REVIEWS Nature Reviews Cancer | AOP, published online 31 January 2008; doi:10.1038/nrc2325

Mechanisms linking physical activity with cancer Anne McTiernan

Abstract | About 25% of cancer cases globally are due to excess weight and a sedentary lifestyle. Physical activity may decrease risk for various cancers by several mechanisms, including decreasing sex hormones, metabolic hormones and inflammation, and improving immune function. The level of physical activity might also be associated with prognosis among individuals with cancer. Randomized clinical trials have shown that physical activity interventions can change biomarkers of cancer risk. Observational studies can also provide useful information on mechanisms that might link physical activity to cancer.

Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, Seattle, Washington 98109, USA. e-mail: [email protected] doi:10.1038/nrc2325 Published online 31 January 2008

The International Agency for Research on Cancer estimated in 2002 that up to one-third of cancers of the colon, breast, endometrium, oesophagus (adenocarcinoma) and kidney (renal cell) could be attributed to overweight or obesity and insufficient physical activity. They further estimated that, in the European Union, 21,000 cases of colon cancer and 13,000 cases of breast cancer could be avoided annually by maintaining a normal body weight1. Regular moderate-intensity or greater physical activity is associated with a reduced risk of several types of cancer, including cancers of the breast, colon and endometrium, and new evidence suggests that similar associations might exist for other cancers. For several cancers a dose–response relationship exists, such that engaging in longer exercise sessions, or exercising with greater intensity or for more years, produces greater reductions in cancer risk2. Understanding the mechanisms that link physical activity with cancer is useful for several reasons. First, if plausible mechanisms can be identified and replicated in experimental studies, then putative associations between physical activity and cancer can be supported. This can then provide evidence for implementing physical activity interventions on individual, clinical and public-health levels. Second, the identification of mechanisms can help with defining physical activity prescriptions for people without and with cancer. Third, identifying mechanisms can give new clues to cancer biology, which might help in designing other cancer prevention and treatment modalities.

Studies linking physical activity and cancer risk Few meta-analyses or combined analyses of physical activity and cancer aetiology have been published, probably owing to the diverse nature of the physical activity measures used by the different studies (BOX 1).

nature reviews | cancer

This Review focuses on mechanisms that have been tested in sufficiently powered human exercise intervention controlled clinical trials. Such trials have largely focused on mechanisms related to breast and colon cancers. Review papers suggest that active individuals have a relative risk for colon cancer of ~0.5–0.6 compared with sedentary persons3, and that women who exercise for 3–4 h/week at moderate-intensity or greater have a 30–40% reduced risk for breast cancer compared with sedentary women. Recent studies, including work from our institution, found a more moderate effect, in the order of a 20% risk reduction with the equivalent of walking approximately 45–60 min/day, 5–6 days/week4–6. Some studies suggest that the greatest cancer risk reduction from increased physical activity might be observed in normal-weight or overweight, versus obese, women6, and others have found that the protective effect of physical activity is limited to those without a family history of breast cancer4,5. Several new studies have pointed to an association between increased physical activity level and improved cancer prognosis, although data are limited to breast and colon cancer survivors. There are two published breast cancer survivor studies: one in a subset of the Nurses’ Health Study, and one in controls from the Women’s Healthy Eating and Living trial. In both studies, women who engaged in more than 9 metabolic equivalent (MET)·h/week of recreational physical activity after breast cancer diagnosis, which corresponds to approximately 2–3 h/week of brisk walking, had a significantly lower risk of breast cancer recurrence and all-cause mortality than women who were inactive7,8. Also in both studies, a decreased risk of recurrence and death associated with physical activity was observed in premenopausal and postmenopausal women and was independent of body mass index (BMI). advance volume online8 publication | march 2008 | 

© 2008 Nature Publishing Group

REVIEWS At a glance • Regular moderate-intensity or greater physical activity is associated with a reduced risk of several types of cancer, including cancers of the breast, colon and endometrium. For several cancers, engaging in longer exercise sessions, or exercising with greater intensity or for more years, produces greater reductions in cancer risk. • Recent findings indicate that women with a history of breast cancer who engage in more than 9 metabolic equivalent (MET)·h/week of recreational physical activity after breast cancer diagnosis, which corresponds to approximately 2–3 h/week of brisk walking, had a significantly lower risk of death or breast cancer recurrence than women who were inactive. Similarly, men or women who are active after a diagnosis of colon cancer appear to have improved prognosis compared with sedentary individuals. • Physical activity effects on menstrual function and sex hormones might explain a link between increased physical activity and reduced risk for breast and endometrial cancers. Athletes and physically active premenopausal women have delayed onset of menses, fewer ovulatory cycles and lower circulating levels of oestrogen and progesterone, all factors that are related to breast or endometrial cancer risk. • Studies in postmenopausal women indicate that physical activity might affect postmenopausal breast cancer and endometrial cancer risk by reducing body fat, thereby lowering circulating levels of oestrogens and androgens. • Insulin resistance, hyperinsulinaemia, hyperglycaemia and type 2 diabetes have been linked to increased risk of breast, colon, pancreas and endometrial cancers. Physical activity improves insulin resistance, reduces hyperinsulinaemia and reduces risk for diabetes, which could explain the link between increased physical activity and reduced risk for these cancers. • Increased levels of pro-inflammatory factors and decreased levels of antiinflammatory factors have been linked with increased cancer risk. Physical activity might reduce systemic inflammation alone or in combination with reduction in body weight or composition through reducing inflammatory cytokines in adipose tissue. • Support for an effect of physical activity on the cancer process comes from smaller randomized clinical trials. If definitive evidence from larger clinical trials is obtained to show that increased physical activity can prevent certain cancers, or can improve prognosis, it would be an excellent public-health intervention for reducing the impact of cancer at relatively low cost and risk.

