Assessing Testosterone Epidemiology Using Distributional Approaches: Evidence From Large-Scale Longitudinal...

Testosterone is the main circulating androgen in adult males and serves a crucial role in reproductive function, skeletal health, metabolic regulation, and overall physiological stability. In recent decades, concern has emerged regarding possible secular changes in serum testosterone concentrations, independent of aging. However, findings across populations remain inconsistent, and most evidence originates from high-income settings.

This study aimed to evaluate long-term temporal patterns in serum testosterone levels among adult men, with particular emphasis on distributional shifts and population-wide hormonal instability, rather than solely on a mean decline.

A retrospective observational analysis was conducted using anonymized laboratory-based serum testosterone measurements from adult males aged ≥18 years between 2008 and 2025. Only the earliest measurement per individual was included to reduce repeat testing bias. Annual descriptive statistics were calculated, and temporal trends were assessed using linear regression. Age-stratified analysis, percentile drift evaluation, and reference interval shifts were used to explore changes across the testosterone distribution over time.

A total of 104,559 testosterone measurements were analysed over 17 years. Linear regression demonstrated a statistically significant but clinically minimal association between calendar year and testosterone concentration (β = +1.97 ng/dL/year, p < 0.001; R2 = 0.002). Substantial overlap in testosterone distribution was observedacross age groups, with notable inter year variability. Percentile drift analyses indicated parallel movement across the distribution, suggesting population-wide hormonal fluctuation rather than a uniform secular decline.

Mean testosterone levels did not demonstrate a consistent age-independent decline. Instead, findings support a temporal instability and persistent low exposure subgroups, emphasising the importance of distributional approaches in understanding contemporary hormonal epidemiology.