Association of Risky Sexual Behavior with Alcohol Consumption and Tobacco Use among Youth in Two Developing Countries: Haiti and Uganda

Anastasia J. Gage, Tulane University
Chiho Suzuki, Tulane University

Adolescent sexual activity and its possible linkages with other health risk behaviors such as alcohol and tobacco use are perhaps the single most important issue on the global youth health agenda. Many studies allude to rising rates of premarital intercourse among adolescents which, combined with later marriage have lengthened the period of risk for unintended pregnancies and sexually transmitted diseases among adolescents. In many countries of the developing world, unintended birth rates are high and adolescents account for a substantial proportion of new HIV infections. Some studies have suggested that the use of alcohol and other substances increases risky sexual behavior while others have documented positive correlations between sexual involvement, cigarette smoking and drug use. In spite of these concerns, research describing the potential linkages between various high-risk activities among developing country youth is quite limited. This is partly due to the lack of nationally or regionally representative data describing a variety of adolescent health behaviors and a general focus on reproductive health (specifically, sexual activity and contraceptive use) and HIV/AIDS outcomes (such as condom use). Most existing studies have investigated a single health behavior rather than a spectrum of related health risk behaviors. Consequently much of what is known about the linkages between behaviors such as tobacco use, alcohol use, and sexual activity among adolescents come from studies conducted in Western nations. The linkages between adolescent sexual activity, alcohol consumption and cigarette smoking are of concern for several reasons. Adolescence is a time when unhealthy behaviors are initiated; hence youth constitute a population toward which a high proportion of preventive activities can be targeted. Even though the prevalence of specific health risk behaviors may be much less frequent among developing country adolescents than among their counterparts in western countries, rapid urbanization, greater internet connectivity and increasing exposure to outside influences may create environments conducive to the adoption of multiple behaviors that put adolescents at risk. Problem Behavior Theory posits that adolescent risk behaviors are manifestations of common underlying factors. If this holds true in developing country settings, then health professionals and program managers working to reduce sexual risk-taking and promote safe sex and sexual responsibility among youth would benefit from information about the prevalence, patterns and correlates of other health risk behaviors among youth. Early use of cigarettes and alcohol may indicate the existence of cognitive and social factors that predispose adolescents to risk-taking behaviors. Therefore, assessing the linkages between tobacco use and sexual activity forms an important part of any global or national effort to prevent health endangering behaviors among youth. Additionally, tobacco use is one of the chief preventable causes of morbidity and mortality worldwide. The specific objectives of the study are to: 1) Analyze the linkages between risky sexual behavior and tobacco and alcohol use among adolescents aged 15-24 2) Explore whether the linkages between these behaviors differ by gender 3) Identify a set of common risk and protective factors that predict the adolescent behaviors under examination The analysis is guided by three main theoretical perspectives: problem behavior theory, social influence theory, and stress theory. Problem behavior theory suggests that adolescent high-risk behaviors are all manifestations of common underlying factors. Social influence theory suggests that adolescent behaviors derive largely from interaction with others and modeling. Thus, if adolescents live in families or communities where individuals engage frequently in high risk activities, adolescents would be more likely to do so themselves. Stress theory suggests the participation in high-risk behaviors is a response to stress, conflict and instability in adolescents’ immediate family and social environments. The data come from the 2000-2001 Uganda Demographic and Health Survey and the 2000 Haiti Enquête Mortalité, Morbidité et Utilisation des Services. The risky sexual behaviors examined are those that have been identified as increasing the risk of contracting AIDS or STDs. These include (1) transactional sex (2) more than one sexual partner (3) sex with a casual partner and (4) non-use of condoms. The other health risk behaviors examined are ever use of tobacco and use of alcohol in the past three months. Each variable is score 0 if the adolescent never engaged in the behavior and 1, otherwise. Problem behavior suggests there is a significant overlap in the set of unobservable variables that affect each outcome. This overlap in the unobservable variables means that the error terms of the three equations (alcohol use, tobacco use, and risky sexual behavior) are correlated. The analytical strategy used, therefore, is joint estimation/simultaneous equation modeling for the three outcome variables. Since all the dependent variables are dichotomous, trivariate probit models are used for the multivariate analysis. These models estimate not only the regression coefficients but also the error correlations. The analysis is performed using LIMDEP. To eliminate problems of intrahousehold correlation, the final sample for the multivariate analysis includes only one youth per household. Such intrahousehold correlation, if it exists would bias estimates of the standard error of the coefficients, although the coefficients themselves would be consistent. Our expectations with regard to the study findings are as follows: (1) Risk-taking behavior will be singular among younger adolescents (that is, younger adolescents are more likely to participate in one risk-taking activity than in several); (2) There will be a strong positive correlation between alcohol use, tobacco use and risky sexual behavior. This correlation will be stronger among older male youth and out-of-school youth. (3) The likelihood of engaging in risk-taking behavior will be higher with exposure to violence in the home, consumption of alcohol and/or tobacco by household members, and residence in communities with a high prevalence of risky behaviors among youth. (4) Low socioeconomic status will be inversely related to drinking and smoking (due to affordability issues) but positively related to sexual risk-taking.