Abstract

The impact of total physical activity level on cancer risk has not been fully clarified, particularly in non-Western, relatively lean populations. The authors prospectively examined the association between daily total physical activity (using a metabolic equivalents/day score) and subsequent cancer risk in the Japan Public Health Center-based Prospective Study. A total of 79,771 general-population Japanese men and women aged 45–74 years who responded to a questionnaire in 1995–1999 were followed for total cancer incidence (4,334 cases) through 2004. Compared with subjects in the lowest quartile, increased daily physical activity was associated with a significantly decreased risk of cancer in both sexes. In men, hazard ratios for the second, third, and highest quartiles were 1.00 (95% confidence interval (CI): 0.90, 1.11), 0.96 (95% CI: 0.86, 1.07), and 0.87 (95% CI: 0.78, 0.96), respectively (p for trend = 0.005); in women, hazard ratios were 0.93 (95% CI: 0.82, 1.05), 0.84 (95% CI: 0.73, 0.96), and 0.84 (95% CI: 0.73, 0.97), respectively (p for trend = 0.007). The decreased risk was more clearly observed in women than in men, especially among the elderly and those who regularly engaged in leisure-time sports or physical exercise. By site, decreased risks were observed for cancers of the colon, liver, and pancreas in men and for cancer of the stomach in women. Increased daily physical activity may be beneficial in preventing cancer in a relatively lean population.

A number of investigators have reported beneficial effects of physical activity on the risk of cancer at certain specific sites, and physical activity is now regarded as an important target for cancer prevention. The second report of the World Cancer Research Fund/American Institute for Cancer Research recently concluded that all forms of physical activity protect against some cancers, including colon cancer, postmenopausal breast cancer, and endometrial cancer, in relation to or independently of weight gain, overweight, and obesity (1).

To date, however, the association between physical activity and total cancer risk has been relatively poorly investigated. Given that exercise and physical activity probably affect cancer development at different sites via the same mechanism or closely similar mechanisms, at least to some degree, it is reasonable to assess the preventive effect of physical activity not only on cancer at specific sites but also on all cancers in aggregate. Further, from a public health point of view, an understanding of the preventive effect of physical activity on total cancer risk will provide concrete clues in estimating the effect of physical activity measures in health policy planning. For the latter case, evidence from populations with similar general lifestyle backgrounds is indispensable. Evidence for an association between physical activity and total cancer risk is limited (2–10), however; most studies have targeted mortality (4–10) rather than incidence (2, 3) and have been carried out in Western populations (2–8). Evidence from other populations is sparse (9, 10).

Here, we examined the association between daily total physical activity and risk of all types of cancer in the Japan Public Health Center-based Prospective Study. Our main purpose was to estimate the magnitude of the effect of overall physical activity, including exercise and nonexercise physical activities, on total cancer risk among Japanese, a population characterized as non-Western and relatively lean. To date, physical activity has been assessed using various types of activity categories, such as leisure-time and non-leisure-time activity, physical exercise or sports, and nonexercise activities, such as occupational activity and household work. However, given recognition of the need for comprehensive evaluation of these physical activities in the aggregate, particularly with regard to nonexercise physical activity (11), here we attempted a quantitative approach to assessment using a common scale for all activities (namely, metabolic equivalents (METs)) to estimate the effect of total physical activity level.

MATERIALS AND METHODS

Study population

The Japan Public Health Center-based Prospective Study was started in 1990–1994. It targeted all registered Japanese inhabitants in 11 public health center areas who were aged 40–69 years at the beginning of the baseline survey (12).

The study protocol was approved by the institutional review board of the National Cancer Center, Japan. In the present analysis, one public health center area was excluded, since data on cancer incidence were not available.

The participants in the present study were subjects in the Japan Public Health Center-based Prospective Study who responded to a self-administered 5-year follow-up questionnaire in 1995–1999 at age 45–74 years. Initially, at baseline, 133,323 subjects were identified as being in the study population. After excluding 241 persons with non-Japanese nationality (n = 51), duplicate enrollment (n = 4), a late report of emigration occurring before the start of the follow-up period (n = 180), or ineligibility due to an incorrect birth date (n = 6), a population-based cohort of 133,082 subjects was established. After exclusion of the 13,663 persons who had died, moved out of the study area, or been lost to follow-up before the starting point, the remaining 119,419 subjects were considered eligible for the present study. A total of 96,566 subjects responded to the questionnaire, yielding a response rate of 81 percent.

Questionnaire

The questionnaire included items on demographic factors, personal medical history, physical activity, smoking and alcohol drinking, other lifestyle factors, and diet (via a validated food frequency questionnaire containing questions on 138 food items and 14 supplementary questions (13)). Persons who had been diagnosed with cancer before the starting point (n = 2,153) or who had missing data for physical activity-related factors (n = 6,346) or other factors included in the multivariate model (n = 8,296) were excluded. Finally, 79,771 eligible subjects (37,898 men and 41,873 women) were included in the analysis.

Follow-up

Subjects were followed from the starting point until December 31, 2004. Residence status, including survival, was confirmed through the residential registry. Inspection of the resident registry is available to anyone under the resident registration law. Among the study subjects, 5,271 died, 3,166 moved out of the study area, one withdrew from the study, and 239 (0.3 percent) were lost to follow-up within the follow-up period. Information on the cause of death for deceased subjects was obtained from death certificates (provided by the Ministry of Health, Labour, and Welfare with the permission of the Ministry of Internal Affairs and Communications), on which cause of death is defined according to the International Classification of Diseases, Tenth Revision (14). Resident registration and death registration are required by law in Japan, and the registries are believed to be complete.

Incident cancers were identified through notification from the major hospitals in the study area and through data linkage with population-based cancer registries. Death certificates were used as a supplementary information source. The site and histology of each case were coded using the International Classification of Diseases for Oncology, Third Edition (15). In our cancer registry system, the proportion of cases for which information was available from death certificates only was 3.7 percent. For the present analysis, the earliest date of diagnosis was used in cases with multiple primary cancers diagnosed at different times. A total of 4,334 newly diagnosed cancer cases were identified.

Physical activity levels

The main exposure of interest in the present study was daily total physical activity level. In our questionnaire (see Appendix), subjects were asked about the average amount of time spent per day in three types of physical activity: heavy physical work or strenuous exercise (none, <1 hour, or ≥1 hour), sitting (<3, 3–<8, or ≥8 hours), and standing or walking (<1, 1–<3, or ≥3 hours). The following values were assigned as time scores for each activity: heavy physical work or strenuous exercise—0 for none, 0.5 for <1 hour, and 3 for ≥1 hour; sitting—1.5 for <3 hours, 5.5 for 3–<8 hours, and 7.5 for ≥8 hours; standing or walking—0.5 for <1 hour, 2 for 1–<3 hours, and 8.5 for ≥3 hours. The midpoint of the time range for each category was assigned when minimum and maximum values were presented on the questionnaire, and arbitrary values considered to have the highest validity from the validation study were assigned for the highest category. MET-hours/day were estimated by multiplying the daily time score for each activity by the MET intensity of that activity (16): for heavy physical work or strenuous exercise, 4.5; for standing or walking, 2.0; for being sedentary, 1.5; and for sleep or other passive activity, 0.9. After data were summed across all activities, subjects were grouped by sex into four exposure levels according to quartile of total METs/day score. Because the question on MET calculation incorporated all activities, including occupation, housework, leisure-time sports, etc., a separate question on the frequency of leisure-time sports and physical exercise was not included in the estimation of total physical activity level.

The validity of the total METs/day score was assessed among 108 eligible samples (53 men and 55 women) derived from 110 original volunteer subjects from the cohort using 4-day, 24-hour physical activity records (Sunday or another day off plus three weekdays) in two different seasons (namely, harvesting and one other season in a single year). The mean number of total METs/day for physical activity obtained from the self-report was 33.5 in men and 33.4 in women, while the mean from the 24-hour physical activity record was 39.5 in men and 40.8 in women. Energy expenditure estimated in METs showed little difference by area. Spearman's rank correlation coefficient for the correlation between the total METs/day score and the physical activity records was 0.46 when the average of two seasons was taken (men, 0.53; women, 0.35).

Analysis

The number of person-years in the follow-up period was counted from the starting point (i.e., the date of response to the 5-year follow-up questionnaire) to the date of occurrence of any cancer, emigration from the study area, death, or the end of the study period, whichever came first. For subjects who withdrew from the study or were lost to follow-up, the date of withdrawal or the last confirmed date of presence in the study was used as the date of censoring.

Hazard ratios and 95 percent confidence intervals were used to characterize the relative risk of cancer occurrence associated with daily total physical activity level. Daily total physical activity was assessed in quartiles of total METs/day score. The median METs/day value for each quartile was used when the linear association was assessed. To investigate whether the effect on the outcome differed by type of physical activity, we also assessed risk by the frequency of leisure-time sports or physical exercise (≤1–3 days/month, 1–2 days/week, 3–4 days/week, or almost every day), in addition to the amount of time spent per day in heavy physical work or strenuous exercise (none, <1 hour, or ≥1 hour) and in standing or walking (<1, 1–<3, or ≥3 hours). Ordinal values were used to assess linear trends for these variables.

