International Journal of Epidemiology ©International Epidemiological Association 1990
Vol. 19, No. 4 Printed in Great Britain
Sunlight and Breast Cancer Incidence in the USSR EDWARD D GORHAM*, FRANK C GARLAND* AND CEDRIC F GARLANDM
There is some indication from descriptive epidemiological studies,1"5 and from laboratory investigations6"'4 that vitamin D may play a role in reducing the occurrence of breast cancer. Mortality and incidence rates of breast cancer vary tremendously worldwide415"'7 and reveal an intriguing geographical pattern. The lowest rates of breast cancer generally occur in countries close to the equator. As latitude increases, reported breast cancer incidence and mortality rates also increase.4'5'5"23 This pattern is detectable within the US where age-adjusted mortality rates for breast cancer in the South are 1.5-2.0 times lower than those in the North.3-24 Laboratory studies suggest that vitamin D may reduce risk or severity of human breast cancer.6"14 Vitamin D receptors have been described for human breast cancer cell lines grown in tissue culture, and inhibition of in vitro growth by vitamin D and increased cell differentiation have been reported for several of these cell lines.10"1214 Casual exposure to sunlight is the primary, source of vitamin D for adults in the US,25 where the average daily dietary intake of vitamin D is only 80IU, less than half the Recommended Daily Allowance of 200 IU.
The USSR extends 5000 kilometres from North to South and the climate varies from subtropical to arctic. The opportunity for sunlight exposure sufficient to create vitamin D in populations living in this range of latitudes could be expected to vary considerably. For this reason, the relationship between ground-level sunlight energy and breast cancer incidence in the USSR was evaluated. METHODS Total average annual ambient sunlight energy measured in calories per cm2 per day as reported by Landsberg et al26 was plotted on a map of the USSR along with average annual age-adjusted (world standard) breast cancer incidence rates per 100 000 for the 15 republics of the USSR during the years 1969 to 1971." The association between sunlight levels and breast cancer incidence rates was analysed using simple linear regression. The degree of association was determined by calculation of the Pearson correlation coefficient (R), a line of least squares was plotted, and statistical significance was assessed using the F-test. Results of some case-control and descriptive epidemiological studies3'28"29 suggest that a positive relationship exists between socioeconomic status and breast cancer. While exact measures of the socioeconomic status in the republics of the USSR are scarce,30 the number of doctors per 1000 population was used to provide an approximate measure of socio-
'Department of Community and Family Medicine, University of California, San Diego, School of Medicine (M-007) La Jolla, California 92093, USA. tCancer Center Epidemiology ProgTam, University of California, San Diego, School of Medicine, La Jolla, California 92093, USA.
820
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
Gorham E D (Department of Community and Family Medicine, University of California, San Diego, La Jolla, California 92093, USA), Garland F C and Garland C F. Sunlight and breast cancer incidence in the USSR. Internationa/Journal of Epidemiology 1990, 19: 820-824. Epidemiological and laboratory evidence suggest that vitamin D may play a role in reducing risk of breast cancer. Lack of exposure to ultraviolet sunlight can increase the prevalence of vitamin D deficiency, and may place some populations at higher risk of breast cancer. The association between total average annual sunlight energy striking the ground and age-adjusted breast cancer incidence rates in the USSR was evaluated. Breast cancer had a threefold range of incidence. Sunlight levels varied from 210 to 400 calories per cm2 per day. A statistically significant negative association was found between breast cancer incidence rates and total sunlight levels (R - -0.75, p =• 0.001). The slope of the regression line corresponded to two additional cases per 100 000 per year for each reduction of 35 calories per cm2 of sunlight. The pattern of increased breast cancer incidence in regions of low solar radiation in the USSR is consistent with the geographical pattern seen for breast cancer mortality in the US and worldwide. A positive relationship between socioeconomic status and breast cancer incidence was also present in the Soviet Union, based on an approximate socioeconomic measure, the number of doctors per 1000 population (R = +0.89, p - 0.0001). The possibility that correlates of socioeconomic status, such as dietary, ethnic, or behavioural factors, could account for the association is discussed.
821
SUNLIGHT AND BREAST CANCER
economic status in various regions of the USSR. This information was available for all but one (Moldavia) regional tumour registry in the USSR.27
Estonia Latvia Lithuania Georgia Armenia Ukraine Russian SFSR Moldavia
23.5 22.8 20.2 18.0 17.5 16.9 15.6 14.4
levels of 280,325 and 375 calories per cm per day, and had age-adjusted incidence rates of 16.9,12.3, and 11.2 per 100 000 respectively. Finally, republics with clear skies and strong sunlight, such as Uzbekistan and Tadjikistan, located in a mountainous desert near Afghanistan, and Turkmenistan, located on the Iranian border, had sunlight levels of 375, 390 and 400 calories per cm2, and age-adjusted incidence rates of 9.0, 7.8 and 8.5, respectively, which were the lowest rates in the USSR. The negative association between breast cancer incidence rates and sunlight levels in the USSR was statistically significant (R = -0.75, p = 0.001) (Figure 2). The slope of the regression line (y = —0.05x + 30.7, where x = sunlight in calories per cm2 per day, and y = average annual age-adjusted incidence rate per 100 000 population for 1969 through 1971) corresponded to approximately two additional cases per 100 000 per year for each reduction of 35 calories per cm2 of sunlight.
