|Home | About | Journals | Submit | Contact Us | Français|
Evidence for a relation between physical activity and renal cell cancer has been inconsistent. The authors examined physical activity in relation to renal cell cancer in a large, prospective US cohort study of 482,386 participants (289,503 men and 192,883 women) aged 50–71 years at baseline (1995–1996). At baseline, participants reported their frequency of exercise of at least 20 minutes' duration, intensity of daily routine activity, and frequency of physical activity during adolescence. During 8.2 years of follow-up (through December 2003), 1,238 cases of renal cell cancer were ascertained. In multivariate Cox regression models adjusted for renal cell cancer risk factors, the authors observed that current exercise, routine physical activity, and activity during adolescence were associated with a reduced risk of renal cell cancer. The multivariate relative risks for the highest activity level as compared with the lowest were 0.77 (95% confidence interval (CI): 0.64, 0.92; ptrend = 0.10) for current exercise, 0.84 (95% CI: 0.57, 1.22; ptrend = 0.03) for routine physical activity, and 0.82 (95% CI: 0.68, 1.00; ptrend = 0.05) for activity during adolescence. The authors conclude that increased physical activity, including activity during adolescence, is associated with reduced risk of renal cell cancer.
Renal cell carcinoma is the seventh-leading incident cancer among US men and the 12th-leading cancer among US women (1). The incidence rates of renal cell cancer increase with age and are higher among men than among women (1). Established risk factors for renal cell cancer include smoking, high body weight, and high blood pressure (2).
Physical activity has been hypothesized to decrease the risk of renal cell cancer through reductions in body fat, blood pressure (3, 4), and concentrations of circulating growth factors (5). However, evidence that physical activity, or lack thereof, is related to renal cell cancer is inconclusive (6–17). Reasons for the inconsistent findings may include the relatively small sample sizes in existing prospective cohort studies and potential selection and recall bias in case-control studies. In addition, the inverse relation may be strongest for physical activity performed at younger ages (11), which few investigators have been able to examine. In order to address these issues, we examined renal cell cancer risk in a large, prospective cohort study of 482,386 AARP (formerly the American Association of Retired Persons) members with available data on physical activity at cohort entry and during adolescence.
Men and women in this study were enrolled in the National Institutes of Health (NIH)-AARP Diet and Health Study, the design of which has been previously described (18). Briefly, the cohort study was initiated in 1995–1996 when a baseline questionnaire eliciting information on usual dietary intake, physical activity, and other health-related behaviors was sent to 3.5 million AARP members aged 50–71 years who resided in one of six US states (California, Florida, Pennsylvania, New Jersey, North Carolina, and Louisiana) or two US metropolitan areas (Atlanta, Georgia, and Detroit, Michigan). A total of 617,119 persons responded, among whom 567,169 persons completed the questionnaire to a satisfactory degree. In late 1996, a supplementary questionnaire was mailed to participants who satisfactorily completed the baseline questionnaire, who still lived in the study area, and who did not have prevalent cancer of the colon, breast, or prostate. The supplementary questionnaire inquired about history of hypertension and weight at age 18 years, among other health-related questions. In total, 334,908 participants responded to the supplementary questionnaire.
Of the 567,169 respondents to the baseline questionnaire, we excluded participants who returned duplicate questionnaires (n = 179), who had died or moved out of the study area prior to baseline (n = 582), who withdrew from the study (n = 6), who had questionnaires completed by proxy respondents (n = 15,760), who had a previous diagnosis of cancer (n = 51,205) or end-stage renal disease (n = 985), or who were missing information on physical activity (n = 16,066). After these exclusions, data for 482,386 participants (289,503 men and 192,883 women) were available for analysis, including 298,246 persons who completed the supplemental questionnaire.
The NIH-AARP Diet and Health Study was approved by the Special Studies Institutional Review Board of the National Cancer Institute. All participants provided written informed consent.
