Eur J Epidemiol (2008) 23:273280 DOI 10.

1007/s10654-008-9232-8

CANCER

Cancer in the Sami population of Sweden in relation to lifestyle and genetic factors
lander Henrik Gro nberg Sven Hassler Per Sjo Robert Johansson Lena Damber

Received: 9 August 2007 / Accepted: 12 February 2008 / Published online: 6 March 2008 Springer Science+Business Media B.V. 2008

Abstract The reindeer herding Sami of Sweden have low incidences of cancer. The aim of the present study was to investigate the cancer risk in a large cohort of Swedish Sami, containing Sami with different lifestyle and genetic Sami heritage. A cohort of 41,721 Sami identied in ofcial national registers between 1960 and 1997, was divided into two sub-populations - reindeer herding Sami (RS) and non-reindeer herding Sami (NRS). A demographically matched non-Sami reference population (NS) was used as standard when incidence and mortality ratios were calculated. Incidence and mortality data were obtained from the Swedish Cancer and Cause of Death Registers for the period 19612003. For Sami men, lower risks were found for cancers of the colon and prostate, and for malignant melanoma and non-Hodkins lymphoma, but higher for stomach cancer. The Sami women showed higher risks for cancers of the stomach and the ovaries, but lower risk for

cancer of the bladder. The RS demonstrated lower relative cancer risks compared with the NRS. The lowest relative risk was found among the RS men, while the highest were observed among the NRS women. The RS men who had adopted a more westernized lifestyle showed a similar relative risk for prostate cancer as that of the NS living in the same region. Most of these differences in cancer risks could probably be ascribed to differences in lifestyle. It is concluded that the traditional Sami lifestyle contains elements, e.g. dietary contents and physical activity that may protect them from developing cancer. Keywords Sami Reindeer herders Cancer incidence Genetic factors Lifestyle factors Cohort study

Introduction The Sami people are natives of northern Scandinavia. Their ` pmi, stretches over the northern traditional homeland, Sa parts of Norway, Sweden, Finland and the Kola Peninsula. Out of an estimated total Sami population of about 100,000130,000 individuals, approximately 20,000 40,000 inhabits Sweden [1]. The large majority of the Swedish Sami (9095%) is assimilated and has adopted an ordinary western lifestyle [e.g. 2]. Yet, there are a small population of reindeer herding Sami who still live more traditionally, e.g. a semi-nomadic lifestyle with dietary habits characterized by high consumption of reindeer meat and sh, and low intake of vegetables and fruit [3, 4]. In general, there is a shortage of knowledge on the health and living conditions of the Sami, particularly of the Sami populations of Sweden and Russia [2]. However, as a consequence of the radioactive fallout from nuclear weapon tests in the 1950s and 1960s and from the

lander (&) S. Hassler P. Sjo Southern Lapland Research Departement, 91232 Vilhelmina, Sweden e-mail: per.sjolander@vilhelmina.se S. Hassler Department of Nursing, Health and Culture, University West, Trollhattan, Sweden lander P. Sjo vle, Centre for Musculoskeletal Research, University of Ga Umea, Sweden nberg H. Gro Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden R. Johansson L. Damber Department of Radiation Sciences/Oncology, University of , Umea, Sweden Umea

