Background Sexually transmitted infections (STIs), including HIV, lead to significant reproductive morbidities worldwide. The association between risky sexual behaviours and increased STI/HIV prevalence has been well-documented. In addition, low self-esteem appears to lead to higher-risk sexual behaviours. However, the direct association between self-esteem and STI/HIV has not been well studied. We aimed to examine whether self-esteem directly affects STI/HIV prevalence, after adjusting for risky sexual behaviour.
Methods We conducted a secondary analysis of a cross-sectional study of sugar plantation residents in Tanzania. The 2004 study included a self-administered survey with self-esteem assessment and testing for syphilis, herpes simplex virus type 2 (HSV-2) and HIV. We restricted our analysis to individuals with valid STI/HIV results and complete self-esteem data. Through principal component analysis, the 8-item self-esteem scale was reduced to a single variable. This variable was further categorised as low, medium, and high self-esteem and was the primary exposure variable in a multivariable log binomial model with a combined outcome representing any prevalent STI/HIV (syphilis, HSV-2, or HIV).
Results From the full study sample of 556 residents, 431 (78%) individuals had both complete STI/HIV data and self-esteem assessment and were included in the analysis. The analysis population was 51% male and 49% female. Nearly 60% (n=254) had at least one STI/HIV. We observed no significant heterogeneity by gender in the effect of self-esteem on STI/HIV prevalence. Gender, transactional sex and alcohol use were retained as confounders in the final model; no other demographic or behavioural variables met our a priori statistical criteria for confounding. The adjusted prevalence ratios (PR) for STI/HIV for individuals with low self-esteem compared to those with medium self-esteem and high self-esteem were 0.95 (95% CI 0.80 to 1.13) and 0.92 (95% CI 0.67 to 1.27), respectively.
Conclusion Self-esteem was not independently associated with STI/HIV prevalence in plantation residents in Tanzania. Our findings suggest that the role of self-esteem in reproductive health may be limited to its association with increased sexual risk behaviour, which subsequently affects STI/HIV prevalence. Interventions aimed at improving self-esteem may only be appropriate if they reduce likelihood of participation in risky behaviours.