Alcohol Drinking and Risk of Small for Gestational Age Birth

ORIGINAL ARTICLE Alcohol drinking and risk of small for gestational age birth

F Chiaffarino1, F Parazzini1,2, L Chatenoud1, E Ricci2, F Sandretti2, S Cipriani1, D Caserta3 and L Fedele2

1Istituto di Ricerche Farmacologiche ‘Mario Negri’, Milano, Italy; 2II Clinica Ostetrico Ginecologica, Universita` di Milano, Milano, Italy and 3Dipartimento di Scienze Ginecologiche, Universita` La Sapienza, Rome, Italy

Objective: To assess if alcohol drinking is a risk factor for small for gestational age (SGA) birth. Methods: Case–control study. Cases were 555 women (mean age 31 years, range 16–43) who delivered SGA babies at the Clinica Luigi Mangiagalli and the Obstetric and Gynecology Clinic of the University of Verona. The controls were 1966 women (mean age 31 years, range 14–43) who gave birth at term (X37 weeks of gestation) to healthy infants of normal weight at the hospitals where cases had been identified. Results: No increase in the risk of SGA birth was observed in women drinking one or two drinks/day in pregnancy, but three or more per day increased the risk: odds ratios (OR) were 3.2 (1.7–6.2) for X3 drinks during the first trimester, 2.7 (1.4–5.3) during the second and 2.9 (1.5–5.7) during the third. Conclusions: The study shows an increased risk of SGA births in mothers who drink X3 units/day of alcohol in pregnancy. European Journal of Clinical Nutrition (2006) 60, 1062–1066. doi:10.1038/sj.ejcn.1602419; published online 22 February 2006

