Frequency of daily conversation and risk of dementia: A population-based cohort study in Japan

OBJECTIVES: This population-based study evaluated the association between frequency of daily conversation and risk of dementia, with consideration to living arrangement (living alone or living with someone) and sex. METHODS: Participants aged 50 to 79 years who reported their frequency of daily conversation in 2000-2003 within the Japan Public Health Center-based Prospective Study were followed from 2006 to 2016 for incident dementia using certification records for national long-term care insurance in Japan. Hazard ratio (HR) and 95 % confidence interval (CI) were estimated using a Cox proportional hazards model with adjustment for factors such as lifestyle and medical history. Subgroup analyses were performed by living arrangement and sex. RESULTS: Among 35,488 participants, 3334 were diagnosed with dementia. Fully adjusted HRs compared with a conversation frequency of almost every day were 0.80 (95 % CI: 0.69-0.93) for every day with many people, 0.88 (0.80-0.97) for every day with several people, 1.18 (1.06-1.31) for 1-4 times/week, 1.17 (0.97-1.42) for 1-3 times/month, and 2.06 (1.49-2.85) for <once/month (P for trend <0.001). The association was not modified by living arrangement. Men living alone showed a high HR, at 1.71 (1.26-2.32) for high conversation frequency (almost every day or more) and 2.60 (1.71-3.95) for low conversation frequency (1-4 times/week or less), while women living alone did not show a high risk. CONCLUSIONS: Lower conversation frequency was associated with higher risk of dementia. Living alone was associated with a higher risk in men, but not in women.