. Author manuscript; available in PMC: 2015 Sep 17.

Published in final edited form as: Am J Obstet Gynecol. 2011 Dec 30;206(4):327.e1–327.e8. doi: 10.1016/j.ajog.2011.12.030

Table 4.

Association between occupational exposures during the first trimester and early versus late spontaneous abortion

Occupational Exposures	Odds Ratio (95% Confidence Interval)
Occupational Exposures	Early Spontaneous Abortion (<12 weeks)^a n=575	Late Spontaneous Abortion (12–20 weeks) ^a,^b n=200	Difference of effect estimates between early and late^c P-value
Anesthetic gases
<1 hour per day (reference)	1.0	1.0
1+ hours per day	0.94 (0.66–1.33)	0.79 (0.44–1.42)	0.6
Antineoplastic agents
<1 hours per day (reference)	1.0	1.0
1+ hours per day	2.13 (1.39–3.27)	1.39 (0.68–2.84)	0.3
Antiviral agents
<1 hours per day (reference)	1.0	1.0
1+ hours per day	0.76 (0.50–1.18)	1.35 (0.75–2.44)	0.1
Sterilizing agents
<1 hour per day (reference)	1.0	1.0
1+ hours per day	1.13 (0.80–1.60)	2.10 (1.29–3.41)	0.04
X-ray radiation
<1 hour per day (reference)	1.0	1.0
1+ hours per day	1.31 (1.01–1.71)	0.98 (0.63–1.53)	0.3

Adjusted for all work exposures, age, parity, shift work categories, and hours worked per week.

Excludes the 575 pregnancies ending in early spontaneous abortion

To assess statistically significant differences between early and late spontaneous abortion models, we calculated p-values for common effects with a chi-square test statistic using the maximum likelihood estimates from the logistic regressions.