Tobacco Documents Online

UfNNEFBSITY~OF~~CAUFtORMIA ~ Los Angeles The Relationship of Passive Smoking to Various. Health Outcomes among Seventh-day Adventists in California I IA dissertation submitted in partial satisfaction of'the requirements for the degree Doctor of, Public Health by Terrence Leslie Butler 1988 1g,

CHAPTER 6: SMOICI11Cti RELATED CANCER~ RESULTS In this chapter we investigate the relationship between passive smoking exposure and the outcome of smoking related cancers for both our populations. Stratified analyses using both the Mantel-Haenszel method and the conditional maximum 6ke6hood method are ernployed'to calculate the adjusted measures of effeet: These cancers are : combined as a group since cigarette smoking is considered one of the major risk factors for each of, these outcomes. Lung cancer cases are induded in some analyses but are excluded in others to detemnine what differences, there are in the overa0 measures of effeci. In the female spouse pairs separate analyses were conducted on the cervical cancer outcome. Although the major group of' interest was the non-smokers, alli females in the spouse pairs„ both smokers and non-smokers, were treated as a group for analysis. 6.1 SPOUSE'PAIRS COHORT Among i the female spouse pairs cohort; 50 incident cases of smoking i related cancers were diagnosed in the time period 1977-1982. Table 6.1 shows the distribution of these cases.by'site: 114 I

TABLE 6.1 SPOWSE' PAIRS - FEMALE SITE SPEC1FiC DISTAtIBUl1OM OF SMOf(IN"a' RELATED CAQNCERS! Site Cases Up, oral, pharynx (ICD 140-149) , 3' Pancreas (ICD 157) 9 1 Trachea, bronchus, Iung i pCD 162) 9 1 Cervix, uteri (lCD 180) : 17' Bladden (lCD 188) : 7' fGdney (ICD 189) 5 1 Table 6.2 presents the distribution of incidenC cases,, person years, and incidbnce rates by age group according to subjea and spouse smo status. Overall there appears to be an increased risk for females who are married to smokers compared to females married to non-smokers. There is an additional incidence rate elevation for females who have smoked' versus those who have not smoked: However, the trend is notconsistenf over all strata. This may be due largely to the small number of outcome events. Initialfy Mantel-Haenszef' stratified analyses were performed for all females, smokers and non smokers combined but excluding the nine cases 115'

TABLE 6.2 SPOUSE PAIRS - FEMALES ALL SMOKING RELATED CANCERS (1977-1982) DISTRIBUTION OF INCIDENT CASES. PERSON YEARS AND INCIDENCE RATES BY AGE GROUP AND HUSBANDS SMOKING STATUS IN MARRIAGE WIFE NON-SMOKER WIFE SMOKER HUSBAND SMOKER HUSBAND NON-SMOKER HUSBAND SMOKER HUSBAND NON-SMOKER Casesf Rate/ E',ases/ Rate/ Casosl Rate/ Cases/ Rate/ Age Group Porsoo years 10,000 PY Porson years 10.000 PY Person years 10,000 PY Person years 10,000 PY 25-34 1 942 10.62 1 8004 1.25 1 718 13.92 0 836 0.00 35-44 0 1552 0.00 2 8637 2.32 2 999 20.02 1 816 12.25 45-54 1 2190 4.57 4' 9520 4.20 1.1011 9.89 1 842 11.83 55-64 6 3000 16.67 10 8578 11.66 1,1240 8.06 0 676 0.00 65-74 4 2159 18.53 7 6980 11.71 1: 629 15.90 0 305 0.00 75-84 0 682 0.00 6 2112 28.41 0 70 0.00 0 39 0.00 65•94 0 53 0.00 0 220 0.00 0' 0 0.00 0 6 0.00 Total 11 10579 10.40 30• 43052 6.97 64668 12.85 23519 5.68 9r44C.4JC.0z

