Cancer in Migrants to Australia: Extending the Descriptive Epidemiological Data1

[CANCER RESEARCH 48, 751-756, February I, 1988] Cancer in Migrants to Australia: Extending the Descriptive Epidemiological Data1

Anthony J. McMichael and Graham G. Giles Department of Community Medicine, University of Adelaide, Adelaide, South Australia 5000 [A.J.M.J, and Cancer Epidemiology Centre, I Rathdowne Street, Carlton, Victoria 3053 [G. G. GJ, Australia

ABSTRACT pursue etiological hunches or hypotheses with the more labor- intensive, more expensive, analytical epidemiolÃ³gica! studies, Australia experienced a large influx of European migrants during the gathering data de novo at the level of the individual migrant. period 1950-1975. The descriptive epidemiolÃ³gica! data on cancer rates Paradoxically, few such studies have been done despite the fact within the major migrant groups, reviewed here, provide strong evidence that the extent of exposure variation (particularly for labile of environmental and behavioral influences on the etiology of various lifestyle factors such as dietary habits) is likely to be very much cancers. The opportunity to extend the conventional type of data analysis, greater in the first-generation migrant population than in the to include an examination of the effect of duration of residence upon cancer risk, provides further insight into cancer etiology. The possibilities general population of the host country. One important practical of further exploring the descriptive data are also discussed, and both the difficulty may be the usual lack of any migrant population desirability and the timeliness of mounting analytical studies of the register to serve as a framework for either cohort or case- Southern European migrants to Australia are emphasized. control studies. In this paper we present a selective review of the cancer experience of the major migrant groups in Australia, the great INTRODUCTION majority of whom have arrived since World War II. Where a Descriptive data on morbidity and mortality in migrant pop finer cut in the descriptive data is available, using the sort of ulations are a potent source of circumstantial evidence about subdivision shown in Fig. 1, we present and discuss it. In the relative importance of environmental factors in disease conclusion, we identify some of the potential studies that are etiology. When examined in relation to other ecological meas either planned or should be considered in the next decade to ures, such data can provide important clues about specific capitalize on the experiment of opportunity offered by this etiological factors. unique population of first generation migrants. Previously reported studies of cancer mortality in migrants have predominantly compared three ratesâ€”the Ro,2 Rh, and CANCER MORTALITY, 1962-1971 Rml. In some studies it has been possible to compare the rate in the first generation migrants (Rm 1) with that of their off Comprehensive analysis of cancer mortality in migrants to spring, the second generation (Rm2); in particular, the studies Australia has been carried out for the period 1962-1971, using of first and second generation Japanese migrants to the United routine mortality data files from the Australian Bureau of States (1-4) have demonstrated considerable variation in the Statistics (5). Some selected results of that analysis have been extent of postmigration change in the risk of different cancers. summarized in Table 1. For cancers of the large bowel, breast, These often-reported comparisons are shown in the top half and prostate, each of which has been associated, in a range of of Fig. 1. The figure also reminds us that the subset of the epidemiological studies, with the Western (affluent) lifestyle, eounlry-of origin population that migrates is likely to be a mortality rates have been consistently low in migrants from selected group with respect to various factors that bear on Southern Europe (i.e., Italy, Greece, Yugoslavia, and Malta). subsequent health risks. However, in view of the various hy These analyses of the early mortality experience of recent potheses that postulate the importance of age at which exposure immigrants to Australia have also demonstrated a general tend occurs or the likelihood that exposure factors (e.g., dietary ency for the rates to converge upon those of the Australian- habits) influence the later stages of carcinogenesis, it would be born population with increasing duration of residence in Aus desirable to be able to classify members of the migrating pop tralia. That tendency had previously been reported, more spe ulation according to such characteristics. Comparison of the cifically, for heart disease mortality in Italian migrants to cancer rates occurring in those subgroups would then yield Australia (6) and for gastrointestinal cancers in SEMs to Aus additional evidence about the likely etiological factors. tralia (7). Such convergence was apparent after 5 to 10 yr of In the context of descriptive, group-based analyses, certain residence in Australia. characteristics may be available for subdivision of the migrant population, e.g., age at migration, duration of residence in the CANCER MORTALITY, 1980-1982 host country, and extent of cultural change (perhaps estimable indirectly from such sources as membership lists of ethnic/ An analysis of more recent mortality in migrants to Australia cultural associations or traditional religious groups). These has been published for 1980-1982 (8). Although the disease possibilities are illustrated in Fig. 1. and country of birth categories were much broader than those Inevitably, there are few, if any, preexisting bases for subdi used in the 1962-1971 analyses, some comparisons were able vision of the migrant population. It is therefore necessary to to be drawn for all cancers and major groups of cancer. The mortality advantage of the SEMs (excluding Yugoslavs) was Received 4/20/87; revised 8/11/87; accepted 11/4/87. The costs of publication of this article were defrayed in part by the payment still apparent 10 yr later than the earlier data. The SMRs for of page charges. This article must therefore be hereby marked advertisement in all malignant neoplasms and three major groupings are sum accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1Presented at the Fifth Symposium on Epidemiology and Cancer Registries marized in Table 2 for all SEMs and for the four countries of in the Pacific Basin, November 16-21, 1986, Kauai, HI. Southern Europe with sizable migrant populations in Australia. 2The abbreviations used are: Ro, reported rate in the country of origin; Rh, It is apparent that Greeks and Italians have sustained their reported rate in the host country to which the migrants have moved; Rml, observed rate in migrating group; Rm2, second-generation migrants; SEM, South mortality benefit for longer than either the Yugoslavs or ern European migrant; SMR, standardized mortality ratio. Maltese. 751

