Congenital heart defects occur frequently, affecting approximately 1 in 120 newborns, 1–7 yet their causes remain elusive. Although genetic conditions 8 or causal environmental factors (for example, maternal diabetes 9 or rubella 10) can be identified in some cases, for most babies who are born with a heart defect the causes remain unknown.
Maternal febrile illness is a potential risk factor for heart defects because hyperthermia and some infections are teratogenic in many animal species and because febrile illness is considered a likely teratogen in humans, 11–13 particularly for neural tube defects. 14–20 Febrile illness is also of interest because it occurs frequently (thus even relatively small increases in risk would produce many cases) and because, at least in some cases, it might be preventable.
In this study we had two objectives: (1) to evaluate the association between febrile illness and congenital heart defects, by type of illness and heart defect, and (2) to assess whether multivitamin use decreased the risk associated with febrile illness, as multivitamin use seems to decrease the risk for heart defects in the general population, according to some studies. 21–24
Subjects and Methods
We analyzed data from the Atlanta Birth Defects Case-Control Study (ABDCCS), which was conducted in 1982 and 1983. Information was gathered from parents of infants who were born from January 1, 1968, through December 31, 1980. This study was designed to find causes of birth defects, so it explored many exposures in relation to many congenital anomalies. 25–27 Detailed descriptions of the ABDCCS, 25,26 including the complete questionnaire, 26 have been published. Some preliminary or partial findings on febrile illness from this dataset were previously published. 27,28
Case infants were ascertained through the Metropolitan Atlanta Congenital Defects Program, a population-based registry of infants with birth defects born to mothers residing in the five-county metropolitan Atlanta area. Registry staff actively ascertained information on structural birth defects among liveborn and stillborn infants whose defects were diagnosed during the first year of life. Sources of information included vital records and medical records from birthing hospitals, pediatric and specialty wards, and cytogenetic laboratories.
Control infants were a 1% stratified random sample of infants without birth defects born during the same period and frequency-matched to the case infants by hospital of birth, calendar quarter of birth, and race.
We selected liveborn and stillborn infants with a cardiac defect (International Classification of Diseases, 9th revision, codes 745.00–747.49), and we assigned to each an anatomic diagnosis and an overall clinical diagnosis. Each baby received one anatomic diagnosis on the basis of a hierarchical classification (Table 1). 1,29 When more than one heart defect was present, the baby was assigned to the diagnostic category listed highest in Table 1. 1,29 Each baby was also given one clinical diagnosis: (1) isolated, if the cardiac defect was the only major anomaly present; (2) multiple, if the cardiac defect was part of a pattern of multiple congenital anomalies of unknown cause; or (3) syndromic, if the cardiac defect was part of a genetic or teratogenic syndrome (for example, trisomy 21 or congenital rubella). We excluded infants with syndromic diagnoses.
The study included 905 infants with nonsyndromic heart defects and 3,029 control infants (Table 1). Participation rates were 71% and 68% for mothers of control and case infants, respectively.
Exposure information was obtained through telephone interviews. 26 Information on febrile illnesses was obtained from questions on illnesses occurring from 3 months before pregnancy through the third month of pregnancy. In four questions, mothers were asked whether they had had (1) “flu” lasting 2 or more days, (2) rubella, (3) kidney infections, or (4) “any other illness” (those who answered positively to this question were asked to specify the illnesses). Mothers were asked to indicate, for each illness, the number of episodes, the month during which each episode occurred, and whether fever had been present. We classified the reported illnesses into six categories: “flu” (influenzalike illness), upper respiratory infections (for example, pharyngitis or “colds”), lower respiratory infections (for example, bronchitis or pneumonia), kidney and urinary tract infections, gynecologic infections (for example, vaginal infections or salpingitis), and “other” infections (for example, mumps, chickenpox, or dental infections). We excluded mothers who reported having had rubella during pregnancy.