The Cox proportional hazards model was employed to control for potentially confounding factors, namely age at the starting point (5-year categories), area (10 public health center areas), history of diabetes (no, yes), smoking status (never smoking, past smoking, or 1–19, 20–29, or ≥30 cigarettes/day), alcohol intake status (almost none, occasional, or regular), body mass index (weight (kg)/height (m)2; 14–<20, 20–<27, or ≥27), and total energy intake (in quintiles, estimated by semiquantitative food frequency questionnaire). These variables, obtained from the questionnaire, are either known or suspected risk factors for cancer that have been identified in previous studies. We treated age, area, and total energy intake as strata to allow for a different baseline hazard for each stratum. In testing of the proportional hazards assumption by Schoenfeld residuals and scaled Schoenfeld residuals, we found no violation of proportionality. In addition, we evaluated whether the effect of total physical activity was influenced by age, body mass index, or frequency of leisure-time sports or physical exercise using a test of interaction, by entering into the model multiplicative terms for interaction between the respective factors. Since the effect of total physical activity was significantly influenced by sex (p for interaction ≤ 0.001), all analysis were conducted by sex. All statistical analyses were performed using Stata 10 (Stata Corporation, College Station, Texas) (17).

RESULTS

During 599,117 person-years of follow-up (average follow-up period, 7.5 years) for the 79,771 subjects (37,898 men and 41,873 women), 4,334 newly diagnosed cases of cancer (2,704 in men and 1,630 in women), including skin cancer (n = 53; 1.2 percent), were identified and included in the analyses. In men, gastric cancer was the most common cancer (n = 621; 23.0 percent), followed by cancers of the lung (n = 388; 14.3 percent), colon (n = 328; 12.1 percent), and prostate (n = 279; 10.3 percent). In women, breast cancer was the most common (n = 294; 18.0 percent), followed by cancers of the stomach (n = 232; 14.2 percent), colon (n = 228; 14.0 percent), and lung (n = 144; 8.8 percent).

Characteristics of the study subjects according to physical activity level are shown in table 1. The median values in the lowest, second, third, and highest quartiles of total METs/day score were 25.45, 31.85, 34.25, and 42.65, respectively, in men and 26.10, 31.85, 34.25, and 42.65, respectively, in women. Men who were more physically active were more likely to report regular drinking, a higher frequency of leisure-time sports or physical exercise, and higher daily mean energy consumption and were less likely to report a history of diabetes mellitus and liver disease. No difference in body mass index was observed between groups by physical activity level. In women, similar trends were observed, except that the differences in the proportion of regular drinkers were not significant.

TABLE 1.

Baseline characteristics of study subjects according to daily total physical activity level (n = 79,771), Japan Public Health Center-based Prospective Study, 1995–2004

Characteristic Quartile of physical activity level (quartile of METs*/day score)
† 
Lowest Second Third Highest 
Men (n = 37,898)     
    No. of subjects 12,966 7,822 7,579 9,531 
    Quartile median value (METs/day score) 25.45 (21.60–27.10)‡ 31.85 (27.25–31.85) 34.25 (32.40–36.05) 42.65 (36.25–46.25) 
    Mean age (years) 56.7 56.4 56.9 56.1 
    Mean body mass index§ 23.66 23.62 23.56 23.49 
    History of diabetes mellitus (%) 10.1 8.6 8.3 7.6 
    History of liver disease (%) 3.7 2.9 2.7 2.6 
    Current smoking (%) 47.2 46.9 47.2 48.8 
    Regular alcohol drinking (≥1 day/week) (%) 64.6 68.0 68.1 71.0 
    Regular leisure-time sports or physical exercise (≥3–4 days/week) (%) 9.0 10.8 13.1 12.0 
    Mean total energy intake¶ (kcal/day) 2,040.8 2,141.5 2,168.6 2,299.5 
Women (n = 41,873)     
    No. of subjects 13,277 10,838 9,663 8,095 
    Quartile median value (METs/day score) 26.10 (21.60–27.10) 31.85 (27.25–31.85) 34.25 (32.75–34.25) 42.65 (35.45–46.25) 
    Mean age (years) 57.3 56.4 56.5 56.0 
    Mean body mass index 23.58 23.41 23.40 23.49 
    History of diabetes mellitus (%) 5.0 3.8 3.5 3.9 
    History of liver disease (%) 1.3 1.3 1.5 1.0 
    Current smoking (%) 5.9 5.8 5.6 5.5 
    Regular alcohol drinking (≥1 day/week) (%) 12.8 13.4 13.7 13.2 
    Regular leisure-time sports or physical exercise (≥3–4 days/week) (%) 9.4 9.7 11.5 14.6 
    Mean total energy intake¶ (kcal/day) 1,840.3 1,886.4 1,882.3 1,972.2 
Characteristic Quartile of physical activity level (quartile of METs*/day score)
† 
Lowest Second Third Highest 
Men (n = 37,898)     
    No. of subjects 12,966 7,822 7,579 9,531 
    Quartile median value (METs/day score) 25.45 (21.60–27.10)‡ 31.85 (27.25–31.85) 34.25 (32.40–36.05) 42.65 (36.25–46.25) 
    Mean age (years) 56.7 56.4 56.9 56.1 
    Mean body mass index§ 23.66 23.62 23.56 23.49 
    History of diabetes mellitus (%) 10.1 8.6 8.3 7.6 
    History of liver disease (%) 3.7 2.9 2.7 2.6 
    Current smoking (%) 47.2 46.9 47.2 48.8 
    Regular alcohol drinking (≥1 day/week) (%) 64.6 68.0 68.1 71.0 
    Regular leisure-time sports or physical exercise (≥3–4 days/week) (%) 9.0 10.8 13.1 12.0 
    Mean total energy intake¶ (kcal/day) 2,040.8 2,141.5 2,168.6 2,299.5 
Women (n = 41,873)     
    No. of subjects 13,277 10,838 9,663 8,095 
    Quartile median value (METs/day score) 26.10 (21.60–27.10) 31.85 (27.25–31.85) 34.25 (32.75–34.25) 42.65 (35.45–46.25) 
    Mean age (years) 57.3 56.4 56.5 56.0 
    Mean body mass index 23.58 23.41 23.40 23.49 
    History of diabetes mellitus (%) 5.0 3.8 3.5 3.9 
    History of liver disease (%) 1.3 1.3 1.5 1.0 
    Current smoking (%) 5.9 5.8 5.6 5.5 
    Regular alcohol drinking (≥1 day/week) (%) 12.8 13.4 13.7 13.2 
    Regular leisure-time sports or physical exercise (≥3–4 days/week) (%) 9.4 9.7 11.5 14.6 
    Mean total energy intake¶ (kcal/day) 1,840.3 1,886.4 1,882.3 1,972.2 
*

METs, metabolic equivalents.

Sum of the scores for reported amount of time per day spent in each physical activity multiplied by the MET value for each activity.

Numbers in parentheses, range.

§

Weight (kg)/height(m)2.

Adjusted for age.

Associations between daily total physical activity level by total METs/day score and total cancer incidence are shown in table 2. Upon multivariate adjustment, compared with subjects in the lowest quartile, increased daily total physical activity was significantly associated with a decreased risk of cancer incidence in both men and women. In men, hazard ratios in the second, third, and highest quartiles were 1.00 (95 percent confidence interval (CI): 0.90, 1.11), 0.96 (95 percent CI: 0.86, 1.07), and 0.87 (95 percent CI: 0.78, 0.96), respectively (p for trend = 0.005); in women, they were 0.93 (95 percent CI: 0.82, 1.05), 0.84 (95 percent CI: 0.73, 0.96), and 0.84 (95 percent CI: 0.73, 0.97), respectively (p for trend = 0.007). Our estimates also showed that the risk decreased by 7 percent in men and 10 percent in women with each 10-MET/day increase in physical activity level. The results did not differ substantially after exclusion of early cancer cases—those occurring within 3 years of the starting point—or after further exclusion of subjects with very low physical activity levels (<23 METs/day; 2 percent of subjects), considered to result from poor physical condition. On further estimation of the population attributable fraction (18) from our results, 4.5 percent of cases in men and 5.5 percent of cases in women were considered to have been preventable if the persons in the lowest physical activity category had increased their activity to a higher level.

TABLE 2.

Hazard ratios for total cancer incidence according to daily total physical activity level (n = 79,771), Japan Public Health Center-based Prospective Study, 1995–2004

Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within first 3 years
 
No. of cases HR1*† 95% CI* HR2‡ 95% CI No. of cases HR1 95% CI HR2 95% CI 
Men (n = 37,898)   (n = 2,704) (n = 1,804) 
    Lowest 12,966 92,421 921 1.00 Reference 1.00 Reference 604 1.00 Reference 1.00 Reference 
    Second 7,822 57,957 575 1.00 0.90, 1.10 1.00 0.90, 1.11 381 0.98 0.86, 1.11 0.98 0.86, 1.11 
    Third 7,579 56,512 574 0.96 0.86, 1.06 0.96 0.86, 1.07 386 0.95 0.83, 1.08 0.95 0.83, 1.08 
    Highest 9,531 72,841 634 0.87 0.79, 0.96 0.87 0.78, 0.96 433 0.86 0.76, 0.97 0.86 0.76, 0.98 
        p for trend    0.006 0.005  0.015 0.017 
    Per 1-MET increase    0.99 0.99, 0.998 0.99 0.99, 0.998  0.99 0.99, 0.999 0.99 0.99, 0.999 
    Per 10-MET increase    0.93 0.88, 0.99 0.93 0.88, 0.99  0.93 0.87, 0.996 0.93 0.87, 0.997 
Women (n = 41,873)   (n = 1,630) (n = 1,056) 
    Lowest 13,277 99,385 569 1.00 Reference 1.00 Reference 368 1.00 Reference 1.00 Reference 
    Second 10,838 83,644 428 0.92 0.81, 1.04 0.93 0.82, 1.05 290 0.94 0.81, 1.10 0.94 0.81, 1.10 
    Third 9,663 74,073 350 0.84 0.73, 0.96 0.84 0.73, 0.96 222 0.80 0.68, 0.95 0.79 0.67, 0.94 
    Highest 8,095 62,284 283 0.83 0.72, 0.96 0.84 0.73, 0.97 176 0.78 0.65, 0.93 0.78 0.65, 0.94 
        p for trend    0.004 0.007  0.002 0.002 
    Per 1-MET increase    0.99 0.98, 0.997 0.99 0.98, 0.997  0.98 0.97, 0.995 0.98 0.97, 0.995 
    Per 10-MET increase    0.89 0.82, 0.97 0.90 0.82, 0.98  0.85 0.77, 0.95 0.85 0.77, 0.95 
Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within first 3 years
 
