Infection with the human immunodeficiency virus type 1 (HIV-1) is characterized by disruption in the normal functions of the immune and metabolic systems. At the immune system level, HIV-1 causes its deleterious effects in the host mainly through the progressive loss of CD4⁺ T cells, eventually leading to the development of Acquired Immune Deficiency Syndrome (AIDS) [1]. The effects of HIV-1 are also evident at the metabolic level, specifically in terms of oxidant/antioxidant balance. Under normal conditions a homeostatic balance exists between the production and elimination of Reactive Oxygen Species (ROS). It is well known that many diseases involve the dysregulation of this system by the disruption of antioxidants and the overload of oxidants. HIV-1 infection promotes the shifting of the oxidant/antioxidant system, causing oxidative stress due to the depletion of the tissue antioxidant defense system [2-4].

Because of the availability of antiretroviral therapy, having the HIV-1 infection has changed from being a sure death sentence to a manageable, chronic disease [5]. Compared to the general population, the treated HIV-infected population is showing increased longevity but also is exhibiting greater development of some non-AIDS comorbid diseases. Mental health comorbidities among the population with infectious diseases are getting much importance in primary health care [6]. The medical burden of depression is increasing and studies have shown that people who are infected with HIV are more likely than the general population to develop depression, the most common psychiatric diagnosis in people living with HIV/AIDS [7-9]. HIV/AIDS causes psychological trauma and adverse effects to the nervous system leading to the development of mania, depression, and other cognitive disorders [10]. In the USA, the estimated prevalence of depression is 2-10 times higher in HIV/AIDS patients than it is in the general population [9]. According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-V), depression is classified as a mood disorder characterized by the presence of a severely depressive mood for at least 2 weeks. Some of the most important changes in patients who suffer from major depression are changes related to appetite, weight, sleep patterns, and psychomotor behavior, and those symptoms can both increase and decrease among patients [11]. Interestingly, immune system function changes are present in major depression too [12], and it has been suggested that these immune changes may be more apparent in depressed patients infected with HIV-1 since their immune systems are already compromised [12]. Recently, Song and Wang reviewed the correlation between pro-inflammatory cytokines, free radical species, and oxidants in the etiology of depression [13,14]. Erlandson and colleagues demonstrated that development of impaired physical function and frailty are associated with markers for immune activation such as Interleukin-6 (IL-6), high sensitive C Reactive Protein (hs-CRP) and Tumor Necrosis Factor alpha (TNF-α), and inflammation [15]. Another study revealed that depressed patients had significantly increased production of oxidative stress, autoimmune and inflammatory markers against Lipopolysaccharide (LPS) than do the controls [16]. Furthermore, decreased intracellular levels of free Glutathione (GSH) and increased oxidized GSH were found to be correlated with increased levels of pro-inflammatory cytokines and Malondialdehyde (MDA) in macrophages from HIV-infected patients compared to controls [17].

As there is strong evidence for the dysregulation of inflammatory response and oxidative imbalance in HIV-1 as well as in depressive disorders, in the present study we wanted to assess markers related to the immunological, pro-inflammatory, and oxidant/antioxidant balance in HIV-1 infected subjects with and without depressive symptoms. The findings of this study led us to hypothesize that the successful management of depression would lead to increased antiretroviral treatment adherence and overall enhancement/re-establishment of physical and immunological status in individuals diagnosed with HIV/AIDS.

Ethics statement

Before our obtaining any information or biological samples, patients completed an informed consent to participate, were briefed on the study, and were encouraged to ask questions. The intervention was designed to take place on the same day as the patients’ routine medical visit, to minimize any discomfort. The Institutional Review Board at Ponce School of Medicine and Health Sciences (PSMHS) approved the study protocol.

Participants

A total of 23 participants from the Ryan White HIV/AIDS Program (Part A) were recruited at Special Health Clinic, Juana Díaz, Puerto Rico. This program is administered by the U.S. Department of Health and Human Services and provides funds to those communities that do not have sufficient health care coverage or financial resources for the treatment of HIV disease. In our study, we included males and females who were =21 years of age, whose HIV sero-status was confirmed by Western Blot and who were currently undergoing antiretroviral treatment. We excluded any participant with a history of a significant pre-existing neurological brain disease, including Alzheimer's disease, stroke, seizure disorder, or traumatic brain injury; a chronic psychiatric illness involving psychosis (e.g., schizophrenia), according to DSM-V criteria; end-stage disease (life expectancy<12 months); and evidence of opportunistic CNS infections (toxoplasmosis, progressive multifocal leukoencephalopathy, neoplasm). We also excluded those who had sero-converted to negative on HIV testing and were no longer on antiretroviral drugs.

