- Email: [email protected]
Reproductive disorders among male and female greenhouse workers
Available online at www.sciencedirect.com
Reproductive Toxicology 25 (2008) 107–114
Reproductive disorders among male and female greenhouse workers Reini W. Bretveld ∗ , Mari¨ette Hooiveld, Gerhard A. Zielhuis, Annelies Pellegrino, Iris A.L.M. van Rooij, Nel Roeleveld Department of Epidemiology and Biostatistics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Received 29 November 2006; received in revised form 11 June 2007; accepted 30 August 2007 Available online 11 September 2007
Abstract The aim of this study was to evaluate reproductive disorders in male and female greenhouse workers. In 2002, data were collected from 4872 Dutch greenhouse workers and 8133 referents through postal questionnaires with detailed questions on reproductive disorders of the most recent pregnancy, lifestyle habits, and occupational exposures (e.g. pesticides) prior to conception. Different reproductive outcome measures were compared between 957 male and 101 female greenhouse workers and 1408 referents by means of logistic regression analyses. The analyses of primigravidous couples showed a slightly elevated risk of prolonged TTP (ORwomen = 1.9; 95% CI: 0.8–4.4) and an increased risk of spontaneous abortion among female greenhouse workers (ORwomen = 4.0; 95% CI: 1.1–14.0). A decreased risk of preterm birth was found among male greenhouse workers (ORmen = 0.1; 95% CI: 0.03–0.5). This study may offer some evidence for the hypothesis that pesticide exposure affects human reproduction leading to spontaneous abortion and possibly to prolonged time-to-pregnancy. © 2007 Elsevier Inc. All rights reserved. Keywords: Pesticides; Greenhouse workers; Time-to-pregnancy; Spontaneous abortion; Reproductive disorders
1. Introduction Human reproduction is a delicate process that can be influenced by many factors. Maternal, and to lesser extent, paternal age [1–4], smoking [2,5–9], alcohol and coffee intake [1,9–11], socio-economic status [2], genetic factors, and hormonal imbalance [12] can all affect a couple’s reproductive capability. Hormonal balance of sexual hormones in particular is an important factor in maintaining fertility and regulating reproductive processes. Exogenous substances, such as environmental endocrine disruptors, may disturb the hormonal balance and thereby cause reproductive disorders [13,14]. Relatively high levels of exposure to environmental endocrine disruptors in the form of pesticides occur among people working in agriculture [15]. Some pesticides are able to influence the synthesis, storage, release, recognition, or binding of hormones, which may lead to alterations in reproductive hormone levels [16].
∗ Corresponding author at: Department of Epidemiology and Biostatistics (HP 133), Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Tel.: +31 243610357; fax: +31 243613505. E-mail address: [email protected] (R.W. Bretveld).
0890-6238/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.reprotox.2007.08.005
Several studies have been performed on occupational exposure to pesticides and adverse effects on human reproduction, including endpoints such as prolonged time-to-pregnancy (TTP), spontaneous abortion or stillbirth, low birth weight, and developmental disorders. An overview by Hanke and Jurewicz [17] showed that the effects of employment in agriculture on TTP are not unequivocal, but the majority of studies suggest a relationship between pesticide exposure and decreased fertility [18–20]. Associations between occupational pesticide exposure and birth weight, prematurity, and sex-ratio are inconsistent, but parental exposure in agriculture may increase the risk of congenital malformations in offspring [17]. Studies on other reproductive disorders suggest increased risks of spontaneous abortion and stillbirths for women occupationally exposed to pesticides [21–25]. Because these studies did not provide conclusive evidence to support the hypothesis that pesticide exposure adversely affects human reproduction, we performed a retrospective cohort study in The Netherlands to examine the effects of pesticide exposure on reproduction in male and female greenhouse workers. We focused our study on flower greenhouses, where large amounts of pesticides, such as abamectine, imidacloprid, methiocarb, deltamethrin, and pirimicarb are used [26]. The average amount of pesticide use is 28 kg per hectare
108
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
per year (range of 0.06–249 kg/hectare) and a large percentage of active ingredients are classified as fungicides. The specific aim of this study was to assess which part of the reproductive process is predominantly affected by occupational pesticide exposure. 2. Materials and methods 2.1. Population We used two different approaches to obtain the study population. From the database of a major trade union, which registers information about a variety of Dutch workers, we identified 1536 male and 604 female greenhouse workers and a random sample of 2350 male and 949 female cleaners of reproductive age (born in 1956–1979) in 2002. As only a small percentage of all greenhouses workers are members of a trade union in The Netherlands, we also obtained addresses of 2414 greenhouse owners through the national database of an agricultural marketing company with coverage of more than 85%. These owners were born in 1950–1979 or their date of birth was unknown. For the reference group, addresses of 4691 shopkeepers and market stallholders were obtained through the Chamber of Commerce, which covers all companies in The Netherlands. We only selected companies in which exposure to pesticides or other reproductive toxicants was unlikely (e.g. builder’s merchants and shops for household goods). This reference group of cleaners, retail shopkeepers, and market stallholders was chosen because of assumed comparability with greenhouse workers and owners with respect to educational level, socioeconomic status, and working conditions such as standing and carrying heavy loads.
2.2. Data collection All potential participants were sent an introductory letter, an information leaflet, and a self-administered questionnaire. Owners of a company received two questionnaires: one for themselves and one for their partners, if he or she worked in the same company. Owners were also asked for addresses of employees so we could send them questionnaires as well. Alternatively, the owners could choose to personally distribute questionnaires to their employees. Only a few owners used these options. To increase response rates, reminders were sent after 2 and 6 weeks. To avoid selective non-response and information bias, the study was presented in general terms, i.e. as a study on work, lifestyle, and fertility. We focused on the most recent pregnancy to avoid dependency issues of multiple pregnancies, assuming that the last pregnancy would be remembered best.
2.3. Questionnaires The design of the questionnaires was based on the literature and on previous reproductive studies performed by our group [27–29]. Separate questionnaires were used for women and men, who were asked whether they or their female partner had ever been pregnant. If so, detailed questions were asked about work, exposure to pesticides, and lifestyle (e.g. smoking habits, coffee, and alcohol consumption) of the respondents and their partners. All questions related to a period of 6 months before the start of the most recent pregnancy. In addition, questions were asked about time-to-pregnancy and adverse pregnancy outcomes.
2.4. Outcome definitions We defined time-to-pregnancy (TTP) as the number of months of unprotected intercourse leading to pregnancy. In accordance with the clinical definition of subfertility, the cutoff point for prolonged time-to-pregnancy was 12 months. Only planned pregnancies, regardless of pregnancy outcome, were included in the TTP analyses. Spontaneous abortion was defined as natural termination of pregnancy prior to the 20th week of gestation. Fetal deaths from 20 weeks of gestation onwards were defined as stillbirths. A delivery of a live-born infant of less than 37 weeks was classified as preterm. The cutoff point for low birth weight was 2500 g. The analyses of birth weight were always adjusted for gestational age. Sex-ratio was defined as the ratio between male and female live-born infants. For birth defects, we excluded genetic defects (e.g. heriditary metabolic disorders) and minor defects, such as hypotonia or strabismus, from the analyses.
2.5. Exposure Exposure was qualified by job title reported for the 6-month-period prior to the most recent pregnancy or pregnancy attempt. All participants were classified as ‘greenhouse workers’ (greenhouse and other horticulture workers with likely exposure to pesticides), ‘referents’ (cleaners, market and retail workers, and other workers not exposed to pesticides) or ‘not working’. After classification, we selected couples with at least one person who worked and formed two different exposure groups: female greenhouse workers (without pesticide exposed partners) and male greenhouse workers (with or without assisting female partners). In the reference group neither of the partners were exposed to pesticides.
2.6. Potential confounding factors Smoking habits, alcohol and coffee consumption, maternal vitamin use, oral contraceptives in the 12 months preceding the most recent pregnancy, and primiparity were used dichotomously in the analyses. Educational level was categorized as low (no education and primary, lower vocational, or intermediate vocational education) or high (intermediate secondary, higher secondary, higher vocational, or university education). Job status for each couple was divided into three categories (only the woman worked, only the man worked, and both partners worked). Both male and female age were used as continuous variables in the analyses.
2.7. Non-response We randomly selected 150 male and 150 female greenhouse workers and 150 male and 150 female cleaners from the trade union, 200 greenhouse owners from the marketing company, and 200 shopkeepers from the Chamber of Commerce who did not return the questionnaire. These non-participants were called and asked for the reason why they did not respond and whether they had any children. When they did, we tried to ask questions about TTP and outcome of the most recent pregnancy.
