Presented by: Chaoran Ma
View Abstract
Background: Abnormal gastrointestinal motility and perturbations in the gut microbiome may contribute to cognitive impairment and dementia. We examined the associations of bowel movement (BM) frequency with objective and subjective cognitive function in 139,978 women and men from Nurses’ Health Study (NHS), Nurses’ Health Study II (NHSII) and Health Professionals Follow-Up Study (HPFS), and explored the mediating role of the gut microbiome in a subsample.
Methods: Participants in all three cohorts reported their BM frequency in 2012/2013 and subsequently (2014-2017) self-reported subjective cognitive dysfunction. In NHSII, 14,586 participants completed the CogState neuropsychological battery for objective cognitive assessment annually 2014-2018. We profiled the gut microbiome using shotgun metagenomics in fecal samples collected from a subpopulation of 515 women and men.
Results: The average age was 76 years in NHS, 61 in NHSII, and 75 in HPFS at the first cognitive assessment. After adjustment for potential confounders, BM frequency was associated with global objective cognitive function and its subdomain,learning and working memory, in an inverse U-shape dose-response manner (both P quadratic=0.02). Compared to those with daily BM, participants in the highest (>twice/day) and lowest categories (every 3+ days) of BM frequency had 0.03 and 0.06 standard units (equivalent to 1-2 additional years of age) lower in global cognition, respectively. We found similar inverse U-shape dose-response associations of BM frequency with overall subjective cognitive function and its subdomains (all P quadratic <0.001). Compared to daily BM, the odds ratios (95% CI) of cognitive decline were 1.18 (1.08, 1.30) for every 3+ days, 1.08 (1.01, 1.15) for every 2 days, and 1.10 (1.05, 1.15) for ≥ twice/day (P quadratic <0.001). The association between BM frequency and cognition was consistent across subgroups defined by age, laxative use, antibiotic use, hypertension, hypercholesterolemia, diabetes, smoking status, weight status, and APOE4 genotype. BM frequency and subjective cognition were each significantly associated with the overall variation of the gut microbiome (P <0.001 for BM frequency; P = 0.02 for cognition) and specific microbial species. Methanogenic bacteria, e.g., Methanobrevibacter smithii, were enriched in participants with less frequent bowel movements. Pro-inflammatory bacteria, e.g., Ruminococcus gnavus, were enriched at one of the extreme ends of BM frequency. Many microbial features, e.g., increased Veillonella parvula, were associated with both more frequent BM and worse cognition. Our mediation analysis suggested that the gut microbial taxonomy explained 81% of the association between BM frequency and cognition.
Conclusions: Abnormal bowel movement frequency was associated with worse cognitive function. The association may be mediated through differences in gut microbiome composition.