BACKGROUND: Experimental evidence suggests pesticides may be associated with hypospadias.
OBJECTIVE: Examine the association of hypospadias with residential proximity to commercial agricultural pesticide applications.
METHODS: The study population included male infants born from 1991 to 2004 to mothers residing in 8 California counties. Cases (n = 690) were ascertained by the California Birth Defects Monitoring Program; controls were selected randomly from the birth population (n = 2195). We determined early pregnancy exposure to pesticide applications within a 500-m radius of mother’s residential address, using detailed data on applications and land use. Associations with exposures to physicochemical groups of pesticides and specific chemicals were assessed using logistic regression adjusted for maternal race or ethnicity and age and infant birth year.
RESULTS: Forty-one percent of cases and controls were classified as exposed to 57 chemical groups and 292 chemicals. Despite >500 statistical comparisons, there were few elevated odds ratios with confidence intervals that excluded 1 for chemical groups or specific chemicals. Those that did were for monochlorophenoxy acid or ester herbicides; the insecticides aldicarb, dimethoate, phorate, and petroleum oils; and adjuvant polyoxyethylene sorbitol among all cases; 2,6-dinitroaniline herbicides, the herbicide oxyfluorfen, and the fungicide copper sulfate among mild cases; and chloroacetanilide herbicides, polyalkyloxy compounds used as adjuvants, the insecticides aldicarb and acephate, and the adjuvant nonyl-phenoxy-poly(ethylene oxy)ethanol among moderate and severe cases. Odds ratios ranged from 1.9 to 2.9.
CONCLUSIONS: Most pesticides were not associated with elevated hypospadias risk. For the few that were associated, results should be interpreted with caution until replicated in other study populations.
- BPA —
- British Pediatric Association
- CI —
- confidence interval
- DOC —
- date of conception
- DPR —
- Department of Pesticide Regulation
- DWR —
- Department of Water Resources
- EPA —
- US Environmental Protection Agency
- LMP —
- last menstrual period
- OR —
- odds ratio
- PLSS —
- public land survey section
What’s Known on This Subject:
Some studies suggest a contribution of environmental exposures such as pesticides to risk of hypospadias, whereas others do not. One of the challenges that has limited current knowledge is the lack of detailed exposure data.
What This Study Adds:
This study examined a more detailed assessment of exposure to pesticides than previous studies. Exposure assignments, whether to groups of chemicals, specific chemicals, or a composite involving a number of chemicals, showed a general lack of association with hypospadias.
Hypospadias is a congenital malformation in which the urethral opening is on the ventral side of the penis rather than the tip. It affects 4 to 6 male infants per 1000 male births.1–3 In most cases the cause is unknown.4,5 The public health impact is significant, given that it usually involves surgical correction and may be associated with impaired sexual function and psychosocial difficulties later in life.6,7
Hypospadias can be induced experimentally by exposures that interfere with androgen and estrogen synthesis and signaling pathways during sexual differentiation.8–11 Thus, concern exists that endocrine disruptors (ie, exogenous substances that interfere with hormones) may cause hypospadias in humans, although the applicability of animal studies to humans is uncertain. Pesticides, some of which are endocrine disruptors, have received the most attention. A meta-analysis indicated that maternal occupational pesticide exposure was associated with a modest increase in hypospadias risk (odds ratio [OR] 1.4, 95% confidence interval [CI] 1.0–1.8).12 Three subsequent large maternal occupation–based studies have not suggested elevated risk.13–15 Studies examining maternal serum levels of persistent pesticides DDE16–20 and chlordane21 and 1 study that estimated exposure based on residential proximity to certain crops22 also do not provide evidence for association. These studies are limited by relatively crude measures of exposure, small sample sizes, or investigation of a limited set of specific compounds. Few studies have investigated pesticide exposures in pregnant women’s residential environments as possible risks for abnormal fetal development, in particular hypospadias.
We examined whether residential proximity to commercial pesticide applications during early pregnancy was associated with elevated hypospadias risk. Analyses incorporated a large range of pesticides, in the spirit of being as comprehensive as possible with regard to capturing this aspect of the “exposome.” The study population was from the Central Valley of California, one of the highest pesticide use areas in the United States.
