2016 Fiscal Year Research-status Report
Global Perspectives of Seasonal Changes of Suicide and Associations between Suicide and Weather
| Project/Area Number |
16K19773
|
|---|
| Research Institution | Nagasaki University |
|---|
Principal Investigator |
金 允姫 長崎大学, 熱帯医学研究所, 助教 (40746020)
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Keywords | suicide / temperature / season / time-series / multi-country |
|---|
| Outline of Annual Research Achievements |
Principle investigator (PI) collected daily time-series of suicide and temperature from 299 communities in 10 countries with study periods from 4 to 41 years. The number of locations increased by approximately 100 communities from the initial plan and the study periods were extended in some communities.
PI has checked the quality of the data and merged the suicide with weather to construct the community-specific daily time-series data to be ready for statistical analysis. As expected, the data in some countries had a substantial number of days without suicide due to their low suicide rates.
For descriptive analyses, seasonal changes in suicide were evident with the spring peak in some countries (e.g. Japan, South Korea, Taiwan, and the UK), whereas there was weak evidence for the seasonal effect on suicide in others (e.g. Brazil). The countries with evident seasonal patterns tended to have clearer patterns of day-of-week on suicide.
PI reviewed statistical methods to estimate the non-linear temperature-suicide association, although planned to be done next year and found the more flexible approach than the two statistical methods initially planned, which is conditional Poisson regression model. The new statistical method was tested using the UK data.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Although the primary data of suicide and temperature have been successfully collected and covered more locations and longer periods than the initial plan, some minor variables (i.e. methods of suicide, sunshine duration, and humidity) have been collected in some of the communities.
The suicide data consist of subgroup of sex and five age groups except for Taiwan, and a half of countries provided the method of suicide. While temperature data were available for all countries, a half of countries provided sunshine duration and relative humidity. In addition, the data of regional characteristics did not be collected yet.
However, a review of statistical methods, initially planned next year, has been done and some of the data were tested by using a possible statistical method.
|
| Strategy for Future Research Activity |
The research goals next fiscal year are to estimate weather-suicide associations and to find a critical range of temperature with the highest risk of suicide.
Community-specific analysis will be implemented to estimate temperature-suicide association for the 299 communities. Although it was planned to use time-series analysis and case-crossover design separately, the more flexible approach, as found the first year, will be used. The approach has advantages of both methods, called as a conditional Poisson regression model.
The association between sunshine duration and suicide will also be estimated using the collected data in a half of countries. In addition, city- or country-level characteristics (e.g. latitude, population, or socioeconomic status) will be collected by websites (e.g., the UN and WHO).
|
| Causes of Carryover |
It was expected to spend the budget to obtain a part of the weather data. However, it did not cost to obtain the data so far.
|
| Expenditure Plan for Carryover Budget |
The remained budget will be used to extend possible travel expenses for research meetings discussing statistical methodology.
|
