Because alcohol exposure can have devastating effects on the fetus, the Centers for Disease Control and Prevention (CDC) recently embarked on a public health education effort with the primary aim of decreasing alcohol exposure in pregnancy.1 Noting that 53% of U.S. women in a large CDC study had consumed alcohol within the previous 30 days and that more than half of pregnancies in the United States are unplanned,2,3 the CDC recommends that all women who are planning a pregnancy or not using reliable contraception abstain from alcohol use.2,3 In response, women, the media, policy pundits, international health commentators, and health care providers criticized the program as patronizing social engineering and asked pointed questions regarding the primacy of improved contraceptive access and lack of normative data about typical behavior.4,5 The guidance implies that women desiring a pregnancy will abstain from alcohol, but previous studies provide conflicting evidence on the effect pregnancy intention has on alcohol consumption.6,7 Despite being the focus of several large studies, consistent predictors of alcohol exposure in pregnancy have yet to be determined.8,9 Research also provides conflicting findings of which maternal characteristics are predictors of risky drinking behaviors in pregnancy such as binge drinking, defined as consumption of five or more drinks on one occasion.9,10
Our objective was to examine the independent association of pregnancy intention with maternal alcohol consumption from the periconception period (30 days before missed menses) through the first trimester of pregnancy. Additionally, we sought to describe maternal characteristics associated with continued use of alcohol in pregnancy and binge drinking.
MATERIALS AND METHODS
Right From the Start is a prospective, community-based cohort study of women enrolled between 2000 and 2012 before 12 weeks of gestation. Participants were recruited from multiple metropolitan areas in North Carolina, Texas, and Tennessee. To be eligible for the cohort, a woman must be older than 18 years of age, speak English or Spanish, and not use assisted reproductive technology. Specific details about participant recruitment have been published previously.11 Right From the Start consisted of three phases with multiple sources of funding and a wide range of aims. All phases share common methods and data elements.
A telephone enrollment interview captured baseline demographic and previous pregnancy information and was administered at a mean of 48±13 days of gestation. An extensive first-trimester computer-assisted telephone interview was conducted near the end of the trimester (mean±standard deviation 85±21 days). This interview assessed medical history, prior reproductive history, symptoms and events during the current pregnancy, health behaviors (caffeine, tobacco, illicit drug, and alcohol use), medication and supplement use, and physical activity. Participants had an early first-trimester ultrasonogram to confirm gestational dating (mean gestational age at ultrasonography 9 weeks). Of the 6,105 pregnancies enrolled in Right From the Start, 5,036 were eligible for inclusion in this analysis (Fig. 1). The institutional review board of Vanderbilt University School of Medicine, Nashville, Tennessee, approved this study.
Participants voluntarily provided information about alcohol consumption before and during pregnancy in the first-trimester interview. Information regarding pattern of use, frequency, quantity, and type of alcohol was also obtained. Women with a history of alcohol use within the 4 months before the interview were considered exposed. Information was also obtained on frequency of binge drinking, defined as consuming five or more drinks on any single occasion. For the purpose of the study, an alcoholic drink was defined as 14 g ethanol.12 Some women reported alcohol use in the periconception and early pregnancy period but did not provide information on specific characteristics such as quantity or frequency (n=458). These women were considered exposed and included in the primary analyses but were excluded from the secondary analyses that required more specific measures of exposure.
Pregnancy intention was assessed by maternal interview at enrollment using the intendedness of pregnancy items developed by the CDC.13 A pregnancy was considered intended if the participant affirmed that she desired to have a baby, or another baby, at some time in her life and became pregnant in what she deemed was “about the right time” or later than desired. Pregnancies considered sooner than desired or not desired entirely were classified as unintended.
Maternal demographic, reproductive, and behavioral data were collected in telephone interviews at enrollment and in the first trimester. Covariates, determined a priori based on literature review and biological plausibility, consisted of maternal education level (high school or less, some college, college graduate or more), maternal age, race–ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other), marital status (married, other), first-trimester body mass index, maternal prenatal vitamin use, history of spontaneous and therapeutic abortion, household income, smoking status, and illicit drug use (use of any illegal substance in the past 4 months) were also obtained to assess association between these variables and alcohol consumption during pregnancy.
