An epidemic of confusion
One of the fundamental premises that guides thinking and research at SIRC is that people have the right of access to accurate and balanced health and lifestyle information, on the basis of which they can make informed decisions about how they lead their lives. They may choose, of course, to ignore the evidence and its implications. In liberal democracies we must concede that people are entitled to have bad habits. But when the facts are clearly and fairly presented they cannot claim that they have acted in ignorance or have been misled.
The problem, however, is that we are increasingly exposed not to balanced presentations of the evidence but rather to an ever-changing diet of conjecture and distortions of reality motivated by doctrine or personal advancement. While the popular media may, and indeed should, be singled out as being responsible for misleading coverage of health issues, they are not the only culprits. The former editor of the BMJ, for example, has recently argued that medical journals are no more than "an extension of the marketing arm of pharmaceutical companies" — a strong claim but one which the present-day editor of the BMJ has been obliged to consider very seriously. Government departments and agencies may also bow to pressure from narrowly focused advocacy groups and put a politically convenient spin on research results.
The end result of this is inevitably a distinct sense of confusion and distrust. Our own research, for example, indicates that around one third of people in the UK find health reports contradictory or confusing and tend not to pay attention to them — considerably more than those who rely on such reports to inform their personal lifestyles. And nowhere is this confusion more evident than in the context of obesity.
Just over a year ago the Centers for Disease Control and Prevention announced that around 400,000 people would die every year in the United States because they were obese — almost as many as those who die from smoking-related diseases. Despite the surprise expressed by some leading specialists about the size of the alleged impact of obesity, the statistics were treated as a basis for world-wide alarm about 'epidemics' that had potentially catastrophic consequences. It also provided the essential basis for classifying obesity as an illness and, therefore, treatable using medical resources. The Washington Times noted:
"Obesity is now an illness and can be covered by Medicare, the federal health-insurance program for the elderly and disabled — CDC first requested removal of the obesity language [defining it as not an illness] in 2001. Now the second-biggest cause of preventable deaths in the United States, obesity contributes to 400,000 annual deaths, according to the CDC. With this new policy, Medicare will be able to review scientific evidence in order to determine which interventions improve health outcomes'."
An acceleration of the medicalisation of obesity rapidly followed in other countries, including the UK, as similar estimates of the apparent death toll of obesity were made. The Commons Health Committee, for example, noted in the introduction to their much criticised report on obesity that:
"The devastating consequences of the epidemic of obesity are likely to have a profound impact over the next century. Obesity will soon supersede tobacco as the greatest cause of premature death in this country."
Very soon after the CDC announcement, however, serious doubts about the statistics were increasingly being expressed. Both Science magazine and the Wall Street Journal noted that a number of anonymous sources within the CDC itself were worried about the political pressure that had been applied to ensure that the findings were consistent with the US government's public health policies on obesity. Others pointed to serious statistical errors in the CDC analysis that rendered the conclusions unreliable.
By November 2004, just 8 months after the original report was published, the CDC were already conceding that some 'computational errors' had occurred in the analyses. The Washington Post noted:
"Federal health officials said yesterday they had overestimated in a high-profile study the number of Americans dying from being overweight. Officials at the federal Centers for Disease Control and Prevention said they will submit a correction to the Journal of the American Medical Association, which published the paper March 10, to set the record straight. In the hope of producing more accurate estimates in the future, the agency is reviewing the methods it uses to calculate the health effects of being overweight."
Meanwhile in the UK, SIRC's own analysis of the obesity statistics for England indicated that the widespread assumptions about 'exponential' rises in obesity and the 'epidemic' scale of the problem were not justified by the data. Obesity was a problem much more associated with middle-age and later life than it was with children and teenagers. The SIRC report also commented on the somewhat arbitrary nature of the definition of obesity in childhood when rapid changes in height, weight and body shape occur. The use of cut-off points established in 1990 is widely seen, not least by the International Task Force on Obesity, as presenting a potentially distorted picture. The most common alternative is the International Standard, which enables more meaningful cross-national comparisons to be made. It also generates figures for the prevalence of childhood obesity that are typically less than half of those than result from the use of the old UK Standard.
This move to a more meaningful discussion of the scale of obesity, however, was ignored in a report commissioned by the UK Department for Health and Published in April 2005. Still using the UK Standard they claimed that the prevalence of obesity in children under the age of 11 had risen from 9.9% in 1995 to 13.7% in 2003 — twice as many as would be classified by the international method.
A month later in May 2005, back in the United States, the true scale of the CDC's 'computational errors' was revealed in a new analysis of the data by a team led by Katherine Flegal, a researcher at the CDC itself, and published in the Journal of the American Medical Association. The previous estimate of 400,000 deaths per year from obesity was revised down to 112,000 after the impact of confounding variables had been taken into account properly — an almost four-fold reduction. But there was rather more in Flegal's paper to suggest that we had previously been seriously misled about the impact of weight on health. The mortality rates for overweight people (BMI 25-30) were actually lower than those for people of 'normal' weight — 86,000 fewer deaths. Taking this into account, the true picture was that fewer than 26,000 deaths per year resulted from being above normal weight.
It was also clear that the negative effect of obesity was confined mostly to those with a BMI of 35 or above (the cut-off point for obesity itself is a BMI of 30). As an indication of how big you need to be to be included in this group, a man who is 5' 10" tall (1.778 metres) would need to weigh in excess of 17 stones and 5 pounds (110.64 kilos).
So what are we to make of all of this? Within the space of just over a year we have gone from obesity threatening to overtake smoking as the biggest avoidable killer to the idea that being overweight might actually be good for you. We have also been led to believe that the levels of obesity in children are twice has high as the standards preferred by the anti-obesity organisations themselves would indicate. And we wonder why people increasingly distrust reports from health 'experts' and become immune to their coercions?
What we can learn from all of this is that things are not always what they seem to be. The data themselves rarely lead to a single, simple conclusion. The ways in which the data are defined, analysed and interpreted can radically alter their meaning, especially when the researchers are strongly motivated to demonstrate that they are right and others are wrong, or to support some cause that is more 'worthy' than good science. And the peer-review process, despite all of its merits, does not always detect such subtle sleights of hand.
There is now a need, we feel, for additional checks and balances to be introduced to reduce the damaging lack of trust that results from contradictory and, in some cases, just plain wrong pronouncements that emanate from government research departments and the weight-loss NGOs. Yes, we need to be concerned about risks to our health, and especially to the health of our children. But only when we have some better form of quality control in the stream of so-called 'information' to which we are exposed on a daily basis will we be in a position to discriminate between fact and fiction and between balanced assessments and doctrinal spin.
Peter Marsh — 22 May 2005