Table 2.

Binary regression model establishing association between greater patient-reported childbirth trauma and postpartum depression, over and above covariates

Postpartum depression (time 3)
Covariates	B	SE	p	OR	95% CI
Maternal age (years)	.001	.02	.98	1.00	.96, 1.04
Socioeconomic position	−.07	.11	.54	.93	.75, 1.12
Parity	.17	.08	.04	1.19	1.004, 1.37
Black vs. White	−.07	.31	.82	.93	.51, 1.56
Asian vs. White	−.07	.34	.85	.94	.47, 1.67
Others vs. White	.04	.54	.95	1.04	.35, 2.56
Latinx vs. Non-Latinx	.03	.42	.94	1.04	.46, 2.06
Married/living with partner (time 1)	−.08	.28	.77	.92	.54, 1,45
Breastfeeding (time 3)	−.12	.11	.30	.89	.72, 1.07
Prenatal anxiety (time 1)	.93	.27	.001	2.53	1.49, 3.95
Prenatal depression (time 1)	1.17	.31	<.001	3.22	1.75, 5.93
Predictor
Patient-reported childbirth trauma (time 3)	.29	.10	.003	1.33	1.10, 1.60

Significant results are bolded. OR, odds ratio. Patient-reported childbirth trauma was modeled as a latent factor with three indicators: “during your delivery or immediately after delivery, how much did you fear for your own life?”, “during your delivery or immediately after delivery, how much did you fear for your baby’s life?”, and “how traumatic did you find your birth experience?” (see Supplemental Figure S2). We modeled pathways between covariates and both patient-reported childbirth trauma and PPD, as well as modeling covariances between all covariates. Estimates were similar using a continuous measure of PPD (total scores for EPSD; Table S2) and after removing women who had PTB (<37 weeks; Table S3). Estimates were also similar when controlling for mode of childbirth (i.e., vaginal vs. cesarean; Table S4). Mode of childbirth was also not a significant moderator of the association between patient-reported childbirth trauma and risk for PPD (Table S5; i.e., higher patient-reported childbirth trauma was associated with greater risk for PPD similarly among individuals who had a vaginal or cesarean birth).