The study population included all male infants born from 1991 to 2004 to mothers who were residents of 8 California Central Valley counties (Fresno, Kern, Kings, Madera, Merced, San Joaquin, Stanislaus, Tulare). California Birth Defects Monitoring Program staff actively ascertained cases by reviewing medical records at hospitals and genetic centers in study counties.23
Cases were classified by severity based on reported anatomic position of the urethral meatus. Mild cases were those for which the meatus was limited to the coronal or glanular penis (British Pediatric Association [BPA] codes 752.605, 752.625), moderate if it was on the penile shaft, severe if it was at the penoscrotal junction or perineal area (BPA codes 752.606, 752.607, 752.626, 752.627), and not otherwise specified if the location was insufficiently described (BPA codes 752.600, 752.620). Assignment of severity was finalized based on review by a medical geneticist (E.J.L. or Dr Cynthia Curry).24 Mild cases without chordee (ie, no penile curvature, BPA 752.605) were not ascertained during the study period, except for 2004; thus, mild cases in this study are primarily those with chordee. Cases with known single-gene disorders or chromosomal abnormalities were excluded. Cases were linked with birth certificates using personal identifiers. A total of 772 cases were available for study.
The underlying study population included 415 347 nonmalformed (ie, had no major congenital malformations) liveborn male infants. We randomly selected 2460 as male controls, in proportion to the underlying birth population for each year, to give an approximate 3:1 ratio of controls to cases.
Information on the following covariates was derived from birth certificates: maternal race or ethnicity, education, age, and parity; plurality; and infant birth weight and gestational age at delivery. Birth certificate accession numbers were used to request access to maternal residential addresses at delivery.
Pesticide Exposure Assessment
We assigned a time window of exposure for each study subject relative to reported date of conception (DOC), derived from reported last menstrual period (LMP) plus 14 days. The time window evaluated was 1 to 14 weeks of embryonic age (ie, 1–98 days after DOC). This window approximately encompasses the time of genital tubercle development (the anlage of the penis) and urethral development and closure,25 as well as several weeks preceding it, given the potential physiologic persistence of some pesticides. LMP was derived primarily from birth certificates, with replacement from medical record data for 35 cases and 45 controls. LMP was considered missing if it resulted in calculated gestational age ≥308 days (44 weeks). We imputed LMP for 31 cases and 201 controls as the median among cases or controls with nonmissing LMP (ie, 275 days for cases, 277 for controls). Pesticide data were available starting in 1991. Therefore, deliveries that occurred in 1991 but had a DOC in 1990 were excluded (47 cases, 155 controls).
The California Environmental Health Tracking Program Geocoding Service geocoded subjects’ addresses.26 Geocoding was successful for 95.2% of cases (690/725) and 95.6% of controls (2203/2305).
To estimate pesticide applications, we obtained Pesticide Use Reporting records from the California Department of Pesticide Regulation (DPR) describing agricultural pesticide applications within the study area after January 1, 1991. These data are spatially referenced to public land survey sections (PLSSs). Following the method of Rull and Ritz,27 we spatially refined PLSS polygons through the overlay of matched land use survey field polygons provided by the California Department of Water Resources (DWR); that is, we refined the pesticide application to a specific field polygon, which is smaller than the 1-square-mile area of the PLSS polygon. We matched each DPR record to the land use survey conducted closest in time to the application date (DWR surveys are conducted roughly every 5–7 years in each county). Matching is based on location and crop type, as specified in both the DPR and DWR records. Infrequently rotated crops, such as orchard crops and vineyards, were matched 1-to-1, whereas frequently rotated crops, such as field and truck crops, were grouped together in a single category, and nonagricultural land uses were subtracted from PLSS polygons when no DPR crop types were matched to available DWR polygons. Of the total applications (and active ingredient poundage) recorded by the DPR spanning 1991 to 2004, 86.7% (91.8% by poundage) were successfully linked to DWR land use survey polygons: 31.7% (51.5% by poundage) were matched on individual crop, 52.3% (37.5% by poundage) were under the “frequently rotated” category, and 2.6% (2.7% by poundage) were refined, subtracting nonagricultural land use polygons from PLSS polygons. For the remaining 13.3% of applications (8.2% by poundage), no field polygon in the specified PLSS grid was identified with an appropriate DWR land use survey or land use polygon, so no spatial refinement was possible; that is, we used the spatial reference of the 1-square-mile PLSS grid for these applications. We determined temporal proximity by comparing dates of applications in the DPR data set (accurate within a few days) with the time window of exposure for each subject.
