smoking, and cocaine abuse/dependence and lower levels/rates for age and statin use.
Depressive Symptoms, Antidepressant Use, and Monocyte Activation
Among the HIV-positive participants (see Table 2), the PHQ-9 total, cognitive/affective, and somatic scores were all associated with higher sCD14 in Model 1 adjusting for demographic factors. The somatic score remained positively related to sCD14 in our primary model (Model 2) adjusting for several potential confounders, although the total and cognitive/affective scores did not. The exp(b) of 1.02 indicates that every 1-SD increase in the somatic score (2.7 points on a 0–9 scale) was associated with a 2% increase in sCD14 on average [(1.02 – 1) x 100]. Model 3, which added the antidepressant use variables, revealed that SSRI use was negatively related and TCA use was positively related to sCD14. SSRI use (versus no use) was associated with a 5% decrease in sCD14 on average, whereas TCA use was associated with a 7% increase in sCD14 on average. The somatic score remained positively related to sCD14 in Model 3. In Model 4 further adjusting for HIV factors, the pattern of results was similar, although the associations for the somatic score and SSRI use were slightly attenuated.
Table 2.
Linear Regression Models Examining Associations of Depressive Symptoms and Antidepressant Use with Soluble CD14 Separately among HIV-Positive and HIV-Negative Veterans in the VACS Biomarker Cohort
| HIV-Positive (n=1,546) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Model 1: Demographics-Adjusted Modela | Model 2 (Primary Model): Model 1 + Confoundersb | Model 3: Model 2 + Antidepressants | Model 4: Model 3 + HIV Factorsc | |||||
| exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | |
| PHQ-9 Total | 1.02* | 1.01–1.04 | 1.01† | 1.00–1.03 | 1.01 | 1.00–1.03 | 1.01 | 0.99–1.02 |
| PHQ-9 Cognitive/affective | 1.02* | 1.00–1.03 | 1.01 | 0.99–1.02 | 1.01 | 0.99–1.02 | 1.00 | 0.99–1.02 |
| PHQ-9 Somatic | 1.03* | 1.01–1.04 | 1.02* | 1.00–1.03 | 1.02* | 1.00–1.03 | 1.01† | 1.00–1.03 |
| SSRI Use | 0.95* | 0.91–1.00 | 0.96* | 0.92–1.00 | ||||
| TCA Use | 1.07* | 1.03–1.12 | 1.07* | 1.02–1.11 | ||||
| Other Antidepressant Use | 1.04† | 1.00–1.09 | 1.04† | 1.00–1.09 | ||||
| HIV-Negative (n=843) | ||||||||
| PHQ-9 Total | 1.01 | 0.99–1.02 | 0.99 | 0.98–1.01 | 0.98* | 0.96–1.00 | ||
| PHQ-9 Cognitive/affective | 1.00 | 0.99–1.02 | 0.99 | 0.97–1.01 | 0.98* | 0.96–1.00 | ||
| PHQ-9 Somatic | 1.01 | 0.99–1.03 | 1.00 | 0.98–1.01 | 0.99 | 0.97–1.00 | ||
| SSRI Use | 1.04† | 0.99–1.10 | ||||||
| TCA Use | 1.06* | 1.00–1.11 | ||||||
| Other Antidepressant Use | 1.01 | 0.96–1.07 | ||||||
Note. Soluble CD14 is log transformed. PHQ-9 variables are z scored. Statistically significant effects are bolded. HIV = human immunodeficiency virus. VACS = Veterans Aging Cohort Study. exp(b) = exponentiated regression coefficient. CI = confidence interval. PHQ-9 = Patient Health Questionnaire-9. SSRI = selective serotonin reuptake inhibitor. TCA = tricyclic antidepressant.
Adjusted for age, sex, and race/ethnicity.
Additionally adjusted for cardiovascular disease, diabetes, hypertension, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, statin use, hepatitis C infection, renal function, hemoglobin, BMI, smoking, alcohol use, and cocaine abuse/dependence.
Additionally adjusted for CD4+ T-cell count, HIV-1 RNA level, and ART use.
p < .05
p = .05-.10
Among the HIV-negative participants (see Table 2), the PHQ-9 scores were not associated with sCD14 in Models 1 or 2. Model 3 showed that TCA use was positively related to sCD14, with TCA use being associated with a 6% increase in sCD14 on average. The total and cognitive/affective scores were negatively related to sCD14 in Model 3; however, these associations may be unreliable, as they were not observed in Models 1 or 2.
The HIV x SSRI use interaction was significant (p = 0.023), indicating that the relationship between SSRI use and sCD14 was stronger among the HIV-positive (negative association) versus the HIV-negative (no association) participants. The interactions between HIV and PHQ-9 total score, cognitive/affective score, somatic score, TCA use, and other antidepressant use were not significant (all ps >0.18), demonstrating that these relationships did not differ by HIV status.