Relative Risk Estimation
We selected as the exposure window the period from 1 month before conception through the third month of pregnancy. As in other studies of febrile illness, 15 the reference group was that of mothers who reported no illness during that time period. We used univariate and multivariate methods to evaluate the relative risks. We constructed logistic regression models to estimate the aggregate contribution of potential confounders. For case groups with three or more exposed infants, we present effect estimates adjusted for maternal education, race, age, smoking, alcohol use, chronic diseases (including diabetes mellitus), period of infant’s birth, and multivitamin use. To assess the independent and joint effects of febrile illness and multivitamin use (one of the objectives of the study), we contrasted periconceptional multivitamin use (regular use from 3 months before conception through the third month of pregnancy) to no use (during the same period). 24 Regular multivitamin use was defined as taking multivitamin supplements three or more times per week.
The most frequent defects were ventricular septal defects, obstructive defects, and outflow tract defects (Table 1). Compared with case mothers, control mothers were more likely to be white and to have delivered their baby during the earlier years of the study, and were less likely to report chronic illnesses (Table 2). Febrile illnesses, most commonly respiratory infections, were reported by 6.4% of control mothers and 11.2% of case mothers (Table 3).
Febrile illness was associated with a 1.8-fold increased risk for having a child with a major heart defect (Table 3). This association seemed driven by febrile respiratory infections, in particular flu [odds ratio (OR) = 2.1]. The association with febrile illness in general(Table 4) and with febrile flu (Table 5) was strongest for transposition of the great arteries, for some left obstructive defects, for tricuspid atresia, and for ventricular septal defects.
Effect estimates for febrile illness (Table 6) and for flu-associated fever (Table 7) were generally lower among mothers who used multivitamin supplements compared with those who did not use them. For example, the OR for heart defects associated with febrile illness was 1.1 for multivitamin users compared with 2.3 for non-users (Table 6). Because of the small number of subjects within outcome categories (small sample sizes), we studied only the larger case groups and present unadjusted effect estimates.
In this population-based study, we found that a febrile illness around the time of conception or in early pregnancy was associated with an approximately twofold increased risk for major heart defects. The most common febrile illnesses in this study were respiratory infections, frequently reported as “flu.” The risk associated with febrile illness seemed lower among mothers who used multivitamins from before conception.
The study has several strengths. It was population-based and relatively large. It also had information on numerous covariates and potential confounders. Multivariate analysis showed increased risk for heart defects even after controlling for many potential confounders.
A major concern in this study is the possibility for recall bias, which could readily account for effects of the magnitude we report. Recall bias is certainly possible, given the long delay (2–12 years) between infant’s birth and maternal interview and the limitations of exposure assessment (for example, lack of information such as core temperature, peak temperature, or fever duration). We also might have missed some cases of febrile illness, and we did not ask about other heat exposures such as sauna or hot tub bathing. Although we had some information on medications used for the illnesses, the numbers were too small to help disentangle the effect of medications from that of the underlying illness. Finally, study participation was incomplete, increasing the potential for bias, and some case groups were small, decreasing precision. Many of these factors could have contributed to recall bias. Nevertheless, some findings, in particular the interaction with multivitamin use, argue against recall bias, because it seems unlikely that recall bias could produce this pattern.
Findings from previous studies can offer further insights, although direct comparisons are sometimes difficult because of methodologic differences (for example, in exposure definitions). Two case-control studies report risk estimates associated with febrile illness for all heart defects combined. In the first study, from Finland, 30 researchers reported data from which we calculated a 1.8-fold increased risk (95% confidence interval = 1.2–2.6), similar to our estimate. In the second study, from China, 31 where the exposure was febrile “cold,” the increase was 1.4-fold (95% confidence interval = 0.7–2.9).
