• Job demands
• suicide ideation
• data-extraction error
• longitudinal analysis

I read with great interest the meta-analysis paper by Milner et al 1 about the associations between psychosocial job stressors and suicidality in working populations. The authors have compiled  and investigated 22 epidemiological studies on chronic job stressors (job control, job demands, job strain, colleague/supervisor support, effort–reward imbalance, job insecurity, role conflict and working hours/shift work) and suicide ideation/death in the literature, which is a very timely and important review that will contribute to the primary prevention of suicide in working populations.

However, I would like to point out one important data extraction error in their paper. The authors1 reported that high psychological job demands increased the risk for suicide death in the original paper by Ostry et al 2: OR=1.32 (95% CI: 1.31 to 1.33). They also reported a gender difference in their meta-analysis that high psychological job demands were significantly and positively associated with suicide death in male workers (OR=1.20, 95% CI: 1.10 to 1.31). However, as a matter of fact, in the original paper by Ostry et al 2 the direction of the association was the opposite: that is, high psychological job demands ‘decreased’ the risk for suicide death in Canadian male sawmill workers: OR=0.76 (95% CI: 0.67 to 0.85). It is very likely that the authors’ finding of a positive association between high psychological job demands and suicide death in male workers in the paper1 resulted from the data extraction error.

Ironically, the aforementioned data extraction error by the authors leads us into an important knowledge gap in the literature between cross-sectional and longitudinal studies with regard to the association between high psychological job demands and suicidality. Most cross-sectional studies included in the paper by Milner et al 1 supported a positive association between high psychological job demands and suicide ideation. However, in previous two longitudinal studies in Canada and Japan,2 3 high psychological job demands were negatively associated with suicide death. Given that suicide ideation increases the risk for suicide attempt or death,4 it’s reasonable to expect a positive association between high psychological job demands and suicide death in longitudinal studies. But there is no longitudinal study supporting the expected positive association in the literature. More longitudinal studies are warranted in future for the association between high psychological job demands and suicide death.