Is Exercise Really Good For You?

I spend a lot of my time looking at claims made for various cosmetic products.  People selling products are easily convinced of the benefits of those products, and are eager to share those benefits with potential purchasers.  But it’s very often the case that the claims they make are based on the flimsiest of evidence.  That is how the world works.  The more you spend time looking at cosmetic claims, the more cynical you get about all claims across the board.  So when someone tells me that exercise is good for you my first reaction is to be suspicious, even though this claim is often made, seems reasonable and doesn’t seem to be especially controversial.  And it is one that I think most people assume is backed up by solid science.

I am a big fan of the scientific method, but I am well aware that it isn’t perfect. In fact that it isn’t perfect is one of its points.   You try to find fault with you current ideas to see if there are better ones out there. It is quite common for notions that enjoy a complete consensus amongst scientists to turn out to be quite wrong when new data and better ideas come along.   Could the exercise equals good health idea be one that won’t stand the test of time?

Of course the situation here is quite different to  when someone is trying to sell something.  Whether I exercise of not is not going to directly benefit anyone’s pocket.  So when you hear  specific claims about what exercise can do for you, they are not obviously self serving.  But that doesn’t mean that they are necessarily right.  So let’s have a look at just how strong the case for exercise is.  This is what the UK’s National Health Service has to say –

[Exercise] can reduce your risk of major illnesses, such as heart disease, stroke, type 2 diabetes and cancer by up to 50% and lower your risk of early death by up to 30%.

The good thing about the NHS is it isn’t a commercial organisation, and if anything it should be biased towards keeping people healthy.  So maybe this advice can be taken at face value.   Maybe.  But those numbers do look suspicious.   The “up to 50%” trope in particular is a marketing favourite.  Saying something can work up to a given figure is actually putting a limit on its efficacy.  So it can work up to 50% but no more and usually less.  Our brains though don’t work that logically when left to themselves.  We see the number and take it as a promise.  It doesn’t actually say that exercise will halve my chance of having a heart attack – but that is the way most of us will read it.  I can’t stop myself feeling the need to investigate what the actual evidence is.

Exercise and Heart Disease

Let’s look at heart disease first.  Luckily this has already been reviewed by the Cochrane group, who looked at 16 studies and found modest improvements in how long heart attack survivors lived if they took some exercise.  This is all good stuff.  If I had suffered from a heart attack I’d certainly be interested in even a marginal improvement in my prospects.

Given how interesting this area is, it is surprising how little actual research seems to have been done into it.  The question seems to me to be an important one.  If I need to exercise to be healthy then I’ll give it a go.  If I just need to lose weight then I probably don’t need to bother.  One rather small study indicated that when obese men exercised then they had better health as indicated by some measures of their fitness – though they didn’t look at actual health outcomes like how long they lived and what diseases they got.  What we really need is a large trial that quantifies what kind of health benefit you might expect from a given amount of exercise.  This is a tall order – you need a lot of people, a  lot of monitoring and you need to run it for a long time so you can get an idea of just how much longer people live when they exercise.  Any other measure is not direct, and is subject to a lot of inherent variability.  Death doesn’t have much to recommend it generally, but it is at least something that can be measured pretty accurately and over which there is not much room for debate.

Given the difficulties of conducting such a study I was not optimistic of being able to find one that met my criteria.  But to my surprise there was one that pretty much matched exactly what I was looking for. This found that  150 minutes of exercise a week corresponded to living an extra 3.4 years. At 75 minutes a week the life extension was 1.8 years.  These data certainly support the notion that exercise is good for you.  3.4 years is just over 4% of the current UK male life expectancy of 81.5 years.   So you could say that regular exercise makes you about 4% healthier.  This was however the only paper that I could find which really allowed such a conclusion to be drawn.  You wouldn’t approve a drug on the basis of one study.  So while the notion of exercise being good for your health is supported by data I wouldn’t say that the case is yet overwhelming.  And I don’t think that we can say much about the optimal amount of exercise yet.  There is going to be an amount of exercise that would be unhealthy.  We know that Olympic athletes live about two and half years longer than the general population.  But there is no difference between high and low intensity sporting disciplines.  This suggests to me that there is a plateau in the benefits that accrue from exercise.

So exercise is good, but only up to a point, and its benefits are well worth having but not spectacular.  And even then the evidence in the form of hard data is not yet compelling.  I don’t see any sign of the 50% risk reduction that the NHS was suggesting.

Exercise and Diabetes

Heart disease is a life threatening condition but is one that you don’t really notice until you have a heart attack. Another disease that it has been suggested exercise can help with is diabetes.  This is one that rarely kills quickly, but has a big impact on the quality of your life.  So if you can reduce your chances of succumbing to it through exercise then that is worth knowing.  It seems to be a consensus amongst people who work in this area that exercise does help.  But what is there in the way of formal studies to justify this belief?

I found a very good study that was pretty conclusive that a combination of diet and exercise could help with the symptoms of diabetes and might well prevent its onset.  It was an analysis of 8 previous trials and so is particularly valuable.  If eight different groups are finding the same thing that is encouraging.   The trouble was that it was considering diet and exercise together as a single factor.  It isn’t impossible that they work together and that either on their own wouldn’t work as well.  But it is more likely that both individually are beneficial.  But which one is the big one?  I failed to find any objective evidence that exercise alone can be shown to be beneficial in diabetes – though I don’t dismiss the opinion of people who work in the area.  It is quite normal for something to be suspected long before it is demonstrated.

