Association between alcohol abuse during pregnancy and birth weight

OBJECTIVE: To assess the association between alcohol abuse during gestation and low birth weight. METHODS: Cross-sectional, population-based nested study from a cohort of 957 pregnant women who received prenatal assistance through Sistema Único de Saúde (National Health System) in the city of Pelotas, Southern Brazil, and delivered their babies between September 2007 and September 2008. The mothers were interviewed at two distinct moments: prenatal and postpartum periods. In order to verify alcohol abuse, the CAGE (Cut down, Annoyed by criticism, Guilty and Eye-opener) scale was used. Bivariate analyses were carried out, as well as multiple logistic regression adjusted by the variables prematurity and alcohol abuse. The level of signifi cance that was adopted was 95%. RESULTS: Of the women who participated in the study, 2.1% abused alcohol during pregnancy and, among these, 26.3% had low birth weight children. There was an association between alcohol abuse and low birth weight (p<0.038). CONCLUSIONS: The fi ndings indicate that alcohol abuse during pregnancy is associated with low birth weight. DESCRIPTORS: Pregnant Women. Alcohol Drinking, adverse effects. Infant, Low Birth Weight. Cross-Sectional Studies.


INTRODUCTION
Birth weight is an important parameter related to perinatal and child morbidity and mortality. 18Low birth weight is defi ned by the World Health Organization (WHO) as weight at birth of less than 2,500 grams, 20 and this is one of the factors that determine neonatal mortality, higher risk of infections, greater propensity to growth delay and postnatal neuropsychological defi cit.In addition, it may infl uence the health conditions of the adult. 8The rates undergo great variations in diverse regions of the world, with evident disadvantages to the less developed countries, because they are associated with unfavorable socioeconomic conditions. 3,9A Brazilian study has shown a 10% incidence of low birth weight. 2 Low birth weight is strongly related to harmful maternal habits, mainly the consumption of alcoholic beverages. 15The literature shows higher risk of malformations, spontaneous miscarriage, low birth weight, prematurity, asphyxia and perinatal mortality, besides several physical and mental problems deriving from the Fetal Alcohol Syndrome. 18During the gestational period, the prevalence of inadequate alcohol use varies from 2.0% to 40.6%.The alcohol that the mother ingests during gestation crosses the placental barrier and the fetus is exposed to this substance that is present in the maternal blood.Fetal exposure is greater due to its slow metabolism and elimination system; thus the amniotic fl uid remains impregnated with unmodifi ed alcohol (ethanol) and acetaldehyde (a metabolite of ethanol). 7hanol induces the formation of oxygen free radicals that are capable of damaging cellular proteins and lipids, which increases apoptosis or "programmed cellular death" and negatively affects cellular divisions and specifi cations.Ethanol also inhibits the synthesis of retinoic acid, a substance that regulates embryonic development. 7The consumption of 20 grams of alcohol seems to be enough to suppress fetal breathing and movements. 4cohol consumption may be underdiagnosed during gestation, probably due to the health professionals' unpreparedness to investigate adequately or to give importance to complaints compatible with the habit of drinking. 15 addition to alcohol abuse during pregnancy, other factors may be associated with low weight, like low maternal age, not receiving prenatal care, female newborn, prematurity, low socioeconomic condition and smoking during pregnancy. 17,19us, the present study aimed to analyze the association between alcohol abuse during gestation and low birth weight.

