- 日本生気象学会雑誌 (ISSN:03891313)
- vol.23, no.2, pp.65-70, 1986-10-15 (Released:2010-10-13)
Seasonal variation of suicidal deaths in Japan for 1900-41 and 1947-82 was studied. As suicide seasonality index, we used the sum of deviations of the monthly suicidal rates from the annual average, which was normalized to 100. The long-term trend of the suicide seasonality was negatively correlated to the logarithm of the per capita GNP. The correlation coefficient between them was -0.83 for men and -0.88 for women, respectively (p<0.001) . Elimination of the effect of GNP on suicide seasonality by using the regression line of the latter on the former lead to the “corrected” suicide seasonality. The short-term fluctuations of five-year moving averages of the corrected suicide seasonality resembled to those of annual hours of sunshine. The correlation coefficient between them was 0.69 for men and 0.64 for women, respectively (p< 0.001) . The corrected suicide seasonality was similarly correlated to the sunshine seasonality, and the correlation coefficient between them (each, five-year moving averages) was 0.51 for men and 0.41 for women, respectively (p<0.001) . Although these values are smaller than those obtained for the correlation of the corrected suicide seasonality to the annual hours of sunshine, the differences are not significant. One possible explanation is that annual hours of sunshine and/or sunshine seasonality influence the suicide seasonality and that the two sunshine factors are proportional to each other. With a moving average period more than three years, we obtained significant correlation coefficient between the corrected suicide seasonality and annual hours of sunshine. We discussed the limitation of this study and proposed future studies on the association of suicide with sunshine.