A Prospective Longitudinal Model of Substance Use Onset Among South African Adolescents

Abstract

Substance use onset among Colored adolescents between eighth and ninth grades in an urban area of Cape Town, South Africa was examined using latent transition analysis. Longitudinal self-report data regarding substance use (N= 1118, 50.9% female) were collected in 2004 and 2005. Results indicated that the pattern of onset was similar across genders; adolescents first tried either alcohol or cigarettes, followed by both, then dagga (cannabis), and then inhalants. The prevalence of lifetime cigarette use was slightly greater for females; dagga (cannabis) and inhalant use were greater for males. The similarity of developmental onset in the current sample to previous international work supports the promise of adapting prevention programs across contexts. The study's limitations are noted.

Keywords:

• Substance use onset
• gateway drugs
• South Africa
• latent transition analysis
• gender differences
• adolescent drug use

Notes

¹With the advent of artificial science and its theoretical underpinnings (chaos, complexity and uncertainty theories) it is now posited that much of human behavior is complex, dynamic, multi-dimensional, level/phase structured, non-linear, law-driven and bounded (culture, time, place, age, gender, ethnicity, etc.). “Gateway drugs,” however these are defined, would be such a phenomena. The term and posited process has yet to be adequately delineated in terms of its dimensions. Nor do we currently know, in a predictable sense, the critical conditions (endogenous as well as exogenous ones) which are necessary for this process to operate or not to operate. Lastly “gateway drugs” are, in a sense, mystified and empowered while the human organism or other living organisms who also “use” drugs, are disempowered. This is not a semantic issue. There are at least two important issues to consider and which are derived from this: (1) Using linear models/tools to study non-linear processes/phenomena can and does result in misleading conclusions and can therefore also result in inappropriate intervention; (2) the concepts prediction and control have different meanings and dimensions for non-linearity than they do in the more traditional linear ‘cause and effect’ paradigms. (Buscema, M. (1998), Artificial Neural Networks, Substance Use and Misuse, 33(1–3)). Editor's note.