3
Cigarette, Alcohol, and Coffee Consumption and Congenital Defects Alison D. McDonald, MD, Ben G. Amnstrong, PhD, and Maigaret Sloan, BA Inltdwton A report that smoking in pregnancy was associated with congenital heart de- fects, was not confirmed,2- even in two large series of births.5 Heavy alcohol con- sumption has been reported to cause re- tarded physical and mental development and certain facial and neurological abnor- malities (fetal alcohol syndrome). Moder- ate alcohol consumption has also been re- ported to affect fetal growth and morphogenesis.6 Some concern has been expressed that heavy coffee consumption may be teratogenic,7 although a case-con- trol study in Finland was negative.8 We investigated the relationships between cigarette, alcohol, and coffee consump- tion and congenital defects, using data from a survey of occupational factors in pregnancy conducted in Montreal from 1982 to 1984 (see the first paper in this series). In this survey no association was found between congenital defects and chemical exposures estimated from job ti- tle,9 but in a case-referent study an asso- ciation with aromatic solvents was found.10 The excess defects were mainly renal and urinary. The association of cig- arette, alcohol, and coffee consumption with congenital defects has now been an- alyzed for all current and previous preg- nancies of the 56 000 women inter- viewed-104 000 pregnancies in all. Medhds The study design and methods are briefly descnbed in the first of these re- ports, and have been described in detail elsewhere.-1 All current and previous pregnancies of more than 20 weeks' du- ration, whether or not the woman was employed during the pregnancy, were in- cluded in the present analysis. Therapeu- tic abortions for fetal defect were included irrespective of gestation length. Informa- tion on cigarette and alcohol consump- tion, and on potential confounding factors was available for 89 317 of these pregnan- cies. As questions on coffee consumption in previous pregnancies were introduced after the survey began, only 80 319 re- mained for study of this factor. Congenital defects were ascertained (blind) in current pregnancies from pedi- atric records.8 For previous pregnancies, the mother was asked whether any abnor- mality had been noted in the child. If a defect was suspected, further inquiries were made. The defects were classified into 90 groups. Minor abnormalities such as hydrocele, glandular hypospadias, un- descended testes, skin tags, nevi, and her- nias were omitted, as were known inher- ited defects, those that were poorly described (less than 1%), and those that became apparent only at a later stage of development, such as mental retardation, epilepsy, cerebral palsy, and deafness. In- terauricular and interventricular septal de- fects were retained if accompanied by an- other heart abnormality. Cases with multiple defects were classified under the one that appeared to be most serious. Eight types of congenital defect (shown in Tables 1 and 2) were entered into the pre- sent analyses. Possible confounding by maternal age, educational level, and ethnic (color and language) group was controlled by in- cluding these factors together with ciga- rette, alcohol, and coffee consumption in logistic regression analyses, which were performed separately for current and pre- vious pregnancies. Regression coeffi- cients for cigarette, alcohol, and coffee consumption from the two series were combined as a weighted mean to obtain a single odds ratio, as they did not signifi- cantly differ for any factor at a 5% prob- ability level. Results There was no consistent evidence of an association between cigarette con- sumption and any defect group, after The authors are with the School of Occupa- tional Health, McGill University, Montreal, Quebec. Requests for reprints should be sent to Ben G. Armstrong, PhD, School of Occupa- tional Health, McGill University, 1130 Pine Av- enue West, Montreal, Quebec, Canada H3A 1A3. This paper was submitted to the journal July 19, 1990, and accepted with revisions May 6, 1991. American Journal of Public Health 91

Cigarette, Alcohol, and Coffee Consumption and Congenital Defects

Embed Size (px)

Citation preview

Page 1: Cigarette, Alcohol, and Coffee Consumption and Congenital Defects

Cigarette, Alcohol, and CoffeeConsumption and Congenital Defects

Alison D. McDonald, MD, Ben G. Amnstrong, PhD, and Maigaret Sloan, BA

InltdwtonA report that smoking in pregnancy

was associated with congenital heart de-fects, was not confirmed,2- even in twolarge series of births.5 Heavy alcohol con-sumption has been reported to cause re-tarded physical and mental developmentand certain facial and neurological abnor-malities (fetal alcohol syndrome). Moder-ate alcohol consumption has also been re-ported to affect fetal growth andmorphogenesis.6 Some concern has beenexpressed that heavy coffee consumptionmay be teratogenic,7 although a case-con-trol study in Finland was negative.8 Weinvestigated the relationships betweencigarette, alcohol, and coffee consump-tion and congenital defects, using datafrom a survey of occupational factors inpregnancy conducted in Montreal from1982 to 1984 (see the first paper in thisseries). In this survey no association wasfound between congenital defects andchemical exposures estimated from job ti-tle,9 but in a case-referent study an asso-ciation with aromatic solvents wasfound.10 The excess defects were mainlyrenal and urinary. The association of cig-arette, alcohol, and coffee consumptionwith congenital defects has now been an-alyzed for all current and previous preg-nancies of the 56 000 women inter-viewed-104 000 pregnancies in all.

