Looking for Causes of Neural Tube Defects:Where Does the Environment Fit In?Lowell E. SeverHealth Risk Assessment Department, Battelle Pacific Northwest Laboratories, Richland, Washington

The neural tube defects anencephaly and spina bifida are important causes of infant mortality and morbidity. Recent studies suggest that many ofthese defects can be prevented by the periconceptional use of folic acid. At the same time, we do not know what causes most cases of neural tubedefects and there is evidence to suggest that they are etiologically heterogeneous. Additional research needs to be directed toward the role ofoccupational and environmental exposures in the etiology of these defects. Importantly, studies need to examine embryologically and anatomicallyspecific types of defects and develop accurate information on biologically relevant exposures. Exposures toward which attention needs to bedirected include organic solvents; agricultural chemicals, including pesticides; water nitrates; heavy metals such as mercury; ionizing radiation; andwater disinfection by products. We also recommend that additional attention be paid to mechanisms of neural tube closure and to the potential roleof genetic heterogeneity in the absorption and metabolism of xenobiotics and in their effects on the neural tube. - Environ Health Perspect103(Suppl 6):165-171 (1995)

Key words: anencephaly, environmental exposure, neural tube defects, parental occupation, risk factors, spina bifida

Introduction

Among many populations, defects ofneural tube closure-anencephaly, spinabifida cystica or meningomyelocele, andencephalocele-are significant causes ofinfant mortality and morbidity. Commonlyconsidered as a group and currently most

often referred to as neural tube defects(NTDs), these defects have been studiedextensively; yet their causes remainunknown. There is overwhelming evidencesupporting a multifactorial etiology for thisgroup of defects. In addition, there isincreasing evidence that these defects are

heterogeneous. As discussed below, more

attention needs to be paid to the possibilityof etiologic differences among the defectsthan often has been the case in the past.

The determination of the role of peri-conceptional folic acid supplementation inreducing the risk of NTDs must be consid-ered a major public health accomplishment(1). There is now convincing evidence thatfolic acid supplementation prior to concep-tion and early in gestation reduces both theoccurrence and recurrence of NTDs.Oakley (1) has recently introduced theterm and concept of folic acid-preventable

This paper was presented at the Symposium onPreventing Child Exposures to EnvironmentalHazards: Research and Policy Issues held 18-19March 1994 in Washington, DC. Manuscript received:December 5, 1994; accepted: May 15, 1995.

Address correspondence to Dr. Lowell E. Sever,Battelle Seattle Research Center, 4000 NE 41stStreet, Seattle, WA 98105. Telephone (206) 528-3348. Fax (206) 528-3553.

spina bifida and anencephaly. It has beenestimated that approximately 50% ofNTDs are preventable by taking a supple-ment containing 0.4 mg of folic acid priorto conception and early in gestation. Itmust be kept in mind that the success ofthis preventive intervention is likely to varyamong populations. This is supported byfindings from the case-control study of vit-amin supplementation and NTDs inIllinois and California that did not showprotective effects (2).

Although the evidence that folic acidsupplementation reduces the risk of anen-cephaly and spina bifida is compelling, itmust be emphasized that we do not knowhow folic acid works. In addition, it isimportant to realize that although thereare no overwhelming data to support arole for environmental agents in the etiol-ogy of NTDs, there are data that suggestthat extrinsic agents need to be evaluatedfurther.

My thesis is that additional attentionneeds to be paid to the potential roles ofenvironmental and occupational agents inthe etiology of NTDs. As part of thisprocess, it is necessary to increase thespecificity of categorizing both outcomesand exposures. The key issues to beaddressed in this presentation are theimportance of recognizing etiologic hetero-geneity-and examining specific types ofdefects-and improving assessment ofexposure in attempting to evaluate the roleof exogenous agents in disrupting neuraltube closure.

Heterogeneity ofNeural Tube DefectsBecause of their epidemiologic and embry-ologic similarities, anencephaly and spinabifida have commonly been referred to as a

single etiologic entity. This has beenreflected in the use of collective terms to cat-

egorize the defects: historically, central ner-

vous system malformations and, more

recently, neural tube defects. I believe thatthis approach obscures the possibility ofimportant etiologic differences betweenanencephaly and spina bifida, the two majordefects subsumed in those groupings. Inaddition, some studies of central nervous

system malformations have included non-

neural tube defects such as hydrocephalyand microcephaly among the defects thatwere so characterized. While for some pur-

poses the grouping of defects may be useful,I maintain that in trying to determine etiol-ogy, it is important to look at defects by as

specifically defined anatomic/embryologiccategories as possible.

There is a growing literature on the eti-ologic heterogeneity of the NTDs. Earlydiscussions focused on the importance oflooking at defects associated with chromo-somal or genetic factors separately fromthose expected to be due to multifactorialetiology (3-5). Later discussions focusedon epidemiologic differences betweenanencephaly and spina bifida (6) and on

potential etiologic differences betweenNTDs with and without malformations inother organ systems-often referred to as

single versus multiple (7-9) and upperversus lower NTDs (10,11). While not all

Environmental Health Perspectives 165

L. E. SEVER

investigators agree with the concept of etio-logic heterogeneity characterized by theabove discussion (12,13), it is my opinionthat the recent paper by Van Allen et al.(14) presenting evidence for multisiteneural tube closure clearly supports theimportance of looking at individual cate-gories of defects, as we stressed in an earlypaper (6).