The Nurses’ Health Study included 2,987 women with stage I–III breast cancer, who were followed for a median 96 months. Among women with a BMI > 25 kg/m2, which indicates overweight to obese, the relative risk (RR) of death from breast cancer for women who engaged in > 24 MET·h/week of physical activity (compared with < 3 MET·h/week) was 0.52 (95% confidence interval (CI) 0.26–1.06, P for trend 0.01)7. Among women with a BMI < 25.0 kg/m2, that is, those who were normal weight, the corresponding RR of death from breast cancer was 0.61 (95% CI 0.37–0.99, P for trend 0.10). Physical activity was also shown to be beneficial in women with all stages of disease, but was particularly beneficial in women diagnosed with higher stages of breast cancer. For women with stage III breast cancer who had engaged in ≥9 MET·h/week of physical activity compared with <9 MET·h/week, the RR was 0.36 (95% CI 0.19–0.71). Recently, in the Women’s Healthy Eating and Living study, an observational analysis of 1,490 women with early-stage breast cancer who were randomized to the control group of a dietary trial and followed for a mean 6.7 years, a combination of intake of five or more vegetables or fruits daily and the equivalent physical activity level of walking 30 min/day, 6 days/week was associated with a significant survival advantage (hazard ratio 0.56,  | advance march 2008 online | volume publication 8

95% CI 0.31–0.98)8. However, univariate associations showed stronger associations of physical activity with mortality than vegetable or fruit intake with mortality. The associations of these health behaviours with prognosis were limited to women with oestrogen receptorpositive tumours, which suggests that these behaviours might in part work through affecting oestrogen or related factors that can fuel breast tumour cell growth. A prospective study of 832 patients with stage III colon cancer followed for a mean of 3.8 years after enrolment in an adjuvant treatment trial also looked at the effects of exercise. The hazard ratio for diseasefree survival was 0.51 (95% CI 0.26–0.97) for patients engaging in 18–26.9 MET·h/week of exercise, with similar results for those exercising at greater levels, compared with sedentary patients (<3 MET·h/week)9. In a follow-up of 573 women from the Nurses’ Health Study cohort who developed stage I–III colon cancer and were followed for a mean 9.6 years from diagnosis, the hazard ratio for disease-free survival was 0.39 (95% CI 0.18–0.82) for patients engaging in ≥18 MET·h/ week of exercise, compared with sedentary patients (<3 MET·h/week)10. The explanation for the observed improved prognosis in these cancers with increased physical activity is unknown. It is not clear whether the ability to do physical activity is a marker for a higher health status in the patient, if physical activity has biological effects that can counter cancer progression, if physical activity increases the effectiveness or tolerability of cancer therapies or if the physical activity is reflective of a generally healthy lifestyle such as eating more fruit and vegetables, eating less meat and fat, drinking less alcohol and smoking less. If biological mechanisms are at play, the proposed connections between physical activity and cancer risk might also be germane for the links between physical activity and cancer progression11. There are several possible biological mechanisms through which physical activity could affect cancer risk or progression, as depicted in FIG. 1. Several mechanisms may act through the effects of physical activity on obesity, with resulting changes to circulating adipokines and cytokines, insulin resistance and blood insulin levels, and sex hormone production (Table 1). Other mechanisms could be mediated through physical activity effects on other body structures. For instance, in skeletal muscle physical activity results in reduced insulin resistance and reduced hyperinsulinaemia, even without changes in body composition12. Physical activity affects colon motility, leading to decreased transit time and, perhaps, reduced carcinogen exposure in the colon13. In addition, physical activity has been hypothesized to affect various tissues, leading to reduced carcinogenic prostaglandin production14. Another potential mechanism is through the postulated beneficial effects of chronic physical activity on DNA oxidative damage or repair15. The effects of physical activity on carcinogenesis are likely to be multifactorial and might be affected by many factors, such as age, gender and adiposity, in addition to physical activity-specific factors such as the type, duration, frequency and intensity of physical activity. www.nature.com/reviews/cancer