No. of cases HR1*† 95% CI* HR2‡ 95% CI No. of cases HR1 95% CI HR2 95% CI 
Men (n = 37,898)   (n = 2,704) (n = 1,804) 
    Lowest 12,966 92,421 921 1.00 Reference 1.00 Reference 604 1.00 Reference 1.00 Reference 
    Second 7,822 57,957 575 1.00 0.90, 1.10 1.00 0.90, 1.11 381 0.98 0.86, 1.11 0.98 0.86, 1.11 
    Third 7,579 56,512 574 0.96 0.86, 1.06 0.96 0.86, 1.07 386 0.95 0.83, 1.08 0.95 0.83, 1.08 
    Highest 9,531 72,841 634 0.87 0.79, 0.96 0.87 0.78, 0.96 433 0.86 0.76, 0.97 0.86 0.76, 0.98 
        p for trend    0.006 0.005  0.015 0.017 
    Per 1-MET increase    0.99 0.99, 0.998 0.99 0.99, 0.998  0.99 0.99, 0.999 0.99 0.99, 0.999 
    Per 10-MET increase    0.93 0.88, 0.99 0.93 0.88, 0.99  0.93 0.87, 0.996 0.93 0.87, 0.997 
Women (n = 41,873)   (n = 1,630) (n = 1,056) 
    Lowest 13,277 99,385 569 1.00 Reference 1.00 Reference 368 1.00 Reference 1.00 Reference 
    Second 10,838 83,644 428 0.92 0.81, 1.04 0.93 0.82, 1.05 290 0.94 0.81, 1.10 0.94 0.81, 1.10 
    Third 9,663 74,073 350 0.84 0.73, 0.96 0.84 0.73, 0.96 222 0.80 0.68, 0.95 0.79 0.67, 0.94 
    Highest 8,095 62,284 283 0.83 0.72, 0.96 0.84 0.73, 0.97 176 0.78 0.65, 0.93 0.78 0.65, 0.94 
        p for trend    0.004 0.007  0.002 0.002 
    Per 1-MET increase    0.99 0.98, 0.997 0.99 0.98, 0.997  0.98 0.97, 0.995 0.98 0.97, 0.995 
    Per 10-MET increase    0.89 0.82, 0.97 0.90 0.82, 0.98  0.85 0.77, 0.95 0.85 0.77, 0.95 
*

MET(s), metabolic equivalent(s); HR, hazard ratio; CI, confidence interval.

Adjusted for age (stratified, 5-year categories) and area (stratified, 10 public health center areas).

Adjusted for age (stratified, 5-year categories), area (stratified, 10 public health center areas), total energy intake (stratified, quintiles), history of diabetes (no, yes), smoking status (never smoking, past smoking, or 1–19, 20–29, or ≥30 cigarettes/day), alcohol intake status (almost none, occasional, or regular), body mass index (weight (kg)/height (m)2; <20, 20–<27, or ≥27), and leisure-time sports or physical exercise (<1, 1–2, or ≥3–4 days/week).

In both sexes, the degree of risk decrease was attenuated among persons with increasing body mass index. In contrast, it was strengthened among the elderly and among persons who regularly engaged in leisure-time sports or physical exercise; this relation appeared more clearly in women. No significant interaction was observed for age, obesity status, or frequency of leisure-time sports and physical exercise (table 3). No particularly significant associations were identified in analysis by type of physical activity (table 4).

TABLE 3.

Hazard ratios for total cancer incidence according to daily total physical activity level and body mass index or frequency of leisure-time sports or physical exercise (n = 79,771), Japan Public Health Center-based Prospective Study, 1995–2004

Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within the first 3 years
 
No. of cases HR*† 95% CI* No. of cases HR† 95% CI 
Men (n = 37,898)         
    Age (years)         
        <60         
            Lowest 8,239 61,181 364 1.00 Reference 259 1.00 Reference 
            Second 5,063 38,860 239 1.00 0.85, 1.18 174 1.00 0.83, 1.22 
            Third 4,709 36,624 219 0.94 0.79, 1.12 161 0.94 0.77, 1.15 
            Highest 6,301 49,823 269 0.86 0.73, 1.01 202 0.87 0.72, 1.06 
                p for trend    0.049  0.135 
        ≥60         
            Lowest 4,727 31,240 557 1.00 Reference 345 1.00 Reference 
            Second 2,759 19,096 336 0.99 0.86, 1.14 207 0.96 0.80, 1.14 
            Third 2,870 19,887 355 0.97 0.85, 1.11 225 0.96 0.81, 1.14 
            Highest 3,230 23,018 365 0.87 0.76, 1.00 231 0.85 0.72, 1.01 
                p for trend    0.051  0.064 
                p for interaction    0.505  0.976 
    Body mass index‡         
        <20         
            Lowest 2,316 15,737 196 1.00 Reference 121 1.00 Reference 
            Second 1,409 10,180 118 0.93 0.73, 1.17 69 0.85 0.63, 1.16 
            Third 1,407 10,194 131 0.97 0.77, 1.22 89 1.02 0.77, 1.36 
            Highest 1,772 13,162 126 0.79 0.63, 1.00 71 0.69 0.51, 0.94 
                p for trend    0.063  0.031 
        20–<27         
            Lowest 9,081 65,122 632 1.00 Reference 420 1.00 Reference 
            Second 5,493 40,888 386 0.97 0.86, 1.11 264 0.99 0.83, 1.14 
            Third 5,325 39,896 397 0.96 0.85, 1.09 263 0.92 0.79, 1.08 
            Highest 6,779 52,341 451 0.87 0.77, 0.98 324 0.89 0.77, 1.04 
                p for trend    0.026  0.118 
        ≥27         
            Lowest 1,569 11,562 93 1.00 Reference 63 1.00 Reference 
            Second 920 6,889 71 1.16 0.84, 1.62 48 1.23 0.83, 1.84 
            Third 847 6,422 46 0.84 0.58, 1.22 34 0.94 0.60, 1.46 
            Highest 980 7,339 57 0.93 0.66, 1.32 38 0.96 0.63, 1.47 
                p for trend    0.501  0.713 
                p for interaction    0.515  0.797 
    Frequency of leisure-time sports or physical exercise (days/week)         
        <1         
            Lowest 10,378 74,547 723 1.00 Reference 479 1.00 Reference 
            Second 6,077 45,423 453 1.02 0.91, 1.15 309 1.01 0.88, 1.17 
            Third 5,704 42,999 443 1.00 0.88, 1.12 303 0.98 0.85, 1.14 
            Highest 7,497 57,786 499 0.88 0.79, 0.99 343 0.87 0.75, 1.00 
                p for trend    0.032  0.044 
        ≥1         
            Lowest 2,588 17,875 198 1.00 Reference 125 1.00 Reference 
            Second 1,745 12,534 122 0.90 0.72, 1.14 72 0.84 0.63, 1.13 
            Third 1,875 13,513 131 0.84 0.67, 1.06 83 0.84 0.63, 1.12 
            Highest 2,034 15,055 135 0.78 0.62, 0.99 90 0.82 0.62, 1.09 
                p for trend    0.034  0.190 
                p for interaction    0.766  0.566 
Women (n = 41,873)         
    Age (years)         
        <60         
            Lowest 7,946 61,385 279 1.00 Reference 184 1.00 Reference 
            Second 7,053 55,628 261 1.03 0.87, 1.22 184 1.09 0.88, 1.33 
            Third 6,271 48,932 202 0.90 0.75, 1.08 131 0.86 0.69, 1.08 
            Highest 5,501 43,242 188 0.95 0.79, 1.15 120 0.91 0.72, 1.14 
                p for trend    0.419  0.241 
        ≥60         
            Lowest 5,331 38,000 290 1.00 Reference 184 1.00 Reference 
            Second 3,785 28,016 167 0.81 0.67, 0.98 106 0.78 0.61, 0.996 
            Third 3,392 25,141 148 0.77 0.63, 0.95 91 0.72 0.56, 0.93 
            Highest 2,594 19,042 95 0.71 0.56, 0.90 56 0.63 0.47, 0.86 
                p for trend    0.001  0.001 
                p for interaction    0.667  0.396 
    Body mass index         
        <20         
            Lowest 2,896 20,823 116 1.00 Reference 72 1.00 Reference 
            Second 2,383 17,909 86 0.92 0.68, 1.22 64 1.08 0.76, 1.54 
            Third 2,096 15,459 69 0.87 0.64, 1.18 45 0.92 0.63, 1.36 
            Highest 1,598 12,009 47 0.76 0.54, 1.09 35 0.92 0.60, 1.40 
                p for trend    0.119  0.623 
        20–<27         
            Lowest 8,467 63,889 370 1.00 Reference 238 1.00 Reference 
            Second 7,117 55,220 283 0.91 0.78, 1.06 190 0.92 0.76, 1.12 
            Third 6,453 49,990 239 0.82 0.70, 0.97 149 0.76 0.62, 0.93 
            Highest 5,515 42,597 192 0.81 0.68, 0.97 116 0.73 0.58, 0.92 
                p for trend    0.009  0.002 
        ≥27         
            Lowest 1,914 14,673 83 1.00 Reference 58 1.00 Reference 
            Second 1,338 10,516 59 1.05 0.74, 1.48 36 0.94 0.61, 1.44 
            Third 1,114 8,624 42 0.82 0.56, 1.20 28 0.79 0.49, 1.25 
            Highest 982 7,678 44 0.96 0.65, 1.41 25 0.76 0.46, 1.25 
                p for trend    0.643  0.223 
                p for interaction    0.839  0.137 
    Frequency of leisure-time sports or physical exercise (days/week)         
        <1         
            Lowest 10,837 81,716 464 1.00 Reference 297 1.00 Reference 
            Second 8,773 68,595 354 0.95 0.83, 1.10 236 0.96 0.81, 1.14 
            Third 7,521 58,563 274 0.84 0.72, 0.98 174 0.80 0.66, 0.97 
            Highest 5,811 45,696 223 0.92 0.78, 1.08 139 0.87 0.70, 1.06 
                p for trend    0.140  0.065 
        ≥1         
            Lowest 2,440 17,670 105 1.00 Reference 71 1.00 Reference 
            Second 2,065 15,049 74 0.80 0.59, 1.09 54 0.85 0.59, 1.22 
            Third 2,142 15,510 76 0.81 0.59, 1.09 48 0.74 0.51, 1.08 
            Highest 2,284 16,587 60 0.61 0.44, 0.84 37 0.55 0.37, 0.83 
                p for trend    0.003  0.003 
                p for interaction    0.158  0.105 
Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within the first 3 years
 