Environmental Factors andthe Specificity of Neural TubeDefectsOne particularly important point in look-ing at environmental agents and NTDs isthat when clusters occur, rates most fre-quently increase for either anencephaly orspina bifida but not for both defects. Weexamined this topic several years ago andour findings are included in Table 1.

Of particular interest are two additionalclusters with which I have been involved.The first of these is a cluster observed in aLos Angeles County hospital in 1976 to1977. During a 3-month period, seven of923 babies were born with NTDs and sixof the cases were anencephaly; of those,three were craniorachischisis. All these babieswere born to Hispanic mothers. Rates ofanencephaly in the United States are usuallyless than one per 1000 births, with evenlower rates in Los Angeles County, althoughrates among Hispanics are higher than thoseamong non-Hispanics (6).

Of more contemporary concern is thecluster of anencephaly in Cameron Countyin south Texas. In April 1991, a nurse inBrownsville, Texas, recognized that some-thing was wrong: in one 36-hr period,three babies were born with anencephaly.The Brownsville nurse's observation wasthe first indication of this cluster. A subse-quent investigation by the Texas Depart-ment of Health and the Centers for DiseaseControl and Prevention (CDC) revealed acluster of anencephaly cases in 1990 and1991, along with endemically high rates ofthe defect in the predominantly Hispanicpopulation (15).

Table 1. Anencephaly and spina bifida cases inreported clusters of neural tube defects.

Location Anencephaly Spina bifida Ratio

Jacksonville, FL 2 12 1:6Kanawha County, WV 23 15 1.5:1Pulaski County, KY 7 3 2.3:1Pineville, KY 1 6 1:6Fitchburg, MA 1 5 1:5Antioch-Pittsburg, CA 8 2 4:1

A key feature of the south Texas clusteris that the rate of anencephaly in particularwas found to be increased. Information islacking on the anatomic details of thecases. As shown in Table 2, the rate ofother NTDs (spina bifida predominantly)appears to be more consistent over time(15). As noted earlier, anencephaly wasalso observed to have increased in severalother clusters, suggesting the possibility ofdefect-specific etiologic agents (6). Valproicacid, for example, increases the risk ofspina bifida but not anencephaly (16).

The Brownsville anencephaly clusterhas triggered increased concern regardingthe potential role(s) of environmentalagents in the etiology ofNTDs. Preliminarystudies by the Texas Department of Healthand the CDC examined the possible role ofenvironmental contamination in theCameron County cluster but failed toidentify any significant associations betweenenvironmental contaminants and NTDs(15). The community continues to be con-cerned that the excess of anencephaly isrelated either to environmental pollutionfrom the maquiladoras across the RioGrande in Matamoros, Mexico, or to pesti-cides applied to agricultural fields in thearea. It is my position that, while there isno compelling evidence for environmentalpollutants causing anencephaly, additionalresearch needs to be carried out in this areato more fully evaluate the possible role ofenvironmental factors in this cluster (17).

The Environment andNeural Tube DefectsThroughout much of the history of theepidemiologic study ofNTDs a very broadconcept of the environment was used. The"environment" was considered to embraceall nongenetic aspects of etiology. Thus,environmental factors included maternalage, parity, social class, metabolic diseases,etc. In addition, we often thought of theintrauterine environment and the extraso-matic (ambient) environment as heuristi-cally appealing characterizations. Otherthan concern about a few specific agents,

Table 2. Rates of neural tube defects by time of con-ception, Cameron County, Texas (15).

Other neuralAnencephalya tube defects

Years Cases Ratec Cases Ratec

1986-1989 23 9.6 12 5.01990-1991 24 19.7 12 7.4

"Anencephaly p-value = 0.01. bOther neural tubedefects p-value = 0.5. CRates are per 10,000 births.

for example potato blight (18), soft water(19), selected occupational groups (20), ordrugs as teratogens (16), little attentionwas paid to the possibility of agents fromthe ambient or occupational environmentsplaying etiologic roles in these defects. Toillustrate this, it is interesting to note thatin many of the classic studies of NTDs,occupation, particularly a baby's father'soccupation, was used as an indicator ofsocial class or socioeconomic status (21).In earlier papers we discussed the impor-tance of social class in the epidemiology ofNTDs and a variety of variables were usedto define it, including occupation, educa-tion, mean census track income, etc. (21).For an extensive review of the epidemiol-ogy ofNTDs, see Elwood et al. (22).

During the last 10 to 15 years, therehas been increasing concern about the pos-sible role of occupational exposures andexposures to chemicals in the ambientenvironment in the etiology of adversereproductive outcomes, including NTDs(23). While it is extremely unlikely that aspecific chemical or physical agent in theenvironment is responsible for the majorityof NTDs, evidence suggests that additionalattention needs to be paid to this issue.

In the remainder of this article I willbriefly review the studies that I believe sug-gest possible associations between occupa-tional and environmental exposures andNTDs. I will then close by consideringsome of the research needs relevant to thistopic, focusing on four issues: a) theimportance of considering specific defectsin etiologic studies; b) the importance ofimproving methods of exposure assess-ment; c) the need to explore commonmechanisms through which multipleagents could contribute to the etiology ofNTDs; and d) the need to pay additionalattention to genetic variability in responseto xenobiotic agents. I suggest that we needto develop a coordinated approach to theseissues, encouraging multicenter, multidisci-plinary studies.

