Association between weight change over a period of 4 years and mortality in Korean older adults

Aim This study aimed to evaluate the association between weight change and mortality in Korean older adults. Methods We collected data pertaining to National Screening Program participants aged ≥65 years from the Korean National Health Insurance Corporation records during 2005–2011. To this data, we included mortality data, such as the date of death, of these patients (up to and including 2017) from Statistics Korea. We defined weight change as a difference in bodyweight measured in the National Screening Program over a period of 4 years. Cox proportional hazards models were used to evaluate the association between weight change and mortality. Results The study population consisted of 1 100 256 participants, and a total of 46 415 deaths were observed during a mean follow‐up period of 3.2 ± 0.8 years (maximum 5.0 years). For 3 531 585 person‐year follow up, the mortality rate for stable weight was 10.79 per 1000 person‐years (PY). Weight loss increased the mortality rate by 68%, whereas weight gain increased the rate by 10% compared with stable weight (weight loss: mortality rate 20.28 per 1000 PY, adjusted hazard ratio 1.68, 95% confidence interval 1.65–1.72; weight gain: mortality rate 12.86 per 1000 PY, adjusted hazard ratio 1.10, 95% confidence interval 1.07–1.13). However, in subgroup analysis, for participants who were underweight at baseline, current cigarette smokers or heavy alcohol drinkers, weight gain did not increase the mortality rate. Conclusions In Korean older adults, regardless of the risk factors, both weight loss and weight gain increased the mortality rate compared with stable weight. Geriatr Gerontol Int 2020; 20: 474–481.