Documentation for LBW.DAT col variable code ----- ----------------- ------------------------------------------- 1- 2 Low birth weight # of cases (with low birth weight) in group 3- 6 Number of subjects # in group 7- 8 Social class 0 class I, II 1 class III 2 class IV, V 9-10 Alcohol use 0 heavy 1 moderate 2 light 11-12 Smoking 0 non-smoker 1 smoker -------------- There are 18 records Abstracted by jh from TABLE I - NUMBER OF BABIES WITH BIRTHWEIGHT LESS THAN 10TH CENTILE BY SMOKING CLASS AND SOCIAL CLASS in the article Alcohol Consumption, Pregnancy and Low Birthweight by Wright JT. Waterson EJ. Barrison IG. Toplis PJ. Lewis IG. Gordon MG. MacRae KD. Morris NF. Murray-Lyon IM. in Lancet. 1(8326 Pt 1):663-5, 1983 Mar 26. The fetal alcohol syndrome Ñ a combination of growth retardation, facial abnormalities, mental retardation, and other congenital anomalies Ñ was first described in English by Jones and co- workers[] and many cases have since been recognised throughout the world[]. This syndrome is found only in the offspring of mothers who regularly consume more than 80 g alcohol a day in pregnancy and seems to be rare in the United Kingdom[]. Whether more moderate alcohol consumption in pregnancy is harmful (for example, mid trimester abortion[], low birthweight, delayed achievement of developmental goals) is less clear, but the data were considered sufficiently suggestive for both the Royal College of Psychiatrists in Britain and the Surgeon General in the U.S.A. to recommend that pregnant women should abstain from alcohol. This advice, however, is controversial[]. The relation between drinking before or during pregnancy and birthweight has been investigated in various centers with conflicting results[]. None of these studies, however, has completely allowed for other confounding variables such as maternal age, smoking, and socioeconomic status. We have examined this question as part of a continuing study of drinking habits and pregnancy outcome in women attending the Charing Cross Hospital department of obstetrics in the West London Hospital. Patients and Methods Drinking histories were obtained from all women attending the antenatal booking clinic (usually between 8 and 14 weeks' amenorrhoea) by the interviewing doctor, using a questionnaire that we had validated previously[]. This concerned drinking before the patient knew she was pregnant. She was then asked to complete a further questionnaire (available on request from the authors) concerning alcohol consumption during pregnancy and, in particular, in the week preceding the visit to the clinic. This was completed at home and returned by post. This second questionnaire also contained questions on drug use and past occupations and included a shortened version of the Michigan Alcohol Screening Test[]. In addition, a dietary, drug, smoking, obstetric, and medical history was taken in the antenatal clinic. Routine blood tests included blood count and gamma- glutamyl transpeptidase. We have reported an evaluation of these screening tests for drinking in pregnancy. On the basis of the pregnancy questionnaire, the patients were classified into "heavy" drinkers, consuming 100 g or more alcohol a week; "moderate" drinkers, consuming 50-100 g alcohol a week; "light" drinkers, consuming less than 50 g alcohol a week, including teetotallers. No effort was made to influence drinking habits and the information was not available to the clinician looking after the patient. After delivery, the case-notes were reviewed and the pregnancy outcome was recorded. From the initial interview group of 1122 patients (delivery data available in 1108),we have studied all 900 white patients giving birth to a singleton fetus after 28 weeks' gestation, 85% of whom completed the second questionnaire. From the birthweight, gestation, and sex of the infant, centile weight was calculated from the Castlemead centile weight chart[]. Social class was assessed from the father's occupation according to the Registrar General's classification. Results 10.9% of the babies had a birthweight on or below the 10th centile. Table I shows the interrelations between level of drinking social class, cigarette smoking, and low birthweight. Overall, 20% of the mothers were heavy drinkers and 26% smoked during pregnancy. Effect of Reducing Alcohol During Pregnancy Only 31 (26%) of the non-smoking heavy drinkers continued to drink heavily during pregnancy and 5 (17%) of these produced a light baby. Of the remaining 14% who reduced their drinking to 100 g alcohol a week or less, 10% produced a baby whose weight was on the 10th centile or less. The difference in low birthweight outcome between these two groups does not approach statistical significance (X2 1.11; p=0.29). In the smoking group, 24 (38%) of heavy drinkers continued to drink heavily during pregnancy and 8 (33%) of those produced a light baby. Of the 62% of heavy drinkers who reduced their consumption to less than 100 g week, 30% produced a light baby (X2 = 0.06; p=0.81). These data do not suggest any benefit from reduction of heavy drinking in the later stages of pregnancy in either smokers or non-smokers. ... ------------------- See also Wacholder S. Binomial regression in GLIM: estimating risk ratios and risk differences American Journal of Epidemiology. 123(1):174-84, 1986.