Learn more: PMC Disclaimer | PMC Copyright Notice
Editor's choice
Maternal Genitourinary Infections and the Risk of Gastroschisis
Abstract
Genitourinary infections (GUIs) have been associated with increased risk of gastroschisis in 2 studies. Using data collected in the Slone Epidemiology Center Birth Defects Study, we examined the association between GUI and gastroschisis. From 1998 to 2010, mothers of 249 gastroschisis cases and 7,104 controls were interviewed within 6 months of delivery about pregnancy events, including vaginal infections, genital herpes, urinary tract infections (UTIs), and other sexually transmitted diseases (STDs). Women were considered exposed if they reported at least 1 instance of a GUI in the first trimester. Logistic regression models were used to calculate odds ratios and 95% confidence intervals. Women who reported having any GUI had an adjusted odds ratio of 1.8 (95% confidence interval (CI): 1.3, 2.4). The highest risk was seen among women who reported a UTI only (adjusted odds ratio = 2.3, 95% CI: 1.5, 3.5), while the odds ratio for an STD only was slightly elevated (adjusted odds ratio = 1.2, 95% CI: 1.0, 1.5). Among women under 25 years of age, the odds ratio for UTI only was 2.6 (95% CI: 1.7, 4.0), and among older women it was 1.8 (95% CI: 0.6, 5.9). When we considered the joint association of UTIs and young maternal age, a synergistic effect was observed. The results of this study add further evidence that UTIs may increase the risk of gastroschisis.
Gastroschisis is a rare congenital malformation that occurs in the first trimester of pregnancy and results in loops of bowel protruding from the abdominal wall of the infant (1). The prevalence of gastroschisis has been increasing worldwide (2), though its etiology is not well understood. There is some indication that environmental or infectious factors may be responsible (3, 4). The strongest evidence for this hypothesis is the consistent association between young maternal age and gastroschisis (3, 5), which has been suggested to result from a higher prevalence among younger mothers of negative health behaviors such as cigarette smoking and sexually transmitted diseases (STDs). While some of these risk factors have been associated with gastroschisis, none have been shown to explain the increasing prevalence in younger mothers (6, 7).
Gastroschisis has also been associated with first-trimester exposure to a number of medications, including pseudoephedrine, phenylpropanolamine, aspirin, ibuprofen, and acetaminophen (8–13). While these substances vary in their pharmacological actions, each can be indicated for an underlying infection or inflammation in the mother. Genitourinary infections (GUIs) are one type of infection that has been associated with an increased risk of birth defects (6, 13–16), and 2 studies that considered GUIs both linked them to gastroschisis (6, 13). One, from the United Kingdom, found a 2.6-fold increased risk of gastroschisis for women with any history of gynecological infection or disease prior to the current pregnancy (13). The other, from the United States, found that the greatest risk, a 4-fold increase, was among women who reported both an STD and a urinary tract infection (UTI) in the first trimester (6).
The objective of this analysis was to examine the association between GUIs and gastroschisis using data collected in a large case-control surveillance study of birth defects. In addition, we evaluated the independent associations of STDs and UTIs with GUI, as well as whether maternal age modified the association between GUIs and gastroschisis.
METHODS
Study design
The Slone Epidemiology Center Birth Defects Study (also known as the Pregnancy Health Interview Study) is an ongoing multicenter case-control surveillance study designed to assess risk factors in relation to birth defects. The study began in 1976 and has been described in detail elsewhere (17–19). The present study was restricted to women interviewed between 1998 and 2010. During this time, subjects were identified at birth hospitals and tertiary-care hospitals in the greater metropolitan areas of Philadelphia, Pennsylvania; San Diego, California; and Toronto, Ontario, Canada, as well as through statewide birth defects registries in Massachusetts and New York. Malformed infants were identified by review of discharge lists at participating hospitals and through the birth defects registries. Study staff did not consistently ascertain elective terminations or fetal deaths occurring after 20 weeks, but mothers of cases with these outcomes were eligible for inclusion. Nonmalformed liveborn controls were ascertained at the same birth hospitals as malformed infants and from statewide birth records in Massachusetts and New York. The study protocol was approved by institutional review boards at Boston University and other institutions, as appropriate.