Behavioral assessment

A professional therapist from the Clinical Psychology Program (PSMHS) interviewed each patient to obtain sociodemographic data (age, gender, town of origin, time since HIV diagnosis, antiretroviral treatment, method of infection, and psychological treatment) and information related to alcohol and/or drug abuse and depressive symptoms. This was accomplished through the administration of the Patient Health Questionnaire-9 (PHQ-9); treatment adherence was determined through the administration of the Immunological Status Report.

PHQ-9

The PHQ-9 is a 9-item questionnaire used for the assessment of depressive symptoms in primary care settings [18]. This questionnaire evaluates the presence of depressive symptoms over the 2 weeks prior to the test’s being filled out. Each of the items can be scored from 0 (not at all), to 3 (nearly every day). The general score can range from 0 to 27 and is interpreted as follow: a score of 0 to 4 means that the subject has minimal or no symptoms, 5 to 9 signifies mild depression, 10 to 14 signifies moderate depression, 15 to 19 signifies moderately severe depression, and 20 to 27 signifies that the subject has severe depression. Its validity and reliability as a diagnostic measure, as well as its utility in assessing depression severity and monitoring treatment response are well-established [18-22].

Treatment adherence

The percentage of treatment adherence was measured by patient self-reporting using the Immunological Status Report. This instrument was developed by the Health Psychology Program’s team. The level of adherence was calculated by dividing the number of doses taken by the number of prescribed doses over the 2 weeks previous to the visit.

Blood collection

Trained personnel from the clinic drew blood samples for the assessment of immunological, blood chemistry, pro-inflammatory and oxidative stress parameters. Samples were collected from overnight-fasted participants by venipuncture and using blood collection tubes containing heparin. Blood samples were centrifuged at 1000 × g for 10 minutes at 4°C. Plasma was isolated and stored at -80°C until assayed.

Viral load and CD4⁺ T-cell counts

Viral load analyses were done by trained personnel from the Immunology Reference Laboratory at PSMHS using the COBAS AmpliPrep/TaqMan HIV-1 v2.0 Test, according to manufacturer’s protocol.

CD4⁺ T-cell counts

Analyses were done by trained personnel from the Immunology Reference Laboratory at PSMHS using the BD Multitest CD3/CD8/CD45/CD4 and analyzed in a BD FACSCalibur™ system according to the manufacturer recommendations.

Plasma level of cytokines

Human cytokines and chemokines were quantified simultaneously using a panel of 39 fluorescent-labeled magnetic beads based technology from MILLIPLEX^® MAP (MILLIPLEX HCYTMAG-60K-PX39, Biosource International Inc., Carlsbad, CA, USA), according to the manufacturer’s instructions. Results were obtained with the MAGPIX^® with xPONENT software program (xPONENT 4.1) and analyzed with the MILLIPLEX Analyst (EMD Millipore, Chicago, IL) using a standard curve derived from recombinant cytokine and chemokine standards.

Superoxide dismutase

Superoxide Dismutase (SOD) activity was measured using the Superoxide Dismutase Assay Kit (Cayman Chemical Company, Ann Arbor, MI), according to the manufacturer recommendations. Briefly, 200 µl of diluted radical detector and 10 µl of standards or samples (diluted 1:5 with sample buffer) were added to their corresponding well. To initiate the reaction, 20 µl of diluted xanthine oxidase was added as quickly as possible to each well and the plate was shaken for a few seconds. At this point the plate was covered and then incubated in an orbital shaker for 20 min at room temperature. Absorbance was read at a wavelength of 450 nm, using a plate reader.

Catalase

Catalase activity was measured using the Catalase Assay Kit (Cayman Chemical Company, Ann Arbor, MI), according to the manufacturer recommendations. Briefly, 100 µl of diluted assay buffer and 30 µl of methanol were added to each well being used. Then, 20 µl of standard, catalase control, or plasma samples (diluted 1:5 with sample buffer) were added to the corresponding well. To initiate the reaction, 20 µl of diluted hydrogen peroxide was added and the plate was covered and incubated in an orbital shaker for 20 min at room temperature. To terminate the reaction, 30 µl of diluted potassium hydroxide and then 30 µl of Catalase Purpald (Chromagen) was added to each well. The plate was covered and incubated for 10 min at room temperature in an orbital shaker. Finally, 10 µl of catalase potassium periodate was added to each well, after which the plate was covered and incubated for 5 minutes at room temperature in an orbital shaker. Absorbance was read at a wavelength of 540 nm, using a plate reader.

