Abstract
Sedentary behaviors have been associated with the risk of dementia in older adults. Whether driving and computer use are associated with the risk of dementia in older adults is an important research question. The participants of a longitudinal cohort study that included European middle- and old-aged adults at the baseline (2006-2010) who had not been diagnosed with dementia before 5 years after the baseline and had not died within 5 years after the baseline were followed up (until 2018) and analyzed. The associations between driving and non-occupational computer use time measured by the questionnaire at the baseline and incident dementia 5 years after the baseline were analyzed after correcting for confounding variables. Each analysis included approximately 370,000 participants and 1,000 cases. According to Cox proportional hazard models that divide subjects into four groups of habit duration levels [(a) 0 h; (b) less than 1 h, 1 h; (c), 2 h, 3 h; (d) 4 h or more, per/day)], the group with 0 h < driving time ≤1 h at the baseline exhibited a significantly lower risk of incident dementia than the other groups. In addition, in the analysis of non-occupational computer use duration, the 0 h group exhibited a significantly higher risk than the other groups. Our results indicate that different sedentary behaviors have different associations with dementia risk over time and have no simple dose-response relationship with dementia risk. The sedentary behavior risk assessments must consider these factors.</p>