Cases of gastroschisis were confirmed by medical records when available and were further classified as “isolated” or “associated.” Isolated cases had gastroschisis with no other major malformation except those that may be secondary to gastroschisis (e.g., intestinal atresia), while associated cases had at least 1 other unrelated major malformation. Infants with chromosomal anomalies, Mendelian inherited disorders, or a known syndrome were excluded.
Telephone interviews were conducted in English or Spanish with the mothers of cases and controls within 6 months of delivery by trained study nurses. After consent was obtained, women were asked about demographic factors, vitamin use, reproductive and medical history, illnesses and medications, cigarette smoking, alcohol and caffeine intake, and diet using a standardized questionnaire.
Exposure
Detailed information was obtained on illnesses and medications from 2 months prior to the last menstrual period through the end of pregnancy. A tiered approach was used to obtain information on illnesses and medications (20). Women were first questioned about specific illnesses; for GUIs, these included “UTIs,” “yeast/vaginal infections,” “genital herpes,” and “any other STDs.” If a woman responded positively, she was asked about the dates of illness, whether she had used any medications for the illness, whether fever had occurred in conjunction with the illness, and whether she had seen a physician or health-care provider. Next, independent of their responses to the illness questions, women were asked about any medications they took and were prompted with a list of specific medications that included Zovirax (GlaxoSmithKline Mississauga, Ontario, Canada) and acyclovir. If a woman reported using a medication, the dates and indication for use were recorded. Additionally, comments added by interviewers were reviewed for reports of GUIs.
Mothers were considered exposed if they reported having had at least 1 instance of a GUI anytime in the first 3 lunar months after the last menstrual period (i.e., the first trimester). Four mutually exclusive exposure categories were created: 1) a UTI alone, 2) an STD alone, 3) both a UTI and an STD, and 4) no UTI and no STD (reference group). The STD category included reports of yeast/vaginal infections, genital herpes, and any other STDs. In addition, we collapsed categories 1–3 to consider reports of either a UTI or an STD. Mothers were excluded if they reported GUIs with dates that were unclear or unknown.
The classes of medications reported for GUIs were compared between cases and controls, as were the types of medications, which included prescription and over-the-counter medications and alternative therapies (e.g., herbal remedies, vitamins, and douches). Reports of certain unidentified medications were considered prescription drugs; these included urinary tract medication, STD medication, herpes medication, antibiotic medication, and UTI medication. In contrast, unidentified vaginal antiinfective medication, antifungal medication, vaginal cream, and vaginal suppositories were categorized as unknown, because such products are available both over the counter and by prescription.
Statistical analysis
Odds ratios and 95% confidence intervals were calculated for each exposure group, with women who reported no UTIs and no STDs serving as the reference category. Confounders that were considered included: study center (Massachusetts, New York State, Philadelphia, Toronto, or San Diego), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian, or other), maternal education (less than high school, high school, or more than high school), maternal age (<20, 20–24, 25–29, 30–34, or ≥35 years), intentionality of pregnancy (planned or unplanned), body mass index (weight (kg)/height (m)2; <18.5, 18.5–24.9, 25–29.9, or ≥30), maternal smoking (never smoked or smoked only before pregnancy vs. smoked in the first trimester), alcohol consumption (never consumed alcohol or consumed it only before pregnancy vs. consumed alcohol in the first trimester), illicit drug use (never used illicit drugs or used them only before pregnancy vs. used illicit drugs in the first trimester), multivitamin use (no use of multivitamins vs. use in the first trimester), and any fevers reported in the first trimester (yes/no). Covariates were added to the model one at a time, and those that changed the effect estimate by more than 10% were retained in the final models.
We conducted a stratified analysis to assess whether effect-measure modification by maternal age was present. In addition, to further explore the possibility of interaction on the additive scale (i.e., biological interaction), we calculated the relative excess risk due to interaction (21), with the reference group being comprised of older mothers (≥25 years of age) without exposure (no UTIs or STDs).
Fever (≥38.3°C or ≥101°F) can indicate a more severe infection; therefore, using fever as a proxy for severity, we conducted a subanalysis to examine the association between severe GUI and risk of gastroschisis. In the subanalysis, women who reported fevers from other indications (e.g., respiratory illnesses or influenza) were excluded. We performed a second subanalysis to assess whether first-trimester multivitamin use (either multivitamins or prenatal vitamins) might alter the association with GUIs, given that various vitamins have been found to enhance immune response (22) and reduce the risk of birth defects (18). Due to the wording of the questionnaire, we could not separate vaginal infections from yeast infections; therefore, we conducted a separate subanalysis excluding vaginal infections from the STD group, since some vaginal infections are considered to be STDs (e.g., trichomoniasis), while others are not (e.g., yeast infections). In addition, a separate analysis was restricted to only isolated cases of gastroschisis.
