Chronotype and All‐Cause Mortality in <scp>US</scp> Middle‐Aged and Older Adults: Results From the <scp>NHANES</scp>

We examined the association between behavioural chronotype estimates and all-cause mortality in a nationally representative sample of middle-aged and older adults. We studied 2261 participants aged ≥ 50 years from the National Health and Nutrition Examination Survey (NHANES) who wore an ActiGraph on their hip for 7 days and were instructed to remove it for sleep. We estimated the average midpoint of time-in-bed (TIB) on weekends and categorised participants based on two sets of chronotype cutoffs. Chronotype I was defined as follows: intermediate-type I (≥ 3:30 AM and ≤ 4:30 AM), early-type I (≥ 11:00 PM and < 3:30 AM) and late-type I (> 4:30 AM and ≤ 11:00 AMm). In chronotype definition II, we expanded the intermediate category by 1 h on each side. We examined associations between chronotype and all-cause mortality in three age groups (50-65, n = 1055; 66-80, n = 895; 81+, n = 311) using survey-weighted Cox models. Six-hundred-fifty deaths occurred. Following adjustment for age, sex, race/ethnicity, poverty income ratio, BMI, smoking and drinking status, comorbidities and average TIB, among participants aged 81+, late-type I was associated with 123% greater mortality risk (HR = 2.23, 95% CI = 1.33, 3.74) versus intermediate-type I. Among participants aged 50-65, late-type II was associated with a 107% greater mortality risk (HR = 2.07, 95% CI = 1.09, 3.91) versus intermediate-type II. Findings suggest, in adults 50-65 and those aged 81+, that compared to individuals with intermediate chronotype, those with late chronotype have a higher risk of all-cause mortality; however, these associations may differ as a function of the temporal boundaries by which late chronotype is defined.