Skip to content

Familial Linkage between Neuropsychiatric Disorders and Intellectual Interests

When I spoke at Princeton last year, I talked with neuroscientist Sam Wang, who told me about a project he did surveying incoming Princeton freshmen about mental illness in their families. He and his coauthor Benjamin Campbell found some interesting results, which they just published:

A link between intellect and temperament has long been the subject of speculation. . . . Studies of the artistically inclined report linkage with familial depression, while among eminent and creative scientists, a lower incidence of affective disorders is found. In the case of developmental disorders, a heightened prevalence of autism spectrum disorders (ASDs) has been found in the families of mathematicians, physicists, and engineers. . . .

We surveyed the incoming class of 2014 at Princeton University about their intended academic major, familial incidence of neuropsychiatric disorders, and demographic variables. . . . Consistent with prior findings, we noticed a relation between intended academic majors and ASDs. Looking for relations between other neuropsychiatric disorders and academic interest we also noted a heightened prevalence of bipolar disorder, major depressive disorder and substance abuse in the families of those pursuing the humanities. A composite score based on these four heritable disorders was strongly correlated with a student’s intended academic major. Thus, familial risk toward a spectrum of psychopathologies can predict propensity toward technical versus humanist interests.

When I spoke with Sam last year we discussed various ways to analyze the data as well as various interpretations of the results, but I don’t actually remember any of our conversation except for the bit where he described to me how they conducted their study.


  1. TicoG says:

    As if humanities majors didn’t already get enough criticism! Interesting to see social sciences had intermediate rates between humanities and technical.

  2. David Shor says:

    I get a bit iffy measuring very low probability events via survey data. Contamination is a big issue.

  3. mpledger says:

    I suspect that the family member with autism is a sibling and the family member with the other disorders are, by and large, parents (except, perhaps, drug addiction) –
    given that the average age of onset for non-austisic/non-drug disorders listed above is older than the students themselves and that you’d expect half the siblings to be younger than the student i.e. the sibling haven’t had a chance to develop it yet.

    That makes for very different dynamics within a family about how much family help is asked for/is necessary from the student and how much time they have available to master technical subjects necessary to go on and do STEM at college.

    You can do well in some humanties subjects at College without any High School background in them, it’s pretty hard to do that with STEM.

  4. revo11 says:

    Interesting correlations to report, but the population composition of disciplines are different in so many ways.

  5. MAYO says:

    I haven’t had a chance to read the paper, but the conclusion strikes me as way too strong.

    Firstly they seem to imply that whether you major in science or humanities, there’s an increased tendency toward some psychological disorder or other? But even more questionable is the following:

    “Thus, familial risk toward a spectrum of psychopathologies can predict propensity toward technical versus humanist interests.”

  6. K? O'Rourke says:


    Great to see you are going forward with something many will point out you can’t/shouldn’t do (like that meta-analysis of polls).

    When you are ready for another distraction …

    There was this physical machine that Galton constructed to carry out Bayesian analysis physically.

    “Darwin, Galton and the Statistical Enlightenment.” Journal of the Royal Statistical
    Society, Series A, 2010, Vol. 173 (Part 3, July), pp. 469-482. (Just the simple two stage Quincunz)

    Now over coffee at the airport, it did not seem impossible (n=2)
    that brain physiology could implement something like that (perhaps not achieving independence of prior and _likelihood_ as in Galton’s machine).

    Other than ruining the fun by pointing out why it would be impossible is there any accessible literature for those who do not know better – to gaze at?