Women referred for occupational risk assessment in pregnancy have no increased risk of adverse obstetric outcomes

The prevalence of long-term absence from work during pregnancy is high in the Nordic countries [1]. For example, among hospital employees 31% were on sick leave for at least 10% of their scheduled working time. Pregnancy-related medical disorders account only for a small part of the sick leave during pregnancy. On the other hand, discomfort related to strenuous work seems also to be important.

According to the Danish Executive Order on the Performance of Work (Executive Order No. 559 of 17 June 2004), the employer is responsible for ensuring a safe working environment. If a harmful reproductive hazard is present and if preventive measures or reassigning the worker to a safe job is not possible, healthy pregnant women may obtain social benefits. Some 3-4% of all cases of pregnancy leave in Denmark are due to such situations [2]. A manual from the Danish Working Environment Authority (AT-guidance A.1.8 of January 2009) describes the impacts that are likely to endanger pregnancy [3]. If the general practitioner or midwife assesses that the woman’s working conditions may pose a risk to the pregnancy, the pregnant woman may be referred to the regional clinic of occupational health for occupational risk assessment. This usually happens between 8 and 12 weeks of gestation. The referred pregnant woman should be repositioned or be granted sick leave until the risk assessment has been carried out.

Several reproductive workplace hazards have been identified [4, 5]. Shift work, long working hours, heavy lifting, standing, and exposure to endocrine-disrupting chemicals and other chemicals have been associated with an increased risk of adverse pregnancy outcomes [6-8], although not with strong effects [9]. For example, there is increasing evidence that exposure to polychlorinated biphenyls (PCBs) is associated with inhibited foetal growth, even at low doses [10].

Assuming that women referred to counselling at an occupational health clinic have a more hazardous occupational work environment than pregnant women in general, we examined the risk of preterm birth, low birthweight, and small for gestational age in this group.

METHODS

The occupational medical database

An occupational medical database hosted by two regional occupational university departments located in the Eastern (Copenhagen) and Western (Aarhus) regions of Denmark included 1,358 pregnant women who were referred for risk assessment and management during the 26-year period from 1984 to 2010 (referred
women’s referred pregnancies, RWRP). An internal reference group consisted of the referred women’s 2,042 non-referred pregnancies, RWNP (the women served as their own reference). Furthermore, an external reference group consisted of all births among women who were gainfully employed when pregnant in the period from 1984 to 2010 in the same geographical regions and who were not referred for risk assessment – in total 770,605 (Danish working population, DKWORKPOP). We introduced restrictions as to those who were working at the onset of their pregnancy, which left 1,273/1,683 and 610,873 in the RWRP/RWNP and DKWORKPOP groups, respectively. Further restrictions were: mother’s age at delivery 16-45 years, birthweight 1,000-7,000 g and gestational age 154-310 days. These restrictions excluded 15/29/8,987 from the RWRP/ RWNP and DKWORKPOP groups, respectively. Since pregnancies of multiparae cannot be considered independent, one birth was
chosen by random sampling among women with two or more deliveries in each of the three groups, which left 1,202 women in the RWRP group (261 from Copenhagen, 941 from Aarhus), 1,077 in the RWNP group (181 from Copenhagen, 896 from Aarhus) and 345,467 in the DKWORKPOP group. Data from the eastern region of Denmark were collected only from one out of three hospitals, so the number of patients from this region was relatively small compared with the number from the western region. The date of consultation and the physicians’ registrations of occupational exposure, occupation and industry were extracted from the occupational medical database.

The Danish Civil Registration System

Persons with a permanent address in Denmark have a unique ten-digit personal identification number (PIN), which was used to link the occupational medical database to the Danish Civil Registration System, the Medical Birth Register, and to public registries developed by Statistics Denmark [11, 12].

Outcomes and confounders

Data on pregnancy outcomes in terms of birthweight, gestational age, previous pregnancies and smoking status were drawn from the Medical Birth Register. Data on pre-pregnancy body mass index (BMI), defined as the weight in kg divided by the squared height in meters, was available only for the period 1997-2010. Data on the country of origin and current home region were drawn from Statistics Denmark [11, 12].

Low birthweight (LBW) was defined as a birthweight below 2,500 g. Gestational age at the time of consultation was calculated as the date from conception to the date of consultation divided by 7. Small for gestational age (SGA) was defined as a birthweight below the gestational week-specific 10th percentile based on the reference population.