Azerbaijan Byelorussia Kazakhstan Kirghizia Uzbekistan Turkmenistan Tadjikistan
13.5 13.1 12.3 11.2 9.0 8.5 7.8
FIGURE i Average annual age-adjustedbreast cancerincidence rate per100000 females, 1969-1971 and average annual total solar energy (calories per cm2 per day), USSR.
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
RESULTS The 15 republics of the USSR had a threefold range of incidence for breast cancer (Figure 1). Total sunlight levels varied in the range of 210 to 400 calories per cm2 per day (Figure 2). Two republics which received the least sunlight, Estonia and Latvia, had sunlight levels of 210 and 225 calories per cm2 per day2* and ageadjusted breast cancer incidence rates per 100 000 of 23.5 and 22.8 respectively. These were the highest reported rates in the USSR (incidence rates were not reported for Siberia, except as part of the Russian SFSR).27 Republics with intermediate amounts of sunlight had intermediate incidence rates. For example, the Ukraine, Kazakhstan, and Kirghizia had sunlight
2
822
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY 24 1 22
o o o d" o Q.
£ CD
20
Estonia
°
R = - 0.75 p value = 0.001
o Latvia
^^
oL1tbuan1a
^ ^ - ^ ^
18
o
Georgia
^Armenia
^\olUkraine o Russian SFSR ^ ^ - ^ ^ 0 Moldavia^^""^-^^
16 14
^Byelorussia
O
12
—
10.
Q AzerDai
^
stan-^CInjbizi. Uzbekistan--^.
200
220
240
260
280
300
320
340
Sunlight (calories per cm
360
380
TadJ1k1st« 400
420
per day)
FIGURE 2 Average annual age-adjusted breast cancer incidence rate per 100000 females, 1969-1971, versus average annual total solar energy (calories per cm2 per day), USSR.
Breast cancer incidence was positively associated with an approximate measure of socioeconomic status, the number of doctors per 1000 population (R = +0.89, p = 0.0001). The northwestern republics of Georgia, Latvia, and Estonia reported the greatest number of doctors per 1000 population and ranked fourth, second and first, respectively, in breast cancer incidence (Table 1). Southeastern republics such as Tadjikistan, Kirghizia, and Uzbekistan tended to have fewer doctors per 1000 population, and also had the lowest incidence of breast cancer (Table 1). DISCUSSION The inverse pattern of breast cancer incidence with respect to total solar energy present in the USSR is consistent with the general worldwide geographical pattern, and supports the idea that some factor associated with sunlight may inhibit the development of breast cancer in populations. Countries within 20 degrees of the equator have extremely high levels of sunlight and low incidence rates for breast cancer.4-26 Countries in the Caribbean and South and Central America such as Nicaragua, Honduras, El Salvador, and the Dominican Republic each get at least 440 calories per cm2 of sunlight each day.26 The same is true for countries in North Africa and Southern Asia. All have incidence rates of breast cancer which are four to six times lower than rates in North America and Northern Europe, where breast cancer is the third leading cause of cancer death.1617 One exception to this general pattern occurs in
Japan, a country with very low breast cancer mortality15 but situated at relatively high latitude. The Japanese diet is extremely rich in vitamin D, however, because fish popular in Japan contain large amounts of it. Japanese eat an average of 1000 grams of fish each week,31 and estimated mean daily Japanese intake of vitamin D varies from 800-1000 IUs per day,32 approximately ten times higher than the average adult intake in North America.25 No information was available on dietary vitamin D TABLE 1 Average annual age-adjusted breast cancer incidence rale per 100000 females, 1969-1971, average annual total solar energy (calories per square centimetre per day), and number of doctors per 1000 population, USSR. Breast cancer incidence rate
Sunlight (Cal/cm2)
Doctors per 1000 population
Armenia Azerbaijan Byelorussia Estonia Georgia
17.5 13.5 13 1 23.5 18.0
385 395 245 210 280
3.5 2.9 3.0 3.7 4.1
Kazakhstan Kirghizia Latvia Lithuania Moldavia
12.3 11.2 22.8 20.2 14.4
325 375 225 230 270
2.7 2.4 3.9 3.4 not reported
Russian SFSR Tadjikistan Turkmenistan Ukraine Uzbekistan
15.6 7.8 8.5 16.9 90
230 390 400 280 375
3.5 2.1 2.6 3.2 2.5
Soviet republic
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
o
823
SUNLIGHT AND BREAST CANCER
It is likely that a factor or group of factors which vary by social class may be responsible for the positive correlation between socioeconomic status and breast cancer risk observed in this and other studies. Factors which have been suggested to explain the positive association between socioeconomic status and breast cancer risk include differences in diet, alcohol intake,