2.8. Analyses Associations between each outcome measure and greenhouse work were studied by means of logistic regression analyses. Initially, crude odds ratios (OR) with 95% confidence intervals (95% CI) were calculated, for odds ratios based on small numbers (less than 10) using the mid-p exact method [30]. Following, multivariable analyses were performed with confounder correction and checks for interaction between the exposure variable and all relevant covariates listed in Table 2. We used two criteria to evaluate whether factors were potential confounders: the distribution among greenhouse workers and referents and the associations of the potential confounder with each outcome measure in the reference group. The confounder set was then evaluated for each outcome measure separately by reducing the full model with all potential confounders to gain precision without loosing validity. Only variables that did not change the OR by more than 0.05 upon removal were permanently deleted from the model. Because in non-experimental research of reproductive failure most pregnancies are planned and past pregnancy experience is used in planning [31], we performed the analyses for all most recent pregnancies, irrespective of gravidity, and for primigravidous couples only. In the latter analysis, the results are not influenced by the reproductively unhealthy worker effect bias either [32].
3. Results The composition of the study population with basic characteristics of the final participants is shown in Fig. 1. Of the 13,005 employees and company owners invited for the study, at least 837 men and 300 women proved to be ineligible. We received completed questionnaires from 1222 male and 924 female greenhouse workers and from 1235 men and 1058 women in the
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
109
Fig. 1. Composition of the study population, response, and pregnancies among the final participants. a Not a member of the trade union anymore, no company anymore, other occupation, or too old.
reference group, including 752 couples in which both partners filled out the study questionnaires. Participation rates varied from 19.0% among male referents to 52.8% among female greenhouse workers. From the 4439 initial participants, 619 men and 312 women who had never tried to get pregnant were excluded. All other participants and their partners, except 140 couples for which occupation was unknown for at least one partner, were placed in one of the two exposure groups or the reference group (Table 1). Because many women were partners of greenhouse owners, the group of female greenhouse workers consisted of only 102 women who had been pregnant at least once and 15 women who tried to get pregnant but did not conceive. The group of male greenhouse workers (n = 985) consisted of greenhouse employees (53.6%) and male greenhouse owners with or without assisting female partners (46.4%). The other participants and their partners formed the reference group (n = 1545). Couples who did not get pregnant
(n = 112), couples in which neither one of the partners worked in the preconceptional period (n = 60), and couples with missing values on all outcome measures (n = 9) were excluded from the analyses. As a result, the final study population consisted of 101 couples with female greenhouse workers, 957 couples with male greenhouse workers, and 1408 referents. Their general characteristics are shown in Table 2. Most of the lifestyle factors differed between the three groups with the largest differences for smoking and alcohol consumption and maternal vitamin use. The average number of children also varied between the three groups from 2.6 for male greenhouse workers to 2.0 for the reference group and 1.7 for female greenhouse workers. Among couples with female greenhouse workers, far more primigravidea were counted (42.6%) than among couples with male greenhouse workers (10.5%) and the reference group (20.0%). Male greenhouse workers were less often part of a couple in which
Table 1 Number of couples for which occupation was known for both man and woman in the last 6 months prior to the most recent pregnancy or pregnancy attempt
One or more pregnancies Pregnancy attempt a
Female greenhouse workers (n = 117)
Male greenhouse workers (n = 985)
Reference group (n = 1545)a
102 (87.2%) 15 (12.8%)
959 (97.4%) 26 (2.6%)
1474 (95.4%) 71 (4.6%)
In total, 140 couples were missing because of unknown occupation.
110
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
Table 2 General characteristics of the final study population consisting of couples with one or more pregnancies of which the woman or the man was a greenhouse worker (exposure groups) or both partners were not exposed to pesticides (reference group) Characteristic
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Reference group (n = 1408)
Age men (years) (mean, S.E.) Age women (years) (mean, S.E.) Educational level men (low (%)) Educational level women (low (%)) Smoking men (yes (%)) Smoking women (yes (%)) Alcohol use men (yes (%)) Alcohol use women (yes (%)) Coffee consumption men (yes (%)) Coffee consumption women (yes (%)) Maternal vitamin use (yes (%)) Use of oral contraception (yes (%)) Primiparity (yes (%)) Number of children (mean, S.E.) Job status Only woman worked (%) Only man worked (%) Both worked (%)
33.0 (5.8) 30.0 (4.8) 57 (57.6) 64 (63.4) 47 (46.5) 28 (28.0) 48 (48.0) 49 (49.0) 75 (75.0) 74 (74.0) 45 (44.6) 67 (67.7) 43 (42.6) 1.7 (0.8)
32.5 (4.2) 30.2 (4.1) 468 (49.1) 426 (44.7) 312 (23.5) 225 (23.5) 727 (76.1) 488 (51.0) 849 (88.9) 758 (79.4) 143 (15.0) 554 (58.2) 100 (10.5) 2.6 (1.1)
31.1 (5.2) 28.3 (4.7) 786 (56.5) 829 (59.4) 774 (55.1) 613 (43.6) 780 (55.5) 558 (39.7) 1176 (83.8) 1086 (77.2) 233 (16.6) 826 (58.9) 280 (20.0) 2.0 (1.0)
4 (4.0%) 0 97 (96.0%)
0 693 (72.4%) 264 (27.6%)
both partners worked (27.6%) than female greenhouse workers or referents (96.0% and 63.7%, respectively). Tables 3 and 4 show the crude and adjusted odd ratios with 95% CI for each outcome measure. Table 3 presents the results for all most recent pregnancies and Table 4 the results for primigravidous couples only. We will concentrate on the results for the primigravidous couples, which are the most valid [31,32].
21 (1.5%) 490 (34.8%) 897 (63.7%)
3.4. Low birth weight The results for primigravidous couples showed an elevated risk of low birth weight among female greenhouse workers, but a slightly decreased risk of low birth weight among male greenhouse workers. However, due to the wide confidence intervals which include the null value, no firm conclusion can be drawn regarding low birth weight.
3.1. Prolonged TTP 3.5. Sex ratio Among primigravidous couples, a more than twofold increased risk of prolonged TTP was observed for female greenhouse workers in the crude analysis which was still seen, although slightly reduced, after correction for confounding. The ORs for prolonged TTP among male greenhouse workers were slightly increased as well. 3.2. Spontaneous abortion In the crude analysis among primigravidous couples, a threefold increased risk was observed for spontaneous abortion among female greenhouse workers, which was more pronounced after correction for confounding (ORwomen adjusted = 4.0; 95% CI: 1.1–14.0). Because of small numbers, we did not analyse stillbirths separately. Combining spontaneous abortion and stillbirths did not change the above-mentioned results (data not shown). 3.3. Preterm birth For male greenhouse workers, a decreased risk was observed for preterm birth among primigravidous couples (ORmen adjusted = 0.1; 95% CI: 0.0–0.5). Among female greenhouse workers, no risk of preterm birth was seen.
No increased risks of skewed sex ratio distributions were observed among male and female greenhouse workers. 3.6. Birth defects No increased risks of birth defects were observed for male and female greenhouse workers either. For all outcomes, the results for all most recent pregnancies (Table 3) were comparable to the results for primigravidous couples only, except for prolonged TTP which did not show an increased risk among female greenhouse workers and a decreased risk among male greenhouse workers. 4. Discussion For primigravidous couples only, a slightly elevated risk of prolonged time-to-pregnancy and an increased risk of spontaneous abortion were observed among female greenhouse workers, as well as a decreased risk of preterm birth among male greenhouse workers. The results for all most recent pregnancies were comparable except for TTP. Before further discussing these results, we need to address some methodological issues,
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
111
Table 3 Crude and adjusted odds ratios for reproductive outcomes among female and male greenhouse workers compared to the reference group for all most recent pregnancies Outcome variable
Reference group (n = 1408)
Crude
Adjusted
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Prolonged TTPa (>12 months) 1.0 ORb 95% CI ref
1.40 0.76–2.61
0.69 0.51–0.95
1.15c 0.61–2.17
0.68c 0.49–0.94
Spontaneous abortiond (<20 weeks) OR 1.0 95% CI ref
3.52 1.72–7.22
0.69 0.42–1.11
2.62e 1.24–5.56
0.89e 0.52–1.54
Preterm birthf (<37 weeks) OR 1.0 95% CI ref
1.14 0.57–2.31
0.47 0.35–0.64
1.14g 0.57–2.31
0.47g 0.35–0.64
Low birth weightf,h (≤2500 g) OR 1.0 95% CI ref
1.63 0.65–4.06
0.64 0.42–0.98
1.85i 0.73–4.64
0.72i 0.46–1.12
Sex-ratiof (girls) OR 95% CI
1.0 ref
0.81 0.48–1.34
0.95 0.79–1.13
0.73j 0.43–1.23
0.97j 0.80–1.17
Birth defectsf (combined) OR 1.0 95% CI ref
1.02 0.25–4.13
0.55 0.30–1.01
1.16k 0.27–4.96
0.62k 0.33–1.14
a
Number of observations in the analysis of TTP: 1097 referents, 85 female greenhouse workers, and 796 male greenhouse workers; 334 pregnancies were unintended and excluded and 154 TTPs were missing. b OR: odds ratio; CI: confidence interval. c Adjusted for age women, smoking men, and maternal vitamin use. d Number of observations in the analysis of spontaneous abortion: 1351 referents, 79 female greenhouse workers, and 899 male greenhouse workers; 141 couples were still pregnant at the time of study. e Adjusted for educational level men, maternal vitamin use, and job status. f Number of observations in the analysis of live births: 1253 referents, 65 female greenhouse workers, and 858 male greenhouse workers; 153 pregnancies ended in a spontaneous abortion, stillbirth, or other unfavorable outcome (e.g. ectopic pregnancy). g No confounders were identified for preterm birth. h Adjusted for gestational age. i Adjusted for age women and smoking women. j Adjusted for alcohol consumption women and job status. k Adjusted for age women.
