NODDING syndrome is a form of epilepsy that strikes children, mostly between the ages of five and 15. Despite the innocuous name, it is debilitating. It robs its victims of their mental capacity, stunts their growth and causes both the characteristic “nodding-off” motion which gives its name and more serious seizures, often when a child is being fed. The exact death rate is unknown, but it is high.

The syndrome is also something of a medical mystery. The first cases were identified in Tanzania in the 1960s. Now it has spread to parts of Uganda and South Sudan. No one knows how many people are affected, but it is thousands, at least. Nor has anyone been sure what causes the disease. But Tory Johnson, of America’s National Institutes of Health, and her colleagues have a theory. As they describe in a paper just published in Science Translational Medicine, they suspect that nodding syndrome is an “autoimmune” disease caused by sufferers’ attempts to fight off infection by a parasitic worm.

The worm in question is Onchocerca volvulus, a tiny nematode spread by the bites of black flies that is best known for causing river blindness. Epidemiologists had already drawn a link between nodding syndrome and areas infested by O. volvulus, but whenever people have looked, they have failed to find traces of the worm in sufferers’ brains, or in the cerebrospinal fluid (CSF) that bathes their brains and spinal cords. This led to the suspicion that, if the worm is indeed responsible, it is doing its harm indirectly.

To investigate, Dr Johnson and her colleagues analysed blood and CSF from children with nodding syndrome in both Uganda and South Sudan. They were looking for antibodies. These are proteins produced by the immune system which bind to and disable specific molecules on the surfaces of invading viruses, bacteria and the like, thus damaging or destroying the invader. One of the antibodies they discovered was tailored to a protein called leiomodin-1, which is produced by mammalian nerve cells. In particular, studies of mice suggest it is found in both the cerebellum (a region of the brain which, among other things, helps control muscle function) and in the cerebral cortex (where abstract thinking happens).

Trials in a Petri dish confirmed that the leiomodin-1 antibody Dr Johnson isolated is toxic to human nerve cells. That suggests nodding syndrome is, indeed, autoimmune: the victims’ immune systems are attacking their own brains. It does, though, leave the question of why infection with O. volvulus should cause this antibody to be produced in the first place.

Dr Johnson and her colleagues think they know the answer to that. When they looked at proteins produced by the worm, they found one, called tropomyosin, that is strikingly similar to leiomodin-1. This similarity suggests antibodies intended to attack the worm’s proteins could end up inflicting collateral damage on the human versions, too. And it might not just be tropomyosin that is involved. The researchers found a handful of other worm proteins that were chemically similar to their human counterparts.

It is an elegant chain of reasoning. But the study is not conclusive. For one thing, only half of those with nodding syndrome seemed to be producing the antibody to leiomodin-1. That, though, might be explained by the fact that many of the samples tested came from people who had been infected years before, and who may have thrown off the parasitic infection (which is susceptible to treatment with drugs) while still suffering the neurological effects. More difficult to explain is that a third of nodding-free people seemed to be making the antibody too. But perhaps it does not always attack human proteins.

Whatever the details, though, Dr Johnson’s hypothesis is tantalising. If she is right, then nodding syndrome may not be a separate disease at all, but, like river blindness, simply another symptom of infection with O. volvulus.