© 2008 Nature Publishing Group

REVIEWS Box 1 | Measurement and classification of physical activity Physical activity is defined for the purposes of this Review as purposive activity done as part of daily living, occupation, recreation or sports. It can be measured in a variety of ways, including self-report through questionnaires or logs, direct observation or objective measures such as step counters and tests of cardiopulmonary fitness. Numerous reviews of physical activity assessment have been conducted in the last decade, including two specific to cancer studies53,54. The most precise self-reported measurement methods collect information on type, intensity, frequency, duration and circumstance (for example, walking for pleasure versus transportation) of activity. Although objective measures such as step counters or fitness measures provide more precise estimates of activity, their costs are often prohibitive for epidemiological studies. Physical activity is typically quantified, for example, as h/week of activity or as metabolic equivalents (MET)·h/week, where MET is the estimated intensity of an activity. There is no consensus on what defines sedentary behaviour for either observational or intervention studies. In some studies, the bottom quartile of activity level defines sedentary, whereas in others people not meeting the U.S. Surgeon General’s recommended 150 min/week of moderate activity55 are considered sedentary.

Sex steroid hormones Women with increased levels of oestrogens and androgens have an increased risk of developing breast cancer16, and those with increased oestrogen concentrations (unopposed by progesterone) are at an increased risk for endometrial cancer17. In men, anti-androgen therapy improves prostate cancer survival18 and reduces the overall incidence of the disease when tested as a preventive agent19. Premenopausal women. The effects of physical activity on the age at menarche, menstrual cycle function and the level of endogenous sex steroid hormone levels in girls and young women are often cited as potential mechanisms for reduced breast cancer risk13. Previous literature suggests that exercise may cause minor shifts in the hormonal milieu of premenopausal women. However, a programme of significant intensity and volume, or exercise combined with dietary caloric restriction, is needed to induce menstrual dysfunction sufficient to result in significantly decreased exposure to sex steroid hormones. Recent research suggests that exercise has no disruptive effect on reproductive function beyond that of its effect on energy availability20–22. Observational research points to a delayed age at menarche, a continuum of menstrual dysfunction (amenorrhoea, anovular cycles and luteal phase deficiency), longer menstrual cycle length and lower progesterone and oestradiol levels in athletes compared with non-athletes or sedentary women13. However, the individual effects of weight control and physical activity are difficult to discern20,21. Prospective intervention studies in premenopausal women have been limited in number and used small sample sizes; results were mixed. Two of these studies found that a moderate-intensity running intervention did not disrupt reproductive function 23,24. Other studies reported minor changes in measures of reproductive function, including decreased ovulation or lengthening of menstrual cycle25,26, or induced menstrual dysfunction with significant exercise intensity and volume27. nature reviews | cancer

Postmenopausal women. In postmenopausal women, higher physical activity has been associated with lower serum concentrations of oestradiol, oestrone and androgens after adjustment for BMI in most reported observational studies28–30. The positive effect of physical activity is closely linked to body composition as the primary source of oestrogen in postmenopausal women is from aromatization of androgen precursors in peripheral, mainly adipose, tissue. In a subsample from the Women’s Health Initiative Dietary Modification Trial we found that women with low self-reported physical activity had higher levels of oestrone, oestradiol and free oestradiol, and lower levels of sex hormone-binding globulin (which binds oestradiol, making less of it available to the target tissue) than active women29 (FIG. 2). The highest levels of oestrogen were observed in women who were both below the median level for physical activity and above the median BMI. We conducted a physical activity trial in 173 overweight, sedentary postmenopausal women (BMI ≥ 24 kg/m2 and percentage body fat > 33%), in which women were randomized to a moderate-intensity aerobic exercise, 45 min/day, 5 days/week for 12 months or to a control group. A significant decrease in oestradiol, oestrone and free oestradiol was seen from baseline to 3 months, with an attenuation of the effect at 12 months31. However, in those women who lost body fat, the exercise intervention resulted in a statistically significant reduction in these oestrogens at both 3 and 12 months (FIG. 3a). Similarly, in women who lost body fat, there was a statistically significant decrease in testosterone and free testosterone in exercisers compared with controls32 (FIG. 3b). These results suggest that both increased physical activity and reduced body fat will produce the greatest protection against breast cancer by producing the greatest decrease in serum sex hormones. Men. Few observational studies have examined the association of physical activity levels with sex hormones in men. Chronically lowered testosterone concentrations have been reported in athletes, but this finding might require a threshold amount or intensity of physical activity to occur33. Excess levels of some androgens have been postulated to increase risk for prostate cancer, and the testosterone-reducing medication finasteride reduces the incidence of prostate cancer, although the picture is mixed as finasteride also increased the incidence of high-grade prostate tumours19. Furthermore, the association of obesity with lowered testosterone in men further complicates the issue of the effect, if any, of physical activity on testosterone34. The effects of a moderate-intensity aerobic exercise intervention on sex steroid hormones in previously sedentary middle-aged or older men are not known; however, studies are underway to examine this question. We recently tested the effect of aerobic exercise on androgens in 102 sedentary men, aged 40–75 years, whom we randomly assigned to a 12 month moderate/vigorous intensity aerobic exercise intervention (60 min/day, 6 days/week) or a control group (no change in activity)35. Dihydrotestosterone (DHT) increased 14.5% in exercisers advance volume online8 publication | march 2008 | 

