Genetic Linkage for Schizophrenia?

To the Editor: I would like to comment on the article by Stephen J. Glatt, Ph.D., et al. (1), which used meta-analysis in an attempt to demonstrate a genetic linkage for schizophrenia on the catechol O-methyltransferase (COMT) gene. I find this trend of using meta-analysis to resurrect largely negative genetic linkage studies disturbing. It appears to be nothing more than a manipulation of data to obtain a desired result.

Of the 18 studies looked at in this meta-analysis, only four originally had positive linkage results. Two of these showed a positive linkage between schizophrenia and the Met allele of the COMT gene, while the other two showed a positive linkage to the Val allele. These are contradictory results. Common sense should dictate that these contradictory positive studies might be nothing more than random false positives. Surprisingly, Dr. Glatt and colleagues never even considered this possibility.

Instead, they parsed the data in an effort to find positive results by first dividing the studies into those that used a case-control method versus those that were family based, then they further divided the family-based studies by the ethnicity of the subjects. When looking at study data after the fact, one can always find subgroups that give the appearance of a positive result. It is hard to imagine a reason that one would expect opposite results for case-controlled studies versus family-based studies. Moreover, the authors’ suggestion that the Val allele confers some risk factor for schizophrenia in those of European descent but not in those of Asian descent has no scientific basis whatsoever and, in my opinion, enters the realm of eugenics.

I understand that the search for genetic linkages to schizophrenia has been frustrating. As Dr. Glatt et al. pointed out, “Genetic linkage studies have failed to endorse schizophrenia linkage universally with any chromosomal region” (p. 469). The same could be said for genetic linkage studies of any mental disorder. I suggest that it is the genetic linkage studies themselves that are the problem, by giving frequent false positive results without successfully identifying any genetic linkages to date. Meta-analyses of false positive results will not garner any more merit for the original studies.