Skip Navigation LinksHome > January 1, 2005 - Volume 38 - Issue 1 > The Role of Depressive Symptoms in Predicting Sex With Multi...
JAIDS Journal of Acquired Immune Deficiency Syndromes:
Epidemiology and Social Science

The Role of Depressive Symptoms in Predicting Sex With Multiple and High-Risk Partners

Williams, Chyvette T.; Latkin, Carl A. PhD

Free Access
Article Outline
Collapse Box

Author Information

From the Johns Hopkins Bloomberg School of Public Health, Baltimore, MD.

Received for publication June 17, 2003; accepted April 5, 2004.

Supported by the National Institute on Drug Abuse (DA 13142).

Reprints: Carl A. Latkin, Johns Hopkins Bloomberg School of Public Health, 624 N. Broadway, Room 737, Baltimore, MD 21205 (e-mail:

Collapse Box


Existing research examining the role of psychopathology on HIV risk behaviors has relied heavily upon cross-sectional data, making it difficult to assess causal relationships. This paper uses longitudinal data to examine the relationship between depressive symptoms and sexual risk behaviors in a community sample of 332 inner-city drug users. Results from repeated measures analysis using generalized estimating equations suggest that high depressive symptoms may be causally related to having sex with multiple partners (odds ratio = 1.55, CI: 1.12, 2.14) and having sex with injection drug users (odds ratio = 1.57, CI: 1.11, 2.22) and crack users (odds ratio = 1.37, CI: 1.02, 1.86). Findings support greater attention to depressive symptoms in HIV prevention interventions for illicit drug users to reduce sexual risk behaviors.

Preventing further spread of HIV continues to depend on the adoption of preventive behaviors for those with and those at risk for the disease. In the United States, drug and sexual risk behaviors represent the primary means of acquiring and transmitting HIV; therefore, identifying malleable determinants of risk behavior remains an important task. For researchers studying drug-using populations with high rates of psychiatric co-morbidity, attention to psychopathology as a determinant of HIV risk behavior is critical.

Research has shown that psychiatric disorders, primarily depression, antisocial personality, and anxiety, are more prevalent among drug users than in the general population.1-4 The manifestations of psychiatric disorders can have significant implications for the practice of HIV preventive behaviors among drug users. For example, several studies have shown a significant relationship between depression and injection behaviors.5-7 Although sharing contaminated drug equipment is the primary mode of HIV transmission among injection drug users (IDUs), studies suggest that sexual transmission is also high in drug-using groups.8,9 Reducing risky sexual behaviors of drug users has proven to be more challenging than adopting safer needle practices.9,10 Sex with IDUs has become a significant mode of HIV transmission, for minorities and women in particular.11 Additionally, high-risk sex has also been shown among users of cocaine and crack.12-14 As such, inquiries into the role of psychopathologies in predicting engagement in risky sexual behaviors for drug users are of great consequence.

Research examining the effects of psychopathologies on sexual risk behaviors has been mixed. In a study of opioid users entering treatment, psychologic dysfunction significantly increased the odds of sex with multiple partners, unprotected sex with an IDU, and exchanging sex for money or drugs.13 A meta-analytic review examining the association of negative affect and HIV sexual risk behavior identified 9 of 34 studies that used samples of substance abusers. Across these 9 samples, they found modest effect sizes (only 1 with effect size ≥0.10) and a great deal of heterogeneity (95% CI: 2 0.36, 0.38). The authors concluded that there was insufficient evidence to support an association. Results from this review suggest that associations observed in past studies may be the result of methodologic artifacts, including cross-sectional research design.15,16 Data collected over time provide the best means of determining the degree and direction of associations between depression and risk behaviors.

The present study explores the role of depression in predicting sex risks related to HIV transmission in an inner-city community sample of drug users. Additionally, the present study takes a step toward addressing limitations in past studies by assessing the relationship between depressive symptoms and sexual risk behaviors longitudinally.