© 2008 Nature Publishing Group

REVIEWS Decreased oestrogens and androgens

Improved immune function

Increased physical activity

Reduced adiposity

Decreased risk of cancer

Decreased insulin and glucose

Altered adipocytokines (↑adiponectin, ↓leptin, ↓inflammation)

Figure 1 | Hypothesized mechanisms linking physical activity to cancer risk or prognosis. Physical activity Nature Reviews | Cancer might work through reducing the amount of adipose tissue, which lowers production of sex hormones, insulin, leptin and inflammatory markers, thereby decreasing the exposure to these potentially carcinogenic hormones and peptides and reducing cancer risk.

versus 1.7% in controls at 3 months (P = 0.04), and at 12 months it remained 8.6% above the baseline in exercisers versus a 3.1% decrease in controls (P = 0.03). Sex hormone-binding globulin increased 14.3% in exercisers versus 5.7% in controls at 3 months (P = 0.04), and at 12 months it remained 8.9% above the baseline in exercisers versus 4.0% in controls (P = 0.13). No statistically significant differences were observed for testosterone, free testosterone, 3α androstanediol glucuronide, oestradiol or free oestradiol in exercisers versus controls. Although the associations between blood levels of sex hormones and prostate cancer risk are not established, one recent epidemiological study found low levels of DHT to be associated with decreased prostate cancer survival among men with screening-detected cancer36; further, lower levels of intra-prostatic DHT have been Table 1 | Physical activity and cancer aetiology or prognosis: mechanisms Mechanism

Cancers potentially affected by mechanism

Sex hormones

Breast, endometrium, prostate

Insulin, glucose

Colon, breast, pancreas and several other cancers related to obesity, such as lower oesophagus, renal, thyroid and endometrium

Inflammation

Most cancer types

Immune function

Any, although the links between physical activity and cancers linked to immune function have not been well studied, and are therefore unknown

Adipokines

Several cancers related to obesity, such as colon, postmenopausal breast, lower oesophagus, renal, thyroid and endometrium

 | advance march 2008 online | volume publication 8

associated with greater cancer aggressiveness37. Thus increasing DHT through exercise could potentially be beneficial for prostate health.

Metabolic hormones Insulin resistance has been linked to increased risk of breast, colon, pancreas and endometrial cancers38. Higher cancer incidence and mortality have also been noted in those with type 2 diabetes mellitus or impaired glucose tolerance38,39. Insulin can enhance tumour development by stimulating cell proliferation or inhibiting apoptosis38. It can also regulate the synthesis and biological availability of sex steroid hormones, and inhibit hepatic synthesis of sex hormone-binding globulin38. Acute bouts of physical activity improve insulin sensitivity and increase glucose uptake by skeletal muscle, even in those with type 2 diabetes mellitus, for up to 12 h40. Chronic exercise training results in prolonged improvements in insulin sensitivity as we12 (FIG. 3c) and others41,42 have shown. Although body composition has been strongly associated with insulin sensitivity, exercise-induced changes in insulin sensitivity can occur independently of the changes in body weight or body composition that are associated with physical activity12,40,43 (FIG. 3c). A greater effect of resistance training compared with aerobic activity with similar levels of energy expenditure to improve insulin sensitivity and glycaemic control has also been proposed, as skeletal muscle is a primary site of insulin resistance40,44. However, similar improvements in insulin resistance for diet-induced or www.nature.com/reviews/cancer