MedhdsThe study design and methods are

briefly descnbed in the first of these re-ports, and have been described in detailelsewhere.-1 All current and previouspregnancies of more than 20 weeks' du-ration, whether or not the woman was

employed during the pregnancy, were in-cluded in the present analysis. Therapeu-tic abortions for fetal defect were includedirrespective of gestation length. Informa-tion on cigarette and alcohol consump-tion, and on potential confounding factorswas available for 89 317 of these pregnan-cies. As questions on coffee consumptionin previous pregnancies were introducedafter the survey began, only 80 319 re-

mained for study of this factor.

Congenital defects were ascertained(blind) in current pregnancies from pedi-atric records.8 For previous pregnancies,the motherwas asked whether any abnor-mality had been noted in the child. If adefect was suspected, further inquirieswere made. The defects were classifiedinto 90 groups. Minor abnormalities suchas hydrocele, glandular hypospadias, un-descended testes, skin tags, nevi, and her-nias were omitted, as were known inher-ited defects, those that were poorlydescribed (less than 1%), and those thatbecame apparent only at a later stage ofdevelopment, such as mental retardation,epilepsy, cerebral palsy, and deafness. In-terauricular and interventricular septal de-fects were retained if accompanied by an-other heart abnormality. Cases withmultiple defects were classified under theone that appeared to be most serious.Eight types of congenital defect (shown inTables 1 and 2) were entered into the pre-sent analyses.

Possible confounding by maternalage, educational level, and ethnic (colorand language) group was controlled by in-cluding these factors together with ciga-rette, alcohol, and coffee consumption inlogistic regression analyses, which wereperformed separately for current and pre-vious pregnancies. Regression coeffi-cients for cigarette, alcohol, and coffeeconsumption from the two series werecombined as a weighted mean to obtain asingle odds ratio, as they did not signifi-cantly differ for any factor at a 5% prob-ability level.

ResultsThere was no consistent evidence of

an association between cigarette con-sumption and any defect group, after

The authors are with the School of Occupa-tional Health, McGill University, Montreal,Quebec.

Requests for reprints should be sent to

Ben G. Armstrong, PhD, School of Occupa-tional Health, McGill University, 1130 Pine Av-enue West, Montreal, Quebec, Canada H3A1A3.

This paper was submitted to the journalJuly 19, 1990, and accepted with revisions May6, 1991.

American Journal of Public Health 91

Page 2: Cigarette, Alcohol, and Coffee Consumption and Congenital Defects

Public Health Brefs

accounting for possible confounding (Ta-ble 1). Odds ratios (relative to nonsmok-ers) for mothers who smoked one to ninecigarettes per day were 1.84 for neuraltube, 1.53 for renal-urinary, and 1.14 forall defects. These values were on the bor-derline of statistical significance (at aboutP = .05); however, odds ratios werelower in heavier smokers.

There was some evidence of an in-creased risk ofmusculoskeletal defects inwomen who took alcohol, with risk in-creasing with the amount drunk (Table2). The odds ratio of 1.82 in women who

took seven or more drinks per week wassignificant at P = .05. Among the 13 de-fects in this group were 6 cases of con-genital dislocation of the hip, whereas 2.3were expected based on the proportion ofmusculoskeletal defects that were of thistype overall. The specific risk for con-genital dislocation of the hip (6/2.3 = 2.61) was thus somewhat higherthan for musculoskeletal defects overall,but because of small numbers it is notstatistically significant (P > .05). Theother 7 musculoskeletal defects werecraniosynostosis, scaphocephaly, cranial

abnormality, hemivertebrae, limb reduc-tion deformity (2), and fibular abnormal-ity. Facial deformities were not ob-served.