TABLE 6.3 SPOUSE PAIRS - FEMALES AGE ADJUSTED RATE RATIOS (RR) OF SMOKING RELATED CANCER (EXCLUDING LUNG CANCER) FOR SELECTED EXPOSURE FACTORS (1977-1982) Factor Levels • Cases/ Person Years + qa_ te/ 10s PY Crude RR Adjusted RR" (95% C.L) 1 fusband Smoking Never 27•47235 572 Ref. Ref. Status in Past 1111610 947 1.66 1.38 (0.68-2.82) Marriage Current 2 3930 609 0.89 1.02 (0.24-4.30) Husband Smoked No 27,47235 572 Ref. Ref. in Marriage Yes 13'15540 837 1.46 1.31 (0.67-2.67) Female Smokod Non-smoker 34'64839 620 Re/. Ref. Smoker 7. 8432 830 1.34 1.66 (0.69-3.54) Education High School 26'20994 1238 Ref. Ref. College • 15~43232 347 0.28 0.34 (0.18-0.65) Age Married 20 + years 1842043 428 Ref. Rel. < 20 years 23•'22378 1028 2.40 2.70 (1.45-8.02) Spouso White collar 1294645 346 Raf. Ref. Occupation Blue collar 2426077 920 2.66 2.34 (1.17-4.70) Bota Carotene High 14+22380 626 Ref. Ref. Index Medium 1119867 554 0.88 0.94 (0.42-2.06) Low 16'21295 704 1.13 1.19 (0.57•2.48) Mantel Haenszel summary rate ratio (RR) - adjusted for age Total Person Years - 64.420 missing values in exposure factors_ Total Subjects - 11.060 Rate per 1 million person years Cancer Cases - 41 - _6ssrepancies in cases and person years are duo to - p=0.61 p0.41 p- 0.29 p 0.00 p 0.00 p-0.01 p=0.78 4V4E%z9zoz

of' lung cancer from the : analyses. The results for selected exposures are in Table 6.3. There is an increased risk for subjects who had been smokers; RR=1.56{959: C:I. (169,3:54), p'=0.29, and 6kewise an increased risk for subjects married to husbands who smoked, RR =1.31 (95~'p C.I. 0:67-2.57), p=0.41.. Females with a higher education level'had a reduced risk, RR=0.34 (95% C.I. &18-0:66), p-0.00, while subjects married to husbands who had blue collar occupations had an increasedl risk ofi disease, RR=2.34 (95% C.I. 1.17-4.70), p=0.01. Women who married at less than, 20 years of age were also at greater risk than those who married aU 20 years of age or more, RR!=2.70 (95% C.L 1.45-5;02), p=0.00: These three results probably indicate exposures related to socio-economic status. Further stratified analyses were conducted after including the nine lung cancer, cases. Table 6.4 indicates that the results are similar to those for the smoking related cancers excluding i lung cancers. As might be expected with the inclusion of' additional cases more strongly influenced by cigarette: smoke exposure, the rate ratio: for the variable, husband smoked l in marriage, increased slightly, RR!=1'.43 (:95% C.I. 0:79-2.59)4 . p= 0!22. On the i other, hand, the adjusted rate ratio for the variable female smoked decreased sGghtly from 1.56 in ithe previous table to 1.39. 118

TABLE 6.4 SPOUSE PAIRS - FEMALES AGE ADJUSTED RATE RATIOS (RR) OF ALL SMOKING RELATED CANCER FOR SELECTED EXPOSURE FACTORS (1977-1982) sVUZzszoz Factor Levels ' Cases/ +Rate/ f:rude Adjusted Person Years 106 PY RR RR- (95% C.L) Husband Smoking Never 32:47227 678 Status in Past 1411607 948 Marriage Current 3• 3929 764 Husband Smoked No 32'47227 678 in Marriage Yes 1716536 1094 - Fomale Smoked Nen-smoker 4264827 766 Smoker 8• 8431 949 Education High School 30'20991 1429 College + 20:43223 463 Age Married_ 20+ years 26'42031 618 < 20 years 24•'22377 1073 Spouse White collar 1434644 404 Occupation Blue collar 30-26066 1151 Beta Carotene High 1822372 805 Index Medium 1519864 755 Low 16 21294 751 Discrepancies in cases and person years are due to missing values in exposure factors - + Rate per 1 million person years Mantel Haenszet summary rate ratio (RR) • adjusted for age Ref. 1.40 1.13 Ref. 1.61 Ref. 1.24 Ref. 0.32 Ref. 1.74 Ref. 2.83 Ref. 0.94 0.93 Ref. 1.47 (0.78•2.79) 1.27 (0.39-4.15) p - 0.44 Re/. 1.43 (0.79-2.59) p - 0.22 Re/. 1.39 (0.65-2.98) p -0.40 Ref. 0.40 (0.22•0.71) p  0.00 Ref. 1.89 (1.08-3.30) p = 0.02 Ref. 2.56 (1.35-4.85) p a 0.00 Ref. 1.00 (0.50-1.99) 1.00 (0.61•1.98) p=0.99

