Frontiers in Public Health · Nov 18, 2026

Self-Reported Aerobic and Resistance-Exercise Patterns and Metabolic Health in Adults Attending a Chinese Health-Examination Center: A Cross-Sectional Analysis

By Yue Shi, Guangdong Cheng, Cai Wang

Background: Aerobic and muscle-strengthening activities are complementary guideline components, but the metabolic profile associated with self-reported exercise patterns in single-center routine-examination samples remains uncertain. We compared aerobic-only, resistance-frequency-only, and combined exercise categories using continuous and multidimensional measures of metabolic health. Methods: In this single-center cross-sectional study, 753 of 3,411 screened health-examination participants in 2026 fulfilled the eligibility, questionnaire and complete-data requirements. Exercise was characterized using a self-designed, nonvalidated questionnaire asking about a typical week during the preceding three months. Aerobic attainment was operationalized as at least 150 equivalent min/week; resistance attainment was a frequency-only proxy (at least two reported days/week), not verified full guideline compliance. The primary outcome was sex-standardized cMetS. Multivariable HC3-robust regression, restricted cubic splines, exploratory B-spline response surfaces, secondary-outcome FDR analyses and sensitivity analyses were conducted. Results: Of 753 adults, 446 (59.23%) were in the neither group, 155 (20.58%) aerobic only, 92 (12.22%) resistance-frequency only and 60 (7.97%) combined. Compared with neither, fully adjusted cMetS differences were -0.19 (95% CI, -0.30 to -0.08), -0.18 (-0.33 to -0.03), and -0.33 (- 0.48 to -0.17), respectively (all P≤0.02). Direct combined-versus-single-mode contrasts were nonsignificant after Holm correction. The prespecified three-knot aerobic spline showed an overall association (Wald P=0.0078) but not nonlinearity (P=0.69); the original reported overall P=0.21 was not reproduced. The revised 12-degree-of-freedom surface interaction was nonsignificant (P=0.61). Among 30 exploratory secondary contrasts, FDR adjustment attenuated several nominal results. Conclusions: In this single-center health-examination sample, self-reported aerobic exercise and a frequency-only measure of resistance exercise were each associated with lower cMetS than the neither category. A numerically lower cMetS estimate in the combined group did not establish an incremental benefit over either individual category. Cross-sectional measurement, nonvalidated exposure assessment, residual confounding and small combined-group numbers preclude causal or generalized exercise-prescription claims.

The paperJournal sitePopulations

AgingResearch.News