There is a common observation, going back to Dalton, that families who are quite bright have more madness within. The Scandiavians have managed to do a correlation of individual school achievement and family cognitive achievement from one of their very large cohorts. This kind of work only happens in Scandinavia, where your school records, your military IQ, and your hospital admissions and assessments are all on public databases and all accessible.

What you get is interesting. Poor School achievement at entry high school (Grade 9: the test was used to stream kids to high schools) was predictive of a later diagnosis of psychosis (they were using diagnoses of schizophrenia or bipolar disorder) in later life.

However, it is worth noting that this is a deviation from where the family is. Note the lowest and highest SD. In the lowest SD group, being cognitively slow does not have that great a penalty: the family is slowed. However, being of normal intelligence was protective. Among intellectual families, however, not being in the top 5% — being normal — predicted illness.

A couple of caveats. The baseline rate of these disorders is around 0.5% of the population (from their figures). Being at higher risk is still a low risk. The reason for relative poor achievement cannot be explored: these kids may have been injured, abused, neglected or be born later, which is also a risk factor.

Dalton, as usual, is half wrong. It may not be coming from a family that is bright that is the risk factor, but the return to the mean.