This would have been expected since only one of, the nine lung cancer cases had previously smoked. The stratified analyses were repeated i for the 9,378 non-smokers and the results are reported I in, Table 6.5 and TaNe 6.6. for, smoking related cancers exduding and I induding lung cancer cases respectively. The age adjusted rate ratios for the main exposure variable of interest were lower than those reported in the previous table for aA females. For the outcome exduding lung cancers the adjusted RR was 1.06 (95% C.L 0.47- 2.36) and for al smoking related cancers including lung cancer the adjusted'rate ratio was 1.22 (959G Ca. 0.61-2.44), p=0:57: The conditional maximum likelihood estimate of the rate ratio of all smoking related cancers for non-smoking females married to smokers compared 1 to those married to non-smokers was 1.18 (95% C.1. 0.57-2.33), p=M63. Table 6.7 compares this conditionallmaximum likelihood estirnate and' exact: confidence intervals with the Mantel-Haenszel RR estimate and confidence intervals. Based on these results it would appear that passive smoking exposure in this cohort of non-smokers has minimal effect on the combined smoking related I cancers. However, some factors associated with education level', spouse occupation, and age married l appear to influence the risk of these cancers. 120. 0

TABLE 6.5 SPOUSE PAIRS - NON-SMOKING FEMALES AGE ADJUSTED JUSTED RATE RATIOS (RR) OF SMOKING RELATED CANCER (EXCLUDING LUNG) FOR SELECTED EXPOSURE FACTORS (1977-1982) ZS4Czz9zoz Factor Levels • Cases/ Person Years Jate/ 10 PY Crude RR Ad)usted RFi' (95% C.L) Husband Smoking Never 25%43000 581 Ref. Ref. Status in Past 6 8081 742 1.28 0.95 (0.39-2.32) _ Marriage Current 2! 2485 805 1.39 1.49 (0.35-6.27) Husband Smoked No 2543000 581 Ref. Ref. in Marr)age Yes 8• 10565 757 1.30 1.06 (0.47-2.36) Education High School 2016027 1248 Ref Ref College + - 14'38654 362 . 0.29 . 0.39 (0.19-0.79) Spouso White collar 11 31196 353 Ref. Ref. Occupation Blue collar 19'20734 916 2.60 2.14 (1.01-4.63) Ago Married 20 + years 15'37642 398 Ref. Ret. < 20 years 19'17197 1105 2.78 3.16 (1.60-6.24) ' Discrepancies in cases and person years are due to missing values in exposure factors Total Subjects - 9.379 + Rate per I million person years Cancer Cases - 34 - Mantel Haenszel summary rate ratio (RR) Total Person Years - 64.839 p 0.86 p 0.89 p 0.00 p0.04 p 0.00

TABLE 6.6 SPOUSE PAIRS • NON-SMOKING FEMALES AGE ADJUSTED RATE RATIOS (RR) OF SMOKING RELATED CANCER (INCLUDING LUNG) FOR SELECTED E(POS,URE FACTORS (1977•1982) Factor Levels ' Cases/ Person Years +(~ate/ 10 PY Crude RR Adjusted RR- (95% C.L) Husband Smoking Never 3042993 698 Ref. Ref. Status In Past 8 8078 990 1.42 1.07 (0.49•2.34) Marriago Current 3 2484 1208 1.73 1.83 (0.56-5.99) Husband Smoked No 3042993 698 Ref. Ref. in Marriage Yes 11'10562 1041 1.49 1.22 (0.61 •2.44) Education High School 23'16024 1435 Ref. Ref. College + 19•38645 492 0.34 0.45 (0.24•0.85) Spouse While collar 1331195 417 Ref. Ref. Occupation Blue collar 25..'20724 1206 2.89 2.47 (1.25•4.87) Age Marriod 20 + years 23.37631 611 Ref. Ref. < 20 years 19'1I197 1105 1.81 1.98 (1.07-3.64) ' Discrepancies in cases and person years are due to missing values in exposure factors Total Subjects • 9.378 + Rate per 1 million person years Cancer Cases - 42 -- Mantel Haenszel summary rate ratio (RR) Total Person Years • 54,827 p- 0.57 p- 0.00 p 0.01 p-0.02 zS4EZZ9zoz