Objective
To determine the association between dyspnea at entry into the PAQUID cohort and 13-year mortality, taking into consideration BMI and other mortality-related factors
Methods: Design
Longitudinal study
Setting
In Dordogne and Gironde, South Western France
Subjects
A total of 3646 French community dwellers aged 65 years old and over from the PAQUID study were included. Main outcome measures: dyspnea measured on 5-grades scale, mortality measured over 13 years of follow-up. Adjustment variables: age, gender, BMI (kg/m2), antecedent of ischemic heart disease, antecedent of stroke, hypertension, smoking history and diabetes
Results
The study sample included 3646 subjects out of whom 54.11% died at 13 years of follow-up, 57.3% of participants were women and mean age was 75.3 (SD 6.8) years. Univariate analysis showed that dyspnea was associated with 13-year mortality. Death occurred in 45.6% of non-dyspneic subjects, 51.8% in those with level 1 of dyspnea, 65.6% in level 2 and 80.6% in level 3 and 4 (P<10−4). The median survival was at 13.26 (SD 0.20) years for level-0 of dyspnea, 12.33 (SD 0.31) years for level-1 of dyspnea, 9.28 (SD 0.44) years for level-2 and 6.43 (SD 0.45) years for level-3 and 4 (P=10−3). In the multivariate analysis, the risk of mortality for level1 compared to level-0 was HR=1.13 (CI95%=[1.01–1.26]); this risk increases to HR=1.42 (CI95%=[1.25–1.63]) for level-2 and to HR=1.90 (CI95%=[1.61–2.25]) for level-3 and 4ConclusionThese findings suggest that the relationship between long-term mortality and dyspnea is strong, consistent and independent of other covariates in the elderly