Occupational StudiesWhat evidence is there to suggest thatexposure to specific chemicals or physicalagents may play a role in the etiology ofNTDs in general or anencephaly specifi-cally? I will briefly consider studies thathave examined maternal occupationalexposures, next consider paternal occupa-tional exposures, and then discuss expo-sures from the ambient environment. Someoccupations and occupational exposures

Environmental Health Perspectives166

NEURAL TUBE DEFECTSAND THE ENVIRONMENT

suggested to be associated with increasedNTD risks include:

* Maternal occupations such as nursing* Paternal occupations such as

- farmer and farmworker- painter- food and beverage processing

* Maternal occupational exposuresinclude- solvents- ionizing radiation- anesthetic gases

* Paternal occupational exposuresinclude- solvents- pesticides- ionizing radiation- mercury.

Space does not allow a thorough review ofall studies of occupational exposures; theinterested reader is referred to Elwood etal. (22) and Sever (23) for additionalinformation.

Some early studies suggested associa-tions between maternal solvent exposureand risk of central nervous system malfor-mations (24). In a subsequent study in thissame population, the association was nolonger present (25).

Roeleveld et al. (26) state that there islittle evidence that structural or functionaldefects of the central nervous system arerelated to parental occupational exposureto organic solvents but that they should beregarded as potentially hazardous to thedeveloping brain. An important considera-tion is the grouping of defects, since oftenneural tube closure defects are combinedwith central nervous system abnormalitieswith different pathogenetic mechanisms.In a review of organic solvent exposure andadverse pregnancy outcomes, Taskinen(27) interprets the literature as suggestingthat maternal exposure to organic solventsduring pregnancy may have adverse effectson offspring.A case-control study of congenital mal-

formations and parental employment inhealth care occupations was publishedrecently by Matte et al. (28). The cases andcontrols for this study came from theAtlanta Birth Defects Case-Control Study.Totals of 4915 case babies and 3027 con-trols were included in this analysis.

Mothers employed in nursing occupa-tions had statistically significant excess riskof having a child with anencephaly or spinabifida. Possible exposures to the followingspecific agents were also evaluated: anes-thetic gases, X-irradiation, and mercury.Among mothers potentially exposed to X-

irradiation, there was a statisticallysignificantly excess ofNTDs based on onlythree cases. Potential maternal exposure toanesthetic gases was also significantlyassociated with spina bifida, again based onthree cases.

The authors (28) discuss their findingsin relation to earlier work and suggest thatadditional studies of maternal nursingoccupations are indicated. They downplaythe association they found between poten-tial for maternal X-irradiation exposureand NTDs since it was based on a verysmall number of cases; this finding deservesfurther examination. In a study of congeni-tal malformations and parental occupa-tional exposure to ionizing radiation,statistically significant associations wereobserved between NTDs and parental pre-conception radiation dose (29). As in theAtlanta study, this was based on a smallnumber of cases and these findings werenot interpreted as causal. Since, however,statistically significant associations werefound with both paternal preconceptiondose and combined parental doses on thebasis of tests for trend, the issue of precon-ception occupational radiation exposureand NTDs requires further evaluation.A potential role for male-mediated fac-

tors in the development ofNTDs has beensuggested recently. Studies of occupationalexposures and birth defects often look atjob title or occupation and industry in anattempt to identify associations. For exam-ple, in a study in Montreal, an excess ofNTDs was observed among offspring offathers employed in the processing of foodand beverages (30). A similar but statisti-cally nonsignificant association wasobserved in British Columbia (31). Olshanet al. (31) did observe a statistically signifi-cant association between paternal employ-ment as a painter and spina bifida risk.

Polednak and Janerich (32) reported anincreased risk for anencephaly related topaternal employment as a farmer or farm-worker. Brender and Suarez (33) carriedout a case-control study of paternal occu-pation and anencephaly in Texas. Caseswere identified from vital records for 1981to 1986 and were compared to a series ofcontrols selected from live births duringthe same period and frequency matched tothe cases by race, ethnicity, and year ofbirth. Parental occupations were obtainedfrom the vital records, and occupationswith potential for exposure to pesticidesand solvents were identified. For the groupof paternal occupations associated with sol-vent exposure there was a significantly

increased odds ratio, as there was forpainters. Taskinen and colleagues (34)noted that paternal exposure to organicsolvents before conception may haveadverse effects on pregnancy and offspring.

Brender and Suarez (33) also examinedthe association between anencephaly andpaternal pesticide exposure. For the groupof paternal occupations with estimatedexposure to pesticides, the odds ratio was1.28, which was not statistically signifi-cantly elevated. For farmers and ranchworkers, the odds ratio was 1.73, again anincrease that was not statistically signifi-cant. Brender and Suarez (33) review someof the studies that suggest possible associa-tions between NTDs, including anen-cephaly, and pesticide exposure. I share theirassessment that additional studies need to becarried out examining this possible associa-tion. The use of solvents in pesticide formu-lations and the suggestions of associationsbetween both solvent and pesticide expo-sures and NTDs increases the importance ofsuch studies, particularly in areas whereNTD rates are high.

Louik and Mitchell (35) reportedsignificant associations between anen-cephaly and paternal mercury exposure,basing exposure status on a job-exposurematrix. This finding was based on a smallnumber of cases. Of greater interest is thefact that they also observed increased risksof NTDs associated with paternal solventexposure. For spina bifida there was astatistically significant association withxylene, and for anencephaly and spinabifida there was a statistically significantassociation with benzene. These findings,contained in an unpublished report toNIOSH (35), are consistent with the asso-ciations observed between paternal solventexposure and anencephaly reported byBrender and Suarez (33) and, similarly,with an association between paternalemployment as a painter and spina bifida,as observed by Olshan et al. (31).