123

274

S. Hassler et al.

Chernobyl accident in 1986, which contaminated large areas in northern Scandinavia, two studies on the cancer risk among Swedish reindeer herding Sami have been published [5, 6]. Both studies were based on a cohort of about 2,000 members of reindeer herding households identied in 1960, but with a longer follow-up period in the study by Hassler and co-workers [6]. Signicantly lower overall cancer risks were found for the reindeer herding Sami, particularly in comparison with the Swedish general population [5, 6], but also compared with a demographically matched reference population of non-Sami [6]. Lower risks than in the Swedish general population were observed for cancers of the colon, respiratory organs, female breast, prostate, kidneys, urinary bladder and skin, and for malignant lymphoma. Higher risk was only found for stomach cancer, whereas no differences in risks were found for cancers particularly sensitive to radiation, i.e. cancer of the thyroid gland and leukaemia. Studies from Norway and Finland on mixed populations of reindeer herding and non-reindeer herding Sami have showed lower risks for cancer in general, and especially for cancers of the colon, lung, bladder and the prostate [7, 8]. Thus, low incidences of cancer appear to be a common trait among Sami regardless of nationality. This conclusion is supported by Norwegian and Swedish mortality data demonstrating signicantly lower death rates due to cancer among the Sami, at least among men, as compared with demographically matched reference populations [9, 10]. The low cancer rates among the Sami has been suggested to be caused by dietary factors such as high intake of selenium from reindeer meat, and high consumption of omega-3 fatty acids from wild sh and reindeer [58]. It has also been indicated that physical activity may contribute to the lower cancer incidences, particularly among the reindeer herders [6, 8]. Finally, genetic protective factors have been discussed as potential causes of the lower cancer risk [69]. Taken together, it is known that the risk for developing cancer is reduced among reindeer herding Sami in Sweden, as well as in Sami populations in Finland and Norway.
Table 1 Age and gender distribution of the Sami cohort, separately shown for reindeer herding (RS) and non-reindeer herding Sami (NRS), and for the demographically matched reference population of nonSami (NS) RS Men N Median age (years) Age distribution (%) 020 years 2140 4160 61+ 29.2 33.1 24.3 13.5 4,451 31

However, we lack knowledge on whether the cancer incidence is reduced within the majority, non-herding, Sami population of Sweden. There is also a shortage of incidence data from sub-groups of Sami with different lifestyle and genetic heritage, as well as data on incidence changes over time. The aim of the present study was to analyse the cancer incidence in the Sami population of Sweden. The study was designed to enable gender and site specic comparisons of incidence data for reindeer herding Sami (RS), non-reindeer herding Sami (NRS) and non-Sami (NS). Other objectives were to analyse changes of the cancer risks over the follow-up period 19612003, and to shed light on the relative importance of Sami lifestyle and genetic heritage for their low cancer incidence.

Material and methods The study was approved by the ethics committee at the and conforms to the principles of the University of Umea Declaration of Helsinki, the International Ethical Guidelines for Biomedical Research Involving Human subjects, and the International Guidelines for ethical review for epidemiological studies. A previously constructed cohort of 41,721 Sami identied in ofcial national registers between 1960 and 1997, was used [9, 11]. The Sami cohort was divided into two sub-populations: (1) RS, containing a total of 7 482 reindeer herding Sami with family members (spouses and children), and (2) NRS, containing 34 239 non-reindeer herding Sami (Table 1). A non-Sami (NS) reference population was used for comparison with the Sami cohorts. The NS was constructed using National Kinship Registers in a similar way as in the construction of the Sami cohort [9, 11]. For each indexSami, four non-Sami index-persons were randomly selected from the National Population and Housing Census Registers to match the index-Sami regarding age, gender, area of residency (i.e., by parish or municipality) and date of identication. Forefathers, full siblings and descendants
NRS Women 3,031 25 37.4 38.8 18.5 5.3 Men 17,416 21 47.1 28.2 16.1 8.6 Women 16,823 22 43.8 28.8 17.9 9.6 NS Men 75,899 30 35.0 30.9 22.6 11.5 Women 69,031 30 35.1 31.6 22.3 10.9

123

Cancer in the Sami population of Sweden

275

of these non-Sami index-persons were identied in the National Kinship Register, thereby generating a reference population that was demographically similar to the Sami cohort. The NS population contained a total of 144,930 persons (Table 1). The observed cancer incidences in the NS population, for each sex, age group, year and diagnose, served as the standard for comparisons with the Sami data. Expected number of cases was calculated by applying the sex-, year- and cancer specic incidence rates within veyear age categories. Standard incidence ratios (SIR) and standard mortality ratios (SMR) were calculated as the ratio between observed and expected number of cancer cases. A Poisson distribution was assumed in the calculation of 95% condence intervals (95% CI) for the incidence ratios [12]. Comparisons between different calendar periods and sub-populations of particularly strong/ weak Sami lifestyle and genetic Sami heritage were based on a Poisson regression model adjusting for age and gender differences [13]. A P-value \0.05 indicated statistically signicant differences. The unique personal identication numbers of the Swedish citizens made it possible to extract person-specic data from the Swedish Cancer Register, Causes of Death Register, Population and Housing Census Registers, and Kinship Register (for details, see [11]). All diagnosed cases of malignant cancer, classied according to the International Classication of Diseases, Injuries and Causes of Death, Seventh Revision (ICD-7), was identied. The follow-up period started January 1, 1961, and ended December 31, 2003. To estimate the relative importance of lifestyle and genetic factors for the low cancer risk among the Sami, the risk for cancer of the prostate was analysed in separate subpopulations of RS that were expected to differ substantially regarding Sami lifestyle and genetic Sami heritage. In general, the lifestyle in the sparsely populated mountain regions in northern Scandinavia is more inuenced by the traditional Sami lifestyle than that in other parts of Scandinavia where the lifestyle is more westernized (cf. [10]). Thus, RS living in municipalities in the mountain region of Sweden are supposed to have a more traditional lifestyle whereas RS living in other regions in Sweden are assumed to have adopted a more westernized lifestyle. The genetic Sami heritage was indirectly assessed by analysing the occurrence of Sami and non-Sami ancestors among the RS. Through kinship data, genealogical pedigrees were constructed and RS with pedigrees that did not overlap with any of the NSs were dened as having strong genetic Sami heritage, and those with pedigrees that did overlap as having weak Sami heritage. Two pairs of sub-populations were dened: (1) traditional Sami lifestyle vs. western lifestyle, and (2) strong genetic Sami heritage vs. weak genetic Sami heritage (Table 3).

