ILLNESSES ARE often named after those who discovered them. An exception is a motor-neuron disease called amyotrophic lateral sclerosis (ALS), which is widely known as Lou Gehrig’s disease after a famous baseball player of the 1920s, who died of it. As the name of its class suggests, ALS kills motor neurons—the cells through which the brain controls so-called voluntary muscles, including those for moving, eating and breathing. Some 10% of ALS cases are inherited. What causes the rest is unknown. A long-held hypothesis is that strenuous exercise has a role, because the prevalence of the condition among athletes and those serving in the armed forces is several times that in the general population (where the lifetime chance of diagnosis is one in 300).
A paper in this month’s EBioMedicine backs up that hypothesis. Thomas Julian of Sheffield University, in England, and his colleagues used a method called Mendelian randomisation, named after Gregor Mendel, a Moravian friar who worked out, in the 1850s, the mathematics of the inheritance of genetic traits.
Actually proving that a particular behaviour causes a disease would mean designing an experiment that assigned some people at random to engage in that behaviour and others not to. This clearly has ethical problems. Mendelian randomisation, however, is a pretty good substitute. It takes advantage of the fact that people have numerous genetic variations which, combined, make them more—or less—likely to engage in behaviours like smoking, drinking, eating lots of sweets and even strenuous exercise. These gene combinations work in mysterious ways. They might, for example, make people more or less sensitive to a substance, so that some binge on it while others find it unpleasant. Such genes are randomly distributed in a population, just like the genes for blond hair, blue eyes and so on. Comparing people with the relevant behaviour genes to those without them is therefore a reasonable stand-in for randomly assigning some people to engage in the behaviour.
The study drew on the UK Biobank, a repository of the genetic, health and lifestyle data of 500,000 people. Dr Julian and his colleagues pored through these to identify genes particularly common in those engaging in regular strenuous exercise. Then, in a separate sample of Europeans, they found that people with such genes were more likely to develop ALS.
To investigate how exercise might be damaging neurons, the researchers took blood samples from people before and after bursts of physical activity, and measured chemicals regulated by genes known to be associated with ALS. Half of these genes changed their activity in response to exercise. Lastly, they studied a particular variant, which had previously been linked with ALS, of a gene called C9ORF72. The more those with this variant exercised, the earlier they developed ALS. But exercise made no difference to people without it.
All this suggests that some people with a genetic predisposition to ALS could be pushed over the edge and develop the disease if they engage in regular strenuous exercise. This revelation cannot immediately be used to give practical advice. One in 300 people have the C9ORF72 variant concerned, and most would not develop ALS. The lifetime risks of heart disease and cancer—which exercise can stave off—are one in three and one in two respectively, so stopping exercising in order not to develop ALS would be foolish. But it is possible to imagine screening those who wish to pursue a sporting career for relevant variants, so that they could be advised of the potential risks. Everyone else, meanwhile, can stick to gentler forms of exercise, to keep fit into a ripe old age.
This article appeared in the Science & technology section of the print edition under the headline “A question of sport?”