Keywords: small for gestational age birth; alcohol; case–control study

Introduction during 1988–1991, light or mild alcohol consumption during the first month of pregnancy was associated with a Maternal alcohol abuse during pregnancy has been consis- significant protective effect against SGA babies (Lundsberg tently associated with reduced birth weight (IARC, 1988). et al., 1997). Likewise, in a Canadian survey, light consumers However, despite its clinical significance, the effect of of alcohol had a small but significant reduction in the risk of moderate alcohol drinking in pregnancy on the risk of small SGA birth relative to total abstainers (McDonald et al., 1992). for gestational age (SGA) birth is controversial (Day et al., Other studies have found no effect of alcohol intake on SGA 1989; Shu et al., 1995; Windham et al., 1995; Passaro et al., births risk. (Sulaiman et al., 1988; Faden and Graubard, 1996). 1994). Also unclear is the potentially different role of alcohol In an American prospective study based on 633 women, a drinking before pregnancy or in the different trimesters. marked increase in the frequency of SGA births was noted We analyzed the association between alcohol drinking with increased alcohol intake, rising to 27% among heavy before and during the three trimesters of pregnancy using drinkers (daily average of more than 45 ml of absolute data from a case–control study conducted in Italy. This study alcohol) (Oulette et al., 1977). However, other studies provides the opportunity to analyze the effect of alcohol reported a protective effect of low–moderate alcohol con- intake on the risk of SGA birth, preterm or at term, and the sumption on SGA birth risk. In a prospective investigation potential interaction between alcohol drinking and risk based on 2714 singleton live births conducted in the USA factors for SGA birth. A specific point of interest is that the study analyzes the association between alcohol drinking and SGA birth in a Southern European population, where Correspondence: Dr F Parazzini, Istituto di Ricerche Farmacologiche ‘Mario moderate alcohol drinking is more widespread than in North Negri’, Via Eritrea, 62 20157 Milano, Italy. American or European populations, where there may be a E-mail: [email protected] Received 11 April 2005; revised 21 December 2005; accepted 11 January lower prevalence of drinkers, but a higher frequency of heavy 2006; published online 22 February 2006 drinkers. Alcohol drinking and risk of small for gestational age birth F Chiaffarino et al 1063 Materials and methods To account simultaneously for the effects of several potential confounding factors, we used conditional multiple We conducted a case–control study on risk factors for SGA logistic regression (age as matching variable), with max- birth from February 1989 to October 1999. Cases were 555 imum likelihood fitting, to obtain odds ratios (OR) and their women (mean age 31 years, range 16–43) who delivered SGA corresponding 95% CI (Breslow and Day, 1980). Included in babies at the Clinica Luigi Mangiagalli (the largest obstetric the regression equations were terms for education, plus hospital in Milan) and the Obstetric and Gynecology Clinic terms significantly associated in this data set with the risk of of the University of Verona. Small for gestational age was SGA birth. These factors are listed in the footnotes to tables. defined as birth weight less than the 10th percentile for the infant’s gestational age in comparison with the birth weight expected for the same gestational age and sex, according to Results Italian standards (Parazzini et al., 1995); 143 births were preterm SGA and 412 at term. Table 1 shows the distribution of cases and controls The controls were women who gave birth at term (X37 according to age, gestational week at birth and selected weeks of gestation) to healthy infants on randomly selected factors. Smoking during the first trimester of pregnancy, days at the hospitals where cases had been identified. history of SGA birth and gestational hypertension increased The interviewers surveyed the obstetric wards on days the risk of SGA birth. Parous women and women with established at random to interview controls whose age was maternal body mass index (BMI) at conception between 21 comparable with cases. Cases and controls were not matched and 25 kg/m2 were at lower risk. for age, but the interviewers were instructed to select No significant association emerged between marital status controls comparable with cases for quinquennia of age. and risk of SGA birth. Likewise, no association emerged A total of 1966 controls (mean age 31 years, range 14–43) between coffee intake in the third trimester of pregnancy were interviewed. and risk of SGA birth. In comparison with women without Each control was interviewed within 1 month the case was nausea, the OR of SGA birth was 0.8 (95% CI 0.6–1.1) in interviewed. women with nausea in the first trimester. Overall participation rate was over 95% for cases and Alcohol use before pregnancy and in the three trimesters controls. of pregnancy and risk of SGA birth are considered in Table 2. Among the cases surveyed, no fetal alcohol syndrome was When pregnancy was confirmed, women tended to stop observed. drinking, the prevalence of non-drinkers in the control Information was obtained on general socio-demographic group rising from 54.6% before pregnancy to 66.0% in the habits, personal characteristics and habits, gynecological and first trimester. obstetric history, smoking and alcohol consumption in Small for gestational age births were more common among pregnancy. women reporting three or more drinks/day during preg- Data on nausea in pregnancy were collected for each nancy: the OR was 3.2 (95% CI 1.7–6.2) for three or more trimester of pregnancy in terms of intensity (none, low, drinks/day during the first trimester, 2.7 (95% CI 1.4–5.3) moderate or serious) but only for 286 cases (51.5%) and 1232 during the second and 2.9 (95% CI 1.5–5.7) during the third. controls (62.7%) because this information was only collected No association emerged between SGA birth risk and lower from 1993. We defined gestational hypertension as diastolic levels of alcohol drinking. pressure above 90 mm Hg on at least two occasions 24 h Table 3 considers the effect of alcohol drinking on the risk apart, without proteinuria. The presence of hypertension in of SGA birth during the third trimester of pregnancy in strata pregnancy was checked with clinical records. of selected covariates. No appreciable heterogeneity or effect Specific attention was paid to obtaining information from modification was found across strata of maternal age, cases and controls on alcohol consumption at conception education, parity and gestational hypertension and smoking, and during the trimesters of pregnancy. Questions on although the association was apparently stronger for smo- alcohol included the number of days per week each type of kers, where the OR tended to be higher. alcoholic beverage (red and white wine, beer and spirits) was consumed before pregnancy and in each trimester, the average number of drinks per day and the duration of the Discussion habit in years. From these data, we estimated the total daily average alcohol intake assuming a comparable pure alcohol Potential limitations of this study should be considered. The content in each type of drink (125 ml red wine ¼ 333 ml choice as controls of women who delivered healthy infants beer ¼ 30 ml spirits ¼ B12 g pure alcohol). Wine alone ac- only at term may have introduced some bias. However, when counted for B90% of the alcohol consumed by women in we considered only cases who delivered after 37 weeks’ this population. All information, apart from birth weight gestation, no differences emerged in the OR. In fact, the OR and week of gestation at birth, were obtained from personal of SGA birth after 37 weeks’ gestation in comparison with interview. non-alcohol drinkers were, respectively, 0.9 (95% CI 0.6–

European Journal of Clinical Nutrition Alcohol drinking and risk of small for gestational age birth F Chiaffarino et al 1064 Table 1 Distribution of 555 cases of SGA birth and 1966 controls according to selected factorsa

SGA Controls OR (95% CI)