Table 3 Standardized mortality ratios: Australia (1980-1982), cancer in migrants from selected birthplaces by period of residence in Australia (8) yrMalesUnited Period of residence in

Rote countryof origin (R_) Rate ho i t country (Ri)

IrelandSouthernKingdom and (selectivemigration)/... Europe"GreeceItalyAsiaFemalesUnited

'. : : : : ' : : Ageat Duration of Extent of IrelandSouthernKingdom and ; â€¢';;;;'..; . ; migration residenceâ€¢â€¢â€¢change cultural ; EuropeGreeceItalyAsia0-47948i_838448â€”â€”585-910474____7610958â€”_7210-14104777870721117475848515+11182778193US707668119

Fig. 1. Schematic representation of changes in cancer (incidence/mortality) Â°Includes Greece, Italy, Malta, Albania, Spain, Portugal, and Cyprus. rates associated with migration. Details of possible comparisons of rates within 4 â€”,fewer than 25 deaths. the migrating population (ml) are shown in the lower half.

Table 1 Cancer in migrants to Australia compared with the Australian-born 150 Iâ€” 147 U.K, IRELAND population: mortality during 1962 to 1971 (5) EUROPE105619Australian-bom74SOUTHERN Cancer site Remarks Colon and rectum Low in southern European, high in O1-sÃˆ_j^ British (especially Scottish) migrants Lung High in British migrants Breast Low in southern European migrants Skin and melanoma Low in all migrants 100oUJN Nasopharynx High in males from Asia (excluding India and Pakistan) and Malta (x9), Holland (x6) Cervix High (x20) in females from Middle and Northern European countries Prostate Low in southern European migrants Thyroid High (x3) in Welsh, Italian, and Oce1 Yugoslav females 50135O

Table 2 Standardized mortality ratios for cancer: Australia (1980-1982), in migrants from Southern Europe aged 15-74 at death (8)

SiteMalesDigestiveRespiratoryGenitourinaryTotalFemalesDigestiveRespiratoryGenitourinary'TotalGreece7777877781â€”8275Italy7595728072416768Malta74137t10376â€”9479Yugoslavia90997086118â€”9893SouthernEurope"7692778174397270

STOMACH COLON RECTUM CANCER SITE Fig. 2. Age-sex-standardized mortality ratios for gastrointestinal cancers in migrants from the United Kingdom/Ireland and from Southern Europe. Austra lian-born population = 100.