such as the possibility of selection bias in several phases of the study. The overall response rate of our study was low, in particular among market stallholders and retail shopkeepers, possibly due to the selection of owners of small companies or market stalls who were too busy to fill out the questionnaire. Alternatively, our questionnaires could have been labeled as junk mail in large companies. Moreover, the company owners were mostly men, who are known to have low response rates in reproductive studies [33]. This may have caused similarly low response rates among their female partners. As selection by age was not possible for the reference group, a relatively large number of people could have been too old to see the relevance of a questionnaire on reproductive issues. However, since the study was presented as a general study on work and lifestyle, not putting any emphasis on pesticide exposure, greenhouse workers with adverse pregnancy outcomes had no other reasons to participate than cleaners, market stallholders, or retail shopkeepers with similar problems. This reduces the chances of selection bias.
Because of the low response rates, a non-response study was performed. By telephone, we successfully contacted 243 male and 81 female greenhouse workers and 200 male and 70 female cleaners and shopkeepers, of which 128 and 30 greenhouse workers and 79 and 31 cleaners and shopkeepers were willing to answer questions about the most recent pregnancy. The median TTPs among both non-responder groups were lower than the median TTPs in all participant groups. Thus, slightly more men and women with a long TTP participated in our study, both among greenhouse workers and in the reference group. Because of low prevalences of the other reproductive disorders, we were not able to compare these outcomes measures. The choice of the reference group, which consisted of men and women working as professional cleaners and in shops and market stalls, is an important point of discussion. We chose this reference group for our population of greenhouse workers because of similar educational level, socioeconomic status, and working conditions (such as standing and carrying loads) apart from pesticide exposure. The participating cleaners mainly
112
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
Table 4 Crude and adjusted odds ratios for reproductive outcomes among female and male greenhouse workers compared to the reference group for primigravidous couples only Outcome variable
Reference group (n = 280)
Crude
Adjusted
Female greenhouse workers (n = 43)
Male greenhouse workers (n = 100)
Female greenhouse workers (n = 43)
Male greenhouse workers (n = 100)
Prolonged TTPa (>12 months) ORb 1.0 95% CI ref
2.28 1.03–5.05
1.27 0.64–2.52
1.90c 0.81–4.44
1.46c 0.79–3.06
Spontaneous abortiond (<20 weeks) OR 1.0 95% CI ref
2.91 0.92–7.63
1.17 0.43–3.16
4.00e 1.14–14.01
1.17e 0.36–3.84
Preterm birthf (<37 weeks) OR 1.0 95% CI ref
0.36 0.11–1.60
0.10 0.03–0.50
0.36g 0.11–1.60
0.10g 0.03–0.50
Low birth weightf,h (≤2500 g) OR 1.0 95% CI ref
2.36 0.52–10.64
0.66 0.18–2.43
2.16i 0.47–9.97
0.64i 0.17–2.40
Sex-ratiof (girls) OR 95% CI
0.79 0.33–1.85
0.71 0.43–1.20
0.68j 0.28–1.66
0.72j 0.43–1.21
1.34 0.17–10.10
0.75 0.16–3.49
1.12k 0.13–9.86
0.87k 0.17–4.25
1.0 ref
Birth defectsf (combined) OR 1.0 95% CI ref
a Number of observations in the analysis of TTP: 218 referents, 38 female greenhouse workers, and 76 male greenhouse workers; 62 pregnancies were unintended and excluded and 29 TTPs were missing. b OR: odds ratio; CI: confidence interval. c Adjusted for age women, smoking men, and maternal vitamin use. d Number of observations in the analysis of spontaneous abortion: 259 referents, 31 female greenhouse workers, and 86 male greenhouse workers; 47 couples were still pregnant at the time of study. e Adjusted for age women and maternal vitamin use. f Number of observations in the analysis of live births: 235 referents, 24 female greenhouse workers, and 79 male greenhouse workers; 38 pregnancies ended in a spontaneous abortion, stillbirth, or other unfavorable outcome (e.g. ectopic pregnancy). g No confounders were identified for preterm birth. h Adjusted for gestational age. i Adjusted for age women and smoking women. j Adjusted for alcohol consumption women. k Adjusted for age women and educational level men.
used standard non-aggressive household detergents. A literature review did not show any evidence for diminished fertility in female professional cleaners [34] or shopworkers. Therefore, selection on outcome measures in the reference population seems unlikely. Because all data were collected by means of questionnaires, there is the probability of misclassification on outcome, exposure, and confounder variables. Most likely, the prevalence of e.g. spontaneous abortion (4%) is underreported in our study, which may have resulted in non-differential misclassification. We have no reason to assume differential misclassification on outcome measures or general population characteristics (Table 2) leading to severe information bias or residual confounding. Concerning exposure, no distinctions could be made in amounts and kinds of pesticides used, as many different pesticides were reported, some of which may not be reproduction toxic. Consequently, the effects of pesticide exposure on reproductive disorders may have been underestimated. The questions on occupation and exposures pertained to the 6 months period before pregnancy. In this
period, pesticides may influence time-to-pregnancy, conception, implantation, and early pregnancy and may cause spontaneous abortion, whereas some of the other reproductive disorders (e.g. low birth weight and preterm birth) may be susceptible to pesticide exposure later in pregnancy. In The Netherlands, however, women normally stay at work until 4 weeks before delivery. Although the analyses among primigravidous couples only were based on small numbers, we have more confidence in these results than in those of all pregnancies, in which past pregnancy experiences may have been used in planning [31]. Adjusting for parity is insufficient, because it does not take into account desired family size and results of previous pregnancies. Pregnancy history, which contains this information, was not collected in our study. However, the average number of children was higher among male greenhouse workers than among referents. This may explain the different ORs for TTP among male greenhouse workers when we analysed most recent pregnancies (OR = 0.68) or first pregnancies only (OR = 1.46). All this favors the analyses based on primigravi-
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
dous couples only, although the power in these analyses is not enough to detect small differences in reproductive disorders associated with pesticide exposure. Nevertheless, these analyses showed increased risks of spontaneous abortion and possibly of prolonged time-to-pregnancy among female greenhouse workers. Prolonged TTP indicates reproductive loss at any of several different stages, including gametogenesis, transport of gametes in both male and female reproductive tracts, fertilization, migration of zygote to uterus, implantation, and early survival of the conceptus. Pesticides may disrupt the female hormonal balance and cause any of these effects, including spontaneous abortion. Our results on spontaneous abortion are in accordance with a number of studies which reported that the risk of spontaneous abortion seemed to be increased among women occupationally exposed to pesticides and/or working in the agricultural sector [21,35]. For prolonged TTP we used a cutoff point of 12 months, in accordance with the clinical definition of subfertility. This may not be as good a measure as TTP on a continuous scale to detect adverse effects on fertility due to pesticides. Nevertheless, the results of both analyses point in the same direction of an increased risk of prolonged TTP among primigravidous greenhouse workers [36]. Sallmen et al. also showed that the effects of pesticide exposure on fertility were more severe among primiparous families compared to parous families [20]. In a recent overview, Hanke and Jurewicz point towards an association between decreased fecundability and pesticide exposure as well [17]. The results of the analyses on preterm birth surprisingly indicated a decreased risk among male greenhouse workers. It should be noted, however, that the prevalence of preterm birth in the reference group (20%) is much higher than expected (less than 10%), [37] whereas the prevalences of preterm birth among female and male greenhouse workers (8% and 3%, respectively) are within the ‘normal’ range for first pregnancies. We do not have an explanation for the high rate of preterm births in the reference group. In contrast to other studies, [17] we did not find any associations between birth defects and pesticide exposure, but our final analyses were based on very small numbers of birth defects. Another explanation could be that not all greenhouse workers in our study were exposed to relevant concentrations of pesticides. Consequently, the effects of pesticide exposure on reproductive outcomes may have been underestimated. Furthermore, greenhouse workers in The Netherlands appear to have better general health, reflected in better life expectancy in regions with high greenhouse activity, [38] which may also result in better reproductive capacity than the general population. If this were the case, the effects of greenhouse work are underestimated. In conclusion, this study may offer some evidence for the hypothesis that pesticide exposure affects human reproduction leading to spontaneous abortion and possibly to prolonged time-to-pregnancy. As a hormonal mechanism may be assumed for these reproductive failures, this study also points towards endocrine disrupting properties of pesticides used in flower greenhouses.