To assign exposure, we used the California Environmental Health Tracking Program Pesticide Linkage Tool, a custom-developed Java (Oracle, Redwood Shores, CA) application that incorporates the GeoTools Java GIS Toolkit, version 2.7 (open source, http://geotools.codehaus.org/) for geographic information system data management and spatial analysis.28 We calculated the sum of pounds of pesticides used during the relevant time window within a 500-m radius of a geocoded point,29 intersecting DWR land use or PLSS polygons with the buffer and assuming homogeneous distribution of pesticides within each of these polygons.
Selection of Pesticide Compounds
We assessed whether subjects were exposed to 566 individual chemicals and 71 physicochemical groupings of chemicals that had the same chemical classification and proven or putative mechanism of action (eg, organophosphates) and were applied at >100 lb in any county in any study year.30 Low-toxicity chemicals such as biopesticides (eg, microbial pesticides, soaps, essential oils), low-toxicity inorganic compounds (eg, sulfur), and other compounds determined by the US Environmental Protection Agency (EPA) to have low toxicity31 were excluded. In addition, compounds were flagged as having reproductive or developmental toxicity based on the California Proposition 65 list32 or as endocrine disruptors,33–35 to create exposure scores. Chemicals with a US EPA–determined reference dose based on a toxicological study with a reproductive or developmental endpoint were also flagged.31
Risks associated with pesticide exposures were estimated using logistic regression models. Models included maternal race or ethnicity and age and infant birth year as potential confounders. Maternal race or ethnicity and age were included because they were associated with any pesticide exposure among controls (data not shown). Year was included to account for potential changes in the prevalence of hypospadias or use of particular pesticides over time. The models included 690 cases (170 mild; 173 moderate to severe, analyzed together because of sample size considerations; and 347 not otherwise specified) and 2195 controls with complete covariate data. To focus on comparisons likely to have the most precise estimates yet use data fully, we did the following. For pesticides with ≥15 exposed cases, risks were estimated that compared tertiles of exposure versus no exposure, based on distributions among controls. For pesticides with 5 to 14 exposed cases, risks were estimated for any versus no exposure. Risks were not estimated for pesticides with <5 exposed cases. We created overall exposure scores by summing the total number of chemicals or groups, endocrine disruptors, Proposition 65 chemicals, or EPA reproductive and developmental toxicants to which each subject was exposed. We examined the association of hypospadias with these scores specified as continuous and categorical variables (exposed subjects were divided into tertiles based on the control distributions).
Relative to controls, mothers of case infants were more likely to be non-Hispanic white, have a higher education, and be nulliparous, and case infants were more likely to be low birth weight, delivered <37 weeks, and multiple births (Table 1). Eighty-six percent (n = 590) of cases were isolated (ie, involved no other major nongenital congenital anomalies).
Subjects were exposed to 57 groups of chemicals and 292 individual chemicals during the first 14 weeks of gestation within 500 m of their residence. A total of 41.4% of cases (286/690) and 41.0% of controls (899/2195) were classified as exposed to any pesticides. The 5 groups to which controls were most frequently exposed were polyalkyloxy compounds (25%), glyphosate (22%), organophosphorus insecticides (22%), simple alcohols or ethers (19%), and petroleum derivatives (15%). The 5 chemicals to which controls were most frequently exposed were the isopropylamine salt of the herbicide glyphosate (22%), isopropyl alcohol (19%, a simple alcohol used as a solvent), nonyl-phenoxy-poly(ethylene oxy)ethanol (18%, a polyalkyloxy compound used as an adjuvant in pesticide formulations), paraquat dichloride (11%, a bipyridylium herbicide), and oxyfluorfen (11%, a diphenyl ether herbicide).
