Lifetime physical activity and the risk of renal cell cancer
Article first published online: 31 JAN 2007
Copyright © 2006 Wiley-Liss, Inc.
International Journal of Cancer
Volume 120, Issue 9, pages 1977–1980, 1 May 2007
How to Cite
Tavani, A., Zucchetto, A., Maso, L. D., Montella, M., Ramazzotti, V., Talamini, R., Franceschi, S. and La Vecchia, C. (2007), Lifetime physical activity and the risk of renal cell cancer. Int. J. Cancer, 120: 1977–1980. doi: 10.1002/ijc.22438
- Issue published online: 28 FEB 2007
- Article first published online: 31 JAN 2007
- Manuscript Accepted: 6 OCT 2006
- Manuscript Received: 26 JUL 2006
- Italian Association for Cancer Research
- Italian League against Cancer
- Italian Ministry of Education. Grant Number: PRIN 2005
- case-control studies;
- physical activity;
- renal cell cancer;
- risk factors
The relation between lifelong physical activity at work and during leisure-time and the risk of renal cell cancer (RCC) was analyzed in a case-control study conducted in Italy between 1992 and 2004. Cases were 767 subjects with incident, histologically confirmed RCC, and controls were 1,534 patients hospitalized for acute nonneoplastic conditions. Odds ratios (OR) and 95% confidence intervals (CI) for RCC were computed by multiple logistic regression models, conditioned on study center, sex and age, and adjusted for main covariates. Compared to the lowest level of occupational physical activity, the multivariate OR of RCC for the highest level were 0.65 (95% CI 0.49–0.87) at age 12 years, 0.67 (95% CI 0.53–0.84) at age 15–19, 0.74 (95% CI 0.59–0.93) at age 30–39 and 0.71 (95% CI 0.55–0.92) at age 50–59 years, with significant inverse trends in risk. The inverse association was consistent in strata of sex, age at diagnosis, body mass index, smoking habit and alcohol drinking. No significant association was found for leisure-time physical activity. The inverse association between occupational physical activity and RCC risk, if real, may be related to the effects of insulin-like growth factors, or lipid peroxidation and about 9% of cases of RCC in Italy could be avoided by increasing physical activity. However the inverse association might involve confounding by indirect mechanisms, such as body composition or other social class correlates. © 2006 Wiley-Liss, Inc.
Physical exercise is an important correlate of cardiovascular disease and several neoplasms.1 Epidemiological evidence on the relation of physical activity and renal cell cancer (RCC) risk is scanty and mostly based on small studies, and the results are inconclusive. Out of 3 cohort studies considering either occupational or leisure-time exercise, a Swedish one suggested no relation with both.2 The Netherlands Cohort Study on Diet and Cancer found an inverse, but nonsignificant association with either occupational or leisure-time exercise,3 and the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study found an inverse association for leisure-time physical activity, but not for occupational one.4 Of 2 American case-control studies, 1 found no association,5 and the other found an inverse association with recent recreational activity in either sexes.6 However, the International Renal Cell Cancer Study, a multicenter case-control study including 1,732 cases, found no association in either sex with either occupational or leisure-time exercise at 20 or 40 years.7
We analyzed the role of physical activity at various ages on RCC risk, using data from a large case-control study conducted in Italy.
Material and methods
A case-control study of RCC was conducted from 1992 to 2004 in 4 Italian areas: the provinces of Gorizia and Pordenone and the greater Milan area in the North Italy, the provinces of Latina and Naples in the centre-South Italy.8 Cases were 767 patients (494 men and 273 women, median age 62 years, range 24–79) with incident, histologically confirmed RCC, admitted to major teaching and general hospitals of the study area. Controls were 1,534 subjects (988 men and 546 women, median age 62 years, range 22–79) admitted to the same hospitals of cases for a wide spectrum of acute nonneoplastic conditions, not related with smoking, urinary or genital tract diseases and conditions which can cause a modifications in diet. Controls were frequency matched with cases by 5-year age groups, sex and study center. Among controls, 26% were admitted for traumas, 32% for other orthopaedic disorders, 14% for surgical conditions and 28% for other illnesses.