Occupational studies suggest that pesti-cide exposure and exposure to solvents mayincrease the risk of having a child with anNTD. I emphasize that these studies areonly suggestive. Of particular interest is thefact that the observed associations are notrestricted to gestational exposures withdirect effects on the embryo but includepaternally mediated effects.

Ambient ExposuresThe possibility of effects on neural tubeclosure by exposure of both males andfemales to chemical agents is important as

Volume 103, Supplement 6, September 1995 167

L. E. SEVER

we turn from the occupational environmentto the ambient environment. Environ-mental exposure levels are usually lowerthan occupational exposure levels. Thus, itis important to realize that an increasedrisk may be present but not demonstrablebecause of inadequate statistical power todetect a low risk and also because of theincreased probability of exposure mis-classification at low exposure levels.Suggested associations between environ-mental contaminants and NTD risksinclude:

* Vinyl chloride* Environmental pollution* Hazardous waste sites

- Solvents- Metals

* Agricultural chemicals* Water nitrates* Organic solvents* Water disinfection by-products.

Important exposure pathways include airand water.

Ambient air and drinking water are themost common sources of environmentalexposure. From the perspective of NTDs,relatively more attention has been paid to apotential role for drinking water. This hasincluded consideration of water hardnessand mineral constitution, water nitrates,organic solvents, and water disinfection by-products. We will briefly consider an air-borne pollutant-vinyl chloride-and thendiscuss the latter three categories of drink-ing water contamination.

The possible importance of airbornepollutants in NTDs can be illustrated bystudies of vinyl chloride. Several studieshave been conducted of potential exposureto vinyl chloride monomer from industrialsources and NTDs. These studies grew outof suggestions that central nervous systemmalformation rates were high in communi-ties with polyvinyl chloride polymerizationplants (36). Studies in West Virginia (37)and Quebec (38) tested the suggested rela-tionship using case-control studies andfailed to demonstrate statistically signifi-cant associations. A more recent study inNew Jersey found increased odds ratios forcentral nervous system defects in proximityto two vinyl chloride polymerization facili-ties, but the increases were not statisticallysignificant (39). These studies can be usedto illustrate some of the epidemiologicapproaches and problems in studying envi-ronmental reproductive hazards (40).

High rates of anencephaly in Cubatao,Brazil, were suggested to be associated withenvironmental pollution in the area (41).

To our knowledge, this is the only reportregarding general environmental pollutionand NTDs. There may be parallels with theoccurrence of anencephaly in Brownsville,Texas.A recent case-control study in upstate

New York found increased odds ratios forNTDs associated with several indicators ofpotential exposure to hazardous waste sites(42). This study was based on cases fromthe New York State Congenital Malfor-mation Registry and information on toxicwaste sites from the New York StateHazardous Waste Site Inspection Program.A statistically significant odds ratio wasobserved for nervous system malformationsand residential proximity to selected haz-ardous waste sites. In addition, significantlyelevated odds ratios were observed betweennervous system malformations and sitescontaining solvents and metals. The cate-gory "nervous system malformations"included IDC-9 rubrics 740-742, which ismore inclusive than NTDs. The potentialassociations between solvents and metalsand NTDs need to be evaluated further,since, as noted, there are occupationalstudies that suggest these exposures may berelated to NTD risk (23).

Increased prevalence of NTDs hasbeen observed in areas with a high use ofagricultural chemicals (29,43). White etal. (44) studied associations between envi-ronmental chemical exposures and NTDsin New Brunswick. No associations werefound between anencephaly or spina bifidaand pesticides used in forestry. Since noinformation was available on the specificapplication of agricultural chemicals, theauthors developed an agricultural chemicalexposure opportunity index. No associa-tions were observed between anencephalyor all spina bifida cases combined andpotential exposure to agricultural chemi-cals. A significant association wasobserved, however, between the exposureindex and spina bifida without hydro-cephalus. The exposure assessment in thisstudy was quite crude, with a strongpotential for exposure misclassification.While the authors felt that the fact thatonly spina bifida without hydrocephaluswas associated with the exposure indexdetracted from the plausibility of biologi-cal meaningfulness for the association, therecent evidence regarding multiple neuraltube closure sites (14) puts this findinginto a different context. Thus, these datamay be more suggestive of a biologicallymeaningful association than has beenthought to be the case (44).

Probably more attention has been paidto a possible role for water contaminants inthe etiology of NTDs than to any otherenvironmental factor. One of the first areasof concern was nitrates in water. Nitratecontamination of water supplies may resultfrom agricultural (fertilizer) runoff, sewer-age, or industrial waste.

