Breathlessness, phlegm and mortality: 26 years of follow-up in healthy middle-aged Norwegian men

Objectives. It is well known that pulmonary function is associated with all-cause and cardiovascular (CV) death. Less is known about the association between respiratory symptoms and mortality and whether such an association is independent of physical fitness. In this study, we assessed the association of breathlessness and productive cough with CV and all-cause mortality over 26 years. Design. Prospective occupational cohort study. Setting and subjects. In 1972-75, 1999 apparently healthy men aged 40-59 years were recruited to the study from five companies in Oslo, Norway. At study entry clinical, physiological and biochemical parameters including respiratory symptoms, spirometry, and an objective assessment of physical fitness were measured in all subjects, of whom 1623 had acceptable spirometry. The data was analysed using Cox proportional hazards analysis, adjusting for age, lung function, physical fitness, and other possible confounders, with mortality until 2000. Results. After 26 years (range 25-27), 615 men (38%) had died, of whom 308 (50%) from CV deaths. In multivariable proportional hazards models, 'having phlegm winter mornings' [hazard ratio (HR) 1.30, P = 0.01], 'breathlessness when hurrying/walking uphill' (HR 1.43, P = 0.005) and combinations of the two symptoms remained significant predictors of all-cause mortality. None of six respiratory symptoms were significant predictors of CV mortality in multivariable models. Conclusions. Phlegm, breathlessness and combinations of them were associated with all-cause mortality, even after adjusting for physical fitness, known CV and other risk factors such as smoking, and lung function. The finding of an association also after adjustment for physical fitness is new. In contrast, none of the six respiratory symptoms individually or in combination were associated with CV mortality in multivariable analysis.