No. of cases HR*† 95% CI* No. of cases HR† 95% CI 
Men (n = 37,898)         
    Age (years)         
        <60         
            Lowest 8,239 61,181 364 1.00 Reference 259 1.00 Reference 
            Second 5,063 38,860 239 1.00 0.85, 1.18 174 1.00 0.83, 1.22 
            Third 4,709 36,624 219 0.94 0.79, 1.12 161 0.94 0.77, 1.15 
            Highest 6,301 49,823 269 0.86 0.73, 1.01 202 0.87 0.72, 1.06 
                p for trend    0.049  0.135 
        ≥60         
            Lowest 4,727 31,240 557 1.00 Reference 345 1.00 Reference 
            Second 2,759 19,096 336 0.99 0.86, 1.14 207 0.96 0.80, 1.14 
            Third 2,870 19,887 355 0.97 0.85, 1.11 225 0.96 0.81, 1.14 
            Highest 3,230 23,018 365 0.87 0.76, 1.00 231 0.85 0.72, 1.01 
                p for trend    0.051  0.064 
                p for interaction    0.505  0.976 
    Body mass index‡         
        <20         
            Lowest 2,316 15,737 196 1.00 Reference 121 1.00 Reference 
            Second 1,409 10,180 118 0.93 0.73, 1.17 69 0.85 0.63, 1.16 
            Third 1,407 10,194 131 0.97 0.77, 1.22 89 1.02 0.77, 1.36 
            Highest 1,772 13,162 126 0.79 0.63, 1.00 71 0.69 0.51, 0.94 
                p for trend    0.063  0.031 
        20–<27         
            Lowest 9,081 65,122 632 1.00 Reference 420 1.00 Reference 
            Second 5,493 40,888 386 0.97 0.86, 1.11 264 0.99 0.83, 1.14 
            Third 5,325 39,896 397 0.96 0.85, 1.09 263 0.92 0.79, 1.08 
            Highest 6,779 52,341 451 0.87 0.77, 0.98 324 0.89 0.77, 1.04 
                p for trend    0.026  0.118 
        ≥27         
            Lowest 1,569 11,562 93 1.00 Reference 63 1.00 Reference 
            Second 920 6,889 71 1.16 0.84, 1.62 48 1.23 0.83, 1.84 
            Third 847 6,422 46 0.84 0.58, 1.22 34 0.94 0.60, 1.46 
            Highest 980 7,339 57 0.93 0.66, 1.32 38 0.96 0.63, 1.47 
                p for trend    0.501  0.713 
                p for interaction    0.515  0.797 
    Frequency of leisure-time sports or physical exercise (days/week)         
        <1         
            Lowest 10,378 74,547 723 1.00 Reference 479 1.00 Reference 
            Second 6,077 45,423 453 1.02 0.91, 1.15 309 1.01 0.88, 1.17 
            Third 5,704 42,999 443 1.00 0.88, 1.12 303 0.98 0.85, 1.14 
            Highest 7,497 57,786 499 0.88 0.79, 0.99 343 0.87 0.75, 1.00 
                p for trend    0.032  0.044 
        ≥1         
            Lowest 2,588 17,875 198 1.00 Reference 125 1.00 Reference 
            Second 1,745 12,534 122 0.90 0.72, 1.14 72 0.84 0.63, 1.13 
            Third 1,875 13,513 131 0.84 0.67, 1.06 83 0.84 0.63, 1.12 
            Highest 2,034 15,055 135 0.78 0.62, 0.99 90 0.82 0.62, 1.09 
                p for trend    0.034  0.190 
                p for interaction    0.766  0.566 
Women (n = 41,873)         
    Age (years)         
        <60         
            Lowest 7,946 61,385 279 1.00 Reference 184 1.00 Reference 
            Second 7,053 55,628 261 1.03 0.87, 1.22 184 1.09 0.88, 1.33 
            Third 6,271 48,932 202 0.90 0.75, 1.08 131 0.86 0.69, 1.08 
            Highest 5,501 43,242 188 0.95 0.79, 1.15 120 0.91 0.72, 1.14 
                p for trend    0.419  0.241 
        ≥60         
            Lowest 5,331 38,000 290 1.00 Reference 184 1.00 Reference 
            Second 3,785 28,016 167 0.81 0.67, 0.98 106 0.78 0.61, 0.996 
            Third 3,392 25,141 148 0.77 0.63, 0.95 91 0.72 0.56, 0.93 
            Highest 2,594 19,042 95 0.71 0.56, 0.90 56 0.63 0.47, 0.86 
                p for trend    0.001  0.001 
                p for interaction    0.667  0.396 
    Body mass index         
        <20         
            Lowest 2,896 20,823 116 1.00 Reference 72 1.00 Reference 
            Second 2,383 17,909 86 0.92 0.68, 1.22 64 1.08 0.76, 1.54 
            Third 2,096 15,459 69 0.87 0.64, 1.18 45 0.92 0.63, 1.36 
            Highest 1,598 12,009 47 0.76 0.54, 1.09 35 0.92 0.60, 1.40 
                p for trend    0.119  0.623 
        20–<27         
            Lowest 8,467 63,889 370 1.00 Reference 238 1.00 Reference 
            Second 7,117 55,220 283 0.91 0.78, 1.06 190 0.92 0.76, 1.12 
            Third 6,453 49,990 239 0.82 0.70, 0.97 149 0.76 0.62, 0.93 
            Highest 5,515 42,597 192 0.81 0.68, 0.97 116 0.73 0.58, 0.92 
                p for trend    0.009  0.002 
        ≥27         
            Lowest 1,914 14,673 83 1.00 Reference 58 1.00 Reference 
            Second 1,338 10,516 59 1.05 0.74, 1.48 36 0.94 0.61, 1.44 
            Third 1,114 8,624 42 0.82 0.56, 1.20 28 0.79 0.49, 1.25 
            Highest 982 7,678 44 0.96 0.65, 1.41 25 0.76 0.46, 1.25 
                p for trend    0.643  0.223 
                p for interaction    0.839  0.137 
    Frequency of leisure-time sports or physical exercise (days/week)         
        <1         
            Lowest 10,837 81,716 464 1.00 Reference 297 1.00 Reference 
            Second 8,773 68,595 354 0.95 0.83, 1.10 236 0.96 0.81, 1.14 
            Third 7,521 58,563 274 0.84 0.72, 0.98 174 0.80 0.66, 0.97 
            Highest 5,811 45,696 223 0.92 0.78, 1.08 139 0.87 0.70, 1.06 
                p for trend    0.140  0.065 
        ≥1         
            Lowest 2,440 17,670 105 1.00 Reference 71 1.00 Reference 
            Second 2,065 15,049 74 0.80 0.59, 1.09 54 0.85 0.59, 1.22 
            Third 2,142 15,510 76 0.81 0.59, 1.09 48 0.74 0.51, 1.08 
            Highest 2,284 16,587 60 0.61 0.44, 0.84 37 0.55 0.37, 0.83 
                p for trend    0.003  0.003 
                p for interaction    0.158  0.105 
*

METs, metabolic equivalents; HR, hazard ratio; CI, confidence interval.

Adjusted for age (stratified, 5-year categories), area (stratified, 10 public health center areas), total energy intake (stratified, quintiles), history of diabetes (no, yes), smoking status (never smoking, past smoking, or 1–19, 20–29, or ≥30 cigarettes/day), alcohol intake status (almost none, occasional, or regular), body mass index (weight (kg)/height (m)2; <20, 20–<27, or ≥27), and leisure-time sports or physical exercise (<1, 1–2, or ≥3–4 days/week).

Weight (kg)/height (m)2.