Depressive Symptoms, Antidepressant Use, and Systemic Inflammation
Among the HIV-positive participants (see Table 3), the PHQ-9 total and cognitive/affective scores were not related to IL-6 in Models 1 or 2. Although the somatic score was associated with higher IL-6 in Model 1, adjustment for potential confounders in Model 2 eliminated this relationship. Model 3 revealed that SSRI use was associated with 14% decrease and TCA use was associated with 14% increase in IL-6 on average. Model 4 results were similar to those of Model 3.
Table 3.
Linear Regression Models Examining Associations of Depressive Symptoms and Antidepressant Use with Interleukin-6 Separately among HIV-Positive and HIV-Negative Veterans in the VACS Biomarker Cohort
| HIV-Positive (n=1,546) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Model 1: Demographics-Adjusted Modela | Model 2 (Primary Model): Model 1 + Confoundersb | Model 3: Model 2 + Antidepressants | Model 4: Model 3 + HIV Factorsc | |||||
| exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | |
| PHQ-9 Total | 1.04† | 1.00–1.08 | 1.00 | 0.96–1.04 | 1.00 | 0.96–1.04 | 0.99 | 0.95–1.02 |
| PHQ-9 Cognitive/affective | 1.03 | 0.99–1.07 | 0.99 | 0.95–1.03 | 0.99 | 0.95–1.03 | 0.97 | 0.94–1.01 |
| PHQ-9 Somatic | 1.06* | 1.02–1.10 | 1.01 | 0.98–1.05 | 1.02 | 0.98–1.06 | 1.01 | 0.97–1.05 |
| SSRI Use | 0.86* | 0.76–0.96 | 0.86* | 0.77–0.96 | ||||
| TCA Use | 1.14* | 1.02–1.28 | 1.13* | 1.01–1.27 | ||||
| Other Antidepressant Use | 1.12† | 0.99–1.26 | 1.13† | 1.00–1.27 | ||||
| HIV-Negative (n=843) | ||||||||
| PHQ-9 Total | 1.11* | 1.05–1.18 | 1.08* | 1.02–1.14 | 1.05† | 0.99–1.12 | ||
| PHQ-9 Cognitive/affective | 1.10* | 1.04–1.16 | 1.07* | 1.01–1.13 | 1.04 | 0.98–1.10 | ||
| PHQ-9 Somatic | 1.13* | 1.06–1.19 | 1.09* | 1.03–1.15 | 1.06* | 1.00–1.12 | ||
| SSRI Use | 1.12 | 0.94–1.33 | ||||||
| TCA Use | 1.05 | 0.88–1.24 | ||||||
| Other Antidepressant Use | 1.07 | 0.90–1.27 | ||||||
Note. Interleukin-6 is log transformed. PHQ-9 variables are z scored. Statistically significant effects are bolded. HIV = human immunodeficiency virus. VACS = Veterans Aging Cohort Study. exp(b) = exponentiated regression coefficient. CI = confidence interval. PHQ-9 = Patient Health Questionnaire-9. SSRI = selective serotonin reuptake inhibitor. TCA = tricyclic antidepressant.
Adjusted for age, sex, and race/ethnicity.
Additionally adjusted for cardiovascular disease, diabetes, hypertension, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, statin use, hepatitis C infection, renal function, hemoglobin, BMI, smoking, alcohol use, and cocaine abuse/dependence.
Additionally adjusted for CD4+ T-cell count, HIV-1 RNA level, and ART use.
p < .05
p = .05-.10
Among the HIV-negative participants (see Table 3), the PHQ-9 total, cognitive/affective, and somatic scores were all associated with higher IL-6 in Models 1 and 2. In Model 2, every 1-SD increase in the PHQ-9 total (7.1 points on a 0–27 scale), cognitive/affective (4.7 points on a 0–18 scale), and somatic score (2.9 points on a 0–9 scale) was associated with a 8%, 7%, and 9% increase in IL-6 on average, respectively. In Model 3, the somatic score remained positively related to IL-6, although no antidepressant use variable was associated with IL-6.
The interactions between HIV and PHQ-9 total score (p = 0.013), cognitive/affective score (p = 0.015), somatic score (p = 0.025), and SSRI use (p = 0.005) were significant. These results indicate that (a) relationships between the PHQ-9 scores and IL-6 were stronger among the HIV-negative (positive associations) versus the HIV-positive (no associations) participants and that (b) the relationship between SSRI use and IL-6 was stronger among the HIV-positive (negative association) versus the HIV-negative (no association) participants. The HIV x TCA use and HIV x other antidepressant use interactions were not significant (both ps >0.41).
Depressive Symptoms, Antidepressant Use, and Altered Coagulation
Among the HIV-positive participants (see Table 4), the PHQ-9 total, cognitive/affective, and somatic scores were all associated with higher D-dimer in Model 1. However, only the relationship with the somatic score remained significant in Model 2, with every 1-SD increase in that score (2.7 points on a 0–9 scale) being associated with a 6% increase in D-dimer on average. In Model 3, no antidepressant use variable was associated with D-dimer. The association for the PHQ-9 somatic score was slightly attenuated in Models 3 and 4, falling short of significance.
