Benjamin Weiner <[email hidden]> wrote in message news:<40306a8f$1@darkstar>...
Quoted message said:Carl Fogel said:Tim McNamara <[email hidden]> wrote in message news:<[email hidden]>...
Quoted message said:Quoted message said:ISTR a reference that in the 1970's some worldwide health study found that the life expectancy
of pro cyclists was well below average (like 53 years). I've never been able find the original
data, though.
Quoted message said:procycling.comnews main.aspOpen ↗
Quoted message said:It might be hard to find because it was French. It sounds as if it was a study of post-war Tour
de France riders.
Quoted message said:I'm curious whether such studies connect the wide-spread drug use in professional bicycling with
shortened life-spans.
This topic is occasionally debated in rec.bicycles.racing. Nobody has ever given a reference to a
trustworthy study that found that ex-pro cyclists have a short life expectancy, and there are
several reasons why studies like the above are flawed. You can Google search rbr for "life
expectancy" for details. Robert Chung has explained a couple of times why such studies would be
quite hard to do right. Here's a pithy one:p://groups.google.com/groups?q=life+expectancy+winners+group:rec.bicycles.racing&hl=en&lr=&ie=UTF-
8&scoring=r&selm=3C57EAEC.1B5C47B1%40nospam.com&rnum=3
A couple of obvious examples: The number of pro cyclists has increased greatly over the 20th
century (faster than the population as a whole I expect). The majority of post-war Tour de France
cyclists are still alive. Therefore the sample of dead cyclists is biased towards people who died
young. You can't just compare to the population as a whole without controlling for the difference
in distributions of birthdates. (See Chung's post on the difference between longitudinal and
cohort studies.)
I believe that the reference remembered by Tim is either an urban legend or statistically flawed.
Dear Benjamin,
The logic behind those arguments may need more explanation.
The number of pro cyclists may have increased (or decreased or remained the same) and may have
changed at a different rate than the population as a whole--but how does this affect their predicted
longevity?
The majority of post-war Tour de France cyclists may still be alive, but then so are the majority of
the people in the population as a whole who were born in the same year as the cyclists.
If pro cyclists born in 1956 are dying at a significantly different rate than my high school class,
then they're dying at a different rate.
A proper study is certainly tricky and tedious, but but given roughly 100 to 200 pros riding in the
post-war Tour every year for the last fifty years, it hardly seems impossible.
My spreadsheet shows an average of 151 riders starting the Tour each year from 1946 to 2003.
Obviously, many of the 8611 starters were duplicates, veterans riding the Tour again. But even if we
assume an absurdly high 10-Tour average, there would be 861 riders over 57 years.
Much of what I saw in those posts consisted of examples and counter-examples limited to Tour winners--
a group of only 30 riders over 57 years, too small to be significant because a single occurrence may
skew things wildly. (To cite the obvious example, testicular cancer does not strike 1 out of 30 men--
the incidence is roughly 0.5% by the age of 75, 1 victim out of 200.)
Life insurance companies and pension funds do this sort of thing well enough to make money at it.
They have a fairly clear notion of the differences that you mention, so it would be interesting if
they've done a study. Unfortunately, professional bicyclists are a very small group, so they may
have been ignored as not being worth the trouble.
Incidentally, professional cyclists on the Tour probably do have a lower life-expectancy than the
population as a whole--simply because they're all men. Women are slightly more than half the
population and live 5-7 years longer, on average, in Europe and America.
Carl Fogel
