Table 1

Correlations between prevalence and odds ratios of risk factors for suicide in general population samples and people in contact with mental health services

VariableGeneral population studiesMental health servicesCombinedCombined take over (excluding age-specific studies)†
Spearman rank correlation coefficient (95% CI)Spearman rank correlation coefficient (95% CI)Spearman rank correlation coefficient (95% CI)Spearman rank correlation coefficient (95% CI)
UnemploymentN=22
−0.55**
(–0.17 to −0.79)		N=7
−0.93**
(−0.59 to −0.98)		N=29
−0.73**
(–0.50 to −0.87)		N=22
−0.65**
(−0.18 to −0.88)
Ethnic minorityN=14
−0.71*
(−0.11 to −0.93)		N=5
−0.60
(0.81 to −0.99)		N=19
−0.68**
(−0.18 to 0.90)		N=15
−0.66**
(−0.05 to −0.91)
Living aloneN=18
−0.28
(0.36 to −0.74)		N=4
−0.95
(0.63 to −1.00)		N=22
−0.46*
(−0.05 to −0.74)		N=15
−0.34
(0.37 to −0.79)
Unmarried†N=28
−0.35
(0.15 to −0.71)		N=9
0.15
(0.83 to −0.72)		––

  • ↵* p <0.05

  • ↵** p <0.01.

  • ↵† Correlation coefficients for the relationship between odds ratio and prevalence for being unmarried are not provided because of differences in the direction of correlations in mental health and general population samples. Findings are from a reanalysis of data having excluded 11 studies that focussed exclusively on younger or older adults.

  • N=number of studies used to generate each correlation.