Malondialdehyde

Malondialdehyde (MDA) levels for lipid peroxidation were measured using the TBARS Assay Kit (Cayman Chemical Company, Ann Arbor, MI), according to the manufacturer recommendations. Five mL polypropylene screw-cap tubes were used to prepare standards and samples. Briefly, 100 µl of SDS solution and 4 mL of color reagent were added to 100 µl of standard and plasma samples. Samples were boiled for 1 hour, incubated on ice for 10 minutes, and centrifuged for 10 minutes at 1,600 × g at 4°C. One hundred and fifty µl (in duplicate) of each sample was loaded in a 96-well plate. Absorbance was read colorimetrically at a wavelength of 540 nm, using a plate reader.

8-Isoprostane

Free 8-isoprostane fraction was measured using the 8-Isoprostane EIA Kit (Cayman Chemical Company, Ann Arbor, MI). Plasma was collected and stored at -80°C with 0.005% Butylated Hydroxytoluene (BHT) until assayed. Briefly, 100 µl and 50 µl of EIA buffer were added to Non-Specific Binding (NSB) and maximum binding (B₀) wells, respectively. Then, 50 µl of standards (starting with 8) and samples were added to their corresponding wells. Fifty µl of AchE tracer and 50 µl of 8-isoprostane EIA antiserum were added to the corresponding wells and the plate was incubated for 18 hours at 4°C. The second day, wells were washed and rinsed with wash buffer prior to the addition of Ellman’s reagent to each well. Five µl of tracer were added to the total activity well. The plate was covered from the light and incubated for 90-120 min at room temperature. Absorbance was read at a wavelength of 405 nm, using a plate reader until the B₀ wells reached 0.3 absorbance units (A.U.).

Statistical analyses

Data were analyzed by means of descriptive statistics (including percentages, means, and standard deviations–SD). The data were expressed as mean ± SD. Difference between groups was assessed using a 2-sided Wilcoxon–Mann–Whitney test and correlation analyses with a 2-sided Spearman test. Results with a p = 0.05 were considered statistically significant.

Clinical and demographic parameters for study participants

The information depicted in Table 1 presents the demographic and clinical characteristics for the subjects in the HIV-positive cohort. The mean age of the 23 subjects included in these analyses was 53.1 years old (range, 40-69) for subjects with no depression symptoms and 48.5 years old (range, 41-61) years for subjects with mild-moderate to severe depression symptoms. All of the subjects lived in Puerto Rico, and 70% of them were male. Based on the assessment of depression symptoms using the PHQ-9, 10 subjects showed no-depressive symptoms, while 13 exhibited mild-moderate to severe depression symptoms. On average, subjects had a self-reported combination antiretroviral treatment (CART) adherence of 90% (SD = 13%).

Table 1: Demographic and Clinical Characteristics of HIV Positive Subjects in the Study.

However, subjects experiencing depressive symptoms reported a significantly decreased adherence to the antiretroviral treatment. There were no statistical differences in mean CD4+ T-cell counts and HIV viral load between the two groups. A total of 35% of the subjects had an undetectable HIV viral load. More than 50% of the subjects in both groups were infected through heterosexual contact (Table 1). A total of 14 subjects were also con-infected with Hepatitis C Virus (HCV). Although there was no statistical difference in the distribution of HCV co-infection or viral load between HIV⁺ non-depressed and depressed subjects, we observed higher levels of alaline transaminase (ALT) in HIV/HCV co-infected subjects with depression symptoms (Table 1).

Plasma pro-inflammatory cytokine levels correlate with depression symptoms in HIV-positive subjects

Depression affects not only the brain and behavior but also the rest of the body. However, the underlying mechanisms by which depression influences HIV/AIDS disease progression remain to be fully understood. Evidence suggests that a chronic inflammatory state plays a key role in regulating HIV infection and likely depression [23,24].

