Fig. 3

The range of lithium levels in drinking water and crude suicide rates (per 100,000 population). Studies with an association: these studies showed a significant association between lithium levels in tap water and suicide; if lithium levels increased, suicide rates decreased. Studies without an association: these studies did not show a significant association between lithium levels in tap water and suicide. The range of lithium levels: max value minus min value. The dotted lines show minimum levels of the range of lithium levels and crude suicide rates among studies with an association (male). Areas A, B, C, and D are divided by dotted lines (the horizontal line shows crude suicide rates 25.8/100,000, and the vertical line shows the range of lithium levels 35.05 μg /L). We used lithium levels from previous studies [9, 10, 13,14,15,16,17, 31], as well as crude suicide rates (per 100,000 population) from statistics on suicide provided by the Ministry of Health, Labor and Welfare in Japan and WHO [32]. If crude suicide rates of the survey year were not available, we substituted crude suicide rates of the nearest survey year