113
Acknowledgements We are most obliged to the participating trade union, the Chamber of Commerce, and the agricultural marketing company for providing the study sample and to the participants for filling out the questionnaires. The study was financially supported by the Netherlands Organisation for Health Research and Development; grant no. 904-63-116. References [1] Dunson DB, Colombo B, Baird DD. Changes with age in the level and duration of fertility in the menstrual cycle. Hum Reprod 2002;17(5):1399– 403. [2] Joffe M, Li Z. Male and female factors in fertility. Am J Epidemiol 1994;140(10):921–9. [3] Noord-Zaadstra BM, Looman CW, Alsbach H, Habbema JD, te Velde ER, Karbaat J. Delaying childbearing: effect of age on fecundity and outcome of pregnancy. BMJ 1991;302(6789):1361–5. [4] Kazaura M, Lie RT, Skjaerven R. Paternal age and the risk of birth defects in Norway. Ann Epidemiol 2004;14(8):566–70. [5] Hull MG, North K, Taylor H, Farrow A, Ford WC. Delayed conception and active and passive smoking. The avon longitudinal study of pregnancy and childhood study team. Fertil Steril 2000;74(4):725–33. [6] Bolumar F, Olsen J, Boldsen J. Smoking reduces fecundity: a European multicenter study on infertility and subfecundity. The European study group on infertility and subfecundity. Am J Epidemiol 1996;143(6):578–87. [7] Alderete E, Eskenazi B, Sholtz R. Effect of cigarette smoking and coffee drinking on time to conception. Epidemiology 1995;6(4):403–8. [8] Florack EI, Zielhuis GA, Rolland R. Cigarette smoking, alcohol consumption, and caffeine intake and fecundability. Prev Med 1994;23(2):175– 80. [9] Savitz DA, Schwingl PJ, Keels MA. Influence of paternal age, smoking, and alcohol consumption on congenital anomalies. Teratology 1991;44(4):429–40. [10] te Velde ER, Eijkemans R, Habbema HD. Variation in couple fecundity and time-to-pregnancy, an essential concept in human reproduction. Lancet 2000;355(9219):1928–9. [11] Golding J. Reproduction and caffeine consumption—a literature review. Early Hum Dev 1995;43(1):1–14. [12] Ma WG, Song H, Das SK, Paria BC, Dey SK. Estrogen is a critical determinant that specifies the duration of the window of uterine receptivity for implantation. Proc Natl Acad Sci USA 2003;100(5):2963–8. [13] Bretveld RW, Thomas CM, Scheepers PT, Zielhuis GA, Roeleveld N. Pesticide exposure: the hormonal function of the female reproductive system disrupted? Reprod Biol Endocrinol 2006;4:30, 30. [14] Bretveld R, Brouwers M, Ebisch I, Roeleveld N. Influence of pesticides on male fertility. Scand J Work Environ Health 2007;33(1):13–28. [15] Maroni M, Fait A, Colosio C. Risk assessment and management of occupational exposure to pesticides. Toxicol Lett 1999;107(1–3):145–53. [16] Andersen HR, Vinggaard AM, Rasmussen TH, Gjermandsen IM, Bonefeld-Jorgensen EC. Effects of currently used pesticides in assays for estrogenicity, androgenicity, and aromatase activity in vitro. Toxicol Appl Pharmacol 2002;179(1):1–12. [17] Hanke W, Jurewicz J. The risk of adverse reproductive and developmental disorders due to occupational pesticide exposure: an overview of current epidemiological evidence. Int J Occup Med Environ Health 2004;17(2):223–43. [18] de Cock J, Westveer K, Heederik D, te VE, van Kooij R. Time-to-pregnancy and occupational exposure to pesticides in fruit growers in The Netherlands. Occup Environ Med 1994;51(10):693–9. [19] Petrelli G, Figa-Talamanca I. Reduction in fertility in male greenhouse workers exposed to pesticides. Eur J Epidemiol 2001;17(7):675–7. [20] Sallmen M, Liesivuori J, Taskinen H, Lindbohm ML, Anttila A, Aalto L, et al. Time-to-pregnancy among the wives of Finnish greenhouse workers. Scand J Work Environ Health 2003;29(2):85–93.
114
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
[21] Arbuckle TE, Lin Z, Mery LS. An exploratory analysis of the effect of pesticide exposure on the risk of spontaneous abortion in an Ontario farm population. Environ Health Perspect 2001;109(8):851–7. [22] Savitz DA, Whelan EA, Kleckner RC. Self-reported exposure to pesticides and radiation related to pregnancy outcome-results from National Natality and Fetal Mortality Surveys. Public Health Rep 1989;104(5):473–7. [23] Savitz DA, Whelan EA, Kleckner RC. Effect of parents’ occupational exposures on risk of stillbirth, preterm delivery, and small-for-gestational-age infants. Am J Epidemiol 1989;129(6):1201–18. [24] Goulet L, Theriault G. Stillbirth and chemical exposure of pregnant workers. Scand J Work Environ Health 1991;17(1):25–31. [25] Pastore LM, Hertz-Picciotto I, Beaumont JJ. Risk of stillbirth from occupational and residential exposures. Occup Environ Med 1997;54(7):511–8. [26] Tielemans E, Bretveld R, Schinkel J, van Wendel de JB, Kromhout H, Gerritsen-Ebben R, et al. Exposure profiles of pesticides among greenhouse workers: implications for epidemiological studies. J Expo Sci Environ Epidemiol 2007. [27] Blatter BM, Roeleveld N, Zielhuis GA, Mullaart RA, Gabreels FJ. Spina bifida and parental occupation. Epidemiology 1996;7(2):188–93. [28] Kersemaekers WM, Roeleveld N, Zielhuis GA. Reproductive disorders among hairdressers. Epidemiology 1997;8(4):396–401. [29] Bonde JP, Joffe M, Danscher G, Apostoli P, Bisanti L, Giwercman A, et al. Objectives, designs and populations of the European Asclepios study on occupational hazards to male reproductive capability. Scand J Work Environ Health 1999;25(Suppl. 1):49–61.
[30] Rothman KJ. Computation of exact confidence intervals for the odds ratio. Int J Biomed Comput 1975;6(1):33–9. [31] Olsen J. Options in making use of pregnancy history in planning and analysing studies of reproductive failure. J Epidemiol Community Health 1994;48(2):171–4. [32] Baird DD, Wilcox AJ, Weinberg CR. Use of time-to-pregnancy to study environmental exposures. Am J Epidemiol 1986;124(3):470–80. [33] Lindbohm ML. Effects of parental exposure to solvents on pregnancy outcome. J Occup Environ Med 1995;37(8):908–14. [34] Florack EI, Zielhuis GA, Rolland R. The influence of occupational physical activity on the menstrual cycle and fecundability. Epidemiology 1994;5(1):14–8. [35] Arbuckle TE, Savitz DA, Mery LS, Curtis KM. Exposure to phenoxy herbicides and the risk of spontaneous abortion. Epidemiology 1999;10(6):752–60. [36] Bretveld R, Zielhuis GA, Roeleveld N. Time-to-pregnancy among female greenhouse workers. Scand J Work Environ Health 2006;32(5):359– 67. [37] Mozurkewich EL, Luke B, Avni M, Wolf FM. Working conditions and adverse pregnancy outcome: a meta-analysis. Obstet Gynecol 2000; 95(4):623–35. [38] Hoeymans N, Mulder YM. In: RIVM, editor. Levensverwachting in goed ervaren gezondheid 1995–1999. Bilthoven: Volksgezondheid Toekomst Verkenning, Nationale Atlas Volksgezondheid; 2004.
