Weekday-weekend sleep patterning and testosterone outcomes in US men: a cross-sectional analysis of NHANES 2011-2014

Weekday-weekend differences in sleep duration are common, but their associations with testosterone and the specificity of corrected weekend catch-up sleep (CUS) as a sleep-pattern measure remain uncertain. We examined corrected CUS and related weekday-weekend sleep-pattern measures in relation to low and continuous total testosterone in US men.

This cross-sectional analysis included 3,042 men aged ≥ 20 years from NHANES 2011-2014. Corrected CUS was defined as max (mean weekend sleep duration - mean weekday sleep duration, 0). Survey-weighted logistic and linear regression models examined biochemically low total testosterone and continuous total testosterone, with multiple imputation for missing covariates.

Low total testosterone was present in 883 participants (weighted prevalence, 28.0%). Compared with CUS = 0, only the highest positive corrected-CUS category (> 1.346 h) was associated with low total testosterone (OR 1.37, 95% CI 1.06-1.77) and lower continuous testosterone (- 28.90 ng/dL, 95% CI - 46.96 to - 10.84). The categorical low-testosterone association was not retained in complete-case, strict valid-night, or upper-tail-excluded analyses, whereas the corresponding continuous-testosterone contrast remained evident in the strict and upper-tail-excluded cohorts. Directional weekday-weekend and sleep-irregularity analyses did not support an association specific to weekend sleep extension.

Corrected CUS showed modest cross-sectional associations with testosterone outcomes, but the broader pattern of results did not establish specificity to weekend sleep extension. Corrected CUS may instead reflect weekday-weekend sleep-pattern mismatch or broader sleep irregularity. These findings do not establish causality or support clinical use.