Many of the detrimental effects of prenatal alcohol exposure may depend on the achieved maternal and fetal blood alcohol concentration and duration of exposure.1–3 Intake of large doses of alcohol on a single occasion (binge drinking) leads to a high blood alcohol level, and such drinking behavior in the prerecognized phase of pregnancy is frequent among women in societies where recreational binge drinking is common among young women. Figures from Denmark show that 25–50% of pregnant Danes experience at least one episode of binge drinking, most often in the very early part of pregnancy.4,5 Figures from the United States show that binge drinking is increasing among both pregnant and nonpregnant women.6–8
Animal studies indicate that exposure to alcohol levels equivalent to binge drinking may induce chromosomal anomalies and increase the risk of spontaneous abortion,9–13 but evidence from humans is still sparse. To our knowledge, only three studies address the association between binge drinking and spontaneous abortion, and these studies showed no evidence of an association (Jones KL, Chernoff GF, Kelly CD. Outcome of pregnancy in women who “binge” drink during the first trimester of pregnancy [abstract]. Clin Res 1984;32:114A).14,15 However, all three studies were small, and two of them had no information on the frequency or timing of binge drinking during pregnancy. One study defined binge drinking as drinking to the point of feeling drunk 1–3 times during the first trimester. Lack of or imprecise information on timing of binge drinking may mask a possible detrimental effect of binge drinking if the fetus is particularly susceptible during certain periods of development.5 For binge drinking to induce chromosomal aberrations, it has to occur during the follicular cycle leading to conception or at the time of the early cleavage. Later binge drinking may have direct lethal effects on the embryo/fetus and thereby lead to chromosomally normal, but maybe otherwise abnormal, fetal deaths. In the present study, we used data from the largest existing maternal-child cohort, the Danish National Birth Cohort, to assess the association between binge drinking in the first 16 weeks of gestation and risk of early, as well as late, fetal death.
MATERIALS AND METHODS
The Danish National Birth Cohort is a population-based cohort of children born by pregnant women who intended to carry their pregnancy to term.16,17 During 1996–2002, pregnant women were approached at the first antenatal visit, and participants provided information on exposures during pregnancy by means of a computer-assisted telephone interview scheduled to occur around the 12th through 16th weeks of gestation. If the woman experienced a fetal loss before answering the interview, she was offered a slightly modified version of this interview shortly after the fetal loss (retrospectively collected data). The retrospectively collected data concerned only the first 16 weeks of pregnancy. The Danish Scientific Ethics Committee has approved the Danish National Birth Cohort, and before initiating the present study, approval by the Danish Data Protection Board was obtained.
In this study we included 92,717 women who participated in either the early pregnancy interview (n=90,165) or the case interview (n=2,552). We excluded pregnancies that were diagnosed as hydatidiform mole or ectopic (n=50). In addition, we excluded participants with no information on gestational age at recruitment or at the end of pregnancy (n=40), no information on binge drinking (n=367), or any of the covariates included in the regression models (n=3,059), leaving 89,201 pregnancies eligible for the analyses.
The women were asked how many times they had consumed five or more alcoholic drinks on a single occasion since the onset of pregnancy, defined as the first day of the last menstrual period. The definition of a drink followed the definition of the Danish National Board of Health, one drink containing 12 g or 15 mL of pure alcohol, the equivalent of one normal beer, one glass of wine, or 4 cl of spirits. The frequency of binge drinking episodes was recorded, and for each episode the pregnancy week in which the binge drinking occurred was recorded. We categorized the frequency of binge drinking episodes into 0, 1, 2, and 3 or more episodes. Furthermore, we categorized the timing of the binge drinking as 0 versus 1 or more episodes in each of the following time periods: the preconceptional period (pregnancy weeks 0–2), the period of fertilization and implantation (pregnancy weeks 3–4), the embryonic period (pregnancy weeks 5–10), and the early fetal period (pregnancy weeks 11–16), respectively.
