Research reportThe sales of antidepressants and suicide rates in Norway and its counties 1980–2004
Introduction
Mental health disorders are important causes of suicide (Harris and Barraclough, 1997), and particularly important are affective disorders, because different subtypes are found among about 60% of suicides in the general population (Cavanagh et al., 2003). The risk of suicide also increases with the severity of the depressive episode (Kessing, 2004). Treatment of such conditions is thought to be beneficial in the prevention of suicide. The introduction of newer antidepressants, and especially the selective serotonin reuptake inhibitors (SSRIs), at the start of the 1990s accelerated the diagnosis and treatment of depression. Far more people were exposed to drugs that could alleviate depression. Together with this increase in sales of antidepressants, many western countries experienced a decline in their national suicide rates. Several researchers, particularly Isacsson (Isacsson, 2000), have promoted the idea that these two phenomena are related and increased treatment of depression has led to the observed decline.
Although suicide is a major public health issue, it is a rare phenomenon at an individual level. Only very large, possibly unsustainable, randomized controlled trials (RCTs) would, by themselves, have enough power to demonstrate any difference in suicide rate between a treatment and a placebo. Meta-analysis of such studies could, to some extent, overcome this (Khan et al., 2003, Gunnell et al., 2005). The follow-up time of RCTs included in the meta-analysis is, however, too short to capture the whole at-risk period for suicide in connection with depression. Furthermore, suicidality is usually an exclusion criterion in such trials, because of ethical constrains and practical difficulties.
Ecological investigations into the relationship between the sales of newer antidepressants and suicide rates have been performed in several countries, with varying time spans, different age groups, and various types of co-variates and statistical approaches (Table 1). With the exception of the studies in Iceland, Italy and Slovenia, which have not shown a decline in suicide rates, most of these studies find that the increased sales of newer antidepressants seem to be related to the decrease in suicide rate. This is also found in other communications from Hungary (Rihmer et al., 2001, Rihmer, 2003) and Britain (Gunnell and Ashby, 2004). What separates them are the result interpretations, ranging from a belief in a causal relationship to stating that sales figures are indirect measures of better diagnosis of depression and more treatment optimism, or even that the relationship is purely spurious.
The aim of the current study was to investigate the relationship between antidepressant sales and suicide rates as seen in Norway as a whole, and in its counties, on the basis of observations made from 1980 to 2004. As a completed suicide is not just a matter of mental illness and its treatment, we also included data on alcohol consumption and unemployment as confounders, both of which have been found to be related to national suicide rates (Rossow, 2005, Platt and Hawton, 2006).
Section snippets
Sources of data
Suicide rates were provided by Statistics Norway. The figures are given in deaths by suicide per 100,000 inhabitants. Suicide rates were available for the country as a whole from 1980 to 2004, but only up to 2003 when dealt with by the county. Statistics Norway also supplied data on sales of alcohol for the years 1980–2004, although data from 1998 were missing. Alcohol sales data were available only for the country as a whole. Figures are given in liters of pure ethanol sold through official
Model for sales of antidepressants
Three models using drug sales figures as independent variables and suicide rate as dependent variable were used:
- 1.
The relationship between suicide rates and sales figures for all antidepressants.
- 2.
The relationship between suicide rates and sales figures for non-TCAs.
- 3.
The relationship between suicide rates and sales figures for SSRIs.
Only the main effects were studied. The model giving the best fit was option 2: the relationship between suicide rate and sales of non-TCAs. We chose this model because
Discussion
This ecological study of the increase in sales of antidepressants and fall in suicide rate in Norway and its counties found a clear, statistically significant association between these two variables. Suicide rates seemed to fall mostly in association with an increase in the sales of non-TCAs. The relationship was prominent for sales figures that were below average, and not for higher sales figures, indicating that an initial increase in sales was related more to a decrease in suicide rate.
Conclusion
The fall in suicide rates in Norway and its counties was related to the increased sales of non-TCAs. The effect was mostly the result of a sales increase in the lower sales segment, indicating that an explanation might be a change from the more toxic TCAs or heightened GP awareness of depression and its treatment.
Acknowledgements
This work was performed without any external source of finance. The authors would like to thank Berit Olsen and Tove Granum at the Norwegian Institute of Public Health for the help in collecting the wholesale data on drugs.
Conflicts of interest: JGB and AT do not have any commercial associations that might pose a conflict of interest in connection with the manuscript. FAW has received an honorarium from Pfizer and H. Lundbeck for lectures on suicide prevention in general.
References (41)
- et al.
Antidepressant prescribing and suicide rate in Northern Ireland
Eur. Psychiatr.
(2003) - et al.
Do antidepressants t(h)reat(en) depressives? Toward a clinical judicious formulation of the antidepressant–suicidality FDA advisory in light of the declining national suicide statistics from many countries
J. Affect. Disord.
(2006) - et al.
Depression and suicide on Gotland. An intensive study of all suicides before and after a depression-training programme for general practitioners
J. Affect. Disord.
(1995) Preventing suicide and premature death by education and treatment
J. Affect. Disord.
(2001)Alcohol use and suicide rates
Med. Hypotheses
(2005)- et al.
Antidepressant drug use in Italy since the introduction of SSRIs: national trends, regional differences abd impact on suicide rates
Soc. Psychiatry Psychiatr. Epidemiol.
(1999) - et al.
Behandling av alcohol — och narkotikaproblem
- et al.
Antidepressant medication and suicide in Sweden
Pharmacoepidemiol. Drug Saf.
(2001) - et al.
Psychological autopsy studies of suicide: a systematic review
Psychol. Med.
(2003) - et al.
Application of least squares regression to relationships containing auto-correlated error terms
JASA
(1949)