TABLE AS1L—Summary of methods used in retrospeetvie stiuties of tubacco use and cancer of the esophagua Toy Author, year, Chasen Contruls country, wm ee ~ reference Sex Number Method of aclection Number Method of selection Collection of data Sadowsky ct al, M. 104 White patients admitted during 1938-43 to G16 White patients with Hincaaes other than Obtained by ¢ apeclatly 1953, selected hospituls in New York City, cancer admitted to same group of hos. tralned Jay interviewers. ULS.A. (232), Missouri, New Orleans, and Chicago, pitals during same period, 242 records out of a total of 2.847 excluded bes cause of incomplte or Questionable smoking histortes. Sanghvietal., M. 73 Consecutive clinic admissions to Tata me- 288 Consecutive clinic admissions of paticnta By means of “detailed 1955, morial Hospital, Bombay, Without cancer, queationary.” No other India (262), 107 Consecutive admissions of patients with details given, cancery other than intraoral or eso- phagus. Wynder et al., M. 39 Patients admitted to Radiumhemmet, 115 Patients admitted to same hospital with 1957, F, 35 Stockholm, during 1952-55. 166 cancer of skin, head and neck region Sweden (s22). Other than squamoua cell cancer, leu- kemia, colon, and other aitea. No matching. Staszewskl, M, 24 Patients admitted to Oncological Institute 912 Other patients sent to Institute with No detalls given on 1960, during 1957-59, symptoms probably not etlologically method of data callee- Poland (260), connected either with smoking or with diseases of esophagus, atomach or du- odenum, tion, Nu age adjust- Ment or matching. Av. erage age of canccr Patients, 60.5; controls, 53. cO¥ TABLE A31.-—Summary of methods used in retrospective studies of tobacco use and cancer of the esophagus (cont.) wees Cases Controls country, teference Sex Number Method of selection Number Method of selection Collection of data Schwarts et al, M, 362 Admissions to hospitals in Paris and a 362 Healthy Individuals admitted to same hos- Interviewed by team of 1961, few large provincia] tities sin¢e 1964, pitul because of work or traMe aeci-+ specially trained inter France (248), dents—matched by 6 year age group viewers who interviewsasl and time of admisyion. the loxgryt proportion possible of al cancer patients, Cascs and matched controls inter. viewed by aame peraon. Wynder and M, 160 Cancer patients seen in Memorial Hospi- 160 Paticnts seen in same hospitals during Data collected by trained Bross, tal, New York City, and Kingsbridge same time period with other tumors, interviewers. 1961, and Brooklyn VA Hospitals during 64%-malignant tumor; 36S¢-ebenign con- ULS.A. ($10). 1950-69 (86% white). ditions. Matched by age with canter patients. F, 37 Same hospitals: and-same time period as 37 Same as with regard to male controls, male patients (86% white). 43% had malignant and &7% benign - tumors, Wynder and M. 67 Admitted to Tata Memorial Hospital Bom- 134 Patients with other forms of cancer ex- Interviewed by one per- Drona, F. at bay, cept for oral cavity and Jungs: ag well gon, 1961, ay various benign Ulscaace, 10% uf male and 40%. of India ($70). fomule cancer cancy hintolowleally conflemed. Takano ctal, M. 167 Patients with esophageal cancer, 167 Patients with canecrous and non-can- Interviews at various LDCR, F, ag 33 cerous divcases of non-digeative organs, hospitals, Caxew and Japan (#78), controle ageematohed, €0% TABLE A31.