Table 4.
Linear Regression Models Examining Associations of Depressive Symptoms and Antidepressant Use with D-Dimer Separately among HIV-Positive and HIV-Negative Veterans in the VACS Biomarker Cohort
| HIV-Positive (n=1,546) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Model 1: Demographics-Adjusted Modela | Model 2 (Primary Model): Model 1 + Confoundersb | Model 3: Model 2 + Antidepressants | Model 4: Model 3 + HIV Factorsc | |||||
| exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | exp(b) | 95% CI | |
| PHQ-9 Total | 1.09* | 1.03–1.15 | 1.05† | 1.00–1.11 | 1.04 | 0.99–1.10 | 1.03 | 0.97–1.08 |
| PHQ-9 Cognitive/affective | 1.07* | 1.02–1.13 | 1.04 | 0.99–1.10 | 1.03 | 0.98–1.09 | 1.01 | 0.96–1.07 |
| PHQ-9 Somatic | 1.10* | 1.04–1.16 | 1.06* | 1.00–1.11 | 1.05† | 0.99–1.11 | 1.04 | 0.98–1.09 |
| SSRI Use | 0.98 | 0.84–1.13 | 0.99 | 0.86–1.15 | ||||
| TCA Use | 0.97 | 0.84–1.13 | 0.98 | 0.84–1.13 | ||||
| Other Antidepressant Use | 1.14 | 0.97–1.33 | 1.15† | 0.98–1.34 | ||||
| HIV-Negative (n=843) | ||||||||
| PHQ-9 Total | 1.02 | 0.96–1.07 | 1.01 | 1.00–1.02 | 1.00 | 0.94–1.06 | ||
| PHQ-9 Cognitive/affective | 1.01 | 0.96–1.07 | 1.00 | 0.95–1.06 | 1.00 | 0.94–1.06 | ||
| PHQ-9 Somatic | 1.02 | 0.96–1.08 | 1.00 | 0.95–1.06 | 1.00 | 0.94–1.06 | ||
| SSRI Use | 1.09 | 0.92–1.29 | ||||||
| TCA Use | 0.97 | 0.83–1.15 | ||||||
| Other Antidepressant Use | 0.95 | 0.81–1.13 | ||||||
Note. D-dimer is log transformed. PHQ-9 variables are z scored. Statistically significant effects are bolded. HIV = human immunodeficiency virus. VACS = Veterans Aging Cohort Study. exp(b) = exponentiated regression coefficient. CI = confidence interval. PHQ-9 = Patient Health Questionnaire-9. SSRI = selective serotonin reuptake inhibitor. TCA = tricyclic antidepressant.
Adjusted for age, sex, and race/ethnicity.
Additionally adjusted for cardiovascular disease, diabetes, hypertension, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, statin use, hepatitis C infection, renal function, hemoglobin, BMI, smoking, alcohol use, and cocaine abuse/dependence.
Additionally adjusted for CD4+ T-cell count, HIV-1 RNA level, and ART use.
p < .05
p = .05-.10
Among the HIV-negative participants (see Table 4), no PHQ-9 score or antidepressant use variable was associated with D-dimer across models.
The interactions between HIV and PHQ-9 total score, cognitive/affective score, somatic score, SSRI use, TCA use, and other antidepressant use were all not significant (all ps >0.15).
Discussion
We report novel independent associations between depression-related variables and monocyte activation, systemic inflammation, and altered coagulation – which are all implicated in HIV-CVD (7). In HIV-positive veterans, greater somatic depressive symptoms were related to higher sCD14 and D-dimer in our primary model adjusting for demographics and potential medical and behavioral confounders. Further adjustment for antidepressant use and HIV factors slightly attenuated these relationships. In addition, past-year SSRI use was linked to lower and past-year TCA use was linked to higher sCD14 and IL-6 in HIV-positive veterans.
The observed relationships are small (2%−14% biomarker change), and their clinical relevance is unclear. Nevertheless, because we found relationships unlikely due to chance, our results raise the possibilities that (a) monocyte activation and altered coagulation may be two of the likely many mechanisms through which depressive symptoms may increase HIV-CVD risk and that (b) TCA treatment may elevate and SSRI treatment may attenuate HIV-CVD risk due to their possible effects on monocyte activation and systemic inflammation. Future prospective studies, ideally randomized controlled trials, are needed to rigorously evaluate these possibilities.