We identified the following potential confounders according to Directed Acyclic Graphs18: maternal age in years (less than 25, 25–29, 30–34, or 35 or more), parity (0 or 1 or more), number of previous spontaneous abortions (0, 1, 2, or 3 or more), time until pregnancy in months (unplanned, 0–5, 6–12, and 13 or more), daily number of cigarettes smoked during pregnancy (0, 1–10, or 11 or more), daily number of cups of coffee consumed during pregnancy (0,–3, 4–7, or 8 or more), occupational status (higher-grade professionals, middle-grade professionals, skilled workers, unskilled workers, students, or unemployed for more than 1 year), and prepregnancy body mass index (less than 18.5, 18.5–24.9, 25–29.9, or 30 or more).
Fetal death was defined as spontaneous abortion (less than 22 completed weeks of gestation), stillbirth (22 or more completed weeks of gestation), and induced abortion on medical indication. We coded multiple pregnancies (n=1,892) as live births if all (n=1,872) or just one (n=20) of the fetuses resulted in a liveborn infant. By use of the mother’s civil registry number and the link between mother and child, we identified all live births and stillbirths through linkage to the Civil Registration System and the Danish Medical Birth Registry. From the National Hospital Discharge Registry, we identified all other pregnancy outcomes, such as spontaneous abortion, hydatidiform mole, and ectopic pregnancy. In more than 99% of the pregnancies, we identified the outcome through register-linkage, and in the remaining pregnancies we used information obtained from the mothers.
We used SAS 8 (SAS Institute Inc., Cary, NC) to conduct Cox regression models with gestational age (days since last menstrual period) as the underlying time variable. We followed the pregnancies from the day of signing the informed consent form until the date of live birth, induced abortion, emigration, or fetal death, whichever came first. Initially, we estimated the gestational age adjusted hazard ratios of fetal death associated with the frequency and timing of binge episodes, treated as time-dependent variables. To account for any differences in the frequency of binge drinking in the analyses of an association between the timing of binge drinking and fetal death, we simultaneously included variables for timing and for number of episodes (1, 2, 3 or more). By including an interaction term between each variable for binge drinking and gestational age at fetal death, indicating early spontaneous abortion (12 completed weeks or less), late spontaneous abortion (13–21 completed weeks), and stillbirth (22 or more completed weeks), we assessed the effect of binge drinking by gestational age. Second, we fitted multiple Cox regression models to adjust for potential confounding from other covariates. Third, we addressed interactions between the number of binge drinking episodes and weekly alcohol consumption, parity, smoking, and coffee consumption during pregnancy, respectively. We considered the interactions to be significant if P<.10. Finally, to test the robustness of the obtained results all analyses were restricted to the following subpopulations: women with exclusively prospective collected exposure (n=86,752), women who reported the date of all the reported binge episodes (n=88,310), and first-time pregnant women who had waited less than 1 year to become pregnant (n=25,866). Additionally, we repeated all analyses without considering induced abortion on medical indication as fetal death, using only the first enrolled pregnancy of women who participated with more than one pregnancy (n=80,721) and with gestational age based on information from the National Hospital Discharge Registry.
A total of 3,714 (4.2%) pregnancies ended in fetal death: 3,270 in spontaneous abortion and 444 in stillbirth. Exposure data were collected retrospectively for virtually all of the early, as well as the majority of the late, spontaneous abortions (Table 1). During the first 16 weeks of pregnancy, 21,349 (23.9%) of the women had at least one episode of binge drinking, most of which occurred in the first 6 weeks of pregnancy. Less than 1% of the women reported binge drinking in each week from pregnancy week 7 and onward. The proportion of women who were primiparous, smokers, and coffee-consumers increased with the number of binge episodes, but the proportion of women who were obese and had experienced previous spontaneous abortions decreased as the frequency of binge episodes increased (Table 2). The average alcohol consumption during pregnancy was higher among women who reported a high frequency of binge drinking during the first 16 weeks of pregnancy and/or binge drinking in the later part of pregnancy.