METHODS
The study is nested to a cohort that monitored families since the prenatal period until 12 months in the postpartum period, with the objective of evaluating psychiatric disorders in mothers and fathers and child developmental delay.The sample was constituted by 1,062 women who received prenatal care through the Sistema Único de Saúde (Brazil's National Health System) of the city of Pelotas, Southern Brazil, and delivered their babies between September 2007 and September 2008.The women were captured by means of the Sistema de Acompanhamento do Programa de Humanização no Pré-Natal e Nascimento (System to Accompany the Prenatal and Birth Humanization Program, SISPRENATAL) of the Municipal Department of Health.The SISPRENATAL is a government program that stimulates primary care units to provide prenatal assistance.It provides funds for the unit every time a pregnant woman is enrolled.Data such as address, phone number and gestational age (calculated based on the date of the last menstruation) were obtained from the program's medical record.The only exclusion criterion was to reside in the rural zone of the city of Pelotas (n = 44).
The women were interviewed at their homes between 30 and 45 days after delivery by means of a questionnaire about economic conditions, maternal characteristics, prenatal care and smoking during pregnancy.Alcohol abuse was verifi ed by means of the CAGE scale (Cut down, Annoyed by criticism, Guilty and Eye-opener), validated to Brazil by Castells & Furlanetto. 5 This scale considers that two or more affi rmative answers are indicative of alcohol abuse.The child's weight and sex data, as well as gestational age at delivery, by means of which prematurity was verifi ed, were obtained from the baby's health card.
The newborns up to 2,499 grams were considered low weight.The families' socioeconomic level was classifi ed into fi ve levels (A, B, C, D and E), according to Associação Brasileira de Empresas de Pesquisa (ABEP -Brazilian Survey Companies Association), a which is based on the accumulation of material goods and on the level of schooling of the head of household.In the present study, the levels were merged into three (A and B; C; D and E).Maternal age was divided into three categories: up to 19 years, 20 to 34 and 35 or older.
After the collection, the data were double-entered in the program EpiInfo 6.04 for comparison and elimination of inconsistencies.The statistical package SPSS 10.0 for Windows was used for data analysis.Univariate analysis by simple frequency was conducted to learn about the sample's characteristics, and bivariate analysis (chi-square test), for comparison between proportions.To control for possible confounding factors, logistic regression was used, based on the following hierarchical model: in the first level, socioeconomic classifi cation and maternal age; in the second level, prenatal care, smoking during gestation, newborn's sex, prematurity, and alcohol abuse; and as the outcome, low birth weight.
To verify the power of association, a signifi cance level of 95% was used, as well as power of association of 80%.The initial sample was constituted by 1,062 women.There was 11% (105) of losses or refusals in the second part of the study and the fi nal sample was constituted by 957 mothers.With a sample of 957 women, it was possible to establish a power of association of 99%.
The investigation was approved by the research ethics committee of Universidade Católica de Pelotas according to the norms that were in force at the time (CONEP-Res196/96), under the protocol no.2007/95.
Alcohol abuse was verifi ed in 19 (2.1%) women.There was association between prematurity and low weight (p<0.001) and alcohol abuse and low birth weight (p<0.038), and 26.3% of the mothers who reported alcohol abuse in gestation delivered low birth weight babies.The variables that presented association with birth weight in the bivariate analyses with level of signifi cance <0.20 were included in the regression models.
In the multiple regression analysis, after adjustments for prematurity and alcohol abuse, low birth weight remained associated with prematurity (p<0.001) and alcohol abuse in gestation (p<0.020).The preterm babies were 11 times more likely to be born at low weight than those who were born at term, even controlling for alcohol abuse (p<0.020).The likelihood of a baby being born at low weight was four times higher among those whose mothers abused alcohol during gestation (OR= 4.20; 95%CI: 1.25;14.13).