Of the odds ratios for defects inbabies born to coffee drinkers (Table3), only that for heart defects amongthe children of women who drank threeor more cups a day (1.52) was signifi-cantly elevated (P = .02). No specifictype of heart defect was overrepresentedin these 58 cases compared with defectsin babies born to women who did notdrink coffee.

92 American Joumal of Public Health January 1992, Vol. 82, No. 1

Page 3: Cigarette, Alcohol, and Coffee Consumption and Congenital Defects

Pubfic Health Biiefs

E. X, X,

..............M M M ............................................................. .........................

:-,...................

...........

Y ..........

I.....................

%......W..........................................

...............................................................................

..........

.................................

................. ..................

.......................................%%--.-----.-,--

...............................................................

............

...

..............

.................

W.. I'M..........

X

.....................

.............................

............................

............

..................

...................... .............................................C............................

X..........................

DiscussionGiven that over 70 odds ratios were

tested, the three that were significantly(P < .05) raised could have occurred bychance. The one association with smok-ing, namely an increase of neural tubedefects in babies born to women whosmoked 1 to 9 cigarettes per day, wouldhave been more convincing if an excesshad been seen in the heavier-smokinggroups. Even quite heavy cigarettesmoking was common in this survey-in30% of the pregnancies the mothersmoked during the first trimester of preg-nancy, and in 15% the mother smoked 20or more cigarettes per day. Thus ourstudy had the power to detect importantrisks due to smoking, if such risks werepresent. The risks we can exclude arespecified by the upper confidence limitsof the odds ratios, which are quite low forsmoking.

Alcohol consumption, on the otherhand, although perhaps underreported,was generally low. Only3% of thewomensurveyed reported taking one or moredrinks per day (seven perweek), and 0.3%reported taking three or more drinks per

day. The power of our study to detect in-creased risk at these levels was thus verylimited, as is evident from the wide con-fidence limits for the odds ratios shown inTable 2. No case ofsuspected fetal alcoholsyndrome was noted, although we werenot specifically looking for the syndrome.The association of alcohol with muscu-loskeletal defects has not been reportedelsewhere, and was of low statistical sig-nificance.

Coffee consumption was associatedonly with heart defects, and the evidencewas not strong. In Finland, where coffeeconsumption is high (29% of those sur-veyed reported taking three or more cupsof coffee a day, compared with 12% inMontreal), no association was found be-tween coffee consumption and clefts ofthe lip and palate or defects of the centralnervous, cardiovascular, or musculoskel-etal systems. LI

References1. Fredrick J, Alberman E, Goldstein H. Pos-

sible teratogenic effect of cigarette smok-ing. Nature. 1971;231:529-530.

2. Comstock GW, Shah FK, Meyer MB, Ab-beyH. Low birth weight and neonatal mor-

tality rate related to maternal smoking andsocioeconomic status.AmJ Obstet Gyne-col. 1971;111:53-59.

3. Kullander S, Kallen, B. A prospectivestudy of smoking and pregnancy. ActaObstet Gynecol Scand. 1971;50:83-94.

4. Andrews J, McGarry JM. A communitystudy of smoking in pregnancy. J ObstetGynecol BC 1972;79:1057-1073.

5. Shiono PH, KlebanoffMA, Berendes HW.Congenital malformations and maternalsmoking during pregnancy. Teratology.1986;34:65-71.

6. Hanson J, Streissguth A, Smith D. The ef-fects of moderate alcohol consumptionduring pregnancy on fetal growth and mor-phogenesis. JPediatr. 1978;92:457-460.

7. Soyka L. Caffeine ingestion during preg-nancy: in utero exposure and possible ef-fects. Sernin Pennatol. 1981;5:305-309.

8. Kurppa K, Holmberg P, Kuosmer E,Saxen L. Coffee consumption during preg-nancy and selected congenital malforma-tions: a nationwide case-control study.AmJ Public Health. 1983;73:1397-1399.

9. McDonald AD, McDonald JC, ArmstrongB, et al. Congenital defects and work inpregnancy. BrJIndMed. 1988;45:581-588.

10. McDonald JC, Lavoie J, Cote R, McDon-ald AD. Chemical exposures at work inearly pregnancy and congenital defect: acase-referent study. Br J Ind Med.1987;44:527-533.

11. McDonald AD, McDonald JC, ArmstrongB, et al. Occupation and pregnancy out-come. BrJ Ind Med 1987;44:521-526.

January 1992, Vol. 82, No. 1 American Journal of Public Health 93