Snoring and risk of dementia: a prospective cohort and Mendelian randomization study

Abstract

Study objectives: The association between snoring, a very common condition that increases with age, and dementia risk is controversial. We aimed to investigate the observational and causal relationship between snoring and dementia, and to elucidate the role of body mass index (BMI).

Methods: Using data from 451 250 participants who were dementia-free at baseline, we examined the association between self-reported snoring and incident dementia using Cox proportional-hazards models. Causal relationship between snoring and Alzheimer's disease (AD) was examined using bidirectional two-sample Mendelian randomization (MR) analysis.

Results: During a median follow-up of 13.6 years, 8325 individuals developed dementia. Snoring was associated with a lower risk of all-cause dementia (hazard ratio [HR] 0.93; 95% confidence interval [CI] 0.89 to 0.98) and AD (HR 0.91; 95% CI 0.84 to 0.97). The association was slightly attenuated after adjusting for BMI, and was stronger in older individuals, APOE ε4 allele carriers, and during shorter follow-up periods. MR analyses suggested no causal effect of snoring on AD; however, genetic liability to AD was associated with a lower risk of snoring. Multivariable MR indicated that the effect of AD on snoring was primarily driven by BMI.

Conclusions: The phenotypic association between snoring and lower dementia risk likely stems from reverse causation, with genetic predisposition to AD associated with reduced snoring. This may be driven by weight loss in prodromal AD. Increased attention should be paid to reduced snoring and weight loss in older adults as potential early indicators of dementia risk.

Keywords: Alzheimer’s disease; Mendelian randomization; body mass index; dementia; multivariable Mendelian randomization; prodrome; sleep apnea; snoring; vascular dementia.