Table 3. Association of Epidural Analgesia Use During Labor and Delivery With Risk of Autism Spectrum Disorder (ASD) in Unmatched and Matched Sibling Cohorts.

	Epidural analgesia during labor and delivery, No. (%)		Absolute RD, % (95% CI)	Unadjusted HR (95% CI)	Adjusted HR (95% CI)a
	Exposed	Not exposed			Model 1b	Model 2c	Model 3d
Unmatched analysis in sibling cohorte
No. of deliveries	59 033	174 449	233 482	233 482	218 132	218 128	218 124
ASD	938 (1.59)	2140 (1.23)	0.36 (0.25-0.48)	1.30 (1.20-1.41)	1.29 (1.18-1.40)	1.12 (1.03-1.22)	1.10 (0.99-1.20)
Matched analysis in sibling cohortf
No. of deliveries	1659	4587	6246	6246	5465	5465	5462
ASD	839 (50.6)	1905 (41.5)	9.04 (6.25-11.84)	1.57 (1.36-1.81)	1.29 (1.10-1.52)	1.05 (0.89-1.24)	1.07 (0.87-1.30)

Abbreviations: HR, hazard ratio; RD, risk difference.

^aMissing data for body mass index were imputed using multiple imputation. The missing data for other variables were not imputed and were assumed missing at random.

^bAdjusted for year of birth, maternal and co-parent age, neighborhood income quintile, and community size.

^cAdjusted for items in model 1 plus gestational diabetes, preexisting diabetes, pregnancy-induced hypertension, other hypertension, parity, smoking during pregnancy, and body mass index.

^dAdjusted for items in models 1 and 2 plus induction of labor, gestational age, sex, small or large for gestational age, and congenital anomaly.

^eLogistic regression models were run comparing deliveries among the cohort of deliveries born to mothers with 2 or more deliveries during the study period.

^fConditional logistic regression models were run matching a woman to herself and comparing across her deliveries. These models only make use of offspring that were discordant in ASD status born to the same woman. Siblings may have had different biological fathers.