We assessed the levels of pro-inflammatory cytokines in plasma samples of HIV-infected individuals who were not experiencing depression symptoms (n=10) and compared them to the levels in HIV-infected individuals who were experiencing depression symptoms (n=13). The plasma pro-inflammatory cytokine levels of IL-15 (Figure 1A), IP-10 (Figure 1C), and IL-12 p40/p70 (Figure 1E) were significantly elevated in subjects with mild-moderate to severe depression symptoms compared with those of subjects having no depression symptoms (p = 0.044 and p = 0.025, respectively), indicating an enhanced inflammatory state in this group of participants. Moreover, the plasma levels of IL-15 (Figure 1B), IP-10 (Figure 1D), and IL-12 p40/p70 (Figure 1F), significantly correlated with the level of depression symptoms (p = 0.011; p = 0.050; p = 0.036, respectively), which correlation is indicative of a relationship between mental health status and the inflammatory response in HIV infection. We also analyzed the plasma levels of G-CSF (Figure 1G) and observed significantly elevated levels of this cytokine in subjects with depression symptoms compared to the levels observed in their non-depressed counterparts (p = 0.011). The levels of G-CSF also correlated with the level of depression symptoms (Figure 1H; p < 0.0001). Other cytokines (IL-1β, IL-1α, IFN-γ, MCP-1, MIP-1α, MIP-1β, among others) assessed within the MILLIPLEX HCYTMAG-60K-PX39 panel were detected at higher levels in individuals who were experiencing depression symptoms however, no significant differences were observed between the two groups (data not shown).

Figure 1: Correlations between blood cell number and depression symptoms. Two-sided Spearman correlation tests between the white blood cell numbers and levels of depression symptoms. Panels: A= total White Blood Cells (WBC); B= Lymphocytes (LYM); C= Monocytes (MO); D= Neutrophils (NE); E= Eosinophils (EO); F= Basophils (BA). CB: Confidence Band. The cut-off for significance was p = 0.05.

Table 2 shows the clinical information related to the blood parameters of the HIV-infected subjects enrolled in this study. According to the analyses, the depressed subjects have significantly decreased white blood cells (WBCs; p = 0.023) and platelets (p = 0.048) as compared to their non-depressed counterparts, and there was a slight inverse correlation of the former with depressive symptoms (rs = -0.382, p = 0.044). Within the WBCs, the neutrophils and the basophils were the subpopulations that were significantly decreased in depressed subjects. However, only the total WBCs (rs = -0.382, p = 0.044) and neutrophils (rs = -0.423, p = 0.0281) inversely correlated with depression symptoms (Figure 2, panels A and D). Other blood parameters, including RBCs, Hb, and HCT were not significant when comparing the two groups (Figures 2E and 2F).

Table 2: Blood Parameters of HIV Positive Subjects in the Study.

Figure 2: HIV-depressed patients have higher plasma IL-15, IP-10, IL-12 p40/p70 levels. Panels A, C, E, and G: Cytokines (panel A: IL-15; panel C, IP10; panel E: IL-12 p40/p70; panel F: G-CSF) were measured simultaneously using the Luminex’s MAGPIX (with xPONENT 4.1 program) and a cytokine magnetic bead panel (MILLIPLEX HCYTMAG-60K-PX39). Significance was calculated using Wilcoxon-Mann-Whitney test. Solid black lines represent the median of each group and dashed boxes denote the first to third quartiles. Observed values greater than 1.5*IQR from the quartiles are highlighted as possible near outliers (+). Panels B, D, F and H: 2-sided Spearman correlation tests between the levels of IL-15 (panel B), IP-10 (panel D), IL-12 p40/p70 (panel F), or G-CSF (panel H) and the levels of depression symptoms. CB: Confidence Band. The cut-off for significance was p = 0.05.

HIV depressed and non-depressed patients have significant differences in lipid peroxidation

Overload of oxidative stress is commonly found in the development and progression of chronic diseases. It has been shown that lipids are the most susceptible biomolecules affected by oxidative stress [25]. Thus, to analyze the redox status between HIV-infected subjects who have/do not have depression symptoms we performed analysis for plasma levels of lipid peroxidation and antioxidant activity. We measured lipid peroxidation with two different assays.

