Smoking, respiratory symptoms, lung function and life expectancy: A longitudinal study of ageing.

BACKGROUND AND OBJECTIVE

Prognostic indices have been developed to predict various outcomes, including mortality. These indices and hazard ratios may be difficult for patients to understand. We investigated the association between smoking, respiratory symptoms and lung function with remaining life expectancy (LE) in older adults.

METHODS

Data were from the 2004/05 English Longitudinal Study of Ageing (ELSA) (n = 8930), participants aged ≥50-years, with mortality data until 2012. Respiratory symptoms included were chronic phlegm and shortness of breath (SOB). The association between smoking, respiratory symptoms and FEV/FVC, and remaining LE was estimated using a parametric survival function and adjusted for covariates including age at baseline and sex.

RESULTS

The extent to which symptoms and FEV/FVC predicted differences in remaining LE varied by smoking. Compared to asymptomatic never smokers with normal lung function (the reference group), in never smokers, only those with SOB had a significant reduction in remaining LE. In former and current smokers, those with respiratory symptoms had significantly lower remaining LE compared to the reference group if they had FEV/FVC <0.70 compared to those with FEV/FVC ≥0.70. Males aged 50-years, current smokers with SOB and FEV/FVC <0.70, had a remaining LE of 19.2 (95%CI: 16.5-22.2) years, a decrease of 8.1 (5.3-10.8) years, compared to the reference group.

CONCLUSION

Smoking, respiratory symptoms and FEV/FVC are strongly associated with remaining LE in older people. The use of remaining LE to communicate mortality risk to patients needs further investigation.