Three studies provide data for specific heart defects and allow comparisons with our finding of an increased risk for tricuspid atresia, left obstructive defects, transposition of the great arteries, and ventricular septal defects. Tricuspid atresia was strongly associated with maternal fever also in the Baltimore-Washington Infant Study, 32 and their effect estimate (OR = 5.1) was similar to ours (OR = 5.2). Hypoplastic left heart, a left-sided obstructive defect, was associated with febrile illness in the study from Finland 30 (recalculated OR = 4.0; 95% confidence interval = 1.9–8.5). In the Baltimore-Washington Infant Study, hypoplastic left heart and aortic stenosis (but not aortic coarctation) were also possibly associated with “influenza,” which presumably included forms with and without fever (OR = 1.5, 1.9, and 0.5, respectively). 32 In the same study, influenza was also associated with transposition of the great arteries, 32 with an effect estimate (OR = 2.2) similar to what we observed with flu-associated fever (OR = 2.1, Table 5). In the Finnish study, 30 outflow tract defects were positively associated with upper respiratory infections (with or without fever) (OR = 2.2), 30 although a separate estimate for transposition of the great arteries was not available. Finally, in a cohort of 64 liveborn infants exposed prenatally to high fever, 17 one child had transposition of the great arteries. 17 These epidemiologic studies also have limitations, however, and the relatively consistent findings might simply indicate similar recall bias. The frequency among controls of self-reported febrile illness was similar in most studies, although the similarity is difficult to interpret because of differences in exposure definitions. In our study, such frequency was 6.4%, whereas it was 6.3% in the Finnish study (exposure = fever ≥38°C), 6.0% in the Chinese study (exposure = cold with fever), and 4.6% in the Baltimore-Washington Infant Study (exposure = fever).
Animal studies suggest that hyperthermia in itself might be a cardiac teratogen. In chick embryos, for example, hyperthermia causes defects of endocardial lining with pathologic leakage, 33,34 malformations of the cardiac bulb (the outflow tract of the heart), and stenosis of the ventral aorta and aortic arches, 33 defects that are similar to those reported in epidemiologic studies of fever and heart defects. In humans, however, where the exposure is febrile illness rather than hyperthermia, it is difficult to disentangle the effects of temperature elevation, underlying infection, or medications. Common infections such as flulike illnesses have been associated with an increased risk for congenital anomalies even in the absence of fever. 15 Although effect estimates are often higher, for example, for flu with fever compared with flu alone, 15,31 such apparent increases may be due to greater severity of the underlying illness, of which fever may be but a hallmark.
Both fever and infection, however, have documented biological effects on specific developmental pathways. Apoptosis, for example, is affected by both hyperthermia 35,36 and viruses, 37 including influenza viruses. 38 Apoptosis is known to be involved in cardiac morphogenesis, for example in the development of the cardiac outflow tract, 39 and it has been suggested that altered apoptosis may cause birth defects. 36 Folic acid, contained in many multivitamin supplements, might also affect apoptosis 40–44 and seems to rescue apoptotic cells that are folate deficient. 42–44 The effect of nutrients such as folic acid on common developmental processes, such as cell proliferation and apoptosis, may explain, in part, our findings of an apparent interaction between maternal fever, multivitamin use, and congenital heart defects in the offspring.
We thank Yecai Liu for preparing and managing the datasets.
1. Ferencz C, Loffredo CA, Rubin JD, Magee CA. Epidemiology of Congenital Heart Disease: The Baltimore-Washington Infant Study 1981–1989. Perspectives in Pediatric Cardiology. vol. 4. Mount Kisco, NY: Futura Publishing Company, 1993.
2. Fyler DC. Report of the New England Regional Infant Cardiac Program. Pediatrics 1980; 65: 375–461.
3. Grabitz RG, Joffres MR, Collins-Nakai RL. Congenital heart disease: incidence in the first year of life. The Alberta Heritage Pediatric Cardiology Program. Am J Epidemiol 1988; 128: 381–388.
4. Johnson KC, Rouleau J. Temporal trends in Canadian birth defects birth prevalences, 1979–1993. Can J Public Health 1997; 88: 169–176.
5. Layde PM, Dooley K, Erickson JD, Edmonds LD. Is there an epidemic of ventricular septal defects in the USA? Lancet 1980; 1: 407–408.
