Heart failure is associated with obesity, but the relationship between weight change and heart failure is inconsistent. We examined the relationship between weight change and the incidence of heart failure in the Korean population.


We used the National Health Insurance System health checkup data from 2005 to 2015. A total of 11,210,394 subjects (6,198,542 men and 5,011,852 women) greater than 20 years of age were enrolled in this study. Weight change over 4 years was divided into seven categories from weight loss ≥ 15% to weight gain ≥ 15%. The hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of HF were analysed using multivariable Cox’s proportional hazard models.


The HR for HF shows a reverse J-shaped curve and > 15% weight loss shows the highest HR (HR and 95% CI = 1.647;1.546-1.735) followed by -15 to -10% weight loss (HR = 1.444 and 1.408, respectively). In underweight subjects, HR decreased as weight increased. HR showed a reverse J-shaped curve in normal and overweight subjects, a U-shaped curve in obesity stage I, and a J-shaped curve in obesity stage II. Weight gain ≥ 15% in obesity stage II showed the highest HR (HR; 95% CI = 2.97; 1.79-4.928). Stable weight for 4 years in the underweight and obesity stage I and II increased the incidence of HF (HR = 1.402, 1.092, and 1.566, respectively).


Both weight loss and weight gain increased HR for heart failure. Sustained weight in the obesity or underweight categories increased the incidence of heart failure.