Women who reported binge drinking 1, 2, or 3 or more times during the first 16 weeks of pregnancy had a risk of fetal death essentially similar to that of women with no binge episodes (Table 3). The risk of early as well as late spontaneous abortion was unrelated to the number of binge drinking episodes. Women with three or more binge episodes had an increased risk of stillbirth compared with nonbinge drinkers, hazard ratio 1.56 (95% confidence interval [CI] 1.01–2.40). Women with one binge episode in either of the four periods of pregnancy (pregnancy weeks 1–2, 3–4, 5–10, and 11–16) had a similar risk of fetal death as did women who did not binge drink in the same periods (Table 4). Binge drinking in pregnancy weeks 11–16 was associated with an elevated risk of late spontaneous abortion, but adjustment for maternal age, parity, previous spontaneous abortion, coffee consumption, smoking, time until pregnancy, prepregnancy body mass index, and occupational status slightly attenuated the risk, and the lower confidence bound was 0.97 (Table 4). The association between binge drinking in pregnancy weeks 11–16 and late spontaneous abortion was stronger in the analyses restricted to women with complete information on the timing of binge drinking, women with prospectively collected data, the first enrolled pregnancy, or primiparous women who had waited less than 1 year to become pregnant (results not shown). The same applied to the analyses in which induced abortion on medical indication was excluded from the definition of fetal death and in which gestational age was based on information from the National Hospital Discharge Registry (results not shown).
Average alcohol consumption during pregnancy modified the association between frequency of binge drinking and fetal death (P=.06 for interaction). Average alcohol consumption during pregnancy was positively associated with the risk of fetal death, independently of binge drinking (Table 5). The lack of association between the frequency of binge drinking and fetal death was stable for different levels of average alcohol intake. However, for women with an average intake of three or more drinks per week, two or more binge drinking episodes seemed to be associated with a higher risk (hazard ratio 2.20, 95% CI 1.73–2.80) than for women with the same average intake who did not binge drink (hazard ratio 1.76, 95% CI 1.50–2.06). In the analyses restricted to women with prospectively collected data, there was no evidence for an interaction between the number of binge drinking episodes and average alcohol consumption (P=.27), and neither was there an association between average alcohol consumption and fetal death (results not shown). The estimates did not change substantially in any of the other analyses performed to check the robustness of the results.
In this large cohort, we found no association between binge drinking in the first 16 weeks of pregnancy and the risk of spontaneous abortion in clinically recognized pregnancies. We did, however, find an elevated risk of stillbirth for mothers who had three or more episodes of binge drinking, and women with a weekly consumption of three or more drinks and at least two binge drinking episodes had twice the risk of fetal death as women who abstained from alcohol intake during pregnancy. The confidence intervals are narrow for most of the estimates, providing some statistical support for the negative results.
High blood-alcohol level in early pregnancy increases the incidence of embryonic resorption and fetal death in animal settings. The exact biological mechanisms for this remains unknown, and several mechanisms may be involved.10,11 Alcohol may influence inflammatory factors, impair the nutritional or hormonal status, cause mutations in the DNA, interfere with the functioning of the mitotic/meiotic spindle apparatus, or initiate cell damage causing abnormalities that are incompatible with life.9,19 In addition, alcohol may increase the synthesis and excretion of prostaglandins. Prostaglandins suppress the rate of cell division to a degree proportional to the ingested amount of alcohol, and increased levels of prostaglandins may increase the risk of chromosomally normal fetal death.20 An alternative pathway through which alcohol could cause fetal death may be an effect of paternal alcohol consumption on sperm,21,22 since women who drink a lot are more likely to have a partner who also drinks.21
The lack of an association between binge drinking and fetal death among women with a weekly average consumption of less than three drinks could indicate a threshold for an effect of alcohol in early pregnancy. If the average alcohol intake is generally low, a single or two occasions of binge drinking in our data and the data of others (Jones et al [abstract], 1984).14,15 do not indicate any strong effect on fetal survival, but the evidence remains sparse, especially for early fetal loss. For the time being, women have no reason to fear that a single episode of moderate binge drinking in the prerecognized part of pregnancy will impact fetal survival. The vast majority of women reduce their alcohol consumption once they know they are pregnant. Women who continue to drink alcohol after they know they are pregnant, especially in large amounts, including binge drinking, may increase the risk of fetal death, and clinicians need to pay special attention to these women.