Age (years) o25 60 (10.8) 180 (9.2) 25–29 176 (31.7) 637 (32.4) 30–34 207 (37.3) 767 (39.0) X35 112 (20.2) 382 (19.4)

Gestational week at birth (mean range) 37.4 (28–43) 39.6 (37–44)

Education (years) o6 24 (4.3) 53 (2.7) 1b 7–13 415 (74.8) 1439 (73.2) 0.8 (0.4–1.4) 414 116 (20.9) 473 (24.1) 0.7 (0.4–1.4)

Marital status Married 541 (97.5) 1934 (98.4) 1b Unmarried 14 (2.5) 32 (1.6) 1.2 (0.6–2.4)

Parity 0 404 (72.8) 1103 (56.1) 1b 1 118 (21.3) 693 (35.3) 0.4 (0.3–0.5) X2 33 (6.0) 170 (8.7) 0.4 (0.2–0.6)

Smoking during the third trimester of pregnancy No 414 (74.6) 1711 (87.0) 1b Yes 141 (25.4) 255 (13.0) 2.4 (1.9–3.1)

Coffee consumption during the third trimester of pregnancy No 226 (40.7) 750 (38.2) 1b Yes 329 (59.3) 1216 (61.9) 0.9 (0.7–1.1)

Previous SGA birthc No 133 (88.1) 853 (98.8) 1b Yes 18 (11.9) 10 (1.2) 10.6 (4.6–24.1)

Maternal BMI at conception (kg/m2) p20 255 (46.2) 776 (39.6) 1b 21–25 248 (44.9) 982 (50.1) 0.7 (0.6–0.9) 425 49 (8.9) 201 (10.3) 0.7 (0.5–1.0)

Hypertension in pregnancy No 435 (78.4) 1863 (94.8) 1b Yes 120 (21.6) 103 (5.2) 5.1 (3.8–6.9)

Nausea during the first trimester of pregnancy No 126 (44.1) 441 (35.8) 1b Yes 160 (55.9) 791 (64.2) 0.8 (0.6–1.1)

Abbreviations: BMI, body mass index (kg/m2); CI, confidence interval; OR, odds ratios adjusted in turn for education, parity and smoking during the third trimester of pregnancy, gestational hypertension, history of SGA birth; SGA, small for gestational age. aIn some cases, the sum does not add up to the total because of some missing values. bReference category. cParous women only.

1.2), 1.2 (95% CI 0.8–1.8) and 2.9 (95% CI 1.4–6.1) for Information on alcohol drinking was retrospectively women drinking one, two, and three or more portions of collected after delivery. The assessment of alcohol consump- alcohol a day in the third trimester of pregnancy (data not tion was based on self-reporting, so some underestimates shown). More in general, the interest of this analysis was to could have occurred. However, in Italy, alcohol consumption define the effect of alcohol consumption on the risk of SGA is socially accepted and recommendations to avoid alcohol birth in comparison with normal pregnancies, and the in pregnancy have not received widespread attention and are inclusion in the control group of women with adverse not routinely advocated by gynecologists. Good reporting of pregnancy outcomes (i.e. preterm birth) potentially related alcohol intake at post partum interview by women deliver- to alcohol drinking might spuriously affect the potential ing normal or low birth weight infant has been documented relation between alcohol and SGA. (Delgado-Rodriguez et al., 1995).

European Journal of Clinical Nutrition Alcohol drinking and risk of small for gestational age birth F Chiaffarino et al 1065 Table 2 OR of SGA birth according to alcohol drinkinga Table 3 OR of SGA birth according to alcohol drinking during the third trimester of pregnancy and selected covariatesa SGA Controls OR (95% CI) No. (%) No. (%) Alcohol drinking during the third trimester of pregnancy (units/day)

