Association of secular trends in unemployment with suicide in Taiwan, 1959-2007: A time-series analysis

S. S. Chang*, J. A C Sterne, W. C. Huang, H. L. Chuang, D. Gunnell

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

43 Citations (Scopus)


Objective: Despite the wealth of research investigating the association of unemployment with suicide in the West, few studies have investigated the association in non-Western countries. This study aimed to investigate the relationship between secular trends in unemployment and suicide in Taiwan. Study design: Time-series analysis. Methods: Overall and age-specific suicide rates (1959–2007) for Taiwanese men and women aged 15 years or above were calculated from national population and mortality statistics. The association of secular trends in unemployment with suicide was investigated graphically and using time-series modelling (Prais-Winsten regression). Results: Rises in unemployment were associated with an increase in male suicide rates, but evidence for an association in females was limited. In the model controlling for changes in gross domestic product (GDP) per capita, GDP growth, divorce and female labour force participation, for every 1% rise in unemployment, male suicide rates increased by 3.1 (95% confidence interval 1.4–4.8) per 100,000. There is no evidence for a difference in the strength of association between men of different ages. Conclusion: Trends in suicide, particularly for adult males, appear to be influenced by unemployment. The results have implications for suicide prevention, in particular for societies facing acute rises in unemployment during recessions.
Original languageEnglish
Pages (from-to)49-54
Number of pages6
JournalPublic Health
Issue number1
Publication statusPublished - Jan 2010


  • Suicide
  • Taiwan
  • Time-series analysis
  • Unemployment


Dive into the research topics of 'Association of secular trends in unemployment with suicide in Taiwan, 1959-2007: A time-series analysis'. Together they form a unique fingerprint.

Cite this