The longitudinal effect of sleep apnea on changes in brain structure in a population-based sample

STUDY OBJECTIVES: Increasing evidence points towards sleep apnea as a modifiable risk factor for neurodegeneration. This study investigates the longitudinal effect of sleep apnea on brain structures in a general population. METHODS: Data from the Study of Health in Pomerania with baseline laboratory-based polysomnography and magnetic resonance imaging (MRI) (2008-2012) and 7-year follow-up MRI (2016-2019) were used. Sleep apnea was characterized by the apnea-hypopnea index, oxygen desaturation index, mean sleep oxygen saturation, and arousal index; self-reported sleep quality using the Epworth Sleepiness Scale and Pittsburgh Sleep Quality Index. Total brain volume, regional and total gray matter thickness, white matter hyperintensity burden, and brain age were derived from 1.5 tesla T1 and fluid-attenuated inversion recovery sequences. Associations between sleep apnea and follow-up MRI measures were examined using linear regression adjusted for age, sex, intracranial volume, total sleep time, follow-up duration, and baseline MRI values. Restricted cubic splines were used in all continuous variables to model non-linear effects. RESULTS: 387 participants were included (age 50.7 ± 12.4 years, 47.3 per cent female). Neither apnea-hypopnea index nor oxygen desaturation index showed significant longitudinal effects. Lower mean sleep oxygen saturation was associated with reduced total brain volume, while a higher arousal index was linked to increase brain age and decreased global and regional gray matter thickness. Self-reported sleep quality correlated with total brain volume changes. CONCLUSIONS: Unlike prior cross-sectional findings, traditional sleep apnea indices were not associated with structural brain changes over 7 years. Instead, sleep fragmentation emerged as a key factor in brain atrophy. Targeting sleep fragmentation may offer neuroprotective benefits.