12X3 Alcohol drinking before conception (Units/day) 0 303 (54.6) 1132 (57.6) 1b 0.5 60 (10.8) 245 (12.5) 0.9 (0.7–1.3) Maternal age (years) 1 97 (17.5) 333 (16.9) 1.1 (0.8–1.5) p30 0.9 (0.6–1.5) 1.7 (0.9–3.0) 2.1 (0.8–5.2) X 2 69 (12.4) 207 (10.5) 1.2 (0.9–1.7) 31 0.8 (0.6–1.3) 0.8 (0.4–1.3) 5.1 (1.8–14.5) X3c 26 (4.7) 49 (2.5) 1.7 (1.0–2.9) Education (years) First trimester p11 0.8 (0.5–1.3) 1.5 (0.8–2.8) 3.0 (1.0–8.5) b X 0 366 (66.0) 1345 (68.4) 1 12 0.9 (0.6–1.4) 0.9 (0.5–1.6) 3.0 (1.2–7.4) 0.5 63 (11.4) 187 (9.5) 1.1 (0.8–1.6) 1 65 (11.7) 281 (14.3) 0.9 (0.6–1.2) Parity 2 38 (6.9) 131 (6.7) 1.1 (0.7–1.7) 0 0.8 (0.6–1.2) 1.1 (0.7–1.9) 2.1 (0.9–5.1) c X X3 23 (4.1) 22 (1.1) 3.2 (1.7–6.2) 1 1.0 (0.6–1.6) 1.0 (0.5–1.9) 4.3 (1.5–12.1)

rd Second trimester Smoking during 3 trimester 0 371 (66.9) 1353 (68.8) 1b No 0.8 (0.6–1.2) 1.0 (0.7–1.7) 2.0 (0.9–4.8) 0.5 61 (11.0) 190 (9.7) 1.1 (0.8–1.5) Yes 1.1 (0.6–2.0) 1.2 (0.5–2.6) 7.6 (1.8–32.2) 1 66 (11.9) 273 (13.9) 0.9 (0.7–1.2) 2 38 (6.9) 127 (6.5) 1.1 (0.7–1.7) Gestational hypertension X3d 19 (3.4) 23 (1.2) 2.7 (1.4–5.3) No 0.9 (0.6–1.2) 1.1 (0.7–1.7) 3.2 (1.6–6.2) Yes 0.9 (0.4–2.1) 0.8 (0.2–3.3) 1.2 (0.1–23.9) Third trimester b Reference category: no alcohol drinking. 0 372 (67.0) 1319 (67.1) 1 a 0.5 58 (10.5) 204 (10.4) 0.9 (0.7–1.3) Adjusted in turn for age, education, parity, smoking during the third trimester 1 68 (12.3) 292 (14.9) 0.8 (0.6–1.2) of pregnancy and gestational hypertension.Abbreviation: OR, odds ratio. 2 38 (6.9) 129 (6.6) 1.1 (0.7–1.6) X3d 19 (3.4) 22 (1.1) 2.9 (1.5–5.7)

Abbreviations: CI, confidence interval; OR, odds ratios adjusted in turn for education, parity and smoking during the third trimester of pregnancy, birth weight approximately 150 g less than those whose gestational hypertension, history of SGA birth; SGA, small for gestational age. mothers abstained during pregnancy. No effect was seen in a In some cases, the sum does not add up to the total because of some missing infants born to women who drank weekly in early preg- values. b nancy, less than weekly or not at all (Passaro et al., 1996). Reference category. cWomen drinking 10 Units/day or more: two cases of SGA and four controls. Shu et al. (1995) found that alcohol consumption in the dWomen drinking 10 Units/day or more: one case of SGA and two controls. first trimester (average four drinks per week) was associated with a 155 g reduction in fetal weight, even after adjustment for number of cigarettes smoked. In a recent US survey of Other sources of bias, including selection or confounding women who delivered a live infant (Whitehead and factors, are unlikely to have produced marked effects, Lipscomb, 2003), SGA births were more common among especially considering that cases and controls were inter- moderate (4–13 drinks/week) and heavy (14 or more drinks/ viewed in the same institution and that participation was week) drinkers during the last 3 months than non-drinkers. practically complete. Further, the findings were largely Smoking and drinking three or more drinks/day during the consistent when the analysis was run separately for cases third trimester have a marked effect on SGA birth. Some and controls collected in the two hospitals or by different other studies showed an association between alcohol intake interviewers or when the models included terms for center and smoking, and the effect of smoking on birth weight was and interviewers (data not shown). Finally, the questionnaire independent of but additive with the effect of alcohol. In a was satisfactorily reproducible (D’Avanzo et al., 1997). Californian study on 1233 women, moderate drinkers (X3 We did not analyze information on nutritional status, but drinks/week) who also smoked more than half a pack/day inclusion of BMI in the model did not change the estimated had a greater weight decrement than was expected on the OR (data not shown). basis of additivity of effects (Windham et al., 1995). In a Our study suggests that there is no risk of SGA birth when population-based case–control study in New York on 701 one or two drinks a day are taken, and that there is, at most, cases of SGA birth, smoking increased the risk of SGA birth, an effect at higher levels. These findings are in general which was even greater for heavy alcohol consumption agreement with other studies, but some differences in the (Yang et al., 2001). study design and in the definitions of alcohol consumption The biological mechanism through which alcohol drink- make the results difficult to compare. ing affects fetal growth is not completely understood. In a British cohort study, infant born to women who Alcohol or its metabolites may affect fetal development by reported three or more drinks daily had an adjusted mean causing hypoxia, impairing cell proliferation or affecting