Table 2 shows the number of chemical groups and chemicals that met sample size criteria for risk estimation. Tables 3 and 4 show results for associations that had an OR ≤0.5 or ≥2.0 or for which the CI excluded 1. The only chemical groups with CIs that excluded 1 were: monochlorophenoxy acid or ester herbicides (OR 2.6 for highest tertile versus no exposure) among all cases, 2,6-dinitroaniline herbicides (OR 2.1 for lowest tertile) among mild cases, and chloroacetanilide herbicides (OR 2.8 for any exposure) and polyalkyloxy compounds (OR 1.9 for lowest tertile) among moderate to severe cases.
Specific chemicals with CIs excluding 1 were the insecticides aldicarb (OR 2.7 for lowest tertile exposure), dimethoate (OR 2.5 for lowest tertile exposure), phorate (OR 2.8 for any exposure), and paraffin-based petroleum oil (MCIAL 401) (OR 1.9 for middle tertile exposure) and adjuvant polyoxyethylene sorbitol (OR 3.4 for any exposure) among all cases; oxyfluorfen (OR 2.0 for middle tertile exposure) and copper sulfate (OR 2.9 for any exposure) among mild cases; and aldicarb (OR 2.5 for any exposure), acephate (OR 2.6 for any exposure), and nonyl-phenoxy-poly(ethylene oxy)ethanol (OR 2.0 for lowest tertile) among moderate to severe cases.
Table 5 shows ORs for sums of specific types of pesticide exposures. Results did not provide evidence that hypospadias risk increased with exposure to increasing numbers of pesticide groups, endocrine disruptors, reproductive or developmental toxicants, or Proposition 65 listed chemicals.
We examined the association of hypospadias with residential proximity to agriculture-related pesticide applications in the Central Valley of California. Exposure assignments, whether to groups of chemicals, specific chemicals, or a composite involving total number of chemicals, showed a general lack of association with hypospadias.
The general lack of association agrees with recent studies indicating that hypospadias is not associated with maternal occupations that involve pesticide exposure13–15 or serum levels of the persistent pesticides chlordane21 and DDE.16–20 We are aware of 1 other study that examined residential proximity to agriculture-related pesticide applications.22 It too did not find an association with specific chemical exposures, but its exposures were derived from annual statewide summaries of applications to specific crop types.
A few pesticide groups were associated with elevated hypospadias risk, but most were not. Among all cases, monochlorophenoxy acid or ester herbicides were associated with elevated risk. These compounds have short half-lives in humans (<48 hours) and tend to be known reproductive or developmental toxicants,36 but exposure was rare (1.6% of controls). The 2,6-dinitroaniline herbicides (which include benfluralin, trifluralin, ethalfluralin, oryzalin pendimethalin, and prodiamine) were associated with elevated risk of mild hypospadias. These compounds are moderately volatile, with vapor pressures higher than 105 mm Hg at 25°C (ie, high potential for exposure via volatilization drift), they are lipophilic, their use was frequent (11% of controls were exposed), and trifluralin,37–39 prodiamine,40 and pendimethalin40,41 are suspected endocrine disruptors. Chloracetanilide herbicides (metolachlor, acetochlor, and alachlor) and polyalkyloxy compounds used as adjuvants were associated with elevated risk of moderate to severe hypospadias. Many of the chemicals in these groups are suspected endocrine disruptors.34,39–42
Results for several specific chemicals also had elevated ORs and CIs that excluded 1. Among all cases, insecticides aldicarb, dimethoate, phorate, and paraffin-based petroleum oil and the adjuvant polyoxyethylene sorbitol were associated with elevated risk. The first 2 chemicals are known endocrine disruptors and reproductive or developmental toxicants.34,39 Phorate is an organophosphate, and polyoxyethylene sorbitol is a polyalkyloxy compound; phorate is a reproductive and developmental toxicant. Paraffin-based petroleum oil was part of the petroleum distillates group, one of the highest-use groups in California, with typically 20 to 30 million pounds used annually. Among mild cases, oxyfluorfen and copper sulfate were associated with elevated risk; neither is a known endocrine disruptor or reproductive or developmental toxicant.34,39 Among moderate and severe cases, aldicarb, acephate, and nonyl-phenoxy-poly(ethylene oxy)ethanol were associated with elevated risk, and all 3 are known endocrine disruptors.34,43,44 Among these chemicals, only 2 had exposure >5% among the controls: 11% were exposed to oxyfluorfen and 18% to nonyl-phenoxy-poly(ethylene oxy)ethanol. These compounds have not previously been reported to be associated with hypospadias.