6. Lin AE, Herring AH, Amstutz KS, Westgate MN, Lacro RV, Al-Jufan M, Ryan L, Holmes LB. Cardiovascular malformations: changes in prevalence and birth status, 1972–1990. Am J Med Genet 1999; 84: 102–110.
7. Wilson PD, Correa-Villaseñor A, Loffredo CA, Ferencz C. Temporal trends in prevalence of cardiovascular malformations in Maryland and the District of Columbia, 1981–1988. The Baltimore-Washington Infant Study Group. Epidemiology 1993; 4: 259–65.
8. Burn J, Goodship J. Congenital heart disease. In: Rimoin DL, Connor JM, Pyeritz RE, eds. Emery and Rimoin’s Principles and Practice of Medical Genetics. vol. 1. New York: Churchill Livingstone, 1996;767–828.
9. Shepard TH. Diabetes. In: Shepard TH, ed. Catalog of Teratogenic Agents. 8th ed. Baltimore: The Johns Hopkins University Press, 1995; 131–134.
10. Cooper LZ, Preblud SR, Alford CA. Rubella. In: Remington JS, Klein JO, eds. Infectious Diseases in the Fetus and Newborn Infant. 4th ed. Philadelphia: WB Saunders, 1995; 268–311.
11. Edwards MJ, Shiota K, Smith MS, Walsh DA. Hyperthermia and birth defects. Reprod Toxicol 1995; 9: 411–425.
12. Edwards MJ. Hyperthermia and birth defects (Editorial). Cornell Vet 1993; 83: 1–7.
13. Graham JM Jr, Edwards MJ. Teratogen update: gestational effects of maternal hyperthermia due to febrile illnesses and resultant patterns of defects in humans. Teratology 1998; 58: 209–221.
14. Layde PM, Edmonds LD, Erickson JD. Maternal fever and neural tube defects. Teratology 1980; 21: 105–108.
15. Lynberg MC, Khoury MJ, Lu X, Cocian T. Maternal flu, fever, and the risk of neural tube defects: a population-based case-control study. Am J Epidemiol 1994; 140: 244–255.
16. Smith MS, Upfold JB, Edwards MJ, Shiota K, Cawdell-Smith J. The induction of neural tube defects by maternal hyperthermia: a comparison of the guinea-pig and human. Neuropathol Appl Neurobiol 1992; 18: 71–80.
17. Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. Maternal fever and birth outcome: a prospective study. Teratology 1998; 58: 251–257.
18. Shaw GM, Todoroff K, Velie EM, Lammer EJ. Maternal illness, including fever and medication use as risk factors for neural tube defects. Teratology 1998; 57: 1–7.
19. Shiota K. Neural tube defects and maternal hyperthermia in early pregnancy
: epidemiology in a human embryo population. Am J Med Genet 1982; 12: 281–288.
20. Halperin LR, Wilroy RS Jr. Maternal hyperthermia and neural-tube defects (Letter). Lancet 1978; 2: 212–213.
21. Czeizel AE, Toth M, Rockenbauer M. Population-based case control study of folic acid supplementation during pregnancy
. Teratology 1996; 53: 345–351.
22. Shaw GM, O’Malley CD, Wasserman CR, Tolarova MM, Lammer EJ. Maternal periconceptional use of multivitamins and reduced risk for conotruncal heart defects and limb deficiencies among offspring. Am J Med Genet 1995; 59: 536–545.
23. Czeizel AE. Periconceptional folic acid containing multivitamin supplementation. Eur J Obstet Gynecol Reprod Biol 1998; 78: 151–161.
24. Botto LD, Mulinare J, Erickson JD. Occurrence of congenital heart defects in relation to maternal multivitamin use. Am J Epidemiol 2000; 151: 878–884.
25. Erickson JD, Mulinare J, McClain PW, Fitch TG, James LM, McClearn AB, Adams MJ Jr. Vietnam veterans’ risks for fathering babies with birth defects. JAMA 1984; 252: 903–912.
