Table 2. Associations between early-life stress and α-diversity.

	β (95 % CI)	p-value	q-value
Overall early-life stress
Richness	0.003 (−0.029; 0.035)	0.872	0.872
Shannon diversity	− 0.017 (−0.062; 0.029)	0.472	0.707
inverse Simpson diversity	− 0.031 (−0.077; 0.016)	0.194	0.581
Life events
Richness	0.001 (−0.032; 0.029)	0.939	0.963
Shannon diversity	0.009 (−0.035; 0.053)	0.685	0.963
inverse Simpson diversity	− 0.001 (−0.046; 0.044)	0.963	0.963
Contextual risk
Richness	− 0.048 (−0.086; − 0.011)	0.010	0.031
Shannon diversity	− 0.063 (−0.116; − 0.010)	0.021	0.031
inverse Simpson diversity	− 0.051 (−0.106; 0.003)	0.064	0.064
Parental risk
Richness	0.032 (0.000; 0.063)	0.048	0.079
Shannon diversity	0.045 (−0.000; 0.090)	0.053	0.079
inverse Simpson diversity	0.036 (−0.010; 0.082)	0.122	0.122
Interpersonal risk
Richness	0.023 (−0.011; 0.057)	0.190	0.571
Shannon diversity	− 0.004 (−0.053; 0.045)	0.881	0.881
inverse Simpson diversity	− 0.015 (−0.065; 0.035)	0.549	0.823
Direct victimization
Richness	− 0.002 (−0.034; 0.029)	0.889	0.889
Shannon diversity	− 0.011 (−0.056; 0.034)	0.623	0.889
inverse Simpson diversity	− 0.015 (−0.061; 0.031)	0.520	0.889

Models were adjusted for child sex, age, 5 genetic PCs, the time in mail, season of production, batch (DNA Isolation batch and sequencing batch), and number of reads. In addition, models with the risk domains (life events, contextual risk, parental risk, interpersonal risk, and direct victimization) were mutually adjusted for the other risk domains. Q-values are FDR-adjusted within each stressor, over the three α-diversity indices. Results are pooled estimates from 30 imputed datasets.