FENAUGHTY, ANDREA M. PhD; FISHER, DENNIS G. PhD
DRUG USERS are at considerable risk for human immunodeficiency virus (HIV) and acquired immune deficiency syndrome (AIDS) because of both unsafe drug use (e.g., using previously used injection equipment1–3) and unsafe sexual behavior (e.g., having multiple sex partners and rarely using condoms2,4–8). Prevention efforts aimed at drug users have focused on changing both risky drug use and risky sexual behavior to slow the spread of HIV infection9; however, sexual behavior appears to be more difficult to change.9–11 One reason for the minimal effectiveness of interventions targeting sexual risk reduction among drug users12–15 may be that these interventions typically treat all participants similarly. However, it is clear from the drug abuse literature that there is substantial heterogeneity among drug users.16–20
A typology based on alcohol use may be useful in distinguishing among levels of behavioral risk for HIV. There has been much evidence in support of a link between alcohol and high‐risk sexual behavior,21–23 with explanations ranging from pharmacological disinhibition24–25 to personality characteristics.26 For example, one study found that the association between alcohol use and risky sexual behavior disappeared after controlling for personality characteristics such as sensation seeking and impulsivity.26
Relatively little research on the link between alcohol and sexual risk behavior has been conducted using samples of drug users; however, the extant research does indicate an association between alcohol and risky sex among samples of injection drug users (IDUs)27–28 and crack smokers and IDUs.29–30 The purpose of this study was to develop a typology of drug users based on alcohol use variables and then determine the utility of this typology for predicting high‐risk sexual behavior, controlling for the personality traits of sensation seeking and risk proneness.
Data were collected as part of the National Institute on Drug Abuse (NIDA) Cooperative Agreement for AIDS Community‐Based Outreach/Intervention Research Program (CA). The data were collected between October 1994 and January 1996 from out‐of‐treatment IDUs and cocaine smokers in Anchorage, AK. Participants (N = 283) were recruited via targeted sampling. Targeted sampling identifies samples of hidden populations by targeting specific geographic regions within which to recruit.31 Each participant had to meet the following criteria: (1) be at least 18 years of age; (2) selfreport smoking cocaine within the last 48 hours and produce positive urinalysis for cocaine metabolites or self‐report injecting illicit drugs within the last 30 days and present needle track marks indicative of recent injection drug use; and (3) not have been in drug or alcohol treatment in the prior 30 days. Because most of the risk behaviors involved sexual behavior, only participants who reported having at least one sex partner in the last 30 days were included. Urinalysis was performed using the ONTRAK system (Roche Diagnostics, Nutley, NJ).
Data for this study were obtained from the Risk Behavior Assessment32 (RBA) and the Antisocial Personality/Sensation Seeking Questionnaire (ASP/SS). The RBA is a structured interview developed by a team of principal investigators participating in the CA. The RBA is used to assess sexual‐ and drug‐related risk behaviors, as well as demographics and other health‐related information, and has been demonstrated to have acceptable reliability33–34 and validity.35–36 The ASP/ SS is composed of the Antisocial Personality Disorder section (ASPD) of the Semi‐Structured Assessment for the Genetics of Alcoholism (SSAGA), Form V of the Sensation‐Seeking Scale37 (SSS), the Impulsivity and Risk‐Taking scale38 (IRT), and three items that assess recent alcohol use (described below). Data from the ASPD were not analyzed for this study.
Individual differences in propensity for engaging in high‐risk activities were assessed by the Sensation Seeking Scale (SSS), developed as a measure of individual differences in optimal levels of stimulation.39 The SSS has four 10‐item subscales: Thrill and Adventure Seeking (TAS), Experience Seeking (ES), Disinhibition (Dis), and Boredom Susceptibility (BS).40 Degree of risk proneness was measured by the Impulsivity and Risk‐Taking Scale (IRT).
Alcohol Use and Sexual Risk Behavior Variables
Frequency and amount of alcohol consumed and frequency of drinking to the point of inebriation were assessed with items from the ASP/SS. The following sexual risk behavior items were extracted from the RBA: proportion of protected sex, any use of condoms, having multiple sex partners (without consistent condom use), having an IDU sex partner (without consistent condom use), STD history, and history of trading sex for drugs. (Details are available from the author.)
On meeting eligibility requirements, participants were assured of confidentiality and were asked to give written evidence of informed consent. Participants were administered the battery of questionnaires, including the RBA and the ASP/SS. Participants also received NIDA‐standardized pretest and posttest counseling for HIV testing, had their blood drawn, and were paid for time spent in research.
