Abstract 1482: Quality of Life Questionnaire Responsiveness and Lack of Relationship to Functional Capacity: Findings From the Exercise Rehabilitation Trial (EXERT)
Background: There have been many clinical trials evaluating exercise training (ET) in heart failure (HF) that assessed quality of life (QOL) as an outcome. Many of these trials have not detected significant changes in QOL scores despite exercise capacity changes. This analysis evaluated the responsiveness of four different QOL questionnaires administered in the Exercise Rehabilitation Trial (EXERT).
Methods: 181 patients (age 66±1 years, 81% men, 98% NYHA class II/III) were randomized to ET or control. The questionnaires completed were the Chronic HF Questionnaire (CHFQ), Minnesota Living with HF (MLHFQ), MOS Short Form-36 (SF-36), and Standard Gamble (SG). The effect size (ES), defined as the ratio of the mean difference between groups to the standard deviation, was calculated based on changes from baseline to 3 months. An ES of 0.3 is considered clinically significant. Spearman correlation coefficients were calculated between functional capacity (6 minute walk distance (6MWd), peak VO2, arm-curl, knee-extension and leg-press strength) and QOL score at baseline.
Results: The CHFQ Dyspnea and Fatigue scales were the most responsive to treatment, with a difference in ES of 0.316 and 0.153 between groups. The MLHFQ Total score had a small difference in ES of −0.144 (low score is healthier). The SF-36 Pain scale displayed the largest difference of treatment compared to all of the SF-36 scales, an ES of 0.226. The SG showed almost no ES. The CHFQ and MLHFQ weakly correlated with the 6MWd (ρ= 0.15– 0.26, p<0.05). The SF-36 Physical Function scale correlated strongest of all the questionnaires with the 6MWd (ρ= 0.342, p<0.01). None of the questionnaires significantly correlated with muscle strength or peak VO2.
Conclusions: The disease-specific questionnaires showed the greatest response to ET with the CHFQ being the most responsive. The SF-36, a generic measure of QOL, was less sensitive to change and the SG, a single-item utility index, showed no treatment effect. The CHFQ, based on these data, may be the best tool to measure the effect of ET on QOL in HF. It is not surprising that the relationship between QOL and functional capacity is weak as there are many factors beyond functional capacity that determine QOL.