Synergistic Role of Alcohol for Esophageal Cancer Numerous investigators have found a synergistic relationship between the use of tobacco in various forms, alcohol consumption, and the development of cancer of the esophagus (119, 132, 143, 241, 243, 263, 299, 307, 323). Some investigators report that tobacco is a more important carcinogen than alcohol, but others report that the reverse is true. Most of the studies report a synergism with the combined use of tobacco and alcohol, resulting in higher rates of cancer of the esophagus than would be observed by the addition of the two exposures. The mechanisms by which these two factors interact are not known. Alcohol may act as a solvent for carcinogenic hydrocarbons in the tobacco smoke or may alter microsomal enzymes in the mucosal cells of the esophagus (306). This hypothesis has received support from experimental observations (150). It has been noted, however, that alcoholism may be accompanied by severe nutritional deficiencies, which also may predispose an individual to certain diseases (271). Experimental Studies There is experimental evidence that benda]pyrene is able to penetrate the cell membranes of the esophageal epithelium, produc- ing papillomas and squamous cell carcinoma. These studies and others are presented in the Part of this Report on mechanisms of carcinogenesis. Conclusion 1. Cigarette smoking is a major cause of esophageal cancer in the United States. Cigar and pipe smokers experience a risk of esophageal cancer similar to that of cigarette smokers. 2. The risk of esophageal cancer increases with increased smoke exposure, as measured by the number of cigarettes smoked daily, and is diminished by discontinuing the habit. 3. The use of alcohol in combination with smoking acts synergisti- cally to greatly increase the risk for esophageal cancer mortality. Cancer of the Urinary Bladder Pntroduction It is estimated that in 1982 in the United States there will be 37,100 new cases and 10,600 deaths from cancer of the bladder (2). The average annual incidence for males is almost three times that for females. 101 Cancer of the bladder resulted in 6,401 deaths in 1950 and 9,812 deaths in 1977 in the United States. The age-adjusted rate fell from 3.7 to 2.9 per 100,000. The age-adjusted mortality rate fell in all four color-sex groups (Figure 39). The rate for white males, who had the highest mortality from this disease, decreased by 5.7 percent between 1950 and 1977. Among other than white males, who had the second highest mortality rate from this disease, mortality declined by 2.6 percent. In contrast, the age-adjusted death rate for white females decreased by 36.4 percent, and that of other than white females fell 25.9 percent. White males between 45 and 74 years of age had lower death rates from cancer of the bladder in 1977 than in 1960, but older males had higher mortality. Among white females 45 years of age and older, mortality decreased over the study period. The death rate increased in other than white males 65 years of age or older and in other than white females 75 years of age or older (Figures 40 and 41). The age-specific death rates show no significant increases in either white males or white females when plotted on a three-dimensional graph for the period 1950-1977 (Figures 42 and 43). Most cancers of the bladder are transitional or squamous cell carcinomas. Unless these produce hematuria or obstruct the bladder outlet, they remain undiagnosed until quite late, making cure less likely. Five-year survival rates range from 4 percent for individuals with distant metastasis, to 21 percent for individuals with regional involvement, and to 72 percent with localized disease (2). For patients diagnosed with bladder cancer from 1960 to 1973, the overall 5-year survival rate was approximately 60 percent for whites and 30 percent for other than white (313). Certain occupational exposures are associated with an elevated risk for bladder cancer. Many of these are related to the exposure to certain aromatic amines in the work place. The first report of an association between cigarette smoking and human bladder cancer in the United States was based on a retrospective study of 321 men with bladder cancer (157). In the ensuing 35 years, other epidemiological and experimental data have established an association between cigarette smoking and bladder cancer. Several authors have conservatively calculated the percentage of bladder cancers that can be attributed to cigarette smoking. One study (313) estimated that 40 percent of male bladder cancers and 31 percent of female bladder cancers in the United States may be attributed to smoking cigarettes. This, is in agreement with the estimate by Cole et al. (48) of 39 percent in males and 29 percent in females. A Canadian study reported a population-attributable risk of bladder cancer due to cigarette smoking of 61 percent in males and 26 percent in females (129). 