Our study has several strengths, including the large sample size, population-based design, careful case ascertainment, and exposure assessment that was highly detailed, was spatially and temporally specific, and captured a broad spectrum of pesticides. Exposure assessment was not subject to recall bias, because it was derived from addresses. We expect selection bias to be minimal, given that exposure assessment was successful for 95% of subjects. Several limitations also merit discussion. Degree of severity was not available for many cases. Exposure assessment was based on address at delivery rather than conception. Misclassification of exposure could have occurred for women who moved. If moving is unrelated to case status, results would be biased toward the null; if not, the direction is unpredictable. Our assessment of residential proximity to pesticide applications was thorough but does not take into account other factors such as qualities of the pesticides and individuals’ metabolism or behaviors that would affect exposures (eg, chemical half-lives and vapor pressure, wind patterns, cumulative exposures over time, an individual’s ability to metabolize chemicals, and other exposure sources such as occupation or home use). However, it is also notable that most pesticides are prone to drift and detectable in air samples at locations beyond the application site,45,46 and residential proximity to pesticide-treated fields has been associated with household dust and urine levels.47–50 For many exposures, analysis of dose was not feasible because of limited sample size. For those for which dose was analyzed, results tended to suggest stronger associations at lower rather than higher exposure levels; the explanation for this finding is uncertain. However, it is known that endocrine disruptors may exert more potent affects at lower rather than higher doses and have nontraditional dose–response curves.51 The breadth of chemicals we examined was a strength in that we captured as much of the pesticide “exposome” as possible but also a limitation (from a multiple-comparison perspective). Our risk score approach was an attempt to capture cumulative effects of exposure to multiple chemicals, but it was somewhat limited by its simplicity (ie, scores were sums of exposures, not weighted by other properties such as dose or toxicity).
As noted earlier, few previous studies have carefully studied the association of hypospadias with pesticide exposure. Results have been mixed. Our study had some statistically significant findings, but they have not been reported previously and need confirmation. In addition, studies with more direct measurement of pesticide exposure would be helpful, but we acknowledge that they are difficult to conduct.
We consider this study to be hypothesis-generating, given that we evaluated exposures to all reported pesticide applications in the study area rather than limiting evaluation to those for which we had prior knowledge about potential effects on urogenital development. Such effects have not been evaluated previously for most of the studied pesticides. Because of the hypothesis-generating approach, we conducted >500 statistical comparisons. The number of statistically precise findings we highlight is within the number of false positive results one would expect to observe under typical type I error considerations. We tried to maximize power of our comparisons and reduce their number by imposing sample size criteria. Because of the novel nature of our study and because we limited the number of risk estimates based on sample size criteria, we did not correct our results for multiple comparisons or restrict our presentation of results to strict criteria of statistical significance. That said, our “positive” results should be interpreted with caution and need replication in other populations. It should also be reiterated that results did not suggest an association of hypospadias with most of the studied pesticides. Among the few that were associated, exposure tended to be infrequent in our study area, which is known for high pesticide usage.
This study of residential proximity to agriculture-related pesticide applications, which encompassed a broad range of specific chemicals, showed a general lack of association with hypospadias.
We thank the California Department of Public Health Maternal, Child and Adolescent Health Division for providing data. This work would not have been possible without the tremendous intellectual contribution of our dear colleague Craig Wolff, who sadly passed away during the course of this research. We also thank Dr Cynthia Curry for her contributions to case review for some of the study subjects. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the California Department of Public Health.