Our study makes important contributions to the HIV literature examining relationships between depression-related factors and CVD-relevant biomarkers, as it has the largest sample to date, included an HIV-negative group, adjusted for several potential confounders, and examined depressive symptoms clusters and antidepressant use. HIV studies (14–19) have detected depression-immune activation associations, as we did here for sCD14. We extend those findings with evidence that the somatic symptoms may be contributing more to this link than the cognitive/affective symptoms. HIV studies have also reported depression-inflammation associations (19–24), conflicting with the lack of relationships for IL-6 here. Of relevance, Norcini Pala and colleagues (22) examined depressive symptom subtypes, observing elevated IL-6 in the group with severe/moderate somatic symptoms. A possible explanation for these discrepant results is that our models included more extensive adjustment for potential confounders than most studies. Consistent with this idea, greater somatic symptoms were associated with higher IL-6 in our demographics-adjusted model but not in our subsequent models. To our knowledge, our study is the first to report a relationship between greater depressive symptoms and higher levels of a coagulation marker (D-dimer) in HIV. Our findings further suggest that this is another link to which the somatic symptoms may be contributing more than the cognitive/affective symptoms.
There are five plausible, non-mutually exclusive explanations for the relationships we observed. One, elevated depressive symptoms could lead to immune activation and altered coagulation through the putative mechanisms underlying depression-immune function relationships, such as autonomic dysfunction, hypothalamic-pituitary-adrenal axis dysregulation, and increased adiposity (40, 41). In line with our stronger associations of the somatic cluster with sCD14 and D-dimer, others have found the somatic depressive symptoms to be more strongly related to autonomic dysfunction (42) and abdominal obesity (43) than the cognitive depressive symptoms. Two, depression has been associated with markers of increased gut permeability (44, 45), which promotes immune activation due to increased microbial translocation from the gut to circulation (46). Three, a candidate behavioral mechanism specific to HIV is ART nonadherence, as depression is associated with poorer adherence to HIV treatment (47), with resulting poorer control of viremia-associated immune activation. Four, increased release of proinflammatory cytokines in the central nervous system – due, in part, to HIV-related immune activation – could produce elevated depressive symptoms, particularly the somatic symptoms (“sickness behavior”), via their effects on neurobiological systems implicated in depression (48–50). Five, confounders, such as HIV severity, could be contributing to the observed relationship, especially considering the overlap between somatic depressive symptoms and HIV symptoms (51). However, our primary models adjusted for several comorbid conditions, and further adjustment for viral load and CD4+ T-cell count only slightly attenuated associations.
Although we are not aware of previous HIV studies examining relationships between antidepressant use and CVD-relevant biomarkers, meta-analytic evidence from non-HIV studies indicates that IL-6 decreases following SSRI, but not TCA, treatment (26). Furthermore, a systematic review of non-HIV studies concluded that both SSRIs and TCAs appear to reduce inflammatory markers (52). Those prior reviews are consistent with our finding that SSRI use is associated with lower immune activation (sCD14) and systemic inflammation (IL-6) in people with HIV. SSRIs may have direct immunomodulary effects or indirect effects through improved depression and HIV treatment adherence (53, 54). SSRIs may also have antimicrobial effects and, thus, may restore gut microbiota balance (55), thereby decreasing gut permeability and the consequent bacterial translocation, immune activation, and systemic inflammation. Those prior reviews, however, are inconsistent with our finding that TCA use is related to higher sCD14 and IL-6 levels. One possible explanation for this discrepancy is that TCA use is acting as a marker of treatment-resistant depression, as SSRIs are typically recommended as first-line antidepressants (56). Another possible explanation is that TCA use is acting as a proxy for conditions not in our models, as these medications are often prescribed for other conditions linked with immune/inflammatory activation, such as insomnia and chronic pain (33).
We also observed noteworthy associations in HIV-negative veterans. First, we detected positive associations of total, cognitive/affective, and somatic depressive symptoms with IL-6. Second, we found that TCA use was linked to higher sCD14. Our findings are in line with a recent meta-analysis of non-HIV studies showing that depression is associated with higher inflammatory markers predictive of CVD (57). We also detected some differences in relationships by HIV status. SSRI use was more strongly associated with lower sCD14 and IL-6 in HIV-positive veterans, raising the possibility that SSRI therapy (currently first-line antidepressants) may be a promising approach to reducing monocyte and inflammatory activation in depressed people with HIV. In addition, PHQ-9 scores were more strongly related to higher IL-6 in HIV-negative veterans.
Some limitations deserve attention. First, our cross-sectional data prevents us from drawing inferences regarding directionality of associations, and both directions are plausible. More specifically, because the immune activation, inflammatory, and coagulation markers were assessed at only one time point in the VACS Biomarker Cohort, we could not examine associations between baseline depression variables and changes over time in these biomarkers. Second, our yes/no past-year antidepressant use variables lack precision, which could have led us to underestimate their true effect sizes. Incorporating dose, duration, and recency of use would provide a more nuanced assessment. In addition, we do not know the treatment indication for the detected antidepressant use. To illustrate, 28% of HIV-positive and HIV-negative veterans classified as not depressed had past-year antidepressant use – a group that likely consists of those whose prior depression had been successfully treated and those taking antidepressants for other indications, such as anxiety, insomnia, or pain conditions (33). Third, it is unknown whether our findings extend to women. The VACS Biomarker Cohort, especially the HIV-positive group, contains a low percentage of women, which may explain the observed lower prevalence of depression (23%) in veterans with HIV compared to prior estimates (3).
