# Association Between Suicide and Physical Illness in Japan: An Analysis Using Statistical Adjustment Based on Regional Attributes

## TL;DR

This study finds that cerebral infarction and diabetes are linked to higher suicide rates in Japan, emphasizing the need to consider regional socioeconomic factors in suicide prevention.

## Contribution

The study introduces a novel approach by adjusting for regional attributes when analyzing the relationship between physical illness and suicide.

## Key findings

- Suicide was significantly associated with cerebral infarction and diabetes after adjusting for regional factors.
- The proportion of older people and household assets were identified as significant regional factors affecting suicide rates.
- Adjusting for regional attributes revealed significant associations that were not apparent without adjustment.

## Abstract

Background: Suicide is a serious social problem in Japan, and suicides related to health problems are predominant. To date, no studies have considered and adjusted for suicide-associated factors based on regional attributes when examining the relationship between physical illness and suicide.

Objective: Using open data, we examined the relationship between the number of outpatient visits by disease and the number of suicides per prefecture, while adjusting for regional confounding factors.

Methods: We conducted stepwise multiple regression analysis on data from 47 Japanese prefectures, adjusting the number of suicides per prefecture for suicide-associated factors based on regional attributes, such as the proportion of older people, household assets, happiness level, and sunlight hours. We analyzed six diseases potentially associated with suicide: neoplasms, gastric/duodenal ulcers, liver disease, glomerular disease/renal failure, diabetes, and cerebral infarction.

Results: Suicide was significantly associated with cerebral infarction and diabetes (scaled estimates: 1.10 and 0.842; adjusted power: 83.2% and 85.7%, respectively). When analyzing without adjustment for regional attributes, no significant associations were found, highlighting the importance of controlling for confounding factors. The proportion of older people (estimate: 0.343, p = 0.002) and household assets (estimate: −0.138, p = 0.004) were identified as significant regional factors affecting suicide rates. These findings suggest that both clinical factors and socioeconomic conditions must be considered in suicide prevention efforts.

Conclusions: Our findings highlight the importance of monitoring patients with cerebral infarction and diabetes for suicide risk, while considering regional socioeconomic factors. Primary healthcare workers, who are available, accessible, knowledgeable, and committed to providing care, may play a crucial role in suicide prevention by paying particular attention to these patient populations. These results provide valuable insights for developing targeted suicide prevention strategies in clinical settings.