There are several reasons for choosing the cancer of the prostate for this analysis: (1) it is the most common site for cancer among men in Scandinavia, (2) the incidence of prostate cancer has been found to be signicantly lower among Sami in Sweden, Norway and Finland as compared with non-Sami reference populations [58], and (3) development of cancer of the prostate is related to a combination of heredity, ethnicity and lifestyle factors such as dietary habits and exercise [1418].

Results Between 1961 and 2003 the average annual incidence of cancer per 100,000 persons was 349 among the Sami (332 for men and 365 for women), 366 among the NS (376 for men and 356 for women), and 464 in the general Swedish population (473 for men and 455 for women). As in the general Swedish population the incidence of cancer among the Sami has increased over the last decades. This is illustrated in Fig. 1 where changes of the overall incidence are illustrated separately for RS, NRS, NS and the general Swedish population, and for men and women. It can be

600

Men

Incidence per 100000 persons

500

400

300

200
Sweden NS NRS RS

100

0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

600

Women

Incidence per 100000 persons

500

400

300

200
Sweden NS NRS RS

100

0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

Fig. 1 Overall incidence of cancer among reindeer herding Sami (RS), non-reindeer herding Sami (NRS), demographically matched non-Sami (NS), and the general Swedish population. Incidences shown as 10-years oating averages, 19702003

123

276

S. Hassler et al.

seen that the incidences among the Sami, both the RS and the NRS, were clearly lower than that of the general Swedish population, but only marginally lower than that of the NS living in the same area as the Sami. There was a tendency for these differences to decrease over time among the women, while the relative differences appear to be rather stable among the men. Over the entire follow-up period, a total of 1,964 cases of cancer were identied in the Sami cohort, 1 562 among the NRS and 402 among the RS (Table 2). Compared with the NS, the men in the Sami cohort showed a signicantly lower overall cancer risk, while no statistically signicant difference was found among the women (see All Sami in Table 2). Higher risks for stomach cancer were shown for both Sami women and men. Sami women demonstrated higher risk for cancer of the ovary and lower risk for bladder cancer. For Sami men, statistically signicant lower risks were found for cancers of the colon and the prostate, as well as for non-Hodgkins lymphoma. In relation to the NS reference population, the RS showed generally lower cancer risks than the NRS (Table 2). The RS men showed a signicantly lower risk for cancer in general and for cancer of the prostate, but higher risk for stomach cancer. Among the NRS men, lower risk was only observed for cancer of the colon. No signicant differences in risks were observed for the RS women compared with the NS women. The NRS women showed a signicantly higher total cancer risk and higher risks for cancer of the stomach and the ovaries, but lower risk for cancer of the bladder. For most of the cancers, no differences in relative risk (i.e. risk relative to the NS) were found between the two periods 19611982 and 19832003. However, Sami women (RS and NRS together) showed a statistically signicant increased relative risk for cancers of the stomach (from SIR 0.95, 95% CI 0.531.57, to SIR 2.06, 95% CI 1.442.87) and of the breast (from SIR 0.67, 95% CI 0.470.92, to SIR 1.11, 95% CI 0.961.27). An elevated relative risk for stomach cancer was also demonstrated for Sami men (RS and NRS together; from SIR 0.98, 95% CI 0.711.32, to SIR 1.53, 95% CI, 1.16 1.97). The Sami men also showed an increasing relative risk for developing leukaemia over the follow-up period (from SIR 0.33, 95% CI 0.110.77, to SIR 1.09, 95% CI 0.741.54). The analysis of prostate cancer in relation to lifestyle and genetic Sami heritage, showed that the risk for both developing and dying from prostate cancer were signicantly lower among Sami living in the mountain areas, where the traditional Sami lifestyle is expected to be strongest (Table 3). No statistically signicant differences in relative risks were observed between Sami with strong and weak genetic Sami heritage (Table 3).

