Average fitness has (surprise) increased since the 1970s


As of September 2017, new Sweat Science columns are being published at www.outsideonline.com/sweatscience. Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Here’s an interesting graph from a new paper from the Cooper Institute, the famous fitness institute founded by Kenneth Cooper, the “inventor” of aerobics:

The paper appears in this month’s issue of Medicine & Science in Sports & Exercise, and it shows the results of 52,785 fitness tests performed on men at the Cooper Institute between 1970 and 2009. Each patient appears only once in the data, so it’s not a longitudinal study of how individuals got more or less fit; it’s a cross-sectional study showing how the initial fitness of patients at the institute has changed over the years. The trend is pretty straightforward: a big jump from the 1970s to the 1980s, then levelling off to the 1990s, then a slight decline into the most recent decade.

So… at first glance, I found this data pretty surprising. My naive guess would have been that fitness (as represented in this graph by VO2max determined in an incremental treadmill test) would have declined steadily from the 1970s to the present. After all, we’re always hearing about how society has never been less fit. Instead, it seems that we’re fitter than we were 30 years ago, and other data in the paper suggests that we’re more active too. So does this mean that our current health woes have nothing to do with physical activity level? After all, the data also shows that average weight in the subjects has increased by 8 kg since 1970, while height has stayed the same. So we’re more active, but fatter — a pretty good indicator that diet, rather than exercise, is driving the rise in obesity.

Of course, there are a few caveats. For one thing:

Cooper Clinic patients are self-referred or referred by their employers and are primarily healthy college educated middle to upper income non- Hispanic whites who have access to medical care.

So there’s really no way of knowing whether the average patient who decides to go to the Cooper Clinic is comparable to the average patient from the 1970s. It could be, for example, that the clinic’s clientele has changed as it became famous, so that it now attracts people who are already a bit more serious about fitness. But that’s just speculation. The data may well be a fair representation of societal trends — and if so, I need to revise some of my assumptions.

15 Replies to “Average fitness has (surprise) increased since the 1970s”

  1. My sense is that your original assumption is validated, by the chart at least. Unless there’s some compelling reason not to, drop the 70s…and the results become exactly what we assume: male fitness has declined over the last ~30 years. More sad is the larger decrease in the youngest cohort. Cross your fingers that adequate advocacy can prevent this from becoming a trend.
    And a question: if the first two sets of numbers are solid, what was involved in the cross-the-board increase?

  2. It seems rather dubious… The average VO2max not expressed per kg has gone up even more significantly than the graph shows (a 10% increase in VO2max in all bar the A group, and the 8kg weight increase would show up as a declining VO2max score.

  3. It’s an average. Probably top end has increased quite a bit, and low end doesn’t get any lower.

  4. @DKoehn:

    “Unless there’s some compelling reason not to, drop the 70s…”

    Well, sure — all data conforms to my expectations if I ignore the data that DOESN’T conform to my expectations! 🙂

    “And a question: if the first two sets of numbers are solid, what was involved in the cross-the-board increase?”

    Here’s what the authors of the paper suggest:

    “Some key events in the past four decades relating to PA and health include 1) publication of Aerobics by Dr. Kenneth H. Cooper in 1968; 2) the emergence of the jogging/running phenomenon in the 1970s and 1980s influenced by popular icons such as Olympic gold medal winner Frank Shorter and running enthusiast Jim Fixx; 3) the PA [physical activity] guidelines by the American College of Sports Medicine in 1978, 1986, and 1995; 4) release of the US Surgeon General’s Report on Physical Activity and Health in the 1990s; and 5) the collaborative development of national PA guidelines in 2008. The timeline of these events parallels the observed changes in CRF demonstrated in this cohort.”

    Now, you might wonder what the surgeon general’s report in the 1990s and the national physical activity guidelines in 2008 have to do with the observed trends. The authors suggest that guidelines starting in the 1970s and 1980s emphasized vigorous and intense exercise, but since the 1990s guidelines have been watered down, with more of an emphasis of moderate activity — which is why gains have levelled off and even started to reverse.

    I still find all this rather hard to swallow, but that’s what they say.

  5. @Jim: Yes, as you point out, the absolute VO2max has jumped even more dramatically. In truth, looking at absolute VO2max probably overstates the real increase in fitness (since it’s dragged up by people getting fatter), while looking at relative VO2max (divided by body weight) understates the increase in fitness. If we consider allometric scaling, we should really be looking at VO2max divided by body mass to the power of (2/3) to get a sense of how fitness, independent of body size, has changed.

    Still, the basic point remains the same no matter how we scale the data: fitness in this cohort has increased since the 1970s. And the basic question also remains the same: is this cohort really representative of the general population?

  6. As mentioned earlier if the Vo2 max suggests a sub 7 minute mile for a 40 year old it is most certainly not representative of the general population. You make the key point at the end of your post. Without knowledge of the potential variables influencing self-selection it is not wise to draw conclusions in trends over time.

    There is no reason to assume the factors that influence selection stay steady over time. Especially since selection is based both on the individual and also on employee referrals.

    Has the ratio of individual vs employee referral even stayed steady? Are one of those groups more fit than the other? Are the trends over time similar for each of these groups? The could be numerous motivations to go to the institute. The employrers who do the referring could change over time and their motivations or their workers population characteristics could be very different. Way too many unknowns.

  7. For all the possible selection biases, I’d say that there is one big factor that explains the increase in fitness since the seventies: smoking.

  8. That’s actually not a bad guess. The percentage of patients who smoked was 22.7% in the 70s, 17.7% in the 80s, 14.9% in the 90s and 15.2% in the 00s — so that trend is indeed the mirror image of the fitness trend.

  9. I think that you can’t over-emphasize the selection bias here.

    The only reasonable conclusion is that “the average patient at the Cooper Institute has gotten fitter over the past 30 years”. Any inference beyond that to the general population is pure speculation. While the title is attention grabbing: “Average fitness has increased…” it is somewhat misleading.

    A fun study to spur discussion for sure. But anyone inferring broader applicability has nothing to base their generalization on since the demographic (as you describe it) at the Cooper Institute is so different from the ‘average citizen’.

  10. @npm

    The averages seem to be in line with the averages for non athletes quoted here: http://www.sport-fitness-advisor.com/VO2max.html. They seem to be quoted from a respectable source (though unfortunately no page number. It´s a 600 page handbook).

    In my opinion the 6:49 mile is a false comparison. The calculator also gives a 17 hour 200 KM. I don’t suppose the authors of the calculator mean to say that anyone with an average VO2max can finish a 200 km in 17 hours without any training at all. The correct comparison would be that an average 42 year old with a VO2max of 42 can do a 6:49 mile after a reasonable amount of training.

Comments are closed.