We hypothesize that depression will increase the likelihood of engaging in sexual risk behaviors. Additionally, we hypothesize that having a main sexual partner will be protective of sex with multiple partners. Finally, we hypothesize that HIV-positive participants will be less likely to engage in risky sexual behaviors than HIV-negative participants. Several studies have found that risky behaviors decrease among individuals who are HIV infected.10,17

Back to Top | Article Outline


Sample and Design

Targeted outreach and word-of-mouth were used to recruit participants for the SHIELD Study, an experimental network-oriented HIV prevention intervention targeting risky behaviors among drug users in Baltimore, MD. A more detailed description of the study is documented elsewhere.18 Study eligibility included being 18 years of age or older, having regular contact with members of the drug-using community, and willingness to educate drug users about HIV. Eligible members provided informed consent before beginning the study and were administered baseline and follow-up surveys by a trained interviewer. Three waves of data were used for this study (baseline and 2 follow-ups) and were collected between 1997 and 2002. The first follow-up occurred between January 1998 and February 2001; 77% of first follow-ups were conducted within 1 year of baseline. The second follow-up occurred between April 2000 and June 2002. The final number of participants at baseline was 1637. Of these, 1110 were given priority status for a second interview; and 896 (81%) were interviewed. Priority status for receiving a second interview was given to those who attended majority of the group sessions or brought in a network member. By wave 3, 708 (79% of 896) had completed surveys for all 3 waves. Surveys covered a range of topics including sociodemographic characteristics, physical, and mental health measures, drug and sexual behaviors, and self-reported HIV status. Data from participants who completed the first 3 waves and who reported being sexually active in the past 90 days for all 3 waves are included in the present study (n = 332).

Back to Top | Article Outline
Depressive Symptoms

For these analyses, we were interested in a longitudinal of depressive symptoms as a predictor of sexual risk behaviors. The Center for Epidemiologic Study-Depression scale (CES-D) was used to measure depressive symptoms in the study sample.19 The CES-D is a normed and validated 20-item, 4-point scale developed for use in the general population; however, it has been used successfully in previous studies with drug users.20,21 For example, Grella et al21 found that higher scores on the CES-D were related to HIV risk behaviors among female IDUs. Scores on the CES-D can range from 0-60, with higher scores indicating greater probability of depression. A CES-D score was calculated for each respondent by summing all items in the scale, after reverse-coding positive items. Cronbach a was ≥0.90 for the 3 waves of CES-D score. Then, we dichotomized CES-D score using the widely accepted value of 16 as the cut-off score for probable clinical depression.22 We chose to dichotomize the scale because we thought the interpretation would be more clinically meaningful than with the continuous scale. Prior studies have used a dichotomized construction of the CES-D scale as a measure of probable depression.20,23

Back to Top | Article Outline
Sexual Risk Behaviors

Multiple partners and sex with high-risk partners were selected as outcomes of interest. Engagement in these high-risk sexual behaviors has significance for HIV transmission independent of frequency of condom use.24 Further, research has shown infrequent condom use among sexually active drug users.25,26 Given that the frequency of condom use is low among drug users, these outcomes represent an important avenue of inquiry for HIV transmission. Participants who reported having only one sexual partner in the last 90 days were considered monogamous. High-risk sexual partners were IDUs and crack smokers. Variables were coded 1 if participants reported having an IDU or crack smoker as a sexual partner and coded 0 otherwise.

Back to Top | Article Outline
Statistical Methods

Longitudinal data analysis using generalized estimating equations (GEE) was employed for 996 observations. This statistical technique groups repeated measures for each subject and accounts for the correlation that may occur from multiple observations within subjects. The 3 outcome variables were dichotomous, so the logit link was specified. Examination of the within-group correlation suggested that an exchangeable correlation structure was appropriate for the data. Depressive symptoms was the primary independent variable of interest and was measured longitudinally along with the 3 outcome variables: monogamy, sex with an IDU, and sex with a crack smoker. Covariates included in the model were also measured longitudinally, except for gender. We adjusted for age (continuous), gender, self-reported HIV status, injection status (yes/no), cocaine/crack use, and time by wave. We constructed a 4-category cocaine/crack use variable: 0 = no cocaine or crack use, 1 = crack use only, 2 = cocaine use only (sum of inject and snort, but no crack use), and 3 = cocaine and crack use. This was done to tease apart potentially separate effects of cocaine and crack on sexual behavior. Finally, having a main sexual partner (yes/no) was added to the model examining monogamy.