Epidemiologic studies do not provideconclusive evidence that pregnant womenwho consume low levels of nitrates indrinking water are at an increased risk forhaving adverse reproductive outcomes. InSouth Australia an excess of birth defectsled to a case-control study examining therelationship between maternal drinkingwater source (groundwater versus rainwa-ter) and risk of congenital malformations(45). The risk of having a malformedinfant was increased among women whodrank groundwater. Risks for centralnervous system malformations and oralclefts were particularly increased. Adose-response relationship was foundusing estimated nitrate concentrations.Strengths of the study include the com-pleteness of case ascertainment and themonitoring of water nitrates during thestudy period. Limitations include theassumption that water concentrations wereconstant during monitoring intervals andthat the subjects used the same drinkingwater source throughout pregnancy.Another assumption was that nitratesrather than some unmeasured drinkingwater contaminant were responsible. Therecould be unrecognized confounding by athird variable. For example, seasonal varia-tion in malformation risks suggests thatdietary, nutritional, or other environmentalfactors may have contributed to theincreased malformation rates. The authorsdid not believe the association was relatedto chlorination by-products, a topic thatwe discuss below.A Canadian case-control study found

an increased risk for delivering an infantwith a central nervous system malforma-tion associated with exposure to nitratesthrough water from private wells (46). Theopposite was found with drinking waterobtained from other sources. To assessexposure, the investigators analyzed nitratesin water samples collected at addresseswhere study subjects lived at the time ofdelivery. This raises the important issue ofdetermining exposures at relevant stages inpregnancy; recent studies have shown ahigh degree of residential mobility duringpregnancy. The study was also limited bythe lack of information about other

Environmental Health Perspectives168

NEURAL TUBE DEFECTSAND THE ENVIRONMENT

possible water contaminants. The oppositerisks associated with drinking water source,independent of nitrate concentration, alsosuggest that other factors contributed tothe observed effects.

Relationships between congenital mal-formations and water contamination wererecently studied in New Jersey (47-49).Based on cases from the New Jersey BirthDefects Registry and databases with infor-mation on environmental pollutant levels,a number of statistically significant associa-tions were demonstrated. Exposure assess-ments in the studies were based on analysisof existing environmental databases(47,49) and maternal interviews (48). Acomplete review of these studies is beyondthe scope of this article, but these studiessuggest the need for additional attention towater contamination and NTDs. No asso-ciations were found between NTDs andtoxic air emission data, agricultural pesti-cide applications, or proximity to hazardouswaste sites (49). Extensive studies of drink-ing water quality and birth defects, how-ever, showed associations between severaldrinking water contaminants and NTDs.On the basis of a cross-sectional study,significant associations were observedbetween carbon tetrachloride concentra-tions and total trihalomethanes and NTDs(47). Weaker associations were observedfor trichloroethylene, nitrates, and mixedwater sources. More specific informationon the breakdown by types ofNTDs is notincluded. Trihalomethanes are by-productsofwater disinfection by chlorination (50).

A case-control study was carried outthat obtained more specific information onwater exposures and other risk factors (48).Statistically significant associations werefound for total trihalomethanes and withperchloroethylene and nitrates, when thedata were adjusted for the confoundingeffects of total trihalomethanes. Craun(51) has reviewed the New Jersey studiesand suggested that the association betweenNTDs and trihalomethanes "appears ques-tionable." In addition, the studies havebeen critiqued by a panel for the U.S. EPAand the International Life SciencesInstitute Risk Sciences Institute (50). Anumber of important points regarding expo-sure assessment methods and classificationof outcomes are raised in the latter review.

Little (20) cites an unpublished PhDdissertation by Rausch, who studied preg-nancy outcome in upstate New York com-munities served by different types of watersupplies. A highly significant associationbetween anencephaly and chlorinated surface

water was observed, but there was no associa-tion with use of chlorinated groundwatercompared to use of nonchlorinatedgroundwater. The issue of potential associ-ations between water disinfection by-prod-ucts and NTDs deserves further research.

Research NeedsThe studies described above suggest thatadditional attention needs to be directedtoward a potential role for occupationaland environmental exposures to specificagents in the etiology of NTDs in somepopulations. As part of the research agendato address this important public healthtopic, I want to emphasize the importanceof considering specific types of defects,rather than lumping together all neuraltube or central nervous system defects, insearching for etiology. The recent seminalpaper by Van Allen and colleagues (14),supported by earlier discussions of the het-erogeneity of NTDs (3,4,6), establishesclearly the necessity of, at a minimum,assessing anencephaly and spina bifida sep-arately when attempting to identify envi-ronmental etiologic agents. If, in fact, thesetwo defects have distinct embryologicbases, as the evidence for the multisite clo-sure model suggests (14), then lumpingthem together would represent outcomemisclassification that would tend to reduceany association between an environmentalagent and one of the defects toward thenull. Since clusters of anencephaly seem tobe more common than clusters of spinabifida, it may be that environmental oroccupational chemicals play potentiallygreater roles in the etiology of these defectsthan of spina bifida.

The second research need is for addi-tional attention to specificity in exposureassessment. Exposure misclassification isrecognized as a potential problem in stud-ies of both occupational and ambient expo-sures. While what we are actually interestedin is a biologically effective dose to eitherthe conceptus or the gametes of the parents,depending on whether we are evaluating adirect teratogenic effect or a mutageniceffect, usually at best what we have is someindication of parental exposure. In mostcases, we do not even have a direct measureof parental exposure but a surrogate such asjob tide, occupation, or residential location,perhaps tied to an environmental database.Thus, we are often dealing with a surrogateof a surrogate. Some approaches to expo-sure assessment commonly used in occupa-tional studies and for ambient exposuresinclude:

* Occupational studies- Job titles- Work histories- Industrial hygiene monitoring- Personal dosimetry- Biomonitoring- Biological markers

* Ambient studies (population levelbased on geographic location)- History of releases/applications- Environmental monitoring- Environmental modeling

* Ambient studies (individual level)- Questionnaires- Biomonitoring- Biological markers

It is important to emphasize that manyof the limitations of current studies relateto the fact that it is often necessary to useavailable exposure information that is* non-specific with regard to an individual agentof concern. New approaches to exposureassessment using exposure biomarkers areneeded in developmental toxicity, andinformative studies are now under way(52). An additional issue of concern is theability to estimate a biologically effectivedose at a biologically relevant time. Thereis a great need for additional research sup-port for efforts in retrospective exposureassessment (53).

