Dietary amino acid intake and sleep duration are additively involved in future cognitive decline in Japanese adults aged 60 years or over: a community-based longitudinal study

BACKGROUND: Sleep duration and amino acid intake are independently associated with cognitive decline. This study aimed to determine the longitudinal association between sleep duration and cognitive impairment incidence and to examine the involvement of diet, particularly amino acid intake, in these associations in community dwellers. METHODS: In this longitudinal study in a community-based setting, we analyzed data from 623 adults aged 60–83 years without cognitive impairment at baseline. Sleep duration was assessed using a self-report questionnaire. Amino acid intake was assessed using 3-day dietary records. Cognitive impairment was defined as a Mini-Mental State Examination score ≤ 27. Participants were classified into short-, moderate-, and long-sleep groups according to baseline sleep duration (≤ 6, 7–8, and > 8 h, respectively). Using moderate sleep as a reference, odds ratios (ORs) and 95% confidence intervals (CIs) of short- and long-sleep for cognitive-impairment incidence were estimated using the generalized estimating equation. Participants were classified according to sex-stratified quartiles (Q) of 19 amino acid intake: Q1 and Q2–Q4 were low- and middle to high-intake groups, respectively. Using middle- to high-intake as a reference, ORs and 95% CIs of low intake for cognitive impairment incidence were estimated using the generalized estimating equation in each sleep-duration group. Follow-up period, sex, age, body mass index, depressive symptoms, education, smoking status, employment status, sleep aids use, physical activity, medical history, and Mini-Mental State Examination score at baseline were covariates. RESULTS: Mean follow-up period was 6.9 ± 2.1 years. Adjusted ORs (95% CIs) for cognitive impairment in short- and long-sleep groups were 0.81 (0.49–1.35, P = 0.423) and 1.41 (1.05–1.87, P = 0.020), respectively. Particularly in long sleepers (i.e., > 8 h), cognitive impairment was significantly associated with low cystine, proline, and serine intake [adjusted ORs (95% CIs) for cognitive impairment were 2.17 (1.15–4.11, P = 0.017), 1.86 (1.07–3.23, P = 0.027), and 2.21 (1.14–4.29, P = 0.019), respectively]. CONCLUSIONS: Community-dwelling adults aged ≥ 60 years who sleep longer are more likely to have cognitive decline, and attention should be paid to the low cystine, proline, and serine intake. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12877-023-04359-2.