We conducted a sensitivity analysis to assess the possibility of recall bias by examining exposures outside the etiologically relevant time period. Mothers who reported a GUI in only the second or third trimester were considered exposed and were compared with those mothers who reported no GUIs. The expectation was that recall bias would be unlikely to affect reporting of exposure timing; therefore, concern about recall bias would be diminished if the effect were confined to the etiologically relevant time period.
Due to methods of case and control ascertainment, we performed a second sensitivity analysis to assess the possibility of selection bias, since controls could be identified at hospitals from which there were no gastroschisis cases; if this were so, the control population might not represent the base population that gave rise to the cases. Therefore, the sensitivity analysis restricted cases and controls to those chosen from the same birth hospitals.
RESULTS
There were 20 controls and 3 cases excluded because of GUI dates that were unclear or missing, leaving a total of 249 cases and 7,104 controls. Compared with cases, control mothers were more likely to be non-Hispanic white, older, and more educated, to be nonsmokers, to have planned their pregnancy, and to have had previous pregnancies (Table 1). Reports of any GUI were more common in case mothers (29%) than in control mothers (15%) (Table 2). Mothers of cases also reported a higher percentage of only STDs (13%) and only UTIs (13%) than mothers of controls (10% and 4%, respectively). Among the 40 case mothers who reported an STD, 5% had genital herpes, 63% had vaginal/yeast infections, 23% had other STDs, and 10% had a combination of STD types. Among the 756 control mothers who reported an STD, 15% had genital herpes, 77% had vaginal/yeast infections, 5% had other STDs, and 3% had a combination of STD types. The other STD category included reports of genital warts, chlamydia, human papilloma virus, trichomoniasis, hepatitis B, group B streptococcus, gonorrhea, and not-otherwise-specified STD. Ninety-five percent of mothers with a UTI reported seeing a health-care professional for their condition; for STDs, the proportion was 80%. Notably, many of the mothers in the STD group who did not seek care had chronic conditions (e.g., herpes) that may have been monitored or had yeast/vaginal infections that may have been treated with over-the-counter medications.
Table 1.
Characteristics of Women Whose Infants Had Gastroschisis and Women Whose Infants Had No Major Malformations, United States and Canada, 1998–2010
| Cases (n = 249) | Controls (n = 7,104) | |||
|---|---|---|---|---|
| No. | % | No. | % | |
| Maternal race/ethnicity | ||||
| Non-Hispanic white | 135 | 54.2 | 5,013 | 70.6 |
| Non-Hispanic black | 28 | 11.2 | 529 | 7.4 |
| Hispanic | 68 | 27.3 | 1,025 | 14.4 |
| Asian | 9 | 3.6 | 385 | 5.4 |
| Other | 9 | 3.6 | 146 | 2.1 |
| Missing data | 0 | 0.0 | 6 | 0.1 |
| Maternal age, years | ||||
| <20 | 109 | 43.8 | 500 | 7.0 |
| 20–24 | 93 | 37.3 | 1,006 | 14.2 |
| 25–29 | 30 | 12.0 | 1,841 | 25.9 |
| 30–34 | 12 | 4.8 | 2,424 | 34.1 |
| ≥35 | 5 | 2.0 | 1,313 | 18.5 |
| Missing data | 0 | 0.0 | 20 | 0.3 |
| Education, years | ||||
| <12 | 65 | 26.1 | 650 | 9.1 |
| 12 | 93 | 37.3 | 1,302 | 18.3 |
| >12 | 91 | 36.5 | 5,147 | 72.5 |
