Work ability (WA) is an important concept in occupational health research and for over 30 years assessed worldwide with the Work Ability Index (WAI). In recent years, criticism of the WAI is increasing and alternative instruments are presented. The authors postulate that theoretical and methodological issues need to be considered when developing alternative measures for WA and conclude that a short uni-dimensional measure is needed that avoids conceptual blurring. The aim of this contribution is to validate the short and uni-dimensional WAI components WAI 1 (one item measuring "current WA compared with the lifetime best") and WAI 2 (two items assessing "WA in relation to the [mental/physical] demands of the job"). Cross-sectional and 12-month follow-up data of two large samples was used to determine construct validity of WAI 1 and WAI 2 and to relate this to respective results with the WAI. Data sources comprise nurses in Europe investigated in the European NEXT-Study (Sample A; Ncross-sectional = 28,948 and NLongitudinal = 9462, respectively) and nursing home employees of the German 3Q-Study (Sample B) where nurses (N = 786; 339, respectively) and non-nursing workers (N = 443; 196, respectively) were included. Concurrent and predictive validity of WAI 1 and WAI 2 were assessed with self-rated general health, burnout and considerations leaving the profession. Spearman rank correlation (ρ) with bootstrapping was applied. In all instances, WAI 1 and WAI 2 correlated moderately, and to a similar degree, with the related constructs. Further, WAI 1 and 2 correlated with WAI moderately to strongly with ρ ranging from 0.72-0.76 (WAI 1) and 0.70-0.78 (WAI 2). Based on the findings and supported by theoretical and methodological considerations, the authors confirm the feasibility of the short measures WAI 1 and WAI 2 for replacing WAI at least in occupational health research and employee surveys.
Keywords: WAI; measurement; occupational epidemiology; occupational health; work ability; work ability index.