Table 2.
Associations between maternal characteristics and top 5 phyla and top 10 genera: Grouped by BMI
| Taxa | Group effect p-value | Pairwise comparison | IRR | 95% CI | Pairwise comparison p-value |
|---|---|---|---|---|---|
| Pre-pregnancy BMI | |||||
| Phylum | |||||
| Proteobacteria | 0.019 | Obese vs overweight | 0.62 | 0.43–0.90 | 0.012 |
| Overweight vs healthy | 1.23 | 1.00–1.50 | 0.045 | ||
| Bacteroidetes | 0.0051 | Obese vs overweight | 3.70 | 1.61–8.48 | 0.002 |
| Obese vs healthy | 2.56 | 1.27–5.17 | 0.0086 | ||
| Genus | |||||
| Staphylococcus | 0.011 | Obese vs overweight | 2.50 | 1.09–5.72 | 0.031 |
| Obese vs healthy | 3.15 | 1.47–6.76 | 0.0032 | ||
| Corynebacterium | 0.0003 | Obese vs overweight | 5.13 | 1.79–14.70 | 0.0023 |
| Obese vs healthy | 4.98 | 2.11–11.74 | 0.0002 | ||
| Brevundimonas | < 0.0001 | Overweight vs healthy | 8.72 | 3.24–23.48 | < 0.0001 |
| < 0.0001 | Unscheduled C-section vs vaginal | 16.70 | 5.99–46.57 | < 0.0001 | |
| Scheduled C-section vs unscheduled C-section | 0.071 | 0.011–0.46 | 0.0053 | ||
| 3-month post-partum BMI | |||||
| Phylum | |||||
| Actinobacteria | 0.0058 | Obese vs overweight | 2.34 | 1.38–3.98 | 0.0017 |
| Obese vs healthy | 2.02 | 1.18–3.46 | 0.010 | ||
| Genus | |||||
| Corynebacterium | < 0.0001 | Obese vs overweight | 4.84 | 2.19–10.72 | 0.0001 |
| Obese vs healthy | 7.77 | 2.95–20.43 | < 0.0001 | ||
| Brevundimonas | 0.0005 | Obese vs overweight | 8.89 | 2.29–34.57 | 0.016 |
| Obese vs healthy | 9.56 | 2.17–42.22 | 0.0029 | ||
| < 0.0001 | Unscheduled C-section vs vaginal | 13.01 | 4.01–42.20 | < 0.0001 | |
| Scheduled C-section vs unscheduled C-section | 0.08 | 0.013–0.62 | 0.015 | ||
Separate Poisson regression models were run for pre-pregnancy BMI and 3-month post-partum BMI, while adjusting for maternal glucose tolerance status, mode of delivery, DNA extraction batch, and PCR sequencing batch. Statistically significant main group effect findings shown only (group effect: p ≤ 0.022 for phylum, p ≤ 0.017 for genus; pairwise comparison: p < 0.05). An interaction term between pre-pregnancy BMI and maternal glucose tolerance status was found to be statistically significant for Gemella. No other statistically significant interactions were found between maternal pre-pregnancy BMI and glucose tolerance status. Gemella showed an increased incidence among mothers with an overweight BMI (versus healthy BMI) and concurrent gestational diabetes IRR, CI (5.96 [1.85–19.21], p = 0.0028). An increased incidence of Gemella was also observed in mothers with an obese BMI and concurrent impaired glucose tolerance versus both mothers with overweight (11.42 [1.49–87.67], p = 0.019) and healthy weight BMIs (4.04 [1.63–10.01], p = 0.0026). Abbreviations: confidence interval CI, incidence rate ratio, IRR