Cluster analysis was used to identify a typology or clustering of drug users based on the set of alcohol variables. Cluster analysis is a tool used to identify groups of individuals who respond similarly on one set of variables. One can then examine differences among these clusters on another set of variables.41 Accordingly, cluster analysis was used to identify clusters of drug users with similar patterns of alcohol use. Bivariate tests, including chi‐squares and t tests, were then used to examine the bivariate association between the typology and each of the sexual risk behavior variables. Logistic and linear regression were used to examine the association between the alcohol typology and each dichotomous or continuous sexual risk behavior, respectively, while controlling for sensation seeking and risk proneness. Clustan (Edinburgh, Scotland)42 was used to perform the cluster analyses; for all other analyses, SAS (Cary, NC) was used.43
The sample was 66% male, 44% white, 26% black, and 24% American Indian/Alaska Native (Native American). Compared with the ethnic distribution of the men, women were less likely to be black and more likely to be Native American, χ2(2) = 18.31, p <. 001. Men were significantly older (M = 36.59, SD = 7.45) than women (M = 33.73, SD = 7.19, t(281) = 3.10 p <. 01). Twenty percent of the sample never graduated from high school; there were no gender differences in level of education. The median monthly income for men was less than $500; the median monthly income for women was between $500 and $999 (z = 2.67, p <. 01, median 2‐sample test).
Alcohol Use and Sexual Behavior
Alcohol was consumed a median of 7 of the 30 days before assessment. The median age of first alcohol use was 14. Alcohol was used during sex a median of two times in the last month. Median frequency of alcohol consumption was once or twice a week, with a median amount of alcohol used per time of 3 to 4 drinks. Seventy‐one percent of the sample reported drinking to intoxication at least once in the prior 6 months.
In the 30 days before assessment, condoms were used by only 32% of the sample. The mean proportion of sex acts for which condoms were used was 19. Twenty‐seven percent reported having multiple sex partners and not always using condoms, and 25% reported having at least one IDU sex partner and not always using condoms. Fifty‐three percent of the sample reported having had an STD in the last 6 months, with a median of one STD during that time. The median number of sex partners in the 30 days before assessment was one. There were no significant gender differences in any of the above‐listed alcohol or sexual risk behavior variables. Forty percent of the women reported having given sex to receive money or drugs in the 30 days before assessment; 19% reported having done so in the prior 6 months. Forty‐one percent of the men reported having given drugs or money to receive sex. Owing to the lack of gender differences, the following analyses were performed using a combined sample of men and women.
A complete linkage with squared Euclidean distance cluster solution was selected to define the typology. (Details are available from the author.) Each of the six variables used in the cluster analysis had a different scale; as a result, to facilitate interpretation of the mean profiles, all scores were transformed into T scores (M = 50, SD = 10) following the cluster analysis (see Figure 1). Relative to cluster 1 (n = 203), cluster 2 (n = 80) was high on amount and frequency of alcohol consumed in the last 6 months, frequency of being drunk, number of days in the past month in which alcohol was consumed, and proportion of sexual acts that involved the use of alcohol. Cluster 2 also began using alcohol at an earlier age than cluster 1. Thus, this alcohol typology can be characterized as distinguishing between high‐ and low‐risk patterns of alcohol use, with cluster 2 being the high‐risk group.
Results of the bivariate analyses between alcohol typology and each of the sexual risk behavior variables are presented in Table 1. Although no association was shown between typology and whether condoms had been used at all in the last month, drug users in the high‐risk group (cluster 2) were significantly more likely than those in the low‐risk group to have had multiple sex partners and/or an IDU sex partner with inconsistent use of condoms. Those in the high‐risk group reported a lower proportion of protected sexual acts in the last 30 days (M = .13) than those in the low‐risk group (M = .22), but this difference was only marginally significant (p = .06). High‐risk drug users had more sex partners in the last 30 days (M = 5.38) and were more likely to have ever had an STD (66%) than low‐risk drug users (M = 2.27 sex partners and 48%, respectively); the difference in total number of lifetime STDs approached significance (p = .055). Finally, high‐risk women were significantly more likely than low‐risk women to have traded sex for drugs or money both in their lifetime and in the last 30 days.
