Predictors of weight regain after weight loss in obesity: roles of birthweight, diet, and metabolic factors

Weight regain after weight loss is common in obesity. However, risk factors for weight regain remain uncertain, particularly with respect to developmental and dietary factors. This study aims to identify clinical, metabolic, developmental, and dietary predictors of weight regain. We analysed data from the Japan Obesity and Metabolic Syndrome Study (JOMS), a multicentre prospective cohort of adults with body mass index (BMI) ≥ 25 kg/m 2 . Among 729 participants, 192 individuals (108 women, mean age 49.9 years; BMI 31.7 kg/m 2 ) who achieved > 3% weight loss at 3 months and had 6-month follow-up data were included. Weight regain between 3 and 6 months was observed in 61 participants (30 women) (regain group), while 131 participants (78 women) maintained or continued to lose weight (no-regain group). Longitudinal assessment of 49 anthropometric and clinical laboratory parameters was conducted for nearly all participants. Fifty-four participants (34 women) reported their exact birthweights. One hundred and thirty-one participants (78 women) responded to a 20-item dietary habit questionnaire, and the similar number of participants responded to a 12-item lifestyle questionnaire. The regain group showed significantly lower birthweight in women (Cohen’s d = 0.879, p = 0.026) and less frequent meat consumption in men (odds ratio (OR) = 0.194, p = 0.021) than the no-regain group. At baseline, higher HbA1c was observed in men who subsequently regained weight ( d = 0.737, p = 0.010). Two-way repeated measures mixed-effects model analysis showed that higher total cholesterol (T-Chol) levels in men (post-hoc test, d = 0.594, p = 0.017) and women ( d = 0.639, p = 0.018), higher ankle–brachial index (ABI) in women ( d = 0.586, p = 0.019), and higher HbA1c in men ( d = 0.603, p = 0.037) at 3 months were associated with subsequent weight regain. In sex-stratified multivariable logistic regression models adjusted for age, baseline BMI, percentage weight change from baseline to 3 months, ABI in women, HbA1c in men, and T-Chol in both sexes, only higher baseline HbA1c in men (per 1SD; OR = 2.4, p = 0.005) and higher 3-month T-Chol in men (per 1SD; OR = 1.9, p = 0.020) and women (per 1SD; OR = 2.5, p = 0.012) were independently associated subsequent weight regain. Developmental factors (low birthweight), dietary patterns (low meat intake), and metabolic markers (higher HbA1c and total cholesterol) were associated with weight regain after weight loss. These findings support a risk-stratification approach to personalised obesity management. Further studies with larger sample sizes and validated dietary assessments are needed to confirm these results. Intervention studies to determine whether personalised treatments targeting these factors would improve outcomes are warranted.