Longitudinal association between body weight change and brachial-ankle pulse wave velocity in patients with treated hypertension or diabetes mellitus: A post hoc analysis of the FMD-J study B.

We investigated cross-sectional associations between body mass index (BMI) and brachial-ankle pulse wave velocity (baPWV) and longitudinal associations between body weight change and baPWV over 3 years in 874 patients with treated hypertension or diabetes mellitus without coronary artery disease (475 men; mean age, 61.9 ± 9.5 years). baPWV was measured at baseline, 1.5 years, and 3 years. At baseline, BMI was inversely correlated with baPWV (r = -0.14, P < 0.001), and multivariate analysis confirmed this association (β = -12.5, P < 0.001). In generalized linear mixed-effects models adjusted for concurrent systolic blood pressure, heart rate, and other covariates, neither continuous percentage change in body weight (β = -1.2, P = 0.08) nor categorical body weight change (≥5% increase: β = -10.6, P = 0.45; ≥5% decrease: β = 18.0, P = 0.15) was associated with longitudinal changes in baPWV, with no interaction across baseline BMI categories. The modest body weight changes observed in this cohort were not significantly associated with longitudinal changes in baPWV.