In sum, our findings suggest that (a) monocyte activation and altered coagulation may be two pathways through which depression increases HIV-CVD risk and that (b) TCAs may elevate and SSRIs may attenuate HIV-CVD risk by influencing monocyte activation and systemic inflammation. Ultimately, elucidating the mechanisms underlying the depression to HIV-CVD association could identify additional targets (beyond depression itself) for novel HIV-CVD prevention efforts. Moreover, determining the effect of various classes of antidepressants on HIV-CVD risk and relevant biomarkers could lead to the development of antidepressant algorithms to simultaneously treat depression and lower CVD risk in people with HIV.
Acknowledgments
Conflicts of Interest and Sources of Funding: Dr. Stewart reports funding from the National Institutes of Health and Indiana University. Dr. So-Armah reports funding from the National Institutes of Health and the Providence/Boston Center for AIDS Research. Dr. Gupta reports funding from the National Institutes of Health, Indiana University, advisory board fees from Gilead Sciences and GlaxoSmithKline/ViiV, and travel support to present data at scientific conferences from Gilead Sciences. Dr. Freiberg reports funding from the National Institutes of Health. For the remaining authors, no conflicts of interest were declared. The Veterans Aging Cohort Study was funded by grant U10AA13566 from the National Institute on Alcohol Abuse and Alcoholism and Veterans Health Administration Public Health Strategic Health Core Group. This analysis was funded, in part, by grant R01HL126557 from the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or Department of Veterans Affairs.
Acronyms:
- ART
antiretroviral therapy
- AUDIT-C
Alcohol Use Disorders Identification Test
- CI
confidence interval
- CV
coefficients of variability
- CVD
cardiovascular disease
- eGFR
estimated glomerular filtration rate
- HDL
high-density lipoprotein
- HIV
human immunodeficiency virus
- ICD-9
International Classification of Diseases, Ninth Revision
- IL-6
interleukin-6
- LDL
low-density lipoprotein
- NNRTI
nonnucleoside reverse-transcriptase inhibitors
- NRTI
nucleoside reverse-transcriptase inhibitors
- PHQ-9
Patient Health Questionnaire-9
- PI
protease inhibitors
- sCD14
soluble CD14
- SD
standard deviation
- SSRI
selective serotonin reuptake inhibitor
- TCA
tricyclic antidepressant
- VA
Veterans Affair
- VACS
Veterans Aging Cohort Study
References
- 1.Shah ASV, Stelzle D, Lee KK, Beck EJ, Alam S, Clifford S, Longenecker CT, Strachan F, Bagchi S, Whiteley W, Rajagopalan S, Kottilil S, Nair H, Newby DE, McAllister DA, Mills NL. Global burden of atherosclerotic cardiovascular disease in people living with HIV. Circulation. 2018;138:1100–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Freiberg MS, Chang CC, Kuller LH, Skanderson M, Lowy E, Kraemer KL, Butt AA, Bidwell Goetz M, Leaf D, Oursler KA, Rimland D, Rodriguez Barradas M, Brown S, Gibert C, McGinnis K, Crothers K, Sico J, Crane H, Warner A, Gottlieb S, Gottdiener J, Tracy RP, Budoff M, Watson C, Armah KA, Doebler D, Bryant K, Justice AC. HIV infection and the risk of acute myocardial infarction. JAMA Intern Med 2013;173:614–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Nanni MG, Caruso R, Mitchell AJ, Meggiolaro E, Grassi L. Depression in HIV infected patients: a review. Curr Psychiat Rep 2015;17. [DOI] [PubMed]
- 4.Van der Kooy K, van Hout H, Marwijk H, Marten H, Stehouwer C, Beekman A. Depression and the risk for cardiovascular diseases: systematic review and meta analysis. Int J Geriatr Psychiatry 2007;22:613–26. [DOI] [PubMed] [Google Scholar]
- 5.Khambaty T, Stewart JC, Gupta SK, Chang CCH, Bedimo RJ, Budoff MJ, Butt AA, Crane H, Gibert CL, Leaf DA, Rimland D, Tindle HA, So-Armah KA, Justice AC, Freiberg MS. Association between depressive disorders and incident acute myocardial infarction in human immunodeficiency virus-infected adults: Veterans Aging Cohort Study. JAMA Cardiol 2016;1:929–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.White JR, Chang CCH, So-Armah KA, Stewart JC, Gupta SK, Butt AA, Gibert CL, Rimland D, Rodriguez-Barradas MC, Leaf DA, Bedimo RJ, Gottdiener JS, Kop WJ, Gottlieb SS, Budoff MJ, Khambaty T, Tindle HA, Justice AC, Freiberg MS. Depression and human immunodeficiency virus infection are risk factors for incident heart failure among veterans: Veterans Aging Cohort Study. Circulation. 2015;132:1630–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Vachiat A, McCutcheon K, Tsabedze N, Zachariah D, Manga P. HIV and ischemic heart disease. J Am Coll Cardiol 2017;69:73–82. [DOI] [PubMed] [Google Scholar]
