Respiratory symptoms and mortality in four general population cohorts over 45 years

This study assessed the association between respiratory symptoms and mortality in four cohorts of the general population in Norway aged 15–75 years and in selected subgroups in the pooled sample. The study comprised 158,702 persons, who were drawn randomly from the Norwegian population register. All subjects received a standardized, self-administered questionnaire on 11 respiratory symptoms between 1972 and 1998, with follow-up of death until December 31, 2017. Analyses were performed on 114,380 respondents. The hazard of death was closely associated with sex, age, and education. The hazard ratios (HR) for death and the 95% confidence intervals (CI) by risk factors were similar in the four cohorts. After adjustment for demographic and environmental, modifiable factors, the HR for death was 1.90 (95% CI 1.80–2.00) for breathlessness score 3, 1.28 (1.21–1.37) for cough/phlegm score 5 and 1.09 (1.05–1.14) for attack of breathlessness/wheeze score 2 compared to the referent (no symptom), respectively. The cough/phlegm score was associated with death in current smokers but not in never smokers or ex-smokers. Breathlessness score was associated with death in men and women. Among persons aged 45–75 years, respiratory symptoms were significant predictors of all cause mortality. Education and smoking habits influenced only the associations between coughing and mortality. The associations were independent of study sites. •This study assessed the association of respiratory symptoms and mortality in four general population in Norway.•All subjects received a standardized, self-administered questionnaire on 11 respiratory symptoms between 1972 and 1998•The subjects had follow-up of death until 31 December 2017.•The hazard of death was closely associated with gender, age, education and smoking habits.•The hazard ratios (HR) for death and the 95 % confidence intervals (CI) by risk factors were similar in the four cohorts.•The study gives new information from stratified analyses and longer follow-up than previous studies.