Skip to main content
. 2017 Jun 8;7:3048. doi: 10.1038/s41598-017-03106-z

Table 2.

Multivariable linear regression analyses of the association between maternal urinary chromium (μg/g cr) and gestational age (days).

Chromium All infants Male infants Female infants p-value
n β (95% CI)a β (95% CI)b n β (95% CI)a β (95% CI)b n β (95% CI)a β (95% CI)b
Continuous ln-Cr 7290 −1.05(−1.19, −0.91) −0.68(−0.88, −0.48) 3891 −1.11(−1.29, −0.92) −0.71(−0.98, −0.44) 3399 −0.95(−1.16, −0.74) −0.65(−0.95, −0.36) 0.24d
Low (≤1.09) 2428 reference reference 1292 reference reference 1136 reference reference
Medium (1.09–3.76) 2435 −0.95(−1.43, −0.49) −0.67(−1.14, −0.20) 1263 −1.20(−1.86, −0.54) −0.95(−1.60, −0.29) 1172 −0.71(−1.39, −0.03) −0.39(−1.08, −0.28)
High (>3.76) 2427 −3.49(−3.97, −3.02) −2.30(−2.93, −1.67) 1336 −3.78(−4.43, −3.13) −2.50(−3.37, −1.62) 1091 −3.12(−3.81, −2.43) −2.10(−3.01, −1.20)
p for trendc <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 0.21e

CI = confidence interval.

aUnadjusted confounders.

bAdjusted for maternal age, education, BMI, parity, pregnancy-induced hypertension and other metals.

c p-values for trend across tertiles of maternal urinary chromium.

d p-value for ln-Cr*sex interaction.

e p-value for Cr tertile*sex interaction.