- 8.Shive CL, Jiang W, Anthony DD, Lederman MM. Soluble CD14 is a nonspecific marker of monocyte activation. AIDS 2015;29:1263–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bas S, Gauthier BR, Spenato U, Stingelin S, Gabay C. CD14 is an acute-phase protein. J Immunol 2004;172:4470–9. [DOI] [PubMed] [Google Scholar]
- 10.Hileman CO, Longenecker CT, Carman TL, Milne GL, Labbato DE, Storer NJ, White CA, McComsey GA. Elevated D-dimer is independently associated with endothelial dysfunction: a cross-sectional study in HIV-infected adults on antiretroviral therapy. Antivir Ther 2012;17:1345–9. [DOI] [PubMed] [Google Scholar]
- 11.Hsu DC, Ma YF, Hur S, Li D, Rupert A, Scherzer R, Kalapus SC, Deeks S, Sereti I, Hsue PY. Plasma IL-6 levels are independently associated with atherosclerosis and mortality in HIV-infected individuals on suppressive antiretroviral therapy. AIDS 2016;30:2065–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Longenecker CT, Jiang Y, Orringer CE, Gilkeson RC, Debanne S, Funderburg NT, Lederman MM, Storer N, Labbato DE, McComsey GA. Soluble CD14 is independently associated with coronary calcification and extent of subclinical vascular disease in treated HIV infection. AIDS 2014;28:969–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Duprez DA, Neuhaus J, Kuller LH, Tracy R, Belloso W, De Wit S, Drummond F, Lane HC, Ledergerber B, Lundgren J, Nixon D, Paton NI, Prineas RJ, Neaton JD. Inflammation, coagulation and cardiovascular disease in HIV-infected individuals. PLoS One 2012;7:e44454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Evans DL, Ten Have TR, Douglas SD, Gettes DR, Morrison M, Chiappini MS, Brinker-Spence P, Job C, Mercer DE, Wang YL, Cruess D, Dube B, Dalen EA, Brown T, Bauer R, Petitto JM. Association of depression with viral load, CD8 T lymphocytes, and natural killer cells in women with HIV infection. Am J Psychiatry. 2002;159:1752–9. [DOI] [PubMed] [Google Scholar]
- 15.Hellmuth J, Colby D, Valcour V, Suttichom D, Spudich S, Ananworanich J, Prueksakaew P, Sailasuta N, Allen I, Jagodzinski LL, Slike B, Ochi D, Paul R, on behalf of the RV254/SEARCH 010 Study Group. Depression and anxiety are common in acute HIV infection and associate with plasma immune activation. AIDS Behav 2017;21:3238–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kemeny ME, Weiner H, Duran R, Taylor SE, Visscher B, Fahey JL. Immune-system changes after the death of a partner in HIV-positive gay men. Psychosom Med 1995;57:547–54. [DOI] [PubMed] [Google Scholar]
- 17.Kemeny ME, Weiner H, Taylor SE, Schneider S, Visscher B, Fahey JL. Repeated bereavement, depressed mood, and immune parameters in HIV-seropositive and seronegative gay men. Health Psychol 1994;13:14–24. [DOI] [PubMed] [Google Scholar]
- 18.Schroecksnadel K, Sarcletti M, Winkler C, Mumelter B, Weiss G, Fuchs D, Kemmler G, Zangerle R. Quality of life and immune activation in patients with HIV-infection. Brain Behav Immun 2008;22:881–9. [DOI] [PubMed] [Google Scholar]
- 19.Warriner EM, Rourke SB, Rourke BP, Rubenstein S, Millikin C, Buchanan L, Connelly P, Hyrcza M, Ostrowski M, Der S, Gough K. Immune activation and neuropsychiatric symptoms in HIV infection. J Neuropsych Clin N 2010;22:321–8. [DOI] [PubMed] [Google Scholar]
- 20.Fumaz CR, Gonzalez-Garcia M, Borras X, Munoz-Moreno JA, Perez-Alvarez N, Mothe B, Brander C, Ferrer MJ, Puig J, Llano A, Fernandez-Castro J, Clotet B. Psychological stress is associated with high levels of IL-6 in HIV-1 infected individuals on effective combined antiretroviral treatment. Brain Behav Immun 2012;26:568–72. [DOI] [PubMed] [Google Scholar]
- 21.Musinguzi K, Obuku A, Nakasujja N, Birabwa H, Nakku J, Levin J, Kinyanda E. Association between major depressive disorder and pro-inflammatory cytokines and acute phase proteins among HIV-1 positive patients in Uganda. BMC Immunol 2018;19. [DOI] [PMC free article] [PubMed]