Trained interviewers questioned cases and controls in the hospital setting using a structured questionnaire. The questionnaire collected information on socio-demographic characteristics, anthropometric measures, lifestyle habits, including alcohol drinking and cigarette smoking, medical history, and family history of cancer in first-degree relatives. The subjects' diet was assessed using an interviewer-administered food frequency questionnaire, including the weekly frequency of consumption of 78 foods and food groups and beverages.9, 10
The section on physical activity included questions on self-reported levels of activity at work and during leisure-time.11 Interviewers asked patients how they would describe the level of physical activity at age 12, 15–19, 30–39 and 50–59 years. For occupational physical activity, patients were asked whether their jobs were “very heavy,” “heavy,” “average,” “standing,” “mainly sitting,” according to a structured scale classified into 5 categories (scored between 1 and 5). The scores of the 2 highest and the 2 lowest level of occupational physical activity were combined, as there were too few subjects in the categories with the highest and the lowest activity. Physical activity during leisure-time was defined on the basis of the number of hours per week. The 4 levels considered corresponded to <2, 2–4, 5–7 and >7 hr of physical activity per week. No information was available on the intensity or amount of activity.
Odds ratios (OR) and the corresponding 95% confidence intervals (CI) were derived by multiple logistic regression models conditioned on study center, sex and age (quinquennia), and adjusted for year of interview (<1998, ≥1998), smoking habit (never, ex since at least 4 year, current smokers of <20, ≥20 cigarettes/day), body mass index 1 year before diagnosis/interview (<25, 25 to <30, ≥30 kg/m2) and history of treated hypertension (ever/never),12 i.e., the major covariates associated with RCC in our study.8
The distribution of cases and controls, and the corresponding multivariate ORs of RCC according to occupational and leisure-time physical activity at different ages, are given in Table I. Compared to subjects in the lowest level of occupational physical activity, the ORs of RCC in the highest level was 0.65 at age 12, 0.67 at age 15–19, 0.74 at age 30–39 and 0.71 at age 50–59 years. The trends in risk were significant. The corresponding ORs after further adjustment for education (<12 years/≥12 years) were 0.81 (95% CI 0.63–1.03, p = 0.08) at age 15–19 years, 0.85 (95% CI 0.67–1.07, p = 0.16) at age 30–39 years, and 0.80 (95% CI 0.61–1.04, p = 0.09) at age 50–59 years. The ORs for leisure-time physical activity were close to unity at all the considered ages.
|Selected ages||Cases||Controls||OR [95% CI]|
|Occupational physical activity|
|At age 12 years2|
|Low||498 (65.2)||927 (60.6)||13|
|Medium||177 (23.2)||359 (23.5)||0.88 [0.70–1.10]|
|High||89 (11.7)||243 (15.9)||0.65 [0.49–0.87]|
|χ2 trend||8.46 (p < 0.01)|
|At age 15–19 years2|
|Low||359 (46.9)||602 (39.3)||15|
|Medium||208 (27.2)||452 (29.5)||0.75 [0.60–0.94]|
|High||198 (25.9)||477 (31.2)||0.67 [0.53–0.84]|
|χ2 trend||12.54 (p < 0.01)|
|At age 30–39 years4|
|Low||283 (37.2)||471 (31.0)||13|
|Medium||258 (33.9)||552 (36.3)||0.80 [0.64–1.00]|