TABLE 4.

Hazard ratios* for total cancer incidence according to type of physical activity (n = 79,771), Japan Public Health Center-based Prospective Study, 1995–2004

 No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within the first 3 years
 
 No. of cases HR† 95% CI† No. of cases HR 95% CI 
Men (n = 37,898)         
    Heavy physical work or strenuous exercise (hours/day)         
        None 22,235 161,694 1,670 1.00 Reference 1,093 1.00 Reference 
        <1 5,165 38,119 324 0.95 0.84, 1.07 229 1.02 0.88, 1.18 
        ≥1 10,498 79,918 710 0.89 0.81, 0.98 482 0.89 0.80, 1.00 
            p for trend    0.014  0.071 
    Standing or walking (hours/day)         
        <1 8,243 59,839 564 1.00 Reference 369 1.00 Reference 
        1–<3 9,143 65,023 649 1.04 0.92, 1.17 425 1.04 0.90, 1.21 
        ≥3 20,512 154,869 1,491 0.99 0.89, 1.11 1,010 0.99 0.87, 1.13 
            p for trend    0.787  0.764 
    Sitting (hours/day)         
        <3 17,251 128,076 1,230 1.00 Reference 821 1.00 Reference 
        3–<8 17,472 128,067 1,247 0.97 0.89, 1.06 835 0.97 0.88, 1.08 
        ≥8 3,175 23,588 227 1.02 0.87, 1.18 148 0.97 0.80, 1.16 
            p for trend    0.839  0.599 
    Leisure-time sports or physical exercise (days/week)         
        <1 29,656 220,754 2,118 1.00 Reference 1,434 1.00 Reference 
        1–2 4,095 30,011 240 0.92 0.80, 1.05 155 0.87 0.74, 1.03 
        ≥3–4 4,147 28,965 346 1.12 0.998, 1.26 215 1.09 0.94, 1.26 
            p for trend    0.158  0.519 
Women (n = 41,873)         
    Heavy physical work or strenuous exercise (hours/day)         
        None 31,286 238,962 1,266 1.00 Reference 832 1.00 Reference 
        <1 4,097 30,583 138 0.91 0.76, 1.09 89 0.90 0.72, 1.12 
        ≥1 6,490 49,840 226 0.93 0.80, 1.07 135 0.84 0.70, 1.01 
            p for trend    0.200  0.043 
    Standing or walking (hours/day)         
        <1 6,077 45,688 259 1.00 Reference 164 1.00 Reference 
        1–<3 9,828 73,552 410 1.00 0.85, 1.18 266 1.02 0.84, 1.25 
        ≥3 25,968 200,146 961 0.89 0.77, 1.04 626 0.90 0.75, 1.09 
            p for trend    0.054  0.128 
    Sitting (hours/day)         
        <3 18,981 144,501 724 1.00 Reference 463 1.00 Reference 
        3–<8 20,184 153,659 785 0.98 0.88, 1.09 509 0.97 0.85, 1.11 
        ≥8 2,708 21,226 121 1.05 0.86, 1.29 84 1.10 0.86, 1.41 
            p for trend    0.896  0.748 
    Leisure-time sports or physical exercise (days/week)         
        <1 32,942 254,570 1,315 1.00 Reference 846 1.00 Reference 
        1–2 4,338 31,712 136 0.91 0.76, 1.09 85 0.91 0.73, 1.15 
        ≥3–4 4,593 33,104 179 1.05 0.89, 1.23 125 1.20 0.99, 1.45 
            p for trend    0.883  0.160 
 No. of subjects Person-years of follow-up Total
 
Excluding cases diagnosed within the first 3 years
 
 No. of cases HR† 95% CI† No. of cases HR 95% CI 
Men (n = 37,898)         
    Heavy physical work or strenuous exercise (hours/day)         
        None 22,235 161,694 1,670 1.00 Reference 1,093 1.00 Reference 
        <1 5,165 38,119 324 0.95 0.84, 1.07 229 1.02 0.88, 1.18 
        ≥1 10,498 79,918 710 0.89 0.81, 0.98 482 0.89 0.80, 1.00 
            p for trend    0.014  0.071 
    Standing or walking (hours/day)         
        <1 8,243 59,839 564 1.00 Reference 369 1.00 Reference 
        1–<3 9,143 65,023 649 1.04 0.92, 1.17 425 1.04 0.90, 1.21 
        ≥3 20,512 154,869 1,491 0.99 0.89, 1.11 1,010 0.99 0.87, 1.13 
            p for trend    0.787  0.764 
    Sitting (hours/day)         
        <3 17,251 128,076 1,230 1.00 Reference 821 1.00 Reference 
        3–<8 17,472 128,067 1,247 0.97 0.89, 1.06 835 0.97 0.88, 1.08 
        ≥8 3,175 23,588 227 1.02 0.87, 1.18 148 0.97 0.80, 1.16 
            p for trend    0.839  0.599 
    Leisure-time sports or physical exercise (days/week)         
        <1 29,656 220,754 2,118 1.00 Reference 1,434 1.00 Reference 
        1–2 4,095 30,011 240 0.92 0.80, 1.05 155 0.87 0.74, 1.03 
        ≥3–4 4,147 28,965 346 1.12 0.998, 1.26 215 1.09 0.94, 1.26 
            p for trend    0.158  0.519 
Women (n = 41,873)         
    Heavy physical work or strenuous exercise (hours/day)         
        None 31,286 238,962 1,266 1.00 Reference 832 1.00 Reference 
        <1 4,097 30,583 138 0.91 0.76, 1.09 89 0.90 0.72, 1.12 
        ≥1 6,490 49,840 226 0.93 0.80, 1.07 135 0.84 0.70, 1.01 
            p for trend    0.200  0.043 
    Standing or walking (hours/day)         
        <1 6,077 45,688 259 1.00 Reference 164 1.00 Reference 
        1–<3 9,828 73,552 410 1.00 0.85, 1.18 266 1.02 0.84, 1.25 
        ≥3 25,968 200,146 961 0.89 0.77, 1.04 626 0.90 0.75, 1.09 
            p for trend    0.054  0.128 
    Sitting (hours/day)         
        <3 18,981 144,501 724 1.00 Reference 463 1.00 Reference 
        3–<8 20,184 153,659 785 0.98 0.88, 1.09 509 0.97 0.85, 1.11 
        ≥8 2,708 21,226 121 1.05 0.86, 1.29 84 1.10 0.86, 1.41 
            p for trend    0.896  0.748 
    Leisure-time sports or physical exercise (days/week)         
        <1 32,942 254,570 1,315 1.00 Reference 846 1.00 Reference 
        1–2 4,338 31,712 136 0.91 0.76, 1.09 85 0.91 0.73, 1.15 
        ≥3–4 4,593 33,104 179 1.05 0.89, 1.23 125 1.20 0.99, 1.45 
            p for trend    0.883  0.160 
*

The model included age (stratified, 5-year categories), area (stratified, 10 public health center areas), total energy intake (stratified, quintiles), history of diabetes (no, yes), smoking status (never smoking, past smoking, or 1–19, 20–29, or ≥30 cigarettes/day), alcohol intake status (almost none, occasional, regular), body mass index (weight (kg)/height (m)2; <20, 20–<27, or ≥27), heavy physical work or strenuous exercise (none, <1 hour, or ≥1 hour/day), sitting (<3, 3–<8, or ≥8 hours/day), standing or walking (<1, 1–<3, or ≥3 hours/day), and leisure-time sports or physical exercise (<1, 1–2, or ≥3–4 days/week).

HR, hazard ratio; CI, confidence interval.

Results from analyses of specific cancer sites are shown in table 5. Significantly decreased risks were observed for colon, liver, and pancreatic cancer in men and for stomach cancer in women. In additional analyses for these cancers stratified by age, body mass index, and frequency of leisure-time sports or physical exercise, larger risk reductions were observed in persons with a lower body mass index, persons with frequent leisure-time sports or physical exercise, and the elderly for female stomach cancer and in persons with lower body mass index and persons with infrequent leisure-time sports or physical exercise for male colon cancer. For male liver and pancreatic cancers, we did not detect any significant difference or tendency in risk between stratified groups. In the analysis of breast cancer, the null association was not influenced by menopausal status.

TABLE 5.