Discussion The present study shows that the overall cancer risk is signicantly lower among Sami men compared with NS men living in the same area, while the overall risk for Sami women was similar to that of NS women. For Sami men, lower risks were found for cancers of the colon and prostate, and for malignant melanoma and non-Hodkins lymphoma, but higher for stomach cancer. The Sami women showed higher risks for cancers of the stomach and the ovaries, but lower risk for cancer of the bladder. These results are basically in agreement with previous results on Swedish RS and on Norwegian and Finnish Sami [58], and reinforce the conclusion that low cancer risk is a common feature among indigenous people in arctic regions (cf. [1921]). The present paper is the rst study where the cancer incidence in Swedish RS and NRS populations has been analysed and compared with data from a demographically matched population of NS. The design of the study permits assessment of alterations in cancer incidences over a period embracing more than four decades. In general, the RS demonstrated lower cancer risks compared with the NRS. The lowest relative risk was found among the RS men, while the highest were observed among the NRS women. In the total Sami population there was a statistically signicant trend over the follow-up period towards increasing relative risks for developing cancers of the stomach (both men and women), the breast (women only) and for leukaemia (men only). Together with results from a previous study showing increased mortality risks from cancer among Swedish Sami women [9], the present data indicate that the historically low cancer risks among the Sami have declined over the last decades, at least among the Sami women. Most of the differences in cancer risks observed in this study could probably be ascribed to differences in lifestyle between Sami and NS, between RS and NRS, and between Sami men and women. The traditional Sami lifestyle holds several factors that may reduce the risk for cancer. Dietary factors such as e.g. high intake of reindeer meat and wild sh rich in selenium, vitamin A and omega-3 fatty acids, and low consumption of dairy products [4, 22, 23], have been suggested to be partly responsible for the generally low cancer risk among more traditionally living Sami [5 8]. However, some of the characteristic dietary habits of the traditional Sami lifestyle, such as high intake of red meat, are also regarded as risk factors for developing gastrointestinal and prostate cancers [17, 18, 24, 25]. Moreover, it has been suggested that the high risk for stomach cancer among the Sami might be associated with low intake of fruit and vegetables and/or with high intake of smoked and salted meat and sh [37]. The radioactive fall-out

123

Table 2 Observed cases of cancer among reindeer herding Sami (RS) and non-reindeer herding Sami (NRS), together with expected cases and standard incidence ratios (SIR) with 95% condence intervals (in brackets) based on comparisons with a demographically matched reference population of non-Sami. Follow-up period 19612003 RS Men Obs Exp SIR Obs Exp SIR Obs Exp SIR Obs Exp SIR Women Men Women NRS All Sami Men SIR Women SIR

Cancer in the Sami population of Sweden

Site (ICD 7*)

All sites 35 20 17 9 22 35 8 46 100 0.46 (0.340.61) 9 18 3 1 7 6 10 12 0.82 (0.391.51) 3 6 0.95 (0.352.07) 0 13 0.52 (0.211.08) 2 4 0.51 (0.061.84) 2 0.00 (0.002.16) 4 0.77 (0.162.26) 2 0.53 (0.012.97) 1 3 0.33 (0.001.82) 5 21 20 28 8 0.37 (0.081.09) 2 4 0.48 (0.051.72) 11 21 0.87 (0.521.38) 0 3 0.29 (0.001.64) 55 12 0.77 (0.351.47) 1 3 0.29 (0.001.64) 27 7 1.08 (0.472.14) 189 208 0.91 (0.791.05) 27 0.99 (0.661.45) 43 1.27 (0.951.86) 17 0.66 (0.331.19) 5 1.09 (0.352.53) 30 0.70 (0.441.08) 13 1.49 (0.912.30) 32 0.89 (0.591.28) 21 9 18 15 26 9 24 44 0.80 (0.561.12) 10 0.92 (0.421.75) 24 0.92 (0.581.40) 6 5 3 2.13 (0.784.63) 5 0.98 (0.322.28) 17 38 22 0.76 (0.441.22) 50 0.76 (0.541.04) 24 25 60 16 1.05 (0.611.68) 5 4 1.31 (0.423.05) 27 33 0.81 (0.531.17) 29 23 0.86 (0.521.33) 7 9 0.80 (0.321.65) 32 47 0.68 (0.470.96) 63 24 1.43 (1.001.90) 6 4 1.59 (0.583.47) 68 59 1.15 (0.891.46) 44