We tested for bivariate differences in maternal characteristics between women with intended and unintended pregnancies using the χ2 test. Mean and median alcohol consumption was calculated as the average number of drinks per week from the periconception period through the first-trimester interview. If the study participant reported a change in pattern of alcohol consumption during this time, mean and median consumption was calculated for before and after the change taking into account gestational week in which change in consumption was reported; χ2 tests were used to examine patterns of use by intendedness. We used univariate and multivariable logistic regression models to investigate the independent association of pregnancy intention with alcohol use in early pregnancy. All covariates determined a priori were included in the final logistic regression model. Interaction terms were generated for pregnancy intention and maternal age at enrollment, history of spontaneous or therapeutic abortions, parity, and maternal education to test for effect modification. No effect modification was noted (P>.20). Univariate and multivariable analyses were also generated for the given maternal demographic and reproductive characteristics to determine the association with alcohol consumption in pregnancy. All analyses were conducted in R 3.1.2 (www.r-project.org) or Stata 14.0.
The final cohort population consisted of 5,036 women. Among them, 70% reported the pregnancy was intended. Intended pregnancies were more common among women who were older (P<.001), white (P<.001), college-educated or greater (P<.001), married (P<.001), and of higher income (P<.001) (Table 1).
Fifty-five percent of women in the study reported using alcohol in the 12 weeks preceding the first-trimester interview, with use occurring in 55% and 56% of intended and unintended pregnancies, respectively (P=.32). Six percent of all women reported current use of alcohol at the first-trimester interview, and this did not differ by pregnancy intention (P=.20). In the youngest age group (ages 18–20 years) who are not old enough to drink legally, 35% consumed alcohol at some point during pregnancy and this was significantly more likely among those with unintended pregnancies (39% compared with 27%, P<.001).
Women who recently quit using alcohol had higher median weekly alcohol consumption before quitting compared with women who continued to drink throughout pregnancy in both intended and unintended pregnancies (Table 2). Among women who stopped drinking during pregnancy, those with intended and unintended pregnancies consumed a median of two drinks per week (interquartile range 1.0–4.0 for both groups, P=.96). In the group that continued to drink alcohol at the first-trimester interview, women with intended and unintended pregnancies consumed a median of one drink per week (interquartile range for both 0.25–2.0, P=.25). During pregnancy, 90% of women with alcohol consumption reported a change in drinking pattern. Of these women with a change in pattern, 90% ceased consumption, 8% reduced consumption, and 2% of women increased alcohol intake. In women who changed drinking pattern, the median number of alcoholic beverages consumed was two per week before the change (interquartile range 0.92–4.0). At the first-trimester interview, the median number of alcoholic beverages consumed among those still drinking was 0.92 drinks per week (interquartile range 0.23–2.0) and did not differ based on pregnancy intention (P=.36). The median gestational age at which a change in drinking status occurred was 29 days (interquartile range 15–35) (Fig. 2). When stratifying this change in alcohol use by pregnancy intention, women with intended pregnancies made the change an average of 3 days earlier (28 days, interquartile range 14–34, compared with 31 days, interquartile range 17–38).
After adjusting for maternal age, race–ethnicity, marital status, maternal education level, income, parity, body mass index, prenatal vitamin use, spontaneous or elective abortion history, smoking, and drug use, women with intended pregnancies had 31% lower odds of alcohol consumption in pregnancy compared with women with unintended pregnancies (adjusted odds ratio [OR] 0.69, 95% confidence interval [CI] 0.60–0.81). Women who were older, white, college-educated, had higher incomes, and having their first pregnancy were most likely to use alcohol in pregnancy (Table 3).
Women with intended pregnancies were 32% less likely (adjusted OR 0.68, 95% CI 0.54–0.86) to have at least one episode of binge drinking at any point in pregnancy compared with participants with unintended pregnancies. Women with intended pregnancies were 39% (adjusted OR 0.61, 95% CI 0.41–0.92) less likely to have five or more episodes of binge drinking in pregnancy compared with those with unintended pregnancies (P=.02). Characteristics associated with binge drinking included younger age, being unmarried, past or current smoking, and illicit drug use.
Although women with intended pregnancies are less likely to report using alcohol in early pregnancy than women with unintended pregnancies, alcohol use is still prevalent among women intending to conceive. The use of alcohol among women with intended pregnancies in the cohort was 55%, which parallels the CDC findings of alcohol use among approximately 50% of women of childbearing age.2
The majority of women with periconception alcohol exposure stop or curtail use around the time of a positive pregnancy test as indicated by mean gestational age at which women stopped or decreased alcohol use in this study (mean±standard deviation 23.4±19 days). This mean time of change is shortly after conception suggesting that women hoping to be pregnant may reduce use until missed menses and then return “to normal” if they are not pregnant. If pregnancy is achieved, the woman may maintain abstinence from alcohol or decrease use shortly after pregnancy recognition. Taking into account confounders, continuing to use alcohol in pregnancy is less likely but still occurs among those with intended pregnancies.
