Abstract
BACKGROUND: Both weight loss and cognitive impairment are common in late-life, but it remains unknown whether weight change is associated with risk of incident dementia among elderly Japanese. Our study aimed to investigate the association between long-term weight change since midlife and risk of incident disabling dementia using a community-based cohort study of elderly Japanese. METHODS: In 2006, we conducted a cohort study of 6,672 disability-free Japanese adults aged ≥65 years. In both 1994 and 2006, the participants reported their weight using a self-reported questionnaire. Based on weight obtained at these two time points, participants were classified into: stable weight (-1.4 to +1.4 kg), weight gain (≥+1.5 kg), and weight loss of -2.4 to -1.5 kg, -3.4 to -2.5 kg, -4.4 to -3.5 kg, -5.4 to -4.5 kg, and ≥-5.5 kg. Incident disabling dementia was retrieved from the public Long-term Care Insurance database. Participants were followed-up for 5.7 years (between April 2007 and November 2012). Cox proportional hazards model was used to estimate multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for incident disabling dementia. RESULTS: During 32,865 person-years of follow-up, 564 participants were ascertained as having incident disabling dementia. Compared with stable weight, the multivariable-adjusted HRs were 0.97 (95% CI, 0.70-1.34) for weight loss of -2.4 to -1.5 kg, 0.98 (95% CI, 0.70-1.38) for -3.4 to -2.5 kg, 1.28 (95% CI, 0.91-1.81) for -4.4 to -3.5 kg, 1.27 (95% CI, 0.92-1.77) for -5.4 to -4.5 kg, and 1.64 (95% CI, 1.29-2.09) for ≥-5.5 kg. CONCLUSION: Our study suggested that a ≥-3.5 kg weight loss over 12 years might be associated with higher risk of incident disabling dementia among elderly Japanese.