Questions of the specificity of exposureare particularly pertinent to considerationsof associations between defects and haz-ardous waste sites (42) or water contami-nation (48). Severe criticisms have beenleveled at some of the studies discussed ear-lier with regard to the adequacy of theexposure assessment methods used.

Implicit in the discussion of exposure isthe issue of exposure to whom. Whilemuch of the concern about occupationalreproductive hazards in particular hastended to focus on exposures of pregnantwomen during sensitive periods early inembryogenesis, the growing evidence forpaternally mediated developmental effectsstrongly suggests that exposures of bothparents need to be taken into account (54).The studies we have reviewed suggest a rolefor male-mediated effects in the etiology ofNTDs. If this is based on heritable effects,as the studies of radiation exposure andneural tubes suggest, then this indicates thateffects are not necessarily sex-specific.

It is my opinion that additional atten-tion needs to be paid to both maternal andpaternal exposures to solvents and agricul-tural chemicals, including pesticides.Studies should include approaches to esti-mated biologically relevant doses to both

Volume 103, Supplement 6, September 1995 169

L. E. SEVER

parents and to the embryo using biologicmarkers. In addition, concerns regardingthe possible effects of solvents, agriculturalchemicals, and water disinfection by-prod-ucts in the ambient environment indicatethe need for developing and applying moresophisticated approaches to exposure assess-ment than have been possible using envi-ronmental monitoring data routinelycollected for other purposes. To avoidexposure misclassification in epidemiologicstudies, the specificity and accuracy ofexposure assessment must be considered apriority in future epidemiologic research.

A third area for research priority isattention to mechanisms of neural tubeclosure. The work of Van Allen and col-leagues (14) provides insight into some ofthe embryologic and mechanistic issuesthat need attention. An important questionis, what can we learn from the effect offolic acid supplementation on neural tubeclosure that may be relevant to environ-mental and occupational agents? While ourhypothesis is that anencephaly and spinabifida are more etiologically distinct thanhas been appreciated in the past-and thatdifferent agents may play varying roles inaffecting closure-it is also logical to try tounderstand common anomalies at the cel-lular or intracellular level.

Recent interest in a potential role forfree radicals and free radical scavengersmay be one place to search for commonmechanisms. For example, Graf andPippenger (55) have recently published the

results of a preliminary study that showedincreased rates of glutathione peroxidasedeficiency in families with neural tubedefects. Glutathione peroxidase is a freeradical scavenger, and this study suggeststhat there may be an inherited vulnerabilityto peroxidation stress or reduced antioxida-tive protection mechanisms at the time ofembryogenesis. There is growing interest inthe potential relationship between thedevelopment of birth defects and free radi-cal utilization. Interestingly, in light of thediscussion of trihalomethanes, trihalo-methane chloroform frequently is found indrinking water and leads to the productionof free radicals (56).

Finally, additional attention needs to bepaid to issues of genetic heterogeneity inmaternal absorption and metabolism ofxenobiotics and in embryonic response. Aspart of the consideration of mechanisms ofaction of both protective agents such asfolic acid and teratogenic agents such asvalproic acid, it is important to evaluategenetic heterogeneity. This applies to envi-ronmental teratogens as well, and the pre-ceding discussion of glutathione peroxidaseis relevant here. Since there is a geneticcomponent in the etiology of both anen-cephaly and spina bifida and evidence issuggested for male-mediated developmen-tal effects that imply a genetic mechanism,it is important that additional researchemphasis be placed on issues such asgenetic heterogeneity in the metabolism ofxenobiotic agents.

ConclusionBirth defects are the single leading cause ofinfant mortality in the United States, andthe two major defects of neural tube clo-sure, anencephaly and spina bifida, make asignificant contribution to this total (57).The periconceptional use of folic acid isanticipated to have a dramatic impact onthe incidence of these two defects. At thesame time, some data suggest an etiologicrole for occupational and environmentalagents. It is essential that while publichealth programs for increasing folic acidintake by potentially pregnant women areinitiated, additional research be directedtoward this latter possibility. This shouldinclude large-scale multicenter case-controlstudies that pay particular attention toexamining the role of agents such as sol-vents, agricultural chemicals, and waterdisinfection by-products in the develop-ment of specific types of defects. Attentionmust be directed toward reducing mis-classification of both outcomes and expo-sures and toward mechanisms throughwhich agents affect neural tube closure.

As we try to reach our objectives ofreducing infant mortality and improvingthe health of children, it is essential that weexpand our study of the role of the envi-ronment in the etiology of birth defects.For too long those concerned with chil-dren's health have ignored the problem ofbirth defects. The need is clear, the need isgreat, and the time is now for a commit-ment to these issues.

REFERENCES

1. Oakley GP Jr. Folic acid-preventable spina bifida and anen-cephaly. JAMA 269:1292-1293 (1993).

2. Mills JL, Rhoads GG, Simpson JL, Cunningham GC, ConleyMR, Lassman MR, Walden ME, Depp R, Hoffman HJ. Theabsence of a relation between the periconceptional use of vita-mins and neural tube defects. N Engl J Med 321:430-435(1989).

