[Predictive value of the quality of life questionnaire in the elderly].

Objective: To evaluate the predictive value of the quality of life (QOL) questionnaire at 8 years and analyse which of its 5 dimensions better predict adverse results in health.

Design: Descriptive, longitudinal, and retrospective study.

Setting: Three basic health areas in Cuenca province, Spain.

Participants: We randomly selected 519 subjects over 64 years, taking addresses in which more than 1 person over 64 years resided as a sample marker.

Main Measurements: The QOL questionnaire was used to evaluate the quality of life of the subjects in 1994 and in 2002. Main results. Between 1994 and 2002, 25% of the participants died, 2.6% were institutionalised and there 3.6% losses. The QOL questionnaire has a mortality predictive capacity at 8 years, the cut off point of its global index being situated around 13 points. The sensitivity of the model is poor (41.5%), with a high specificity (93.5%) and a good classification percentage of 80.3%. The positive and negative probability ratios were 6.36 and 0.62, respectively. Using a Cox regression model, it is seen that of the 5 dimensions of the QOL test, activities of daily living show a relationship with mortality (risk ratio or OR=1.3; 95% confiance interval [CI], 1.049-1.611), mental health (OR=1.048; 95% CI, 1.054-1.620), and economic resources (OR=0.806; 95% CI, 0.648-1.003).

Conclusions: The QOL questionnaire has a mortality predictive capacity at 8 years.