—Summary of methods used in retrospective s tudies of tobaceo use and cancer of the esophagus (cont.) Author, year, Cases Controls country, reference Sex Number Method of selection Number Method of selection Collection of data Bradshaw and M. 98 Patients with esophageal cancer, 341 Patients with non-malignant disease, Hospital interviews by Schonland, trained African social 1969, workers. South Africa (41). Martinez, M. 120 Patients with confirmed epidermoid eso- 360 120 male, 59 female patients in same hoa- Interviews by trained 1969, F, 59 phageal cancer diagnosed in 1966, WT pital with non-cancerous diugnoses. peraunnel, Puerto Rico 240 male, 118 female members from same (183). community. 707 TABLE ASla—Summary of results of retrospective studies of tobacco use and cancer of the esophagus Rolutlve rink ratio, Author, Percent inhalers All smokers to yenr, Pereent nonamokere Percent heavy smokers among smokers nonamokerg country, Cases Controls Al Heavy reference Cases Controls Cases Controls smokers amokers Sadowaky et al., 1983, 3.8 13.2 —_ _ _ _ 4.0 — U.S.A. (252). Sangvhi et al, 5.5 17.3 Average number of _ _ 1985, bidia smoked India (241), 15.3 14,1 Wynder et al., M 13.0 24.0 1987, F (about) 85.0 (about)s2.9 - ~ - _ _ Sweden (322), _ - ~ _ 2 = Staszewaki, - 18.0 95.8 59.0 87.6 80.0 > _ 1960, Poland (260), —_———_. Schwartz et al,, 3.0 17.0 Totalaumount amoked 39.0 38.0 6.6 _ 1961, daily (cigarettes) France (249), 16, 16.0 Wynder and Bross, American males 6.0 16.0 48.0 33.0 ~ _ 3.4 44 1961, U.S.A. and American females 41.0 78.0 27.0 16.0 _ ~ a 3.2 India (370), Indian males 13.0 28.0 - - _ _ 2.6 - Indian females 78.0 94.0 - - _ _ 4.5 _ Takano etal., 17.0 23.0 _ _ - _ 1.3 - 1963, Japan (272), Bradshaw and Schonland, 18.3 31.7 31.6 69 ~ — 2.6 Mal 1969, South Africa (41). Martinez, 1969, 14.0 23.6 79 4.6 _ _ eB 3.6 Puerto Rico (183). SO” TABLE A32,—Atypical nuclei in basal cells of epithelium of esophagus of males, by smoking habits and age TSections with some epithelium present. Souree: Auerbach, O. et al (75), Never smoked Current regularly Cigarettes Ex-cigarettes Pipe, cigar Other Atypical nuclei Num- Per. Num- Per- Num- Pers Num- Per. Num- ber cent ber cent ber cent ber cent ber A. All men: Number men pecans 9t _ 779 _- 18 _— 89 od 62 Total sections! te 187 100.0 6,762 100.0 1,586 100.0 766 100.0 622 No atypical nuclei 733 93.1 167 2.5 770 48.5 ag 6.9 198 Sume but <60 percent atypical 52 6.6 5,389 79.8 165 44.3 Obs OY ai? G0 percent aur morentypical ......0, 2 0.3 1,196 int bl 3.2 25 33 10 B. Men under age 50: Number men 26 ~~ 236 —_ on — 9 —~ 1 Total sections 223 100.0 2,039 100.0 254 100.0 17 t0u.0 83 No atypical nuclei soeeeneeer 190 85.2 TM b4 56 21.7 1 1.3 4 Some but <60 percent atypical ...., 33 14.8 1,853 9U.0 105 75.6 74 96.1 46 60 percent or more atypical ~ — 135 6.6 7 2.7 2 2.6 3 C. Men aged 50-69: Number men 44 — 445 _ 109 — 3R — 34 Total sections oc... viee cece aes 379 100.0 3,553 100.0 953 100.0 310 100.0 265 No atypical nuclei eee ee 37 98.4 $3 22 461 48.4 37 41.9 74 Some but <60 percent atypical ....,, 4 ht 2,915 75.6 462 dtd 261 84.2 178 60 percent or more atypical 2 0.5 855 22.2 40 4,2 12 3.9 4 D. Men aged 70 or older: Number men .. 21 —_ UN ~- ad — 42 ~~ a4 Tota! sections Pees eee eee eee ee 185 100.0 K4G 100.0 S75 100.0 379 100.0 213 No atypient nuctei bette eee 170 91.9 V3 1S 268 oT4 1S 4.0 M7 Some but <60 percent atypical » 15 8.1 621 74.0 IR 415 353 oa 95 G0 percent or more atypical - _ 206 24.5 4 Mt iM 249 3 100.0 bo.8 6&7 100.0 28.9 69.5 1.6 100.0 Gay 14 907 TABLE A33.—Atypical nuclei in basal cells of eptthelion of esophagus of males, by amaunt uf sntoking and age Current cigarette smok ers Never smoked regularly <1 pack 1-2 packs >2 packs Cells with atypical nuclei Number Percent Number Percent Number Pereent Nionber A. Allages o........0.. 