Risk of fracture according to temporal changes of low body weight changes in adults over 40 years: a nationwide population-based cohort study

BACKGROUND: Low body weight is associated with an increased risk of fractures. However, the effect of temporal changes in the low body weight status on the risk of fracture remains unknown. This study aimed to evaluate the relationships between temporal changes in low body weight status and the risk of fractures in adults over the age of 40 years. METHODS: This study included data on adults over 40 years old who underwent two biannual consecutive general health examinations between January 1, 2007 and December 31, 2009 extracted from the National Health Insurance Database, a large nationwide population database. Fracture cases in this cohort were monitored from the time of the last health examination to the end of the designated follow-up period (from January 1, 2010 to December 31, 2018) or the participant's death. Fractures were defined as any fracture resulting in hospitalization or outpatient treatment claim after the date of general health screening. The study population was then separated into four groups based on the temporal changes in low body weight status as follows: low body weight to low body weight (L-to-L), low body weight to non-low body weight (L-to-N), non-low body weight to low body weight (N-to-L), and non-low body weight to non-low body weight (N-to-N). The hazard ratios (HRs) for new fractures, depending on weight changes over time, were calculated using Cox proportional hazard analysis. RESULTS: Adults in the L-to-L, N-to-L, and L-to-N groups had a substantially increased risk of fractures after multivariate adjustment (HR, 1.165; 95% confidence interval [CI], 1.113–1.218; HR, 1.193; 95% CI, 1.131–1.259; and HR, 1.114; 95% CI, 1.050–1.183, respectively). Although the adjusted HR was greater in participants who changed into having a low body weight, followed by those with consistently low body weight, those with low body weight remained to have an elevated risk of fracture independent of weight fluctuation. Elderly men (aged over 65 years), high blood pressure, and chronic kidney disease were significantly associated with an increase in fractures (p < 0.05). CONCLUSION: Individuals aged over 40 years with low body weight, even after regaining normal weight, had an increased risk of fracture. Moreover, having a low body weight after having a normal body weight increased the risk of fractures the most, followed by those with consistently low body weight. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12889-023-15940-0.