Hazard ratios for incidence of cancer at specific sites according to daily total physical activity level (n = 79,771), Japan Public Health Center-based Prospective Study, 1995–2004

Site (International Classification of Diseases for Oncology, Third Edition, code) Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up No. of cases Hazard ratio† 95% confidence interval 
Men (n = 37,898)       
    Stomach (C16) Lowest 12,966 92,421 194 1.00 Reference 
 Second 7,822 57,957 134 1.10 0.88, 1.37 
 Third 7,579 56,512 136 1.10 0.88, 1.37 
 Highest 9,531 72,841 157 1.04 0.84, 1.29 
     p for trend    0.785 
    Colon (C18) Lowest 12,966 92,421 131 1.00 Reference 
 Second 7,822 57,957 72 0.83 0.62, 1.11 
 Third 7,579 56,512 59 0.65 0.48, 0.89 
 Highest 9,531 72,841 66 0.58 0.43, 0.79 
     p for trend    <0.001 
    Rectum (C19–20) Lowest 12,966 92,421 51 1.00 Reference 
 Second 7,822 57,957 41 1.30 0.85, 1.97 
 Third 7,579 56,512 35 1.11 0.72, 1.72 
 Highest 9,531 72,841 35 0.88 0.57, 1.36 
     p for trend    0.464 
    Liver (C22) Lowest 12,966 92,421 82 1.00 Reference 
 Second 7,822 57,957 32 0.69 0.45, 1.06 
 Third 7,579 56,512 44 1.01 0.69, 1.49 
 Highest 9,531 72,841 31 0.62 0.40, 0.96 
     p for trend    0.062 
    Pancreas (C25) Lowest 12,966 92,421 36 1.00 Reference 
 Second 7,822 57,957 20 0.90 0.52, 1.57 
 Third 7,579 56,512 15 0.67 0.36, 1.24 
 Highest 9,531 72,841 16 0.55 0.30, 1.00 
     p for trend    0.038 
    Lung (C34) Lowest 12,966 92,421 108 1.00 Reference 
 Second 7,822 57,957 81 1.22 0.91, 1.63 
 Third 7,579 56,512 103 1.44 1.09, 1.90 
 Highest 9,531 72,841 96 1.10 0.83, 1.45 
     p for trend    0.494 
    Prostate (C61) Lowest 12,966 92,421 77 1.00 Reference 
 Second 7,822 57,957 68 1.39 1.00, 1.94 
 Third 7,579 56,512 63 1.21 0.86, 1.69 
 Highest 9,531 72,841 71 1.13 0.82, 1.57 
     p for trend    0.644 
Women (n = 41,873)       
    Stomach (C16) Lowest 13,277 99,385 91 1.00 Reference 
 Second 10,838 83,644 53 0.74 0.52, 1.04 
 Third 9,663 74,073 54 0.78 0.55, 1.10 
 Highest 8,095 62,284 34 0.63 0.42, 0.94 
     p for trend    0.020 
    Colon (C18) Lowest 13,277 99,385 83 1.00 Reference 
 Second 10,838 83,644 58 0.87 0.62, 1.22 
 Third 9,663 74,073 48 0.74 0.52, 1.07 
 Highest 8,095 62,284 39 0.82 0.56, 1.21 
     p for trend    0.198 
    Rectum (C19–20) Lowest 13,277 99,385 24 1.00 Reference 
 Second 10,838 83,644 24 1.26 0.71, 2.23 
 Third 9,663 74,073 16 1.05 0.55, 2.00 
 Highest 8,095 62,284 22 1.79 0.99, 3.23 
     p for trend    0.077 
    Liver (C22) Lowest 13,277 99,385 29 1.00 Reference 
 Second 10,838 83,644 19 0.96 0.52, 1.78 
 Third 9,663 74,073 19 0.99 0.53, 1.84 
 Highest 8,095 62,284 0.54 0.23, 1.29 
     p for trend    0.248 
    Pancreas (C25) Lowest 13,277 99,385 19 1.00 Reference 
 Second 10,838 83,644 15 0.98 0.50, 1.95 
 Third 9,663 74,073 11 0.83 0.39, 1.76 
 Highest 8,095 62,284 13 1.29 0.62, 2.67 
     p for trend    0.601 
    Lung (C34) Lowest 13,277 99,385 50 1.00 Reference 
 Second 10,838 83,644 37 0.90 0.58, 1.38 
 Third 9,663 74,073 31 0.90 0.57, 1.42 
 Highest 8,095 62,284 26 0.92 0.56, 1.49 
     p for trend    0.686 
    Breast (C50) Lowest 13,277 99,385 85 1.00 Reference 
 Second 10,838 83,644 91 1.24 0.92, 1.66 
 Third 9,663 74,073 67 1.02 0.74, 1.40 
 Highest 8,095 62,284 51 0.91 0.64, 1.29 
     p for trend    0.529 
Site (International Classification of Diseases for Oncology, Third Edition, code) Quartile of physical activity level (quartile of METs*/day score) No. of subjects Person-years of follow-up No. of cases Hazard ratio† 95% confidence interval 
Men (n = 37,898)       
    Stomach (C16) Lowest 12,966 92,421 194 1.00 Reference 
 Second 7,822 57,957 134 1.10 0.88, 1.37 
 Third 7,579 56,512 136 1.10 0.88, 1.37 
 Highest 9,531 72,841 157 1.04 0.84, 1.29 
     p for trend    0.785 
    Colon (C18) Lowest 12,966 92,421 131 1.00 Reference 
 Second 7,822 57,957 72 0.83 0.62, 1.11 
 Third 7,579 56,512 59 0.65 0.48, 0.89 
 Highest 9,531 72,841 66 0.58 0.43, 0.79 
     p for trend    <0.001 
    Rectum (C19–20) Lowest 12,966 92,421 51 1.00 Reference 
 Second 7,822 57,957 41 1.30 0.85, 1.97 
 Third 7,579 56,512 35 1.11 0.72, 1.72 
 Highest 9,531 72,841 35 0.88 0.57, 1.36 
     p for trend    0.464 
    Liver (C22) Lowest 12,966 92,421 82 1.00 Reference 
 Second 7,822 57,957 32 0.69 0.45, 1.06 
 Third 7,579 56,512 44 1.01 0.69, 1.49 
 Highest 9,531 72,841 31 0.62 0.40, 0.96 
     p for trend    0.062 
    Pancreas (C25) Lowest 12,966 92,421 36 1.00 Reference 
 Second 7,822 57,957 20 0.90 0.52, 1.57 
 Third 7,579 56,512 15 0.67 0.36, 1.24 
 Highest 9,531 72,841 16 0.55 0.30, 1.00 
     p for trend    0.038 
    Lung (C34) Lowest 12,966 92,421 108 1.00 Reference 
 Second 7,822 57,957 81 1.22 0.91, 1.63 
 Third 7,579 56,512 103 1.44 1.09, 1.90 
 Highest 9,531 72,841 96 1.10 0.83, 1.45 
     p for trend    0.494 
    Prostate (C61) Lowest 12,966 92,421 77 1.00 Reference 
 Second 7,822 57,957 68 1.39 1.00, 1.94 
 Third 7,579 56,512 63 1.21 0.86, 1.69 
 Highest 9,531 72,841 71 1.13 0.82, 1.57 
     p for trend    0.644 
Women (n = 41,873)       
    Stomach (C16) Lowest 13,277 99,385 91 1.00 Reference 
 Second 10,838 83,644 53 0.74 0.52, 1.04 
 Third 9,663 74,073 54 0.78 0.55, 1.10 
 Highest 8,095 62,284 34 0.63 0.42, 0.94 
     p for trend    0.020 
    Colon (C18) Lowest 13,277 99,385 83 1.00 Reference 
 Second 10,838 83,644 58 0.87 0.62, 1.22 
 Third 9,663 74,073 48 0.74 0.52, 1.07 
 Highest 8,095 62,284 39 0.82 0.56, 1.21 
     p for trend    0.198 
    Rectum (C19–20) Lowest 13,277 99,385 24 1.00 Reference 
 Second 10,838 83,644 24 1.26 0.71, 2.23 
 Third 9,663 74,073 16 1.05 0.55, 2.00 
 Highest 8,095 62,284 22 1.79 0.99, 3.23 
     p for trend    0.077 
    Liver (C22) Lowest 13,277 99,385 29 1.00 Reference 
 Second 10,838 83,644 19 0.96 0.52, 1.78 
 Third 9,663 74,073 19 0.99 0.53, 1.84 
 Highest 8,095 62,284 0.54 0.23, 1.29 
     p for trend    0.248 
    Pancreas (C25) Lowest 13,277 99,385 19 1.00 Reference 
 Second 10,838 83,644 15 0.98 0.50, 1.95 
 Third 9,663 74,073 11 0.83 0.39, 1.76 
 Highest 8,095 62,284 13 1.29 0.62, 2.67 
     p for trend    0.601 
    Lung (C34) Lowest 13,277 99,385 50 1.00 Reference 
 Second 10,838 83,644 37 0.90 0.58, 1.38 
 Third 9,663 74,073 31 0.90 0.57, 1.42 
 Highest 8,095 62,284 26 0.92 0.56, 1.49 
     p for trend    0.686 
    Breast (C50) Lowest 13,277 99,385 85 1.00 Reference 
 Second 10,838 83,644 91 1.24 0.92, 1.66 
 Third 9,663 74,073 67 1.02 0.74, 1.40 
 Highest 8,095 62,284 51 0.91 0.64, 1.29 
     p for trend    0.529 
*

METs, metabolic equivalents.

Adjusted for age (stratified, 5-year categories), area (stratified, 10 public health center areas), total energy intake (stratified, quintiles), history of diabetes (no, yes), smoking status (never smoker, past smoker, or 1–19, 20–29, or ≥30 cigarettes/day), alcohol intake status (almost none, occasional, or regular), body mass index (weight (kg)/height (m)2; <20, 20–<27, or ≥27), and leisure-time sports or physical exercise (<1, 1–2, or ≥3–4 days/week).

DISCUSSION

The health benefits of physical activity are well established for certain cancer sites (1, 19), but the extent to which the grand sum of these effects influences total cancer incidence has not been clarified. Of course, any such association depends to some degree on the background population, namely the site distribution of cancers which are strongly or weakly associated with physical activity. According to recent statistics, in Japan the cancer sites with the highest incidence rates are the stomach, followed by the lung, colon, liver, and prostate, for men and the breast, followed by the stomach, colon, uterus, and lung, for women (20). In this large-scale, population-based cohort study of Japanese men and women, we found a significant inverse association between daily total physical activity level and total cancer incidence. To reduce the potential for spurious associations from reverse causation, we excluded all subjects with a history of cancer at the starting point. Moreover, exclusion of early cases (those occurring within 3 years of the starting point) had no substantial effect on the results.