and reproductive factors. " Another possible explanation is that people living in urban industrialized economies have a diminished opportunity for vitamin D production compared to people living in agriculturally-based economies where exposure to sunlight is more intense. Agriculturally-based economies tend to be more common at lower latitudes, where sunlight levels are high and breast cancer incidence is low. Lack of industry may also mean that ultraviolet light-blocking air pollution is not as common in agricultural regions. Air pollution levels have been previously reported to be positively associated with breast cancer mortality in Canada.2 Although caution must be exercised in interpreting findings based on aggregate populations, we found a strong negative association between solar energy levels and age-adjusted breast cancer incidence rates in the USSR. It is possible that regional variations in dietary vitamin D or other potential risk factors for breast cancer could correlate highly with sunlight measurements. For this reason, the results of this study should be considered provocative rather than definitive until studies of individuals are carried out. ACKNOWLEDGEMENTS Dr Cedric Garland was supported by a National Cancer Institute of the United States Specialized Cancer Center Core Support Grant to the University of California, San Diego Cancer Center (No. CA 23100). REFERENCES Colston K W, Berger U, Coombes R C. Possible role for vitamin D in controlling breast cancer cell proliferation. Lancet 1989; i: 188-91. 2 Gorham E D, Garland C F, Garland F C. Acid haze air pollution and breast and colon cancer mortality in 20 Canadian cities. Can J Pub Hlth 1989; 80: 96-100 3 Blot W J, Fraumeni J F Jr, Stone B J. Geographic patterns of breast cancer in the United States. JNCI1977; 59: 1407-11. 4 Muir C, Waterhouse J, Mack T, « al eds. Cancer incidence in five continents. Vol.5, Lyon: International Agency for Research on Cancer, 1987. (IARC Scientific Publication No 88). ' Facchini U, Camnasio M, Cantaboni A, et al. Geographical variation of cancer mortality in Italy. Int J Epidemiol 1985; 14: 538-48. 'Colston K W, Colston M J, Fieldsteel A H, Feldman D. Breast cancer and bone: presence of 1,25-dihydroxyvitamin D3 receptor in human epithelial cancer cell lines. Cancer Res 1982; 42: 856-9. 'Colston K W, Berger U, Coombes R C, a al. Breast cancer and bone: Presence of 1,25-dihydroxyvitamin D receptors in bone metastases from patients with primary carcinoma of the breast. J Bone Min Res 1988; 39 (suppO: 155 (Abstract). 1 DeLuca H F, Ostrem V. The relationship between the vitamin D system and cancer. Adv Exp Med Biol 1987; 206: 413-29. 'EismanJ A, Martin TJ.MacIntyre I.MoseleyJM. 1,25-dJhydroxy vitamin D receptor in breast cancer cells. Lancet 1979; 2: 1335-6. 1
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
intake within the republics of the USSR, but considerable ethnic differences in diet exist among the republics. Although results of laboratory and descriptive epidemiological studies suggest that a relationship between Vitamin D and breast cancer may exist, we advance this hypothesis as a partial explanation of the threefold variation in breast cancer incidence in the USSR, based on the strong association observed between sunlight and breast cancer incidence. Further study is needed to assess the degree to which the association between sunlight and decreased risk of breast cancer may be affected by ethnic differences in diet or other host characteristics. Total solar radiation was used in this study as a marker for a narrow segment of the ultraviolet spectrum (290-310 nanometres, wavelength) capable of producing vitamin- D in or on the skin.25 Certain tropospheric conditions and some kinds of air pollution can selectively block the portion of the spectrum responsible for vitamin D synthesis, while remaining transparent to other wavelengths of sunlight.2 No direct measures of ultraviolet radiation were available. The number of doctors per 1000 population was used in this study as a correlate of socioeconomic status. Additional measures of other socioeconomic factors would have been desirable, but are difficult to obtain. This measure of socioeconomic status was positively related to breast cancer incidence in the USSR, and indicates the presence of a socioeconomic gradient by latitude which is stronger than those reported in the US or other countries.3'28'29 Blot et al3 sought to determine if the 1.8-fold geographical differences in breast cancer mortality seen in the US resulted from confounding demographic variables, including socioeconomic and reproductive factors. They observed a positive association between breast cancer and socioeconomic status, but concluded that socioeconomic factors could not account for the geographical variation of breast cancer within the US. They found the major indicator of breast cancer mortality to be location of residence, when socioeconomic factors were simultaneously investigated. Similar analyses by Hems28 and MacMahon29 also led to the conclusion that while socioeconomic factors may be of importance, they were not sufficient to account for the geographical variation in breast cancer within the US or around the world.