| Missing data | 0 | 0.0 | 5 | 0.1 |
| Study center | ||||
| Massachusetts | 43 | 17.3 | 3,831 | 53.9 |
| Philadelphia, Pennsylvania | 89 | 35.7 | 1,349 | 19.0 |
| Toronto, Ontario, Canada | 28 | 11.2 | 644 | 9.1 |
| San Diego, California | 51 | 20.5 | 930 | 13.1 |
| New York State | 38 | 15.3 | 350 | 4.9 |
| Maternal body mass indexa | ||||
| <18.5 | 23 | 9.2 | 417 | 5.9 |
| 18.5–<25 | 170 | 68.3 | 4,377 | 61.6 |
| 25–<30 | 46 | 18.5 | 1,367 | 19.2 |
| ≥30 | 7 | 2.8 | 786 | 11.1 |
| Missing data | 3 | 1.2 | 157 | 2.2 |
| Maternal report of smoking in first trimester | ||||
| No | 138 | 55.4 | 5,950 | 83.8 |
| Yes | 110 | 44.2 | 1,148 | 16.2 |
| Unknown | 1 | 0.4 | 6 | 0.1 |
| Maternal report of alcohol use in first trimester | ||||
| No | 157 | 63.1 | 3,875 | 54.5 |
| Yes | 92 | 36.9 | 3,209 | 45.2 |
| Unknown | 0 | 0.0 | 20 | 0.3 |
| Illicit drug use in first trimester | ||||
| No | 230 | 92.4 | 6,978 | 98.2 |
| Yes | 19 | 7.6 | 126 | 1.8 |
| Intentionality of pregnancy | ||||
| Unplanned pregnancy | 162 | 65.1 | 2,624 | 36.9 |
| Planned pregnancy | 87 | 34.9 | 4,480 | 63.1 |
| Multivitamin use in first trimester | ||||
| No | 46 | 18.5 | 742 | 10.4 |
| Yes | 201 | 80.7 | 6,294 | 88.6 |
| Missing data | 2 | 0.8 | 68 | 1.0 |
a Weight (kg)/height (m)2.
Table 2.
Odds Ratio for the Risk of Gastroschisis According to Genitourinary Infection in the First Trimester of Pregnancy, United States and Canada, 1998–2010
| Cases | Controls | Crude | Adjusteda | |||||
|---|---|---|---|---|---|---|---|---|
| No. | % | No. | % | OR | 95% CI | OR | 95% CI | |
| No STD or UTI | 176 | 71 | 6,036 | 85 | 1 | Reference | 1 | Reference |
| STD or UTI | 73 | 29 | 1,068 | 15 | 2.35 | 1.77, 3.10 | 1.76 | 1.31, 2.37 |
| UTI only | 33 | 13 | 312 | 4 | 3.63 | 2.46, 5.35 | 2.31 | 1.53, 3.51 |
| STD only | 33 | 13 | 678 | 10 | 1.29 | 1.07, 1.56 | 1.20 | 0.98, 1.46 |
| STD and UTI | 7 | 3 | 78 | 1 | 1.46 | 1.12, 1.89 | 1.22 | 0.93, 1.61 |
Abbreviations: CI, confidence interval; OR, odds ratio; STD, sexually transmitted disease; UTI, urinary tract infection.
a Adjusted for maternal age (referent: age 25–29 years).
The multivariate models included maternal age, since it was the only covariate that met the criterion for potential confounding. Odds ratios were attenuated after adjustment by maternal age (Table 2); mothers who reported having any GUI had an adjusted odds ratio of 1.8 (95% confidence interval (CI): 1.3, 2.4) as compared with mothers who reported no GUI. The highest risk was observed among women who reported having only a UTI (adjusted odds ratio (OR) = 2.3, 95% CI: 1.5, 3.5). Women who reported having an STD only or an STD in combination with a UTI had adjusted odds ratios of 1.2, with lower 95% confidence limits of 1.0 and 0.9, respectively; however when the 29 cases and 606 controls with vaginal infections were excluded, the adjusted odds ratios increased to 1.4 (95% CI: 1.0, 2.1) and 1.4 (95% CI: 0.8, 2.4), respectively (data not shown).
When the data were stratified by maternal age (<25 years and ≥25 years) (Table 3), none of the potential confounders met the criterion for inclusion in the model; therefore, only crude odds ratios are reported. Younger mothers with a UTI only had a larger risk (OR = 2.6, 95% CI: 1.7, 4.0). All other estimates did not differ markedly between younger and older mothers.
Table 3.
