Objective Comparing the accuracy of single item about mentally tiring work against validated scales, Demand-Control (DC) and the Effort Reward Imbalance (ERI).

Methods We analysed data from the Gazel cohort, where a question about mentally tiring work was administered simultaneously with the DC (in 1997 and 1999) and ERI (in 1998) scales. Correlation and accuracy were studied comparing this single question (8 categories, and recoded into 2 or categories) with DC and ERI scales (without and with recoding into 2 categories based on usual threshold), using sensitivity, specificity, predictive values, and likelihood ratio.

Results For the years considered, 5706 (1998) to 11 304 (1997) workers had responded to the questionnaires. The demand and effort dimensions were moderately correlated with the mentally tiring work question showing a good sensitivity (>0.8), and a negative likelihood ratio (<0.33), with a possible dose-response-relationship. Specificity and positive likelihood ratio were low (respectively <0.5 and<2). Job control, Job strain and ERI were not captured by mentally tiring work, and reward only partly.

Conclusion Though a single question does not replace validated scales as the DC and the ERI scales, these results indicate that it would be possible to use simple measures in questionnaires and non-specialised cohorts for screening purpose.