The thiobarbituric acid reactive substances (TBARS) assay measures malondialdehyde (MDA), a naturally occurring decomposition product of lipid peroxidation. Alternatively, the 8-isoprostane enzyme immuno assay identifies a stable end-product of lipid peroxidation. When we analyzed oxidative stress parameters (Table 3), we did not see any significant differences in catalase and superoxide dismutase activities between the two groups. However, we observed significantly higher plasma levels of lipid peroxidation (malondialdehyde, p = 0.033; 8-isoprostane, p = 0.036) in HIV non-depressed subjects than in the HIV depressed subjects. The levels of 8-isoprostane in plasma inversely correlated with the levels of depression symptoms (rs = -0.560, p = 0.008), whereas there was a trend of inverse correlation between MDA and depression symptoms (rs = -0.302, p = 0.080) (Figure 3).

Table 3: Oxidative stress parameters in depressed versus nondepressed
HIV positive subjects.

Figure 3: Lipid peroxidation levels inversely correlated with depression symptoms. Two-sided Spearman correlation tests between the levels of catalase (panel A), superoxide dismutase (panel B), malondealdehyde (panel C), or 8-isoprostane (panel D) and the levels of depression symptoms. CB: Confidence Band. The cut-off for significance was p = 0.05.

Depression is the most common psychiatric diagnosis in people living with HIV/AIDS [9] and, according to a 2008 literature review by Leserman, almost 50% of HIV/AIDS patients suffering from depression are never diagnosed or treated for that co-morbid condition [26]. Poor management of depression is associated with reduced antiretroviral treatment adherence, which can lead to disease progression [27]. Recent studies have shown a correlation between pro-inflammatory cytokines, free radical species, and oxidants in the etiology of depression [13,14]. In addition, oxidative stress products can contribute to chronic inflammation and the development of pathologies [28]. The current study addressed the correlation of depressive symptoms on the inflammatory state and the oxidant/antioxidant balance in an HIV cohort.

We analyzed the plasma levels of the pro-inflammatory cytokines IL-15, IP-10, IL-12 p40/p70 and G-CSF. Our results demonstrate that participants with depression symptoms exhibit significantly higher levels of the pro-inflammatory cytokines IL-15, IP-10, IL-12 p40/p70 and G-CSF than do those participants without depression symptoms and the levels of these cytokines correlated with the levels of depression symptoms. Our results are in agreement with a previous study that demonstrated significantly higher IP-10 levels in depressed patients when compared to controls [29]. The group also demonstrated that the levels of IP-10 decreased after antidepressant treatment [29]. In addition, there is significant evidence demonstrating that elevated peripheral cytokines (IL-1, IL-6, TNF-α and IFNγ) and other pro-inflammatory mediators (C-reactive protein) influence the development and progression of depression [30-34]. As well, gene expression profiling on post-mortem brain tissue from psychotropic drug-free individuals with a history of major depression revealed up-regulation of a variety of pro- and anti-inflammatory cytokines [35]. Additionally, studies on HIV patients co-infected with HCV revealed higher plasma levels of IL-15 compared to the levels in patients mono-infected with either HIV or HCV [36,37]. However, to our knowledge, this is the first report that correlates higher plasma levels of IL-15 to depression symptoms. IL-12 is a heterodimeric cytokine consisting of the p35 and p40 chains located in separate chromosomes [38]. The bioactive heterodimer (p70) plays an important role in the activities of natural killer cells (NK) and T lymphocytes. IL-12 p40 has been estimated to be secreted in excess as compared to IL-12 p70 by peripheral blood cells [39]. Our results showed that HIV subjects with depression symptoms had IL-12 p40 levels approximately 5 times higher than HIV subjects without depression, whereas the levels of IL-12 p70 were comparable between the two groups. This led to a significantly increased ratio of IL-12 p40/p70 in subjects with depression symptoms as compared to subjects who do not have depression symptoms. Previous studies have shown that overexpression of IL-12 p40 subunits forms more stable homodimers that bind to IL-12 β1 receptors, antagonizing IL-12 p70 subunit [40]. Thus, the increased IL-12 p40/p70 ratio could potentially lead to diminished NK and T cell functions in HIV subject with depression symptoms. We did not find previous studies about the specific role of IL-12 p40/p70 in depression, but interestingly, IL-12 is important for the establishment of a TH1 response and depression is characterized by shift from T naïve to TH1 responses [41]. Although the sample size was small, the levels of the pro-inflammatory cytokines reported in our study correlated with the levels of depression symptoms as we did not observe any significant differences in the percentage of HCV co-infection between the groups.

