| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the HIV Center for Clinical and Behavioral Studies at the New York State Psychiatric Institute and Columbia University, New York City, New York.
Address correspondence and reprint requests to Mark V. Bradley, HIV Center for Clinical and Behavioral Studies, Unit #15, New York State Psychiatric Institute and Columbia University, Unit 15, 1051 Riverside Drive, New York NY 10032. E-mail: mb2032{at}columbia.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: HIV-positive and HIV-negative members of 197 serodiscordant couples (159 male/female, 38 male/male) were assessed using instruments measuring depressive symptoms, sexual risk, and couple satisfaction.
Results: HIV-positive partners with higher depression scores were less likely to be part of couples reporting unprotected sex, and HIV-positive partners' higher depression scores were associated with less unprotected intradyadic sex acts. This decrease in intradyadic sexual risk behavior was partially explained by a decrease in any sexual behavior within the couple. On the other hand, HIV-positive subjects with moderate or higher depression were more likely to have outside partners. Adding the partner satisfaction measure to the models was able to account for the relationship between the HIV-positive subjects' depression scores and outside partners, but not for that between higher depression score and reduced intradyadic sexual risk.
Conclusions: HIV-positive individuals with more depressive symptoms may be less likely to engage in high-risk sexual behavior with their partners than those with less depressive symptoms, but more likely to have sexual partners outside the relationship. These findings suggest that the relationship between depressive symptoms and sexual risk behavior in this population may be mixed and complex, and suggest that clinicians should assess sexual risk behavior across the range of depression symptom severity.
Key Words: depression • HIV • serodiscordant couples • risk behavior • partner satisfaction
Abbreviations: HIV = human immunodeficiency virus; BDI—II = Beck Depression Inventory, Second Edition; BASR = Brief Assessment of Sexual Risk; DAS = Dyadic Adjustment Scale; AIDS = acquired immunodeficiency syndrome.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
As a specific class of psychiatric symptoms, depression demonstrates a complex relationship to sexual risk behavior. Although depressive disorders have been associated with decreased sexual interest and activity (3–5), research regarding the relationship between sexual HIV risk behavior and depression has thus far been mixed. A number of studies have found evidence of increased sexual risk behavior in states of depression (6–14), but other studies have failed to do so (15–19). The reason for these inconsistent findings may relate to the highly disparate populations from which these studies have drawn, including urban men who have sex with men (7,10,12,18), adolescents (11), substance users (8,13,14), sex workers (6), and homeless mentally ill persons (17). Thus, differences in both psychopathology and in social contexts may determine whether individuals respond to depression with either increased or decreased sexual risk behavior. In addition, other research has suggested that individuals vary in the direction of their sexual response to depression. Bancroft et al., studying populations of heterosexual and homosexual men, have found that whereas the most common libinal response to depression is a decrease in sexual interest, a minority reports increases in libido while depressed (20,21). These authors have posited a "dual control" hypothesis for understanding depression's effect on sexual risk behavior, suggesting that sexual interest and activity are shaped at the individual level by varying degrees of sexual excitation and inhibition (22). According to this model, male sexual response is determined by the net result of centrally mediated sexually excitatory and inhibitory tone, with individual variability in degrees of excitation- and inhibition-proneness. This model suggests that although most men will respond to negative mood symptoms with pronounced sexual inhibition, men with high levels of excitatory tone will respond to increased depressive symptoms with heightened sexual responsiveness. Results from a questionnaire developed to measure these propensities suggest two categories of inhibitory tone—the first occurring in response to fears of performance failure, and the second in response to external threats, including conflict within sexual relationships. However, in addition to considering the role of individual differences, the variable findings in all research to date examining depression and sexual risk demonstrate that the relationship between the two must also be understood within the specific contexts in which they occur. Furthermore, research into these contexts must investigate the factors that promote high-risk behaviors among HIV-positive individuals to lay the groundwork for prevention strategies within this population.
