Tobacco Documents Online

C C SMOKING A.D PREGNANCY In recent years, some anti-smoking materials have claimed that smoking during pregnancy leads to adverse effects, particularly in that smokers are more likely to have low-birth-weight (LBW) infants. Some claims have even been made that smoking increases the risk of congenital malformation and perinatal mortality. These claims were reiterated by witnesses who testified at regional "forums" held by an American Cancer Society commission in 1977. The witnesses generally contended that it is unsafe for pregnant women to smoke and that they should be discouraged from smoking by their physicians. However, they did not explain that their claims are based on statistical data which are at best equivocal. Low-Birth-Weight Infants r Smoking opponents have found it easy to attribute the increased probability for LBW infants (2,500 grams or less; about 5.5 pounds) to maternal smoking,. This may be partially explained by the availability of maternal smoking data which are almost routinely collected and the absence of data on other factors which may be associated~with LBW. However, an association between maternal smoking and LBW does not prove that the two are causally related, since statistical associations cannot prove causal relation- ships. 03'74536'7 A biostatistician who examined, and was unable to accept, the causal hypothesis wrote that the data may suggest some other

~: common factor which causes women both to smoke and to have a higher proportion of LBW infants. This theory was advanced by Jacob Yerushalmy in a 1972 report [1] describing data which, he later said, "almost clinch the argument against causation": "This conclusion follows from~the finding that women who eventually became smokers, produced a large proportion of low birth weight infants even before they started to smoke . . ." [2]-Temphais added) ~ To explain these findings, Dr. Yerushalmy speculated that: ". .. the evidence appears to support the hypothesis that the higher incidence of low- birth-weight infants is due to the smoker, not the smoking." [3J A professor of medical physics who examined Dr. Yeru- shalmy's data reached the same conclusion. Professor P. R. J. Burch wrote that: ~ "This collective evidence . . . fails to corroborate the causal hypothesis. Each of its features is remarkably consistent with the view that the smoker, rather than the smoking, is responsible for the high incidence of low birth weight infants." [4] Dr. Yerushalmy's findings are further supported by the recently published results of several research~proj'ects. Two researchers investigating the effects of racial origin on fetal growth in smoking and nonsmoking women in west London~agreed with Dr Yerushalm is "merely an that smokin 0 . y g indicati•on" of other factors which may also be associated with W ~ C11 LBW. [S] They reached this conclusion after determining that, M m in their study group, maternal social class was the major factor -2-

affecting fetal growth. This prompted their suggestion that "there is a case for reopening the file on its [smoking] effects on fetal growth." [6) An Australian researcher who found that intensive anti- smoking advice did not eliminate LBW births, also concluded that his results are "compatible with the hypothesis that smoking in pregnancy does not itself cause low birth weight but that it is an index of some other harmful factor or factors." [7) Relying on his findings, he asserted that "the widely held belief that C smoking is harmful to the fetus .. . is not well based." [8) The need for further research cn the relationship between maternal smoking and LBW was recognized by Silverman who recommended studies be designed to account "for the effects of all variables known to be associated with~smoking and~infant birth weight." [9) She made this recommendation in a report on data collected on over 2,000 births in a Maryland~ county. Her study had been designed to answer: "The critical issue . . . whether smoking causes a reduction in birth weight (the causal hy- pothesis) or whether smokers are a self-selected group that differs from nonsmokers in ways un- related to smoking, including the production of lower-birth-weight babies (the self-selection hypothesis)." [10]' To deal with this issue, she examined birth weight differences in infants born to the same mother. Although Silverman concluded that "at best this study gives a Scottish verdict: not proven," O C.7 ~ she observed that "the direction of the observed differences in ~ U1 _ -3-