275 344 0.80 (0.710.90) 127 146 0.87 (0.731.04) 719 759 0.95 (0.881.02) 843 786 1.07 (1.001.14) 0.90 (0.850.96) 1.04 (0.971.10) 29 1.52 (1.102.04) 1.23 (1.011.50) 1.53 (1.132.01) 50 1.26 (0.971.61) 0.74 (0.550.97) 1.19 (0.931.50) 24 1.23 (0.821.77) 0.89 (0.641.19) 1.24 (0.861.73) 19 1.24 (0.791.85) 0.81 (0.531.19) 1.35 (0.911.93) 27 0.93 (0.601.37) 0.81 (0.621.05) 0.94 (0.631.34) 212 201 1.05 (0.921.21) 38 1.58 (1.222.05) 0.76 (0.670.87) 25 0.85 (0.531.30) 0.93 (0.651.29) 0.78 (0.491.19) 17 0.52 (0.240.98) 1.14 (0.891.43) 0.44 (0.200.84) 18 0.97 (0.581.54) 0.57 (0.310.96) 0.88 (0.541.36) 14 1.04 (0.581.72) 0.93 (0.342.02) 0.92 (0.521.49) 21 1.23 (0.801.80) 0.65 (0.430.94) 1.12 (0.741.61) 12 0.78 (0.351.47) 1.32 (0.861.93) 0.68 (0.311.28) 27 0.89 (0.571.33) 0.87 (0.611.19) 0.88 (0.581.27) 1.01 (0.891.14) 1.51 (1.171.92)

Stomach (151)

Colon (153)

Rectum (154)

Pancreas (157) Lung (162)

Breast (170)

Ovary (175)

Prostate (177)

Kidney (180)

Bladder (181)

Malignant melanoma (190)

Thyroid (194)

Non-Hodgkins lymphomas (200, 202)

Myeloma (203)

Leukaemia (204209)

* Codes of the International Classication of Diseases, 7th revision

277

123

278

S. Hassler et al.

Table 3 Incidences and deaths from prostate cancer, between 1961 and 2003, among reindeer herding Sami with strong and weak Sami lifestyle and genetic Sami heritage, respectively. Standard incidence (SIR) and mortality ratios (SMR) with 95% condence interval (in brackets) based on comparisons with demographically matched non-Sami Incidence Obs Sami lifestyle Strong (n = 3 647) Weak (n = 3 835) Genetic Sami heritage Strong (n = 2 576) Weak (n = 4 906) *P \ 0.05 19 27 45 55 0.42 (0.250.66) 0.49 (0.320.71) 12 8 24 23 0.50 (0.260.88) 0.35 (0.150.69) 21 25 62 38 0.34 (0.210.51) 0.66 (0.430.97)* 4 16 34 13 0.12 (0.030.30) 1.23 (0.702.00)* Exp SIR Mortality Obs Exp SMR