102 .o - +=HHITE HALES X=HHITE FEIIRLES OzNDNWHITE tIALES Cl=NONWHITE FEtlRLES a 10950 1955 1960 1965 1970 1975 CRLENORR YERRS HRLES "t ROE IN YERRS IBY S-YEAR RGE GROUPS) `5 I FEMALES 0 t + =1950-1956 Y-1957-1963 0=1964-1970 cl =1971-1977 5 t 126 100 "0 7t s 3 :: b 8 6( I i. HALES ROE IN YERRS (BY S-YEAR AGE GROUPS) 126 I 100 . +=1950-1956 %:1957-1963 0=1964-1970 clr1971-1977 FEtlALES 26 ROE IN YERRS IBY 6-YEAR AGE GROUPS) __/.H---- _---- _--- -- I I I I I I I I I I I I I I I I I .- 0 I 150 0 FIGURE 42.-Age-specific mortality rates by &year age groups for cancer of the bladder and other urinary glands for white males, United States, 1950-1977 SOURCE: National Cancer Institute ,198). Causal Significance of the Association Consistency, Strength, and Specificity of the Association There have been numerous retrospective studies of the relation- ship between smoking and bladder cancer (3, 46, 48, 55, 75, 139, 141, 157, 159, 188, 247, 253, 267, 313, 325, 327, 330). Almost all of these studies have found an association between smoking and cancer of the 106 FIGURE 43.-Age-specific mortality rates by B-year age groups for cancer of the bladder and other urinary glands for white females, United States, 1950-1977 SOURCE Natmnal Cancer Institute (1%) bladder with relative risk ratios for the smoker averaging two to three times that of the nonsmoker (Table 31). A retrospective population-based study of 470 confirmed cases of transitional cell or squamous cell cancers of the bladder found a positive relationship between cigarette smoking and bladder cancer (48). A dose-response 107 377-310 0 - 82 - 9 relationship was demonstrated for both the number of cigarettes smoked per day and different degrees of inhalation. In the TNCS study (2991, a significant association was found between cigarette smoking and bladder cancer. The Hawaiian study of five ethnic groups (113) also disclosed a positive association between smoking and bladder cancer. In a Canadian population- based retrospective stiady of 632 case-controlled pairs (1291, the relative risk for developing bladder cancer for those who had ever used cigarettes versus those who had never used cigarettes was 3.9 for males and 2.4 for females. A dose-response relationship was demonstrated, and reduced risk was associated with the use of filter cigarettes as compared with the use of nonfilter cigarettes. Several of the retrospective studies found a dose-response relationship of cigarette smoking for bladder cancer, with the risk increasing with increased number of cigarettes smoked per day, duration of cigarette smoking, or lifetime number of cigarettes. Further, a study of successive birth cohorts in four countries, including the United States, found increasing rates of bladder cancer with increasing smoking exposure, for both males and females (128). Several of the large prospective epidemiological studies have examined the relationship between cigarette smoking and bladder cancer and are summarized in Table 32. `On the average, cigarette smokers are twice as likely to die from cancer of the bladder as are nonsmokers. Several of these studies also show a moderate dose response relationship; however, this relationship is not as strong as that noted between smoking and lung, laryngeal, oral, and esopha- geal cancers (Table 33). Comparisons of mortality ratios for selected causes of disease suggest that the specificity of the association is not as great as that noted for the above cancers (Appendix Tables A and B). The American Cancer Society !&State Study (155) reported a reduced risk for bladder cancer among smokers of lower tar and nicotine cigarettes, a reduction which was statistically significant among females but not among males. The lower order of strength and specificity for bladder cancer than for cancers of the lung, larynx, oral cavity, or esophagus suggests that factors other than smoking may also be associated etiologically with bladder cancer. Bladder Cancer Mortaiity and Cessation of Smoking Wynder and Stellman (326) reported that the risk of bladder cancer decreased almost to the level of nonsmokers after about 7 years of cessation (Figure 44). More recent data from the U.S. Veterans and British Physicians prospective studies show bladder cancer mortality ratios for ex-smokers only half those for continuing smokers (68,224). 