C C 2 mean birthweights is more consistent with the self-selection hypothesis."' [11] Although these studies have failed to disprove either the causal or self-selection hypotheses, several have shown that smokers' LBW infants appear to be healthier than nonsmokers'. Dr. Yerushalmy, for example, noted that LBW infants of smokers "are much healthier" than those of the nonsmokers [12] and that the "healthiest" low-weight babies were born to couples in which the wife smoked and the husband did not. [13] The indecisive scientific evidence on the relationship between maternal smoking and LBW indicates smoking has not been proven to affect fetal development. Those who contend otherwise ignore the contradictory evidence which shows that LBW infants of smokers are healthier than LBW infants of nonsmokers. Other Factors Affecting Birth Weight Unlike the ACS forum witnesses who claimed that smoking alone causes LBW, many researchers who have studied fetal development believe it is affected~by multiple maternal characteristics. This belief was expressed by Rantakallio who examined the biological traits and socio-economic circumstances of women in a study of 12,000 deliveries in Northern Finland. She concluded that "the etiology of ... low birth weight is typically multi-factorial." [14] O She also stated~ that: G4~ ~ Cri W ~ O -4-

ti "Birth weight and perinatal mortality are typical variables in human biology, and can only be judged in~the context of the total environment." [15) Rantakallio's conclusion that multiple factors cause LBW is shared by two Swedish researchers who tried to determine whether biological or socio-economic factors were more influential in affecting fetal development. They wrote; "The etiology of low birth weight is multi- facetted [sic) and not to be sought in a single factor but in~the combined effect of various early and existing negative social and biological factors, which vary from one population to another." [16) Despite such conclusions, most studies which are cited to show that smoking causes LBW do not take into account the •numerous environmental, biological and behavioral differences between smokers and nonsmokers which can affect fetal development. For example, women who smoke have been found to change jobs more frequently, drink more coffee and alcoholic beverages and have a higher twinning rate than nonsmokers. [1?) Several researchers have attempted to define the particular maternal characteristics which~may cause LBW. After examining 32 variables in a study on birth weight, including smoking, Hardy concluded that "the amount of weight gained~by the mother duringe the pregnancy and her prepregnant weight showed the strongest correlations with the weight of the infant at birth. ..." [18) `O ~ Nor is smoking always associated with LBW. Two researchers who 4, ~ examined "fetal malnutrition" in black infants with low ratios ~:

of body weight to body length concluded that smoking was "not significantly associated" with this condition~, but that "poor • maternal weight gain, little or no prenatal care, pre-eclampsia and~chronic major illness" were significantly associated. [19]i Other factors which have been found~to be associated~ with LBW include social class [20], maternal weight and height [21], alcohol abuse [22], personal tensions during pregnancies" such as marital discord. [28] The failure of anti-smoking advocates to mention~such altitude [23), hypoglycemia [24]i, diet [25]1, air pollution [26], occupational exposures [27] and "inter- factors in making their claims suggests that they have not considered all the available scientific evidence. Furthermore, additional research must be conducted'on specific maternal characteristics which may affect fetal development. Perinatal Mortality r O W %I ~ C!t with excess fetal or neonatal mortality. .." (34] W ~. -6- Scientific evidence also does not support the claim that maternal smoking during pregnancy is causally associated with increased~perinatal mortality. Several large studies, includ- ing those by Yerushalmy [29], Underwood [30), the Ontario Perinatal Mortality Study Committee [31], Rantakallio [32], and~Targett [33] have found~no increase in the perinatal mortality rate of infants of smoking mothers. As the National Academy of Sciences Committee on Maternal Nutrition concluded in 1970: ". .. smoking is not significantly associated

