Longitudinal body weight dynamics in relation to cognitive decline over 18 years: A prospective cohort study

Abstract Background Weight loss and body weight variability have been related to cognitive decline in old adults, respectively. However, the long‐term and independent associations of the two measures of body weight dynamics with cognitive decline remain unclear. Method In the Health and Retirement Study (HRS), we included 10,258 participants (58.4% women; mean age ± standard deviation (SD), 68.0 ± 9.8 years) free from memory‐related diseases in 1998. Body weight was reported biennially from 1993/1994 to 2016. Linear regression of the longitudinal body mass index for each individual was used to estimate body weight change (BWC, slope, kg/m 2 ·year) and body weight variability (BWV, root mean squared error, kg/m 2 ). Both measures were categorized by tertiles, with participants being classified as stable weight, weight loss, and weight gain (median = 0.0, ‐0.3, and 0.2 kg/m 2 ·year, respectively), and small, moderate, and large BWV (median = 0.5, 1.1, and 2.0 kg/m 2 , respectively). Cognitive function was measured biennially from 1998 to 2016 using the Telephone Interview for Cognitive Status (TICS). We investigated the associations of BWC and BWV with annual rate of cognitive decline using linear mixed models. Result During a median (IQR) follow‐up of 12 (4‐18) years, the average rate of cognitive decline was ‐0.052 z‐score/year. When adjusted for basic demographics, average BMI, and major lifestyle factors, we observed no significant associations of weight gain and moderate BWV with cognitive decline, whereas both weight loss and large BWV were associated with faster cognitive decline in the same model. Compared to those with stable weight, participants with weight loss experienced faster cognitive decline by ‐0.012 (95%CI, ‐0.015, ‐0.009) z‐score/year. Large BWV was also related to faster cognitive decline (β = ‐0.003, 95% CI: ‐0.006, ‐0.001) compared with small BWV. The associations were similar for both episodic memory and working memory, and across prespecified study subgroups. Conclusion Weight loss and large body weight variability were associated with faster cognitive decline in old adults. Our findings suggest the importance of dynamic body weight monitoring and management as a potential strategy to prevent age‐related cognitive decline.