To our knowledge, only two studies have assessed the association between physical activity and total risk of cancer (2, 3); both were carried out in relatively small populations. One, which targeted men only, observed a reduced risk with increased physical activity (2), while the second observed an increased risk with increased nonrecreational physical inactivity (3). Our findings, obtained with a substantially larger sample, accord with those of these previous studies.

Our results showed basically similar risk reductions in men and women. Shephard and Shek (21) suggested that differences between the sexes in benefits associated with regular physical activity are due to the difference in hormonal conditions, which may lead to the failure to adapt activity questionnaires to traditional patterns of physical activity in females. Methodologically, it is commonly noted that men are more likely to be physically active in their jobs and women are more likely to be involved in housework (22). In our questionnaire, rank correlation coefficients for correlation with the 24-hour physical activity record were higher in men than in women. This may have partly resulted from the failure of our questionnaire to suitably account for housework. This type of measurement error may have led to underestimation of the association. Nevertheless, in the present study, a stronger effect of total physical activity among persons who engaged in regular leisure-time sports or physical exercise than among those who did not appears to have been more clearly observed in women. The larger proportion of strenuous work as a fraction of total physical activity in men than in women may be one reason for this discrepancy between men and women.

Our findings also showed that the effect of physical activity was diminished among subjects with a high body mass index, which is accordant with a previous report (3). To a substantial degree, physical activity may affect the risk of cancer by reducing weight and body mass index. We therefore suggest that the effect of physical activity appears less clear in persons with a high body mass index.

By site, our results showed inverse associations for colon, liver, and pancreatic cancer in men and for stomach cancer in women. In our population, we observed a positive association with a high body mass index for colon cancer only (23) and little association for pancreatic cancer (24). A recent evaluation found no association for stomach or liver cancer (1). In addition, nonalcoholic fatty liver disease, an increasingly recognized cause of chronic liver disease across the world, appears to be most strongly associated with central obesity and insulin resistance, and hepatocellular carcinoma has been postulated to arise through the development and progression of nonalcoholic fatty liver disease (25, 26). In the Japanese population, however, most cases of hepatocellular carcinoma are associated with hepatitis virus infection, and attribution to other factors may be small. Therefore, the effect of physical activity on these cancers, if any, appears to be operating not only via any improvement in obesity and related factors but also via other mechanisms.

Discussions on the possible mechanisms by which physical activity protects against cancer remain inconclusive. Various mechanisms have been plausibly associated with various cancers, such as alterations in sex hormones or insulin and insulin-like growth factors, immune modulation, alterations in free radical generation, changes in body fatness, and direct effects on cancer (1, 19, 27–32). Hyperinsulinemia produces an increase in circulating insulin-like growth factor 1, which is thought to play a major role in promoting carcinogenesis, and a decrease in insulin-like growth factor-binding proteins (33). Exercise increases insulin sensitivity and decreases fasting insulin and C-peptide levels (34), which may improve insulin resistance. Exercise-induced changes in the activity of macrophages, natural killer cells, lymphokine-activated killer cells, neutrophils, and regulating cytokines suggest that immunomodulation may contribute to the protective value of exercise (35). Strenuous physical exercise enhances oxygen free radical production, and the increased number of reactive oxygen species that are generated potentially results in damage to lipids, protein, and DNA. The antioxidant defense systems have co-evolved to counteract oxidative damage from oxygen free radicals (24, 36, 37). Moderate physical activity may be of benefit as a means of slowing or stopping the loss of antioxidants, whereas severe exercise might overwhelm the antioxidant system, potentially leading to damage and increased cell mutagenesis (37). Other mechanisms include a decrease in gut transit time, which has beneficial effects on bile content and secretion (1, 38), and have been proposed by site (1).

The major strength of the present study was its prospective design, which enabled us to avoid exposure recall bias. Study subjects were selected from the general population, the sample was large, the response rate to the questionnaire (81 percent) was acceptable for study settings such as this, and the loss to follow-up (0.3 percent) was negligible. Further, the number of exclusions due to missing data on physical activity (7 percent) was not particularly large. Although a difference in the characteristics of subjects with and without missing information had the potential to influence the results, no such difference was seen. In addition, the cancer registry in the study population was of sufficient quality to reduce the possibility of misclassification of the outcome.

In addition to those mentioned above, however, several methodological limitations can be identified. In particular, since assessment of physical activity was based on self-reports, misclassification may have been unavoidable. Nevertheless, because the data were collected before diagnosis, any imprecision is likely to have resulted in underestimation of the association. Changes in physical activity over time may also have caused misclassification, which might have led to underestimation of the association. In addition, some types of cancers or health conditions related to them may have caused low levels of physical activity from the starting point of the study; therefore, we cannot deny the possibility of spurious associations. Further, although adjustment was made for lifestyle factors possibly associated with cancer, unmeasured confounders may not have been controlled. Finally, our results may not be generalizable to populations with a different general lifestyle or a different degree of leanness from the Japanese.

Allowing for these methodological issues, our results suggest that increased daily total physical activity may be beneficial in preventing the development of cancer among Japanese men and women, who are characterized as relatively lean. Further research on the generalizability of our results to other relatively lean populations is warranted.

Abbreviations

    Abbreviations
  • CI

    confidence interval

  • MET(s)

    metabolic equivalent(s)

This study was supported by a Grant-in-Aid for Cancer Research, a Health and Labour Sciences Research Grant for Research on Hepatitis, and a Health and Labour Sciences Research Grant for the Third Term Comprehensive Control Research for Cancer (Ministry of Health, Labour and Welfare of Japan).

Members of the Japan Public Health Center-based Prospective Study Group (Principal Investigator, S. Tsugane): S. Tsugane, M. Inoue, T. Sobue, T. Hanaoka (National Cancer Center, Tokyo); J. Ogata, S. Baba, T. Mannami, A. Okayama, Y. Kokubo (National Cardiovascular Center, Osaka); K. Miyakawa, F. Saito, A. Koizumi, Y. Sano, I. Hashimoto, T. Ikuta (Iwate Prefectural Ninohe Public Health Center, Iwate); Y. Miyajima, N. Suzuki, S. Nagasawa, Y. Furusugi, N. Nagai (Akita Prefectural Yokote Public Health Center, Akita); H. Sanada, Y. Hatayama, F. Kobayashi, H. Uchino, Y. Shirai, T. Kondo, R. Sasaki, Y. Watanabe, Y. Miyagawa, Y. Kobayashi (Nagano Prefectural Saku Public Health Center, Nagano); Y. Kishimoto, E. Takara, T. Fukuyama, M. Kinjo, M. Irei, H. Sakiyama (Okinawa Prefectural Chubu Public Health Center, Okinawa); K. Imoto, H. Yazawa, T. Seo, A. Seiko, F. Ito, F. Shoji (Katsushika Public Health Center, Tokyo); A. Murata, K. Minato, K. Motegi, T. Fujieda (Ibaraki Prefectural Mito Public Health Center, Ibaraki); K. Matsui, T. Abe, M. Katagiri, M. Suzuki (Niigata Prefectural Kashiwazaki and Nagaoka Public Health Center, Niigata); M. Doi, A. Terao, Y. Ishikawa, T. Tagami (Kochi Prefectural Chuo-higashi Public Health Center, Kochi); H. Sueta, H. Doi, M. Urata, N. Okamoto, F. Ide (Nagasaki Prefectural Kamigoto Public Health Center, Nagasaki); H. Sakiyama, N. Onga, H. Takaesu, M. Uehara (Okinawa Prefectural Miyako Public Health Center, Okinawa); F. Horii, I. Asano, H. Yamaguchi, K. Aoki, S. Maruyama, M. Ichii, M. Takano (Osaka Prefectural Suita Public Health Center, Osaka); Y. Tsubono (Tohoku University, Miyagi); K. Suzuki (Research Institute for Brain and Blood Vessels Akita, Akita); Y. Honda, K. Yamagishi, S. Sakurai (Tsukuba University, Ibaraki); M. Kabuto (National Institute for Environmental Studies, Ibaraki); M. Yamaguchi, Y. Matsumura, S. Sasaki, S. Watanabe (National Institute of Health and Nutrition, Tokyo); M. Akabane (Tokyo University of Agriculture, Tokyo); T. Kadowaki (Tokyo University, Tokyo); M. Noda (International Medical Center of Japan, Tokyo); Y. Kawaguchi (Tokyo Medical and Dental University, Tokyo); Y. Takashima (Kyorin University, Tokyo); K. Nakamura (Niigata University, Niigata); S. Matsushima, S. Natsukawa (Saku General Hospital, Nagano); H. Shimizu (Sakihae Institute, Gifu); H. Sugimura (Hamamatsu University, Shizuoka); S. Tominaga (Aichi Cancer Center Research Institute, Aichi); H. Iso (Osaka University, Osaka); M. Iida, W. Ajiki, A. Ioka (Osaka Medical Center for Cancer and Cardiovascular Disease, Osaka); S. Sato (Osaka Medical Center for Health Science and Promotion, Osaka); E. Maruyama (Kobe University, Hyogo); M. Konishi, K. Okada, I. Saito (Ehime University, Ehime); N. Yasuda (Kochi University, Kochi); S. Kono (Kyushu University, Fukuoka).

Conflict of interest: none declared.

APPENDIX

Questions related to physical activity in the 5-year follow-up survey of the Japan Public Health Center-based Prospective Study:

How long on average do you engage in the following activities each day?

