Effects of exercise and liraglutide on vascular health and inflammation during weight loss maintenance: a prespecified secondary analysis of the S-LiTE trial

Figures

• 

  Fig. 1. Effects of exercise (n = 26), liraglutide (n = 36) and the combination (exercise + liraglutide, n = 29) on vascular structure and circulating biomarkers during weight loss maintenance. a, Changes in cIMT in the four intervention groups during the 1 year intervention. b, Changes in inflammation markers compared with placebo (n = 39) during the 1 year intervention. c, Changes in endothelial function markers compared with placebo during the 1 year intervention. Data are presented as estimated treatment ratios (ETR) with 95% confidence intervals (CI), derived from a linear mixed model for repeated measurements, with sex assigned at birth, age (<40 vs ≥40 years), treatment, time (factorial) and a time × treatment interaction as covariates and an unstructured covariance pattern in the per-protocol population (n = 130). For cIMT, the analysis was adjusted for baseline scans. Analysis was performed on a log scale to meet model assumptions. All statistical tests were two-sided. No adjustment for multiple comparisons was applied. Source data

• 

  Extended Data Fig. 1. CONSORT study flow chart. CONSORT flow diagram of the participants enrolled in the S-LiTE trial. apersonal life conditions could be job-related, moving, or death in the family.

• 

  Extended Data Fig. 2. Changes in cIMT and biomarkers according to the 2×2 study design. Effects of exercise (exercise + placebo group and exercise + liraglutide group, n = 55) vs. no exercise (placebo + usual activity and liraglutide + usual activity group, n = 75) in Panel a and liraglutide (liraglutide + usual activity group and exercise + liraglutide, n = 65) vs no liraglutide (placebo + usual activity and exercise + placebo group, n = 65) in Panel b on vascular structure and circulating biomarkers during weight loss maintenance. Estimated treatment ratios (ETR) are derived from a linear mixed model for repeated measurements with sex assigned at birth, age ( < 40 vs ≥40 years), treatment (exercise vs. no exercise, and liraglutide vs. no liraglutide), time (factorial), and a time×treatment interaction as covariates and an unstructured covariance pattern in the per-protocol population (n = 130). Analysis was performed on log scale to meet model assumptions. Estimates are back-transformed and presented as estimated treatment ratios with 95% confidence intervals. For cIMT, the analysis was adjusted for baseline scans. All statistical tests were two-sided. No adjustment for multiple comparisons was…