Latent classes of sleep problems and subjective cognitive decline among middle-aged and older adults in the United States

Previous studies have linked sleep problems to subjective cognitive decline (SCD) using a variable-centered approach (e.g., adding sleep symptoms to form a score); however, sleep problems may cluster differently between individuals. Thus, employing a person-centered approach, we aimed to: 1) identify profiles of self-reported sleep problems among U.S. middle-aged and older adults; 2) examine the cross-sectional association between these classes and SCD.

We studied 33,922 adults aged 45+ years from the 2017 U.S. Behavioral Risk Factor Surveillance System (BRFSS) with data on sleep problems, including short or long sleep duration, trouble falling or staying asleep, unintentionally falling asleep, snoring loudly, and observed apneas. Latent class analysis classified participants based on their responses to sleep items. We then used a subsample from Oregon, the only state that administered both sleep and SCD modules (n = 2,747), to examine the association between class membership and SCD using logistic regression, adjusting for sociodemographic and health-related characteristics.

We identified and labeled four classes: "Healthy Sleep" (48.0 %); "Primarily Apnea" (25.8 %); "Primarily Insomnia" (17.6 %); and "Comorbid Insomnia and Sleep Apnea (COMISA)" (8.6 %). In adjusted models, individuals in the "COMISA" class had almost twice the odds of SCD, compared to those in the "Healthy Sleep" class (OR=1.91, 95 % CI =1.15-3.15).

Compared to U.S. middle-aged and older adults with healthy sleep, those with COMISA were significantly more likely to report SCD, which is a risk factor for dementia. Studies are needed investigating whether sleep interventions delay cognitive decline in these individuals.