20J3 M
824 10
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY 22
Canadian Cancer Society. Canadian Cancer Statistics 1987. Toronto, Canada, CCS, March 1987. SirverbergE,LuberaJA Cancer Statistics 1988. Ca. 1988; 38:5-27. " Mason T J, McKay F W, Hoover R, el al. Atlas of Cancer Mortality for U.S. Counties: 1950-69. Washington DC: US Government Printing Office, 1975 (DHEW Publication No. [N1H] 75-780). 23 Holick M F. In: Norman T, etal. eds. Vitamin D, Chemical and Clinical Update. New York, Walter de Gruyter, 1988. p 925-34. *Landsberg H E, Lippman H, Paffen K, et al. World Maps of Climatology, Second edition. New York, Springer-Verlag, 1965. 27 Napalkov N P, Tserkovny G F, Merabishvili V M, el al. Cancer Incidence in the USSR. (Supplement to Cancer Incidence in Five Continents, Volume 3). Lyon: International Agency for Research on Cancer, 1983 (IARC Scientific Publication No. 48), 59 a Hems G, Stuart A. Breast cancer rates in populations of single women. Br J Cancer 1975; 31: 118-23. B MacMahon B, Cole P, Brown J Etiology of breast cancer: a review. J Nail Cancer Insl 1973; 50: 21-42. "Pollard A P, ed. USSR Facts and Figures Annual, 1989. Vol. 13, London- Academic International Press, 1990; 143-^497. 31 Marcus R, Plimpton J. Guide to Japanese Food Tokyo: Shufuntomo, 1984, p 75-9 32 Lentner C, ed. Geigy Scientific Tables, Vol. I, West Caldwell, New Jersey, Ciba-Geigy, 1981; 259-60. 33 Rohan T E, Bain C J. Diet in the etiology of breast cancer. Epidemiol Rev 1987; 9: 12(M5. 34 Willet W C, Stampfer M J, Colditz G A, et al. Moderate alcohol consumption and risk of breast cancer N EnglJ Med 1987; 316: 1174-80. a
(Revised version received May 1990)
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
Eisman J A, Mclntyre I, Martin T J, Frampton R J. Normal and malignant breast tissue is a target organ for 1,25- (OH) 2 vitamin D,. Clin Endocrinol 1980; 13: 267-72. " Eismaji J A 1,25-dihydroxyvitamin D3 receptors and the role of 1,25-(OH)2-D3 in human cancer cells. In: KumarR,ed. Vitamin D. Basic and Clinical Aspects. Boston, Martinus Nijhoff, 1984. p 365-84. u Eisman J A, Suva L J, Martin T J. Significance of 1,25-dihydroxyvitamin Dj receptors in primary breast cancer. Cancer Res 1986; 46: 5406-8. D Frampton R J, Suva L J, Eisman J A, Findlay D M, Moore G E, Moseley J M, Martin T J. Presence of 1,25-dihydroxyvitamin Dj receptors in established human cancer cell lines in culture. Cancer Res 1982; 42: 1116-9. "Frampton R J, Omond S A, Eisman J A. Inhibitions of human cancer cell growth by 1,25-dihydroxyvitamin Dj metabolites. Cancer Res 1983; 40: 4440-7. u Menck H R, Henderson B E. Cancer incidence rates in the Pacific basin. NCI Monogr 1978; 53: 119-24. " Gardner M J, Winter P D, Taylor C P, el al. Atlas of Cancer Mortality in England and Wales, 1968-1978. London, John Wiley and Sons, 1983; 24. " Canada, Minister of National Health and Welfare. Mortality Atlas of Canada. Hull, Quebec: Canadian Government Publishing Centre, 1980; Map 17: 60. 18 Thein-Hlaing, Thein-Maung-Myint. Risk factors for breast cancer in Burma Int J Cancer 1978, 21: 432-7. "Soini I. Risk factors of breast cancer in Finland Int J Epidemiol 1977; 6: 365-73. ^Tulinius H, Day N E, Johannesson G, el al. Reproductive factors and risk for breast cancer in Iceland. Int J Cancer 1978; 21: 724-30 21 Hirohata T, Shigematsu T, Nomura A M Y, et al. Occurrence of breast cancer in relation to diet and reproductive history: a case-control study in Fukuoka, Japan. Nail Cancer Insl Monogr 1985; 69: 187-90
Vol. 19, No. 4 Printed in Great Britain
Sunlight and Breast Cancer Incidence in the USSR EDWARD D GORHAM*, FRANK C GARLAND* AND CEDRIC F GARLANDM
There is some indication from descriptive epidemiological studies,1"5 and from laboratory investigations6"'4 that vitamin D may play a role in reducing the occurrence of breast cancer. Mortality and incidence rates of breast cancer vary tremendously worldwide415"'7 and reveal an intriguing geographical pattern. The lowest rates of breast cancer generally occur in countries close to the equator. As latitude increases, reported breast cancer incidence and mortality rates also increase.4'5'5"23 This pattern is detectable within the US where age-adjusted mortality rates for breast cancer in the South are 1.5-2.0 times lower than those in the North.3-24 Laboratory studies suggest that vitamin D may reduce risk or severity of human breast cancer.6"14 Vitamin D receptors have been described for human breast cancer cell lines grown in tissue culture, and inhibition of in vitro growth by vitamin D and increased cell differentiation have been reported for several of these cell lines.10"1214 Casual exposure to sunlight is the primary, source of vitamin D for adults in the US,25 where the average daily dietary intake of vitamin D is only 80IU, less than half the Recommended Daily Allowance of 200 IU.