Our exercise/sports assessment was based upon the 1990 recommendation by the American College of Sports Medicine that all persons engage in the equivalent of at least 20 minutes of vigorous exercise three times per week (19). On the baseline questionnaire, participants were asked to report the number of bouts of exercise and/or sports they had engaged in per month during the past year (i.e., current activity) that had lasted at least 20 minutes and had caused sweating or increased breathing or heart rate. Study members were also asked about the frequency of sports and/or exercise they had engaged in when they were 15–18 years of age. Participants selected their level of activity from six preestablished response options (never, rarely, 1–3 times per month, 1–2 times per week, 3–4 times per week, and five or more times per week). Participants in our cohort were also asked to select their level of activity during their current daily routine (including work routine) from five options (sitting; sitting and walking; standing or walking but not lifting or carrying things; carrying light loads or climbing stairs; or doing heavy work).
Our assessment of current exercise was similar to that of a previous questionnaire with demonstrated validity (percentage of agreement = 0.71; kappa = 0.40) based upon comparison with an objective measure (i.e., a computer science and applications activity monitor) (20). Our routine exercise measure resembled questions from the Baecke questionnaire that had previously been validated (21).
Incident cases of first primary renal cell cancer were identified by linkage of the NIH-AARP cohort database with the state cancer registries. In order to ascertain cases among AARP members who changed residence during the follow-up period, we also included linkage to three additional state cancer registries (Arizona, Texas, and Nevada) with large numbers of AARP members. Renal cell cancers were indicated by an International Classification of Diseases for Oncology code of C649 with a histology code consistent with renal cell cancer (8010, 8032, 8140, 8211, 8246, 8260, 8310, 8312, or 8320). Participants were followed from the date of scanning of the baseline questionnaire to the date of diagnosis of first cancer, death, moving out of the cancer registry ascertainment area, or the end of follow-up (December 31, 2003). In a previous validation study, the estimated sensitivity of cancer identification was approximately 90 percent and the specificity was 99.5 percent (22).
Relative risks and 95 percent confidence intervals were estimated using Cox proportional hazards regression models. We collapsed the bottom two categories of exercise and/or sports and activity during adolescence to ensure sufficient numbers of cases in the reference category. For tests of trend, each category was assigned a single value indicating the approximate frequency of physical activity per week (never/rarely = 0.125, 1–3 times per month = 0.5, 1–2 times per week = 1.5, 3–4 times per week = 3.5, and ≥5 times per week = 5.5). For routine activity, the five ordered categories were assigned values of 1, 2, 3, 4, and 5. We constructed linear trend variables based upon the assigned values for each variable and tested their statistical significance in regression models.
Covariates were included in multivariate models if previous studies consistently indicated an association with renal cell cancer or if the covariate was a statistically significant predictor of renal cell cancer in the NIH-AARP Diet and Health cohort. All multivariate models were adjusted for age, sex, body mass index (weight (kg)/height (m)2), height, race/ethnicity, smoking, history of diabetes, protein intake, and history of hypertension. For covariates for which data were missing, we used missing indicator variables. For participants who did not complete the supplementary questionnaire, history of hypertension was modeled as a missing indicator variable. In analyses of the subset of participants who completed the supplementary questionnaire, neither hypertension nor body mass index at age 18 years was a confounder of the relation between physical activity and renal cell cancer. Further adjustment for education, number of children, intakes of red meat, fat, and alcohol, and use of multivitamins did not alter the relation of physical activity to renal cell cancer. All multivariate models met the proportional hazards assumption.
We formally tested for potential interactions with the association between physical activity and renal cell cancer using the likelihood ratio test—that is, comparing the likelihood of models with and without multiplicative interaction terms. Interaction terms were calculated using the cross-product of the physical activity categories and the factor of interest (e.g., sex).
Statistical analyses were performed using the Statistical Analysis System, release 9.1.3 (SAS Institute, Inc., Cary, North Carolina). All p values were based upon two-sided tests.