0, Seay 91 ee 179 _ 413 _ 187 _ Tutatsections! oo... eect e eee 187 100.0 1,544 100.0 3,629 100.0 1,679 100.0 No atypical nuclei oo... 0... cece, 183 93.1 89 5.8 39 hl 39 25 Some but <60 percent atypical ...., 62 6.6 1,341 86.8 2,957 81.5 1,091 69.1 CO percent or more atypical ........ 2 0.3 114 7.4 633 17,4 449 2K.4 B. Men under nage 50: Number men ce cccccceec cesses cease 26 vee 9 _ M2 _ 55 _ Total sections t Severe en eeneees 223 100.0 433 100.0 3,1u9 100.0 457 100.0 Nontypical nuclei ssc ccece cee eee ee 190 85.2 48 Mal 2) Lx 2 0.4 Srme but 60 percent atypienl oo... 33 14.8 382 RK.2 1,049 93.2 Sho a6 fl percent ay more atypical ooo... sae an 3 0.7 69 5,0 73 16.U C. Men aged 60-69; Number men See tee eee eae 44 ee 92 ~ 240 _ Wi - Trtal sections ! Pete eee eee 379 100.0 789 100.9 2.416 100.0 948 luu.d Neatypleal nuchi ...c css ccc sees 373 98.4 30 3.8 1K 0. 45 47 Some but £60 pereont atypleal oo... 4 Ml 604 87.9 1,607 15.9 ol GSK HU percent or nore atypient oy... .. , 2 0.6 65 8.3 491 23.2 209 Sho D. Mou aged Fo on older: Number men ooo. cee ee reece 2h ee aS -- ai _ 19 : Total sections) cece ee cevaes 188 100.0 322 100.0 gas 100.0 17s 100.0 Noatypical nuclei oo... ceca 170 91.9 M1 34 _ _ 2 Lt Some but £60 pereent atypical 2... 15 Bl 265 82.3 261 76.9 uh O47 GU peveent or more atypical ......., tee sae 46 14.3 xs a4. WG dau ‘Sections with some epithelium present, Souree: Auerbach, O. et al, (15). L09 TABLE A86.—Summary of methods used in retrospective studios of smoking aad enneer af the bladder Author, year, Cases Contras country, ee ee ee . wee ee wee ee wee ee — ro reference Sex Number Method of selection whey Method of selection Lilienfeld ct ab, M. 821 Admissions to Roswell Park Memorial Institute, RAT Nou disease patients, 1956, 3045-55 over 45 years of age, U.S.A. (171), rr. lt Same as mules lou Renin bladder conditions. RYT No disease pationts, Schwartz et al., M. 218 Admissions to hospitals in Paris and a few 24 Healthy individuals admitted to same hospitntt 1961, large provincial cities since 1954, because of work or trafic accident, matched France (249), hy & year age geraup. Lockwood, M. 282 All bladder tumors reported to Danish Cancer Qa2 A. From clection rolls matched with cases e+ 1961, KF, Register during 142-50 and living at time ST cording ta sex, age, morital status, ocrupa- Denmark (175), of interview in Copenharen and Fredericks. tion, and residence. burg. (Includes bludder papillomas). BR. Ancther contrul group obtained from game ple of Danish Marhidity Survey (1402, 1953, and VI54) compared with respect te smoke ing histories. Wynder, M. 200 First phase: luo 1963, By 50 Admission te several hospitals in’ New ny Admission ta same hospitals Cexetuded cancer U.S.A. (326), York City during January 1957-Devem- uf respiratory system, upper alimentary tract, ber 1160, myocnvdiol infwretion) matehed by sex and Second phase; are. M. 100 Admission tu same hospitals during 1007, oa Some ous above, K 20 Zo Cobb and Angell, M. 136 Patients udmitted to VA Hospital in Seattle 44a 120 patients with cancer of sigmoid colon, 122 1965, U.S.A. (87), 1951-61, Patients with nun-enceplastic pulmonary dis- vase, 807 TABLE A86.—Summary of methods used in retrospective studies of smoking and cancer of the bladder (cont.) Author, yenur, Crues Controla cauntry, 7 reference Sex Number Method of selection Number Method of actection Stoszewski, M, 150 Patients with histologically confirmed bladder Th0 Undetined source age-matched. 1968, carcinoma, Poland (26!), Deeley and Cuhen, M. 127 Patienta with histologically confirmed bladder 127 Patients in same hospital with nun-cancerous 1966, earcinoma, or pulmunary disease matched for age. Engtand (6¢), Yoshida ct al, M. 163 Patienty with bladder cancer, 163 "Comparison cases.” 