products that have contaminated northern Scandinavia from the tests of nuclear weapons at Novaya Zemlya during the 1950s and 1960s and the nuclear power plant accident in Chernobyl in 1986, have been thought to generate a higher risk for radiation sensitive cancers among RS since radioactive particles are accumulating in lichen which is the main food source for the reindeer during the winter. In spite of exposure to several dietary risk factors, the Sami men demonstrated signicantly reduced susceptibility to cancer. The relative high levels of physical activity [26, 27], in combination with high intake of antioxidants and low intake of dairy products are suggested as the main reasons for the low risk for cancers of the colon and the prostate among the RS [68]. This was supported by the observation that RS who were assumed to have adopted a more westernized lifestyle, by migrating out of the reindeer herding region, appeared to develop a similar risk for prostate cancer as that of the NS living in the same region (Table 3). In a study on Sami in northern Norway a negative relation was reported between the incidence of prostate and colon cancers and the estimated consumption of reindeer meat [8]. Thus, the traditional Sami lifestyle may contains elements that protect them from contracting cancer, elements that appear to be related to dietary habits and physical activity. An important issue that should be addressed in future studies is to identify specic dietary and behavioural factors of the traditional Sami lifestyle that are causally related to the low cancer risks. Over the last decades, the Sami diet has been strongly inuenced by the food habits of the western society. The meat and sh consumption has decreased while the intake of fruit and vegetables, sugar, bread and dairy products has increased [4, 28]. These changes have resulted in a diet with less protein, zinc, phosphate, vitamin B12, vitamin A and selenium, more carbohydrates, vitamin C and betacarotene, and poorer fatty acid and amino acid compositions [4, 28]. The dietary habits have changed more among the NRS than among the RS [8, 23, 27, 28], which to an

important degree could explain the higher cancer risks among the NRS. During the follow-up period, the relative risk for leukaemia increased signicantly for the Sami men. The relative risk was substantially lower than among the NS between 1961 and 1982, but increased to the same level as that of the NS between 1983 and 2003. The diminishing difference in risk for leukaemia between the Sami and the NS reference population could be a result of more similar lifestyle over time. Another possibility is that the risk has increased more among the Sami than the NS as a consequence of higher exposure to radioactive fallout products from nuclear weapon tests in the 1950s and 1960s and the power plant accident in Chernobyl. This hypothesis would be in accordance with ndings of a small exposure related increase in the total cancer incidence in north Sweden as a result of the Chernobyl accidents [29]. However, to clarify to what extent the increased risk for leukaemia among the Sami is related to radioactive contamination, the expected relative risk should be estimated based on known levels of exposure before and after the nuclear weapon tests and the Chernobyl accident. The westernisation of the Sami has not only changed their food habits, but also inferred, for example, fertility control through contraceptives, reduced levels of physical activity and shorter breast feeding periods [3033]. Such lifestyle changes might contribute to the higher overall cancer risk among the Sami women as compared with the men, as well as for the womens increasing risk for ovarian and breast cancers. Thus, it is suggested that in the acculturation process, the Sami women have lost more cancer protective factors of the traditional Sami lifestyle than have the men. One important reason for this is that the RS women, along with motorization and reduced economic outcome of the reindeer herding husbandry, have been forced to take employment outside the Sami communities where they have been more exposed to the western lifestyle [2, 23, 27, 34]. Although most of the differences in cancer risks reported here could be explained by lifestyle differences, we