A set of logistic regression models was used to assess the relationship between alcohol typology and dichotomous sexual risk behavior variables, adjusting for the effects of sensation seeking and risk proneness. Results of these analyses are presented in Table 2. Each of the sexual risk behaviors remained associated with alcohol typology, with the exception of lifetime use of condoms in the last 30 days and (for men) trading money or drugs to get sex. Results of the linear regression analyses using continuous outcomes are presented in Table 3. Proportion of protected sexual acts, number of sex partners, and number of lifetime STDs remained significantly associated with alcohol typology after controlling for sensation seeking and risk proneness.
Among this sample of out‐of‐treatment drug users, a typology based on patterns of alcohol use was found to be associated with high‐risk sexual behavior. Drug users classified as high risk on the basis of their lifetime and current alcohol use patterns were significantly more likely than low‐risk drug users to have had multiple sex partners and inconsistently used condoms, have had at least one IDU sex partner and not consistently used condoms, and have ever had an STD. Female drug users classified as high risk on the basis of the typology were also more likely to have given sex to get drugs or money. The alcohol typology was also significantly related to number of sex partners, the proportion of protected sex acts, and number of lifetime STDs. Furthermore, these associations were maintained even after controlling for the personality traits of sensation seeking and risk proneness.
These findings suggest that a profile based on a relatively small set of alcohol variables can be used to identify drug users likely to be engaging in a range of sexual behaviors that place them at increased risk of HIV and other STD infection. The present study focused on risk behavior and not disease morbidity as an outcome; however, there is some evidence that STD rates among drug users in the Anchorage area are high enough to pose cause for concern. In a comparison of self‐reported history of STDs among 23 sites nationwide, drug users from the Anchorage site reported* the second highest mean number of events for both Chlamydia and genital warts, and the fifth highest mean number of events for genital herpes (unpublished data). The cohort of drug users in Anchorage has an HIV seroprevalence of 2%.
One implication of this study's findings is that alcohol use profiles could be used to target those drug users who are in most need of preventive interventions that focus on sexual risk reduction. For example, drug users enrolled in alcohol treatment programs and out‐of‐treatment drug users who fit the high‐risk alcohol profile may be targeted for HIV risk reduction interventions. Future research should explore the role of such targeting within preventive interventions and assess the extent to which it improves the effectiveness of the intervention.
The limitations of this study need to be noted. The results should not be generalized to non‐drug using populations or to drug users who are currently in treatment. In‐treatment drug users differ from out‐of‐treatment drug users on a number of variables,31 including HIV infection.44 Furthermore, the extent to which the pattern of alcohol use characterizing the high‐risk cluster would generalize to other out‐of‐treatment drug users is unclear. Alcohol use rates in Alaska are extremely high,45 and patterns of alcohol use may vary by region or culture. However, it is not clear that the association between high‐risk alcohol use and risky sexual behavior is dependent on a high mean level of alcohol use. Future research needs to examine the possible moderating effects of level of alcohol use on the association between high‐risk alcohol use and risky sexual behavior.
The results from this study are consistent with previous research showing an association between alcohol use and sexual risk behavior among drug users.27–29,46 Further, the findings lend strength to the argument made by Latkin et al28 that the association between alcohol use and high‐risk sexual behavior was not because of a risk‐prone or sensation‐seeking personality type. There were associations between some of the personality variables and certain sexual risk behaviors (e.g., number of sex partners and impulsivity and risk taking); however, these associations were consistently unable to account for the association between alcohol typology and risk behavior.
There are a number of theoretical frameworks within which the findings of this study may fit. The results are consistent with a causal disinhibitory explanation for the alcohol‐risk behavior relationship. This explanation posits that alcohol consumption leads to cognitive impairment, which in turn leads to disinhibition of certain sexual behaviors.24–25 Although the correlational methods in the current study preclude causal inference regarding the link between alcohol and risk behavior, the observed pattern of data is what one would expect to find if alcohol were playing a role in the disinhibition of sexual activity (e.g., having multiple partners and not using condoms. However, the data are also consistent with a number of other explanations, including the expectancy effects hypothesis,47 which posits that alcohol affects behavior through the expectancies people hold regarding how alcohol should influence behavior. It is likely that no single explanation accounts for the link between risky alcohol use and risky sexual behavior, but rather that multiple mechanisms underlie this effect.21 Regardless of the exact nature of this mechanism or mechanisms, this study adds to the growing body of research indicating that drug users and their sex partners are in need of HIV prevention services that have been developed with both an understanding of the interrelatedness and complexity of the behaviors they seek to change and an acknowledgment of the heterogeneity of the clients they aim to serve. Alcohol use in particular appears to play a significant role in the HIV risk behavior of drug users. A better understanding of this role and its consequences may indicate the need for differential interventions based on alcohol use.
