Association of serum testosterone concentrations with central adiposity in men who inject heroin in Mombasa County, Kenya

Abstract

Introduction: heroin is the most commonly injected drug among people who inject drugs globally. Heroin injection disrupts the hypothalamic-pituitary axis, leading to hormonal imbalances such as altered testosterone concentrations in men. Men who inject heroin concurrently use other substances, which may further affect testosterone levels. Lower testosterone levels are associated with increased central adiposity and decreased sexual function. This study aimed to determine how substance use patterns and adiposity markers are associated with testosterone concentrations among heroin-injecting men in Mombasa County.

Methods: in this analytical cross-sectional study, serum samples from male participants with a documented history of substance use were included. Exclusion criteria included HIV-positive status and incomplete data. Demographic, substance use profiles, and clinical data were extracted from existing records. Serum total testosterone levels were measured using a chemiluminescence immunoassay technique. Descriptive statistics, non-parametric between-group comparisons, Spearman correlations, and multivariable linear regression were used to analyze the data, with statistical significance set at p < 0.05.

Results: relatively higher median testosterone concentrations were observed in heroin-injecting men (3.570 nmol/L) compared to non-heroin-injecting men (0.097 nmol/L) (p < 0.0001). Hypogonadism, defined as testosterone below 8.0 nmol/L, was present in 97.9% of heroin-injecting men and in all non-heroin-injecting men (100.0%), with no significant difference between the two groups (p = 0.712). Heroin-injecting men working in small businesses and transport sectors had relatively higher testosterone levels compared to those working in the hospitality sector (p = 0.001). Hip circumference was positively correlated with testosterone levels in the heroin-injecting men, with a significant positive correlation (Spearman's ρ = 0.173, p = 0.039). Similarly, waist-to-hip ratio showed a significant negative correlation in the heroin-injecting men (Spearman's ρ = -0.204, p = 0.014). On multivariable linear regression, employment in the small business sector (B = 3.147, p < 0.001) and transport sector (B = 3.103, p < 0.001), duration of heroin injection of one year or more (B= 0.802, p = 0.012), and waist-to-hip ratio (B= -6.146, p = 0.006) were independent predictors of serum total testosterone concentrations (Adjusted R2 = 0.201).

Conclusion: this study reports that serum total testosterone concentrations were significantly higher in heroin-injecting men than in non-heroin-injecting men, though both groups fell below the clinical threshold for hypogonadism. Multivariable regression analysis identified employment in the small business and transport sectors, prolonged heroin injection of at least one year, and waist-to-hip ratio as independent predictors of serum total testosterone concentrations. Cigarette smoking was independently associated with lower testosterone after adjustment, though this association warrants cautious interpretation. These findings emphasize the need for integrated interventions that address both the endocrine consequences of heroin injection and the occupational and lifestyle factors that modulate testosterone in this population.