GASTROINTESTINAL CANCERS " Includes Greece, Italy, Malta, Albania, Spain, Portugal, and Cyprus. * â€”,fewer than 25 deaths. Stomach cancer mortality has been consistently higher in ' Includes breast. migrant groups than in the Australian-born population. In view of the well-documented inverse association of this cancer with The SMRs for most migrant groups for all malignant neo socioeconomic class, and since the migrants are predominantly plasms by period of residence in Australia demonstrated a from a working class (i.e., lower socioeconomic class) back ground and, in general, are also from countries with lower monotonie increase (Table 3) with increasing length of stay. standards of living than in Australia, this is an expected finding. One exception was that of Asian males who demonstrated a An early report was of increased stomach cancer mortality in declining SMR during the first 15 yr of residence and an Polish migrants (9), and a recent study has identified Chinese increase thereafter. For SEMs, however, after 15 yr or more of migrants at increased risk (10). The approximately 45% higher residence in Australia, their SMRs for all cancers remained low rates of stomach cancer mortality in both British-Irish and at 82 and 70 for males and females, respectively. Little break Southern European migrants are illustrated in Fig. 2 (11). down was available for subsites of cancer in other than British Recent incidence data from South Australia for the period migrants whose SMRs with increasing residence all exceeded 1977-1984 indicate that these stomach cancer risk differentials 100. After 15 yr of residence, the SMRs for digestive system are persisting unchanged (Table 4); indeed, the rates in migrants cancers were 76 and 75 for male and female SEMs, respectively, from Greece remain more than twice those of the Australian- the SMR for genitourinary and breast cancer in SEM females born population. The effect of early life exposures appears to being only 74 after 15 yr of residence, compared to 119 for be of more importance to stomach cancer risk than factors British and 137 for Asian women. acting later in life. 752

Table 4 Incidence of selected cancers in persons aged 30 plus by sex and country of birth: South Australia, 1977-1984 (standardized to world population) detail in Fig. 3, in which the migrant populations have been subdivided according to duration in residence in Australia (us birthUnited of ing the duration categories supplied by the Australian Bureau CancersiteStomachColonRectumLungBreastCervixUterusOvaryProstateSexMalesKingdom/ of Statistics). After standardization for age and sex, it can be Ireland421750 seen that, for both colon cancer and rectal cancer, there was a general pattern of convergence upon the Australian-born rate FemalesMales 1152 2329253419150 with increasing duration of residence. The fact that this change in risk was clearly evident after 16 yr of residence in Australia FemalesMales 4635 393522171 suggests that, whatever the environmental factors responsible (which almost certainly included dietary changes), they predom FemalesMales 411082211118251794Country inantly influenced the later stages of carcinogenesis.