Women who continue to use alcohol in the range of two to three drinks per week may be overlooked because they defy clinical and cultural biases about who may be most likely to drink during pregnancy (white, older, more educated women with higher income). These predictors are consistent with findings by Tan and colleagues2 using the Behavioral Risk Factor Surveillance System, a state-based random digit telephone survey to collect data on multiple maternal behaviors. Health care providers should not conflate wanting a pregnancy with the likelihood that an individual will avoid all potential risks. This population demonstrates substantial voluntary risk reduction before the onset of prenatal care. However, we do not find that elimination of all alcohol use when planning a pregnancy is the norm before the awareness of pregnancy.
Our study has limitations. Because no reliable marker for alcohol use is readily available, our data on alcohol exposure and timing were obtained from interviews that can be subject to reporting and recall bias. It has been noted in multiple studies that there is always a concern for underreporting of maternal alcohol use in pregnancy, largely as a result of social stigma. We attempted to minimize recall bias by obtaining the exposure information early in pregnancy through a nonclinical interview in which confidentiality was assured. Another important limitation is the self-reported pregnancy intention. It is possible that some women did not desire a pregnancy initially but reported an intended pregnancy after becoming pregnant. Finally, some mothers failed to provide detailed information on quantity and frequency of alcohol consumption.
Our study provides unique insight into the alcohol use patterns of pregnant women. This study, in contrast with previous studies, prospectively examines maternal alcohol use in the periconception period as well as during the first trimester. This cohort was also comprised of a large proportion of intended pregnancies, which provides a unique ability to observe alcohol use in women who wish to become pregnant.
As a public health strategy, our data suggest that the majority of women, regardless of pregnancy intention, stop or decrease alcohol use around the time of a positive pregnancy test. This implies that promoting earlier pregnancy awareness by testing near the time of anticipated menses may be a more beneficial strategy for decreasing alcohol use in pregnancy than encouraging abstinence from alcohol among all women who could conceive. The latter does not seem likely to be achieved, whereas the former results in relatively prompt reductions in alcohol consumption. Any remaining use, and evaluation for high-risk behaviors such as underage alcohol use and binge drinking (a separate phenomenon), can then become the focus of early clinical encounters but must be extended to all pregnant women, not just those perceived as high risk based on traditional socioeconomic markers of risk for poor pregnancy outcomes. Reducing unintended pregnancy is another important public health endeavor and can be accomplished by enhanced contraceptive access, improved counseling, and anticipatory guidance.
2. Tan CH, Denny CH, Cheal NE, Sniezek JE, Kanny D. Alcohol use and binge drinking among women of childbearing age—United States 2011–2013. Morb Mortal Wkly Rep 2015;64:1042–6.
4. Oster E. ‘No alcohol’ in pregnancy is just another shame battle in the mommy wars. Time October 21, 2015.
5. Petri A. The CDC's incredibly condescending warning to young women. Washington Post February 3, 2016.
6. Cheng D, Schwarz EB, Douglas E, Horon I. Unintended pregnancy and associated maternal preconception, prenatal and postpartum behaviors. Contraception 2009;79:194–8.
7. Terplan M, Cheng D, Chisolm M. The relationship between pregnancy intention and alcohol use behavior: an analysis of PRAMS data. J Subst Abuse Treat 2014;46:506–10.
8. Walker M, Al-Sahab B, Islam F, Tamim H. The epidemiology of alcohol utilization during pregnancy: an analysis of the Canadian Maternity Experiences Survey (MES). BMC Pregnancy Childbirth 2011;11:52.
9. McDonald SW, Hicks M, Rasmussen C, Nagulesapillai T, Cook J, Tough SC. Characteristics of women who consume alcohol before and after pregnancy recognition in a Canadian sample: a prospective cohort study. Alcohol Clin Exp Res 2014;38:3008–16.
10. Han JY, Nava-Ocampo AA, Koren G. Unintended pregnancies and exposure to potential human teratogens. Birth Defects Res A Clin Mol Teratol 2005;73:245–8.
11. Promislow JH, Makarushka CM, Gorman JR, Howards PP, Savitz DA, Hartmann KE. Recruitment for a community-based study of early pregnancy: the Right From the Start study. Paediatr Perinat Epidemiol 2004;18:143–52.
© 2017 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
12. Centers for Disease Control and Prevention. Frequently asked questions. Alcohol and public health. Available at: http://www.cdc.gov/alcohol/faqs.htm
. Retrieved February 9, 2017.