© 2008 Nature Publishing Group

REVIEWS

Inflammation Increased levels of pro-inflammatory factors, such as C-reactive protein (CRP), interleukin 6 (IL6), tumour necrosis factor α (TNFα), and decreased levels of antiinflammatory factors, such as adiponectin, have been linked with increased cancer risk48. Physical activity might reduce systemic inflammation alone or in combination with a reduction in body weight or change in body composition by reducing the macrophage or adipose cell production of inflammatory cytokines in adipose tissue, although the exact mechanism(s) are unknown44,49. Although cross-sectional studies support an association between chronic physical activity and lower levels of the inflammatory markers CRP, serum amyloid A (SAA), IL6 and TNFα in both men and women, intervention studies of exercise alone or exercise and diet combined have shown reductions in these inflammatory markers in some but not all studies49. Body composition appears to have a significant influence on markers of inflammation in observational studies and physical activity intervention studies49. Observational studies have reported that lower levels of adiponectin, an anti-inflammatory factor, are associated with higher BMI, higher percentage body fat, larger waist circumference and greater amount of intraabdominal fat44. Increases in adiponectin have been seen with physical activity interventions in the presence of significant weight loss50. Shorter-duration prospective physical activity and weight loss interventions have failed to alter adiponectin levels despite modest changes in body weight and body composition44. Immune function The immune system is suggested to have a role in reducing cancer risk, by recognition and elimination of abnormal cells or through acquired and/or innate immune system components51. The effect of physical activity on immune factors related to cancer has been largely untested, but one hypothesis is that physical activity could improve the number or function of natural killer cells, which have a role in tumour suppression49. Bouts of exercise have been shown to result in acute increases in a number of components of immune function (for example, neutrophils, monocytes, eosinophils and lymphocytes), followed by a dip below pre-exercise levels lasting up to 1–3 h52. For chronic physical activity, an inverted J-shaped dose–response relationship between intensity of physical activity and immune function has been shown. Moderate physical activity results in enhanced immune function, whereas exhausted exercise, nature reviews | cancer

35 Serum concentration (pg/ml)

exercised-induced weight loss groups have been observed in intervention studies, which suggests that weight loss from any mechanism improves insulin resistance and lowers hyperinsulinaemia and hyperglycaemia41,42,45. Several lines of epidemiological and experimental evidence link increased levels of insulin-like growth factors with increased risk for several cancers38. However, observational and experimental evidence in humans does not support an effect of physical activity on these factors46,47.

30 25 20 15 10 5 0

Oestrone

Oestradiol

Low BMI, high PA

High BMI, high PA

Low BMI, low PA

High BMI, low PA

Figure 2 | Associations of body mass index (BMI)|and Nature Reviews Cancer physical activity (PA) with serum oestrogens in postmenopausal women: Women’s Health Initiative (n = 267)29.

overtraining or high-intensity exercise may lead to immunosuppression, such as increased susceptibility to upper respiratory tract infections49. However, the current evidence on moderate-intensity physical activity from randomized, controlled trials is inconclusive, and differences in the components of the innate immune system have been noted in some, but not all, cross-sectional studies comparing exercisers with non-exercisers49.

Future directions Several randomized clinical trials are underway, testing the effect of physical activity on various biomarkers of cancer risk. Two are assessing the effect of physical activity interventions on sex hormones in postmenopausal women (the ALPHA trial, Alberta, Canada, principal investigator C. Friedenreich; and the SHAPE trial, the Netherlands, principal investigator E. Monninkof) and should by 2008 provide additional data to that from the trial described above31,32. A clinical trial is underway to test the effect of 1 year of moderate-intensity physical activity, with and without dietary weight loss, on endogenous sex hormones and other breast cancer biomarkers (the NEW trial, Seattle, Washington, USA, principal investigator A.M.). A total of 503 postmenopausal, sedentary women who are overweight or obese will be randomly assigned to a group with a moderate-intensity exercise programme (45 min/day, 5 days/week), one with a reduced-calorie diet (with goal weight loss of 10% of starting weight), a group given both the exercise and the diet interventions and a control group (no change to diet or exercise). As of December 2007, 353 women have been randomized, and results are intended to be available in 2009–2010. The WISER trial in Minneapolis, Minnesota, USA, (principal investigator M. Kurzer, co-principal investigator K. Schmitz) is examining the effects of aerobic exercise training on various cancer biomarkers among 400 young (18–30 years), premenopausal, eumenorrhoeic women in a randomized, controlled trial. The biomarkers assessed advance volume online8 publication | march 2008 | 

© 2008 Nature Publishing Group

REVIEWS

Change in oestradiol (%)

a

20

5

–5 –10 –15

Change in testosterone (%)

Gained

No change

Lost 0.5–2%

Lost >2%

Gained

No change

Lost 0.5–2%

Lost >2%

20 15 10 5 0 –5 –10 –15 –20

c

Exercisers

0

–20

b

Controls

15 10

20

Change in insulin (%)

15 10 5 0 –5 –10 –15 –20

Gained

Lost 0.5–2% Change in body fat

Lost >2%

Figure 3 | Hormone changes by change in body fat Nature Reviews | Cancer with physical activity in sedentary or overweight postmenopausal women (aged 50–75 years). a | Percent change in oestradiol by percent change in body fat as measured by dual-energy X-ray absorptiometry (DEXA) scan. A statistically significant difference in oestradiol level was observed in exercisers who lost 0.5–2% body fat (P = 0.02) and for those who lost > 2% body fat (P = 0.008) compared with controls at 12 months31. b | Percent change in testosterone by percent change in body fat as measured by DEXA scan. A statistically significant difference in testosterone level was observed in exercisers who lost 0.5–2% body fat (P < 0.05) and for those who lost >2% body fat (P<0.05) compared with controls at 12 months32. c | Percent change in insulin by percent change in body fat as measured by DEXA scan. A statistically significant difference in insulin level was observed in exercisers who lost 0.5–2% body fat and for those who lost > 2% body fat compared with controls at 12 months (P = 0.03)12.

