The Association Between Maternal Alcohol Use and Smoking in Early Pregnancy and Congenital Cardiac Defects

Introduction

The use of alcohol among pregnant women continues to be a significant public health problem, with alcohol use during pregnancy linked to higher rates of numerous adverse infant health indices, including higher rates of infant morbidity and infant mortality and the associated heightened healthcare costs. ^{1 –5}

In 2006, the National Survey on Drug Use and Health found that 53% of US women of childbearing age were current alcohol users (defined as at least one drink in the past 30 days), 24% were binge drinkers (defined as five or more drinks on the same occasion in the past 30 days), and 5% were heavy drinkers (defined as five or more drinks on the same occasion on five or more days of the past 30 days). The same study reported that among pregnant women, 11.8% were current alcohol users, 2.9% were binge drinkers, and 0.7 % were heavy drinkers. ⁶ Because alcohol and other substance use during the early stages of pregnancy, primarily before pregnancy recognition, is harmful to fetal development, these high rates of alcohol use and binge drinking among women of childbearing age are particularly troubling.

A number of studies have examined the role of alcohol use on the outcome of congenital cardiac defects, a group of anomalies that are among the most frequently occurring types of birth defects. To date, these studies have had conflicting findings. ^{1 –3} Many smaller case-control studies have not reported a significant association between maternal alcohol use or binge drinking in early pregnancy and congenital cardiac defects ^{7 –15} ; however, in a relatively large case-control study in Spain, a significant association was reported between women who were reported alcoholics and heavy drinkers (daily consumption of 500 mL or more of wine or 1000 mL or more of beer) and an increase in congenital cardiac defects (odds ratio [OR] 11.93, 95% confidence interval [CI] 1.62-246.). ¹⁶ In addition, a prospective cohort study from Hungary of 24,732 births reported a significant association between regular alcohol use during pregnancy and congenital heart defects (OR 3.5, 95% CI 2.3-5.3). ¹⁷

The purpose of this case-control study was to assess the association between various definitions of maternal alcohol use (i.e., amount of alcohol consumed, frequency of alcohol consumption, and episodes of binge drinking) around the time of early pregnancy/conception and the odds of congenital cardiac defects. In addition, as cigarette use is an established risk factor for alcohol use during pregnancy, we were particularly interested in examining the moderating effect of concurrent alcohol use and cigarette use during pregnancy and the outcome of congenital cardiac defects. ^{18 –22}

Materials and Methods

Results

In the unadjusted models, self-reported alcohol use in the 3 months prior to pregnancy was not related to the odds of congenital cardiac defects. No significant differences between cases of congenital cardiac defects and healthy controls were noted for the measures of alcohol use in the 3 months prior to pregnancy (yes/no), binge drinking in the 3 months prior to pregnancy (yes/no), binge drinking more than once in the 3 months prior to pregnancy (yes/no), or frequent drinking in the three months prior to pregnancy (yes/no) (Table 1). In addition, as shown in Table 1, none of the potential risk factors were significantly different between the case and control groups, including reported cigarette use.

Although any binge drinking during the 3 months prior to pregnancy was not independently related to the odds of congenital cardiac defect in the adjusted models, there was a significant interaction between any reported binge drinking and reported smoking (p<0.01) in the final adjusted model (Table 2). This final model adjusted for maternal age, maternal race, ethnicity, marital status, insurance status, and maternal stress. Smoking was not independently related to congenital cardiac defects in this model.

Stratifying by reported smoking status revealed very interesting findings. A significant odds of congenital cardiac defects was found among mothers who reported both cigarette smoking and any binge drinking in the 3 months prior to pregnancy compared to those mothers who only smoked but did not binge drink in the 3 months prior to pregnancy (OR 12.65, 95% CI 3.54-45.25) (Table 3). Among mothers who did not smoke, we did not find an increase in congenital cardiac defects.

When examining multiple episodes of binge drinking, the odds of congenital cardiac defects among women reporting binge drinking more than once in the 3 months prior to pregnancy was significant in the adjusted model (OR 2.99, 95% CI 1.19-7.51) (Table 4). A significant interaction between smoking and binge drinking more than once was also noted in this adjusted model (p<0.01). Surprisingly, the odds of congenital cardiac defects among women reporting smoking was protective in the final multivariate model (OR 0.40, 95% CI 0.19-0.84) (Table 4).

The analysis stratifying by reported smoking status and controlling for maternal age continued to revealed a significantly increased odds of congenital cardiac defects among mothers who reported both smoking and binge drinking more than once in the 3 months prior to pregnancy compared to mothers who only reported smoking in the 3 months prior to pregnancy (OR 9.45, 95% CI 2.53-35.31) (Table 5). Among mothers who did not smoke, there were no significantly increased odds of congenital cardiac defects.