FemalesFemalesFemalesFemalesFemalesMalesAustralia263913420251882Italy/Greece4778112261362 BREAST CANCER Incidence and death rates from breast cancer in women, within the past decade, showed moderate variation between different migrant groups, expressed relative to the rate in Aus tralian-born women. In South Australia during 1977-1984, the age-standardized, incidence rate of breast cancer in women aged 30 or more was approximately 25% lower in migrants from Italy and Greece than in Australian-born women, whereas it Table 5 National diets in the 1950s: relative consumption (Australia = 10) in was approximately 25% higher in migrants from the United migrant countries of origin versus Australia Kingdom and Ireland. Mortality data for the period 1980-1982 Sources are the Food and Agricultural Organization (1958, 1963, 1966) and showed a similar variation (8). The magnitude of these differ the Finnish Foundation for Alcohol Studies (1977). ences was generally less than between the migrant countries of originComestible Country of origin and Australia (i.e., Ro versus Rh; Fig. 1), suggesting that postmigration changes in environment and lifestyle influenced itemsCerealsPotatoes Kingdom102116996612111391016Ireland142778103514181310618Greece1694221413244226<162NotreportedItaly1310283169274517<11178the risk of breast cancer. Among Italian-born migrant women, mortality from breast tubersPulses,and nutsSugarVegetablesFruitsMeatEggsMilk cancer has been shown to vary substantially as a function of duration of residence in Australia (Fig. 4). These data apply to an earlier period, 1962-1971, but it is unlikely that the pattern illustrated has not continued to apply to more recent immigrant women. In view of the predominant age at death among the andcheeseTotal fatsVegetableanimal breast cancer cases and the fact that most of the observations oilsBeer reported in Fig. 4 occurred within 5-20 yr of arrival within (L)Â°Wine Australia, it is unlikely that a change in age at first completed (L)Spirits, 100% (L)United pregnancy could account for this variation in risk. It is more plausible that a change in diet, perhaps towards a higher intake '(L),. of saturated fats and/or energy, possibly coupled with reduced physical work expenditure of energy, accounted for some of the For mortality from cancers of the large bowel, Fig. 2 shows variation. that, whereas the British-Irish migrants have had rates similar to those of the Australian-born, the migrants from Southern LUNG CANCER Europe have had substantially lower rates. Likewise, the inci dence data from South Australia for the period 1977-1984 Dean's early study of lung cancer mortality in British mi (Table 4) show that colon cancer incidence rates in Southern grants to Australia compared their age-specific mortality rates European migrants are only about 55% those of the Australian with all Australia and with England and Wales and looked at born. differences in cigarette consumption (12). A later study showed The data in Table 5, showing relative food consumption per that British-Irish rates were almost twice those of the Austra person in these countries in the 1950s and 1960s, indicate that lian-born population (13). Subsequent analysis by duration of the British-Irish migrants brought with them a dietary culture residence demonstrated that the rates were lower in those similar to that of Australia (i.e., a variant of the chronology of migrants who had arrived in Australia more than 16 yr earlier earlier Australian colonial history) with the exception of their than in those who had arrived more recently (4). much higher consumption of potatoes, whereas the Southern Data from South Australia for 1977-1984 show that the age- European migrants had a very different dietary background. standardized incidence rates of lung cancer in British-Irish The consumption of cereals, pulses, nuts, vegetables, and migrants compared to the Australian-born population are 58% fruitsâ€”all sources of dietary fiberâ€”was much higher than in and 77% higher for men and women, respectively (Table 6). Australia, while the consumption of animal (saturated) fats was These variations were in accord with those in the prevalence of much lower. If diets high in saturated fats and low in dietary cigarette smoking, estimated from a population survey in 1981 fiber are, as widely postulated, a source of increased risk of and given in Table 8. Migrants from Italy and females from colorectal cancer, then the data in Table 6 may explain the Greece have lung cancer rates which are in accord with the colorectal cancer pattern shown in Fig. 2. estimated prevalence of smoking, which was particular^ low in This facet of migrant cancer experience is examined in more Greek-born women. However, Greek-born men have a 50% 753

1.0 - Australian-born

Fig. 3. Age-sex-standardized relative risk of colorectal cancer mortality in migrants to Australia, 1962-1976, by duration of stay. MIGRANTS < O z IO 111 in Australian born =1.0. FROM: >â€¢ o a _l O o UJ ir < ui O co ce LJ1.5 o 1-Â°