During up to 8.2 years of follow-up, we ascertained 1,238 cases of renal cell cancer (929 in men and 309 in women). As compared with participants who never/rarely engaged in physical activity, participants who engaged in high levels of physical activity were more likely to be male, to be of White race/ethnicity, to have a lower body mass index, and to be free from diabetes (table 1). These findings were generally similar regardless of whether we were examining exercise/sports, routine activity, or activity during adolescence. However, the pattern of participant characteristics differed in certain respects between exercise/sports and routine activity. Frequent exercisers were more likely to be White and to be never or former smokers than persons who never/rarely engaged in exercise or sports. In contrast, participants whose routine activity involved heavy work were more likely to be current smokers but not more likely to be White than persons whose routine involved mostly sitting. The pairwise correlation coefficients of the three activity measures were 0.24 (exercise/sports and routine activity), 0.19 (exercise/sports and activity during adolescence), and 0.06 (routine activity and activity during adolescence).
Study members who engaged in a greater amount of physical activity had a reduced risk of developing renal cell cancer. Persons who engaged in exercise/sports five or more times per week had a 23 percent reduced risk (multivariate relative risk (RR) = 0.77, 95 percent confidence interval (CI): 0.64, 0.92; p for trend = 0.10) as compared with participants who never or rarely engaged in exercise/sports (table 2). Participants who engaged in heavy work during their daily routine had an approximately 16 percent reduced risk (multivariate RR = 0.84, 95 percent CI: 0.57, 1.22; p for trend = 0.03) as compared with those who sat most of the day. Cohort members who had engaged in physical activity five or more times per week during adolescence had an 18 percent reduced risk (multivariate RR = 0.82, 95 percent CI: 0.68, 1.00; p for trend = 0.05) of renal cell cancer in comparison with study members who were inactive as adolescents. In multivariate models, adjusting for body mass index modestly attenuated the relative risks, but no other adjustments substantially affected the estimated relation between physical activity and renal cell cancer (i.e., no other adjustments resulted in a change in beta coefficients of 10 percent or more). Mutual adjustment for baseline exercise/sports, daily routine activity, and activity during adolescence modestly attenuated the observed associations for each measure of activity (e.g., for current exercise/sports, the 23 percent risk reduction became a 19 percent risk reduction and the p value for trend increased from 0.10 to 0.31), indicating some interdependence between activity measures.
In a secondary analysis, we examined the relation between physical activity and renal cell cancer among only those participants who completed the supplementary questionnaire and reported their history of hypertension (approximately 60 percent of the overall cohort). In this subcohort, we observed similar associations between physical activity and renal cell cancer risk, although reductions in risk were of slightly greater magnitude than the reduction observed in the entire cohort. In this subcohort, the multivariate relative risks for frequent exercise/sports (five or more times per week), a high level of routine activity (heavy work), and frequent activity during adolescence (five or more times per week) were 0.74, 0.73, and 0.80, respectively, as compared with 0.77, 0.84, and 0.82 in the entire cohort.
We also conducted analyses stratified according to whether or not participants were overweight (body mass index ≥ 25.0) but did not find that the relation between physical activity and renal cell cancer varied by baseline body mass index (all p's for interaction > 0.05; table 3). Qualitatively, physical activity was associated with greater reductions in renal cell cancer risk among participants who were normal-weight (body mass index < 25) than among those who were overweight, but the difference was not sufficient to reach statistical significance.
To our knowledge, no previous prospective studies have examined the association of physical activity during adolescence with renal cell cancer. Thus, we further explored this relation in several subgroups of potential interest (table 4). Among women, physical activity during adolescence had a graded dose-response relation with renal cell cancer, with a 38 percent reduction in risk being seen among women who were active five or more times per week. Age at menarche neither confounded nor modified this relation (p for interaction = 0.56), indicating that physical activity probably acts through a mechanism other than delayed exposure to estrogens to reduce renal cell cancer risk. Among men, physical activity during adolescence was not associated with reduced risk of renal cell cancer. Overall, however, a test for effect modification did not reveal statistically significant differences by sex (p for interaction = 0.28). Similarly, in analyses stratified by age, smoking status, or history of hypertension, we did not uncover any evidence of important effect modification (all p's for interaction > 0.05).