Odds Ratio for the Risk of Gastroschisis According to Genitourinary Infection in the First Trimester of Pregnancy, by Maternal Age, United States and Canada, 1998–2010
| Maternal Age <25 Years | Maternal Age ≥25 Years | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cases | Controls | Crude OR | 95% CI | Cases | Controls | Crude OR | 95% CI | |||||
| No. | % | No. | % | No. | % | No. | % | |||||
| No STD or UTI | 138 | 68 | 1,211 | 80 | 1 | Reference | 38 | 81 | 4,807 | 86 | 1 | Reference |
| STD or UTI | 64 | 32 | 295 | 20 | 1.90 | 1.38, 2.63 | 9 | 19 | 771 | 14 | 1.48 | 0.71, 3.07 |
| UTI only | 30 | 15 | 102 | 7 | 2.58 | 1.66, 4.02 | 3 | 6 | 210 | 4 | 1.81 | 0.55, 5.90 |
| STD only | 28 | 14 | 164 | 11 | 1.22 | 0.98, 1.52 | 5 | 11 | 512 | 9 | 1.11 | 0.70, 1.78 |
| STD and UTI | 6 | 3 | 29 | 2 | 1.22 | 0.91, 1.65 | 1 | 2 | 49 | 1 | 1.37 | 0.70, 2.68 |
Abbreviations: CI, confidence interval; OR, odds ratio; STD, sexually transmitted disease; UTI, urinary tract infection.
When the joint associations of young maternal age and reported infection were assessed (Tables 4 and and5),5), maternal education and maternal smoking were identified as confounders and thus were included in the adjustment models. A synergistic effect between GUI and maternal age was seen, with the effect being strongest for UTIs only. The adjusted odds ratio for the joint association between young maternal age and UTIs only was 24.8 (95% CI: 14.2, 76.2) in comparison with the reference group of older women with no GUIs, resulting in a relative excess risk due to interaction of 13.2 (95% CI: 1.5, 25.0).
Table 4.
Odds Ratio for the Joint Association of Young Maternal Age (<25 Years) and Urinary Tract Infections With the Risk of Gastroschisis, United States and Canada, 1998–2010
| Young Maternal Age | Urinary Tract Infection Only | Odds Ratioa | 95% Confidence Interval |
|---|---|---|---|
| No | No | 1 | Reference |
| No | Yes | 1.8 | 0.6, 6.0 |
| Yes | No | 10.7 | 7.2, 16.0 |
| Yes | Yes | 24.8b | 14.2, 76.2 |
a Adjusted for education (referent: 12 years) and smoking (referent: no).
b Relative excess risk due to interaction = 13.2 (95% confidence interval: 1.5, 25.0).
Table 5.
Odds Ratio for the Joint Association of Young Maternal Age (<25 Years) and Sexually Transmitted Diseases With the Risk of Gastroschisis, United States and Canada, 1998–2010
| Young Maternal Age | Sexually Transmitted Disease Only | Odds Ratioa | 95% Confidence Interval |
|---|---|---|---|
| No | No | 1 | Reference |
| No | Yes | 1.2 | 0.5, 3.0 |
| Yes | No | 10.8 | 7.3, 16.1 |
| Yes | Yes | 15.5b | 9.0, 26.7 |
a Adjusted for education (referent: 12 years) and smoking (referent: no).
b Relative excess risk due to interaction = 4.5 (95% confidence interval: −2.3, 11.4).
Among mothers with a GUI, the proportion of cases (88%) reporting treatment (medication or alternative therapies) for GUI was slightly higher than that for controls (82%). Antiinfective agents were the most common class of medication reported (Table 6). Mothers of controls were more likely to use an over-the-counter medication (25%) than were mothers of cases (11%). In addition, they were also more likely to report using an antifungal medication (27%) than were cases (14%).
Table 6.
Reported Types of Medications Used for Genitourinary Infections in the First Trimester of Pregnancy, United States and Canada, 1998–2010
| Cases (n = 73) | Controls (n = 1,068) | |||
|---|---|---|---|---|
| No. | % | No. | % | |
| Class of medication reporteda | ||||
| Antiinfective agents | ||||
| Antibacterial | 8 | 11.0 | 88 | 8.2 |
| Antifungal | 10 | 13.7 | 292 | 27.3 |
| Antiprotozoal | 1 | 1.4 | 13 | 1.2 |
| Antiviral | 2 | 2.7 | 26 | 2.4 |
| Urinary | 3 | 4.1 | 33 | 3.1 |
| Unknown | 42 | 57.5 | 481 | 45.0 |
| Analgesic agents | 0 | 0.0 | 5 | 0.5 |
| Alternative therapiesb | 1 | 1.4 | 22 | 2.1 |
| Unknown medication | 2 | 2.7 | 0 | 0.0 |
| Type of medication reporteda | ||||
| Over-the-counter | 8 | 11.0 | 267 | 25.0 |
| Prescription | 45 | 61.6 | 476 | 44.6 |
| Alternative therapiesb | 1 | 1.4 | 22 | 2.1 |
| Unknown typec | 16 | 21.9 | 221 | 20.7 |
a More than 1 type of medication could be reported, and therefore numbers do not sum to the totals.
