Epidemiological association between physical activity and cancers

Which cancers are linked with inactivity?

Evidence linking inactivity and a variety of cancers has grown over the last decade (Thune and Furberg, 2001). The 1996 Surgeon General report on physical activity (US Department of Health and Human Services, 1996) noted that:

• Regular physical activity is associated with a decreased risk of colon cancer
• There was no association between physical activity and rectal cancer
• Data were too sparse to draw conclusions regarding a relationship between physical activity and endometrial, ovarian or testicular cancers
• Despite numerous studies, existing data are inconsistent regarding an association with breast or prostate cancers.

More recently, a number of authors have reviewed the epidemiological evidence around physical activity and specific cancers, and there are an increasing number of reviews of cancer risk in general.

A Canadian workshop on physical activity and cancer prevention in March 2000 presented a systematic review of the published literature on the aetiological role of physical activity in relation to cancer (Marrett, L. et al., 2000). It included an assessment of the frequency, intensity and duration of physical activity associated with cancer risk reduction.

The review divided the overall weight of evidence for each cancer into four categories. These are:

• Convincing – epidemiological studies show consistent associations, with little or no evidence to the contrary. There should be a substantial number of acceptable studies (20+), preferably including prospective designs, conducted in different population groups and controlled for possible confounding factors. Exposure data should refer to the time preceding the occurrence of cancer. Dose response relationships should be supportive of a causal relationship. Associations should be biologically plausible. Laboratory evidence is usually supportive.

• Probable – epidemiological studies showing associations are either not so consistent, with a
  number of studies not supporting the association, or the number or type of studies is not extensive enough to make a more definite judgement. Mechanistic and laboratory evidence is usually supportive.
• Possible – epidemiological studies are generally supportive, but are limited in quantity, quality or consistency. There may not be supportive mechanistic or laboratory evidence. Alternatively, there are few or no epidemiological data, but strongly supportive evidence from other disciplines.
• Insufficient – there are only a few studies, which are generally consistent, but really do no more than hint at a possible relationship. Often, more well-designed research is needed.

These categories are based on an adaptation of the World Cancer Research Fund/American Institute for Cancer Research categories used in a review of nutrition and cancer prevention. The table above summarises this assessment of the strength of the evidence in seven cancers.

Possible mechanisms

Various mechanisms have been suggested as being involved. These include restriction of activity by preexisting disease, dietary influences including overall energy balance, intake and bioavailability of minerals, antioxidant vitamins and fibre and relative proportions of protein and fat ingested.

Links between regular activity and other cancer risks are generally not strong, although endurance athletes are not usually smokers and regular leisure activity is associated with high socio-economic status, which the authors note tends to reduce exposure to airborne carcinogens.

Obesity is a major component of the exercise-cancer relationship (see Chapter 3). A variety of biological mechanisms have been suggested, including changes in hormonal influences, changes in immune functioning and speeding up of passage of faecal matter through the colon (see eg Shephard and Shek,1998; US Department of Health and Human Services, 1996; Thune and Furberg, 2001).

What level of activity is needed?

The general physical activity recommendation is for adults to achieve a minimum of 30 minutes of moderate activity at least five times a week (DH, 1996), and is appropriate as a baseline in reducing the risk of cancers. Reviews of evidence looking at cancer risk at different levels of physical activity have found dose-response effects, showing reduced risks as the level of activity increases. A dose-response effect is one in which a greater amount of what is being studied (here physical activity level) is associated with a larger response. The presence of a dose-response effect is an important step in suggesting that two factors (here physical activity and cancer) are causally related.

The dose-response effect was particularly seen when the subjects took part in activity of at least moderate intensity. This was found particularly when looking at colon cancer (Thune and Furberg, 2001). The review by Marrett et al. (2000) notes that physical activity should comprise at least 30-45 minutes of moderate to vigorous activity on most days of the week.

What is the public health impact of inactivity on cancer?

To understand the impact of a risk factor such as inactivity on the population as a whole it is necessary to know both the relationship between inactivity and health outcomes (in this case cancers) and the prevalence of inactivity in that population. Because of the high prevalence of inactivity (see table below) there is a substantial population-wide benefit from increasing physical activity. This is frequently presented as a population attributable risk (PAR). It is calculated from the prevalence of a risk factor in that population (here the risk factor being inactivity) and the relative risk associated with that factor. Using data from the US, Powell and Blair (1994) estimated the PAR for inactivity in relation to colon cancer to be 32%. As levels of inactivity are similar in this country to those in the US, it is likely that a figure for England would be similar. A full description of the application of population attributable risk estimates in relation to physical activity can be seen in Macera and Powell (2001). summarised into a frequency-duration scale, which took account of average time spent participating in physical activities, and the number of active days in the last week. This summary measure showed that overall participation was lower among girls than among boys. In both sexes there was a decline after about age eight, but the decline was steeper among girls than boys. By age 15, only 36% of girls engaged in physical activities for at least 30 minutes on most days. In contrast, 71% of boys aged 15 participated in physical activities for at least 30 minutes on most days.

Related Articles

• Increasing physical activity
• Features of effective interventions
• Epidemiological association between physical activity and cancers
• Objectives of physical activity interventions