1968, F. 29 69 - Japan ($90). Kida ct al, M. 8B Admissions to 18 hospitals in North Fukuoka Ry Sclected from patients hospltalized in aame re- 1968, F, 26 prefecture, 26 yegion for non-urinary allments and age Japan (144). matched Dunham et al., M. 334 Admissions to New Orleans hospitals with his- 350 Admissions to same hospitals with non-ncoplas- 1968, KF, 189 tologic diagnosis of bladder carcinoma, \7T tic diseases and discases unrelated to genl- U.S.A. (85). tourinary tract. Anthony and Thomas, M. Sst Patients with papilloma and cancer of bladder 275 1970, England (9). at Leeds betweeen 1968-67. Surgical patients without cancer previously In« terviewed for lung cancer study, 604 TABLE ASba.——Summary of results of retrospective studies of smoking and cancer of the bladder Percent elgoretios Relative riek ratio: Author, Percent nonamokers Percent heavy smokers amuked All amokers to nonsmokers year, country, All Heavy Cigarette Comments reference Sex Casea Controls Cases Controls Canes Controls smokers amokers amokers Lilienfeld ct al., M. 16.0 29.0 61.0 44.0 2.8 see 247 Cigarette and other. 1966, F, 87.0 83.0 14 ene vee U.S.A, (971), Schwartz et al, M. 11.0 20.0 : 83.0 70.0 2 , 2.2 Cigarette only. 1961, France (£49). Lockwood, M. 0.0 13.4 30.0 16.0 30.0 18.0 1.6 3.0 3.0 Cigarettes main mode of 1961, F. 56.0 66.0 4.0 4.0 sae see 1.5 1.2 sae smoking. Denmark (178). Wynder ct al,, M. 7.0 18.0 47.0 23.0 85.0 63.0 2.9 5.2 3.3 Phases A and B com- 1963, F. 61.0 86.0 6.0 tee vee vee 8.9 tee aoe bined, ULS.A. (526), Cobb and Ansell, M, 4.6 25.8 79.4 43.3 vee see 1.3 10.3 1965, ULS.A, (67), Staszewski, M. 6.7 16.0 86.7 65.7 87.1 72.2 2.7 8.1 2.9 Cigarettes only. 1966, Polnnd (£61). Deeley and Cohen, M. 24 UA tee see es 3.1 tee see 1956, England (66). OV TABLE A35a,—Summary of results of retrospective studics of smoking and cancer of the bladder (cont.) Percent cigarettes Relative risk rattle: Author, Percent nonsmokers Percent heavy smokers smoked All emokers to nonsmokers year, country, All Heavy Cixarctte Comments reference Sex Cases Contrlos Cases Controls Casue Controls smokcra smokers amokers Yoshida ctal., M. 8.0 22.7 43.4 35,0 — _ 34 3.7 - 1968, F, 62.1 86.4 — _ — _ —_ ~ ~ Janon (330), Kida et al., M. 11.0 11.0 32.0 29.0 _ _ 1.0 _ _ 1068, KF, 16.0 21.0 _ o - _ 4 ~ ~ Japan (144), Dunham etal, M. 8.6 14.5 — ~ 49.4 48.4 1.8 _- 1.8 Clgarettcs only. 1968, F. 62.2 61.5 _ - 32.0 28.2 1.0 - Ma U.S.A. (85), Anthony and F, 6.3 6.3 _ _ 36.5 29.1 1.0 _ 1.3 Cigarettes only. Thomas, More than 15 a diy. 1970, England (8). Chapter 5 Pregnancy Source: 1973 Report, Chapter 4, pages 97 - 149. 411 Contents Introduction 2... 0.0.0.0. 0 0.0002 eee eee Smoking and Birth Weight Epidemiological Studies Cigarette Smoking and the Low-Birth-Weight Infant ..... Evidence for a Causal Association Between Cigarette Smoking and Small-for-Dates Infants ...........-. Evidence for an {Indirect Association Between Ciga- rette Smoking and Smali-for-Dates Infants ......... Experimental Studies Studies in Animals Tobacco Smoke .. 1.2.22... eee ee eee Nicotine .. 2... 2 ee ee ee Carbon Monoxide... 2.2.0.0... 002020220 eee eee Polycyclic Hydrocarbons .. 2... 2.2.26. -+0 ee eee Studies in Humans Carbon Monoxide .. 2.2.02. ee ee ee ee ees Polycyclic Hydrocarbons .......-..-.0202-0-00-- Vitamin By, and Cyanide Detoxification ...........- Vitamin 2. eee Possible Mechanisms 2.00 0 ee Timing of the Influence of Cigarette Smoking on Birth Weight 0.202... ee ee ees Site of Action at the Tissue and Cellular Level .......... Significance of the Association .. 