In addition to its size, our study has several other strengths. 1) Linkage to nationwide registries ensured nearly complete follow-up. 2) Our study combined information on frequency and timing of binge drinking episodes. 3) The average consumption in the Danish National Birth Cohort is low to moderate, and we take into consideration information on average intake. 4) Our study recorded a high number of early fetal deaths. 5) Our study included detailed information on known risk factors for fetal death that allowed for adjustment for confounding. 6) Carrying out the study in a cultural setting where alcohol consumption among women is socially accepted probably makes the data quality more reliable. There is also potential criticism of our study. We cannot rule out unmeasured or residual confounding because it is well known that women who drink a lot also have other unhealthy habits. However, we do not assume that the increased risk of stillbirth is attributable to poor diet or drug abuse because few of such marginalized women accepted the invitation to the study.23 Early spontaneous abortion is difficult to study in a prospective setting unless the cohort recruitment starts at the time of pregnancy planning. Compared with those in other birth cohorts, the women in this study were enrolled during early pregnancy, but we still missed most of the very early abortions and abortions that took place before the pregnancy was recognized. The use of retrospectively collected information for women who miscarried before answering the interview may result in recall bias, and it is difficult to predict the direction. Eighty-five percent of pregnant women in Denmark believe that binge drinking in pregnancy is a hazardous behavior and express concerns for fetal health consequences of binge drinking in the prerecognized part of pregnancy.24 If the women try to find explanations for the fetal demise, this may lead to better recall, but underreporting may be more likely if the women try to avoid victim blaming or self-reproach. On average, women with retrospectively collected data were interviewed with a longer delay from conception than women with prospectively collected data. This may have led to underreporting of binge drinking, because women were more likely to report binge drinking if the interview took place close to the time of exposure.25 Restricting the analyses to women with prospectively collected data resulted in almost identical findings, which indicates that severe differential recall bias between the two sets of information is unlikely. The information on binge drinking was obtained from two questions, which have been found to provide reliable information on binge drinking in pregnancy.5 Still, accurate reports on the timing of binge drinking in pregnancy are difficult to obtain, and this may mask the association between the timing of the binge episodes and fetal death. The susceptible time periods during pregnancy may also vary between individuals, and the ability to time a binge episode correctly may correlate with an exposure that is neurotoxic.
In conclusion, we did not find the frequency and timing of binge drinking during the first 16 weeks of pregnancy to be associated with the risk of spontaneous abortion in clinically recognized pregnancies, but three or more episodes of binge drinking was related to an increased risk of stillbirth. On the basis of a precautionary principle, we would suggest that pregnant women and pregnancy planners should be advised to avoid alcohol consumption, including binge drinking.
1. West JR, Kelly SJ, Pierce DR. Severity of alcohol-induced deficits in rats during the third trimester equivalent is determined by the pattern of exposure. Alcohol Alcohol Suppl 1987;1:461–5.
2. Bonthius DJ, Goodlett CR, West JR. Blood alcohol concentration and severity of microencephaly in neonatal rats depend on the pattern of alcohol administration. Alcohol 1988;5:209–14.
3. Pierce DR, West JR. Blood alcohol concentration: a critical factor for producing fetal alcohol effects. Alcohol 1986;3:269–72.
4. Andersen AM, Olsen J, Gronbaek MN. Did the changed guidelines on alcohol and pregnancy by the National Board of Health and Welfare change alcohol consumption of pregnant women [in Danish]? Ugeskr Laeger 2001;163:1561–5.