include oxidative stress as measured by F2-isoprostanes, luteal phase length or ovulatory status, oestrogen metabolites, body composition (by dual-energy X-ray absorptiometry), insulin, glucose, insulin resistance (homeostasis model assessment index) and insulin-like growth factor  | advance march 2008 online | volume publication 8

axis proteins (insulin-like growth factor 1 (IGF1), IGF-binding protein 1 (IGFBP1), IGFBP2 and IGFBP3). Additional clinical trials are needed to establish the effect of physical activity on other cancer biomarkers. For any cancer biomarker, it would be important to know the effects of specific exercises and patterns, including type, frequency and duration of exercise, as well as the modifying effects of weight loss either through exercise alone or through dietary or other means. It is particularly important to determine whether physical activity acts independently of weight loss or weight control, or whether weight control is required for physical activity effects on cancer risk or prognosis. Our observational and intervention research on physical activity effects on breast cancer and on sex hormones suggests that the optimal physical activity effect is seen in normal weight women or in overweight or obese women who lose weight as part of an exercise intervention. Assessment of other modifying factors, including gender, age, race/ethnicity, level of risk for cancer (determined by, for example, genetics, family history and history of pre-cancerous conditions) and other lifestyle factors, would also be important. In addition, the effect of physical activity on prognosis among persons with cancer should be studied. Funding for the U.S. National Cancer Institute’s Transdisciplinary Research on Energetics and Cancer (TREC) centres began in 2005 to identify and study mechanisms linking energy balance, including physical activity, to cancer. Four clinical centres (Fred Hutchinson Cancer Research Center, Seattle, Washington, USA; University of Southern California, Los Angeles, California, USA; University of Minnesota, Minnesota, USA; and Case Western Reserve, Cleveland, Ohio, USA), a coordinating centre (Fred Hutchinson Cancer Research Center) and several independent and pilot projects have been funded through this programme. The TREC centres are intended to foster collaboration among multidisciplinary teams of scientists with the goal of accelerating progress toward reducing the cancer incidence, morbidity and mortality that is associated with obesity, low levels of physical activity and poor diet. They also aim to provide training opportunities for new and established scientists who can carry out integrative research on energetics and energy balance. Early results from the various TREC projects should be available in 2008.

Conclusions Physical activity in observational studies has been associated with risk and prognosis for various cancers, and several mechanisms have been proposed to explain the links between physical activity and cancer. Support for an effect of physical activity on the cancer process comes from smaller randomized clinical trials, and confirmation of the role of physical activity on cancer risk and prognosis from clinical trial evidence is awaited with interest. If definitive evidence is obtained that increased physical activity can prevent certain cancers, or can improve prognosis, it would be an excellent publichealth intervention for reducing the impact of cancer at relatively low cost and risk. www.nature.com/reviews/cancer

© 2008 Nature Publishing Group

REVIEWS 1.

2.

3.

4.

5.

6.

7.

8.

9.

10. 11. 12.

13.

14. 15.

16.

17.

18. 19. 20. 21.