- 22.Norcini Pala A, Steca P, Bagrodia R, Helpman L, Colangeli V, Viale P, Wainberg ML. Subtypes of depressive symptoms and inflammatory biomarkers: an exploratory study on a sample of HIV-positive patients. Brain Behav Immun 2016;56:105–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Poudel-Tandukar K, Berrone-Johnson ER, Palmer PH, Poudel KC. C-reactive protein and depression in persons with human immunodeficiency virus infection: the Positive Living with HIV (POLH) Study. Brain Behav Immun 2014;42:89–95. [DOI] [PubMed] [Google Scholar]
- 24.Rivera-Rivera Y, Garcia Y, Toro V, Cappas N, Lopez P, Yamamura Y, Rivera-Amill V. Depression correlates with increased plasma levels of inflammatory cytokines and a dysregulated oxidant/antioxidant balance in HIV-1-infected subjects undergoing antiretroviral therapy. J Clin Cell Immunol 2014;5. [DOI] [PMC free article] [PubMed]
- 25.Majd M, Saunders EFH, Engeland CG. Inflammation and the dimensions of depression: a review. Front Neuroendocrinol 2019:100800. [DOI] [PMC free article] [PubMed]
- 26.Hannestad J, DellaGioia N, Bloch M. The effect of antidepressant medication treatment on serum levels of inflammatory cytokines: a meta-analysis. Neuropsychopharmacology. 2011;36:2452–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Justice AC, Dombrowski E, Conigliaro J, Fultz SL, Gibson D, Madenwald T, Goulet J, Simberkoff M, Butt AA, Rimland D, Rodriguez-Barradas MC, Gibert CL, Oursler KA, Brown S, Leaf DA, Goetz MB, Bryant K. Veterans Aging Cohort Study (VACS): overview and description. Med Care. 2006;44:S13–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Armah KA, McGinnis K, Baker J, Gibert C, Butt AA, Bryant KJ, Goetz M, Tracy R, Oursler KK, Rimland D, Crothers K, Rodriguez-Barradas M, Crystal S, Gordon A, Kraemer K, Brown S, Gerschenson M, Leaf DA, Deeks SG, Rinaldo C, Kuller LH, Justice A, Freiberg M. HIV status, burden of comorbid disease, and biomarkers of inflammation, altered coagulation, and monocyte activation. Clin Infect Dis 2012;55:126–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001;16:606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Justice AC, McGinnis KA, Atkinson JH, Heaton RK, Young C, Sadek J, Madenwald T, Becker JT, Conigliaro J, Brown ST, Rimland D, Crystal S, Simberkoff M. Psychiatric and neurocognitive disorders among HIV-positive and negative veterans in care: Veterans Aging Cohort Five-Site Study. AIDS 2004;18 Suppl 1:S49–59. [PubMed] [Google Scholar]
- 31.Kroenke K, Spitzer RL. The PHQ-9: A new depression diagnostic and severity measure. Psychiatr Ann 2002;32:509–15. [Google Scholar]
- 32.Patel JS, Oh Y, Rand KL, Wu W, Cyders MA, Kroenke K, Stewart JC. Measurement invariance of the Patient Health Questionnaire-9 (PHQ-9) depression screener in U.S. adults across sex, race/ethnicity, and education level: NHANES 2005–2014. Depress Anxiety. 2019;36:813–823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wong J, Motulsky A, Eguale T, Buckeridge DL, Abrahamowicz M, Tamblyn R. Treatment indications for antidepressants prescribed in primary care in Quebec, Canada, 2006–2015. JAMA 2016;315:2230–2. [DOI] [PubMed] [Google Scholar]
- 34.Justice AC, Freiberg MS, Tracy R, Kuller L, Tate JP, Goetz MB, Fiellin DA, Vanasse GJ, Butt AA, Rodriguez-Barradas MC, Gibert C, Oursler KA, Deeks SG, Bryant K. Does an index composed of clinical data reflect effects of inflammation, coagulation, and monocyte activation on mortality among those aging with HIV? Clin Infect Dis 2012;54:984–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Butt AA, Fultz SL, Kwoh CK, Kelley D, Skanderson M, Justice AC. Risk of diabetes in HIV infected veterans pre- and post-HAART and the role of HCV coinfection. Hepatol 2004;40:115–9. [DOI] [PubMed] [Google Scholar]
- 36.McGinnis KA, Brandt CA, Skanderson M, Justice AC, Shahrir S, Butt AA, Brown ST, Freiberg MS, Gibert CL, Goetz MB, Kim JW, Pisani MA, Rimland D, Rodriguez-Barradas MC, Sico JJ, Tindle HA, Crothers K. Validating smoking data from the Veteran’s Affairs Health Factors dataset, an electronic data source. Nicotine Tob Res 2011;13:1233–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA, for the Ambulatory Care Quality Improvement Project (ACQUIP). The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Arch Intern Med 1998;158:1789–95. [DOI] [PubMed] [Google Scholar]
- 38.Carrico AW, Cherenack EM, Roach ME, Riley ED, Oni O, Dilworth SE, Shoptaw S, Hunt P, Roy S, Pallikkuth S, Pahwa S. Substance-associated elevations in monocyte activation among methamphetamine users with treated HIV infection. AIDS 2018;32:767–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Swendsen JD, Merikangas KR. The comorbidity of depression and substance use disorders. Clin Psychol Rev 2000;20:173–89. [DOI] [PubMed] [Google Scholar]