C C .41 Congenital Malformation The conflicting opinions of witnesses at the ACS forums on whether smoking causes congenital malformation reflect the inconclusive scientific findings in this area. Although several ACS forum witnesses were convinced that smoking causes birth defects, Dr. James E. Allen stated'at the Philadelphia forum that: ". .. I don~'t think anyone has identified absolute evidence that this [congenital malformation] is the result of the chronic or even acute smoking of the mother." [35] Even the 1976 Report on the Health Consequences his statement: worth's findings that "smoking was more prevalent in normal pregnancies establish a relationship between smoking and congenital malformation. In a study of 51,490 pregnancies for example, the Ontario Perinatal Mortality Study Commission found "no evidence that smoking was associated with a higher incidence of congenital malformations." [37] Yerushalmy's continuing study of California women showed that the risk of congenital malformation in LBWinfants was "consider- 03'7453'73 ably lower for smoking than for nonsmoking mothers.".[38] Hollings- of Smoking supports "Given the considerable variation in study design, study population, sample size, number of affected infants, definitions of malfor- mation, dnd results, no conclusions car: be drawn about any relationship between maternal cigarette smoking and congenital malformation at the present time." [36] Several large-scale population studies have failed to

c than abnormal pregnancies" coincided with those of Yerushalmy. [39] A recent clinical study of 1,000 patients by a German researcher also failed to detect an~increase in congenital malformatiorr among babies born to smoking mothers. [40] relationship between smoking and the outcome of pregnancy, physi- cians frequently urge women to stop smoking. However, an editorial in Nature expressed the concern of some members in the scientific community about the possible undesirable effects of such advice: "Mothers-to-be have always been under pressure to avoid excessive weight gains and this pres- sure, it is well known, frequently causes distress. Cigarettes often keep both weight and nerves under control--it is quite possible j that advice to stop smoking may have exactly the wrong effect on the mother's total health." [4L] ship between smoking and congenital malformation has been proven. Despite the lack of scientific evidence proving a causal Such scientific evidence indicates that no causal relation- Conclusion i Maternal smoking during pregnancy has been associated with decreased infant birth weight, increased infant mortality and a higher risk of congenital malformation in some studies. However, it is doubtful that the association between maternal smoking and LBWreflects a causal relationship. Biological, 037453'74 environmental, and other constitutional differences between smokers and nonsmokers indicate that LBW is due to the smoker, not the smoking. In addition, evidence suggesting that smoking causes -8-

~ increased infant mortality and congenital malformation is minimal and inconclusive. Therefore, the available data do not warrant the con- clusion that smoking has been proven to cause low birth weight, increased infant mortality and congenital malformation~. DRAFT Prepared by: Shook, Hardy & Bacon January 19, 1978

C REFERENCES [1] Yerushalmy, J., "Infants with Low Birth Weight Born Before Their P4others Started to Smoke Cigarettes," American Journal of Obstetrics and Gynecolog 1L2(2)~: 277-284 (January 15, 1972). 1C-[2'] Congressional Record - Senate 119(Part 3): 385 (February 7, 1973). [3] Yerushalmy, supra, p. 283. - [4] Burch, P. R. J., "Smoking and Pregnancy," Nature, p. 177 (November 16, 1973). [5] Alvear, J. and~0. G. Brooke, "Effect of Smoking on Fetal Growth," The Lancet, p. 1158 (May 28, 1977). [6] Ibid. [7] Donovan~, J. W. ,"Randorlized~ Controlled Trial of Anti-Smokinq Advice in Pregnancy," British Journal of Preventive and Social Medicine 31: 10 (1977). [8] Ibid., p. 11. [9]i Silverman, D. T., "Maternal Smoking and Birth Weight," American Journal of Enidemiology 105(6): 520 (1977). [10]' Ibid., p. 513. [11] Ibid., p. 520. [12] Yerushalmy, J., "The Relationship of Parents' Cigarette Smoking to Outcome of Pregnancy--Implications as to the Problem of Inferring Causation from Observed Associations," American Journal of Epidemiology 93(6): 454 (1971). [13] Ibid., p. 443. [14] Rantakallio, P., "Groups at Risk in Low Birth Weight Infants ment 193: 9 (1969). (15] Ibid., p. 6. 03745376 [16] Bjerre, I. and Gunilla Varendh, "A Study of Some Biological and Socio-Economic Factors in Low Birth Weight," Acta Paediatrica Scandinavica 64: 611 (1975). and Perinatal Mortality," Acta Paediatrica Scandinavica, Supple- -1-