Using the 12-item Short Form health survey (SF-12) to measure quality of life among older people.

Aims: To present descriptive data, study psychometric abilities, and investigate the characteristics of non-respondents to SF-12.
Method: The study had a randomized sample (n=4278; age: 75-105) stratified in four age groups. Factor analyses were performed to investigate the factor structure of the instrument, and convergent and discriminant validity was assessed by correlation between summary scores and by means of various measures (such as symptoms and ADL). Mean scores and Cronbach's alpha were computed for the total sample and for each stratum. Respondents and non-respondents to SF-12 were compared regarding various measures. A multiple logistic regression analysis was conducted to identify determinants for respondents/non-respondents.
Results: Mean scores in the total sample were 37.5 SD 11.9 for the physical component summary score (PCS) and 50.3 SD 11.5 for the mental component summary score (MCS). Both PCS and MCS decreased significantly with higher age. The instrument showed acceptable convergent/discriminant validity and Cronbach's alpha, but the factor solution did not follow the intended structure of the instrument. Completion rate for all 12 items was 85.5% in the total sample, and the characteristics of non-respondents to SF-12 were higher age, being female, living alone, living in special accommodation, being worried about health, being in need of help with activities of daily living, reporting fewer symptoms, and having a smaller social network than respondents.
Conclusions: The instrument generally showed acceptable validity and reliability, except for its factor structure. Furthermore, new determinants for non-respondents of SF-12 were found, such as type of living conditions, being worried about health, and social network.