|High||220 (28.9)||498 (32.7)||0.74 [0.59–0.93]|
|χ2 trend||6.73 (p < 0.01)|
|At age 50–59 years5|
|Low||266 (41.6)||455 (35.9)||13|
|Medium||223 (34.8)||463 (36.5)||0.83 [0.65–1.04]|
|High||151 (23.6)||350 (27.6)||0.71 [0.55–0.92]|
|χ2 trend||6.75 (p < 0.01)|
|Leisure-time physical activity (hours/week)|
|At age 12 years2|
|<2||275 (35.9)||575 (37.5)||13|
|2–4||127 (16.6)||283 (18.5)||1.01 [0.77–1.32]|
|≥5||364 (47.5)||674 (44.0)||1.24 [0.99–1.55]|
|χ2 trend||3.79 (p = 0.05)|
|At age 15–19 years2|
|<2||327 (42.7)||663 (43.3)||13|
|2–4||166 (21.7)||305 (19.9)||1.14 [0.89–1.46]|
|≥5||273 (35.6)||564 (36.8)||1.02 [0.82–1.28]|
|χ2 trend||0.04 (p = 0.84)|
|At age 30–39 years4|
|<2||471 (61.9)||933 (61.3)||13|
|2–4||143 (18.8)||306 (20.1)||0.95 [0.75–1.20]|
|≥5||147 (19.3)||283 (18.6)||1.04 [0.81–1.34]|
|χ2 trend||0.04 (p = 0.85)|
|At age 50–59 years5|
|<2||446 (69.8)||858 (67.7)||13|
|2–4||89 (13.9)||215 (17.0)||0.79 [0.59–1.05]|
|≥5||104 (16.3)||195 (15.4)||1.03 [0.78–1.36]|
|χ2 trend||0.07 (p = 0.79)|
The inverse association for RCC risk with occupational physical activity at 30–39 years was consistent in strata of sex, age at diagnosis (<60/≥60 years), body mass index at 30–39 years and 1 year before diagnosis (<25/≥25 kg/m2), smoking habit (nonsmokers/current smokers) and alcohol drinking (nondrinkers/drinkers) (Table II).
|Level of physical activity||Medium||High||χ2 trend|
|Men||0.79 [0.59–1.04]||0.72 [0.55–0.94]||5.94 [p = 0.01]|
|Women||0.85 [0.60–1.22]||0.85 [0.54–1.34]||0.59 [p = 0.44]|
|Age at diagnosis/interview [years]|
|<60||0.73 [0.52–1.03]||0.75 [0.52–1.07]||2.91 [p = 0.09]|
|≥60||0.85 [0.64–1.13]||0.75 [0.55–1.01]||3.63 [p = 0.06]|
|Body mass index at 30–39 years [kg/m2]|
|<25||0.77 [0.59–1.01]||0.79 [0.59–1.04]||2.89 [p = 0.09]|
|≥25||0.90 [0.58–1.39]||0.71 [0.45–1.12]||2.16 [p = 0.14]|
|Body mass index 1 year before diagnosis/interview [kg/m2]|
|<25||0.68 [0.47–0.98]||0.68 [0.46–1.01]||3.97 [p = 0.05]|
|≥25||0.84 [0.64–1.12]||0.73 [0.55–0.98]||4.39 [p = 0.04]|
|Never smokers||0.81 [0.57–1.16]||0.72 [0.49–1.07]||2.74 [p = 0.10]|
|Ever smokers||0.82 [0.61–1.09]||0.79 [0.59–1.05]||2.70 [p = 0.10]|
|Alcohol drinking habit|
|Never drinkers||0.81 [0.47–1.41]||0.56 [0.27–1.15]||2.44 [p = 0.12]|
|Ever drinkers||0.77 [0.60–0.98]||0.75 [0.58–0.96]||5.40 [p = 0.02]|
We found an inverse association between occupational physical activity and the risk of RCC. The strength of the association was lowered after further adjustment for education; however, this represents an overadjustment, as occupational physical activity is highly correlated to education. The inverse association with occupational, but not with leisure-time activity, may involve confounding by other social class correlates. However, the inverse association cannot be attributed to a general healthier lifestyle, as occupational exercise is directly related with lower social classes and education. Obesity and hypertension are established risk factors for renal cell carcinoma, and both can be associated with physical inactivity,13 but adjustment for these 2 covariates did not materially change the risk estimates. When only centre, sex and age were included in the model the OR for occupational physical activity at 30–39 years were 0.77 (95% CI 0.62–0.96) for medium, and 0.72 (95% CI 0.58–0.91) for high activity.
