Table 2.
Effects of Diarrheal Exposures and Covariates on Measures of Bacterial Diversity and Richness in the Fecal Microbiota of the Etiology, Risk Factors, and Interactions of Enteric Infection and Malnutrition and the Consequences for Child Health and Development Peruvian Birth Cohort
| Mean Change in Standardized (βz) and Actual (β) Gut Microbial Diversity Measures (βz [95% Confidence Interval]; β) | ||||
|---|---|---|---|---|
| Predictor Variable | Shannon's Diversity | Simpson's Diversitya | Phylogenetic Diversity | Chao1 Indexb |
| Total V4-16S rDNA reads assigned to amplicon sequence variants per fecal sample (per 1000 reads) | 0.006 (.002 to .01); 0.006** | 0.006 (0 to .011)* | 0.013 (.009 to .017); 0.101*** | 0.013 (.01 to .016); 1.3*** |
| Age (months) | 0.072 (.06 to .085); 0.076*** | 0.082 (.063 to .1)*** | 0.089 (.074 to .103); 0.704*** | 0.056 (.047 to .064); 5.7*** |
| World Health Organization food groups | 0.089 (.062 to .117); 0.094*** | 0.169 (.126 to .212)*** | 0.036 (.009 to .063); 0.288** | 0.073 (.054 to .092); 7.6*** |
| Pathogens per month | 0.223 (.109 to .337); 0.234*** | 0.282 (.102 to .463)** | 0.264 (.148 to .38); 2.087*** | 0.234 (.147 to .321); 26.4*** |
| LAZ at sampling | c | −0.116 (−.223 to −.008)* | c | c |
| Breastfeeding frequency | c | c | c | −0.007 (−.014 to 0); −0.7* |
| Breastfeeding categories (vs weaned) | ||||
| Partial | −0.519 (−.631 to −.407); −0.544*** | −0.637 (−.803 to −.472)*** | −0.47 (−.593 to −.347); −3.721*** | −0.272 (−.351 to −.192); −23.8*** |
| Predominant | −0.442 (−.642 to −.241); −0.463*** | −0.439 (−.776 to −.102)* | −0.517 (−.716 to −.318); −4.092*** | −0.25 (−.395 to −.105); −22.1*** |
| Stunting categories (vs NS/NS) | ||||
| NS/S | 0.024 (−.069 to .117); 0.025 | −0.088 (−.278 to .102) | −0.017 (−.115 to .081); −0.137 | c |
| NS/Sv | −0.021 (−.229 to .187); −0.022 | −0.337 (−.733 to .058) | 0.014 (−.197 to .225); 0.113 | c |
| S/NS | −0.118 (−.309 to .074); −0.123 | −0.154 (−.461 to .154) | −0.163 (−.38 to .055); −1.287 | c |
| S/S | 0.067 (−.117 to .252); 0.071 | −0.054 (−.363 to .255) | −0.062 (−.252 to .128); −0.49 | c |
| S/Sv | −0.319 (−.53 to −.108); −0.334** | −0.783 (−1.197 to −.369)*** | −0.307 (−.544 to −.071); −2.433* | c |
| Diarrheal variables (presented as effects across the entire cohort, regardless of involvement in interactions) | ||||
| Frequency (no. episodes) | −0.02 (−.03 to −.003); −0.02d | −0.02 (−.05 to −.001)d | −0.018 (−.03 to −.005); −0.14** | −0.008 (−.016 to 0); −0.8* |
| Time since last diarrhea (months)e | 0.001 (−.027 to .029); 0.001d | c | 0.017 (0 to .033); 0.133* | 0.012 (.002 to .022); 1.2* |
| Mean duration (days)e | −0.028 (−.046 to −.009); −0.029** | c | c | −0.018 (−.032 to −.003); −1.8* |
| Mean severity (community diarrheal assessment)e | c | −0.07 (−.121 to −.019)** | −0.046 (−.079 to −.012); −0.364** | −0.027 (−.053 to −.001); −2.6* |
| Effects of diarrheal frequency by stunting category (vs NS/NS)f | ||||
| NS/S | −0.016 (−.035 to .004); −0.016 | −0.022 (−.051 to .007) | c | c |
| NS/Sv | 0.026 (−.01 to .062); 0.027 | 0.042 (−.012 to .096) | c | c |
| S/NS | −0.031 (−.065 to .002); −0.033 | −0.049 (−.1 to .002) | c | c |
| S/S | 0.016 (−.03 to .062); 0.017 | 0.028 (−.043 to .099) | c | c |
| S/Sv | −0.057 (−.109 to −.006); −0.06* | −0.099 (−.177 to −.021)* | c | c |
Abbreviations: LAZ, length-for-age z-score; NS/NS, not born stunted and not stunted at the time of sampling; NS/S, not born stunted, stunted but not severely stunted at sampling; NS/Sv, not born stunted, severely stunted at sampling; S/NS, born stunted/not stunted at sampling; S/S, born stunted, stunted but not severely stunted at sampling; S/Sv, stunted and severely stunted at sampling.
aEstimates for Simpson's diversity are presented in units of logit-transformed Simpson's diversity per unit increase of the predictor variable.
bBack-transformed coefficients for Chao1 are interpreted as percent change.
cPredictor variable eliminated during Akaike information criterion–based stepwise model selection.
dEstimates for effects involved in interactions are presented as mean effects without indicators of significance.
eEstimates for time since diarrhea, mean diarrheal duration, and mean diarrheal severity are taken from full models. All others are from frequency-only models.
fStunting categories are written as stunting status at birth/stunting status at sampling: NS = LAZ ≥ −2; S = LAZ < −2); Sv = LAZ < −3.
*, P < .05; **, P < .01; ***, P < .001.