DISCUSSION
The present study revealed a 10% incidence of low birth weight, a value that is similar to the one found in Brazilian studies. 2,8The prevalence of alcohol abuse was of 2.1%; however, other studies have shown variation in alcohol consumption during pregnancy from 2.0% to 40.6%, depending on the utilized investigation method. 1,2,18Moraes & Reichenheim 18 found a 7.3% prevalence of alcohol abuse in gestation.The low percentage of alcohol abuse in the present study may derive from the utilization of the CAGE instrument.Moraes & Reichenheim 18 argue that the cutoff point of this instrument (two affi rmative answers) is high when employed in pregnant women, because it suggests a positive screening only for alcohol abuse or dependence. 18Thus, the CAGE instrument does not take into account lower alcohol ingestions; it only considers them as suspicion of alcohol problems.In gestation, such situation would be of great importance, since safe levels of alcohol consumption in this period are not known.Another important factor that may justify the low prevalence of alcohol abuse during gestation is the question of social acceptability.Pregnant women are usually aware that alcohol consumption is harmful to the fetus, and tend to omit the use of alcoholic beverages due to fear of being disapproved both by society and by the health services.
Nevertheless, even with the low prevalence that was found, the association between alcohol abuse and low birth weight after the adjustment should not be minimized.The literature shows variation from 32.0% to 48.1% in the incidence of low birth weight in children born to mothers who ingested alcohol during gestation. 12,16In the present study, 26.3% of the mothers who abused alcohol delivered low weight babies.
The statistical analysis identifi ed that preterm children are more likely to be born weighing less than 2,500g.
Birth weight is strongly associated with prematurity.
Children born before 37 weeks were almost 35 times more likely to weigh less than 2,500g than those whose gestational age was higher. 3The results found by Carniel et al 3 are similar to the others from the literature that were mentioned above, with a 28.7% incidence among low weight infants who were born premature.
The mechanisms through which alcohol affects intrauterine growth are not well known; therefore, we decided to maintain the adjustment for prematurity in the hierarchical model.Although the causal chain alcoholprematurity-low weight is known, the present study confi rmed that the association between alcohol and low weight occurs even in the absence of prematurity.
Zambonato et al 21 have shown greater association between smoking in gestation and intrauterine growth restriction than with low weight.In the present study, smoking during gestation did not associate significantly with low birth weight, perhaps because it was underdiagnosed, like alcohol abuse, due to social acceptability.The pregnant women, fearing a possible recrimination and disapproval on the part of health professionals, may expose less consumption of these substances or even deny it.
Giglio et al 9 have described association between maternal age and low birth weight, with higher occurrence of low weight among children born to mothers who are in the extremes of reproductive life, that is, younger than 19 years or older than 35 years.This association was not found in the present study, but 69.7% of the pregnant women who were interviewed were between 20 and 34 years old.Kassar et al 13 studied adolescent and adult mothers younger than 35 years and showed that maternal age did not infl uence the variation in birth weight.Thus, the infl uence of maternal age on low birth weight is contradictory, because the mother's chronological age, by itself, is not a good predictor of birth weight.The complexity of the evaluation derives from the consideration of the biological expression of an event that does not occur in isolation from other dimensions.It is fundamental to consider the interaction of many dimensions, such as the socioeconomic, psychological and cultural, over the considered and/or studied biological responses. 13study has shown that poorer children have a 2.4 times higher risk of being born at low weight than those of richer families. 10In the present study there was no association between socioeconomic class and low birth weight; however, it must be considered that the sample was constituted by 54.5% of women from class C, which is considered intermediate.
Prenatal care was not related to low birth weight.This relation has been questioned by studies such as the one by Hueston et al, 11 which did not fi nd lower rates of low weight babies among North-American women who received early prenatal assistance.However, in the present study 88.9% of the women received prenatal assistance, whereas the others, although they denied having received prenatal assistance, attended at least one consultation of this type, as the sample was extracted from a prenatal enrolment program.
A possible limitation of the present study is related to the non-utilization of an instrument considered to be the gold standard to identify the many stages of alcohol use during gestation.The CAGE explores specifi cally the social effects of the inadequate use of alcohol, based on the patient's perception of his/her own habits, and on the perception of his/her relatives and friends, which might cause misunderstandings in the interpretation of the fi ndings. 18Yet, the CAGE is adequate to track alcohol abuse, to the detriment of the clinical interview, which is diffi cult to perform when we are dealing with primary care.Even with the non-ideal characteristics of the instrument, the fi nding of the present study should be valued.The reason is that the cutoff point of the above-mentioned instrument has high specifi city, which makes the 2.1% of women who abused alcohol during pregnancy be relevant in view of the association with low birth weight.
To reduce the consumption of alcoholic beverages during gestation, multidisciplinary work is necessary.When alcohol use is diagnosed, the patient should promptly receive intensive treatment, with psychological support and approaches that motivate her to change.Educational actions since the beginning of pregnancy and home visits increase adherence to the treatment and the chances of reduction or abandonment of alcohol ingestion during gestation.b The mechanisms through which alcohol affects the conceptus have not been fully clarifi ed yet.It is believed that the fetus is exposed to alcohol concentrations that are similar to those of the maternal blood, which makes the environment inadequate to the fetus and favors the incidence of the Fetal Alcohol Syndrome.This is characterized by damage to the central nervous system, which causes neurological and craniofacial anomalies, defi cit in pre-and post-natal growth, behavioral disorders and emotional diffi culties. 6,7e data showed in the present study are a warning so that health professionals track the abusive use of alcohol by pregnant women, since there is an indication that alcohol consumption in gestation is related to factors that may affect delivery, like premature placental abruption, uterine hypertonia, premature labor and meconium-stained amniotic fl uid. 7 b Babor TF, Higgins-Biddle J C. Manual de intervenção breve para indivíduos com problemas decorrentes do uso de álcool atendidos nos serviços de atenção primária.Geneva: World Health Organization; 2004 [cited 2009 Dec 16].Available from: http://apps.einstein.br/alcooledrogas/novosite/atualizacoes/as_203.html

Table .
Distribution of the sample and of low weight newborns, according to maternal characteristics and prematurity condition.Pelotas, Southern Brazil, 2007-2008.