European Journal of Clinical Nutrition Alcohol drinking and risk of small for gestational age birth F Chiaffarino et al 1066 placental development (Abel, 1982). The effects of moderate IARC monographs on evaluation of carcinogenic risks to human maternal alcohol consumption may be multifactorial, and (1988). Alcohol Drinking, vol. 44 Lyon, France. the combined effects with other environmental and genetic Lundsberg LS, Bracken MB, Saftlas AF (1997). Low-to-moderate gestational alcohol use and intrauterine growth retardation, low factors may increase the risk for SGA birth. Smoking may be birthweight, and preterm delivery. Ann Epidemiol 7, 498–508. one of these factors. McDonald AD, Armstrong BG, Sloan M (1992). Cigarette, alcohol, In conclusion, the study shows an increased risk of SGA and coffee consumption and prematurity. Am J Public Health 1992, 82:87–82:90. births only in mothers who drink three units/day or more of Oulette EM, Rosett HL, Rosman P, Weiner L (1977). Adverse effects on alcohol in pregnancy. offspring of maternal alcohol abuse during pregnancy. N Engl J Med 297, 528–530. Parazzini F, Cortinovis I, Bortolus R, Fedele L, Decarli A (1995). Weight at birth by gestational age in Italy. Hum Reprod 10, 1862– References 1963. Passaro KT, Little RE, Savitz DA, Noss J, ALSPAC Study Team (1996). Abel EL (1982). Consumption of alcohol during pregnancy: a review The effect of maternal drinking before conception and in early of effects on growth and development of offpring. Hum Biol 54, pregnancy on infant birthweight. Epidemiology 7, 377–383. 421–453. Shu XO, Hatch MC, Mills J, Clemens J, Susser M (1995). Maternal Breslow NE, Day NE (1980). Statistical Methods in Cancer Research. The smoking, alcohol drinking, caffeine consumpion, and fetal Analysis of Case Control Studies, IARC Publication No. 32, vol. I. growth: results from a prospective study. Epidemiology 6, 115–120. IARC: Lyon, France. Sulaiman ND, Florey CD, Taylor DJ, Ogston SA (1988). Alcohol Day NL, Jasperse D, Richardson G, Robles N, Sambamoorthi U, Taylor consumption in Dundee primigravidas and its effects on outcome P et al. (1989). Prenatal exposure to alcohol: effect on infant of pregnancy. BMJ 296, 1500–1503. growth and morphologic characteristics. Pediatrics 84, 536–541. Whitehead N, Lipscomb L (2003). Patterns of alcohol use before and D’Avanzo B, La Vecchia C, Katsouyanni K, Negri E, Trichopoulos D during pregnancy and the risk of small for gestational age birth. (1997). An assessment, and reproducibility of food frequency data Am J Epidemiol 158, 654–662. provided by hospital controls. Eur J Cancer Prev 6, 288–293. Windham GC, Fenster L, Hopkins B, Swan SH (1995). The association Delgado-Rodriguez M, Gomez-Olmedo M, Bueno-Cavanillas A, of moderate maternal and paternal alcohol consumption with Garcia-Martin M, Galvez-Vargas R (1995). Recall bias in a case– birthweight and gestational age. Epidemiology 6, 591–597. control study of low birth weight. J Clin Epidemiol 48, 1133–1140. Yang Q, Witkiewicz BB, Olney RS, Liu Y, Davis M, Khoury MJ et al. Faden VB, Graubard BL (1994). Alcohol consumption during (2001). A case–control study of maternal alcohol consumption pregnancy and infant birth weight. Ann Epidemiol 4, 279–284. and intrauterine growth retardation. Ann Epidemiol 11, 497–503.

European Journal of Clinical Nutrition