The USSR extends 5000 kilometres from North to South and the climate varies from subtropical to arctic. The opportunity for sunlight exposure sufficient to create vitamin D in populations living in this range of latitudes could be expected to vary considerably. For this reason, the relationship between ground-level sunlight energy and breast cancer incidence in the USSR was evaluated. METHODS Total average annual ambient sunlight energy measured in calories per cm2 per day as reported by Landsberg et al26 was plotted on a map of the USSR along with average annual age-adjusted (world standard) breast cancer incidence rates per 100 000 for the 15 republics of the USSR during the years 1969 to 1971." The association between sunlight levels and breast cancer incidence rates was analysed using simple linear regression. The degree of association was determined by calculation of the Pearson correlation coefficient (R), a line of least squares was plotted, and statistical significance was assessed using the F-test. Results of some case-control and descriptive epidemiological studies3'28"29 suggest that a positive relationship exists between socioeconomic status and breast cancer. While exact measures of the socioeconomic status in the republics of the USSR are scarce,30 the number of doctors per 1000 population was used to provide an approximate measure of socio-
'Department of Community and Family Medicine, University of California, San Diego, School of Medicine (M-007) La Jolla, California 92093, USA. tCancer Center Epidemiology ProgTam, University of California, San Diego, School of Medicine, La Jolla, California 92093, USA.
820
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
Gorham E D (Department of Community and Family Medicine, University of California, San Diego, La Jolla, California 92093, USA), Garland F C and Garland C F. Sunlight and breast cancer incidence in the USSR. Internationa/Journal of Epidemiology 1990, 19: 820-824. Epidemiological and laboratory evidence suggest that vitamin D may play a role in reducing risk of breast cancer. Lack of exposure to ultraviolet sunlight can increase the prevalence of vitamin D deficiency, and may place some populations at higher risk of breast cancer. The association between total average annual sunlight energy striking the ground and age-adjusted breast cancer incidence rates in the USSR was evaluated. Breast cancer had a threefold range of incidence. Sunlight levels varied from 210 to 400 calories per cm2 per day. A statistically significant negative association was found between breast cancer incidence rates and total sunlight levels (R - -0.75, p =• 0.001). The slope of the regression line corresponded to two additional cases per 100 000 per year for each reduction of 35 calories per cm2 of sunlight. The pattern of increased breast cancer incidence in regions of low solar radiation in the USSR is consistent with the geographical pattern seen for breast cancer mortality in the US and worldwide. A positive relationship between socioeconomic status and breast cancer incidence was also present in the Soviet Union, based on an approximate socioeconomic measure, the number of doctors per 1000 population (R = +0.89, p - 0.0001). The possibility that correlates of socioeconomic status, such as dietary, ethnic, or behavioural factors, could account for the association is discussed.
821
SUNLIGHT AND BREAST CANCER
economic status in various regions of the USSR. This information was available for all but one (Moldavia) regional tumour registry in the USSR.27
Estonia Latvia Lithuania Georgia Armenia Ukraine Russian SFSR Moldavia
23.5 22.8 20.2 18.0 17.5 16.9 15.6 14.4
levels of 280,325 and 375 calories per cm per day, and had age-adjusted incidence rates of 16.9,12.3, and 11.2 per 100 000 respectively. Finally, republics with clear skies and strong sunlight, such as Uzbekistan and Tadjikistan, located in a mountainous desert near Afghanistan, and Turkmenistan, located on the Iranian border, had sunlight levels of 375, 390 and 400 calories per cm2, and age-adjusted incidence rates of 9.0, 7.8 and 8.5, respectively, which were the lowest rates in the USSR. The negative association between breast cancer incidence rates and sunlight levels in the USSR was statistically significant (R = -0.75, p = 0.001) (Figure 2). The slope of the regression line (y = —0.05x + 30.7, where x = sunlight in calories per cm2 per day, and y = average annual age-adjusted incidence rate per 100 000 population for 1969 through 1971) corresponded to approximately two additional cases per 100 000 per year for each reduction of 35 calories per cm2 of sunlight.