Discussion

Using multiple years of a large, nationally representative sample of reproductive-age women, we found a significant association between multiple episodes of binge drinking in the 3 months prior to pregnancy and congenital cardiac defects, with a 3-fold increase in congenital cardiac defect cases among women who binge drink multiple times during early pregnancy. Additionally, we found a substantial interaction between women who binge drink and use cigarettes early in pregnancy, with a >12 fold increased odds of congenital cardiac defects among mothers who report any binge drinking and concurrent smoking and a >9-fold increased odds of congenital cardiac defects among mothers who report multiple episodes of binge drinking and concurrent smoking during early pregnancy. These findings point to a potential relationship between binge drinking and, most importantly, binge drinking and concurrent smoking during the early stages of pregnancy and congenital cardiac defects.

The interaction between binge drinking and cigarette use in early pregnancy is synergistic and biologically plausible, illustrating the harmful relationship between frequent exposures to multiple carcinogens and teratogens (i.e., alcohol and cigarette use) during critical periods of early fetal development and the potential risk of congenital cardiac defects. ^{29 –32} This interactive effect was not found among any alcohol use alone, which points to the particular relationship between high, intermittent amounts of alcohol use, characteristic of binge drinking, and concurrent tobacco use in early pregnancy on the development of congenital cardiac defects. Potentially, a threshold exists in the first few weeks of fetal life in which exposure to alcohol in high levels in combination with cigarettes may inhibit or alter crucial biologic processes necessary for proper cardiac and circulatory development. ^{33 –35}

Limited studies have examined the role of binge drinking on other adverse pregnancy or child health outcomes. A longitudinal study of maternal binge drinking in the period before pregnancy found an increase in infant neurobehavior deficits. ³⁶ In other studies, binge drinking has been associated with developmental changes in infant electroencephalograms, ³⁷ longer infant hospital stays, ³⁸ and decreased infant head circumference, ³⁹ but a study of occasional binge drinking among moderate drinkers found no association between binge drinking and infant birth weight, length, head circumference, gestational age, intrauterine growth restriction, or Apgar scores. ³⁹ Another study found that alcohol use at higher levels, particularly heavy use and binge drinking, was associated with increased risk of preterm birth even when drinking is stopped before the second trimester. ⁴⁰ Among childhood outcomes, the effect of binge drinking independent of average alcohol intake has been found to be associated with learning disabilities, deficits in attention and memory, and problem solving at age 7 years, ^41,42 but most studies have shown little or no effect of binge drinking on other health outcomes. ^{43 –46} Much of these null or inconsistent findings may be related to the reported timing of binge drinking exposure, with the information on binge drinking not measuring alcohol use during the highly susceptible period of early pregnancy. ^34,47

In this study, we did not find reported smoking to be a significant, independent risk factor for congenital cardiac defects; however, others have noted a positive relationship. Alverson et al. ²⁹ reported positive associations between first trimester maternal cigarette consumption and the risk of asecundum-type atrial setal defects (OR 1.36, CI 1.04-1.78), right ventricular outflow tract defects (OR 1.35, CI 1.05-1.74), truncus arteriosus (OR 1.90, CI 1.04-3.45), and levo transmission of the great arteries (OR 1.79, CI 1.04-3.10). Malik et al. ¹⁸ found heavy maternal smoking (25 or more cigarettes per day) during pregnancy was associated with both septal and right-sided obstructive defects (OR 2.06, CI 1.20-3.45 and OR 2.35, CI 1.21-4.53, respectively). Kuciene and Dulskiene, ⁴⁸ reviewing studies from Ferencz et al. ⁴⁹ and Torfs and Christianson ⁵⁰ of risk factors for congenital cardiac defects, found smoking to be associated with several isolated cardiac defects—pulmonic stenosis (OR 12.5, CI 3.2-49.4), transposition with ventricular septal defect (OR 2.1, CI 1.2-3.9), atrioventricular canal defects (OR 2.3, CI 1.2-4.5), and atrial septal defects (CI 2.2, CI 1.1-4.3). To date, however, the evidence supporting an independent, causal relationship between smoking and congenital cardiac defects is still limited. In this study, positive report of smoking was defined as any smoking in the 3 months prior to pregnancy, and early prenatal exposure to cigarette smoking was inferred using this question. In future studies, the intensity, duration, and frequency of smoking both before and during early pregnancy should be measured. It is possible that a more valid assessment of prenatal smoking behavior in future studies may reveal a threshold or independent dose-response relationship with congenital cardiac defects.