The disruption of hematological parameters is among the most common complications of HIV disease [42]. Several studies have shown the presence of anemia, leukopenia, and thrombocytopenia within the HIV-infected individuals, and these blood disorders are associated with rapid disease progression and increased mortality [43-48]. Though the participants in both groups were within the normal range of WBC (4-11 cells/µl) and PLT (150-400 cells/µl), we observed that depressed subjects had significantly decreased WBCs and PLTs compared to their non-depressed counterparts. Riviere and colleagues showed that precursors of platelets, also known as megakaryocytes, have high CXCR4 expression [49]. CXCR4 is one of the co-receptors for HIV-1 infection, suggesting that HIV-1 infects megakaryocytes directly and inhibits them to produce platelets [49]. Another study demonstrated that HIV-1 infection can also impair megakaryocyte maturation by reducing the formation of megakaryocytopoietic colony-forming units in infected subjects [50]. Also, a group from the Strategies for Management of Antiretroviral Therapy (SMART) study found that platelet counts increased in most of the HIV subjects after their re-introduction to treatment [51]. Although it may be likely that a combination of the mechanisms described above underlie the significantly lower platelet counts in HIV⁺ depressed subjects in our study, we were not able to establish a correlation between treatment adherence and platelet counts. Self-reported treatment adherence in subjects experiencing depression symptoms ranged from 40% to 100%, and the subjects with lower treatment adherence did not necessarily have the lowest platelet count or the highest HIV viral load. Also, IFN-α based therapy of HCV infection has been reported to negatively affect peripheral platelet counts and induce depression in treated patients [52,53]. However, HIV/HCV co-infected subjects in our study were not under IFN-α based therapy at the time of sample collection. Thus it seems unlikely that the observed altered levels of IP-10, MDA, 8-isoprostane and platelet count be due to IFN-α based therapy of HCV infection.

The activation of the immune system and inflammatory responses occur as a result of the presence of stimuli, and infection with HIV is not an exception. HIV infection is characterized by a persistent chronic inflammatory state [54]. Moreover, it has been shown that during early HIV infection, patients are in oxidative imbalance, characterized by high levels of malondialdehyde (MDA), a product of lipid peroxidation, and reduced antioxidant capacity in serum [55]. We assessed the enzymatic activity and levels of oxidative stress markers in plasma samples from our HIV-infected cohort. We observed a trend of higher catalase and lower SOD levels in HIV subjects with depression symptoms. Catalase is induced by higher hydrogen peroxide (H₂O₂) levels and high H₂O₂ levels may in turn lead to increased peroxidation of lipids. The plasma levels of MDA and 8-isoprostane were significantly lower in the group of depressed subjects. The lower levels of lipid peroxidation may be partially explained by the higher levels of catalase enzyme activity in depressed subjects. In the presence of catalase, an in vitro study estimated that 18% of MDA disappeared [56]. Several other factors could also account for the discrepancy, ranging from differences in gender to differences in environment/pathogens or differences in the phases of neuroendocrine, counter-regulatory systems or severity and stage of HIV disease. In addition, since the non-depressed subjects had significantly higher levels of cART adherence, the antiretroviral therapy itself may be leading to the observed increased in lipid peroxidation. It is known that the antiretroviral treatment increases chemically reactive species in circulation, possibly by producing more oxidized metabolites deriving from the interaction between ROS and infected-cell biomolecules [57-59].

This study does have some limitations. The demographic information and results shown here are from a cross-sectional design. Thus, we need to implement strategies to continue monitoring the mental and physical health statuses of the HIV-infected subjects. In addition, the adherence to antiretroviral medications was self-reported. The actual percentage of adherence could be lower than that which was reported by some of the patients. Regardless, HIV infected subjects experiencing depressive symptoms have worse inflammatory markers demonstrating that it is important to deal with depression because of its physiological effects. Further studies with a larger sample size are needed to confirm whether stabilization of the depression symptoms is accompanied by a reduction in pro-inflammatory cytokines to support the role of cytokines in the pathophysiology of depression.

The authors would like to thank the Clinical Psychology Program, the AIDS Research Infrastructure Program at PSMHS, RCMI Publications Office, and Dr. Rafael Iván Iriarte from the Public Health Program at PSMHS. We also want to acknowledge the efforts of the personnel from Special Health Clinic, Juana Díaz, Puerto Rico, who made a valuable contribution to our study. The project was supported by the following grants from the National Institutes of Health: NCRR-RCMI (RR003050) / NIMHD (G12-MD007579), NIGMS (R25GM084206), NCRR-INBRE (P20-RR016470), and MD007587. The work was also supported by institutional funds from PSMHS.