3. Holmes LB, Driscoll SG, Atkins LA. Etiologic heterogeneity ofneural tube defects. N Engl J Med 294:365-369 (1976).

4. Khoury MJ, Erickson JD, James LM. Etiologic heterogeneity ofneural tube defects: clues from epidemiology. Am J Epidemiol115:538-548 (1982).

5. Simpson JL, Mills J, Rhoads GG, Cunningham GC, ConleyMR, Hoffman HJ. Genetic heterogeneity in neural tubedefects. Ann Genet 34:279-286 (1991).

6. Sever LE, Strassburg MA. Epidemiologic aspects of neural tubedefects in the United States: changing concepts and theirimportance for screening and prenatal diagnostic programs. In:Alpha-fetoprotein and Congenital Disorders (Mizejewski GJ,Porter IH, eds). New York:Academic Press, 1985;243-262.

7. Khoury MF, Erickson JD, James LM. Etiologic heterogeneityof NTD. II. Clues from family studies. Am J Hum Genet34:980-987 (1982).

8. Seller MJ, Kalousek DK. Neural tube defects: heterogeneity

and homogeneity. Am J Med Genet Suppl 2:77-87 (1986).9. Dolk H, De Wals P, Gillerot Y, Lechat MF, Ayme S, Cornel

M, Cuschieri A, Garne E, Goujard J, Laurence KM, Lillis D,Lys F, Nevin N, Owens J, Radic A, Stoll C, Stone D, Ten KateL. Heterogeneity of neural tube defects in Europe: thesignificance of site of defect and presence of other major anom-alies in relation to geographic differences in prevalence.Teratology 44:547-559 (1991).

10. Toriello HV, Higgins JV. Possible causal heterogeneity in spinabifida cystica. Am J Med Genet 21:13-20 (1985).

11. Hall JG, Friedman JM, Kenna BA, Popkin J, Jawanda M,Arnold W. Clinical, genetic, and epidemiological factors inneural tube defects. Am J Hum Genet 43:827-837 (1988).

12. Frecker MF, Fraser FC, Heneghan WD. Are "upper" and"lower" neural tube defects aetiologically different? J MedGenet 25:503-504 (1988).

13. Garabedian BH, Fraser FC. Upper and lower neural tubedefects: an alternate hypothesis. J Med Genet 30:849-851(1993).

14. Van Allen MI, Kalousek DK, Chernoff GF, Juriloff D, HarrisM, McGillivray BC, Yong S-L, Langois S, MacLeod PM,Chitayat D, Friedman JM, Wilson RD, McFadden D, PantzarJ, Ritchie S, Hall JG. Evidence for multi-site closure of theneural tube in humans. Am J Med Genet 47:723-743 (1993).

170 Environmental Health Perspectives

NEURAL TUBE DEFECTS AND THE ENVIRONMENT

15. Texas Department of Health. An investigation of a cluster ofneural tube defects in Cameron County, Texas. A report fromthe Texas Department of Health, 1992.

16. Lammer EJ, Sever LE, Oakley GP Jr. Teratogen update: val-proic acid. Teratology 35:465-473 (1987).

17. Sever LE. The conundrum of birth defect clusters. HealthEnviron Digest 7(10):4-6 (1994).

18. Renwick JH. Hypothesis. Anencephaly and spina bifida areusually preventable by avoidance of a specific but unidentifiedsubstance present in certain potato tubers. Br J Prev Soc Med26:67-88 (1972).

19. Lowe CR, Roberts CJ, Lloyd S. Malformations of central ner-vous system and softness of local water supplies. Br Med J2:357-361 (1971).

20. Little J. Are occupational carcinogens teratogenic? In:Reproductive Toxicology (Richardson M, ed). Weinheim,Germany:VCM, 1993;3-43.

21. Emanuel I, Sever LE. Questions concerning the possible associ-ation of potatoes and neural-tube defects, and an alternativehypothesis relating to maternal growth and development.Teratology 8:325-332 (1973).

22. Elwood JM, Little J, Elwood JH. Epidemiology and Control ofNeural Tube Defects. Oxford:Oxford University Press, 1992.

23. Sever LE. Congenital malformations related to occupationalreproductive hazards. Occup Med 9:471-494 (1994).

24. Holmberg PC, Nurminen MN. Congenital defects of the cen-tral nervous system and occupational factors during pregnancy.Am J Ind Med 1:167-176 (1980).

25. Kurppa K, Holmberg PC, Herenberg S, Rantala R, Riala R,Nurminen T. Screening for occupational exposures and con-genital malformations. Preliminary results from a nation-widecase-referent study. Scand J Work Environ Health 9:89-93(1983).

26. Roeleveld N, Zielhuis GA, Gabreels F. Occupational exposureand defects of the central nervous system in offspring: review.Br J Ind Med 47:580-588 (1990).

27. Taskinen HK. Effects of parental occupational exposures onspontaneous abortion and congenital malformation. Scand JWork Environ Health 16:297-314 (1990).

28. Matte TD, Mulinare J, Erickson JD. Case-control study ofcongenital defects and parental employment in health care. AmJ Ind Med 24:11-23 (1993).

29. Sever LE, Gilbert ES, Hessol NA, McIntyre J. A case-controlstudy of congenital malformations and occupational exposureto low-level ionizing radiation. Am J Epidemiol 127:226-242(1988).

30. McDonald AD, McDonald JC, Armstrong B, Cherry NM,Nolin AD, Robert D. Fathers' occupation and pregnancy out-come. BrJ Ind Med 46:329-333 (1989).