26. Erickson JD, Mulinare J, McClain PW. Vietnam Veterans’ Risk for Fathering Babies with Birth Defects. Atlanta: Centers for Disease Control and Prevention, 1984.
27. Erickson JD. Risk factors for birth defects: data from the Atlanta Birth Defects Case-Control Study. Teratology 1991; 43: 41–51.
28. Adams MM, Mulinare J, Dooley K. Risk factors for conotruncal cardiac defects in Atlanta. J Am Coll Cardiol 1989; 14: 432–442.
29. Clark EB. Pathogenetic mechanisms of congenital cardiovascular malformations revisited. Semin Perinatol 1996; 20: 465–472.
30. Tikkanen J, Heinonen OP. Maternal hyperthermia during pregnancy
and cardiovascular malformations in the offspring. Eur J Epidemiol 1991; 7: 628–635.
31. Zhang J, Cai WW. Association of the common cold in the first trimester of pregnancy
with birth defects. Pediatrics 1993; 92: 559–563.
32. Ferencz C, Loffredo CA, Correa-Villaseñor A, Wilson PD. Genetic and Environmental Risk Factors of Major Congenital Heart Disease: The Baltimore-Washington Infant Study 1981–1989. Perspectives in Pediatric Cardiology. vol. 5. Mount Kisco, NY: Futura Publishing Company, 1997.
33. Nilsen NO. Endothelial changes and microvascular leakage due to hyperthermia in chick embryos. Virchows Arch B Cell Pathol Incl Mol Pathol 1984; 46: 165–174.
34. Nilsen NO. Vascular abnormalities due to hyperthermia in chick embryos. Teratology 1984; 30: 237–251.
35. Mirkes PE, Cornel LM, Park HW, Cunningham ML. Induction of thermotolerance in early postimplantation rat embryos is associated with increased resistance to hyperthermia-induced apoptosis. Teratology 1997; 56: 210–219.
36. Edwards MJ. Apoptosis, the heat shock response, hyperthermia, birth defects, disease and cancer: where are the common links? Cell Stress Chaperones 1998; 3: 213–220.
37. Roulston A, Marcellus RC, Branton PE. Viruses and apoptosis. Annu Rev Microbiol 1999; 53: 577–628.
38. Takizawa T, Ohashi K, Nakanishi Y. Possible involvement of double-stranded RNA-activated protein kinase in cell death by influenza virus infection. J Virol 1996; 70: 8128–8132.
39. Watanabe M, Choudhry A, Berlan M, Singal A, Siwik E, Mohr S, Fisher SA. Developmental remodeling and shortening of the cardiac outflow tract involves myocyte programmed cell death. Development 1998; 125: 3809–3820.
40. James SJ, Basnakian AG, Miller BJ. In vitro
folate deficiency induces deoxynucleotide pool imbalance, apoptosis, and mutagenesis in Chinese hamster ovary cells. Cancer Res 1994; 54: 5075–5080.
41. James SJ, Miller BJ, Basnakian AG, Pogribny IP, Pogribna M, Muskhelishvili L. Apoptosis and proliferation under conditions of deoxynucleotide pool imbalance in liver of folate/methyl deficient rats. Carcinogenesis 1997; 18: 287–293.
42. Koury MJ, Horne DW, Brown ZA, Pietenpol JA, Blount BC, Ames BN, Hard R, Koury ST. Apoptosis of late-stage erythroblasts in megaloblastic anemia: association with DNA damage and macrocyte production. Blood 1997; 89: 4617–4623.
43. Koury MJ, Horne DW. Apoptosis mediates and thymidine prevents erythroblast destruction in folate deficiency anemia. Proc Natl Acad Sci USA 1994; 91: 4067–4071.
44. Huang RF, Ho YH, Lin HL, Wei JS, Liu TZ. Folate deficiency induces a cell cycle-specific apoptosis in HepG2 cells. J Nutr 1999; 129: 25–31.