"New Scientist Magazine reports on findings that suggest that delaying fatherhood may increase the risk of fathering children with disorders such as Apert syndrome, Autism and Schizophrenia. The article reports that 'although there is a big increase in risk for many disorders, it's a big increase in a very small risk. A 40-year-old is about 50 per cent more likely to father an autistic child than a 20-year-old is, for instance, but the overall risk is only about 1 per cent to start with.' In other words: time to start mating before those tadpoles turn into toads."
Some selector is almost always at play. Even if just in sexual selection. While I'm sure age plays a part in the increase in disorders I also wonder how much the parent carrying the disorder is related to having kids at a latter age. People with social disorders, even if mild, may reproduce less likely or when they do, may reproduce later in life. There may be more than one thing at play here.
Natural selection is not just complex, but often counterintuitive. Consider the gene for sickle cell anemia and malaria immunity. If you inherit the gene from one parent, it confers immunity to malaria. If you inherit the gene from both parents, it confers sickle cell anemia. In malaria prone environments, it was common that everyone who survived to adulthood had one copy of the gene. Children of two such parents had a 25% chance of not getting the gene, and likely dying in childhood of malaria, a 25% chance of getting two copies of the gene, and likely dying in childhood of anemia, and a 50% chance of getting one copy of the gene, and maybe living to adulthood. The gene is a mixed blessing at best, but it is strongly selected for in malarial environments, and strongly selected against anywhere else.It is quite possible that any particular undesirable genetic trait is being held in the gene pool by some desirable trait, to which it happens to be linked.