An important category of such contexts is the specific kinds of sexual relationships within which sexual risk might occur. For example, research has consistently shown that most HIV risk behavior occurs in the setting of primary romantic and sexual relationships (23–28). Intradyadic risk behaviors are of particular concern when the partners are of mixed HIV status, representing an inherent risk of HIV transmission. Furthermore, these couples may experience a number of serodiscordance-related stressors that place their constituent members at increased risk of developing a depressive disorder. Previous research has found levels of distress among male serodiscordant couples to be in the mild-moderate range compared with population norms (29). Because of the inherent potential risk of transmission present in the unprotected sexual activity of serodiscordant couples, it is important to understand the factors, including psychiatric symptoms, leading to unsafe behavior within these dyads to develop population-specific prevention strategies. The role of depression in sexual risk behaviors of serodiscordant couples has not been systematically examined to date. Furthermore, to elucidate the impact of depression and other psychiatric symptoms on HIV risk behavior within serodiscordant couples, their effect on these couples' dynamics (in addition to that on their individual members), and the consequent effects on sexual behavior need to be better understood. Depression may exert its specific effects on the sexual behavior of serodiscordant couples by disrupting partner satisfaction and dyadic adjustment. In this study, we hypothesized that greater severity of depressive symptoms would be associated with increased sexual HIV risk behavior among serodiscordant couples. Furthermore, we hypothesized that the relationship of depression and sexual risk behavior would be explained by partner satisfaction within couples.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
Procedures
This is an analysis of baseline data from participants in a behavioral intervention study that addressed improving HIV medication adherence among serodiscordant couples (30). Study data were collected between August 2000 and January 2004 at two HIV/acquired immunodeficiency syndrome (AIDS) outpatient treatment clinics in New York City. Participants were recruited using flyers posted at outpatient HIV medical clinics, private medical practices, and HIV/AIDS service organizations. The eligibility criteria required participants to be HIV-positive adult patients in primary care who demonstrated poor adherence (defined as taking <80% of prescribed doses at regular intervals over a 2-week assessment period) and who reported being in an HIV serodiscordant relationship for 
6 months. Severe cognitive dysfunction or psychotic disorders constituted exclusion criteria.
All eligible participants completed a baseline Audio Computer-Assisted Self-Interview (ACASI) to assess factors that might be associated with medication adherence. The assessment took approximately 2 hours to complete. The participants were paid $20 for the screening appointment and $25 for the baseline assessment. The study protocol was reviewed and approved by the Institutional Review Boards of the New York State Psychiatric Institute and St. Luke's Roosevelt Hospital Center. All participants provided their written consent.
Measures
Depressive Symptoms
The presence of depressive symptoms was assessed in the seropositive and the seronegative partner using the Beck Depression Inventory, Second Edition (BDI—II). The BDI—II consists of 21 items, each of which lists four statements arranged in increasing severity about a particular symptom of depression. The instrument was revised in 1996 to include items regarding sleep and appetite in order to more closely approximate the major depressive episode category from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. The instrument's scoring manual suggests that scores of 0 to 13 represent minimal depression; scores of 14 to 19 correspond to mild depression; scores of 20 to 28 represent moderate depression; and scores of 29 correspond to severe symptoms. The 
for this sample was 0.92.
Sexual Risk Behavior
The Brief Assessment of Sexual Risk (BASR) questionnaire (C. Delezal unpublished work 2000) was used to measure high-risk sexual behavior as it occurred within the primary relationship and to determine whether the HIV-positive partner had sexual activity outside of the primary relationship. Specifically, the BASR was used to ascertain three indicators of sexual risk, occurring over the prior 2 months, which would be used in our analyses. First, we measured intradyadic sexual risk behavior as a dichotomous variable based on whether either partner reported unprotected anal or vaginal sex with the other partner. Second, we measured intradyadic sexual risk behavior as a continuous variable, derived from the mean of the number of unprotected anal or vaginal sex acts reported by both partners in each couple. This variable was rank-transformed before analysis due to its skewed distribution. Finally, we recorded whether the HIV-positive member of each couple reported any sexual partners outside of his or her primary relationship.
Partner Satisfaction. The Dyadic Adjustment Scale (DAS) is a 32-item self-report scale with four subscales (dyadic consensus, affectional expression, dyadic satisfaction, dyadic cohesion) that assess "perception of the adjustment of the relationship as a functioning group." Spanier reported high internal consistency for the DAS total scale as well as for its component subscales (31). As all statistically significant associations in our analysis of depression scores and sexual risk behaviors were found among our HIV-positive subsample, we calculated values for the DAS from this subsample. These were acceptably high: dyadic consensus = 0.85; affectional expression = 0.66; dyadic satisfaction = 0.83; and dyadic cohesion = 0.73.
