Research, evaluation, analysis

Is Suicide Hereditary?

One of the big news stories of the past week (January 22-28, 2000) was a report of a research study which confirmed the idea that suicidal tendencies may be inherited. The story received extensive coverage, but when we look closer at the study we learn, as we often do, that the findings were blown out of proportion.

The researchers, Drs. David Bakish and Pavel Hrdina of Royal Ottawa Hospital, did manage to get some important reservations about their research mentioned in news coverage. Mostly, though, the press played the increasingly popular game of glossing over the facts and concentrating instead on sensational speculation. Let's look at exactly what Bakish and Hrdina found.

They found that the percentage of a group of suicidal depressive patients with a specific gene mutation was higher than the same percentage in a group of people with no mental illness. Incidentally, they did not, as many news stories implied, find that this percentage was higher among suicidal depressives than among non-suicidal depressives. The sample of non-suicidal depressives was small and a statistical test does not find a significant difference between the two groups in this percentage.

So what was the percentage? It was 42% of the group of suicidal depressives. The percentage is based on a sample of 78 suicidal depressives, so 32 of the 78 had the mutation. In other words, more than half did not have the mutation, so leaping to the conclusion that suicide is genetic is clearly unjustified. If suicidal tendencies are usually unrelated to genetic differences, then obviously suicide cannot be described as a genetic phenomenon.

Neither could you draw the less spectacular conclusion that the specific gene mutation predisposes people to suicide without knowing considerably more about the sample. The groups of depressive and non-depressive people were matched for age, sex, and ethnicity, but many other characteristics which might affect the tendency to suicide were not controlled. For example, the press release from Royal Ottawa Hospital mentions alcoholism and abuse in childhood as contributing environmental factors in suicide, and another well-known factor associated with depression is poverty. If the groups differed on these or other factors associated with suicide and depression, then the relationship of the gene mutation to suicide may be indirect or even accidental.

The researchers did manage to impress upon some journalists that the finding needs to be replicated that is, that other researchers have to find similar differences in similar studies. Just as your next door neighbour's purchase of a winning lottery ticket does not imply that your own next purchase will be a winning one, so a single finding in a single study does not imply that the next study will produce the same finding. The finding could just be, like a winning lottery ticket, an accident.

Finally, the way in which this gene mutation is supposed to have its effect is not clear from the information which has been published in the last couple of days (the scholarly article about this research is not to appear till February 7). The press release implies that the mutation increases (sic) the number of serotonin 2A receptors in the brain, but nowhere is it explicitly stated that the mutation is known to have this effect.

In summary, the findings reported by Drs. Bakish and Hrdina seem to be what researchers unblushingly call suggestive. They are certainly worth investigating with further research. However, they do not justify the conclusion that suicide has any genetic origin.

Is Suicide Hereditary? © 2000, John FitzGerald
Home page | Decisionmakers' index | E-mail