2.2... eee ee eee ee Birth Weight Summary 2... ee Cigarette Smoking and Fetal and Infant Mortality Introduction . 2000 ce eee Spontaneous Abortion 00... ee eee Spontaneous Abortion Summary .-.-..-------+-+--- Srillbirth 000 0 te eee Stillbirth Summary .......---.2- 020+ er eee ee cee Late Fetaland Neonatal Deaths .......-..00002222 02% Epidemiological Studies... 2-2-2. - 22-0 - ee ee ee eee ee Comparisons of the Mortality Risks of Low- Birth-Weight Infants Born to Smokers and Nonsmokers .....0.------- 22 eee eee eee Recent Studies ........---.-- De eee eee Analysis of Previously Reported Studies .....-.----- Factors Which Influence Perinatal Mortality Other Than Smoking ...--.--...-.- 000 eee eee 413 Page 417 419 420 424 Experimental Studies Studiesin Animals .....----- 02-20 fete 446 Studiesin Humans ....-.------- 0800 e rrr 447 Significance of the Association... 2. +222 ser rrr ttt 447 Cigratte Smoking and Infant Mortality - Cont. Late Fetal and Neonatal Death Summary ..------ +--+ e000 448 Sex Ratio ..-- eee eet teeter sss 449 Summary... 000000 cece terres tts 449 Congenital Malformations ©... - +++ +2 ser ctr errr 450 Congenital Malformation SUMMANY oo ee 451 Lactation Introduction... 00-0 eters 452 Epidemiological Studies... 62-0000 c ttre srt 452 Experimental Studies... 0000 cer r errr rss ett 452 Studies in Animals Nicotine ...- 2-22 - cee te eet 452 Influence on the Lactation Process ....----- 2 eect 452 Presence of Nicotine in the Milk ..-.---- +e eect 2. 453 Evidence for an Effect Upon the Nursing Off- spring... eee ttt 453 Nitrosamines ....---- 20-2 ¢ cece eter tr 453 Studies in Humans Nicotine and/or Tobacco Smoke ..---- +--+ e src 453 Influence on the Lactation Process -.-- +--+ 7s 5 rr cts 453 Presence of Nicotine in the Milk 2.2.6.2 26-25 s ee 454 Evidence for a Clinical Effect Upon the Off- Od 454 Vitamin ....-- eect cette 455 Lactation Summary ...- 0-000 ccc 455 Preeclampsia... 2-2-0 456 Summary 200-00 e cette tts 456 References ...--ece creer trees ttt rss 456 414 List of Figures Figure 1.—-Mean birth weight for week of cestation according to maternal smoking habit: control week singletons..___.__- Figure 2.—Percentage distribution by birth weight of infants of mothers who did not smoke during pregnancy and of those who smoked 1 pack of cigarettes or more per day_...-____. Figure 3.—Percentage of pregnancies with infant weighing less than 2,500 grams, by cigarette smoking category____.__22_- Figure 4.—Average birth weight by maternal smoking habit (a) before current pregnancy and (b) during current preg- Figure 5.—Percent of low birth weight white infants by smok- ing status of their mothers._....-2--- Figure 6.—Neonatal mortality rates among single white births in hospitals (by detailed birth weight and specified gestation groups: United States)... Figure 7.—Perinatal mortality rate per 1,000 total births by cigarette smoking category_.._.-----6 ee List of Tables Table 1.—Infant birth weight by maternal and paternal smok- ing habits. _____. eee Table 2.—Effect of carbon monoxide exposure of pregnant rabbits on birth weight-_-00--28 Table 3.—Comparison of the perinatal mortality for infants weighing less than 2,500 grams, of smokers and nonsmokers. Table 4.—Effect of carbon monoxide exposure of pregnant rab- bits on birth weight and neonatal mortality... 22 8. Table 5.—Proportion of male infants delivered to smoking and nonsmoking mothers.-_..-.----2 ee Table 6.— Relative risk of congenital malformation for infants of cigarette smokers and nonsmokers, comparing available studies with regard to study design, study population, sample size, number of infants with malformations, and definition of malformation... 