0.5 111 DC 0

MALIGNANT MELANOMA o

oc Australia has the highest reported national rates of cutaneous malignant melanoma in the world. From the incidence data shown in Table 7 it can be seen that the age-standardized rate oc in the Australian-born adult population has been approximately o 25-30 per 100,000 per yr. The fact that migrants from the o UJ United Kingdom and Ireland have had rates approximately half y o those of the Australian born, while having the sort of skin OC < o complexion that predominates in Australia suggests that sun z light exposure in early life may be of particular importance in s setting lifetime risk of malignant melanoma. If that is true, and if a darker (more "olive") complexion also confers protection against this cancer, then it is not surprising that migrants from 0-5 6-16 17Â« Italy and Greece have even lower rates. In a case-control study, RESIDENCE, IN YEARS Holman and co-workers have shown that, in migrants to Aus Fig. 4. Age-standardized mortality ratios for breast cancer in Italian women tralia, the younger the age at arrival the higher has been the migrants, during 1962-1971, by duration of residence in Australia. Australian subsequent risk of malignant melanoma (14). born = 1.0.

Table 6 Age-standardized incidence of lung cancer (Â¡977-1984) and prevalence CANCER IN SOUTHERN EUROPEAN MIGRANTS: A of cigarette smoking (Â¡981)in South Australia, by country of birth and sex RESEARCH OPPORTUNITY United King Differences in Disease. Southern European migrants present dom/ a much different cancer profile than either the Australian-born Australia Ireland Italy Greece or other migrant groups to Australia. Their persistent compar M MF M l MF ative cancer incidence and mortality advantage after more than PrevalencesmokingLung of 15 yr of residence in Australia make them an obvious group to study along the lines suggested in Fig. 1. Unlike other migrant cancerincidence(per 100,000)351082622441713539481401093615564 groups their transition towards the Australian-born cancer rates

Table 7 Age-standardized incidence of cutaneous malignant melanoma, by sex and country of birth: three Australian states, circa 1980 higher lung cancer incidence than Australian men but do not11United have a higher prevalence of smoking. Recent unpublished data from the New South Wales CancerTKingdom/d!-0nDr.State Registry3 indicate that, between the periods 1976-1977 and 1980-1982, the incidence of lung cancer in women has re (period)South F26 mained stable in the Australian-born, has declined by 13% in (1977-1981)NewAustralia 2232 British-Irish migrants, and has increased by 11% in Southern SouthWales(1975-1982)Western 2926 European migrants. Among men, the rates have increased in each of those three groups, by 21%, 18%, and 44%, respectively. Australia(1975-1976)AustraliaM24"

3J. Ford, personal communication. 14.F111213Italy/GreeceM376'F764"754Refers to total continental Europe; seeIrelandM101110Ref.

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1988 American Association for Cancer Research. CANCER IN MIGRANTS TO AUSTRALIA is slower and may now be static for certain cancers. Table 8 Australian dietary survey, 1983: average daily consumption of selected food items Between migrants from different countries within Southern Europe there exist differences with respect to their SMRs and the rate with which they approach the Australian norm. Italian AustralianFood EuropeanMales10994396715388148285234156213252074171936882Females54244048183592182582651182722920845052428 and Greek migrants, who numbered 275,887 and 146,625 at the 1981 census, are more similar to each other, having lower item(g)BreadPastaLeafy SMRs and incidence rates than Yugoslavs and Maltese. The speed with which cancer rates in migrant groups ap greensPotatoesTomatoes proach that of the host country is strong evidence of environ mental effects acting at a late stage of carcinogenesis as pro productsPeasand moters. The SEMs' lower rates and slower transition suggest andbeansCarrotsPumpkin, that there may be some factor(s) in their environment that may etc.Citrus protect against or slow cancer progression. The decreased can fruitBeef cer incidence is common to all age groups. The SEMs' SMRs andvealPoultrySaugesMilkCheesesButterBeverages are lowest for cancers of the colon, rectum, breast, and prostate, all cancers with epidemiological evidence of dietary factors and hormonal/reproductive factors in their etiology.