b Alternative therapies included herbal remedies, vitamins, and treatments such as douches.
c Includes not-otherwise-specified medications which could not be classified into a specific type (e.g., not-otherwise-specified antifungal medication).
With respect to severe UTIs, the age-adjusted odds ratio comparing mothers with a UTI and fever with mothers with no GUIs was 11.2 (95% CI: 1.3, 93.0), though the number of exposed cases and controls was small (n = 2 and n = 3, respectively). There were no older women who reported both a UTI and no use of a multivitamin, so we examined the joint association only among younger women. The odds ratio for the joint association of a UTI and no use of multivitamins was 4.4 (95% CI: 2.1, 9.4) relative to women who reported multivitamin use and no UTIs. This association was greater than expected given the individual associations of no multivitamin use and UTIs (relative excess risk due to interaction = 2.8, 95% CI: 0.6, 6.3) (data not shown).
Only 11 cases had other major malformations, and none were in the exposed group; therefore, the results did not change substantially when the case group was restricted to the 238 isolated cases of gastroschisis (GUI OR = 1.7, 95% CI: 1.2, 2.2). In addition, second- or third-trimester GUIs, in the absence of a first-trimester GUI, did not result in elevated odds ratios (for second-trimester GUIs, adjusted OR = 1.3 (95% CI: 0.8, 2.1); for third-trimester GUIs, adjusted OR = 0.9 (95% CI: 0.5, 1.6)). When the data were restricted to cases and controls from the same birth hospitals, the adjusted odds ratios were further elevated in comparison with the main analysis (GUI OR = 2.0, 95% CI: 1.4, 2.8).
DISCUSSION
This study found that reported UTIs in the first trimester were associated with an increased risk of gastroschisis, whereas the association for reported STDs was close to 1. The odds ratio for UTIs was higher among younger mothers, and a synergistic effect was observed for the joint association of UTIs and young maternal age. In addition, younger women who reported no multivitamin use in the first trimester were at further increased risk of gastroschisis.
Our results are similar to findings from the 2 previous studies that considered GUIs and gastroschisis. In the United Kingdom study, Draper et al. (13) reported an adjusted odds ratio of 2.6 (95% CI: 1.2, 5.6) for gastroschisis in relation to any history of gynecological infection or disease prior to the current pregnancy. While these results identified an elevated risk, the exposure definition was nonspecific with respect to both the type of infection and the timing of infection. In the other study, Feldkamp et al. (6) refined the definition of exposure and narrowed the exposure window to the first trimester. They used data from the US National Birth Defects Prevention Study and found the largest association among mothers who reported both a UTI and an STD, with an odds ratio of 4.0 (95% CI: 1.4, 11.6) after adjustment for maternal age (6). The adjusted odds ratio for STDs only (adjusted OR = 1.3, 95% CI: 0.7, 2.3) was very similar to that in our study; however, their estimate for UTIs, while modestly elevated (adjusted OR = 1.4, 95% CI: 1.0, 2.0), was lower than what we observed. When the data were stratified by maternal age (<20 years and ≥30 years), the authors found that even though younger mothers were more likely to report UTIs than older mothers, the risk estimate for UTIs only and gastroschisis was lower among younger mothers—1.3 (95% CI: 0.8, 2.0) for younger mothers and 2.3 (95% CI: 0.7, 7.6) for older mothers. They reported a similar pattern for STDs only, with odds ratios of 1.3 (95% CI: 0.6, 3.0) and 4.9 (95% CI: 1.1, 21.9) among younger and older mothers, respectively. The higher odds ratio among older mothers could have been due to a lower baseline risk of gastroschisis in the older women, resulting in a smaller denominator and an overall larger odds ratio within the stratum. To address this issue, in the present analysis we both stratified by maternal age and assessed the joint association of maternal age and GUI using a common reference group.