With respect to current exercise/sports and routine activity, we found little evidence that the observed associations varied according to age, smoking status, or history of hypertension (all p's for interaction > 0.05). However, men who engaged in routine physical activity had a reduced risk of renal cell cancer (for heavy work, multivariate RR = 0.78, 95 percent CI: 0.52, 1.18; p for trend = 0.04), whereas women did not (multivariate RR = 0.96, 95 percent CI: 0.33, 2.77; p for trend = 0.42; p for interaction = 0.01). No further statistically significant interactions were observed by sex. For example, men who engaged in exercise/sports five or more times per week had a 23 percent reduced risk of renal cell cancer (multivariate RR = 0.77, 95 percent CI: 0.62, 0.95), which was comparable to the 24 percent reduction observed in women who frequently exercised (multivariate RR = 0.76, 95 percent CI: 0.51, 1.11; p for interaction = 0.28).
In this large prospective cohort study of nearly 500,000 AARP members, we investigated three measures of physical activity—current exercise/sports, routine activity, and activity during adolescence—in relation to renal cell cancer. We found that each activity measure was related to reduced risk of renal cell cancer, with a borderline statistically significant association for exercise/sports and statistically significant associations for routine activity and activity during adolescence. Depending on the specific type of activity assessed, physical activity was associated with risk reductions ranging from 16 percent (for heavy work during daily routine) to 23 percent (for frequent exercise/sports at baseline) relative to inactivity. The associations between physical activity and renal cell cancer appeared to be more pronounced in persons who were normal-weight than in persons who were overweight. This may indicate that physical activity is unable to compensate for the adverse consequences of excess weight with respect to renal cell cancer risk. Alternatively, normal-weight persons may recall physical activity with greater accuracy.
To our knowledge, our study was the first prospective cohort investigation of the association between physical activity during adolescence and renal cell cancer. Among women but not among men, we found a graded inverse relation between activity during adolescence and renal cell cancer. Possible explanations for our findings include that physical activity may inhibit development of potential renal cell cancer precursors such as renal injury (23, 24) or that adolescent physical activity acts as a proxy for activity undertaken during a longer span of time, perhaps including the early years of adulthood.
Several biologic mechanisms have been proposed by which increased physical activity may be linked to lower renal cell cancer risk. In randomized controlled trials, physical activity has been demonstrated to reduce blood pressure (25, 26), a well-established risk factor for renal cell cancer (27). Exercise may also reduce lipid peroxidation (28), a process which may increase the frequency of DNA mutations in renal cells and hence induce renal cell cancer (28). Physical activity is associated with a decrease in serum insulin levels (5). Low insulin levels result in slower proliferation of renal cell cancer in vitro (29). Lower levels of insulin-like growth factor 1 have also been associated with decreased risk of renal cell cancer. Physical activity, through its insulin-reducing effects, may increase the concentration of insulin-like growth factor 1 binding proteins and thereby reduce the pool of bioactive insulin-like growth factor 1 (30). Physical activity may act through its correlate, low body weight, to promote renal health by decreasing oxidative stress (31) and renal atherosclerosis (32).
Despite biologic plausibility, previous prospective cohort studies have not consistently found that physical activity is related to reduced renal cell cancer risk. Reported findings have been as wide-ranging as increased risk (9), no association with risk (13), and reduced risk (10, 12, 17) of renal cell cancer with increasing activity level. In the largest of these prospective cohort studies, comprising 347 cases, physical activity was inversely related to renal cell cancer among women but not among men (17). In the second-largest prospective cohort study, comprising 275 cases, a high level of leisure-time physical activity was related to a nonsignificant reduced risk of renal cell cancer in men (multivariate RR = 0.74, 95 percent CI: 0.44, 1.23) but not in women (multivariate RR = 1.13, 95 percent CI: 0.56, 2.29) (13). Previous prospective cohort studies included relatively few cases, particularly in analyses stratified according to sex, which may have contributed to the inconsistency of their findings.
