Taylor AM, Edwards CG, Khan NA, Holscher HD (2017) FASEB 31 no. 1 Supplement 965.34
Web URL: Read the full article on FASEB here
Background A converging body of literature demonstrates that the gastrointestinal microbiota are implicated in multiple aspects of health including cognitive function and brain health. However, this evidence is largely based on rodent models and there is a paucity of data linking characteristics of the gastrointestinal microbiota profiles and cognitive function in humans.
Objective We aimed to determine the relationship between the human gastrointestinal microbiota and cognition among adult females.
Methods Cross-sectional analyses were conducted to determine relationships between cognitive control and gastrointestinal microbiota of adult 25–45-year-old females (n=34, BMI=32.2 kg/m2). Executive function, specifically, interference control, was assessed using a modified flanker task. Interference scores were calculated for accuracy (congruent – incongruent) and reaction time (incongruent – congruent) with greater interference signifying poor ability to maintain task performance when faced with greater cognitive demand. Gastrointestinal microbiota taxonomy was obtained using a fecal sample. Following fecal DNA isolation, a Fluidigm Access Array was utilized to generate barcoded amplicon pools of archaeal, bacterial, and fungal sequences. High-throughput sequencing was conducted on a MiSeq using version 3 chemistry. Sequence data were analyzed using QIIME 1.9.0 and SAS 9.4. Whole body adiposity (% Fat) was assessed using DXA. Covariates assessed included habitual dietary fiber intake (National Cancer Institute’s Diet History Questionnaire) and IQ (Kaufman Brief Intelligence Test).
Results According to bivariate correlations, the Bacteroidetes:Firmicutes ratio was not related to %Fat (r=0.15, p=0.38) or dietary fiber (r=−0.11, p=0.53). Bacteroidetes and Firmicutes were differentially associated with interference control measures. Accuracy interference was inversely associated with Bacteroidetes (r=−0.37, p=0.03) and the Bacteroidetes:Firmicutes ratio (r=−0.39, p=0.02). Reaction time interference was also negatively correlated with Bacteroidetes (r=−0.37, p=0.02) and the Bacteroidetes:Firmicutes ratio (r=−0.34, p=0.047). Further, accuracy interference was correlated with Firmicutes (r=0.34, p=0.043). According to partial correlations, the relationship between accuracy interference and microbiota was mediated by IQ and dietary fiber/kcal (all p’s>0.07). However, the relationship between reaction interference and Bacteroidetes (r=−0.35, p=0.048 ) persisted even after adjusting for covariates.
Conclusion These results indicate that women with greater relative abundances of Bacteriodetes exhibited greater ability to maintain cognitive performance when faced with greater task demands. These findings are among the first to relate bacterial phylogenetic characteristics to executive function among adult humans. Further study is required to elucidate a causal relationship between dietary manipulation of microbiota composition and changes in selective aspects of cognitive performance.