Vainio, H. & Bianchin, I. F. in International Agency for Research on Cancer Handbooks of Cancer Prevention 249–250 (International Agency for Research on Cancer, Lyon, 2002). Thune, I. & Furberg, A. S. Physical activity and cancer risk: dose–response and cancer, all sites and sitespecific. Med. Sci. Sports Exerc. 33, S530–S550 (2001). Friedenreich, C. M. & Orenstein, M. R. Physical activity and cancer prevention: etiologic evidence and biological mechanisms. J. Nutr. 132, 3456S–3464S (2002). Bernstein, L. et al. Lifetime recreational exercise activity and breast cancer risk among black women and white women. J. Natl Cancer Inst. 97, 1671–1679 (2005). Dallal, C. M. et al. Long-term recreational physical activity and risk of invasive and in situ breast cancer: the California teachers study. Arch. Intern. Med. 167, 408–415 (2007). McTiernan, A. et al. Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women’s Health Initiative Cohort Study. JAMA 290, 1331–1336 (2003). Holmes, M. D., Chen, W. Y., Feskanich, D., Kroenke, C. H. & Colditz, G. A. Physical activity and survival after breast cancer diagnosis. JAMA 293, 2479–2486 (2005). Pierce, J. P. et al. Greater survival after breast cancer in physically active women with high vegetable-fruit intake regardless of obesity. J. Clin. Oncol. 25, 2345–2351 (2007). Meyerhardt, J. A. et al. Impact of physical activity on cancer recurrence and survival in patients with stage III colon cancer: findings from CALGB 89803. J. Clin. Oncol. 24, 3535–3541 (2006). Meyerhardt, J. A. et al. Physical activity and survival after colorectal cancer diagnosis. J. Clin. Oncol. 24, 3527–3534 (2006). McTiernan, A. Physical activity after cancer: physiologic outcomes. Cancer Invest. 22, 68–81 (2004). Frank, L. L. et al. Effects of exercise on metabolic risk variables in overweight postmenopausal women: a randomized clinical trial. Obes. Res. 13, 615–625 (2005). McTiernan, A., Ulrich, C., Slate, S. & Potter, J. Physical activity and cancer etiology: associations and mechanisms. Cancer Causes Control 9, 487–509 (1998). Martinez, M. E. et al. Physical activity, body mass index, and prostaglandin E2 levels in rectal mucose. J. Natl Cancer Inst. 91, 950–951 (1999). Meijer, E. P., Goris, A. H., van Dongen, J. L., Bast, A. & Westerterp, K. R. Exercise-induced oxidative stress in older adults as a function of habitual activity level. J. Am. Geriatr. Soc. 50, 349–353 (2002). Endogenous Hormones and Breast Cancer Collaborative Group. Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J. Natl Cancer Inst. 94, 606–616 (2002). Kaaks, R., Lukanova, A. and Kurzer, M. S. Obesity, endogenous hormones and endometrial cancer risk: a synthetic review. Cancer Epidemiol. Biomarkers Prev. 11, 1531–1543 (2002). Tammela, T. Endocrine treatment of prostate cancer. J. Steroid Biochem. Mol. Biol. 92, 287–295 (2004). Thompson, I. M. et al. The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 349, 215–224 (2003). Loucks, A. B. Energy availability, not body fatness, regulates reproductive function in women. Exerc. Sport Sci. Rev. 31, 144–148 (2003). Loucks, A. B. & Redman, L. M. The effect of stress on menstrual function. Trends Endocrinol. Metab. 15, 466–471 (2004).

22. Williams, N. I., Helmreich, D. L., Parfitt, D. B., CastonBalderrama, A. & Cameron, J. L. Evidence for a causal role of low energy availability in the induction of menstrual cycle disturbances during strenuous exercise training. J. Clin. Endocrinol. Metab. 86, 5184–5193 (2001). 23. Bonen, A. Recreational exercise does not impair menstrual cycles: a prospective study. Int. J. Sports Med. 13, 110–120 (1992). 24. Rogol, A. D. et al. Durability of the reproductive axis in eumenorrheic women during 1 yr of endurance training. J. Appl. Physiol. 72, 1571–1580 (1992). 25. Bullen, B. A. et al. Endurance training effects on plasma hormonal responsiveness and sex hormone excretion. J. Appl. Physiol. 56, 1453–1463 (1984). 26. Keizer, H. A. et al. Effect of a 3-month endurance training program on metabolic and multiple hormonal responses to exercise. Int. J. Sports Med. 8 (Suppl. 3), 154–160 (1987). 27. Bullen, B. A. et al. Induction of menstrual disorders by strenuous exercise in untrained women. N. Engl. J. Med. 312, 1349–1353 (1985). 28. Chan, M. F. et al. Usual physical activity and endogenous sex hormones in postmenopausal women: the European prospective investigation into cancer — Norfolk population study. Cancer Epidemiol. Biomarkers Prev. 16, 900–905 (2007). 29. McTiernan, A. et al. Relation of BMI and physical activity to sex hormones in postmenopausal women. Obesity (Silver Spring) 14, 1662–1677 (2006). 30. Verkasalo, P. K., Thomas, H. V., Appleby, P. N., Davey, G. K. & Key, T. J. Circulating levels of sex hormones and their relation to risk factors for breast cancer: a cross-sectional study in 1092 pre- and postmenopausal women (United Kingdom). Cancer Causes Control 12, 47–59 (2001). 31. McTiernan, A. et al. Effect of exercise on serum estrogen in postmenopausal women: a 12-month randomized controlled trial. Cancer Res. 64, 2923–2928. (2004). 32. McTiernan, A. et al. Effect of exercise on serum androgens in postmenopausal women: a 12-month randomized clinical trial. Cancer Epidemiol. Biomarkers Prev. 13, 1–7 (2004). 33. De Souza, M. J. & Miller, B. E. The effect of endurance training on reproductive function in male runners. A ‘volume threshold’ hypothesis. Sports Med. 23, 357–374 (1997). 34. Abate, N., Haffner, S. M., Garg, A., Peshock, R. M. & Grundy, S. M. Sex steroid hormones, upper body obesity, and insulin resistance. J. Clin. Endocrinol. Metab. 87, 4522–4527 (2002). 35. Hawkins, V. N. et al. Effect of exercise on serum sex hormones in men: a 12-month randomized clinical trial. Med. Sci. Sports Exerc. 40, 223–233 (2008). 36. Kjellman, A., Akre, O., Norming, U., Tornblom, M. & Gustafsson, O. Dihydrotestosterone levels and survival in screening-detected prostate cancer: a 15-yr followup study. Eur. Urol. 53, 106–111 (2007). 37. Nishiyama, T., Ikarashi, T., Hashimoto, Y., Suzuki, K. & Takahashi, K. Association between the dihydrotestosterone level in the prostate and prostate cancer aggressiveness using the Gleason score. J. Urol. 176, 1387–1391 (2006). 38. Kaaks, R. & Lukanova, A. Energy balance and cancer: the role of insulin and insulin-like growth factor-I. Proc. Nutr. Soc. 60, 91–106 (2001). 39. Wolf, I., Sadetzki, S., Catane, R., Karasik, A. & Kaufman, B. Diabetes mellitus and breast cancer. Lancet Oncol. 6, 103–111 (2005). 40. Boule, N. G., Haddad, E., Kenny, G. P., Wells, G. A. & Sigal, R. J. Effects of exercise on glycemic control and body mass in type 2 diabetes mellitus: a meta-analysis of controlled clinical trials. JAMA 286, 1218–1227 (2001). 41. Ross, R. et al. Reduction in obesity and related comorbid conditions after diet-induced weight loss or