Case-control studies have reported similarly varying relations between physical activity and renal cell cancer (7, 11, 14–16). Case-control studies, however, are prone to selection and recall biases, which may explain their inconsistent results. In the case-control studies that found statistically significant associations, an increasing level of physical activity was associated with reductions in renal cell cancer risk ranging from 34 percent (17) to 54 percent (10). With respect to activity during adolescence, two case-control studies found a reduced risk of renal cell cancer with increasing adolescent activity level (11, 16), but another found no such link (7).
A major strength of our study is that it included a substantial number of renal cell cancer cases, thereby allowing us to detect a potentially modest association. The large sample size also allowed us to conduct exploratory analyses stratified by sex, age, race/ethnicity, smoking status, body mass index, and history of hypertension with reasonable statistical power. We were also able to examine multiple physical activity domains (i.e., current exercise/sports, activity during one's daily routine, and activity during adolescence). Additional strengths of our study included the prospective design and the availability of data on many important potential confounders.
Our study also had several important limitations. Because our study was observational, we cannot exclude the possibility that residual confounding explains some or all of the observed reduced risk of renal cell cancer associated with physical activity. In particular, we adjusted for self-reported body mass index as a proxy for adiposity. Consequently, incomplete adjustment for adiposity may partly explain our findings. Our assessments of physical activity, while incorporating elements of previously validated instruments (21), were based on self-reports and may have resulted in some degree of misclassification of physical activity levels. Our question concerning routine activity may not have captured activity arising from housework. Since women commonly engage in housework, this may explain why we observed weaker relations with routine activity among women in our study. Because members of our study were at least 50 years of age at baseline, our inquiry into activity during adolescence required participants to recall activity from a period approximately 30–40 years in the past. Such distant recall may result in imprecise assessment and could result in error in the estimation of relative risks.
We also lacked information on blood pressure history for many of the participants in our study. High blood pressure is a known risk factor for renal cell cancer (27) and is also associated with physical activity and therefore may be a confounding factor. However, in a secondary analysis of only those participants who self-reported data on history of high blood pressure, we found that hypertension neither confounded nor modified the association between physical activity and renal cell cancer.
In conclusion, our study demonstrates that physical activity, including activity undertaken during adolescence, is associated with a reduced risk of renal cell cancer. Our results contribute to the accumulating evidence that physical inactivity predisposes people to cancer and further support the conclusion that physical activity is an important target for cancer prevention efforts.
This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, and by grant TU2CA105666.
Cancer incidence data from the Atlanta, Georgia, metropolitan area were collected by the Georgia Center for Cancer Statistics (Department of Epidemiology, Rollins School of Public Health, Emory University). Cancer incidence data from California were collected by the Cancer Surveillance Section of the California Department of Health Services. Cancer incidence data from the Detroit, Michigan, metropolitan area were collected by the Michigan Cancer Surveillance Program (Community Health Administration, State of Michigan). The Florida cancer incidence data used in this report were collected by the Florida Cancer Data System under contract with the Department of Health. (The views expressed herein are solely those of the authors and do not necessarily reflect those of the contractor or the Department of Health.) Cancer incidence data from Louisiana were collected by the Louisiana Tumor Registry (Louisiana State University Medical Center in New Orleans). Cancer incidence data from New Jersey were collected by the New Jersey State Cancer Registry (Cancer Epidemiology Services, New Jersey State Department of Health and Senior Services). Cancer incidence data from North Carolina were collected by the North Carolina Central Cancer Registry. Cancer incidence data from Pennsylvania were supplied by the Division of Health Statistics and Research of the Pennsylvania Department of Health. (The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.)
Conflict of interest: none declared.