Dietary Differences. Relatively little work has been accom (mL)Orange plished on the dietary content and dietary transition within juiceTeaCoffeeBeerWine Australia's migrant groups. Hopkins' study of Italians in Perth showed that Italians ate more bread and pasta, salami, fish, and red wine than Australians and derived more of their protein (white)Wine and fat intake from vegetables (15). Australians consumed more (red)Males6714141253529192949753018254177474494813585823Females399157424251628544020918316635503459225214Southern beer, tea, and milk and ate more potatoes and beef products than Italians. enees in dietary intakes in SEMs compared to Australian born, With respect to dietary change on migration, Italians admit indicate their value for research. ted to eating more meat and fat after coming to Australia. It is in groups who are in transition from one culture to Studies of Greek migrants to Melbourne (16) and in migrants another that one can expect to find increased heterogeneity of from one island (Levkada) and of their siblings who remained exposure to the dietary factors suspected of either promoting on the island (17) showed that dietary changes occurred soon or protecting against cancer. But before one can take full after arrival and involved the addition of preferred foods, par advantage of the unique research opportunities offered by the ticularly meats (prepared in ways likely to reduce the potential SEMs it is important to find out the degree of heterogeneity in increase in saturated fat consumption). The Greek migrants' their diet. Base-line studies of the current food consumption desire to maintain their culture and cuisine was seen to be very patterns in Greeks and Italians need to be conducted for each strong, demonstrated by the popularity of home wine making group separately using weighed diet records. (The Australian and the continued consumption of olive oil and wild leafy-green dietary survey (18) was based on only one 24-h recall, and the vegetables. numbers of migrants were small.) Preliminary Findings of the 1983 national dietary survey have Given that sufficient dietary heterogeneity can be measured recently been published based on 24-h dietary recall by 6000 in the study population, it would be worthwhile to conduct participants, 375 of whom were migrants from Southern Eu case-control studies to explore the conventional hypotheses rope (18). These are the first population-based dietary data that about protective and promotional factors and test for the effects are available by migrant groups living in Australia. Table 8 of fat, fiber, energy, vegetables, alcohol, and other nutrients illustrates some of the differences in food consumption between while also examining other factors such as parity, obesity, and the SEMs and the Australian born. The findings are consistent exercise. with the migrants maintaining the cuisines of their countries of However, to address the question of diet and cancer ade origin (pasta, tomatoes, green leafy vegetables, citrus fruit, and quately, it is time to explore the problem from the point of view red wine) and enriching them by the addition of more meats, of individual variability. There is little further information to particularly beef and veal. be obtained from ecological studies comparing the mean intakes Directions for Future Research. Haenszel has pointed to the of different migrant groups. The way ahead is to identify exhaustion of research opportunities on migrants to the United populations within which there exist variations in consumption States (with the exception of the Japanese), but he sees promise that are likely to give new information. Case-control studies of in the post-World War II migrants to Australia, Brazil, and such populations might give some increased insight about the Canada (19). Many case-control and cohort studies have failed relationships which could be fruitful to explore, but the defini to confirm international correlations between certain dietary tive analysis will need to adopt a prospective approach. factors such as fat and cancers of the bowel and breast. Reasons The most powerful way to answer dietary questions with that have been suggested for this inconsistency include (a) the respect to the incidence of common tumors of the gut and possibility that the populations studied have been too homo reproductive organs is to mount a cohort study where infor geneous with respect to the exposure measure, (b) our dietary mation and tissue are collected from the migrants while they methods for measurement of exposure are poor, and (c) that are in good health and who are then followed up for the (for case-control studies at least) the value of retrospective data occurrence of cancer and other end points of interest, at which on diet is suspect. However, the observations of differences (to time they are entered into a case-control study from within the their advantage) in incidence and mortality from cancers of the cohort. It is anticipated that the base-line and case-control digestive and reproductive tracts (including breast), and differ- studies currently under way in Australia will culminate in a 755