1. Chitwood DD, Griffin DK, Comerford M, Page JB, Lai S, McCoy CB. Risk factors for HIV-1 seroconversion among injection drug users: a case-control study. Am J Pub Health 1995; 85:1538-1542.
2. Lamothe F, Bruneau J, Coates R, et al. Seroprevalence of and risk factors for HIV-1 infection in injection drug users in Montreal and Toronto: a collaborative study. Can Med Assoc J 1993; 149:945-951.
3. Marmor M, Des Jarlais DC, Cohen H, et al. Risk factors for infection with human immunodeficiency virus among intravenous drug abusers in New York City. AIDS 1987; 1:39-44.
4. Baker A, Kochan N, Dixon J, Wodak A, Heather N. Drug use and HIV risk-taking behavior among injecting drug users not currently in treatment in Sydney, Australia. Drug Alcohol Depend 1994; 34:155-160.
5. Feucht TE, Stephens RC, Roman SW. The sexual behavior of intravenous drug users: Assessing the risk of sexual transmission of HIV. J Drug Issues 1990; 20:195-213.
6. Kim MY, Marmor M, Dubin N, Wolfe H. HIV risk-related sexual behaviors among heterosexuals in New York City: associations with race, sex, and intravenous drug use. AIDS 1993; 7:409-414.
7. Lewis DK, Watters JK. Sexual behavior and sexual identity in male injection drug users. J Acquir Immune Defic Syndr Hum Retrovirol 1994; 7:190-198.
8. Wells EA, Calsyn DA, Saxon AJ, Greenberg DM. Using drugs to facilitate sexual behavior is associated with sexual variety among injection drug users. J Nerv Ment Dis 1993; 181:626-631.
9. van den Hoek JAR, Van Haastrecht HJA, Coutinho RA. Little change in sexual behavior in injecting drug users in Amsterdam. J Acquir Immune Defic Syndr Hum Retrovirol 1992; 5:518-522.
10. Fitterling JM, Matens PB, Scotti JR, Allen JS. AIDS risk behaviors and knowledge among heterosexual alcoholics and non-injecting drug users. Addiction 1993; 88:1257-1265.
11. McKeganey NP, Barnard MA, Watson H. HIV related risk behaviour among a non-clinic sample of injecting drug users. Br J Addiction 1989; 84:1481-1490.
12. McKusker J, Stoddard AM, Zapka JG, Lewis BF. Behavioral outcomes of AIDS educational interventions for drug users in short-term treatment. Am J Pub Health 1993; 83:1463-1466.
13. Watters JK, Downing M, Case P, Lorvick J, Cheng Y-T, Fergusson B. AIDS prevention for intravenous drug users in the community: Street-based education and risk behavior. Am J Commun Psychol 1990; 18:587-596.
14. Donoghoe MC, Stimson GV, Dolan K, Alldritt L. Changes in HIV risk behaviour in clients of syringe-exchange schemes in England and Scotland. AIDS 1989; 3:267-272.
15. Kall K, Olin RG. HIV status and changes in risk behavior among intravenous drug users in Stockholm 1987-1988. AIDS 1990; 4:153-157.
16. Friedman SR, Nealgus A, Des Jarlais DC, et al. Social interventions against AIDS among injecting drug users. Br J Addiction 1992; 87:393-404.
17. Fisher DG, Anglin MD, Weisman CP, Pulliam L. Replication problems of substance abuser MMPI cluster types. Multivar Behav Res 1989; 24:335-352.
18. Swadi H. Psychiatric symptoms in drug abusing adolescents. Drug Alcohol Depend 1992; 31:77-83.
19. Fals-Stewart W. Personality characteristics of substance abusers: An MCMI cluster typology of recreational drug users treated in a therapeutic community and its relationship to length of stay and outcome. J Pers Assess 1992; 59:515-527.
20. Isenhart CE. Motivational subtypes in an inpatient sample of substance abusers. Addictive Behav 1994; 19:463-475.
21. Stall R, McKusick L, Wiley J, Coates TJ, Ostrow DG. Alcohol and drug use during sexual activity and compliance with safe sex guidelines for AIDS: The AIDS Behavioral Research Project. Health Educ Q 1986; 13:359-371.
22. Hingson RW, Strunin L, Berlin BM, Heeren T. Beliefs about AIDS, use of alcohol and drugs, and unprotected sex among Massachusetts adolescents. Am J Public Health 1990; 80:295-299.