Potential limitations exist in this study related to the use of birth certificates for case ascertainment and to the measurement of alcohol use, binge drinking, and smoking using PRAMS data. The reliability and validity of birth certificates may vary by state, but the extent of this variation is not known. However, we believe that the cases identified through birth certificates include the most severe cases of congenital cardiac defects and represent the cases with the highest rate of subsequent morbidity and mortality.

The reliance on self-report to classify frequency of alcohol use, binge drinking, and smoking among pregnant women may influence the study findings. Whitehead and Lipscomb ³ used PRAMS alcohol use data in a study of small for gestational age (SGA) births and found that respondents' interpretation of the PRAMS alcohol use questions varied. When asked about average drinks per week, women may have reported the number of drinks they often consumed on a daily basis rather than the mean number of drinks per week. The authors also noted that women were sometimes confused about the time frame for the 3 months prior to pregnancy. Some women interpreted this time period as the 3 months before conception, whereas other women interpreted this as the 3 months prior to pregnancy recognition. We hypothesize that reliance on self-report to classify alcohol use and binge drinking in this study may cause a nondifferential underestimate of binge drinking in both the case and control groups and subsequently diminish the true relationship between binge drinking and congenital cardiac defects. As we found a significant increase in cardiac defects given reported binge drinking during multiple occasions in the early parts of pregnancy, we posit that these findings are an underestimate of the true relationship between binge drinking and congenital cardiac defects. In the future, studies that use data incorporating a more refined classification of alcohol use and binge drinking and biologic confirmation of alcohol use through urine samples will generate a more precise evaluation of the role of binge drinking in congenital cardiac defects.

The quantification of alcohol use and binge drinking in this study may be subject to recall bias, with women with infants with congenital cardiac defects and the control group differentially recalling their reported use of alcohol and binge drinking during early pregnancy. Recall bias is a limitation that often accompanies case-control studies, particularly case-control studies that rely primarily on self-report to classify exposure. In this situation, women with infants with congenital cardiac defects may be less likely to report their alcohol use or binge drinking, given the stigma related to alcohol use during pregnancy. If this type of differential recall bias did exist, an underreporting of alcohol use and binge drinking in the case group would occur, resulting in an underreporting of the true odds ratio. Thus, our significant findings of a positive relationship between binge drinking and congenital cardiac defects may be an underrepresentation of the true relationship. In contrast, if the cases in this study are more likely to overreport their binge drinking or the controls are less likely to report their binge drinking, the study findings may overrepresent the true relationship between binge drinking and congenital cardiac defects. We believe, given the biologic plausibility of the relationship between early pregnancy exposure to binge drinking and cardiac defects and the recent findings from other studies recognizing the role of binge drinking in other adverse pregnancy and child health outcomes, the overreporting of binge drinking in the case group or the underreporting of binge drinking in the control group is minimal.

The PRAMS survey collects information on alcohol use, binge drinking, and cigarette use during the 3 months prior to pregnancy and during the last 3 months of pregnancy. Because we were interested in examining the outcome of congenital cardiac defects, which occur during embryogenesis, we used responses to questions about the 3 months prior to pregnancy as a proxy for early pregnancy exposure to binge drinking and smoking. Thus, the results of this study are limited to the extent to which behaviors of early pregnancy mirror those occurring before pregnancy recognition. Others have used these same PRAMS questions to estimate early pregnancy exposures. ³ As many episodes of binge drinking during early pregnancy occur before recognition of pregnancy, these questions may be a very accurate approximation of reported binge drinking and smoking in the earliest stages of pregnancy. Finally, as PRAMS data do not differentiate between the types of alcoholic beverage consumed (beer, wine, or liquor), we were unable to determine if different types of alcohol may have produced different effects on the odds of congenital cardiac defects.

These findings, which point to an increased odds of congenital cardiac defects among women who binge drink during the early periods of pregnancy, may be a function of other factors that coexist with binge drinking during early pregnancy, such as poor maternal health, dietary deficiencies, or other types of substance use. Future studies examining the early pregnancy contributors to congenital cardiac defects should examine the contribution of these other factors during early pregnancy.

In conclusion, the present study used a large sample of nationally representative PRAMS data from nine U.S. states during the years1996–2005 to assess the possible association among alcohol consumption, binge drinking during early pregnancy, and congenital cardiac defects. The findings indicate an increased odds of congenital cardiac defects among women who report multiple episodes of binge drinking in the early stages of pregnancy and a highly significant increase in congenital cardiac defects among women reporting multiple episodes of binge drinking and concurrent smoking during early pregnancy. The results indicate that binge drinking and smoking may act synergistically to increase the odds of congenital heart defects. Over the years, the public health community has made strides in educating women of childbearing age about the hazards of smoking during pregnancy, but further education highlighting the prevalence of binge drinking among women of childbearing age and the risks associated with binge drinking alone or binge drinking in combination with cigarette smoking needs to be developed.