22222! Pace 418 419 422 424 427 442 443 425 431 441 447 450 415 Introduction Cigarette smoking is a common habit among women of child-bearing age in the United States. In 1970, approximately one-third of Amer- ican women of child-bearing age were cigarette smokers. The percent- age of U.S. women who smoked throughout pregnancy 1s not definitely known, but is presumably lower. probably in the neighborhood of 20 to 25 percent. With a large fetal population at potential, but prevent- able, risk, the relationship between cigarette smoking and the out- come of preenancy has been the focus of considerable and continuing research. Every investigator who has examined the relationship has confirmed that the infants of women who smoke during pregnancy have a lower average birth weight than the infants of women who do not smoke during pregnancy. Much evidence indicates that cigarette smoking during pregnancy causes this reduction in infant birth weight. Several investigators have demonstrated that the fetal and neonatal mortality rate is sienifcantly higher for the infants of smokers than for the infants of nonsmokers: other investigators have not found higher mor- tality for smokers’ in fants. Studies of the association between maternal cigarette smoking and congenital malformations have produced con- flicting results. The following is a review of work previously reported and recent studies which bear on the relationships bet ween cigarette smoking and different outcomes of pregnancy. In addition, the chapter includes a review of the relationship between cigarette smoking and lactation. Smoking and Birth Weight E' pidemiolo gical St urdiEs” CIGARETTE SMOKING AND THE Low-Birre-Wericur INFant In 1957. Simpson (90), using 2 retrospective study design. deter- mined that among 7.499 women in San Bernardino Connty. Calif. the delivery of infants weishing less than 2,500 grams was nearly twice as 417 frequent among cigarette smokers as among nonsmokers. Subsequently, Lowe (46) studied 2.042 women in Birmingham, England, and dem- onstrated in his retrospective study that the infants of smoking mothers were delivered only slightly earlier (14 days on the average) than those of nonsmokers. He further noted that for gestations of 260 days and over, the infants of smokers were consistently lighter in weight during each week of gestation than those of the nonsmokers. This finding has heen confirmed since, and figure 1 from the British Perinatal Mortality Study (73) provides illustration of this relationship. Given the nearly constant disparity present between the birth weights of the infants of smokers and nonsmokers for gestations of 260 days and over, but absent prior to that time, and given the similar birth weights of infants of nonsmokers and of women who gave up smoking early in pregnancy and did not begin to smoke again, Lowe inferred that the influence of smoking upon birth weight might lie mainly in the later months of pregnancy. He emphasized the tentative nature of this conclusion, since the number of infants with a gestation of less than 260 days and the number of women who gave up smoking early in the pregnancy and did not begin to smoke again were both smal]. Figure 1.—Mean birth weight for week of gestation according to maternal smok- ing habit: control week singletons.’ 3650 - ® 125 _— __ a 0-7-7 OF 3400 g x “or E & 115 7 Non- on or 3150 £ o smokers -7 4 to c oe & = 105 f 7 2 & J. Smokers + 2900 & = & : id = L Cc 5 95 LA 4 2650 § = yo = 85 67 4 2400 75 4 4 ‘ 4 1 ro 2150 36 37 38 39 40 41 42 43+ Gestation in completed weeks 7 This term refers to singleton births In England, Scotland, and Wales occurring during the week of March 3-9, 1958, which are included in the Perinatal! Mortality Survey. These comprise 97 percent of ali births notified in Engtand and Wales or registered in- Scotland during this weak. SOURCE: Butler, N. R., Alberman, E. D. (13). 