Table 2

Effects of PIM on child development modified according to whether it started during pregnancy (121 pairs) or after birth (480 pairs)

Outcome	Unadjusted*				Matched†				Matched with double adjustment‡
	Started during pregnancy		Started after birth		Started during pregnancy		Started after birth		Started during pregnancy		Started after birth		Heterogeneity p-value§
	N=663		N=2534		N=242		N=960		N=242		N=960
	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI
Linear regression for mean BDI score
Total BDI score (SD)	−0.07	−0.27 to 0.14	−0.22	−0.32 to −0.13	0.17	−0.06 to 0.41	−0.06	−0.18 to 0.07	0.19	−0.02 to 0.40	−0.03	−0.15 to 0.10	0.080
Poisson regression for belonging to the group below the 10th BDI percentile
	PR	95% CI	PR	95% CI	PR	95% CI	PR	95% CI	PR	95% CI	PR	95% CI
Low development score	0.76	0.39 to 1.50	1.50	1.15 to 1.96	0.45	0.20 to 1.00	1.14	0.82 to 1.60	0.40	0.18 to 0.89	1.12	0.81 to 1.54	0.020

*Comparison between intervention group and possible controls group without adjustment.
†Paired analysis comparing the intervention group with the matched control group.
‡Paired analysis comparing the intervention group with the matched control group with double adjustment for the following confounders included in the propensity score prediction for starting during pregnancy: neighbourhood violence, mother’s skin colour (others/white), maternal age at birth (>19 years), sex, maternal schooling level (years), paternal schooling level, income at birth (quintiles), asset index at birth (quintiles), people per bedroom in house, number of children with the mother, mother living with partner, mother worked during pregnancy, father’s support during pregnancy, planned pregnancy and older sibling received the PIM; and the following confounders for starting PIM after birth: neighbourhood violence, mother’s skin colour (others/white), maternal age at birth (>19 years), sex, maternal schooling level (years), paternal schooling level, income at birth (quintiles), asset index at birth (quintiles), number of prenatal visits (>5), people per bedroom in house, number of children with the mother, mother living with partner, mother worked during pregnancy, father’s support during pregnancy, planned pregnancy, depressive maternal symptoms (3 months), birth weight (grams), smoking during pregnancy, alcohol use during pregnancy, maternal hypertension during pregnancy, maternal diabetes during pregnancy, gestational age (<37 weeks), main caregiver until age 3 months, childcare attendance up to age 2 years and older sibling received the PIM.
§Cochran’s Q heterogeneity χ2 test.
BDI, Battelle’s Developmental Inventory; PIM, Primeira Infância Melhor; PR, prevalence ratio.