Extended Data Fig. 6: Median eating window time (hours; A), eating window length (hours; B), and percentage of adherent days (%; C) for men among the UC, early TRE, late TRE, and self–selected TRE groups over the 12–week intervention period.

Panel A represents the median eating windows for men across the usual care (UC), early time–restricted eating (TRE), late TRE, and self–selected TRE groups, with the intensity of colors representing adherence to the protocol. Black dotted rectangles indicate group–median eating windows. Panel B shows the duration of the eating window for each intervention group (violin plot), with dots representing the distribution of individual data. Panel C illustrates the percentage of adherent days per group over the 12–week intervention period (adherence defined as an eating window ≥12 hours for usual care and ≤9 hours for TRE groups), with dots representing the distribution of individual data. Data in the Panel B are median, interquartile range, and 1.5x interquartile range. Data in panel C are means with 95% confidence intervals. Statistical analysis applied was analysis of variance with post–hoc Bonferroni analysis for significant differences between groups (P < 0.05). The P values reported are two-sided and Bonferroni-adjusted P values. Groups sharing similar letters denote significant differences. This figure includes only participants with eating window records for more than 60% of the intervention days.