There are multiple mechanisms by which GUIs may act to increase the risk of gastroschisis. The first may be through medications used to treat GUIs, although (as seen in our study) the treatments for GUIs vary widely, from homeopathic remedies to antibacterials to antivirals, making it unlikely that specific treatments may be responsible. The mechanism might also operate through microbe-specific direct effects. However, for UTIs, the wide range of possible causes (e.g., bacteria, viruses, parasites) implies that they would be unlikely to all have the same etiological effect (23, 24). Another possible hypothesis may be a mechanism related to the immune response to the infection, affecting fetal development via maternal antibodies passed through the placenta (25). Alternately, the maternal or fetal immune response to an infection and/or fever could disrupt the embryonic developmental pathways. Evidence of this is found in previous studies where maternal infection, hyperthermia, and fever were associated with changes in maternal vasculature or apoptosis in the embryo (26–29). In our data, fever was not strongly associated with gastroschisis overall (3.2% of cases as compared with 2.5% of controls reported fever in the first trimester), but when a mother reported a UTI associated with fever, the risk of gastroschisis was much stronger than the risk from a UTI alone. Previous studies have found that levels of proinflammatory cytokines, specifically interleukin-6, are higher among febrile patients with UTIs than among nonfebrile patients with UTIs, suggesting an immune response (30–32). In our study, if fever in conjunction with a UTI indicated a more severe infection, it would add support to the hypothesis that the immune response may be involved in the etiology of gastroschisis. It has also been posited that the elevated risk in younger mothers could be the result of an immune response to newly encountered pathogens. In particular, younger mothers would have less cumulative exposure to GUIs and are more likely to have changes in partners between pregnancies, which would expose them to new paternal antigens. Consistent with this hypothesis, 3 studies found that changes in paternity or short sexual cohabitation times increased the risk of gastroschisis, adding further evidence to this hypothesis (33–35).
There were limitations to our study. First, some mothers may have had subclinical (and therefore unrecognized) GUIs, and they would have been included in the unexposed category; however, this type of misclassification would probably lead to underestimation of the true association. We also did not have information on interpregnancy interval, changes in paternity, or cohabitation time with a partner, all of which have been associated with gastroschisis. In addition, we relied on maternal reports of GUIs, which may be susceptible to reporting bias, though the use of a standardized questionnaire reduced the likelihood of this type of bias. In fact, one of the strengths of this study was the tiered approach to obtaining information on medication and illnesses, which resulted in detailed exposure information. Additionally, the short interval between delivery and interview (≤6 months) reduced concerns regarding reporting accuracy. Additional strengths of our study were the use of a sensitivity analysis to assess the possibility of recall bias and selection bias, as well as the assessment of both effect-measure modification and biological interaction by maternal age.
The results of our study add to the evidence that UTIs may increase the risk of gastroschisis. To our knowledge, this study was the first to assess and identify a synergistic relationship between young maternal age and UTIs. Further studies are needed to confirm this synergistic effect, especially given the increasing prevalence of gastroschisis among younger mothers (36–38). In addition, identifying pathogens and measuring markers of inflammatory response might help elucidate the underlying mechanism at play.
ACKNOWLEDGMENTS
Author affiliations: Slone Epidemiology Center at Boston University, Boston, Massachusetts (Mahsa M. Yazdy, Allen A. Mitchell, Martha M. Werler).
This work was supported by grant R01 HD046595-06 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development.
We thank Drs. Marlene Anderka, Veronica Vieira, Ann Aschengrau, and Russell Kirby for their helpful comments. We also thank Dawn Jacobs, Fiona Rice, Rita Krolak, Kathleen Sheehan, Moira Quinn, Clare Coughlin, Laurie Cincotta, Mary Thibeault, Nancy Rodriquez-Sheridan, Ileana Gatica, Laine Catlin Fletcher, Carolina Meyers, Joan Shander, Julia Venanzi, Mark Abcede, and Judy Jean for assistance with data collection; Katherine Kelley for assistance in classifying medications; Nastia Dynkin for computer programming; the staffs of the Massachusetts Department of Public Health Center for Birth Defects Research and Prevention and the Massachusetts Registry of Vital Records; Dr. Charlotte Druschel; the New York State Health Department; and Drs. Christina Chambers and Kenneth Jones of the University of California, San Diego. We thank the medical and nursing staffs at all participating hospitals (see below) for assistance with case ascertainment.