Data Analyses
Regression analyses were used to examine the association between sexual risk behaviors and depression. Logistic regression models were used to examine two dichotomous outcomes: a) whether or not the couple reported unprotected vaginal or anal sex, and b) whether or not the HIV-positive partner had sexual partners outside of the relationship. Linear regression was used to examine one continuous outcome—the number of unprotected vaginal or anal sex acts occurring within the relationship. In separate regressions, depression symptom score was entered into each of these models in two different forms: a) as a continuous BDI—II score, and b) as a dichotomous variable with a BDI—II cut-off of 20, representing a threshold for moderate clinical depression. We examined depression symptoms as both a continuous as well as a dichotomous variable because a) on the one hand, we wanted to test the effect of depression symptoms across the full range of severity (from mild subclinical symptoms to severe symptoms); b) on the other hand, we wanted to test to see whether depression as a clinical phenomenon (i.e., meeting an approximate severity cut-off for clinical depression) had an effect on sexual risk behavior. Regression analysis was then used to test whether partner satisfaction as measured by the DAS explained the relationship between depression and sexual risk behavior. For these analyses, DAS scores were entered into our regression models as independent variables along with the depression variable to determine whether any associations found remained significant. In post hoc analyses, we similarly entered the libido item from the BDI—II into our models along with depression to test whether libido could explain any relationships between depression symptoms and our sexual behavior outcome variables. In these analyses, BDI—II scores were recalculated to exclude the individual libido item, with the threshold for the dichotomous depression variable reduced from a score of 20 to 19.
To examine the whether or not current treatment with antidepressants was associated with sexual risk behavior (as antidepressant medications may either reduce sexual drive through their direct physiologic effects or may improve sexual functioning by reducing depressive symptoms), in further post hoc analyses we tested to see whether current antidepressant use by the HIV-positive partners was associated with our primary sexual risk outcome variables.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
HIV-Positive Participants Reporting Outside Partners
Of the HIV-positive members of the couples, 18% reported having had sexual partners outside the relationship during the past 2 months. The HIV-positive partners in male-male dyads (34.2%) were significantly more likely than those in male-female dyads to have outside sexual partners (p = .003).
Depression Symptoms
BDI—II scores for both the HIV-positive and HIV-negative samples were within the minimally elevated range with mean ± SD values of 11.14 ± 9.05 and 9.83 ± 10.33, respectively. Of the HIV-positive sample, 19% met the BDI moderate-severe cut-off of 20, compared with 15% of the negative partners. There were no significant differences between homosexual or heterosexual couples with respect to whether the HIV-positive or HIV-negative member met the moderate-severe depression cut-off score. Of the HIV-positive sample, 22.3% reported current treatment with antidepressant medication. Antidepressant treatment was not associated with BDI—II score among the HIV-positive subsample. However, a BDI—II score of 19 was significantly associated with decreased libido in both the HIV-positive and HIV-negative subsamples.
Depression Symptoms and Sexual Risk Behavior (Table 2)
|
For both of these regressions demonstrating significant associations, the total frequency of anal or vaginal sex (regardless of condom use) was added to the models to see if the associations remained significant after adjusting for overall frequency of sex. In both of these analyses, the relationships between the depression symptom score variable and intradyadic sexual risk variable were no longer significant after adding total frequency of sex to each model. When rerunning these analyses separately for the male-male and male-female couples, the only change resulting from separating couple types in this manner was that a higher continuous BDI—II score for HIV-negative members was also associated with decreased intradyadic sex among male-female couples.
Depression and HIV-Positive Sexual Behavior Outside the Relationship
HIV-positive participants with a BDI—II score of >20 were more likely to report sexual partners outside their primary relationship than those with a lower score. When the male-male and male-female couples were analyzed separately, we found that a higher BDI—II continuous score was associated with more outside partners for the HIV-positive member of the male-male couples, whereas no such relationship existed among the male-female couples.
Partner Satisfaction as Potentially Explaining the Relationship Between Depressive Symptoms and Lower Risk Behavior
To test whether partner satisfaction as measured by the DAS explained the relationship between higher depression scores of HIV-positive partners and the reduced sexual risk behavior between partners, we entered all four of the subscales of the DAS from the HIV-positive partners into the three logistic and linear regression models above which demonstrated statistically significant relationships. The relationship between BDI—II score and intradyadic risk behavior for HIV-positive participants remained statistically significant after the DAS subscales were entered into the model (from p = .007 to .017), with a change in coefficient value from –1.102 to –1.044, suggesting that partner satisfaction could not explain this relationship. Similarly, the p value in the logistic regression model describing the relationship between BDI—II score and presence or absence of unprotected intradyadic sex demonstrated minimal change with the addition of DAS subscales (from p = .023 to .021), with a small change in OR (0.960 to 0.954). In contrast, in the logistic regression model describing the relationship between BDI—II cut-off of 20 and the presence of outside sexual contacts by the positive partner, the p value became insignificant (from p = .046 to p = .137) after adding the DAS subscales to the model, with a 17% decrease in odds ratio from 2.327 to 1.932.