31. Olshan AF, Teschke K, Baird PA. Paternal occupation andcongenital anomalies in offspring. Am J Ind Med 20:447-475(1991).

32. Polednak AP, Janerich DT. Uses of available record systems inepidemiologic studies of reproductive toxicology. Am J IndMed 4:329-348 (1983).

33. Brender JD, Suarez L. Paternal occupation and anencephaly.Am J Epidemiol 131:517-521 (1990).

34. Taskinen H, Anttila A, Lindbohn M-L, Sallmen M, HemminkiK. Spontaneous abortions and congenital malformationsamong the wives of men occupationally exposed to organic sol-vents. Scand J Work Environ Health 15:345-352 (1989).

35. Louik C, Mitchell AA. Occupational exposures and birthdefects: final performance report RO1OH02598-OlAl.Cincinnati, OH:National Institute for Occupational Safety andHealth, 1992.

36. Infante PF. Oncogenic and mutagenic risks in communitieswith polyvinyl chloride production facilities. Ann NY Acad Sci271:49-57 (1976).

37. Edmonds LD, Anderson CE, Flynt JW, James LM. Congenitalcentral nervous system malformations and vinyl chloridemonomer exposure: a community study. Teratology17:137-142 (1978).

38. Theriault G, Iturra H, Gingras S. Evaluation of the associationbetween birth defects and exposure to ambient vinyl chloride.Teratology 27:359-370 (1983).

39. Rosenman KD, Rizzo JE, Conomos MG, Halpin GJ. Centralnervous system malformations in relation to two polyvinylchloride production facilities. Arch Environ Health44:279-282 (1989).

40. Sever LE, Hessol NA. Overall design considerations in maleand female occupational reproductive studies. Prog Clin BiolRes 160:15-47 (1984).

41. Monteleone-Neto R, Brunoni D, Laurenti R, Jorge HdM,Gotlieb SLD, Lebrao ML. Birth defects and environmentalpollution: the Cubatao example. Prog Clin Biol Res163B:65-68 (1985).

42. Geschwind SA, Stolwijk JAJ, Bracken M, Fitzgerald E, Stark A,Olsen C, Melius J. Risk of congenital malformations associatedwith proximity to hazardous waste sites. Am J Epidemiol135:1197-1207 (1992).

43. Field B, Kerr C. Herbicide use and incidence of neural-tubedefects (letter). Lancet i:1341-1342 (1979).

44. White FMM, Cohen FG, Sherman G, McCurdy R. Chemicals,birth defects and stillbirths in New Brunswick. Can Med AssocJ 138:117-124 (1988).

45. Dorsch, MM, Scragg RKR, McMichael AJ, Baghurst PA, DyerKF. Congenital malformations and maternal drinking watersupply in rural South Australia: a case-control study. Am JEpidemiol 119:473-486 (1984).

46. Arbuckle TE, Sherman GJ, Corey PH, Walters D, Lo B. Waternitrates and CNS birth defects: a population-based case-controlstudy. Arch Environ Health 43:162-167 (1988).

47. Bove F, Fulcomer M, Klotz J, Esmart J, Dufficy E, ZagraniskiR. Report on phase IV-A: public drinking water contaminationand birthweight, and selected birth defects: a cross-sectionalstudy. New Jersey Department of Health, 1992.

48. Bove F, Fulcomer M, Klotz J, Esmart J, Dufficy E, ZagraniskiR, Savrin J. Report on phase IV-B: public drinking water con-tamination and birthweight, and selected birth defects: a case-control study. New Jersey Department of Health, 1992.

49. Fulcomer MC, Bove FJ, Klotz JB, Halpin GJ, Zagraniski RT,Savrin J. Report on phase III: correlational analyses of adversereproductive outcomes and environmental pollution. NewJersey Department of Health, 1992.

50. A review of evidence on reproductive and developmental effectsof disinfection byproducts in drinking water. Research TrianglePark, NC:U.S. Environmental Protection Agency andWashington:International Life Sciences Institute Risk ScienceInstitute, 1993.

51. Craun GF. Epidemiology studies of water disinfectants and dis-infection by-products. In: Safety of Water Disinfection:Balancing Chemical and Microbial Risks (Craun GF, ed).Washington:ILSI Press, 1993;277-301.

52. Lynberg MC, Khoury MJ. Interaction between epidemiologyand laboratory sciences in the study of birth defects: design ofbirth defects risk factor surveillance in metropolitan Atlanta. JTox Environ Health 40:435-444 (1993).

53. Herrick RF, Stewart PA. International workshop on retrospec-tive exposure assessment for occupational epidemiologic stud-ies. Appl Occup Environ Hyg 6:417-420 (1991).

54. Olshan AF, Faustman EM. Male-mediated developmental toxi-city. Ann Rev Publ Health 43:159-181 (1993).

55. Graf W, Pippenger CE. Glutathione peroxidase deficiency infamilies with neural tube defects. Ann Neuro 34:476-477(1993).

56. Blancato JN, Chiu N. Predictive modeling for uptake and tis-sue distribution from human exposures. In: Safety of WaterDisinfection: Balancing Chemical and Microbial Risks (CraunGF, ed). Washington:ILSI Press, 1993;303-316.

57. Lynberg MD, Khoury MJ. Contribution of birth defects toinfant mortality among racial/ethnic minority groups, UnitedStates, 1983. CDC Surveillance Summaries. MMWR 39(No.SS-3):1-12 (1990).

Volume 103, Supplement 6, September 1995 171