Table 4.

Multiple logistic regression analysis of the association between osmolality adjusted maternal BPA exposure divided into tertiles and as continuous ln-transformed and the odds ratio (OR) and 95% confidence intervals (CI 95%) of an ASD-score ≥ 75th percentile compared to < 75th percentile (reference) in boys and girls at 2 and 5 years of age

Osmolality adjusted BPA (ng/ml)	All (N = 38/376) Adjusteda OR (CI 95%)	Boys (N = 25/197) Adjusteda OR (CI 95%)	Girls (N = 13/179) Adjusteda OR (CI 95%)
ASD-score ≥ 75th percentile at both 2 and 5 years
1th tertile (≤0.87)	Reference	Reference	Reference
2nd tertile (0.88–1.96)	0.72 (0.26–2.01)	0.58 (0.15–2.17)	0.95 (0.17–5.52)
3rd tertile (≥ 1.97)	2.05 (0.90–4.64)	2.17 (0.79–6.01)	1.94 (0.44–8.57)
p-trendb	0.06	0.09	0.33
Continuous lnBPA c	1.38 (1.02–1.84)*	1.42 (0.99–2.02)	1.34 (0.79–2.25)

^aAnalyses adjusted for maternal education, maternal age, pre-pregnancy BMI, child age at evaluation, parity. The sex-combined analyses were additionally adjusted for child sex

^bTrend across tertiles tested by inserting the tertile osmolality adjusted BPA as an ordinal indicator variable (0,1,2)

^cBPA inserted as a continuous variable transformed by the natural logarithm

* P-value < 0.05