Azerbaijan Byelorussia Kazakhstan Kirghizia Uzbekistan Turkmenistan Tadjikistan
13.5 13.1 12.3 11.2 9.0 8.5 7.8
FIGURE i Average annual age-adjustedbreast cancerincidence rate per100000 females, 1969-1971 and average annual total solar energy (calories per cm2 per day), USSR.
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
RESULTS The 15 republics of the USSR had a threefold range of incidence for breast cancer (Figure 1). Total sunlight levels varied in the range of 210 to 400 calories per cm2 per day (Figure 2). Two republics which received the least sunlight, Estonia and Latvia, had sunlight levels of 210 and 225 calories per cm2 per day2* and ageadjusted breast cancer incidence rates per 100 000 of 23.5 and 22.8 respectively. These were the highest reported rates in the USSR (incidence rates were not reported for Siberia, except as part of the Russian SFSR).27 Republics with intermediate amounts of sunlight had intermediate incidence rates. For example, the Ukraine, Kazakhstan, and Kirghizia had sunlight
2
822
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY 24 1 22
o o o d" o Q.
£ CD
20
Estonia
°
R = - 0.75 p value = 0.001
o Latvia
^^
oL1tbuan1a
^ ^ - ^ ^
18
o
Georgia
^Armenia
^\olUkraine o Russian SFSR ^ ^ - ^ ^ 0 Moldavia^^""^-^^
16 14
^Byelorussia
O
12
—
10.
Q AzerDai
^
stan-^CInjbizi. Uzbekistan--^.
200
220
240
260
280
300
320
340
Sunlight (calories per cm
360
380
TadJ1k1st« 400
420
per day)
FIGURE 2 Average annual age-adjusted breast cancer incidence rate per 100000 females, 1969-1971, versus average annual total solar energy (calories per cm2 per day), USSR.
Breast cancer incidence was positively associated with an approximate measure of socioeconomic status, the number of doctors per 1000 population (R = +0.89, p = 0.0001). The northwestern republics of Georgia, Latvia, and Estonia reported the greatest number of doctors per 1000 population and ranked fourth, second and first, respectively, in breast cancer incidence (Table 1). Southeastern republics such as Tadjikistan, Kirghizia, and Uzbekistan tended to have fewer doctors per 1000 population, and also had the lowest incidence of breast cancer (Table 1). DISCUSSION The inverse pattern of breast cancer incidence with respect to total solar energy present in the USSR is consistent with the general worldwide geographical pattern, and supports the idea that some factor associated with sunlight may inhibit the development of breast cancer in populations. Countries within 20 degrees of the equator have extremely high levels of sunlight and low incidence rates for breast cancer.4-26 Countries in the Caribbean and South and Central America such as Nicaragua, Honduras, El Salvador, and the Dominican Republic each get at least 440 calories per cm2 of sunlight each day.26 The same is true for countries in North Africa and Southern Asia. All have incidence rates of breast cancer which are four to six times lower than rates in North America and Northern Europe, where breast cancer is the third leading cause of cancer death.1617 One exception to this general pattern occurs in
Japan, a country with very low breast cancer mortality15 but situated at relatively high latitude. The Japanese diet is extremely rich in vitamin D, however, because fish popular in Japan contain large amounts of it. Japanese eat an average of 1000 grams of fish each week,31 and estimated mean daily Japanese intake of vitamin D varies from 800-1000 IUs per day,32 approximately ten times higher than the average adult intake in North America.25 No information was available on dietary vitamin D TABLE 1 Average annual age-adjusted breast cancer incidence rale per 100000 females, 1969-1971, average annual total solar energy (calories per square centimetre per day), and number of doctors per 1000 population, USSR. Breast cancer incidence rate
Sunlight (Cal/cm2)
Doctors per 1000 population
Armenia Azerbaijan Byelorussia Estonia Georgia
17.5 13.5 13 1 23.5 18.0
385 395 245 210 280
3.5 2.9 3.0 3.7 4.1
Kazakhstan Kirghizia Latvia Lithuania Moldavia
12.3 11.2 22.8 20.2 14.4
325 375 225 230 270
2.7 2.4 3.9 3.4 not reported
Russian SFSR Tadjikistan Turkmenistan Ukraine Uzbekistan
15.6 7.8 8.5 16.9 90
230 390 400 280 375
3.5 2.1 2.6 3.2 2.5
Soviet republic
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
o
823
SUNLIGHT AND BREAST CANCER
It is likely that a factor or group of factors which vary by social class may be responsible for the positive correlation between socioeconomic status and breast cancer risk observed in this and other studies. Factors which have been suggested to explain the positive association between socioeconomic status and breast cancer risk include differences in diet, alcohol intake,
