Wednesday November 7 2007
Does weight affect cancer?
Being overweight raises the risk of 10 different types of cancer in women, reported the Daily Express and other newspapers. “A study has found that one in 20 cancers among middle-aged or older women are triggered by their weight”, the newspaper said.
In particular, the effect “is greatest in cancers of the oesophagus (gullet) and endometrium (lining of the womb) where the risks are almost doubled”, stated The Times. The Sun goes on to say that, “Six thousand women develop cancer every year through being overweight or obese”.
The stories are based on a large study in women that examined the link between body-mass index (BMI) and the incidence of cancer. This is the latest in a series of studies that have suggested a link between obesity and cancer. However, due to the design of the study, this latest research cannot prove that obesity causes cancer.
Where did the story come from?
Gillian Reeves and colleagues from the Cancer Epidemiology Unit, Oxford University, conducted this research. The study was funded by Cancer Research UK, the UK Medical Research Council, and the NHS breast screening programme. It was published in the peer-reviewed British Medical Journal.
What kind of scientific study was this?
This was a large cohort study called the Million Women Study, where the researchers studied a large group of women over time to examine the link between the health of the women and various factors, including lifestyle and use of drugs such as hormone replacement therapy (HRT). In this particular study, the researchers looked at the link between BMI and the development of new cases of cancer over a period of time and the death rates due to cancer.
The researchers recruited 1.3m women between 1996 and 2001 who were aged between 50 and 64 and who had been invited for breast screening. The women completed a questionnaire that included personal information, height, weight, and social factors. Three years later, 37% of the women completed a further questionnaire to update these details. During the follow-up period, the researchers obtained information from the NHS registry about new cases of cancer or cancer deaths that occurred among the participating women.
The follow-up, on average, was 5.4 years for cancer incidence follow-up and seven years for cancer deaths. Using the weight and height details provided by the women, the researchers calculated their BMI; women with a BMI between 25 and 29.5 were considered overweight, and women with a BMI over 30 were considered obese (as defined by the World Health Organisation criteria). The researchers then compared how the occurrence of 17 of the most common types of cancer differed among women with different BMIs. The results were adjusted to account for factors that may contribute to cancer risk including age, smoking status, number of children, alcohol use, exercise, years since menopause and use of HRT. Women who were diagnosed with cancer prior to the start of the study were excluded.
What were the results of the study?
The researchers found that there was an overall increase in risk of developing cancer or dying from cancer with increasing BMI. When they looked at the change in the incidence of specific cancers per 10 unit increase in BMI for all women, there was an increased risk of cancer of the womb, oesophagus (in one of two types), kidney, pancreas, breast (in postmenopausal women only), ovary, and also of leukaemia, non-Hodgkin’s lymphoma, and multiple myeloma (a type of blood cancer). The trend was similar for deaths from these cancers. The links were strongest for womb and oesophageal cancer, with more than double the risk of developing or dying from either of these cancers for each additional 10 BMI units.
There was no significant association between BMI and the risk of developing malignant melanoma or cancer of the stomach, colorectum, breast (premenopausal), cervix, bladder, or brain. There was an association between increasing BMI and reduced risk of developing or dying from the other type of cancer of the oesophagus, or of lung cancer.
When researchers looked only at people who had never smoked, the significance of any link to obesity remained for only five of the cancers. When they divided women into two groups based on menopausal status upon enrolment in the study, and looked at the seven cancers for which there were more than 50 new cases during the follow up period (breast, womb, ovary, large bowel, lung, non-Hodgkin’s lymphoma, malignant melanoma), only uterine cancer showed a significantly increased risk with increasing BMI for both pre-and post-menopausal women. The link between breast cancer and BMI was significant only in postmenopausal women and the link between BMI and large bowel cancer was only significant in premenopausal women.
What interpretations did the researchers draw from these results?
The researchers conclude that increased BMI is linked to an increased risk of 10 of the 17 cancers that they examined, particularly for cancer of the womb and a certain form of oesophageal cancer. They say that about 5% of cancers a year in women can be attributed to obesity. They also acknowledge that, “menopausal status is a key factor in the relation between BMI and risk of cancer among women”.
What does the NHS Knowledge Service make of this study?
These findings on the possible links between BMI and risk of cancer are the result of a fairly reliable study of a large number of women across the UK. However, it is important to realise that it cannot prove that obesity is a definite cause of the cancers that were identified.
- As the researchers acknowledged, the change in risk with increasing BMI differed according to whether they looked at all women or split them into pre- or postmenopausal groups. In addition, some types of cancer showed no link, and some showed evidence of reduced risk with increasing BMI. This suggests that the link between BMI and cancer is complex, and may differ in different groups of people, and across different types of cancer.
- Although certain potential confounding factors were accounted for, other unknown factors may have played a role. If an unknown risk factor for cancer was also associated with an increased risk of being overweight, this would appear as though overweight was the causal factor, when it was not.
- It is important to note that although the study looked at a large number of women, the actual number of cases of cancer was relatively small; therefore, the study may not have had the power to calculate a truly reliable risk values.
- In most cases, the women’s weight and height were recorded at one time point only. We cannot reliably assume that their BMI would have remained the same over time (either before or after this measurement). There is also some possibility of inaccurate calculations of BMI due to some women estimating their weight and height, rather than knowing true values.
- This study has examined mainly postmenopausal women and has only followed them over a short period. Much further research would be needed to obtain firmer evidence of this link in other population groups and over longer periods.
Risk factors for different types of cancer are multiple and include both hereditary and environmental factors that all interact in a complex way; some can be modified and others cannot. Although some risk factors may be more clearly established that others, it is not possible to be definite about the exact cause of cancer in an individual.
Sir Muir Gray adds...
Another good reason for walking an extra 3000 steps a day, about 30 minutes. Try 60 minutes if you are overweight and want to lose weight, 30 minutes will then keep you trim.