103 TAB&E 31.-Review of literature on smoking and bladder cancer reported since 1963-retrospective studies country Years of study Authors Relative risk smokers: nonsmokers Number of subjects CaseS Controls Study population U.S.A. 1957-60 U.S.A. 1951-61 Wynder et al. (325l Cobb and Ansell (4s) 3.58 7.3 300 131 3oa 120 Male patients Male VA hospital patients Male patients Male patients Female patients Male patients Male patients Female patients Female patients Bilharzial male patients Nonbilhanial male patients Male patients Female patients Male patients; Austrian population controls Female patients; Austrian population controls Poland 1958-64 Staszewski GXII U.S.A. 1958-64 Dunham et al. (75) 2.7 150 750 1.4' 334 350 1.20 159 177 <1 381 275 1.9 360 381 2.0 106 117 1.6 135 390 1.4 278 278 1.7 87 87 U.K. U.S.A. U.S.A. EBypt 1956-67 Anthony and Thomas (4) 1967-66 Cole et al. (48) 1965-71 Simon et al. (141) 1966-71 Makhyoun (157) Canada 1972-73 Morgan and Jain (288) 6.4b 156 4.4h 74 1.6 150 Austria 1972-75 Flamm et al. (84) 3.0 40 *Recalculated from author's data. b Heavy smokers ( 2 25 cigarettes per day) compared with nonsmokers SOURCE: Wynder and Goldsmith (313). c TABLE 32.-Bladder cancer mortality ratios-prospective studies POpUl8tiOIl Study size NOW All snloIcen eiguette GJmnlentJ SdWS MS lfm= smokenof1~cigatetteJ Males in White Ineludea all winmy 9-stdt.e study Males 1.00 200 tract-. Includes Pmsbte. British Ph- woo Male Dccton 1.00 211 Canadian V&IWlS Acs 25 stat.? study U.S. Veterans mm Genitourinvy anew Male 1.00 1.40 CO~MS~group ~poo Males and 1.00 256 483,ofm 1.00 280 F~Fkk.8 z-=io@J Penon- 1.00 215 Years Wif0da 88,158 Males in 9 Males 1.00 2.89 OCCUptiOllS Japanese study SWedi Study 265,118 Malen and FW&!S WJW Males and Females 1.00 1.00 1.00 1.00 2.00 (Males) 2.55 (Fe&) 1.80 (Males) Bladder+ 1.60 (Females) other urinary OrgMs For male ex-smokers, the risk after 15 years of not smoking was less than one-half that of current male smokers (1.29). Temporal Relationship of the Association Evidence for the temporal relationship of the association is provided by the prospective studies in which populations of initially disease-free subjects were followed for the development of bladder cancer. Reliable histological studies of bladder epithelium in smok- ers compared with nonsmokers have not been reported. 110 TABLE 33.-Bladder cancer mortality ratios by amount smoked-prospective studies Amount Smoked Study Population per Day Ratio Comments U.S. Veterans =mQ Nonsmoker 1.00 l-9 1.22 `Based on E-20 2.18 less than 21-39 2.78 20 deaths 240' 2.29 British Physicians 34,m males Nonsmoker 1.00 1-14 2.20 15-24 2.20 25 + 1.40 Grams of tobacco per day California males in 9 occupations Swedish Study wx@ Nonsmoker 1.00 males about `I* pk 1.52 about 1 pk 2.81 about 1% pk 5.41 Males Females =mQ males NS 1.00 NS 1.M) and l-7 gm/day 1.50 l-7 1.20 females 8-15 1.60 a15 2.10 16 + 2.70 16 + 0.80 5 No 541 1 PRESENT 1-3 44 7-10 11-15 16+ NON- SMOKERS SMOKERS FIGURE 44.--Relative risk of male ex-smokers for cancer of the bladder by years since quitting smoking SOURCE: Wynder and Stellman 1326). Coherence of the Association Dose-Response Relationship The finding of a dose-response relationship in both retrospective 111 and prospective studies (see page 106-107) strengthens the coherence of the association of smoking and bladder cancer. Correlation of Sex Differences in Bladder Cancer With Different Smoking Habits Two investigators (128, 185), reporting 10 years apart, found an association between time trends in smoking patterns and bladder cancer mortality among both males and females. Each found an increasing risk of bladder cancer with increasing smoking exposure. Correlation of Bladder Cancer Among Populations With Different Tobacco Consumption Coherence of the association is also illustrated by data showing a low prevalence of this disease in groups with small proportions of smokers (e.g., Mormons and Seventh Day Adventists) (79, 165, 166, 211,294). Bladder Cancer Mortality and Cessation of Smoking Cessation of smoking decreases the risk of bladder cancer