5. Kesmodel U. Binge drinking in pregnancy: frequency and methodology. Am J Epidemiol 2001;154:777–82.
6. Naimi TS, Brewer RD, Mokdad A, Denny C, Serdula MK, Marks JS. Binge drinking among US adults. JAMA 2003;289:70–5.
7. Ebrahim SH, Luman ET, Floyd RL, Murphy CC, Bennett EM, Boyle CA. Alcohol consumption by pregnant women in the United States during 1988–1995. Obstet Gynecol 1998;92:187–92.
8. Ebrahim SH, Diekman ST, Floyd RL, Decoufle P. Comparison of binge drinking among pregnant and nonpregnant women, United States, 1991–1995. Am J Obstet Gynecol 1999;180:1–7.
9. Kaufman MH. The teratogenic effects of alcohol following exposure during pregnancy, and its influence on the chromosome constitution of the pre-ovulatory egg. Alcohol Alcohol 1997;32:113–28.
10. Randall CL, Taylor J, Walker DW. Ethanol-induced malformations in mice. Alcohol Clin Exp Res 1977;1:219–24.
11. Clarren SK, Bowden DM, Astley SJ. Pregnancy outcomes after weekly oral administration of ethanol during gestation in the pig-tailed macaque (Macaca nemestrina). Teratology 1987;35:345–54.
12. Clarren SK, Astley SJ. Pregnancy outcomes after weekly oral administration of ethanol during gestation in the pig-tailed macaque: comparing early gestational exposure to full gestational exposure. Teratology 1992;45:1–9.
13. Kaufman MH. Ethanol-induced chromosomal abnormalities at conception. Nature 1983;302:258–60.
14. Windham GC, Von Behren J, Fenster L, Schaefer C, Swan SH. Moderate maternal alcohol consumption and risk of spontaneous abortion. Epidemiology 1997;8:509–14.
15. Russell M, Skinner JB. Early measures of maternal alcohol misuse as predictors of adverse pregnancy outcomes. Alcohol Clin Exp Res 1988;12:824–30.
16. Olsen J, Melbye M, Olsen SF, Sorensen TI, Aaby P, Andersen AM, et al. The Danish National Birth Cohort: its background, structure and aim. Scand J Public Health 2001;29:300–7.
17. Andersen AM, Vastrup P, Wohlfahrt J, Andersen PK, Olsen J, Melbye M. Fever in pregnancy and risk of fetal death: a cohort study. Lancet 2002;360:1552–6.
18. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology 1999;10:37–48.
19. Apgar BS, Churgay CA. Spontaneous abortion. Prim Care 1993;20:621–7.
20. Randall CL, Anton RF, Becker HC. Alcohol, pregnancy, and prostaglandins. Alcohol Clin Exp Res 1987;11:32–36.
21. Klonoff-Cohen H, Lam-Kruglick P, Gonzalez C. Effects of maternal and paternal alcohol consumption on the success rates of in vitro fertilization and gamete intrafallopian transfer. Fertil Steril 2003;79:330–9.
22. Henriksen TB, Hjollund NH, Jensen TK, Bonde JP, Andersson AM, Kolstad H, et al. Alcohol consumption at the time of conception and spontaneous abortion. Am J Epidemiol 2004;160:661–7.
23. Nohr EA, Frydenberg M, Henriksen TB, Olsen J. Does low participation in cohort studies induce bias? Epidemiology 2006;17:413–8.
24. Kesmodel U, Schioler Kesmodel P. Drinking during pregnancy: attitudes and knowledge among pregnant Danish women, 1998. Alcohol Clin Exp Res 2002;26:1553–60.
© 2008 The American College of Obstetricians and Gynecologists
25. Strandberg-Larsen K, Nybo Andersen AM, Olsen J, Nielsen RN, Grønbæk M. Do women give the same information on binge drinking during pregnancy when asked repeatedly? Eur J Clin Nutr 2006;60:1294–8.