I have long argued that dissent is the oxygen of science and
that whereas unanimity of opinion is fine for political parties and religious
organisations, it has no place in science. Indeed, I have always held the view
that the unexpected finding in science is its jewel. It is not a view that is
widely held since science tends to be protective of its theories and dislikes
having its conventional wisdom challenged. Right now, a major row has broken
out over the true health risks of being moderately overweight. There are two
dimensions to this squabble, the scientific dimension and the policy dimension.
Let’s start with the science.
Katherine Flegal is a scientist at the US National Center for
Health Statistics and she specialises in the study of the obesity epidemic in
the US. In January of this year, Flegal and
her colleagues published a paper in the Journal of The American Medical
Association[1] in which she showed that
persons who were overweight but not obese, had a statistically lower chance of
dying than persons with a bodyweight deemed to be within the normal range. Her
paper was based on a systematic review of the scientific literature on obesity
in which she searched the literature for any study that involved the use of
standardised classes of body weight (body mass index: BMI) and which involved a
follow up period from baseline measures that allowed for an estimate of the
relative risk of mortality across different standardised classes of BMI. She
included all languages in her search and also excluded studies that failed to
use the standard BMI classes, involved adolescents or which involved subjects
with specific medical conditions. In all, she included 97 studies, which
involved 2.88 million people with a total number of deaths in follow-up of
270,000. The papers were selected initially by Flegal but were included in the
analysis only with the unanimous agreement of three independent reviewers. Top
science here!
Her findings that
overweight people had a 6% lower risk of mortality than normal weight
individuals wasn’t new but it was a massively comprehensive study compared to
all previous studies, which had shown this effect. The Journal commissioned a
guest editorial on the paper from Dr Steven Heymsfield, Director of the
Pennington Institute in Baton Rouge, a center of excellence in obesity research[2]. This editorial notes that: “Body Mass Index
is an imperfect predictor of metabolic risk” and concludes by stating that:
“Establishing BMI is only the first step toward a more comprehensive risk evaluation”.
So, let’s pause here. We have one of America’s leading statisticians on the
obesity epidemic publishing the most comprehensive paper of its kind in this
field in a top academic journal following the normal rigors of peer review and
we have a very supportive guest editorial by the director of the leading
obesity research centre in the US.
However, the dons at
the Harvard School of Public Health were not happy. Led by Professor Walter
Willett, they held a special meeting to line up critics of the Flegal paper. In
a radio interview[3], Willet stated: “This study is really a pile of rubbish and
no one should waste their time reading it”. This in turn led to an
editorial in Nature[4] and an overview article
entitled “The big fat truth” by science journalist Virginia Hughes.[5] She reviews some literature and draws on
data, which shows that the relationship between BMI and mortality varies
greatly with age. In older people, at the lower and at the upper spectrum of
BMI, the risk of mortality is higher than it is for the majority of the
population who are neither too skinny nor too fat. In younger people, the lower
low BMI values are not associated with elevated risk.
Leaving the science
aside, there is a critically important aspect to this row that needs
highlighting. Think back to the BSE crisis. At that point, within the EU we had
the risk assessment process and the risk management process both operated by
the European Commission. That was then amended to take the risk assessment
process away from the Commission and to create a totally science- based
independent body, The European Food Safety Authority, to conduct risk assessment.
The Harvard group is effectively seeking to be both risk assessors and risk managers.
The former is science based and the latter is politically or policy based. If
the two are attended to within the same institute, as the Harvard group seem to
want, then the risk management process will filter the risk assessment process.
Why support a scientific paper, which conflicts with your risk management
goals? Indeed, in this week’s Harvard Gazette which covered this controversy,
Professor Willett is quoted thus: “If you
don’t have the right goal you are very unlikely to end up in the right place”[6]. Clearly, Professor Willett knows what
is “right” and those who differ are “wrong”. This is simply bad for science. As
I said, dissent is the oxygen of science.
