STUDY OBJECTIVES: Insomnia with objective short sleep duration has been previously associated with adverse cardiometabolic health outcomes as well as poorer cognitive performance in otherwise noncognitively impaired adults. However, studies demonstrating an increased prevalence of cognitive impairment (CI) in this insomnia phenotype are lacking. METHODS: We analyzed data from Penn State Adult Cohort (N = 1,524; 48.9 ± 13.4 years; 53.4% women). Self-reported sleep difficulty was defined as normal sleep (n = 899), poor sleep (n = 453), and chronic insomnia (n = 172). Objective short sleep duration was defined as less than 6-h of sleep, based on in-lab, 8-h polysomnography. CI (n = 155) and possible vascular cognitive impairment (pVCI, n = 122) were ascertained using a comprehensive neuropsychological battery. Analyses adjusted for age, sex, race, education, body mass index, apnea/hypopnea index, smoking, alcohol, psychoactive medication, and mental and physical health problems. RESULTS: Participants who reported poor sleep or chronic insomnia and slept objectively less than 6 hours were associated with a 2-fold increased odds of CI (OR = 2.06, 95% confidence limits [CL] = 1.15-3.66 and OR = 2.18, 95% CL = 1.07-4.47, respectively) and of pVCI (OR = 1.94, 95% CL = 1.01-3.75 and OR = 2.33, 95% CL = 1.07-5.06, respectively). Participants who reported poor sleep or chronic insomnia and slept objectively more than 6 hours were not associated with increased odds of either CI (OR = 0.72, 95% CL = 0.30-1.76 and OR = 0.75, 95% CL = 0.21-2.71, respectively) or pVCI (OR = 1.08, 95% CL = 0.42-2.74 and OR = 0.76, 95% CL = 0.16-3.57, respectively). CONCLUSIONS: Insomnia with objective short sleep duration is associated with an increased prevalence of CI, particularly as it relates to cardiometabolic health (i.e. pVCI). These data further support that this insomnia phenotype may be a more biologically severe form of the disorder associated with cardiovascular, cerebrovascular, and neurocognitive morbidity.
Stored only when both papers are in the corpus — 66 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.