23. Leigh BC, Stall R. Substance use and risky sexual behavior for exposure to HIV: Issues in methodology, interpretation, and prevention. Am Psychol 1993; 48:1035-1045.
24. Crowe LC, George WH. Alcohol and human sexuality: Review and integration. Psychol Bull 1989; 105:541-551.
25. Steele CM, Southwick L. Alcohol and social behavior I: the psychology of drunken excess. J Pers Soc Psychol 1985; 48:18-34.
26. Temple MT, Leigh BC, Schafer J. Unsafe sexual behavior and alcohol use at the event level: Results of a national survey. J Acquir Immune Defic Syndr Hum Retrovirol 1993; 6:393-401.
27. Calsyn DA, Saxon AJ, Wells EA, Greenberg DM. Longitudinal sexual behavior changes in injecting drug users. AIDS 1992; 6:1207-1211.
28. Latkin C, Mandell W, Oziemkowska M, Vlahov D, Celentano D. The relationships between sexual behavior, alcohol use, and personal network characteristics among injecting drug users in Baltimore, Maryland. Sex Transm Dis 1994; 21:161-167.
29. Turner S, Paschane D, Johnson M, Fenaughty A, Fisher G. Alcohol consumption by Alaskan drug users not currently in treatment. Poster presented at the 10th International Circumpolar Conference on Health, Anchorage, AK. 1996.
30. Falck RS, Wang J, Carlson RG, Siegal HA. Factors influencing condom use among heterosexual users of injection drugs and crack cocaine. Sex Trasm Dis 1997; 24:204-210.
31. Watters JK, Biernacki P. Targeted sampling: Options for the study of hidden populations. Soc Probl 1989; 36:416-430.
32. National Institute on Drug Abuse. Risk behavior assessment. Rockville, MD: National Institute on Drug Abuse (Community Research Branch), 1991.
33. Fisher DG, Needle R, Weatherby N, et al. Reliability of drug user self-report. IXth International Conference on AIDS. Berlin, Germany, 1993. Abstract (PO-C35-3355).
34. Needle R, Fisher DG, Weatherby N, et al. The reliability of self-reported HIV risk behaviors of drug users. Psychol Addictive Behav 1995; 9:242-250.
35. Dowling-Guyer S, Johnson ME, Fisher DG, et al. Reliability of drug users' self-reported HIV risk behaviors and validity of self-reported recent drug use. Assessment 1994; 1:383-392.
36. Weatherby NL, Needle R, Cesari H, et al. Validity of self-reported drug use among injection drug users and crack cocaine users recruited through street outreach. Eval Prog Plan 1994; 17:347-355.
37. Zuckerman M. Sensation seeking: beyond the optimal level of arousal. Hillsdale, NJ: Lawrence Erlbaum Associates, 1979.
38. Schafer J, Blanchard L, Fals-Stewart W. Drug use and risky sexual behavior. Psychol Addictive Behav 1994; 8:3-7.
39. Zuckerman M, Kolin EA, Price L, Zoob I. Development of a sensation-seeking scale. J Consult Clin Psychol 1964; 28:477-482.
40. Zuckerman M. Dimensions of sensation seeking. J Consult Clin Psychol 1971; 36:45-52.
41. Aldenderfer M, Blashfield RK. Cluster Analysis. Sage University Paper series on Quantitative Applications in the Social Sciences, 07-044. London: Sage Publications, 1984.
42. Wishart D. Clustan user manual, 4th ed. Edinburgh: Program Library Unit, Edinburgh University, 1987.
43. SAS Institute Inc. SAS/STAT User's guide. Version 6, 4th ed., Vols 1 & 2. Cary, NC: SAS Institute Inc., 1989.
44. Watters JK, Lewis DK. HIV infection, race, and drug-treatment history. AIDS 1990; 4:697-702.
45. Tumer S, Paschane D, Johnson M, Fenaughty A, Fisher G. Alcohol consumption by Alaskan drug users not currently in treatment. Poster presented at the 10th International Circumpolar Conference on Health, Anchorage, AK, 1996.
46. Schilling R, Serrano Y, Faruque S, et al. Predictor variables of trading sex among male drug users in Harlem. International Conference on AIDS. Abstracts 8(2), C352 1992.
47. Leigh BC. Beliefs about the effects of alcohol on self and others. J Studies Alcohol 1987; 48:467-475.
*Although self‐report data are not optimal indicators of STD prevalence, there is no reason to assume differential validity of this indicator across the 23 sites. Thus, self‐report gives an indication of STD prevalence in a relative sense. Cited Here...