123

Cancer in the Sami population of Sweden

279 2. Hassler S. The health condition in the Sami population of Sweden, 19612002. Causes of death and incidences of cancer and University Medical Dissertations, cardiovascular diseases. Umea New Series No. 962 (PhD-thesis). glin L. The food and nutrient intake of a Swedish Sami pop3. Ha ulation. Arctic Med Res 1988; 47:13944. 4. Nilsen H, Utsi E, Bonaa K. Dietary and nutrient intake of a Sami population living in traditional reindeer herding areas in north Norway: comparisons with a group of Norwegians. Int J Circumpolar Health 1999; 58:12033. 5. Wiklund K, Holm L-E, Eklund G. Cancer risks in Swedish Lapps who breed reindeer. Am J Epidemiol 1990; 132:107882. lander P, Barnekow-Bergkvist M, Kadesjo A. 6. Hassler S, Sjo Cancer risk in the reindeer herding Saami population of Sweden 19611997. Eur J Epidemiol 2001; 17:96976. rvinen S, Pukkala E. Cancer incidence among Sami 7. Soininen L, Ja in Northern Finland, 19791998. Int J Cancer 2002; 100:3426. 8. Haldorsen T, Tynes T. Cancer in the Sami population of North Norway, 19701997. Eur J Cancer Prev 2005; 14:638. lander P, Gro nberg H, Damber L. 9. Hassler S, Johansson R, Sjo Causes of death in the Sami population of Sweden, 19612000. Int J Epidemiol 2005; 34:6239. 10. Tynes T, Haldorsen T. Mortality in the Sami population of North Norway, 197098. Scand J Public Health 2007; 35:30612. lander P, Ericsson AJ. Construction of a database 11. Hassler S, Sjo on health and living conditions of the Swedish Sami population. ld P, editors. Befolkning och Bosa ttning i Norr In: Lantto P, Sko nser i historiens sken. Centre for Sami etnicitet, identitet och gra University, Miscellaneous publications No. 1, Research, Umea 2004:107124. 12. Armitage P. Statistical methods in medical research. Oxford: Blackwell; 1971. 13. Breslow NE, Day NE, editors. Statistical methods in cancer research. The design and analysis of cohort studies, vol. 2. Lyon, France: International Agency for Research on Cancer; 1987. p. 13142. (IARC scientic publication no. 82). 14. Whittemore AS. Prostate cancer. In: Doll R, Fraumeni JF Jr, Muir CS, editors. Cancer Surveys volume 19/20: trends in cancer incidence and mortality. Plainview, NY: Cold Spring Harbor Laboratory Press; 1994. p. 30922. 15. Shibata A, Whittemore AS. Genetic predisposition to prostate cancer: Possible explanations for ethnic differences in risk. Prostate 1997; 32:6572. 16. Gallagher R, Fleshner N. Prostate cancer: 3. Individual risk factors. CMAJ 1998; 159:80713. nberg H. Prostate cancer epidemiology. Lancet 2003; 17. Gro 361:85964. 18. Wolk A. Diet, lifestyle and risk of prostate cancer. Acta Oncol 2005; 44:27781. 19. Mahoney M, Michalek AM. A meta-analysis of cancer incidence in United States and Canadian native populations. Int J Epidemiol 1991; 20:3237. 20. Nielsen NH, Storm H, Gaudette L, Lanier A. Cancer in circumpolar Inuit 19691988. Acta Oncol 1996; 35:62128. 21. Friborg J, Koch A, Wohlfarht J, Storm HH, Melbye M. Cancer in Greenlandic Inuit 19731997: a cohort study. Int J Cancer 2003; 107:101722. 22. Ringstad J, Aaseth J, Johnsen K, Utsi E, Thomassen Y. High serum selenium concentrations in reindeer breeding Lappish men. Arctic Med Res 1991; 50:1036. lander P, Daerga L. Risk factors 23. Edin-Liljegren A, Hassler S, Sjo for cardiovascular diseases among Swedish Samia controlled cohort study. Int J Circumpolar Health 2004;63 Suppl 2:2927. 24. Leo MA, Lieber CS. Alcohol, vitamin A, and beta-carotene: adverse interactions, including hepatotoxicity and carcinogenicity. Am J Clin Nutr 1999; 69:107185.

cannot of course exclude contributions from genetic factors. In several studies it has been shown that there are distinct genetic differences between the Sami and other European populations (e.g. [3537]). Therefore one would expect the frequency of functional polymorphisms, predisposing to disorders, to show some differences between the Sami and other Scandinavian populations. In studies on the serum protein orosomucoid (ORM) it has been found that the frequency of the alleles controlling the production of ORM is 10 times lower among Sami than NS [38, 39]. Since the ORM is thought to protect the tumour cells against immunological reactions, reduced serum concentrations of ORM may contribute to the low cancer incidence observed among the Sami. It should be emphasised though, that there is a shortage of studies aimed at studying functional polymorphisms in the Sami population, and that there are no data on how genetic and lifestyle factors interact in the genesis and development of cancer among the Sami. The cohort compiled for the present study was based on a previous reconstruction of the Sami population of Sweden [9, 11]. By comparing with a demographically matched reference population, known geographic differences in lifestyle related diseases were controlled for. For instance, in northern Sweden, as compared with in southern parts of the country, the life expectancy is lower and the incidence of cardiovascular diseases higher, but the incidence of cancer is lower [40, 41]. Since most of the Swedish Sami are living in the northern part of the country, comparisons of incidence and mortality rates with geographically matched reference populations, rather than the Swedish general population, enhance the possibility to disclose ethnic differences in morbidity and mortality that are related to lifestyle and genetic factors [9, 11]. An admixture of non-Sami in the Sami cohort, and vice versa, is unavoidable to some extent since mixed marriages have occurred for a long time, although at a very low frequency before the 20th century [42]. Yet, previous attempts to validate the Sami cohort through comparisons with independent sources of Sami ethnicity (i.e. the parish registers and the electoral registers of the national Sami parliament) have shown that the admixture was small in the cohorts [11].
Acknowledgements We are grateful for nancial support from the pmi Norra Norr l 1 Sa sterbotten and EU:s ma County Council of Va land (dnr. SN 1.4211/00).

