[Validity of the Connor-Davidson resilience scale (10 items) in a population of elderly]

To examine the psychometric properties of the Spanish version of the CD-RISC 10-items in a sample of the elderly population of Cuenca, Spain; and to assess if the Spanish version preserves the same factorial structure as the original one.

A cross sectional study was conducted in five health centres in the province of Cuenca. The study included a random sample of 500 people aged 60-75 years who lived within the community.

The sociodemographic variables, CD-RISC-10 (resilience), PSS (perceived stress), SF-12v2, Pfeiffer Short Portable Mental Status questionnaire, GDS (geriatric depression), and MOS (social support), were all used to gather data. The number of factors underlying the CD-RISC 10 items was analysed using exploratory factor analysis. Furthermore, this factor structure was tested by confirmatory factor analysis.

A single underlying factor was found in the CD-RISC 10 items. The factor structure was tested using confirmatory factor analysis and it was found that a single factor model showed acceptable goodness of fit values for both men and women. Convergent validity was performed to test whether the mean scores of the variables were significantly associated with resilience. The Cronbach's α coefficient for the CD-RISC 10 items scale was 0.81. The overall scores of the Spanish version of the CD-RISC-10 items correlated directly with MOS and the physical and mental components of SF-12, and inversely with the PSS and GDS.

The Spanish version of the CD-RISC-10 showed good psychometric properties. Thus, it can be used as a valid and reliable instrument to measure resilience in the non-institutionalized older population.