The mother’s age and parity (number of children) was defined at delivery in the index pregnancy. Occupational status was job status at the time of pregnancy and it was classified according to DISCO 88, the Danish version of the International Standard Classification of Occupation (ISCO). Smoking was categorised as “yes” if the woman had smoked at any time during pregnancy, even though she stopped after her 1st trimester and regardless of the amount of smoking.

Statistical analysis

We compared the women referred to an occupational health clinic with an ongoing pregnancy with their non-referred pregnancies RWNP and the external DKWORKPOP reference group and computed the average difference with a 95% confidence interval (95% CI).

Thanks to the Danish registries with nearly complete data and follow-up, the short follow-up time and the low loss to follow-up due to emigration or death, the risk of the outcomes LBW, birth before 37th week of gestation, and SGA in the RWRP group compared with RWNP and with the DKWORKPOP group, respectively, were assessed by use of multivariate logistic regression. We calculated odds ratios (OR) with 95% confidence intervals (CI) for each outcome and adjusted for the following potential confounders: mother’s age at delivery, parity, smoking, socioeconomic status (SES), ethnicity, and – in further analyses – for children’s year of birth. Data were analysed at Statistics Denmark using SAS version 9.3.

Trial registration: none. The study was conducted using systematically collected data including the refereed women’s occupational exposure codes, which were anonymised and linked to national registries at Statistic Denmark. The Danish Data Protection Agency approved the study (R. no. 2012-41-1267).

RESULTS

As summarised in Table 1, women referred to an occupational health clinic had children earlier than women in the DKWORKPOP group. They were also generally at a lower SES level, and were more often unskilled. Smoking during pregnancy differed only slightly between the groups. During the 1997-2010-period, referred women had a slight tendency towards a higher BMI than the women in the DKWORKPOP group. Most women were referred after year 2000 (n = 643 (53.5%)).

Chemical exposure was the most frequent type of work-related risk factor in the RWRP group (55.7%) followed by ergonomic exposure (25%), Table 2. Referred women on average received their counselling in week 12 of gestation. No significant difference in gestational age or birthweight between pregnancies with chemical,
ergonomic, biological, or psychological workplace exposure was seen in crude analyses (Table 2) or when further adjusting the analyses for possible confounders. Women exposed to chemical factors had no increased risk for either LBW (OR: 0.8; 95% CI: 0.4-1.5), preterm birth (OR: 0.8; 95% CI: 0.5-1.1) or SGA (OR: 0.8; 95% CI: 0.6-1.2) compared with the DKWORKPOP group when adjusted for year of birth, mother’s age at delivery, parity, SES, ethnicity and smoking. Similar findings were recorded for ergonomic exposure.

For the pregnant women who were referred to an occupational health clinic, seen as an entity and as a proxy for being more exposed to work-related risk factors, no association with LBW (OR: 0.81; 95% CI: 0.52-1.26), preterm delivery (OR: 0.8; 95% CI: 0.62-1.64) or SGA (OR: 0.92; 95% CI: 0.72-1.17) was found. Similar results were found when referred pregnancies were compared with the referred women’s non-referred pregnancies. The referred women had children with a higher birthweight than the DKWORKPOP group (difference: 47.8 g; 95% CI: 19.9-75.6) when adjusted for year of birth, mother’s age at delivery, parity, SES, ethnicity and smoking. An additional adjustment for BMI as a continuous variable in the 1997-2010-period, where BMI data were available, showed a non-significant difference
(p = 0.07).

The group of RWRP did not differ from the DKWORKPOP group with respect to gestational age in weeks (difference: 0.05 g, 95% CI: –0.06-0.17). No differences were found between the referred and the non-referred pregnancies of the referred women, either in terms of mean birthweight or gestational age, Table 3 and Table 4.

In sub-analyses, we found that smoking increased the risk of having a child with LBW (OR 2.0, 95% CI 1.9-2.1), preterm delivery (OR: 1.4; 95% CI: 1.4-1.5) and SGA (OR: 2.3; 95% C: 2.2-2.4). Higher salaried employees had a lower risk of LBW, preterm delivery and SGA than unskilled workers.

Analyses with further restriction to primiparae in all three groups showed results consistent with the prior results as referred pregnancies were found to result in children with a significantly higher birthweight than pregnancies among women in the DKWORKPOP group (adjusted difference between means was 41.1 g; 95% CI: 4.7-77.5). No difference between mean gestational ages was found when adjusted for year of birth, mother’s age at delivery, parity, SES, ethnicity and smoking. The birthweight of the children born after occupational counselling did not change across the study period which spanned 26 years. The average birthweight in children of mothers referred from 1984 to 1999 was 3,508 g (95% CI: 3,463-3,552; n = 562), while it was 3,507 g in children born in 2000-2010 (95% CI: 3,462-3,551; n = 640). Further adjustment for the year of birth of the children did not modify the association between birthweight and exposure group.