com- pared to that of continuing smokers. A study of male ex-smokers (129) found a risk of less than one-half that of continuing smokers 15 years after quitting smoking; a similar finding was observed in two of the major prospective studies (68,224). Bladder Cancer and Non-Cigarette Tobacco Use Two prospective studies have noted a relationship between pipe and cigar smoking and cancer of the bladder (68, 131). In the British Physicians Study, a mortality ratio of 1.5 was observed for the combined category of pipe/cigar smokers, whereas in the U.S. Veterans Study, a relationship was noted only for pipe smokers (ratio 1,201. Synergistic Role of Other Substances for Bladder Cancer The relationship between cigarette smoking and occupational exposure(s) is complex and has not been clearly elucidated. A number of carcinogens specific for the human bladder have been identified (45). Some of .these compounds are found in cigarette smoke in very low concentrations. Cigarette smoking probably acts as an independent agent in the development of bladder cancer; however, there may also be additive or synergistic interactions between cigarette smoking and substances present in the work place. Those who work with dye stuffs, rubber, leather, print, paint, petroleum, and other organic chemicals are at higher risk for bladder cancer than workers not exposed. 112 Conclusion 1. Cigarette smoking is a contributory factor in the development of bladder cancer in the United States. This relationship is not as strong as that noted for the association between smoking and cancers of the lung, larynx, oral cavity, and esophagus. The term "contributory factor" by no means excludes the possibili- ty of a causal role for smoking in cancers at this site. Cancer of the Kidney Introduction Over the period 1950-1977, the age-adjusted mortality rate for kidney cancer rose from 2.2 to 2.6. The annual number of deaths due to cancer of the kidney increased from 3,643 to 7,373. It is estimated that in 1982 there will be 18,100 new cases and 8,300 deaths due to kidney and other urinary tract cancers in the United States (other than bladder cancer) (2). The death rate of white males was higher than that of the other three color-sex groups (Figure 45). While age-adjusted death rates increased, although at a decelerating pace, among white males throughout this period, rates among other than white males actually decreased slightly after 1967. Among white females, the age-adjusted rate increased between 1950 and 1957, when it stabilized. Among other than white females, who had the lowest age-adjusted rate of death from this disease, mortality rose from 1.2 to 1.4 per 100,000. In the white population, the mortality sex ratio (male/female) increased from 1.75 in 1950 to 2.24 in 1977, reflecting the rise in the male death rate and the relative stability of the female rate. In the other than white populations, the mortality sex ratio was slightly lower during the 2&year period. White males and white females were at greater risk from this disease than were their counterparts, although the white to other- than-white differential narrowed throughout the study period. In all four color-sex groups, death rates moved generally upward in the population between 45 and 84 years of age (Figures 46 and 47). In 1977, both white and other than white males had higher death rates from this disease than did white and other than white females in the 10-year age group from 35 to 44. The age-specific death rates for cancer of the kidney show an upward trend in the older age groups, without a significant increase in the rates for the younger age groups when plotted on a three- dimensional graph for the period 1950-1977 (Figures 48 and 49). There are four primary histological types of kidney cancer: (1) renal cell carcinoma, (2) nephroblastoma (Wilm's tumor), (3) sarco- 113 5` w p ps* E 80 c Y - s: z P , a2n 1965 ,--- 1970 1975 CALENDRR YERRS --.- CHLtNUHK YERRS AGE IN YEWS (BY S-YERR RGE GROUPS1 30 . 24 10 12 6 FEMRLES +=1950-1956 ?#=1967-1963 cl =1964-1970 (II -1971-1977 0 10 20 30 40 50 60 70 80 RCE IN YERRS IflY 5-YERR RGE GROUPS1 D- 4 . I KRLES 3 2, 2 II ii 7 0 0 10 20 30 40 50 60 70 80 RGE IN YEARS 181 S-YERR RCE CROUPS1 O- 4 FEhRLES + =1950-lY56 H-1957-1963 O-1964-1970 cl:1971-1977 B 8 b 8 12. 0 0 10 20 30 40 50 60 70 80 AGE IN YERRS (BY 5-YEfIR ROE GROUPS1 llllllll1lllliIIIIIIIIlIIl FIGURE 48.