Lower nutritional indicators associated with higher mortality in patients with MCI and dementia—the Amsterdam Ageing Cohort study

Background Our objective was to investigate associations between indicators of nutritional status, including weight and body composition, and survival across cognitive diagnosis groups within a geriatric population. Method From the ongoing Amsterdam Ageing cohort study, 719 patients were included, comprising 111 with subjective cognitive decline (SCD; 76±7 years, 54.1% female, MMSE 28[27‐29]), 213 with mild cognitive impairment (MCI; 79±6 years, 44.6% female, MMSE 26[24‐27]) and 395 with dementia (81±7 years, 53.9% female, MMSE 22[18‐24]). Indicators of nutritional status were measured and dichotomized, including body mass index (BMI), fat percentage (Fat%), skeletal muscle mass index (SMI) and Mini Nutritional Assessment Short‐Form (MNA‐SF) scores. Survival data were requested from the municipality population registry on January 22nd 2020. Associations between indicators of nutritional status and survival among different cognitive diagnosis groups were analyzed using Kaplan Meier analyses, and cox regression models adjusted for age, sex and number of prescribed drugs. Result The overall mortality was 15.8% during a median follow‐up duration of 610 [256‐964] days, with a significant difference between the cognitive diagnosis groups (p<0.0001). Kaplan Meier survival curves stratified for BMI and MNA are presented in Figure 1 and 2. Low BMI (<20 kg/m 2 ) was associated with higher mortality risk in patients with dementia in the adjusted cox regression model (HR 3.1, 95%‐CI 1.7‐5.9). In patients with MCI, the association with BMI was not significant (HR 3.0, 0.6‐14.4). Low MNA‐SF scores (<12) were associated with higher mortality in both patients with MCI (HR 3.55, 1.23‐10.18) and patients with dementia (HR 2.1, 1.3‐3.5). However, there were no significant associations with Fat% and SMI. In patients with SCD, there were no significant associations between the indicators of nutritional status and mortality. Conclusion Lower indicators of nutritional status were associated with higher mortality in patients with MCI and dementia, warranting increased focus on nutritional status in patients with cognitive decline. We plan to repeat these analyses after a longer follow‐up in this ongoing cohort study. Dietary intervention studies aimed at preventing weight loss could shed light on causality in the association between nutritional status and progression of disease.