So yes, exercise might well be helpful in preventing diabetes and this is supported by the people who probably know best.  But again I couldn’t find any really direct compelling evidence.  Given that diabetes is really unpleasant, and exercise is often quite enjoyable this seems like a good bet, but it is disappointing that the potential benefits aren’t more clearly quantifiable.  Again neither the 50% nor the 30% figure are backed up by anything I can find.

Exercise and Cancer

Cancer is no longer the death sentence it used to be, but it is still an unusually unpleasant disease. It strikes apparently at random and once it is settled in it slowly does its work.  Despite a huge amount of research on the subject and huge advances in our understanding of it, it still seems a bit mysterious.  It doesn’t seem impossible to believe that exercise might help, but there isn’t any obvious mechanism linking the two.

So what is the evidence for the benefits of exercise in cancer prevention?  This is a tough thing to prove.  You need to compare the cancer rates of a group of people and see if the ones who get the most exercise have the least incidence of cancer.  Quite apart from the practical difficulties, there are a few ethical issues there.  You can hardly forbid people from taking exercise just to make your stats easier to work out, so this would have to be a trial where you asked people to tell you what they do without giving them any instructions on what they should do.   Or is it even ethical not to encourage them to exercise and so mess  your trial up?

Anyway, I had a look to see what work had been done.  One study looked at post menopausal women and measured the effect of exercise on their hormone levels.  Exercise did indeed alter their hormonal balance to make their profile look less like someone who was prone to breast cancer.   This sounds like evidence that exercise reduced their risk of getting cancer, though it could conceivably be explained in some other ways.

I was really looking for a big review of all the data relating exercise to cancer risk.  I did find one, which basically concluded that data was sparse.  They did find that there were indications that colon cancer was  less prevalent amongst people who were active.   So again, the actual data behind the notion that exercise is good for you as a way of preventing cancer is not especially solid.    One in built bias we all have is that we like to regard things as being either good or bad.  I have written blog posts about this halo effect before.  But there is really no reason to believe that just because exercise is good for diabetes and heart attacks, it also protects against cancer.  It can be good for some things and bad for others.

Exercise and Weight Loss

How about losing weight?  This one sounds on the face of it like an obvious benefit of exercise and one that for some people would make it worth doing for this alone.  Physical activity takes energy, and the only way your body can provide this energy is by burning calories.  So all things being equal, more exercise should lead to losing weight.  And big energy consuming tasks like running a marathon or learning to dance certainly do remove excess pounds.  If you are heavier than you would like to be, get down the gym and sweat it off.  The only problem here is that all things are not equal.  If you take more exercise but simultaneously increase your food consumption to match then you’ll stay much the same as you were before.  And if my personal experience is anything to go by, this is exactly what happens.  When I changed job and had to switch to driving to work instead of a 12 mile cycle, I put on a little  weight.  But it was a tiny amount.  Basically if I didn’t cycle I didn’t feel as hungry and I ate less.

And in fact most of the energy we are using goes to just keeping the body working.  We are burning calories all the time.  We burn them faster when we exercise, but not really that much faster.  Running a mile expends a 100 calories.  That would take me about 20 minutes.  100 calories is about a packet of crisps.  So if I go for a long run and have a packet of crisps at the end as a reward I will have achieved precisely nothing in terms weight loss.  I would do just as well to read a book for half an hour and skip the snack.

So exercise is not a very efficient way of losing weight, which to my mind means you have to really believe in its other benefits that make it worth the effort.

My Attitude To Exercise

I always like to compare what the scientific literature says with my own experience. I have tried for many years to get into the habit of regular exercise.  I find this is something that I have little difficulty with in the winter and particularly when the weather is cold.  Going for a run is a great way to warm up and wake up.   I usually lapse most summers  though because I find that when it is hot my willpower is weak.  But the interesting thing is that I don’t really find any big difference in how I feel when I am exercising and when I am not.  But what I have found is that I having lost a fair bit of weight recently I find exercising much easier.  I almost always follow the same route for my morning run, and as my weight dropped I found I completed the course quicker and could keep up without stopping for a rest for further.

One possible interpretation of this is that exercise is not a cause of good health but a consequence of it.  If you are healthy all your body’s systems are operating effectively to deliver you high levels of energy and your organs are in good order.  So you are able to run around more and go for longer walks.  It is even conceivable that you are damaging your health by exerting yourself.  I don’t think that this is likely.   But it isn’t impossible.  And one fact does bear it out.  As we have adopted more and more labour saving technologies we have begun to live longer.  We definitely know that it is possible to die of overwork under some conditions, so the possibility that any amount of exercise is harmful is not totally out of the question.

I am not acting on this last suggestion.  I continue to make sure I get at least some exercise every day I can.  But I have decided not to aim for heroic amounts of it.  My daily target is a minimum of a quarter of an hour, an ideal of half an hour and up to an hour when I am enjoying it.  That seems like a reasonable approach.

But on the whole, I have a feeling that the biggest factor affecting our health that is actually under our control is probably going to be diet.  I am not even convinced that exercise will ultimately even make it to the second most important factor.  Things such as sleeping properly, not spending too much time sitting down and our general mental well being might prove to be as if not more significant.  But that is all for another blog post.

References

http://www.cochrane.org/CD001800/VASC_exercise-based-rehabilitation-coronary-heart-disease

http://annals.org/aim/article/713672/reduction-obesity-related-comorbid-conditions-after-diet-induced-weight-loss

http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001335

http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD003054.pub3/full

http://ascopubs.org/doi/abs/10.1200/jco.2009.24.9557

http://europepmc.org/abstract/med/14600545