Sir,

Recently a paper suggested that the economic crisis was responsible for a dramatic increase in suicidal rates in Italy1 and the same authors insist that a similar phenomenon is happening in Greece.2 The relevant data are shown in table 1. However, interpretation can be made:

1. In Italy, the reported completed suicidal rates are almost identical with those of attempted suicide. This is highly unlikely and in contrast to the international literature. Registered suicide attempts are reported to be at least 5–10 times higher than completed suicide. With a probable rate of 1% lifetime prevalence of at least one suicidal attempt in the general population aged 18–65,3–5 the number of attempted suicides per year in Italy should be at least around 10 000–15 000 with half of them (7000–10 000) seeking medical help (thus registered).

2. Total suicidal rates remained stable with some fluctuations throughout the decade in both countries.

3. The period covered by these data does not include a dramatic economic crisis period for Italy (at least in terms of unemployment), while for Greece they cover essentially 1 year (2010).

4. Indeed, the rates of suicides due to economic reasons (as reported) doubled in Italy, but this happened in an almost linear way. This increase did not show signs of deceleration during better economic years. In any case, suicides due to economic reasons constitute a very small minority (3%–6% of total).

5. The conclusion that overall the suicidal rates were dropping in Italy before 2008 and increased afterwards is not supported by the data and neither by web appendix 2 published by the authors. There is a peak in 2003 which acts essentially as an outlier and determines the regression line. Linear solutions are not appropriate here.

6. A more realistic conclusion would be that there is an internal rearrangement in the causes of suicide attempts, mainly with reduction of the role of mental disorders and increase in more socially explained causes. Changes are so small that they fall outside statistical significance but they accumulate into the group of economic suicides and produce a disproportional change. Again the linear way of change through the years 2000–2010 suggests that methodological issues are more likely to be responsible for it.