Objective
We sought to determine if periconceptional exposure to agrichemicals was associated with the development of gastroschisis.
Study Design
We conducted a retrospective, case-controlled study using Washington State Birth Certificate and US Geological Survey databases. Cases included all live-born singleton infants with gastroschisis. Distance between a woman’s residence and site of elevated exposure to agrichemicals was calculated. Multivariate regression was used to estimate the association between surface water concentrations of agrichemicals and the risk of gastroschisis.
Results
Eight hundred five cases and 3616 control subjects were identified. Gastroschisis occurred more frequently among those who resided <25 km from a site of high atrazine concentration (odds ratio, 1.6). Risk was related inversely to the distance between the maternal residence and the closest toxic atrazine site. In multivariate analysis, nulliparity, tobacco use, and spring conception remained significant predictive factors for gastroschisis.
Conclusion
Maternal exposure to surface water atrazine is associated with fetal gastroschisis, particularly in spring conceptions.
The prevalence of gastroschisis has more than doubled over the last 30 years. Teratogens, organic chemicals, solvents, and cyclooxygenase inhibitors that are specifically related to the agricultural industry may be associated with the increased prevelance. Atrazine is an organic compound from the triazine family that is used mainly as a herbicide. It is used to stop the spread of weeds among agricultural crops. It does not tend to bioaccumulate in the environment, but its relative persistence (half-life of 35-125 days), especially in surface ground water, increases one’s exposure. Animal data have shown that atrazine has embryotoxic and embryolethal effects in rodents. Furthermore in wildlife, male frogs that are exposed to water that has been polluted with atrazine at levels of >0.1 μg/L show hermaphroditism and retarded gonadal development because of its endocrine disrupting activity. In humans, atrazine can act as an endocrine-disrupting compound with effects on the central nervous system, the endocrine system, and the immune system. 2,4-Dichlorophenoxyacetic acid, a synthetic plant growth regulator, is the most widely used herbicide in the world. Nitrate, another agrichemical, is a salt of nitric acid and, in humans, is metabolized to ammonia that ultimately can lead to the dangerous condition of methameglobinemia. Nitrite, another nitric acid salt, is used frequently for curing meats because it prevents bacterial overgrowth.
See related editorial, page 207
Washington State has >80,000 live births with 2400-3200 reported defects annually. On average, there are 43 cases of gastroschisis in Washington State per year. The economy of eastern Washington counties is predominately agricultural; overall this industry employs >170,000 persons, which is 3% of the total state population. April, aside from the winter months, has the highest mean monthly precipitation in Washington State according to the Western Regional Climate Center. Also, April is the month of highest atrazine use. We hypothesized that an increased exposure in utero to herbicides is associated with the development of fetal gastroschisis. To test this hypothesis, we performed a retrospective, case-control study of pregnancies with and without gastroschisis and evaluated their periconceptional and first-trimester maternal exposure to atrazine and other agricultural chemicals.
Materials and Methods
We used Washington State birth certificate data that were linked with hospital discharge information from all nonfederal hospitals in the Comprehensive Hospital Abstract Reporting System in Washington State. Subjects were selected from all singleton live births in Washington State between 1987 and 2006. Birth records were linked with publicly available data from the United States Geological Survey Data on surface water concentrations of atrazine, 2,4-dichlorophenoxyacetic acid, nitrites, and nitrates in Washington State from 2001-2006. The Institutional Review Board of the Department of Social and Health Services of the State of Washington approved the study.
Data were collected from the Comprehensive Hospital Abstract Reporting System file with the ICD-9 code 756.7 (gastroschisis) or 54.71 (abdominal wall defect repair). Cases included all deliveries of singleton infants with gastroschisis, excluding those cases with coexisting chromosomal anomalies, known genetic syndromes, fetal deaths, and no documented zip code at the time of delivery. Before 2003, abdominal wall defects, gastroschisis, and omphalocele had the same ICD-9 code. Taking this into account, once cases were identified, we increased the accuracy of the diagnosis by performing a brief review of the patients’ hospitalization charts and outpatient records. We made study inclusion decisions based on the combination of ICD-9 codes and clinical evidence, including the results of amniocentesis, fetal echocardiograms, and the descriptive findings at the time of delivery.
Control subjects were selected randomly from all Washington State singleton live births in a ratio of 4 controls per case and frequency matched to cases by year of birth. Gestational age was determined by maternal last menstrual period and calculated in weeks of gestation. Month of last menstrual period was used as a proxy for the time of conception.
