Socioeconomic factors outweigh climate in the regional difference of suicide death rate in Taiwan

Abstract

The present study explored both socioeconomic and climatic factors to obtain a more comprehensive view of the asymmetric regional suicide death rate during 1998–2006 in Taiwan. The annual suicide death rate, population and meteorological data from 19 cities/counties in Taiwan were analysed by multiple regression. Five socioeconomic (sex ratio, no spouse, aged, unemployment and low income) and three climatic (temperature, rainfall and sunshine) factors were identified as significant, explaining 59.0% of the variance in the total suicide death rate. ‘Without spouse’ and ‘aged’ were associated with the highest risk, while ‘low income with financial aids’ was strongly protective. The most influential climatic factor was ‘temperature,’ which was negatively correlated with suicide. ‘Sunshine’ was positively associated with suicide. The socioeconomic and climatic factors contributed 52.7% and 6.8%, respectively, to the variance of the total suicide death rate. Limitations of the study included the fact that no individual events were considered, the study was of relatively short duration and it was confined to the territory of Taiwan. Socioeconomic factors outweighed climatic factors in explaining regional differences in the suicide death rate in Taiwan. Temperature weighed more than sunshine. ‘Thermotherapy’ seems more clinically relevant than the popular light therapy, at least in Taiwan.

Introduction

Suicide rates have been increasing in Taiwan since 1994 (Fig. 1), and suicide ranks among the 10 leading causes of death since 1997. A better understanding of suicidal behaviours may shed light on effective strategies for suicide prevention. The uneven seasonal as well as regional distribution of suicides has been of interest in previous studies. Many studies attributed variations in the suicide rate to climatic factors (Souêtre et al., 1987, Souêtre et al., 1990, Tietjen and Kripke, 1994, Preti, 1997, Preti, 1998, Wang et al., 1997, Marion et al., 1999, Petridou et al., 2003, Deisenhammer et al., 2003, Lambert et al., 2003, Rock et al., 2003, Partonen et al., 2004, Papadopoulos et al., 2005, Heerlein et al., 2006, Lee et al., 2006, Nicholls et al., 2006, Rocchi et al., 2007, Ajdacic-Gross et al., 2007). Among the climatic factors, temperature (Souêtre et al., 1990, Souêtre et al., 1987, Preti, 1998, Marion et al., 1999, Deisenhammer et al., 2003, Partonen et al., 2004, Lee et al., 2006, Ajdacic-Gross et al., 2007, Dixon et al., 2007), sunshine (Grove and Lynge, 1979, Souêtre et al., 1987, Souêtre et al., 1990, Tietjen and Kripke, 1994, Preti, 1998, Petridou et al., 2003, Lambert et al., 2003, Papadopoulos et al., 2005, Heerlein et al., 2006, Ajdacic-Gross et al., 2007) and rainfall (Souêtre et al., 1990, Preti, 1998, Lambert et al., 2003, Nicholls et al., 2006, Ajdacic-Gross et al., 2007) have been the focus of investigative interest. However, the findings of these studies are equivocal, even controversial. For example, Deisenhammer et al. (2003) and Lee et al. (2006) found that temperature was positively associated with suicide. A positive cross-correlation also emerged between suicide and temperature data for the whole time series in the study by Ajdacic-Gross et al. (2007). But they found a major peak in associations for winter frames, and they considered it hardly compatible with the hypothesised role of temperature in suicide seasonality. As far as sunshine was concerned, Papadopoulos et al. (2005), Petridou et al. (2002), and Lambert et al. (2003) believed it was positively correlated with suicide, but Tietjen and Kripke (1994) found the opposite, while Grove and Lynge (1979) found sunshine irrelevant. Similarly, there were conflicting findings regarding the relationship between precipitation and suicide.

Climatic factors have not been universally implicated in the occurrence of suicide and, moreover, could only partially explain suicide within certain social contexts, since socioeconomic factors such as marital status, gender, economic condition, unemployment (Agerbo et al., 2007), education (Lorant et al., 2005) and occupation (Nishimura et al., 2004) have also been considered to contribute to suicide death rates. Although suicidal acts are generally accepted as the results of an interaction of biological, psychological, and socioeconomic factors, most previous studies did not take them into account as a whole. Therefore, the estimated meteorological parameters could be confounded by the socioeconomic factors. This study was designed to simultaneously explore the roles of both socioeconomic and climatic factors in accounting for regional differences in the suicide death rate in Taiwan.