and reproductive factors. " Another possible explanation is that people living in urban industrialized economies have a diminished opportunity for vitamin D production compared to people living in agriculturally-based economies where exposure to sunlight is more intense. Agriculturally-based economies tend to be more common at lower latitudes, where sunlight levels are high and breast cancer incidence is low. Lack of industry may also mean that ultraviolet light-blocking air pollution is not as common in agricultural regions. Air pollution levels have been previously reported to be positively associated with breast cancer mortality in Canada.2 Although caution must be exercised in interpreting findings based on aggregate populations, we found a strong negative association between solar energy levels and age-adjusted breast cancer incidence rates in the USSR. It is possible that regional variations in dietary vitamin D or other potential risk factors for breast cancer could correlate highly with sunlight measurements. For this reason, the results of this study should be considered provocative rather than definitive until studies of individuals are carried out. ACKNOWLEDGEMENTS Dr Cedric Garland was supported by a National Cancer Institute of the United States Specialized Cancer Center Core Support Grant to the University of California, San Diego Cancer Center (No. CA 23100). REFERENCES Colston K W, Berger U, Coombes R C. Possible role for vitamin D in controlling breast cancer cell proliferation. Lancet 1989; i: 188-91. 2 Gorham E D, Garland C F, Garland F C. Acid haze air pollution and breast and colon cancer mortality in 20 Canadian cities. Can J Pub Hlth 1989; 80: 96-100 3 Blot W J, Fraumeni J F Jr, Stone B J. Geographic patterns of breast cancer in the United States. JNCI1977; 59: 1407-11. 4 Muir C, Waterhouse J, Mack T, « al eds. Cancer incidence in five continents. Vol.5, Lyon: International Agency for Research on Cancer, 1987. (IARC Scientific Publication No 88). ' Facchini U, Camnasio M, Cantaboni A, et al. Geographical variation of cancer mortality in Italy. Int J Epidemiol 1985; 14: 538-48. 'Colston K W, Colston M J, Fieldsteel A H, Feldman D. Breast cancer and bone: presence of 1,25-dihydroxyvitamin D3 receptor in human epithelial cancer cell lines. Cancer Res 1982; 42: 856-9. 'Colston K W, Berger U, Coombes R C, a al. Breast cancer and bone: Presence of 1,25-dihydroxyvitamin D receptors in bone metastases from patients with primary carcinoma of the breast. J Bone Min Res 1988; 39 (suppO: 155 (Abstract). 1 DeLuca H F, Ostrem V. The relationship between the vitamin D system and cancer. Adv Exp Med Biol 1987; 206: 413-29. 'EismanJ A, Martin TJ.MacIntyre I.MoseleyJM. 1,25-dJhydroxy vitamin D receptor in breast cancer cells. Lancet 1979; 2: 1335-6. 1
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
intake within the republics of the USSR, but considerable ethnic differences in diet exist among the republics. Although results of laboratory and descriptive epidemiological studies suggest that a relationship between Vitamin D and breast cancer may exist, we advance this hypothesis as a partial explanation of the threefold variation in breast cancer incidence in the USSR, based on the strong association observed between sunlight and breast cancer incidence. Further study is needed to assess the degree to which the association between sunlight and decreased risk of breast cancer may be affected by ethnic differences in diet or other host characteristics. Total solar radiation was used in this study as a marker for a narrow segment of the ultraviolet spectrum (290-310 nanometres, wavelength) capable of producing vitamin- D in or on the skin.25 Certain tropospheric conditions and some kinds of air pollution can selectively block the portion of the spectrum responsible for vitamin D synthesis, while remaining transparent to other wavelengths of sunlight.2 No direct measures of ultraviolet radiation were available. The number of doctors per 1000 population was used in this study as a correlate of socioeconomic status. Additional measures of other socioeconomic factors would have been desirable, but are difficult to obtain. This measure of socioeconomic status was positively related to breast cancer incidence in the USSR, and indicates the presence of a socioeconomic gradient by latitude which is stronger than those reported in the US or other countries.3'28'29 Blot et al3 sought to determine if the 1.8-fold geographical differences in breast cancer mortality seen in the US resulted from confounding demographic variables, including socioeconomic and reproductive factors. They observed a positive association between breast cancer and socioeconomic status, but concluded that socioeconomic factors could not account for the geographical variation of breast cancer within the US. They found the major indicator of breast cancer mortality to be location of residence, when socioeconomic factors were simultaneously investigated. Similar analyses by Hems28 and MacMahon29 also led to the conclusion that while socioeconomic factors may be of importance, they were not sufficient to account for the geographical variation in breast cancer within the US or around the world.
