Associations Between Physical Activity and the Intestinal Microbiome of College Freshmen

Entry into college has been associated with weight gain and increased risk of obesity later in life. Modifiable risk factors such as dietary intake and physical activity have been identified as important contributors to weight gain during this time. Growing evidence suggests that weight gain and obesity are associated with alterations in the gut microbiota. Exercise interventions using animal models have been shown to increase gut microbiome diversity but few studies have assessed associations between physical activity and the intestinal microbiome in humans. This was a secondary analysis of cross‐sectional physical activity data and fecal microbial communities among college freshmen to assess whether associations exist between physical activity and microbial abundance during the first year of college. A total of 89 college students were included in this analysis. Anthropometric data were collected once during the 2014–2015 school year at Arizona State University. At the same time, participating students completed the Godin‐Shepard Leisure‐Time Physical Activity Questionnaire and provided one fecal sample for microbial analysis. Moderate‐to‐vigorous physical activity (MVPA) levels were calculated from survey data by summing the total amount of time spent on strenuous and moderate activity. MVPA was reported in min/day. Fecal microbial community composition was assessed with Illumina miSeq pyrosequencing of PCR amplified 16S rRNA genes. The Firmicutes:Bacteroidetes (F:B) ratio was used to assess phyla‐level differences in microbial abundance by BMI and associations with self‐reported MVPA. The majority (n=51) of participants had a healthy BMI (18.5 – 24.9 kg/m 2 ) at the time of data collection, with 5 classifying as underweight (< 18.5 kg/m 2 ), 21 as overweight (25.0 – 29.9 kg/m 2 ) and 11 as obese (≥ 30.0 kg/m 2 ). Mean F:B ratios for students by BMI category were as follows: 0.398 ± 0.206, 0.439 ± 0.254, 0.523 ± 0.231 and 0.414 ± 0.190 for underweight, normal weight, overweight and obese groups, respectively. No significant differences were observed for daily MVPA (Wilcoxon‐rank sums; p=0.129) or F:B ratio (Wilcoxon‐rank sums; p=0.394) by BMI group. After log‐transforming the F:B ratio to assure the normality of model residuals, no significant association was observed between reported MVPA and the fecal F:B ratio (Linear regression; R 2 =0.030, p=0.104). Principal components analysis was used to reduce the dimensionality of genus‐level microbiota data, from which 12 components explaining 30.8% of the total microbial variation were run in a multivariate regression model with MVPA as the outcome variable. A positive trend was observed between MVPA and component 8 (coefficient=2.303, p=0.098). Major positive contributors to component 8 were Rickettsiales (eigenvalue, 0.162), other Mogibacteriaceae (0.154) and Delftia (0.153) while negative influencers were Flavobacterium (−0.181), other Xenococcaceae (−0.181), and Aggregatibacter (−0.167). Overall, MVPA was not associated with fecal microbiota in this cohort of college freshmen. Slight trends between MVPA and microbial genera suggest that relationships may exist but more work is needed to clearly elucidate the role of physical activity in modulating intestinal microbial composition. Support or Funding Information Project number 1DP5OD017910‐01 (PI: Bruening) was funded by the NIH Office of the Director and the Office of Behavioral and Social Sciences Research.