Body mass index change patterns and risk of hospitalization among populations at high cardiovascular risk: findings from the China PEACE Million Persons Project

Background: Evidence on the impact of dynamic body mass index (BMI) changes on clinical outcomes in high-risk Asian populations is limited. This study aimed to identify BMI change patterns using a data-driven approach and prospectively examine their association with hospitalization risk in individuals at high cardiovascular disease (CVD) risk. Methods: This study included 20,329 high-risk adults from the Anhui region of the China PEACE Million Persons Project (2017–2023). High CVD risk was defined as history of CVD, hypertension, dyslipidemia, or a 10-year CVD risk ≥ 20%. BMI change patterns were identified using K-means clustering based on baseline BMI and annual change rate. Primary outcomes were all-cause and CVD hospitalization. Multivariable Cox models assessed associations. Results: Four change patterns were identified: normal weight-stable (50.9%), rapid gainers (4.4%; normal baseline BMI, median annual increase + 1.80 kg/m²/year), obese-slow losers (39.5%; persistent obesity, minimal loss, median − 0.20 kg/m²/year), and obese-rapid losers (4.9%; highest baseline BMI, rapid loss, median − 2.20 kg/m²/year). Over median 3.6 years, 2,272 all-cause and 617 CVD hospitalizations occurred. Compared with normal weight-stable, rapid gainers (adjusted hazard ratio [aHR] = 1.44, 95%CI 1.10–1.88), obese-slow losers (aHR = 1.17, 95%CI 1.07–1.28), and obese-rapid losers (aHR = 1.39, 95%CI 1.09–1.78) had higher all-cause hospitalization risk. CVD hospitalization risk was significantly elevated in rapid gainers (aHR = 1.86, 95%CI 1.16–2.96) and obese-slow losers (aHR = 1.26, 95%CI 1.06–1.50), was of borderline significance in obese-rapid losers (aHR = 1.57, 95%CI 0.97–2.54). The obese-rapid losers group had higher socioeconomic status but greater comorbidity burden. Conclusions: In a Chinese population at high CVD risk, four heterogeneous BMI change patterns exhibited differential hospitalization risks. Notably, the obese-rapid weight loss pattern displayed an exploratory outcome-specific profile, characterized by elevated all-cause hospitalization risk and a borderline association with CVD hospitalization. These findings suggest that rapid weight loss in obese individuals is not uniformly beneficial, underscoring the need to evaluate the nature of weight loss beyond its magnitude.