Table 3.

Results from multivariate regression models presenting percentage change in 4-hydroxychlorothalonil (HCT) serum concentrations associated with age and parity among pregnant women from Sweden (A), and associated with sociodemographic, pregnancy, environmental and pregnancy variables, among pregnant women from Costa Rica (B)^a.

A
Variables	n = 1806 R2 = 0.09
Intercept
Geometric mean (95% CI) HCT (µg/L)a	4.3 (4.1, 4.5)***
	% Change HCT (95% CI)
Maternal age (years) (tertiles)
≥28 – 32 vs. <28	19 (11, 26)***
≥32 vs. <28	18 (11, 26)***
Parity
1 vs. 0	−27 (−31; −23)***
≥2 vs. 0	−31 (−36, −26)***

B
Variables	First samples, n = 393 R2 = 0.11, R2-adj = 0.08	Second samples, n = 239 R2 = 0.12, R2-adj = 0.08
Intercept
Geometric mean (95% CI) HCT (µg/L)a	16.7 (12.7, 22.1)***	21.5 (13.9, 33.3)***
	% Change HCT (95% CI)	% Change HCT (95% CI)
Maternal age (years) (tertiles)
≥19 – 21 vs. <19	22 (0, 47)*	37 (8, 72)**
≥22 – 27 vs. <19	55 (25, 92)***	43 (11, 84)**
≥28 vs <19	67 (28, 117)***	41 (3, 92)*
Parity
1 vs. 0	−12 (27, 6)	−10 (−28, 11)
2 vs. 0	−29 (−43, −10)**	−24 (−42, −1)*
1 week increase in gestational age	−0.45 (−1.28, 0.39)	−1.06 (−2.33, 0.22)
Residential distance to banana plantations (m) (tertiles)
≥90 – 372 vs. <90	4 (−12, 23)	−2 (−18, 18)
≥373 vs. <90	−19 (−31, −5)*	−27 (−40, −11)**
Woman works in agriculture	27 (−1, 63)	20 (−9, 59)
Partner works in agriculture	−17 (−28, −5)**	−15 (−27, 0)*
Relatively dry season (Feb-Apr or Oct-Nov)	13 (−1, 29)	1 (−13, 18)

^aWe calculated the geometric mean of the intercept by exp(β), and percentage change of variables by [exp(β)-1]*100. For categorical variables this reflects the percentage change as compared to the reference category ( = intercept) and for continuous variables the percent change for each 1-unit increase. ***p < 0.001; **p < 0.01; *p < 0.05.