We collected information on parental demographic characteristics and maternal lifestyle characteristics from the birth certificate database to include age, race, smoking, and county of residence at the time of birth, longitude and latitude of primary residence during pregnancy, years of education, and occupation. Parental occupations were recorded in free text format on the birth certificate in response to the question “indicate type of work done during last year” and mutually exclusively categorized as agricultural, unemployed, <18 years old, and nonagricultural (all other occupations).
The data for herbicide exposure were extracted from the Washington State Department of Agriculture database. We obtained annual surface water concentrations by season of atrazine, nitrates, nitrites, and 2,4-dichlorophenoxyacetic acid from established sites in Washington State and calculated the average concentration for each site between 2001 and 2006. Data for Washington State are collected by the US Geological Survey, which operates 390 data collection sites. Surface water data were not available for the entire study population period; thus, we limited data to include 5 years. Then, we calculated the distance between each woman’s residence using longitude and latitude of her known zip code and the closest geographic site of increased exposure to each of the individual agrichemicals with the use of the Haversine formula, which is an equation important in navigation that calculates great-circle distances between 2 points on a sphere from their longitudes and latitudes. High chemical exposure concentrations in surface water were defined according to the Environmental Protection Agency standards. The exposure limit for atrazine was >3 μg/L, for nitrates was >10 mg/L, for nitrites was >1 mg/L, and for 2,4-dichlorophenoxyacetic acid was >70 μg/L. Spring was defined as March, April, and May.
Characteristics, exposures, and outcomes for cases and control subjects were compared with the use of the Student t test and χ 2 test. Multivariate logistic and linear regressions were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) to evaluate the excess risk of gastroschisis that was associated with increased surface water levels of each agrichemical; adjustment was made for confounders and known risk factors that included age, parity, tobacco use, season of conception, and log transformed distance from exposure. The season of conception was compared for the cases with the use of the Mann-Whitney test. All statistical analyses were performed with statistical software (version 10; Stata Corporation, College Station, TX).
Results
We identified 805 cases and 3616 control subjects who met the inclusion criteria, for a total study population of 4421 infants who were born in Washington State between January 1, 1987, and December 31, 2006 ( Table 1 ). We excluded 215 cases from the original population because of additional anomalies and genetic aneuploidy, which is a clinical diagnosis of omphalocele, and those cases with missing address information. We excluded 464 control subjects based on the same criteria. Within the case population, nulliparity, young maternal age, and tobacco use were associated significantly with gastroschisis. There was no difference between groups when we compared season of conception.
| Characteristic | Gastroschisis status | ||
|---|---|---|---|
| Cases (n = 805) | Control subjects (n = 3616) | P value | |
| Maternal age, y a | 24.4 ± 6.2 (13–45) | 27.4 ± 6.0 (13–46) | < .001 |
| Married, n (%) | 439 (55) | 2606 (72) | < .001 |
| Maternal race, n (%) | .005 | ||
| White | 584 (73) | 2746 (76) | |
| Black | 34 (4) | 129 (4) | |
| Asian | 49 (6) | 260 (7) | |
| Hispanic | 86 (11) | 342 (9) | |
| Native American | 32 (4) | 70 (2) | |
| Missing | 20 (2) | 69 (2) | |
| Maternal education, n (%) | < .001 | ||
| ≤12 y | 386 (48) | 1309 (36) | |
| >12 y | 214 (27) | 1432 (40) | |
| Missing | 205 (25) | 875 (24) | |
| Nulliparity, n | 441 (55) | 1508 (42) | < .001 |
| Smoked during pregnancy, n (%) | 182 (23) | 554 (15) | < .001 |
| Income, $ b | 36,204 (28,819–45,969) | 38,941 (30,780–50,544) | < .001 |
| Paternal age, y a | 27.3 ± 6.9 | 30.3 ± 6.5 | < .001 |
| Season of conception, n (%) | .13 | ||
| Winter | 165 (21) | 797 (22) | |
| Spring | 171 (21) | 715 (20) | |
| Summer | 151 (19) | 843 (23) | |
| Fall | 168 (21) | 818 (23) | |
| Missing | 150 (19) | 443 (12) | |
| Gestational age, wk a | 36.6 ± 2.9 | 39.1 ± 1.7 | < .001 |
| Missing, n (%) | 74 (9) | 345 (10) | |
| Birthweight, g a | 2786 ± 779 | 3444 ± 550 | < .001 |
| Missing, n (%) | 22 (3) | 11 (<1%) | |
| Male sex, n (%) | 443 (55) | 1886 (52) | .14 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