nature reviews | cancer

42. 43.

44. 45.

46.

47. 48.

49.

50.

51. 52. 53.

54.

55.

exercise-induced weight loss in men: a randomized, controlled trial. Ann. Intern. Med. 133, 92–103 (2000). Ross, R. et al. Exercise-induced reduction in obesity and insulin resistance in women: a randomized controlled trial. Obes Res. 12, 789–798 (2004). Duncan, G. E. et al. Exercise training, without weight loss, increases insulin sensitivity and postheparin plasma lipase activity in previously sedentary adults. Diabetes Care 26, 557–562 (2003). Campbell, K. L. & McTiernan, A. Exercise and biomarkers for cancer prevention studies. J. Nutr. 137, 161S-9S (2007). Weiss, E. P. et al. Improvements in glucose tolerance and insulin action induced by increasing energy expenditure or decreasing energy intake: a randomized controlled trial. Am. J. Clin. Nutr. 84, 1033–1042 (2006). McTiernan, A. et al. No effect of exercise on insulinlike growth factor 1 and insulin-like growth factor binding protein 3 in postmenopausal women: a 12month randomized clinical trial. Cancer Epidemiol. Biomarkers Prev. 14, 1020–1021 (2005). Orenstein, M. R. & Friedenreich, C. Insulin-like growth factors and exercise: review of the evidence. J. Phys. Act. Health 1, 291–320 (2004). Il’yasova, D. et al. Circulating levels of inflammatory markers and cancer risk in the health aging and body composition cohort. Cancer Epidemiol. Biomarkers Prev. 14, 2413–2418 (2005). Wetmore, C. M. & Ulrich, C. M. in Cancer Prevention and Management through Exercise and Weight Control (ed. McTiernan, A.) 157–175 (CRC Taylor Francis, Boca Raton, 2006). Esposito, K. et al. Effect of weight loss and lifestyle changes on vascular inflammatory markers in obese women: a randomized trial. JAMA 289, 1799–1804 (2003). Jakobisiak, M., Lasek, W. & Golab, J. Natural mechanisms protecting against cancer. Immunol. Lett. 90, 103–122 (2003). Nieman, D. C. Exercise immunology: practical applications. Int. J. Sports Med. 18 (Suppl. 1), S91–S100 (1997). Ainsworth, B. E., Sternfeld, B., Slattery, M. L., Daguise, V. & Zahm, S. H. Physical activity and breast cancer: evaluation of physical activity assessment methods. Cancer 83, 611–620 (1998). Irwin, M. L. & Ainsworth, B. E. Physical activity interventions following cancer diagnosis: methodologic challenges to delivery and assessment. Cancer Invest. 22, 30–50 (2004). U. S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General (U. S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Atlanta, 1996).

DATABASES Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db=gene CRP | IGF1 | IGFBP1 | IGFBP2 | IGFBP3 | IL6 | SAA | TNFα National Cancer Institute: http://www.cancer.gov/ breast cancer | colon cancer | endometrial cancer | kidney cancer | oesophageal cancer | pancreatic cancer | prostate cancer National Cancer Institute Drug Dictionary: http://www.cancer.gov/drugdictionary/ finasteride

FURTHER INFORMATION Anne McTiernan’s homepage: http://www.fhcrc.org/ science/phs/cprp/faculty/mctiernan.html National Cancer Institute Transdisciplinary Research on Energetics and Cancer (TREC): https://www.compass.fhcrc.org/trec/ All links are active in the online pdf

advance volume online8 publication | march 2008 |  © 2008 Nature Publishing Group