418 Lowe found that the infants whose mothers smoked throughout pregnancy weighed, on the average, 170 grams less than those whose mothers did not smoke. In addition, he noted that the entire distribu- tion of weights of infants of smokers was shifted to the left (toward lower weights) relative to that for the infants of nonsmokers. This finding, too, has been confirmed by other investigators. Figure 2 offers an iNustration from MacMahon, et al. (49). Given that the infants of smokers and nonsmokers differed only slightly with respect to the duration of gestation, Lowe concluded that the lower birth weight of smokers’ infants must be attributed to a direct retardation of fetal growth. In other words, on the basis of his data, the infants of smokers were emall-for-dates rather than truly premature. Many investigators have subsequently confirmed this point (72, 14, 20, 35, 65, 78, 85, 113). Buncher (72), ina study of 49.897 births among U.S. naval wives, in the same population studied by Underwood, et al. (700), found that the infants of smokers were, on the average, de- livered only 1 day earlier than those of nonsmokers. This finding accounted for only 10 percent of the discrepancy in birth weight be- tween the two groups of infants. The remainder of the studies resulted im the detection of either similar variations in gestational length or no average difference. In a recent study, Muleahy and Murphy (56), Figure 2,—Percentage distribution by birth weight of infants of mothers who did not smoke during pregnancy and of those who smoked 1 pack of cigarettes or more per day. INFANT WEIGHT ANO PARENTAL SMOKING HABITS TT ey 10 4 —— Nonsmokers Smokers BEL 4 o Cc BSF 7 4-L “4 2 4 0 ed I ray 4 5 6 7 8 9 10 11 BIRTH WEIGHT (SCALE IN POUNDS; INTERVALS OF 4 OZ.) SOURCE: MacMahon, et al. (49). 419 in « sample of 5,099 Irish mothers, concluded that although the babies born to cigarette smokers were delivered slightly earher than those of nonsmokers, independent of age and parity, the direct effect of smoking in retarding fetal growth was more significant. The following points, based upon the results from many different studies, can be made about the relationship between cigarette smoking during pregnancy and lower infant birth weight: 1. Women who smoke cigarettes during pregnancy have a higher proportion of low-birth-weight infants than do nonsmokers. This excess of low-birth-weight infants among cigarette smokers pre- dominantly consists of infants who are sinall-for-gestational age rather than gestationally premature. 2. The entire distribution of birth weights of the infants of ciga- rette smokers is shifted toward lower weights compared to the birth weights of the infants of nonsmokers. 3. The birth weights of the infants of cigarette smokers are con- sistently lighter than those of the infants of nonsmokers when the birth weights of the two sets of infants are compared within groups of similar gestational age beyond the 36th week of gestation. The results of the studies which have been considered so far identify a relationship between cigarette smoking and lower infant birth weight and illustrate some aspects of that relationship, but do not indicate whether the association is causal or indirect. The succeeding two sections of this chapter contain evaluations of the available evi- dence which bears upon the nature of the association