-Age-specific mortality rates by byear age groups for cancer of the kidney for white males, United States, 1950-1977 SOURCE: National Cancer Institute (198). ma, and (4) epithelial tumors of the renal pelvis. Renal cell carcinomas comprise about 90 percent of kidney tumors and generally affect individuals after age 40 (average 55 to 60) (197). This tumor may be silent until far advanced. The median survival time for kidney cancer in the adult is about 2.7 years for those aged 35 to 54 at the time of diagnosis and 1 year for those 65 or older (197). 117 FIGUFtE 49.-Age-specific mortality rates by 5-year age groups for cancer of the kidney for white females, United States, 195&1977 SOURCE: Natienal Cancer Institute ( 198). Epidemiological studies have established an association between cigarette smoking and kidney cancer. Causal Significance of the Assoqiation Consistency, Strength, and Specificity of the Association Several retrospective studies have examined the relationship between smoking and kidney carcinoma. Data from these studies 118 (Table 34) show a positive association between smoking and kidney cancer with relative risks ranging from 1.06 to over 5, with one study of renal pelvis cancer reporting a tenfold risk for heavy cigarette smokers. Other studies also reported an increasing relative risk of renal adenocarcinoma and cancer of the renal pelvis in cigarette smokers (20, 21, 130, 238); the increase of relative risk of renal adenocarcinoma among cigarette smokers was found for both males and females (320). A significant positive association between ciga- rette smoking and renal cancer was noted in the TNCS study (299) and in the Hawaiian Study of Five Ethnic Groups (113). In most of the prospective studies, cancer of the kidney refers to tumors arising from the renal parenchyma as well as to tumors in the renal pelvis and ureter. In several of the large prospective studies (Table 34), an association was found between cigarette smoking and cancer of the kidney. The mortality ratios for all cigarette smokers varied from 1.20 to almost 3, compared with nonsmokers. Four of the prospective studies have noted a dose- response relationship as measured by the number of cigarettes smoked per day for kidney cancer (68, 105,224,290). Data from these studies are presented in Table 35. Generally, heavy smokers have mortality ratios two to three times greater than nonsmokers. In the U.S. Veterans Study, Rogot and Murray observed a decline in kidney cancer mortality among ex-cigarette smokers with a mortality ratio of 1.21 versus 1.41 for continuing smokers. Thus, the strength of the association of cigarette smoking related to kidney cancer risk is less marked than that for cancer of the other sites discussed above. Chemical elements such as lead and cadmium, hormones, ionizing radiation, genetic susceptibilities, as well as tobacco smoke have each been suggested as potential etiologic factors in this disease (322). Several studies (21, 32, 130, 214) have shown that a substance present in tobacco smoke, di-methylnitrosamine, causes kidney tumors in rats. Temporal Relationship The prospective studies provide support for the temporal relation- ship of the association. Coherence of the Association Dose-Response Relationship The dose-response relationship noted in four of the prospective studies lends support to the coherence of the association between smoking and cancer of the kidney. 119 TABLE 34.-Kidney cancer mortality, ratios and relative risks, prospective and selected retrospective studies Number of kidney Mortality ratio or relative risk ratio Comments Population Study size cancer Non- Cigarette deaths smokers smokers Prospective Studies ACS 9State 188,000 54 Study white males ACS 25-Stat.e Study 440,558 males 104 u. s. VetL?lXlIs 290,ocm 257 1.00 California males in 9 occupations 68,153 males 27 1.00 Japanese 122,281 30 Study males British Physicians 34,ooo 48 1.00 males Bennington renal adenocarcinoma 100 Laubscher 100 casea (20. 21) 190 controls Schmauz and 43 caes of renal Cole pelvis or ureter we) 451 controls 18 1.00 Armstrong GO) Wynder 202 adenocarcinoma 1.00 et al. of kidney (322) 394 controls 1.00 1.00 1.00 1.00 Retmspactive Studies 1.00 106 adenocarcinoma 106 1.00 of kidney 30 carcinoma of 30 1.00 renal pelvis 139 controls 1.58 Based on 54 microscopically proved cases 1.42 1.57 1.41 2.48 1.20 2.88 AH smokers 5.1 10.0 Riik ratio for pipe - 10.3 cigar - 12.9 For smokers of more than 2% pks/day 1.06 1.80 2.00 1.50 (males) (females) Correlation of Sex Differences in Kidney Cancer With Different Smoking Habits There has been an increase in the white male to female ratio of deaths from kidney cancer. This trend does not demonstrate an 120 TABLE 35.