References
lander P, Janlert U. Northern Scandinavia. In: 1. Hassler S, Sjo Young TK, Bjerregaard P, editors. Health transitions in arctic populations, Ch. 6. University of Toronto Press, 2008, pp 0000 [in press].

123

280 25. Kolonel LN, Altshuler D, Henderson BE. The multiethnic cohort study: exploring genes, lifestyle and cancer risk. Natl Rev Cancer 2004; 4:51927. 26. Hermansen R, Njlstad I, Fnneb V. Physical activity according to ethnic origin in Finnmark county, Norway The Finnmark Study. Int J Circumpolar Health 2002; 61:189200. lander P. Work-related muscu27. Daerga L, Edin-Liljegren A, Sjo loskeletal pain among reindeer herding Sami in Sweden-a pilot study on causes and prevention. Int J Circumpolar Health 2004;63 Suppl 2:3438. glin L. The nutrient density of present day and traditional diets 28. Ha and their health aspects: the Sami- and lumberjack families living in rural areas of north Sweden. Int J Circumpolar Health 1999; 58:3043. 29. Tondel M, Hjalmarsson P, Hardell L, Carlsson G, Axelson O. Increase of regional total cancer incidence in north Sweden due to the Chernobyl accident? J Epidemiol Community Health 2004; 58:10116. 30. Mellbin T. The children of Swedish nomad Lapps. A study of their health, growth and development. Acta Paediat 1961; 51:197. 31. Beach H. Reindeer-herd management in transition. The case of Tuorpon Saameby in northern Sweden. Uppsala Studies in Cultural Anthropology, 1981, No. 3. Stockholm (PhD-thesis). 32. Larsson E, Ogaard B, Lindsten R. Rearing of Swedish, Norwegian, and Norwegian Sami children. Scand J Dent Res 1993; 101:3825. pmi in a time of change. A study of Swedish Sami 33. Amft A. Sa living conditions during the twentieth century from a gender and ethnic perspective. Department of Archaeology and Sami studies. University 2000 (PhD-thesis). Umea

S. Hassler et al. lander P, Hassler S, Janlert U. Incidence and mortality of 34. Sjo stroke and acute myocardial infarction in the Swedish Sami populationrelations to income and level of education. Scand J Public Health 2008; 36:8491. 35. Sajantila A, Lahermo P, Anttinen T, Lukka M, Sistonen P, Savontaus M L, Aula P, Beckman L, Tranebjaerg L, Gedde-Dahl T, Issel-Tarver L, DiRienzo A, Paabo S. Genes and language on Europe: an analysis of mitochondrial lineages. Gen Res 1995; 5:4252. 36. Tambets K, Rootsi S, Kivisild T, Help H, Serk P, et al. The western and eastern roots of the Saami-the story of genetic outliers told by mitochondrial DNA and Y chromosomes. Am J Hum Genet 2004; 74:66182. , Vavruch-Nilsson V, Edin-Liljegren A, Sjo lander P, 37. Johansson A Gyllensten U. Linkage disequilibrium between microsatellite markers in the Swedish Sami relative to a worldwide selection of populations. Hum Genet 2005; 116:10513. m C, Beckman G, Beckman L. Orosomucoid 38. Fan C, Sikstro polymorphism in Finns, Swedes and Swedish Saamis. Hum Hered 1993; 43:2725. m C. Population studies in northern Sweden. XIX. 39. Fan C, Sikstro The orosomucoid polymorphism. Hum Hered 1994; 44:1823. 40. Cancer incidence in Sweden 2005. Stockholm, Sweden: National Board of Health and Welfare, 2007. n M, Wall S. Why are Diabetes, Stomach Cancer m L, Rose 41. Nystro and Circulatory Diseases More Common in Northern Sweden? Scand J Prim Health Care 1986; 4:512. lsstrategier. Kulturens Frontlinjer/Ku ld P. Samiska gifterma 42. Sko ns Norr 1998; 13: 3151 (in Swedish). lturgra

123

Copyright of European Journal of Epidemiology is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.