between cigar- ette smoking during pregnancy and the incidence of small-for-dates in fants. Eviwence ror a Cacsa. Association BeTwren CicareTre SMOKING AND Sm 4uu-ror-Dates Ixrants Evidence previously reviewed in the 1971 and 1972 reports on the health consequences of smoking (/01, 102). suggests that cigarette smoking is causally associated with the delivery of small-for-dates infants. The following is a summary of this evidence: 1. The results from all 30 studies in which the relationship between smoking and birth weight was examined have demonstrated a strong association between maternal cigarette smoking and delivery of low- birth-weight infants. On the average. the smoker has nearly twice the risk of delivering a low-birth-weight infant as that of a nonsmoker 420 (3, 13, 17, 20, 25, 29, 35, 42, 43, 46, 47, 49, 47, 58, 59, 65, 70, 72, 73, 77,78, 80, 83, 85, 90, 95, 99, 100, 113, 118). 2. The strong association between cigarette smoking and the de- livery of small-for-dates infants first demonstrated with results from studies of retrospective design (3, 13, 17, 35, 46, 47, 49, 57, 58, 59, 65, 70,72, 73, 77, 80, 85, 90, 95, 99, 100, 118) has been repeatedly confirmed subsequently by data from studies of prospective design (20, 25, 29, 42, 43, 78,83, 113). 3. A strong dose-response relationship has been established between cigarette smoking and the incidence of low-birth-weight infants (25, 43, 46,49, 100, 113). 4. When a variety of known or suspected factors which also exert an influence upon birth weight have been controlled for, cigarette smok- ing has always been shown to be independently related to low birth weight (1, 13,25, 43, 46, 73,78, 83). 5. The association has been demonstrated tn many different coun- tries, among different races and cultures, and in different geographical settings (13, 17, 25, 29, 36. 42. 43,59, 78, 78, 80, 113). 6. Previous smoking does not appear fo influence birth weight if the mother gives up the habit prior to the start of her pregnancy (24, 46.49.13). 7. The infants of smokers experience an necelerated growth rate during the first 6 months after delivery, compared to infants of nonsmokers. This finding is compatible with viewing birth as the re- moval of the smoker's infant from a toxic intlience (83). 8. Data from experiments in animals have documented that ¢x- posure to tobacco smoke or soine of its ingredients results in the delivery of low-birth-weight off NX 3.0 fF 2.0 F XXRXR\RARARRYYAL MOY 0.0 21 pack per day per day ER&&aaw Nonsmoker Number of infants weighing <2,500 grams: 1,3 1,186 793 8 Total births: 28,358 15,328 6,581 (P <0.001) SOURCE: Ontario Department of Health (66). 422 Rantakallio (76) carried out a prospective study of 11,905 single births in Finland. Cigarette smoking mothers had sinificantly more infants weighing less than 2,500 grams than did nonsmokers (P30_.----------------- 1,570 3, 182 213 1,330 3, 393 3 ene t Nonsmoker niitus smoker. Source: Undenvood, ct al. (199). Yerushalmy (727) pointed out that other investigators had found” marked differences between smokers and nonsmokers. In his own study. he found Chat nonsmokers used contraceptives significantly more frequently than did smokers. Moreover, a significantly higher proportion of smokers drank cotfec, beer, and whiskey. However, he did not adjust for these variables in his analysis of the association between cigarette smoking and lower infant birth weight. Other im- vestigaters have also found differences between smokers and non- smokers. Forexample, Frazier. et al. (29) found significant differences in the distribution of parity, work history, education, and‘ psycho- 425