- Accepted August 26, 2013.
- Address correspondence to Suzan L. Carmichael, PhD, Department of Pediatrics, Division of Neonatal and Developmental Medicine, Stanford University, 1265 Welch Road, Room X111, Stanford, CA 94305-5415. E-mail:
Dr Carmichael conceptualized and designed the study and drafted the initial manuscript; Dr Yang carried out the analyses and reviewed and revised the manuscript; Drs Wolff and Roberts conducted the pesticide exposure assessment and reviewed and revised the manuscript; Dr Kegley developed the pesticide exposure matrix and reviewed and revised the manuscript; Drs Guo and English assisted with the pesticide exposure assessment and reviewed and revised the manuscript; Dr Lammer conducted clinical review of cases and reviewed and revised the manuscript; Dr Shaw assisted with the study design and reviewed and revised the manuscript, and all authors approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: This project was partially supported by NIH R01 ES017060 and CDC 6U01DD000489. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclosure.
- Paulozzi LJ,
- Erickson JD,
- Jackson RJ
- Gray LE Jr,
- Ostby J,
- Monosson E,
- Kelce WR
- Noriega NC,
- Ostby J,
- Lambright C,
- Wilson VS,
- Gray LE Jr
- Nassar N,
- Abeywardana P,
- Barker A,
- Bower C
- Longnecker MP,
- Klebanoff MA,
- Brock JW,
- et al
- Shoenwolf GC,
- Bleyl SB,
- Brauer PR,
- Francis-West PH
- ↵California Environmental Health Tracking Program. Geocoding Service. 2012. Available at: http://cehtp.org/geocoding. Accessed May 1, 2013
- ↵California Environmental Health Tracking Program. Agricultural Pesticide Mapping Tool. 2012. Available at: http://cehtp.org/p/tools_pesticide. Accessed May 1, 2013
- Kegley SE,
- Hill BR,
- Orme S,
- Choi AH
- ↵Pesticide Chemical Search EPA. Available at: www.epa.gov/pesticides/chemicalsearch. Accessed May 1, 2013
- ↵California Office of Environmental Health Hazard Assessment. Proposition 65. 2012. Available at: www.oehha.ca.gov/prop65.html. Accessed May 1, 2013
- ↵European-Commission. Towards the Establishment of a Priority List of Substances for Further Evaluation of Their Role in Endocrine Disruption, Appendix 1. 2000. Available at: http://ec.europa.eu/environment/archives/docum/pdf/bkh_annex_01.pdf. Accessed May 1, 2013
- Keith LH
- ↵Colborn T. Widespread pollutants with reproductive and endocrine-disrupting effects. Available at: www.ourstolenfuture.org/basics/chemlist.htm. Accessed May 1, 2013
- ↵EPA. Toxic Chemicals Added to EPCRA Section 313 Under 1994 Chemical Expansion. Table 1: Toxicity data by category. 1994. Available at: www.epa.gov/tri/trichemicals/index.htm. Accessed May 1, 2013
- Rawlings NC,
- Cook SJ,
- Waldbillig D
- ↵European Commission. Towards the establishment of a priority list of substances for further evaluation of their role in endocrine disruption, Annex 13 (List of 146 substances with endocrine disruption classifications prepared in the Expert meeting). 2000. Available at: http://ec.europa.eu/environment/archives/docum/pdf/bkh_main.pdf. Accessed May 1, 2013
- Lemaire G,
- Mnif W,
- Pascussi JM,
- et al
- ↵Kegley SE, Katten A, Moses M. Secondhand Pesticides: Airborne Pesticide Drift in California. Pesticide Action Network. San Francisco, CA: California Rural Legal Assistance Foundation, Pesticide Education Center, and Californians for Pesticide Reform. 2003. Available at: www.pesticideresearch.com/site/docs/SecondhandPcides.pdf. Accessed May 1, 2013
- Copyright © 2013 by the American Academy of Pediatrics