9. Staszewski, J., McCall, M. G., and Stenhouse, N. S. Cancer mortality in unique form of prospective study that will advance our under 1962-66 among Polish migrants to Australia. Br. J. Cancer, 25: 599-610, standing of the role of diet in the etiology of some of our 1971. common cancers. 10. Zhang, Y. O., MacLennan, R., and Berry, G. Mortality of Chinese in New South Wales, 1969-1978. Int. J. Epidemiol. 13: 188-192, 1984. 11. McMichael, A. J., and Bonnet. A. Cancer profiles of British and Southern European migrants. Exploring South Australia's cancer registry data. Med. REFERENCES J. Ausi., 1: 229-232, 1981. 12. Dean, G. Lung cancer in Australia. Med. J. Aust., /: 1003-1011, 1962. 1. Haenszel, W. Cancer mortality among the foreign born in the United States. 13. McCall, M. G., and Stenhouse, N. S. Deaths from lung cancer in Australia. J. Nati. Cancer Inst., 26: 37-132, 1961. Med. J. Aust., /.â€¢524-525,1971. 2. Haenszel, W., and Kurihara, M. Studies of Japanese migrants. I. Mortality 14. HulmÃ¡n, C. D. J., Armstrong, B. K., Heenan, P. J., et al. The causes of from cancer and other diseases in the United States. J. Nail. Cancer Inst., malignant melanoma: results from the West Australian Lions melanoma 40:43-68, 1968. research project. In: R. P. Gallagher (ed.). Epidemiology of Malignant 3. Haenszel, W., Kurihara, M., Segi, M., et al. Stomach cancer among Japanese Melanoma. Berlin: Springer-Verlag, 1986. in Hawaii. J. Nati. Cancer Inst., 49: 969-988. 1972. 15. Hopkins, S., Margetts. B. M., and Armstrong, B. K. Dietary change among 4. Haenszel. W., Berg, J. W., Segi, M., et al. Large bowel cancer in Hawaiian Italians and Australians in Perth. Community Health Stud.. 4: 67-75, 1980. Japanese. J. Nati. Cancer Inst., 51: 1765-1779, 1973. 16. Rutishauser. I., and Wahlquist, M. L. Food intake patterns of Greek migrants 5. Armstrong, B. K., Woodings, T. L., Stenhouse, N. S., and McCall, M. G. to Melbourne in relation to duration of stay. Proc. Nutr. Soc. Aust., 8: 49- Mortality from Cancer in MÂ¡grantstoAustralia, 1962-1971. Perth: NHMRC 55, 1983. Research Unit in Epidemiology and Preventive Medicine, University of 17. Powles, J., Ktenas, D., Sutherland. C., el al. Food Habits in Southern Western Australia, 1983. European Migrants: A Case-Study of Migrants from the Greek Island of 6. Stenhouse, N. S., and McCall, M. G. Differential mortality from cardiovas Levkada. Prahran. Victoria: Department of Social and Preventive Medicine. cular disease in migrants from England and Wales, Scotland and Italy, and Monash Medical School. 1986. native born Australians. J. Chron. Dis., 23:423-431, 1970. 18. Cashel. K., English, R., Bennett, S.. et al. National Dietary Survey of Adults: 7. McMichael, A. J., McCall, M. G., Hartshorne, J. M., and Woodings, T. L. 1983.1. Foods Consumed. Canberra: Commonwealth Department of Health, Patterns of gastro-intestinal cancer in European migrants to Australia: the 1986. role of dietary change. Int. J. Cancer, 25: 431-437, 1980. 19. Haenszel, W. Migrant studies. In: D. Schottenfeld and J. F. Fraumeni (eds.). 8. Young, C. Selection and Survival: Immigrant Mortality in Australia. Can Cancer Epidemiology and Prevention, pp. 194-207. Philadelphia: W. B. berra: Department of Immigration and Ethnic Affairs, 1986. Saunders. 1982.

756

Anthony J. McMichael and Graham G. Giles

Cancer Res 1988;48:751-756.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/48/3/751

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/48/3/751. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.