Reproductive Toxicology 25 (2008) 107–114
Reproductive disorders among male and female greenhouse workers Reini W. Bretveld ∗ , Mari¨ette Hooiveld, Gerhard A. Zielhuis, Annelies Pellegrino, Iris A.L.M. van Rooij, Nel Roeleveld Department of Epidemiology and Biostatistics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Received 29 November 2006; received in revised form 11 June 2007; accepted 30 August 2007 Available online 11 September 2007
Abstract The aim of this study was to evaluate reproductive disorders in male and female greenhouse workers. In 2002, data were collected from 4872 Dutch greenhouse workers and 8133 referents through postal questionnaires with detailed questions on reproductive disorders of the most recent pregnancy, lifestyle habits, and occupational exposures (e.g. pesticides) prior to conception. Different reproductive outcome measures were compared between 957 male and 101 female greenhouse workers and 1408 referents by means of logistic regression analyses. The analyses of primigravidous couples showed a slightly elevated risk of prolonged TTP (ORwomen = 1.9; 95% CI: 0.8–4.4) and an increased risk of spontaneous abortion among female greenhouse workers (ORwomen = 4.0; 95% CI: 1.1–14.0). A decreased risk of preterm birth was found among male greenhouse workers (ORmen = 0.1; 95% CI: 0.03–0.5). This study may offer some evidence for the hypothesis that pesticide exposure affects human reproduction leading to spontaneous abortion and possibly to prolonged time-to-pregnancy. © 2007 Elsevier Inc. All rights reserved. Keywords: Pesticides; Greenhouse workers; Time-to-pregnancy; Spontaneous abortion; Reproductive disorders
1. Introduction Human reproduction is a delicate process that can be influenced by many factors. Maternal, and to lesser extent, paternal age [1–4], smoking [2,5–9], alcohol and coffee intake [1,9–11], socio-economic status [2], genetic factors, and hormonal imbalance [12] can all affect a couple’s reproductive capability. Hormonal balance of sexual hormones in particular is an important factor in maintaining fertility and regulating reproductive processes. Exogenous substances, such as environmental endocrine disruptors, may disturb the hormonal balance and thereby cause reproductive disorders [13,14]. Relatively high levels of exposure to environmental endocrine disruptors in the form of pesticides occur among people working in agriculture [15]. Some pesticides are able to influence the synthesis, storage, release, recognition, or binding of hormones, which may lead to alterations in reproductive hormone levels [16].
∗ Corresponding author at: Department of Epidemiology and Biostatistics (HP 133), Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Tel.: +31 243610357; fax: +31 243613505. E-mail address: [email protected] (R.W. Bretveld).
0890-6238/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.reprotox.2007.08.005
Several studies have been performed on occupational exposure to pesticides and adverse effects on human reproduction, including endpoints such as prolonged time-to-pregnancy (TTP), spontaneous abortion or stillbirth, low birth weight, and developmental disorders. An overview by Hanke and Jurewicz [17] showed that the effects of employment in agriculture on TTP are not unequivocal, but the majority of studies suggest a relationship between pesticide exposure and decreased fertility [18–20]. Associations between occupational pesticide exposure and birth weight, prematurity, and sex-ratio are inconsistent, but parental exposure in agriculture may increase the risk of congenital malformations in offspring [17]. Studies on other reproductive disorders suggest increased risks of spontaneous abortion and stillbirths for women occupationally exposed to pesticides [21–25]. Because these studies did not provide conclusive evidence to support the hypothesis that pesticide exposure adversely affects human reproduction, we performed a retrospective cohort study in The Netherlands to examine the effects of pesticide exposure on reproduction in male and female greenhouse workers. We focused our study on flower greenhouses, where large amounts of pesticides, such as abamectine, imidacloprid, methiocarb, deltamethrin, and pirimicarb are used [26]. The average amount of pesticide use is 28 kg per hectare
108
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
per year (range of 0.06–249 kg/hectare) and a large percentage of active ingredients are classified as fungicides. The specific aim of this study was to assess which part of the reproductive process is predominantly affected by occupational pesticide exposure. 2. Materials and methods 2.1. Population We used two different approaches to obtain the study population. From the database of a major trade union, which registers information about a variety of Dutch workers, we identified 1536 male and 604 female greenhouse workers and a random sample of 2350 male and 949 female cleaners of reproductive age (born in 1956–1979) in 2002. As only a small percentage of all greenhouses workers are members of a trade union in The Netherlands, we also obtained addresses of 2414 greenhouse owners through the national database of an agricultural marketing company with coverage of more than 85%. These owners were born in 1950–1979 or their date of birth was unknown. For the reference group, addresses of 4691 shopkeepers and market stallholders were obtained through the Chamber of Commerce, which covers all companies in The Netherlands. We only selected companies in which exposure to pesticides or other reproductive toxicants was unlikely (e.g. builder’s merchants and shops for household goods). This reference group of cleaners, retail shopkeepers, and market stallholders was chosen because of assumed comparability with greenhouse workers and owners with respect to educational level, socioeconomic status, and working conditions such as standing and carrying heavy loads.
2.2. Data collection All potential participants were sent an introductory letter, an information leaflet, and a self-administered questionnaire. Owners of a company received two questionnaires: one for themselves and one for their partners, if he or she worked in the same company. Owners were also asked for addresses of employees so we could send them questionnaires as well. Alternatively, the owners could choose to personally distribute questionnaires to their employees. Only a few owners used these options. To increase response rates, reminders were sent after 2 and 6 weeks. To avoid selective non-response and information bias, the study was presented in general terms, i.e. as a study on work, lifestyle, and fertility. We focused on the most recent pregnancy to avoid dependency issues of multiple pregnancies, assuming that the last pregnancy would be remembered best.
2.3. Questionnaires The design of the questionnaires was based on the literature and on previous reproductive studies performed by our group [27–29]. Separate questionnaires were used for women and men, who were asked whether they or their female partner had ever been pregnant. If so, detailed questions were asked about work, exposure to pesticides, and lifestyle (e.g. smoking habits, coffee, and alcohol consumption) of the respondents and their partners. All questions related to a period of 6 months before the start of the most recent pregnancy. In addition, questions were asked about time-to-pregnancy and adverse pregnancy outcomes.
2.4. Outcome definitions We defined time-to-pregnancy (TTP) as the number of months of unprotected intercourse leading to pregnancy. In accordance with the clinical definition of subfertility, the cutoff point for prolonged time-to-pregnancy was 12 months. Only planned pregnancies, regardless of pregnancy outcome, were included in the TTP analyses. Spontaneous abortion was defined as natural termination of pregnancy prior to the 20th week of gestation. Fetal deaths from 20 weeks of gestation onwards were defined as stillbirths. A delivery of a live-born infant of less than 37 weeks was classified as preterm. The cutoff point for low birth weight was 2500 g. The analyses of birth weight were always adjusted for gestational age. Sex-ratio was defined as the ratio between male and female live-born infants. For birth defects, we excluded genetic defects (e.g. heriditary metabolic disorders) and minor defects, such as hypotonia or strabismus, from the analyses.
2.5. Exposure Exposure was qualified by job title reported for the 6-month-period prior to the most recent pregnancy or pregnancy attempt. All participants were classified as ‘greenhouse workers’ (greenhouse and other horticulture workers with likely exposure to pesticides), ‘referents’ (cleaners, market and retail workers, and other workers not exposed to pesticides) or ‘not working’. After classification, we selected couples with at least one person who worked and formed two different exposure groups: female greenhouse workers (without pesticide exposed partners) and male greenhouse workers (with or without assisting female partners). In the reference group neither of the partners were exposed to pesticides.
2.6. Potential confounding factors Smoking habits, alcohol and coffee consumption, maternal vitamin use, oral contraceptives in the 12 months preceding the most recent pregnancy, and primiparity were used dichotomously in the analyses. Educational level was categorized as low (no education and primary, lower vocational, or intermediate vocational education) or high (intermediate secondary, higher secondary, higher vocational, or university education). Job status for each couple was divided into three categories (only the woman worked, only the man worked, and both partners worked). Both male and female age were used as continuous variables in the analyses.
2.7. Non-response We randomly selected 150 male and 150 female greenhouse workers and 150 male and 150 female cleaners from the trade union, 200 greenhouse owners from the marketing company, and 200 shopkeepers from the Chamber of Commerce who did not return the questionnaire. These non-participants were called and asked for the reason why they did not respond and whether they had any children. When they did, we tried to ask questions about TTP and outcome of the most recent pregnancy.