20J3 M
824 10
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY 22
Canadian Cancer Society. Canadian Cancer Statistics 1987. Toronto, Canada, CCS, March 1987. SirverbergE,LuberaJA Cancer Statistics 1988. Ca. 1988; 38:5-27. " Mason T J, McKay F W, Hoover R, el al. Atlas of Cancer Mortality for U.S. Counties: 1950-69. Washington DC: US Government Printing Office, 1975 (DHEW Publication No. [N1H] 75-780). 23 Holick M F. In: Norman T, etal. eds. Vitamin D, Chemical and Clinical Update. New York, Walter de Gruyter, 1988. p 925-34. *Landsberg H E, Lippman H, Paffen K, et al. World Maps of Climatology, Second edition. New York, Springer-Verlag, 1965. 27 Napalkov N P, Tserkovny G F, Merabishvili V M, el al. Cancer Incidence in the USSR. (Supplement to Cancer Incidence in Five Continents, Volume 3). Lyon: International Agency for Research on Cancer, 1983 (IARC Scientific Publication No. 48), 59 a Hems G, Stuart A. Breast cancer rates in populations of single women. Br J Cancer 1975; 31: 118-23. B MacMahon B, Cole P, Brown J Etiology of breast cancer: a review. J Nail Cancer Insl 1973; 50: 21-42. "Pollard A P, ed. USSR Facts and Figures Annual, 1989. Vol. 13, London- Academic International Press, 1990; 143-^497. 31 Marcus R, Plimpton J. Guide to Japanese Food Tokyo: Shufuntomo, 1984, p 75-9 32 Lentner C, ed. Geigy Scientific Tables, Vol. I, West Caldwell, New Jersey, Ciba-Geigy, 1981; 259-60. 33 Rohan T E, Bain C J. Diet in the etiology of breast cancer. Epidemiol Rev 1987; 9: 12(M5. 34 Willet W C, Stampfer M J, Colditz G A, et al. Moderate alcohol consumption and risk of breast cancer N EnglJ Med 1987; 316: 1174-80. a
(Revised version received May 1990)
Downloaded from http://ije.oxfordjournals.org/ at Carleton University on June 22, 2015
Eisman J A, Mclntyre I, Martin T J, Frampton R J. Normal and malignant breast tissue is a target organ for 1,25- (OH) 2 vitamin D,. Clin Endocrinol 1980; 13: 267-72. " Eismaji J A 1,25-dihydroxyvitamin D3 receptors and the role of 1,25-(OH)2-D3 in human cancer cells. In: KumarR,ed. Vitamin D. Basic and Clinical Aspects. Boston, Martinus Nijhoff, 1984. p 365-84. u Eisman J A, Suva L J, Martin T J. Significance of 1,25-dihydroxyvitamin Dj receptors in primary breast cancer. Cancer Res 1986; 46: 5406-8. D Frampton R J, Suva L J, Eisman J A, Findlay D M, Moore G E, Moseley J M, Martin T J. Presence of 1,25-dihydroxyvitamin Dj receptors in established human cancer cell lines in culture. Cancer Res 1982; 42: 1116-9. "Frampton R J, Omond S A, Eisman J A. Inhibitions of human cancer cell growth by 1,25-dihydroxyvitamin Dj metabolites. Cancer Res 1983; 40: 4440-7. u Menck H R, Henderson B E. Cancer incidence rates in the Pacific basin. NCI Monogr 1978; 53: 119-24. " Gardner M J, Winter P D, Taylor C P, el al. Atlas of Cancer Mortality in England and Wales, 1968-1978. London, John Wiley and Sons, 1983; 24. " Canada, Minister of National Health and Welfare. Mortality Atlas of Canada. Hull, Quebec: Canadian Government Publishing Centre, 1980; Map 17: 60. 18 Thein-Hlaing, Thein-Maung-Myint. Risk factors for breast cancer in Burma Int J Cancer 1978, 21: 432-7. "Soini I. Risk factors of breast cancer in Finland Int J Epidemiol 1977; 6: 365-73. ^Tulinius H, Day N E, Johannesson G, el al. Reproductive factors and risk for breast cancer in Iceland. Int J Cancer 1978; 21: 724-30 21 Hirohata T, Shigematsu T, Nomura A M Y, et al. Occurrence of breast cancer in relation to diet and reproductive history: a case-control study in Fukuoka, Japan. Nail Cancer Insl Monogr 1985; 69: 187-90