Attrition analysis was conducted to see whether those who were lost to follow-up for any reason (death, incarceration, unable to locate, lost interest) were comparable to those who remained in the study over the 3 waves. No significant differences were noted among the variables included in the analyses. Further, we tested for effects of the intervention and follow-up priority on the sexual outcomes and found none.

Back to Top | Article Outline


Table 1 presents demographic, drug use, and depression characteristics of all participants who had completed 3 waves of survey data (n = 332). The sample is majority African-American, male, and low income. The mean age at baseline was 38 years and approximately 43% of the sample attained at least a high school diploma or GED. Eighty-two percent of the sample had a main sexual partner at baseline, and this remained stable over the 3 waves. Fifty-one participants self-reported HIV-positive status at baseline; an additional 8 participants had reported being HIV positive by wave 3. Drug use characteristics reveal declines in injection drug use and crack smoking over time. Depressive symptoms was relatively high. The mean score at baseline was 18.4 (SD = 11.6) and declined slightly over time. Fifty-four percent of participants had scores > (an indication of depression) at baseline; this declined to 47 and 35% at waves 2 and 3, respectively.

Table 1
Table 1
Image Tools

The percentages of participants reporting sexual risk behaviors by depression status and wave are presented in Table 2. In general, participants with depressive symptom scores <16 reported less sexual risk behavior than participants with scores of ≥16 for each wave. χ2 tests for linear trend were performed for each risk behavior by wave. Trends were noted for sex with an IDU (P < 0.01) and sex with someone who smokes crack (P < 0.05), but not for monogamy (P = 0.93) (data not shown).

Table 2
Table 2
Image Tools

Results from repeated-measures analysis with GEE are reported in Table 3. Depression was a significant predictor of risky sexual behavior in all models fitted. Participants reporting high depressive symptoms were >50% more likely to have >1 sex partner (odds ratio [OR] = 1.55, CI: 1.12, 2.14) and to have sex with an IDU (OR = 1.57, CI: 1.11, 2.22). High depressive symptoms also increased the odds of having sex with a crack user (OR = 1.37, CI: 1.02, 1.86). Several of the covariates entered in the models were also significantly associated with the risk behaviors. Being HIV positive was a significant predictor of having sex with an IDU (OR = 2.18, CI: 1.34, 3.57) but not for the other 2 risk behaviors. Having a main sexual partner was only entered into the model with monogamy and was found to be a significant protective factor against having multiple partners (OR = 0.31, CI: 0.21, 0.45). Another noteworthy finding was the relationship between cocaine/crack use and the sexual risk behaviors. Cocaine use only was not associated with any of the behaviors; but combined cocaine and crack use and crack use only were both significant predictors of having sex with an IDU and having sex with a crack user. Not surprising, this association was stronger for the outcome, sex with a crack user. Finally, injection drug use predicted having sex with a crack user (OR = 1.56, CI: 1.09, 2.24) and was strongly related with having sex with an IDU (OR = 6.38, CI: 4.19, 9.73).

Table 3
Table 3
Image Tools
Back to Top | Article Outline


In this study, we examined longitudinal effects of depression on the sexual risk behaviors of multiple sex partners and sex with risky partners. We found stable and significant effects of high depressive symptoms on these behaviors, adjusting for relevant covariates. Specifically, high depressive symptoms increased the likelihood of engaging in risky sexual behavior for our sample of sexually active drug users.

This is the first study to our knowledge to longitudinally examine the relationship between depression and risk behaviors among drug users. This analytical design lends support for a causal association between depression and sex risks and has significant implications for the health of drug users. Depressive symptoms tend to be more prevalent among drug users than the general population. Mean scores in our sample ranged from 14-18.4 across the 3 waves, while scores for the general population average around 8.27 Further, prevalence ranged from 35-54% in our sample, while lifetime prevalence estimates for the general population range between 7-14%.28 The results from this study suggest that the high prevalence of depressive symptoms in drug users may translate into high-risk sexual behavior, which ultimately has significant implications for the continuing spread of HIV. In addition to the significant effect of depressive symptoms, using crack, being HIV positive, and injecting drugs all increased the risk of having sex with an IDU. This finding is particularly important considering that one-third of new infections are IDU related. The association with drug injection and HIV-positive status are not as surprising as the association with crack use. Crack use increased the risk of having sex with an IDU while injection drug use increased the risk of having sex with a crack user. This connection between injection drug use and crack use suggests the need for additional research.