2.8. Analyses Associations between each outcome measure and greenhouse work were studied by means of logistic regression analyses. Initially, crude odds ratios (OR) with 95% confidence intervals (95% CI) were calculated, for odds ratios based on small numbers (less than 10) using the mid-p exact method [30]. Following, multivariable analyses were performed with confounder correction and checks for interaction between the exposure variable and all relevant covariates listed in Table 2. We used two criteria to evaluate whether factors were potential confounders: the distribution among greenhouse workers and referents and the associations of the potential confounder with each outcome measure in the reference group. The confounder set was then evaluated for each outcome measure separately by reducing the full model with all potential confounders to gain precision without loosing validity. Only variables that did not change the OR by more than 0.05 upon removal were permanently deleted from the model. Because in non-experimental research of reproductive failure most pregnancies are planned and past pregnancy experience is used in planning [31], we performed the analyses for all most recent pregnancies, irrespective of gravidity, and for primigravidous couples only. In the latter analysis, the results are not influenced by the reproductively unhealthy worker effect bias either [32].
3. Results The composition of the study population with basic characteristics of the final participants is shown in Fig. 1. Of the 13,005 employees and company owners invited for the study, at least 837 men and 300 women proved to be ineligible. We received completed questionnaires from 1222 male and 924 female greenhouse workers and from 1235 men and 1058 women in the
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
109
Fig. 1. Composition of the study population, response, and pregnancies among the final participants. a Not a member of the trade union anymore, no company anymore, other occupation, or too old.
reference group, including 752 couples in which both partners filled out the study questionnaires. Participation rates varied from 19.0% among male referents to 52.8% among female greenhouse workers. From the 4439 initial participants, 619 men and 312 women who had never tried to get pregnant were excluded. All other participants and their partners, except 140 couples for which occupation was unknown for at least one partner, were placed in one of the two exposure groups or the reference group (Table 1). Because many women were partners of greenhouse owners, the group of female greenhouse workers consisted of only 102 women who had been pregnant at least once and 15 women who tried to get pregnant but did not conceive. The group of male greenhouse workers (n = 985) consisted of greenhouse employees (53.6%) and male greenhouse owners with or without assisting female partners (46.4%). The other participants and their partners formed the reference group (n = 1545). Couples who did not get pregnant
(n = 112), couples in which neither one of the partners worked in the preconceptional period (n = 60), and couples with missing values on all outcome measures (n = 9) were excluded from the analyses. As a result, the final study population consisted of 101 couples with female greenhouse workers, 957 couples with male greenhouse workers, and 1408 referents. Their general characteristics are shown in Table 2. Most of the lifestyle factors differed between the three groups with the largest differences for smoking and alcohol consumption and maternal vitamin use. The average number of children also varied between the three groups from 2.6 for male greenhouse workers to 2.0 for the reference group and 1.7 for female greenhouse workers. Among couples with female greenhouse workers, far more primigravidea were counted (42.6%) than among couples with male greenhouse workers (10.5%) and the reference group (20.0%). Male greenhouse workers were less often part of a couple in which
Table 1 Number of couples for which occupation was known for both man and woman in the last 6 months prior to the most recent pregnancy or pregnancy attempt
One or more pregnancies Pregnancy attempt a
Female greenhouse workers (n = 117)
Male greenhouse workers (n = 985)
Reference group (n = 1545)a
102 (87.2%) 15 (12.8%)
959 (97.4%) 26 (2.6%)
1474 (95.4%) 71 (4.6%)
In total, 140 couples were missing because of unknown occupation.
110
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
Table 2 General characteristics of the final study population consisting of couples with one or more pregnancies of which the woman or the man was a greenhouse worker (exposure groups) or both partners were not exposed to pesticides (reference group) Characteristic
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Reference group (n = 1408)
Age men (years) (mean, S.E.) Age women (years) (mean, S.E.) Educational level men (low (%)) Educational level women (low (%)) Smoking men (yes (%)) Smoking women (yes (%)) Alcohol use men (yes (%)) Alcohol use women (yes (%)) Coffee consumption men (yes (%)) Coffee consumption women (yes (%)) Maternal vitamin use (yes (%)) Use of oral contraception (yes (%)) Primiparity (yes (%)) Number of children (mean, S.E.) Job status Only woman worked (%) Only man worked (%) Both worked (%)
33.0 (5.8) 30.0 (4.8) 57 (57.6) 64 (63.4) 47 (46.5) 28 (28.0) 48 (48.0) 49 (49.0) 75 (75.0) 74 (74.0) 45 (44.6) 67 (67.7) 43 (42.6) 1.7 (0.8)
32.5 (4.2) 30.2 (4.1) 468 (49.1) 426 (44.7) 312 (23.5) 225 (23.5) 727 (76.1) 488 (51.0) 849 (88.9) 758 (79.4) 143 (15.0) 554 (58.2) 100 (10.5) 2.6 (1.1)
31.1 (5.2) 28.3 (4.7) 786 (56.5) 829 (59.4) 774 (55.1) 613 (43.6) 780 (55.5) 558 (39.7) 1176 (83.8) 1086 (77.2) 233 (16.6) 826 (58.9) 280 (20.0) 2.0 (1.0)
4 (4.0%) 0 97 (96.0%)
0 693 (72.4%) 264 (27.6%)
both partners worked (27.6%) than female greenhouse workers or referents (96.0% and 63.7%, respectively). Tables 3 and 4 show the crude and adjusted odd ratios with 95% CI for each outcome measure. Table 3 presents the results for all most recent pregnancies and Table 4 the results for primigravidous couples only. We will concentrate on the results for the primigravidous couples, which are the most valid [31,32].
21 (1.5%) 490 (34.8%) 897 (63.7%)
3.4. Low birth weight The results for primigravidous couples showed an elevated risk of low birth weight among female greenhouse workers, but a slightly decreased risk of low birth weight among male greenhouse workers. However, due to the wide confidence intervals which include the null value, no firm conclusion can be drawn regarding low birth weight.
3.1. Prolonged TTP 3.5. Sex ratio Among primigravidous couples, a more than twofold increased risk of prolonged TTP was observed for female greenhouse workers in the crude analysis which was still seen, although slightly reduced, after correction for confounding. The ORs for prolonged TTP among male greenhouse workers were slightly increased as well. 3.2. Spontaneous abortion In the crude analysis among primigravidous couples, a threefold increased risk was observed for spontaneous abortion among female greenhouse workers, which was more pronounced after correction for confounding (ORwomen adjusted = 4.0; 95% CI: 1.1–14.0). Because of small numbers, we did not analyse stillbirths separately. Combining spontaneous abortion and stillbirths did not change the above-mentioned results (data not shown). 3.3. Preterm birth For male greenhouse workers, a decreased risk was observed for preterm birth among primigravidous couples (ORmen adjusted = 0.1; 95% CI: 0.0–0.5). Among female greenhouse workers, no risk of preterm birth was seen.
No increased risks of skewed sex ratio distributions were observed among male and female greenhouse workers. 3.6. Birth defects No increased risks of birth defects were observed for male and female greenhouse workers either. For all outcomes, the results for all most recent pregnancies (Table 3) were comparable to the results for primigravidous couples only, except for prolonged TTP which did not show an increased risk among female greenhouse workers and a decreased risk among male greenhouse workers. 4. Discussion For primigravidous couples only, a slightly elevated risk of prolonged time-to-pregnancy and an increased risk of spontaneous abortion were observed among female greenhouse workers, as well as a decreased risk of preterm birth among male greenhouse workers. The results for all most recent pregnancies were comparable except for TTP. Before further discussing these results, we need to address some methodological issues,
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
111
Table 3 Crude and adjusted odds ratios for reproductive outcomes among female and male greenhouse workers compared to the reference group for all most recent pregnancies Outcome variable
Reference group (n = 1408)
Crude
Adjusted
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Female greenhouse workers (n = 101)
Male greenhouse workers (n = 957)
Prolonged TTPa (>12 months) 1.0 ORb 95% CI ref
1.40 0.76–2.61
0.69 0.51–0.95
1.15c 0.61–2.17
0.68c 0.49–0.94
Spontaneous abortiond (<20 weeks) OR 1.0 95% CI ref
3.52 1.72–7.22
0.69 0.42–1.11
2.62e 1.24–5.56
0.89e 0.52–1.54
Preterm birthf (<37 weeks) OR 1.0 95% CI ref
1.14 0.57–2.31
0.47 0.35–0.64
1.14g 0.57–2.31
0.47g 0.35–0.64
Low birth weightf,h (≤2500 g) OR 1.0 95% CI ref
1.63 0.65–4.06
0.64 0.42–0.98
1.85i 0.73–4.64
0.72i 0.46–1.12
Sex-ratiof (girls) OR 95% CI
1.0 ref
0.81 0.48–1.34
0.95 0.79–1.13
0.73j 0.43–1.23
0.97j 0.80–1.17
Birth defectsf (combined) OR 1.0 95% CI ref
1.02 0.25–4.13
0.55 0.30–1.01
1.16k 0.27–4.96
0.62k 0.33–1.14
a
Number of observations in the analysis of TTP: 1097 referents, 85 female greenhouse workers, and 796 male greenhouse workers; 334 pregnancies were unintended and excluded and 154 TTPs were missing. b OR: odds ratio; CI: confidence interval. c Adjusted for age women, smoking men, and maternal vitamin use. d Number of observations in the analysis of spontaneous abortion: 1351 referents, 79 female greenhouse workers, and 899 male greenhouse workers; 141 couples were still pregnant at the time of study. e Adjusted for educational level men, maternal vitamin use, and job status. f Number of observations in the analysis of live births: 1253 referents, 65 female greenhouse workers, and 858 male greenhouse workers; 153 pregnancies ended in a spontaneous abortion, stillbirth, or other unfavorable outcome (e.g. ectopic pregnancy). g No confounders were identified for preterm birth. h Adjusted for gestational age. i Adjusted for age women and smoking women. j Adjusted for alcohol consumption women and job status. k Adjusted for age women.