The following hospitals participate in the Slone Epidemiology Center Birth Defects Study: Baystate Medical Center, Beth Israel Deaconess Medical Center, Boston Medical Center, Brigham and Women's Hospital, Brockton Hospital, Cambridge Hospital Caritas Good Samaritan Medical Center, Charlton Memorial Hospital, Children's Hospital, Holy Family Hospital, Kent Hospital, Lawrence General Hospital, Lowell General Hospital, Melrose-Wakefield Hospital, Metro West Medical Center-Framingham, Mt. Auburn Hospital, New England Medical Center, Newton-Wellesley Hospital, North Shore Medical Center, Rhode Island Hospital, Saints Memorial Medical Center, South Shore Hospital, Southern New Hampshire Medical Center, St. Elizabeth's Medical Center, St. Luke's Hospital, UMass Memorial Health Care, Women and Infants’ Hospital, Abington Memorial Hospital, Albert Einstein Medical Center, Alfred I. duPont Hospital for Children, Bryn Mawr Hospital, Chester County Hospital, Children's Hospital of Philadelphia, Christiana Care Health Services, Community Hospital, Crozer-Chester Medical Center, Doylestown Hospital, Frankford Hospital, Hahnemann University Hospital, The Hospital of the University of Pennsylvania, Lankenau Hospital, Lancaster General Hospital, Lehigh Valley Hospital, Nanticoke Memorial Hospital, Pennsylvania Hospital, Sacred Heart Hospital, St. Christopher's Hospital for Children, St. Mary Medical Center, Temple University Health Sciences Center, Reading Hospital and Medical Center, Thomas Jefferson University Hospital, Grand River Hospital, Guelph General Hospital, Hamilton Health Sciences Corporation, The Hospital for Sick Children, Humber River Regional Hospital-Church Site, Humber River Regional Hospital-Finch Site, Joseph Brant Memorial Hospital, Lakeridge Health Corporation, London Health Sciences Center, Mt. Sinai Hospital, North York General Hospital, Oakville Trafalgar Memorial Hospital, Scarborough Hospital–General Division, Scarborough Hospital–Grace Division, St. Joseph's Health Centre-London, St. Joseph's Health Centre-Toronto, St. Joseph's Healthcare-Hamilton, St. Michael's Hospital, Sunnybrook and Women's College Health Sciences Center, Toronto East General Hospital, Toronto General Hospital, Trillium Health Center, William Osler Heath Centre, York Central Hospital, York County Hospital, Alvarado Hospital, Balboa Naval Medical Center, Camp Pendleton Naval Hospital, Children's Hospital and Health Center, Kaiser Zion Medical Center, Palomar Medical Center, Pomerado Hospital, Scripps Mercy Hospital, Scripps Memorial Hospital-Chula Vista, Scripps Memorial Hospital-Encinitas, Scripps Memorial Hospital-La Jolla, Sharp Chula Vista Hospital, Sharp Coronado Hospital, Sharp Grossmont Hospital, Sharp Mary Birch Hospital, Tri-City Medical Center, and UC San Diego Medical Center.
This research was presented at the annual meeting of the Society for Pediatric and Perinatal Epidemiology Research, Minneapolis, Minnesota, June 25–27, 2012, and at the National Birth Defects Prevention Network Annual Meeting, Atlanta, Georgia, February 25–27, 2013.
Dr. Martha M. Werler sits on the advisory boards of several studies that evaluate reproductive outcomes in association with medications taken for various autoimmune disorders, which are sponsored by the medications’ manufacturers. The products (and manufacturers) are teriflunomide (Sanofi S.A., Paris, France), Enbrel (Amgen, Inc., Thousand Oaks, California), Actemra (Genentech, Inc., South San Francisco, California), Cimzia (UCB S.A., Brussels, Belgium), Humira (Abbott Laboratories, North Chicago, Illinois), and Orencia (Bristol-Meyers Squibb, New York, New York). None of these medications were investigated in the present study, and no conflict from these relationships is perceived. Dr. Allen A. Mitchell is a member of a Biogen-Idec Pregnancy Registry Advisory Committee for Tecfidera (Biogen-Idec, Inc., Weston, Massachusetts), a biological treatment for multiple sclerosis. The Slone Epidemiology Center at Boston University receives research support from various pharmaceutical companies, some of which may make products covered by this study. None of the companies provided support for or had any role in or awareness of the present analysis.