Effect of Libido on Depression-Risk Relationships
When libido scores were added to each regression model, using the recalculated BDI scores (with the individual libido item removed), the associations between BDI and sexual risk were attenuated. All three significant associations became marginally significant (p < .10) after adjusting for libido. Although there was no significant association found between HIV-positive participants' libido scores and whether they had outside sexual relationships, the trend was toward those with outside partners having lower libido scores (p = .10).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Prior research has not demonstrated consistent findings regarding the relationship between depression and HIV risk. Our study suggests that, among serodiscordant couples, depression in the seropositive partner has a complex effect on risk, either increasing HIV risk (in the form of seropositive members' outside partners) or decreasing HIV risk (in the form of decreased unprotected sex within couples). Our findings contradict simplistic notions that psychiatric symptoms are universally and unequivocally related to heightened sexual HIV risk behaviors.
These results may be interpreted as reflecting the impact of depressive symptoms on the functioning of serodiscordant couples and their respective members. For example, anhedonia—including decreased sexual interest—is a common symptom of depression. Decreased sexual interest may lead to reduced sexual behavior—unprotected and otherwise—within a couple. Our study found that decreased overall sexual behavior accounted for the decreased intradyadic risk behavior associated with the HIV-positive partners' higher depression scores. Framing the impact of depressive symptoms on unprotected intradyadic activity as an individual level, symptomatic phenomenon seems plausible in our study, as measures of partner satisfaction do not contribute to this relationship. A mediating effect of partner satisfaction would have suggested that the primary impact of depressive symptoms on intradyadic sexual risk occurs as a couple-level phenomenon; however, this was not found in our study. In contrast, that partner satisfaction explained the relationship between depressive symptoms and outside sexual contacts for seropositive participants suggests that couple relational issues are important determinants of this component of risk behavior within our sample.
An alternative explanation of our findings may be formulated as the impact of sexual behavior on mood status, especially with respect to seropositive members of couples. Couples who refrain from sexual activity, including unprotected sex, may experience less intimacy and pleasure, contributing to depressive symptoms. This may be especially true for HIV-positive individuals if they perceive their serostatus as obstructing sexual intimacy with their partners. However, in light of the fact that partner satisfaction did not explain this relationship, it is unlikely that intimacy problems alone could have accounted for higher depressive symptoms.
Our study is limited on several counts. First, our cross-sectional design precludes conclusions regarding the directionality of our findings. Second, although the BDI—II is a reliable and valid instrument for measuring depressive symptoms, it can only approximate clinical diagnoses of depression by use of cut-off scores. Finally, our measures relied on participants' self-report.
In spite of these limitations, our findings have implications for the clinical care of HIV-positive patients who are in serodiscordant couples. Clinicians may assume that patients with a lower burden of psychiatric symptoms will demonstrate better health behaviors, including sexual risk reduction, than patients who are more severely symptomatic. However, our findings that HIV-positive individuals with lower depression scores are less likely to have extradyadic sexual partners and are also more likely to engage in risky sex with their seronegative relationship partners present a mixed and complex picture regarding depressive symptoms and sexual risk. These findings suggest clinicians must not assume that mild or absent psychiatric symptoms place their HIV-positive patients at a lower likelihood of engaging in sexual risk behavior with their partners. Furthermore, they underscore that risk behavior among HIV-positive patients should be carefully assessed even in the absence of active psychopathology.
Our results suggest that future research would benefit from longitudinal designs to examine the relationships between depression and risk behavior among serodiscordant couples to establish their directionality. In addition, future prevention strategies will require a better understanding of how psychiatric symptoms—or their absence—may increase or reduce the incidence of sexual risk behaviors within these dyads.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This study was funded by Grant R01 MH61173 from The National Institute of Mental Health (R.H.R.) and training grant T32 MH19139 from NIMH (M.V.B.). The HIV Center for Clinical and Behavioral Studies at the New York State Psychiatric Institute and Columbia University is supported by Grant P30-MH43520 from NIMH.
DOI:10.1097/PSY.0b013e3181642a1c
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
