7.1. Seasonal variation of suicides in Finland during 1980-95

All completed suicides in Finland during 1980-95 were investigated in terms of their seasonal distribution. The time trends in the rates and the seasonality of violent and non-violent suicides were also examined.

The time series for suicides were obtained from official statistics of the Finnish, Statistics Centre. They included the monthly numbers of suicides according to the year of death, gender and age group of a suicide victim, as well as the method of a suicide. In further analyses, suicides by hanging, drowning, shooting, wrist cutting or jumping from a height were defined as violent and others as non-violent suicides.

A total of 21279 suicides were completed during the 16-year study period. Male suicides covered 79% of the total suicide rates. The distribution of suicides between age groups (<=39 years, 40-64, >=65 years) was 44%, 42% and 14%, respectively. Violent suicides accounted for 70% of all suicides.

The observed seasonal distribution of the suicides differed statistically significantly from the expected distribution of suicides under the null hypothesis that the occurrence of suicides follows a uniform distribution after adjusting for calendar effect and leap years. A significant excess of total suicides was found during spring/summer (May-July) and a significant trough during winter/spring (December-March) months.

The seasonal pattern of suicides differed significantly between genders. The seasonal distribution of males was unimodal with a peak in spring/summer (May-July) and a trough in winter/spring (December-March). Female suicides have a significant peak in spring (May) and autumn (October), and a trough in winter (December-February). In all age groups, the incidence of suicides was significantly elevated during spring/summer (May-July) and significantly depressed during winter/spring (November-March). In the oldest age groups, there was also a trend toward a secondary peak of suicides in autumn (August-September). The peak incidence for violent suicides was during spring/summer (May-July), while the troughs fell in the winter/spring (December-March). In non-violent suicides, significant excesses of suicides were present in both spring (May) and autumn (October), but no significant trough periods were observed.

The rate of violent suicides per 100 000 mean population increased significantly during the 1980-1990 period, but decreased significantly thereafter. Regarding violent suicides, overall seasonality existed in every successive time period (1980-84, 1985-90, 1991-95) with a peak in spring/summer and a trough in winter. The rate of non-violent suicides increased significantly over the whole study period. The only period, which showed a significant overall seasonality in relation to non-violent suicides, was the period of 1980-84 with a spring peak and a winter trough. However, also in 1985-89 there was a significant winter trough in non-violent suicides.

The fact that the suicide mortality in Finland is one of the highest in the world means that every study, which might be of help in reducing the incidence of suicides, is justified. To investigate the seasonal distribution of suicides is one way to find explanations for this major public health problem. Based on the results of the present study and on the fact that climatic conditions in Finland vary largely between seasons, it was concluded that the observed seasonal distribution of suicides in Finland is connected at least in part with the seasonal changes in the climatic conditions, such as changes in the amount of light and temperature. However, seasonal changes in the psychosocial habits of people and a human’s biological processes are also likely to contribute to the seasonal distributions of suicides. The seasonal distribution of suicides was dependent on gender and age group of a suicide victim, as well as the method of the suicide. Additional studies taking into account also the effect of possible psychological, sociological or biological factors are necessary to identify their putative influences on the seasonal variations of suicides.