While there is a general move to have us lose some of our collective weight, regular readers may have noticed that I’m a less exercised regarding this than most health professionals. Just last week, one of my posts included several links to articles which appear to show that individuals in the ‘overweight’ category have health status at least as good, if not better, than those in the supposedly ‘healthy’ category.
Now this week, the subject of body weight was raised in no small way again as a result of a review on the link between lifestyle factors and cancer. The report, a joint effort by the World Cancer Research Fund (WCRF) and the American Institute for Cancer Research (AICR), makes a number of recommendations designed to ward off cancer. The most prominent of these, is the need to keep our weight within the ‘healthy’ body mass index range of 21-23.
The WCRF/AICR recommendations and other recommendations are the distillate of a 5-year trawl through more than 7000 studies. Big numbers that, it seems, deserve big claims. For instance, I read that Philip James, a member of the report’s panel of experts, says that they are convinced that excess body fat increases risks for cancer of the colon, kidney, pancreas, esophagus and uterus, as well as breast cancer in post-menopausal women. This week’s New Scientist which landed on my doormat yesterday pronounces that: ‘Obesity causes cancer’.
What is striking to me is the apparent certainty with which such statements are made. But does the science support such sureness?
After all, these recommendations come from what are known a epidemiological studies, which look for associations betweens things. Clearly, the panel has found a stack of evidence that links, say, excess weight with cancer. However, this link does not mean that one is causing the other.
Let’s for a moment imagine that a study finds that owning a television is associated with obesity. The question is, does owning a TV cause heart disease? Imagine you bought a TV but left it in its box and stored it in the attic. Would you be at increased risk of heart disease? Probably not. However, watching the TV endlessly for hours might be a problem. However, it’s not really owning a TV, but watching it and therefore being more sedentary that is the likely true cause of an enhanced risk of heart disease. It doesn’t matter whether we have 1 or a hundred such studies ” they still cannot be used to impugn TV ownership, because we just don’t know that it is this or something associated with TV ownership (such as being more sedentary) that is the true cause of heart disease.
The same argument can be made for any association found in epidemiological studies, including those found between body weight and cancer.
Scientists can, however, get closer to the truth by accounting for factors that may also affect the association. So, for instance, in the case of the TV, researchers could look at the relationship between TV ownership and heart disease, but this time they could take into account, say, the activity levels of those who own a TV. Now, if the link between heart disease and TV ownership appeared to remain having taking into account activity levels, then that would strengthen the notion that owning a TV actually causes heart disease. This technique, known as ‘controlling for confounding factors’ is not easy to do with any real accuracy, partly because health and disease tend to be highly ‘multifactorial’, which means potentially important influences may not be accounted for.
And even when such ‘controlling’ is done quite brilliantly, we’re still left with that pesky epidemiological data, which can never prove causality. In certain epidemiological studies, such associations can be tracked over time, which is better than just assessing them on a single occasion. But, still, causality cannot be assumed, however ‘strong’ the evidence.
So, when scientists say they’re ‘convinced’ that excess body weight or body fat increases risk of cancer, what they are offering is an opinion, not scientific certainty. And it’s simply not accurate to say that ‘obesity causes cancer’.
Another major deficiency of the WCRF/AICR report is that it focuses on cancer. I don’t deny for one moment that this is an important condition: t currently affects about 1 in 3 of us in the Western world and kills about 1 in 4. However, when data is available, it’s sometimes better to pan out and examine the influence of a factor not on a defined condition such as cancer, but something with broader implications such as overall risk of death (what scientists call ‘total mortality’). With regard to this, one study published in the Journal of the American Medical Association in 2005 found that individuals in the ‘overweight’ category (BMI of 25-29.9) were actually at a reduced risk of dying compared to those in the ‘healthy’ category [1].
Awareness of this sort of statistic leads me to be somewhat doubtful about the broad value of encouraging individuals to conform to a BMI of 21-23.
Another prominent recommendation made in the WCRF/AICR report is the need to limit red meat and avoid processed meats such as bacon and salami. Again, this advice is based on epidemiological data which is although is not to be dismissed out of hand, provides no certainty either.
And again, the focus is on cancer. When we broaden the debate and focus on overall mortality rates, we find that these are not any higher in meat-eaters than vegetarians. [2]. In light of this, meat somehow doesn’t look quite so ‘deadly’.
The real test of any recommendations or treatment is to apply it to a group of individuals and compare the results with a ‘control’ group who were not subjected to the intervention being tested. Next week, I’ll report in some detail on the results of a recent, large study which assessed the ‘benefits’ of ‘healthy’ eating (broadly similar to those recommended by the WCRF/AICR) with regard to a variety of conditions including cancer. As you’ll learn, the findings of this study were less than impressive.
My aim here is not to pour cold water on the whole of the WCRF/AICR report, but to put it in context. The certainty with which some of its findings have been reported are just not justified by the science. And, taking a broader look at health, it seems that being heavier than is traditionally advised and eating meat just ain’t that bad.
References:
1. Flegal KM, et al. Excess deaths associated with underweight, overweight, and obesity. JAMA. 2005;20;293(15):1861-7.
2. Key TJ, et al. Mortality in British vegetarians review and preliminary results from EPIC-Oxford. Am J Clin Nutr 2003;78(suppl):533S-8S