DISCUSSION

This large follow-up study shows that women receiving counselling at an occupational health clinic because of occupational reproductive hazards do not have an increased risk of LBW, preterm delivery or SGA. One interpretation of the findings is that the occupational counselling and risk management is adequate and given in time to prevent these adverse birth effects. Another interpretation is that the present working environment in Denmark is so well-controlled that even though risk factors do exist, the level of exposure is too low to result in detectable adverse effects. A pregnant woman is most often very concerned and another factor, which may contribute to the results, is that the women may avoid situations they fear may pose a risk.

The study indicated that referred women had children with a higher mean birthweight than the average employee in Denmark. Previous studies have shown that obesity and diabetes mellitus increase the risk of having large-birthweight children [13]. Our analyses with adjustment for BMI showed that the higher birthweight of the referred women’s referred pregnancies can probably be ascribed to a higher BMI average in this group.

No difference between the pregnant women who were referred to occupational risk assessments and the same women’s non-referred pregnancies was found,
either in terms of birthweight, gestational age or in terms of SGA. Still, it is possible that their pregnancies might differ in other aspects, e.g. a woman might have changed smoking habits, job or the father of her children. However, the similarity in birth outcomes points towards the conclusion that either the counselling given at the occupational health clinic is sufficient or that the women are not at an increased risk of delivering preterm or LBW children.

Even though only one third of the referred women are recommended leave [1] and most of the women are found not to be exposed to work-related physical or chemical risk factors during their pregnancy, it is likely that the consultation has a reassuring effect. A number of studies indicate that stress is related to adverse obstetric outcomes [14, 15], and it is therefore beneficial if the counselling helps the woman cope with stressors.

The strengths of this study include the relatively large reference group and reliable register-based data. Furthermore, the expected association between smoking and LBW was demonstrated which indicates the validity of the database [16, 17]. The gradients in birthweight, preterm delivery and SGA by the mother’s working level were also consistent with earlier findings [18, 19].

The limitations of the study include limited statistical power to assess even broad dimensions of occupational exposure as for instance strenuous work and heavy lifting (only 25% of the exposed), as well as subgroups of chemical exposures. The reassuring findings of our analysis are, however, in agreement with recent
meta-analyses. Based upon a large body of high-quality prospective studies, these analyses conclude that the risk of low birthweight and preterm delivery in relation to occupational heavy lifting is limited if at all present [9]. Other limitations are lacking data on miscarriage, pregnancy complications such as preeclampsia and congenital malformations, as well as data on the association between possible exposure status and referral status. For instance, a study including Danish women who received occupational counselling indicated a slightly increased risk of cryptorchidism in sons of greenhouse workers who had been exposed to pesticides [20]. It is therefore important to acknowledge that this study is not informative with respect to risk for all adverse pregnancy outcomes, including specific malformations. All occupational clinics in Denmark were invited to provide data for this analysis. The reason why only two clinics contributed was mainly lack of systematic registration of relevant data in the majority of the clinics. This also explains the low number of pregnancies in the capital area compared with Aarhus, the second largest city in Denmark. Only one of the three clinics in Copenhagen recorded the relevant data except for the last four years. Thus, our data do not represent rural areas where working conditions and exposures for pregnant women may be different.

CONCLUSION

Pregnant women referred for consultation at two large clinics of occupational medicine in Denmark due to a suspected hazardous work environment did not have an increased risk of low birthweight and preterm delivery compared with pregnant women in general in Denmark. These reassuring findings may either reflect that reproductive hazards at the workplace are rare and/or that occupational counselling is effective in reducing risk factors. Since the study does not address miscarriage, pregnancy complications such as preeclampsia and congenital malformations, a continuous focus on potentially reproductive hazards in the work environment is appropriate.

Correspondence: Linda Kaerlev, Research Unit of Clinical Epidemiology, Institute of Clinical Research, University of Southern Denmark, Sdr. Boulevard 29, entrance 101,4, 5000 Odense C, Denmark. E-mail: L.Kaerlev@dadlnet.dk

Accepted: 19 May 2015

Conflicts of interest:Disclosure forms provided by the authors are available with the full text of this article at www.danmedj.dk