Objective: To determine the longitudinal relationship between a reported history of weight cycling and the trajectory of weight in a group of normal-weight to obese women and men.
Research methods and procedures: This was a clinic-based cohort study conducted over 6 years. Subjects were healthy women (n = 141) and men (n = 797) age 20 to 78 years who completed at least four comprehensive medical exams at the Cooper Clinic (Dallas, TX) between 1987 and 2003. Weight loss history was reported, and body weight was measured at all examinations. Weight cycling status was derived from weight loss history and defined as > or = five episodes of weight loss of > or = 2.3 kg/episode. Using linear mixed effects models, weight at each examination was regressed on weight cycling status separately for women and men while controlling for selected covariables.
Results: Baseline BMI was 23 and 21 kg/m2 among cycling and non-cycling women and was 27 and 25 kg/m2 among cycling and non-cycling men, respectively. We observed a non-significant difference (p = 0.09) in women by cycling status (average weight gain = 0.5 and 0.2 kg/yr among cyclers and non-cyclers, respectively) and no difference in weight gain over time among men by cycling status (average weight gain = 0.2 kg/yr among both groups, p = 0.99). Higher baseline cardiorespiratory fitness level and increased fitness over the follow-up were associated with attenuated weight gain in both women and men.
Discussion: In healthy, middle-aged individuals, a history of weight cycling does not seem to increase the risk of long-term weight gain in men; however, this relation needs to be studied further in women.