The hypothesis that having a main sexual partner would be protective of having sex with multiple partners was supported by our results. We had further hypothesized that HIV status would be related to sexual risk behavior. We found a significant association with sex with an IDU, but not with the other 2 outcomes. Additionally, the direction of the association was opposite of what we had hypothesized. We had hypothesized that HIV-positive individuals would be less likely to engage in risky sex, but we found that HIV-positive participants were 2 times more likely to have sex with an IDU. This association is likely due to the strong relationship between injection drug use and HIV status.

Though this study is strong methodologically, there are several noteworthy limitations. First, we were able to use only 3 waves of data for this study. Second, the data we use are self-report and may suffer from social desirability and recall biases that are characteristic of self-report surveys. Several studies, however, have shown that self-reports from drug users have high reliability.29-31 In this paper, we used high CES-D score (≥16) as our measure of depression; thus, we were only able to measure depressive symptoms or probable depression. A diagnostic assessment would be necessary to determine if participants were clinically depressed. Finally, our sample was from a low-income street-recruited population of current and former drug users in the inner city. Thus, generalizations to other populations may not be appropriate.

Despite these limitations, we think our study makes a significant contribution to the debate on the nature of the relationship between depression and sexual risk behavior. The results have implications for the design and development of HIV prevention programs. Results suggest that depressive symptoms are a major contributing factor to sexual risk behavior, and therefore, at minimum, depression should be assessed in intervention studies. Other practices that could be adopted as part of HIV prevention interventions and research are counseling or referrals to psychiatric care. The stress of high crime and poor living conditions may also contribute to depressive symptoms.32 Structural interventions that address these issues may have an indirect effect of reducing depression and hence HIV transmission.

Back to Top | Article Outline


1. Brooner RK, King VL, Kidorf M, et al. Psychiatric and substance use comorbidity among treatment-seeking opioid users. Arch Gen Psychiatry. 1997;54:71-80.

2. Regier DA, Farmer ME, Rae DS, et al. Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) Study. JAMA. 1990;264:2511-2518.

3. Ross HE, Glaser FB, Germanson T. The prevalence of psychiatric disorders in patients with alcohol and other drug problems. Arch Gen Psychiatry. 1988;45:1023-1031.

4. Rounsaville BJ, Weissman MM, Crits-Christoph K, et al. Diagnosis and symptoms of depression in opiate addicts. Arch Gen Psychiatry. 1982;39:151-156.

5. Mandell W, Kim J, Latkin C, et al. Depressive symptoms, drug network, and their synergistic effect on needle-sharing behavior among street injection drug users. Am J Drug Alcohol Abuse. 1999;25:117-127.

6. Hawkins WE, Latkin C, Hawkins MJ, et al. Depressive symptoms and HIV-risk behavior in inner-city users of drug injections. Psychol Rep. 1998;82:137-138.

7. Latkin CA, Mandell W. Depression as an antecedent of frequency of intravenous drug use in an urban, nontreatment sample. Int J Addict. 1993;28:1601-1612.

8. Kral AH, Bluthenthal RN, Lorvick J, et al. Sexual transmission of HIV-1 among injection drug users in San Francisco, USA: risk-factor analysis. Lancet. 2001;357:1397-1401.

9. Strathdee SA, Galai N, Safaiean M, et al. Sex differences in risk factors for HIV seroconversion among injection drug users: a 10-year perspective. Arch Intern Med. 2001;161:1281-1288.

10. Dolezal C, Meyer-Bahlburg HF, Liu X, et. al. Longitudinal changes in sexual risk behavior among HIV+ and HIV− male injecting drug users. Am J Drug Alcohol Abuse. 1999;25:281-303.

11. Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention. Fact Sheets: Drug-Associated HIV Transmission Continues in the United States. National Center for HIV, STD & TB Prevention. September 2000. Available at: Accessed June 12, 2003.

12. Compton WM. Cocaine use and HIV risk in out of treatment drug abusers. Accessed June 12, 2003. Drug Alcohol Depend. 2000;58:215-218.

13. Camacho LM, Brown BS, Simpson DD. Psychological dysfunction and HIV/AIDS risk behavior. J Acquir Immune Defic Syndr Hum Retrovirol. 1996;11:198-202.