Overall epidemiological evidence on the relation between physical activity and the risk of RCC is controversial. Among studies finding an inverse relation, most ones found it with leisure-time exercise3, 4, 6 and only one found a nonsignificant inverse relation with occupational physical activity.3 The reasons for these inconsistencies can be attributed, at least in part, to methodological issues, including, difficulties in measurements of physical activity, lack of longitudinal data on lifetime exercise and inadequate adjustment for confounding.1 However, the lack of association between RCC risk and leisure-time physical activity in this population can be attributable to its low level and frequency in Italy, where regular recreational exercise is a relatively recent phenomenon, mainly restricted to younger people and highest social class. Leisure time physical activity is more difficult to define and quantify than occupational activity, that is practiced over more extended periods during life.14, 15 Moreover, occupational activity was defined by intensity as well as frequency, while leisure-time activity was defined only by frequency. If the association with RCC depends on the level of intensity, then the measure employed for leisure activity would have missed effect.
Our study was not population-based, but cases and controls were interviewed in the same hospital setting, came from comparable catchment areas and participation was almost complete. We excluded from the control group any patient admitted for chronic conditions, particularly those associated with any disorder of the genital or urinary tract. When controls admitted to hospital for traumatic orthopedic conditions were excluded from the analyses, the risk estimates were not materially different (ORs for occupational physical activity at age 30–39 years were 0.82, 95% CI 0.65–1.03 for medium, and 0.77, 95% CI 0.61–0.98 for high activity). The similar interview setting and the consistency of the association across strata of several covariates provides further reassurance against potential information bias,16 and there is no reason to assume different recall on the basis of the disease status. Finally, the estimated level of physical activity was consistent with information on occupation. In our study, information on physical activity was neither validated nor quantified in terms of energy expenditure. However, it has been shown that even simple questions are able to provide useful information.17
There are various mechanisms whereby physical activity may protect against RCC. Physical activity has been reported to decrease levels of estrogens,18, 19 which have been shown to induce kidney cancer in animal models.20, 21 Physical activity might also decrease plasma levels of insulin and insulin-like growth factor-I (IGF-I),22 both of which have been shown to have promoting effects on renal carcinogenesis.21, 23 Another proposed mechanism is the reduction of lipid peroxidation, which, in turn, has been shown to react with renal cell DNA to form adducts, which may lead to mutations.24 Other mechanisms relate to the link between physical inactivity, specific dietary patterns (i.e., a diet rich in refined carbohydrates, fat, dairy foods and high energy intake), propensity to overweight and obesity, which may lead to higher risk. Allowance for body mass index, however, did not appreciably modify the inverse association. However, none of these mechanisms is clearly established.
The attributable risk of low occupational physical activity during adulthood (30–39 years) in this population was 9% (95% CI 3–15%) of cases, suggesting that in Italy increasing physical exercise may have a relevant role in the prevention of RCC.
The authors thank Ms. M.P. Bonifacino for editorial assistance, Ms. O. Volpato and M. Cozzi for study coordination, and Ms. G. Bessega and L. Zaina, Drs. M. Grimaldi and O. Manganelli for their help in data collection. We are deeply thankful to Drs. A Garbeglio and D. Maruzzi for his support in identifying cancer cases, and to Drs. P. Ascierto, G. Chiara, R. Di Lauro, L. Forner, A. Grandi, R. Magri, A. Mele, G. Tosolini and E. Trevisanutto, for providing hospital control patients.
- 12Statistical methods in cancer research, vol. I: The analysis of case-control studies. Lyon, France: IARC Scientific Publication, 1980; 5–338., .