such as the possibility of selection bias in several phases of the study. The overall response rate of our study was low, in particular among market stallholders and retail shopkeepers, possibly due to the selection of owners of small companies or market stalls who were too busy to fill out the questionnaire. Alternatively, our questionnaires could have been labeled as junk mail in large companies. Moreover, the company owners were mostly men, who are known to have low response rates in reproductive studies [33]. This may have caused similarly low response rates among their female partners. As selection by age was not possible for the reference group, a relatively large number of people could have been too old to see the relevance of a questionnaire on reproductive issues. However, since the study was presented as a general study on work and lifestyle, not putting any emphasis on pesticide exposure, greenhouse workers with adverse pregnancy outcomes had no other reasons to participate than cleaners, market stallholders, or retail shopkeepers with similar problems. This reduces the chances of selection bias.
Because of the low response rates, a non-response study was performed. By telephone, we successfully contacted 243 male and 81 female greenhouse workers and 200 male and 70 female cleaners and shopkeepers, of which 128 and 30 greenhouse workers and 79 and 31 cleaners and shopkeepers were willing to answer questions about the most recent pregnancy. The median TTPs among both non-responder groups were lower than the median TTPs in all participant groups. Thus, slightly more men and women with a long TTP participated in our study, both among greenhouse workers and in the reference group. Because of low prevalences of the other reproductive disorders, we were not able to compare these outcomes measures. The choice of the reference group, which consisted of men and women working as professional cleaners and in shops and market stalls, is an important point of discussion. We chose this reference group for our population of greenhouse workers because of similar educational level, socioeconomic status, and working conditions (such as standing and carrying loads) apart from pesticide exposure. The participating cleaners mainly
112
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
Table 4 Crude and adjusted odds ratios for reproductive outcomes among female and male greenhouse workers compared to the reference group for primigravidous couples only Outcome variable
Reference group (n = 280)
Crude
Adjusted
Female greenhouse workers (n = 43)
Male greenhouse workers (n = 100)
Female greenhouse workers (n = 43)
Male greenhouse workers (n = 100)
Prolonged TTPa (>12 months) ORb 1.0 95% CI ref
2.28 1.03–5.05
1.27 0.64–2.52
1.90c 0.81–4.44
1.46c 0.79–3.06
Spontaneous abortiond (<20 weeks) OR 1.0 95% CI ref
2.91 0.92–7.63
1.17 0.43–3.16
4.00e 1.14–14.01
1.17e 0.36–3.84
Preterm birthf (<37 weeks) OR 1.0 95% CI ref
0.36 0.11–1.60
0.10 0.03–0.50
0.36g 0.11–1.60
0.10g 0.03–0.50
Low birth weightf,h (≤2500 g) OR 1.0 95% CI ref
2.36 0.52–10.64
0.66 0.18–2.43
2.16i 0.47–9.97
0.64i 0.17–2.40
Sex-ratiof (girls) OR 95% CI
0.79 0.33–1.85
0.71 0.43–1.20
0.68j 0.28–1.66
0.72j 0.43–1.21
1.34 0.17–10.10
0.75 0.16–3.49
1.12k 0.13–9.86
0.87k 0.17–4.25
1.0 ref
Birth defectsf (combined) OR 1.0 95% CI ref
a Number of observations in the analysis of TTP: 218 referents, 38 female greenhouse workers, and 76 male greenhouse workers; 62 pregnancies were unintended and excluded and 29 TTPs were missing. b OR: odds ratio; CI: confidence interval. c Adjusted for age women, smoking men, and maternal vitamin use. d Number of observations in the analysis of spontaneous abortion: 259 referents, 31 female greenhouse workers, and 86 male greenhouse workers; 47 couples were still pregnant at the time of study. e Adjusted for age women and maternal vitamin use. f Number of observations in the analysis of live births: 235 referents, 24 female greenhouse workers, and 79 male greenhouse workers; 38 pregnancies ended in a spontaneous abortion, stillbirth, or other unfavorable outcome (e.g. ectopic pregnancy). g No confounders were identified for preterm birth. h Adjusted for gestational age. i Adjusted for age women and smoking women. j Adjusted for alcohol consumption women. k Adjusted for age women and educational level men.
used standard non-aggressive household detergents. A literature review did not show any evidence for diminished fertility in female professional cleaners [34] or shopworkers. Therefore, selection on outcome measures in the reference population seems unlikely. Because all data were collected by means of questionnaires, there is the probability of misclassification on outcome, exposure, and confounder variables. Most likely, the prevalence of e.g. spontaneous abortion (4%) is underreported in our study, which may have resulted in non-differential misclassification. We have no reason to assume differential misclassification on outcome measures or general population characteristics (Table 2) leading to severe information bias or residual confounding. Concerning exposure, no distinctions could be made in amounts and kinds of pesticides used, as many different pesticides were reported, some of which may not be reproduction toxic. Consequently, the effects of pesticide exposure on reproductive disorders may have been underestimated. The questions on occupation and exposures pertained to the 6 months period before pregnancy. In this
period, pesticides may influence time-to-pregnancy, conception, implantation, and early pregnancy and may cause spontaneous abortion, whereas some of the other reproductive disorders (e.g. low birth weight and preterm birth) may be susceptible to pesticide exposure later in pregnancy. In The Netherlands, however, women normally stay at work until 4 weeks before delivery. Although the analyses among primigravidous couples only were based on small numbers, we have more confidence in these results than in those of all pregnancies, in which past pregnancy experiences may have been used in planning [31]. Adjusting for parity is insufficient, because it does not take into account desired family size and results of previous pregnancies. Pregnancy history, which contains this information, was not collected in our study. However, the average number of children was higher among male greenhouse workers than among referents. This may explain the different ORs for TTP among male greenhouse workers when we analysed most recent pregnancies (OR = 0.68) or first pregnancies only (OR = 1.46). All this favors the analyses based on primigravi-
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
dous couples only, although the power in these analyses is not enough to detect small differences in reproductive disorders associated with pesticide exposure. Nevertheless, these analyses showed increased risks of spontaneous abortion and possibly of prolonged time-to-pregnancy among female greenhouse workers. Prolonged TTP indicates reproductive loss at any of several different stages, including gametogenesis, transport of gametes in both male and female reproductive tracts, fertilization, migration of zygote to uterus, implantation, and early survival of the conceptus. Pesticides may disrupt the female hormonal balance and cause any of these effects, including spontaneous abortion. Our results on spontaneous abortion are in accordance with a number of studies which reported that the risk of spontaneous abortion seemed to be increased among women occupationally exposed to pesticides and/or working in the agricultural sector [21,35]. For prolonged TTP we used a cutoff point of 12 months, in accordance with the clinical definition of subfertility. This may not be as good a measure as TTP on a continuous scale to detect adverse effects on fertility due to pesticides. Nevertheless, the results of both analyses point in the same direction of an increased risk of prolonged TTP among primigravidous greenhouse workers [36]. Sallmen et al. also showed that the effects of pesticide exposure on fertility were more severe among primiparous families compared to parous families [20]. In a recent overview, Hanke and Jurewicz point towards an association between decreased fecundability and pesticide exposure as well [17]. The results of the analyses on preterm birth surprisingly indicated a decreased risk among male greenhouse workers. It should be noted, however, that the prevalence of preterm birth in the reference group (20%) is much higher than expected (less than 10%), [37] whereas the prevalences of preterm birth among female and male greenhouse workers (8% and 3%, respectively) are within the ‘normal’ range for first pregnancies. We do not have an explanation for the high rate of preterm births in the reference group. In contrast to other studies, [17] we did not find any associations between birth defects and pesticide exposure, but our final analyses were based on very small numbers of birth defects. Another explanation could be that not all greenhouse workers in our study were exposed to relevant concentrations of pesticides. Consequently, the effects of pesticide exposure on reproductive outcomes may have been underestimated. Furthermore, greenhouse workers in The Netherlands appear to have better general health, reflected in better life expectancy in regions with high greenhouse activity, [38] which may also result in better reproductive capacity than the general population. If this were the case, the effects of greenhouse work are underestimated. In conclusion, this study may offer some evidence for the hypothesis that pesticide exposure affects human reproduction leading to spontaneous abortion and possibly to prolonged time-to-pregnancy. As a hormonal mechanism may be assumed for these reproductive failures, this study also points towards endocrine disrupting properties of pesticides used in flower greenhouses.