Heavy physical work or strenuous exercise None <1 hour ≥1 hour 
Sitting <3 hours 3–<8 hours ≥8 hours 
Standing or walking <1 hour 1–<3 hours ≥3 hours 
Heavy physical work or strenuous exercise None <1 hour ≥1 hour 
Sitting <3 hours 3–<8 hours ≥8 hours 
Standing or walking <1 hour 1–<3 hours ≥3 hours 

How often do you participate in sports or physical exercise?

Almost never ≤1–3 days a month 1–2 days a week 3–4 days a week Almost every day 
Almost never ≤1–3 days a month 1–2 days a week 3–4 days a week Almost every day 

References

1.
World Cancer Research Fund/American Institute for Cancer Research
Food, nutrition, physical activity, and the prevention of cancer: a global perspective
2007
Washington, DC
American Institute for Cancer Research
2.
Albanes
D
Blair
A
Taylor
PR
Physical activity and risk of cancer in the NHANES I population
Am J Public Health
 , 
1989
, vol. 
79
 (pg. 
744
-
50
)
3.
Wannamethee
SG
Shaper
AG
Walker
M
Physical activity and risk of cancer in middle-aged men
Br J Cancer
 , 
2001
, vol. 
85
 (pg. 
1311
-
16
)
4.
Lindsted
KD
Tonstad
S
Kuzma
JW
Self-report of physical activity and patterns of mortality in Seventh-Day Adventist men
J Clin Epidemiol
 , 
1991
, vol. 
44
 (pg. 
355
-
64
)
5.
Arraiz
GA
Wigle
DT
Mao
Y
Risk assessment of physical activity and physical fitness in the Canada Health Survey mortality follow-up study
J Clin Epidemiol
 , 
1992
, vol. 
45
 (pg. 
419
-
28
)
6.
Kampert
JB
Blair
SN
Barlow
CE
, et al.  . 
Physical activity, physical fitness, and all-cause and cancer mortality: a prospective study of men and women
Ann Epidemiol
 , 
1996
, vol. 
6
 (pg. 
452
-
7
)
7.
Hu
G
Tuomilehto
J
Silventoinen
K
, et al.  . 
The effects of physical activity and body mass index on cardiovascular, cancer and all-cause mortality among 47 212 middle-aged Finnish men and women
Int J Obes (Lond)
 , 
2005
, vol. 
29
 (pg. 
894
-
902
)
8.
Schnohr
P
Lange
P
Scharling
H
, et al.  . 
Long-term physical activity in leisure time and mortality from coronary heart disease, stroke, respiratory diseases, and cancer. The Copenhagen City Heart Study
Eur J Cardiovasc Prev Rehabil
 , 
2006
, vol. 
13
 (pg. 
173
-
9
)
9.
Matthews
CE
Jurj
AL
Shu
XO
, et al.  . 
Influence of exercise, walking, cycling, and overall nonexercise physical activity on mortality in Chinese women
Am J Epidemiol
 , 
2007
, vol. 
165
 (pg. 
1343
-
50
)
10.
Lam
TH
Ho
SY
Hedley
AJ
, et al.  . 
Leisure time physical activity and mortality in Hong Kong: case-control study of all adult deaths in 1998
Ann Epidemiol
 , 
2004
, vol. 
14
 (pg. 
391
-
8
)
11.
Sesso
HD
Invited commentary: a challenge for physical activity epidemiology
Am J Epidemiol
 , 
2007
, vol. 
165
 (pg. 
1351
-
3
)
12.
Tsugane
S
Sobue
T
Baseline survey of JPHC Study—design and participation rate. Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Diseases
J Epidemiol
 , 
2001
, vol. 
11
 
(suppl)
(pg. 
S24
-
9
)
13.
Tsubono
Y
Takamori
S
Kobayashi
M
, et al.  . 
A data-based approach for designing a semiquantitative food frequency questionnaire for a population-based prospective study in Japan
J Epidemiol
 , 
1996
, vol. 
6
 (pg. 
45
-
53
)
14.
World Health Organization
ICD-10: international statistical classification of diseases and related health problems
Tenth Revision
 , 
1990
1st ed
Geneva, Switzerland
World Health Organization
15.
World Health Organization
International classification of diseases for oncology
2000
Third Edition
Geneva, Switzerland
World Health Organization
16.
Ainsworth
BE
Bassett
DR
Jr
Strath
SJ
, et al.  . 
Comparison of three methods for measuring the time spent in physical activity
Med Sci Sports Exerc
 , 
2000
, vol. 
32
 
(suppl)
(pg. 
S457
-
64
)
17.
Stata Corporation
Stata statistical software, version 10
2007
College Station, TX
Stata Corporation
18.
Rockhill
B
Newman
B
Weinberg
C
Use and misuse of population attributable fractions
Am J Public Health
 , 
1998
, vol. 
88
 (pg. 
15
-
19
)
19.
Lee
IM
Physical activity and cancer prevention—data from epidemiologic studies
Med Sci Sports Exerc
 , 
2003
, vol. 
35
 (pg. 
1823
-
7
)
20.
Marugame
T
Matsuda
T
Kamo
K
, et al.  . 
Cancer incidence and incidence rates in Japan in 2001 based on the data from 10 population-based cancer registries
Jpn J Clin Oncol
 , 
2007
, vol. 
37
 (pg. 
884
-
91
)
21.
Shephard
RJ
Shek
PN
Associations between physical activity and susceptibility to cancer: possible mechanisms
Sports Med
 , 
1998
, vol. 
26
 (pg. 
293
-
315
)
22.
Blair
SN
Cheng
Y
Holder
JS
Is physical activity or physical fitness more important in defining health benefits?
Med Sci Sports Exerc
 , 
2001
, vol. 
33
 
(suppl)
(pg. 
S379
-
99
)
23.
Otani
T
Iwasaki
M
Inoue
M
, et al.  . 
Body mass index, body height, and subsequent risk of colorectal cancer in middle-aged and elderly Japanese men and women: Japan Public Health Center-based Prospective Study
Cancer Causes Control
 , 
2005
, vol. 
16
 (pg. 
839
-
50
)
24.
Luo
J
Iwasaki
M
Inoue
M
, et al.  . 
Body mass index, physical activity and the risk of pancreatic cancer in relation to smoking status and history of diabetes: a large-scale population-based cohort study in Japan—The JPHC Study
Cancer Causes Control
 , 
2007
, vol. 
18
 (pg. 
603
-
12
)
25.
Clark
JM
The epidemiology of nonalcoholic fatty liver disease in adults
J Clin Gastroenterol
 , 
2006
, vol. 
40
 
(suppl 1)
(pg. 
S5
-
10
)
26.
Bugianesi
E
Leone
N
Vanni
E
, et al.  . 
Expanding the natural history of nonalcoholic steatohepatitis: from cryptogenic cirrhosis to hepatocellular carcinoma
Gastroenterology
 , 
2002
, vol. 
123
 (pg. 
134
-
40
)
27.
Westerlind
KC
Physical activity and cancer prevention—mechanisms
Med Sci Sports Exerc
 , 
2003
, vol. 
35
 (pg. 
1834
-
40
)
28.
International Agency for Research on Cancer
Weight and physical activity. (IARC handbooks of cancer prevention, vol 6)
2002
Lyon, France
IARC Press
29.
Moore
MA
Park
CB
Tsuda
H
Physical exercise: a pillar for cancer prevention?
Eur J Cancer Prev
 , 
1998
, vol. 
7
 (pg. 
177
-
93
)
30.
Bullen
BA
Skrinar
GS
Beitins
IZ
, et al.  . 
Induction of menstrual disorders by strenuous exercise in untrained women
N Engl J Med
 , 
1985
, vol. 
312
 (pg. 
1349
-
53
)
31.
Cauley
JA
Gutai
JP
Kuller
LH
, et al.  . 
The epidemiology of serum sex hormones in postmenopausal women
Am J Epidemiol
 , 
1989
, vol. 
129
 (pg. 
1120
-
31
)
32.
Haffner
SM
Newcomb
PA
Marcus
PM
, et al.  . 
Relation of sex hormones and dehydroepiandrosterone sulfate (DHEA-SO4) to cardiovascular risk factors in postmenopausal women
Am J Epidemiol
 , 
1995
, vol. 
142
 (pg. 
925
-
34
)
33.
Nguyen
UN
Mougin
F
Simon-Rigaud
ML
, et al.  . 
Influence of exercise duration on serum insulin-like growth factor and its binding proteins in athletes
Eur J Appl Physiol Occup Physiol
 , 
1998
, vol. 
78
 (pg. 
533
-
7
)
34.
Regensteiner
JG
Mayer
EJ
Shetterly
SM
, et al.  . 
Relationship between habitual physical activity and insulin levels among nondiabetic men and women. San Luis Valley Diabetes Study
Diabetes Care
 , 
1991
, vol. 
14
 (pg. 
1066
-
74
)
35.
Shephard
RJ
Shek
PN
Cancer, immune function, and physical activity
Can J Appl Physiol
 , 
1995
, vol. 
20
 (pg. 
1
-
25
)
36.
Ji
LL
Antioxidants and oxidative stress in exercise
Proc Soc Exp Biol Med
 , 
1999
, vol. 
222
 (pg. 
283
-
92
)
37.
Dreher
D
Junod
AF
Role of oxygen free radicals in cancer development
Eur J Cancer
 , 
1996
, vol. 
32A
 (pg. 
30
-
8
)
38.
Sternfeld
B
Cancer and the protective effect of physical activity: the epidemiological evidence
Med Sci Sports Exerc
 , 
1992
, vol. 
24
 (pg. 
1195
-
209
)