14. Hudgins R, McCusker J, Stoddard A. Cocaine use and risky injection and sexual behaviors. Drug Alcohol Depend. 1995;37:7-14.

15. Crepaz N, Marks G. Are negative affective states associated with HIV sexual risk behaviors? A meta-analytic review. Health Psychol. 2001;20:291-299.

16. Kalichman SC, Weinhardt L. Negative affect and sexual risk behavior: comment on Crepaz and Marks (2001). Health Psychol. 2001;20:300-301.

17. Dolezal C, Ehrhardt AA, Meyer-Bahlburg HF, et al. Sexual risk behavior changes among HIV+ and HIV− female injecting drug users over 4 years. Women Health. 1998;27:1-17.

18. Latkin CA, Forman V. Patterns of needle acquisition and sociobehavioral correlates of needle exchange program attendance in Baltimore, Maryland. J Acquir Immune Defic Syndr. 2001;27:398-404.

19. Radloff L. The CESD scale: a self-report depression scale for research in the general population. Applied Psychol Measures. 1977;1:385-401.

20. Knowlton AR, Latkin CA, Shroeder JR, et al. Longitudinal predictors of depressive symptoms among low income injection drug users. AIDS Care. 2001;13:549-559.

21. Grella CE, Anglin MD, Annon JJ. HIV risk behaviors among women in methadone maintenance treatment. Subst Use Misuse. 1996;31:277-301.

22. Boyd JH, Weissman MM, Thompson WD, et al. Screening for depression in a community sample: understanding the discrepancies between depression symptom and diagnostic sales. Arch Gen Psychiatry. 1982;39:1195-1200.

23. Wingood GM, DiClemente RJ. The influence of psychosocial factors, alcohol drug use on African-American women's high-risk sexual behavior. Am J Prev Med. 1998;15:54-59.

24. Centers for Disease Control and Prevention. Continued sexual risk behavior among HIV-seropositive, drug-using men: Atlanta; Washington, D.C.; and San Juan, Puerto Rico, 1993. Morb Mortal Wkly Rep. 1996;45:151-152.

25. Booth RE, Kwiatkowski CF, Chitwood DD. Sex related HIV risk behaviors: differential risks among injection drug users, crack smokers, and injection drug users who smoke crack. Drug Alcohol Depend. 2000;58:219-226.

26. Kwiatkowski CF, Stober DR, Booth RE, et al. Predictors of increased condom use following HIV intervention with heterosexually active drug users. Drug Alcohol Depend. 1999;54:57-62.

27. Hasin DS, Glick H. Depressive symptoms and DSM-III-R alcohol dependence: general population results. Addiction. 1993;88:1431-1436.

28. Blazer DG, Kessler RC, McGonagle KA, et al. The prevalence and distribution of major depression in a national community sample: the National Comorbidity Survey. Am J Psychiatry. 1994;151:979-986.

29. Petry NM. Reliability of drug users' self-reported HIV risk behaviors using a brief, 11-item scale. Subst Use Misuse. 2001;36:1731-1747.

30. Goldstein MF, Friedman SR, Neaigus A, et al. Self-reports of HIV risk behavior by injecting drug users: are they reliable? Addiction. 1995;90:1097-1104.

31. McElrath K, Chitwood DD, Griffin DK, et al. The consistency of self-reported HIV risk behavior among injection drug users. Am J Public Health. 1994;84:1965-1970.

32. Latkin C, Curry A. Stressful neighborhoods and depression: a prospective study of the impact of neighborhood disorder. J Health Soc Behav. 2003;44:34-44.

Cited By:

This article has been cited 1 time(s).

Sexually Transmitted Diseases
The Influence of Depression on Sexual Risk Reduction and STD Infection in a Controlled, Randomized Intervention Trial
Holden, AE; Shain, RN; Miller, WB; Piper, JM; Perdue, ST; Thurman, AR; Korte, JE
Sexually Transmitted Diseases, 35(10): 898-904.
PDF (243) | CrossRef
Back to Top | Article Outline

depressive symptoms; HIV risk; drug users; sexual risk behaviors

© 2005 Lippincott Williams & Wilkins, Inc.


Article Tools



Article Level Metrics

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.