113
Acknowledgements We are most obliged to the participating trade union, the Chamber of Commerce, and the agricultural marketing company for providing the study sample and to the participants for filling out the questionnaires. The study was financially supported by the Netherlands Organisation for Health Research and Development; grant no. 904-63-116. References [1] Dunson DB, Colombo B, Baird DD. Changes with age in the level and duration of fertility in the menstrual cycle. Hum Reprod 2002;17(5):1399– 403. [2] Joffe M, Li Z. Male and female factors in fertility. Am J Epidemiol 1994;140(10):921–9. [3] Noord-Zaadstra BM, Looman CW, Alsbach H, Habbema JD, te Velde ER, Karbaat J. Delaying childbearing: effect of age on fecundity and outcome of pregnancy. BMJ 1991;302(6789):1361–5. [4] Kazaura M, Lie RT, Skjaerven R. Paternal age and the risk of birth defects in Norway. Ann Epidemiol 2004;14(8):566–70. [5] Hull MG, North K, Taylor H, Farrow A, Ford WC. Delayed conception and active and passive smoking. The avon longitudinal study of pregnancy and childhood study team. Fertil Steril 2000;74(4):725–33. [6] Bolumar F, Olsen J, Boldsen J. Smoking reduces fecundity: a European multicenter study on infertility and subfecundity. The European study group on infertility and subfecundity. Am J Epidemiol 1996;143(6):578–87. [7] Alderete E, Eskenazi B, Sholtz R. Effect of cigarette smoking and coffee drinking on time to conception. Epidemiology 1995;6(4):403–8. [8] Florack EI, Zielhuis GA, Rolland R. Cigarette smoking, alcohol consumption, and caffeine intake and fecundability. Prev Med 1994;23(2):175– 80. [9] Savitz DA, Schwingl PJ, Keels MA. Influence of paternal age, smoking, and alcohol consumption on congenital anomalies. Teratology 1991;44(4):429–40. [10] te Velde ER, Eijkemans R, Habbema HD. Variation in couple fecundity and time-to-pregnancy, an essential concept in human reproduction. Lancet 2000;355(9219):1928–9. [11] Golding J. Reproduction and caffeine consumption—a literature review. Early Hum Dev 1995;43(1):1–14. [12] Ma WG, Song H, Das SK, Paria BC, Dey SK. Estrogen is a critical determinant that specifies the duration of the window of uterine receptivity for implantation. Proc Natl Acad Sci USA 2003;100(5):2963–8. [13] Bretveld RW, Thomas CM, Scheepers PT, Zielhuis GA, Roeleveld N. Pesticide exposure: the hormonal function of the female reproductive system disrupted? Reprod Biol Endocrinol 2006;4:30, 30. [14] Bretveld R, Brouwers M, Ebisch I, Roeleveld N. Influence of pesticides on male fertility. Scand J Work Environ Health 2007;33(1):13–28. [15] Maroni M, Fait A, Colosio C. Risk assessment and management of occupational exposure to pesticides. Toxicol Lett 1999;107(1–3):145–53. [16] Andersen HR, Vinggaard AM, Rasmussen TH, Gjermandsen IM, Bonefeld-Jorgensen EC. Effects of currently used pesticides in assays for estrogenicity, androgenicity, and aromatase activity in vitro. Toxicol Appl Pharmacol 2002;179(1):1–12. [17] Hanke W, Jurewicz J. The risk of adverse reproductive and developmental disorders due to occupational pesticide exposure: an overview of current epidemiological evidence. Int J Occup Med Environ Health 2004;17(2):223–43. [18] de Cock J, Westveer K, Heederik D, te VE, van Kooij R. Time-to-pregnancy and occupational exposure to pesticides in fruit growers in The Netherlands. Occup Environ Med 1994;51(10):693–9. [19] Petrelli G, Figa-Talamanca I. Reduction in fertility in male greenhouse workers exposed to pesticides. Eur J Epidemiol 2001;17(7):675–7. [20] Sallmen M, Liesivuori J, Taskinen H, Lindbohm ML, Anttila A, Aalto L, et al. Time-to-pregnancy among the wives of Finnish greenhouse workers. Scand J Work Environ Health 2003;29(2):85–93.
114
R.W. Bretveld et al. / Reproductive Toxicology 25 (2008) 107–114
[21] Arbuckle TE, Lin Z, Mery LS. An exploratory analysis of the effect of pesticide exposure on the risk of spontaneous abortion in an Ontario farm population. Environ Health Perspect 2001;109(8):851–7. [22] Savitz DA, Whelan EA, Kleckner RC. Self-reported exposure to pesticides and radiation related to pregnancy outcome-results from National Natality and Fetal Mortality Surveys. Public Health Rep 1989;104(5):473–7. [23] Savitz DA, Whelan EA, Kleckner RC. Effect of parents’ occupational exposures on risk of stillbirth, preterm delivery, and small-for-gestational-age infants. Am J Epidemiol 1989;129(6):1201–18. [24] Goulet L, Theriault G. Stillbirth and chemical exposure of pregnant workers. Scand J Work Environ Health 1991;17(1):25–31. [25] Pastore LM, Hertz-Picciotto I, Beaumont JJ. Risk of stillbirth from occupational and residential exposures. Occup Environ Med 1997;54(7):511–8. [26] Tielemans E, Bretveld R, Schinkel J, van Wendel de JB, Kromhout H, Gerritsen-Ebben R, et al. Exposure profiles of pesticides among greenhouse workers: implications for epidemiological studies. J Expo Sci Environ Epidemiol 2007. [27] Blatter BM, Roeleveld N, Zielhuis GA, Mullaart RA, Gabreels FJ. Spina bifida and parental occupation. Epidemiology 1996;7(2):188–93. [28] Kersemaekers WM, Roeleveld N, Zielhuis GA. Reproductive disorders among hairdressers. Epidemiology 1997;8(4):396–401. [29] Bonde JP, Joffe M, Danscher G, Apostoli P, Bisanti L, Giwercman A, et al. Objectives, designs and populations of the European Asclepios study on occupational hazards to male reproductive capability. Scand J Work Environ Health 1999;25(Suppl. 1):49–61.
[30] Rothman KJ. Computation of exact confidence intervals for the odds ratio. Int J Biomed Comput 1975;6(1):33–9. [31] Olsen J. Options in making use of pregnancy history in planning and analysing studies of reproductive failure. J Epidemiol Community Health 1994;48(2):171–4. [32] Baird DD, Wilcox AJ, Weinberg CR. Use of time-to-pregnancy to study environmental exposures. Am J Epidemiol 1986;124(3):470–80. [33] Lindbohm ML. Effects of parental exposure to solvents on pregnancy outcome. J Occup Environ Med 1995;37(8):908–14. [34] Florack EI, Zielhuis GA, Rolland R. The influence of occupational physical activity on the menstrual cycle and fecundability. Epidemiology 1994;5(1):14–8. [35] Arbuckle TE, Savitz DA, Mery LS, Curtis KM. Exposure to phenoxy herbicides and the risk of spontaneous abortion. Epidemiology 1999;10(6):752–60. [36] Bretveld R, Zielhuis GA, Roeleveld N. Time-to-pregnancy among female greenhouse workers. Scand J Work Environ Health 2006;32(5):359– 67. [37] Mozurkewich EL, Luke B, Avni M, Wolf FM. Working conditions and adverse pregnancy outcome: a meta-analysis. Obstet Gynecol 2000; 95(4):623–35. [38] Hoeymans N, Mulder YM. In: RIVM, editor. Levensverwachting in goed ervaren gezondheid 1995–1999. Bilthoven: Volksgezondheid Toekomst Verkenning, Nationale Atlas Volksgezondheid; 2004.