- 40.Irwin MR, Miller AH. Depressive disorders and immunity: 20 years of progress and discovery. Brain Behav Immun 2007;21:374–83. [DOI] [PubMed] [Google Scholar]
- 41.Miller GE, Freedland KE, Carney RM, Stetler CA, Banks WA. Pathways linking depression, adiposity, and inflammatory markers in healthy young adults. Brain Behav Immun 2003;17:276–85. [DOI] [PubMed] [Google Scholar]
- 42.de Jonge P, Mangano D, Whooley MA. Differential association of cognitive and somatic depressive symptoms with heart rate variability in patients with stable coronary heart disease: findings from the Heart and Soul Study. Psychosom Med 2007;69:735–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Wiltink J, Michal M, Wild PS, Zwiener I, Blettner M, Munzel T, Schulz A, Kirschner Y, Beutel ME. Associations between depression and different measures of obesity (BMI, WC, WHtR, WHR). BMC Psychiatry 2013;13:223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Keri S, Szabo C, Kelemen O. Expression of Toll-Like Receptors in peripheral blood mononuclear cells and response to cognitive-behavioral therapy in major depressive disorder. Brain Behav Immun 2014;40:235–43. [DOI] [PubMed] [Google Scholar]
- 45.Maes M, Kubera M, Leunis JC, Berk M. Increased IgA and IgM responses against gut commensals in chronic depression: further evidence for increased bacterial translocation or leaky gut. J Affect Disord 2012;141:55–62. [DOI] [PubMed] [Google Scholar]
- 46.Amar J, Ruidavets JB, Bal Dit Sollier C, Bongard V, Boccalon H, Chamontin B, Drouet L, Ferrieres J. Soluble CD14 and aortic stiffness in a population-based study. J Hypertens 2003;21:1869–77. [DOI] [PubMed] [Google Scholar]
- 47.Gonzalez JS, Batchelder AW, Psaros C, Safren SA. Depression and HIV/AIDS treatment nonadherence: a review and meta-analysis. J Acquir Immune Defic Syndr 2011;58:181–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Del Guerra FB, Fonseca JLI, Figueiredo VM, Ziff EB, Konkiewitz EC. Human immunodeficiency virus-associated depression: contributions of immuno-inflammatory, monoaminergic, neurodegenerative, and neurotrophic pathways. J Neurovirol 2013;19:314–27. [DOI] [PubMed] [Google Scholar]
- 49.Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci 2008;9:46–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Case SM, Stewart JC. Race/ethnicity moderates the relationship between depressive symptom severity and C-reactive protein: 2005–2010 NHANES data. Brain Behav Immun 2014;41:101–8. [DOI] [PubMed] [Google Scholar]
- 51.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 2000;188:662–70. [DOI] [PubMed] [Google Scholar]
- 52.Eyre HA, Lavretsky H, Kartika J, Qassim A, Baune BT. Modulatory effects of antidepressant classes on the innate and adaptive immune system in depression. Pharmacopsychiatry. 2016;49:85–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Shimbo D, Davidson KW, Haas DC, Fuster V, Badimon JJ. Negative impact of depression on outcomes in patients with coronary artery disease: mechanisms, treatment considerations, and future directions. J Thromb Haemost 2005;3:897–908. [DOI] [PubMed] [Google Scholar]
- 54.Sin NL, DiMatteo MR. Depression treatment enhances adherence to antiretroviral therapy: a meta-analysis. Ann Behav Med 2014;47:259–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Macedo D, Filho AJMC, Soares de Sousa CN, Quevedo J, Barichello T, Júnior HVN, Freitas de Lucena D. Antidepressants, antimicrobials or both? Gut microbiota dysbiosis in depression and possible implications of the antimicrobial effects of antidepressant drugs for antidepressant effectiveness. J Affect Disord 2017;208:22–32. [DOI] [PubMed] [Google Scholar]
- 56.Unutzer J, IMPACT Study Investigators. Project IMPACT intervention manual: improving care for depression in late life. Los Angeles: UCLA: NPI Center for Health Services Research; 1999. [Google Scholar]
- 57.Haapakoski R, Mathieu J, Ebmeier KP, Alenius H, Kivimaki M. Cumulative meta-analysis of interleukins 6 and 1 beta, tumour necrosis factor alpha and C-reactive protein in patients with major depressive disorder. Brain Behav Immun 2015;49:206–15. [DOI] [PMC free article] [PubMed] [Google Scholar]