Potential exposure to endocrine disrupting chemicals and selected adverse pregnancy outcomes: a follow-up study of pregnant women referred for occupational counselling

Endocrine disrupting chemicals (EDC) are exogenous substances that cause adverse health effects through interference with the endocrine system [1, 2]. EDC may affect the endocrine system by mimicking or blocking the action of an endogenous hormone or bind to transport proteins. Furthermore, EDC may interfere with the normal endocrine system by changing the normal hormone level through stimulation or inhibition of the production of hormones [1, 2]. Several epidemiological studies suggest that a range of EDC may actually cause preterm birth and/or low birth weight [3, 4]. Many of these studies are concerned with specific substances, such as phthalates [5], some pesticides [6, 7] and poly-chlorinated biphenyls (PCB) [8]. For example, exposure to PCB is believed, with increasing evidence, to inhibit fetal growth [8]. Due to methodological restrictions, the complexity of these chemical substances, and due to preventive measures, the effects are often limited and the results inconsistent [9, 10]. Therefore, it is difficult to determine whether associations between EDC and preterm birth and/or low birth weight are causal [11, 12].

Denmark has had regulation on worker protection since 1873. According to the Danish Working Environment Act, the employer is responsible for ensuring a safe and sound working environment (Executive Order no. 1072 of 7 Sept. 2010). If a potentially harmful reproductive risk is present at the workplace, including exposure to EDC, pregnant women may be referred to an Occupational Health Clinic (OHC) by their general practitioner or midwife for reproductive risk assessment. This usually happens between 8 and 12 weeks of gestation. The clinics perform a risk assessment based upon available evidence in order to assess if the pregnant woman may continue in her job or whether there is a case for intervention in terms of actions to reduce or eliminate hazardous exposures, restructuring of job tasks, reassignment of job functions or – as the ultimate and least desirable solution - maternity leave due to reproductive risks that could not be resolved otherwise [13, 14]. Referred pregnant women are relocated or granted sick leave until the potential risk has been determined. However, keeping the pregnant women in their job, if possible, may be beneficial for the women in order to maintain income and attachment to the labor market.

Assuming that pregnant women referred to an OHC have a more hazardous occupational setting compared to pregnant women in general, since they are referred on the basis of a concrete suspicion of a harmful risk factor, these women constitute a relevant population to investigate. The national and complete coverage of the Danish Birth Register enable comparison of birth weight and gestational age of births by women referred to these clinics to all births in the same geographical regions. The aim of the present study was to examine if potential occupational exposure to EDC during pregnancy is related to preterm birth and/or low birth weight.

Table 1 provides the descriptive characteristics of the study population including the women potentially exposed to EDC, the internal reference group of women referred to an OHC due to exposures other than EDC, and the women in the external reference group of gainfully employed persons. The three groups are similar with respect to characteristics except that mothers referred to an OHC are younger than the external references, and tend to belong to a lower SES group.

Gestational age was approximately equal across the three groups (Table 2) and linear regression analysis with adjustment for the potential confounders showed no significant effect of potential exposure to EDC on gestational age (data not shown).

The mean birth weight of the children born to women in the external reference group was marginally lower than the mean birth weight of the births given by women referred to an OHC, while the mean birth weight in the 40^th week of pregnancy was marginally lower in the group potentially exposed to EDC compared to the external references. However, the confidence intervals overlap, indicating no significant difference in mean birth weight (Table 3).

In analyses adjusting for gestational age and potential confounders, women potentially exposed to EDC had children with a higher birth weight when compared to the external references (in average 63 g higher, 95% CI: 22–101), but not when compared to internal referents (in average 37 g higher, 95% CI: −19–93).

The proportion of preterm births among the women potentially exposed to EDC was marginally larger than the proportion of preterm births among the women in the internal reference group (Table 4), but the difference was not statistically significant. There were no difference in the proportion of preterm births between the women potentially exposed to EDC and the external references.

There was no difference in the proportion of children with low birth weight born to the women potentially exposed to EDC compared to the internal references (Table 5). Compared to the external references, the proportion of children with low birth weight was marginally lower among the women potentially exposed to EDC. This lower risk of low birth weight was likewise not statistically significant.

A sensitivity analysis only including first births did not show any differences of importance for gestational age and birth weight between the three exposure groups (data not shown).

In this prospective follow-up study of pregnant women referred to an OHC for counselling, we found no indications of reduced birth weight or increased risk of preterm birth among women potentially exposed to EDC. Women potentially exposed to EDC had children with a higher birth weight when compared to the external references. This might reflect an anabolic effect of EDC (for instance an estrogenic effect), but without consistent findings in analyses using the internal references, we consider this a chance or spurious result.

Women referred to an OHC are presumed to represent a high risk group, and constitute thereby a preferable basis for investigation compared to a population based study. The present study design enables comparison between women referred to an OHC potentially exposed to EDC and women referred to an OHC but unexposed to EDC. These two groups are highly comparable, since they are expected to be similar in most aspects except exposure status. Furthermore, use of register data provides a large study population contributing to enhanced statistical power.

These reassuring results can be understood in several ways. First of all, Denmark has a long tradition of worker protection regulation and the Danish Working Environment Authorities regularly performs inspections at Danish workplaces. Thus, potentially harmful chemical exposures of pregnant women at Danish workplaces are in general considered well controlled and chemically induced birth defects are rarely notified to the National Board of Industrial Injuries [20]. Therefore, the women defined as potentially exposed to EDC in the present study might not have been exposed to EDC to an extent that is strong enough to cause detectable adverse effects on gestational age and birth weight. A second explanation of the findings might be that the risk imposed by EDC is reduced by occupational risk assessment and counselling. If the risk assessment indicates a harmful exposure, actions to reduce or eliminate the exposure are recommended. Finally, awareness of potential harmful exposure, induced by the referral to an OHC, might cause a behavior change to a more protective lifestyle during pregnancy [13]. However, human embryo development is most vulnerable to toxic substances and endocrine disruption in the early embryonic period during the first 8 weeks of gestation, while counselling at an OHC usually takes place between 8 and 12 weeks of gestation.

One limitation of this study concerns the determination of exposure to EDC. The risk of EDC might go undetected in the present study due to misclassification associated with determination of exposure status. Using a JEM does not necessarily give an accurate picture of exposure since exposure might vary within job titles [7].

The JEM of van Tongeren et al. divides the chemical substances into seven substance categories [1]. In the present study all substances were combined into one group of EDC. By doing so, some accuracy in relation to potential differences between substances is lost. However, these reductions of categories were made to address the purpose of the present study - to investigate the general potential effects of EDC on gestational age and birth weight.

Finally and importantly, EDC are also found in many consumer products and the exposure besides the occupationally induced might be profound which would diminish the association in the present study. We know very little about what factors pregnant women are exposed to at home [21].

Pregnant women with potential occupational exposure to EDC were not at increased risk of delivering preterm or low birth weight babies. Findings may reflect that exposure levels are low without impact on the studied reproductive outcomes or that occupational counseling result in effective preventive actions.