-Kidney cancer mortality ratios by amount smoked per day-prospective studies Amount per Day Study/Ratio Comments Nonsmoker 1-9 l&19 m-39 40+ All smokers Nonsmoker 1-14 lb24 25+ All smokers Nonsmokers l-9 10-20 21+ All smokers Nonsmoker about 10 about 20 Over 30 All smokers U.S. Veterans 1.M) 0.95 1.32 1.63 2.59' 1.41 British Physwians" 1.00 2.66 3.00 3.00 2.66 ACS S-State Study"' 1.00 1.90 1.8 2.94 1.90 California Males in Various Occupations 1.00 0.86 3.30 2.57 2.46 `Less than 20 deaths "Grams of tobacco per day "`Includes genitourinary effect of the later initiation of smoking by females as evidenced so clearly by the recent increases in female lung and laryngeal cancer risks. Correlation of Kidney Cancer Mortality Among Populations With Different Tobacco Consumption The relative risk of kidney cancer is reduced in populations with a low proportion of smokers (79, 16.5, 166, 212, 294), although this reduction is not as great as that observed for lung, larynx, esophageal, and oral cancer. Smoking and Histologic Changes in the Kidney No human autopsy studies have been published which examine histologic changes in the kidney among smokers compared to nonsmokers. 121 Kidney Cancer and Non-Cigarette Tobacco Use An elevated relative risk of from tenfold to twelvefold has been reported for smokers of pipes or cigars in one study (21). The U.S. Veterans Study noted an association for pure pipe smokers (ratio 1.32) and for mixed smokers of pipe and cigars (ratio 1.52) and kidney cancer, but not for pure cigar smokers. Conclusion Cigarette smoking is a contributory factory in the development of kidney cancer in the U.S. The term "contributory factor" by no means excludes the possibility of a causal role for smoking in cancers of this site. Carcinoma of the Pancreas Introduction In 1982, it is estimated that there will be 24,800 new cases and 22,300 deaths from carcinoma of the pancreas in the United States (2). Pancreatic cancer caused the deaths of 8,953 persons in 1950 and 20,465 persons in 1977 (the data for 1977 include deaths coded under ICD No. 157). The age-adjusted death rate rose from 5.3 per 100,000 in 1950 to a peak of 6.8 in 1968, and has remained stable since, at about 6.7. After 1968, the age-adjusted death rate from this disease actually decreased slightly from 6.8 to 6.7 per 100,000. Increases in the age-adjusted rate between 1950 and 1967 resulted from increases in the mortality rates of all four color-sex groups (Figure 50), with white females showing the smallest increase and other than white males showing the largest. In 1950, white males and females had higher death rates from this disease than did males and females of other races. By 1977, the age-adjusted rate for whites was 22 percent lower than the rate for others. The age-adjusted death rate of white males increased from 6.4 to 8.3 per 100,000 over the study period, and that of white females rose slowly from 4.3 to 5.2. Rates nearly doubled in the other populations, rising from 3.4 to 6.6 in females and from 5.3 to 10.5 in males. Among white males 25 to 84 years of age, there was an increase in mortality from 1950 until 1967 (Figure 51). Thereafter, this trend was reversed, except in males 75 or older. Among other than white males, rates rose steadily during the 1950s and early 1960s and then leveled off or declined, except among those 55 or older, whose mortality rates continued to increase through 1977 (Figure 52). Both white and other females of most ages had increasingly higher mortality rates over the entire 1950-1977 period. Generally, the mortality sex ratio decreased with advancing age in both the vvhite and the other than white populations. The age- specific death rates over time show an increase in the older age 122 " - - 6 + =WHITE HALES ^ *m 31 %=WHITE FEHALES ". if zo D=NONWHITE HRLES $ Ek III=NONWHITE FEHALES 3 f if: K W ; #" o.*a* * * -a - *a - * *.a........,.., 1950 1956 1960 1965 1970 1976 m CRLENOAR YEARS URLES 3. I FEMRLES +=1950-1956 m =1957-1963 (3 -1964-1970 R=1971-1977 f = 150 3 I /F 0 0 10 20 30 40 50 60 70 80 RGE IN YEARS (BY 5-YEAR AGE GROUPS) 250 200 50 IIRLES ROE IN YEARS IBY S-YERR ROE GROUPS1 FEtlALES + =1950-1966 * =1967-1953 0 =1964-1970 cJ=1971-1977 AGE IN YEARS IBY S-YERR RGE GROUPS)