Abstract
Dear Sir,
I was present at Professor Kosky's address to the Royal Australian and New Zealand College of Psychiatrists in November 1997 on Durkheim and Suicide. The argument he put forward was wrong then and is wrong now (Australasian Psychiatry, December 1998). Sadly it lacks any basic knowledge of epidemiology. Professor Kosky puts forward four problems in Durkheim's argument.
Firstly there is the problem of definitions. This is a problem with all explanations which try to explain the suicide rate. It doesn't invalidate Durkheim's argument about the role of social forces. Using Kosky's example of the current suicide rate in Turkey being ten times less than in the US, one could just as easily ask the question does this mean that mental illness is ten times less in Turkey?
Secondly, Kosky asks what are the social forces which affect individuals. Durkheim was quite clear about this. He regarded social integration as the social force which affected the suicide rate. Religion was a proxy marker for this. To say he ignored the social revolution of 1848 when he was writing in 1897 is like saying that in trying to explain the suicide rate of different countries today we ought to consider which countries were involved in forming the United Nations!
Thirdly, and most bizarrely, Kosky states that ‘no amount of associations between a rare phenomenon and a common one can establish a causal link between the two’. This is nonsense. Smoking occurs in 27% of the New Zealand population (a common occurrence), about 900,000 people. Lung cancer occurs in less than 31/100,000 people a year (a rare occurrence). Is he saying that smoking is not related to lung cancer? A measure of the burden of illness caused by a risk factor is the population attributable risk; he will find that this figure is high for many social factors such as unemployment, which probably explains Australian Government funding priorities.
Lastly, he discusses the issue of causality. Explaining changes in the suicide rate is different to explaining individual cases of suicide. For instance, high salt diets are associated with high national rates of myocardial infarction but not necessarily with every individual heart attack. This is the advantage of studying populations, it can describe risk factors for an outcome which are readily apparent at a population level but not at an individual level.
Durkheim still has validity today. The social explanation of changes in the suicide rate is still, to my mind, a compelling argument.