Why are kids of older dads at higher risk for mental illness?

Poor sperm cells.

Ever since it was first observed that the children of older fathers are at a higher risk for psychiatric disorders, those tadpole-shaped cells have taken the blame because they can accumulate harmful new mutations. But research published Monday suggests that those accusations may have been false.

Instead, the main culprit might just be plain old, conventional patterns of inheritance — although there could be many elements at play.

“It is likely that there might be multiple causes explaining this,” said Emma Frans, an epidemiologist at the Karolinska Institute in Sweden, who was not involved in the study.

It’s true, sperm cells had seemed like the most plausible suspect. That’s because they are made by special stem cells called spermatogonia, which sit in a man’s testicles, and keep dividing throughout his life. Those cells can either become yet more sperm factories, or they can become sperm cells themselves, replete with tails and rounded heads, raring to fertilize eggs.

But with each spermatogonial division, there is a chance that some bit of DNA will get mangled in the process, causing a new mutation. So the longer a man has been alive, the more likely it is that his sperm factories contain some of these new mutations. And when the factory is contaminated, so is its product.

Scientists had assumed that these new mutations were what caused the higher rates of disorders like autism and schizophrenia among children with older-than-average fathers. After all, the evidence that these mutations were accumulating was very solid.

“There was plausible biology behind the old explanation, but that doesn’t mean that it’s right,” said Steve McCarroll, director of genetics at the Broad Institute’s Stanley Center for Psychiatric Research. “It was so plausible that it became conventional wisdom.”

But a team led by researchers at the University of Queensland, in Australia, wanted to question that conventional wisdom about whether all of those new mutations could actually account for all of the increased risk. So they built some mathematical models. Five, to be exact.

Each one was a kind of universe in which different possible explanations for the older-father effect had different weights. Put in real data — such as the rate of new mutations in older men’s sperm, or common genetic differences that have been passed on for generations — let the model work its magic, and out comes the increased risk that the offspring of fathers 10 years older than the average would have of being diagnosed with psychiatric illnesses if that model were true. The closer that increased risk is to that observed in epidemiological studies, the more likely it is for that model to be the truest of the five.

But no matter how they swung the models, those new mutations that accumulate in sperm could only explain between 10 and 20 percent of that increased risk, said lead study author Jacob Gratten, a neurogeneticist at the University of Queensland, who reported the findings in Nature Genetics.

“It suggests that the story is more complicated than simply ‘age-related mutations equals increased risk,’” said Gratten. “Even though they do contribute, they don’t explain everything, so other factors must be contributing, as well.”

One plausible explanation, he said, could be that those fathers who are genetically predisposed to psychiatric disorders may also happen to be predisposed to having children later, which would increase the chances of these disorders showing up in their children.

Some are tempted to look to epigenetics — the process by which environmental factors change the chemical tags on DNA, thus switching on or off the expression of certain genes. Yet McCarroll, who was not involved in the study, warns against it as an explanation.

“There is no evidence that that ever contributes to schizophrenia,” he said. “It’s an interesting idea that many scientists would like to be true. But it has never, ever been shown to be true.”

Gratten’s models show that even a very weak correlation between the age at which a father start having children and the kids’ likelihood of getting schizophrenia or autism would be enough to explain much of that increased risk.

But, he noted, “This is just a modeling study.”

To confirm these hypotheses, he would need to do huge population-wide studies — so huge that such datasets do not even exist yet.