Phenotypes can be utilised in the clinical management of disorders. Approaches to phenotype disorders have evolved from subjective expert opinion to data-driven methodologies. A previous cluster analysis among working-age subjects with cough revealed a phenotype TBQ (triggers, background disorders, quality-of-life impairment), which included 38% of the subjects with cough. The present study was carried out to validate this phenotyping among elderly, retired subjects with cough.

This was an observational cross-sectional study conducted via email among the members of the Finnish Pensioners’ Federation (n=26?205, 23.6% responded). The analysis included 1109 subjects with current cough (mean±sd age 72.9±5.3 years; 67.7% female). All filled in a comprehensive 86-item questionnaire including the Leicester Cough Questionnaire. Phenotypes were identified utilising k-means partitional clustering.

Two clusters were identified. Cluster A included 75.2% of the subjects and cluster B 24.8% of the subjects. The three most important variables to separate the clusters were the number of cough triggers (mean±sd 2.47±2.34 versus 7.08±3.16, respectively; p<0.001), Leicester Cough Questionnaire physical domain (5.38±0.68 versus 4.21±0.81, respectively; p<0.001) and the number of cough background disorders (0.82±0.78 versus 1.99±0.89 respectively; p<0.001).

The phenotype TBQ could be identified also among elderly, retired subjects with cough, thus validating the previous phenotyping among working-age subjects. The main underlying pathophysiological feature separating the phenotype TBQ from the common cough phenotype is probably hypersensitivity of the cough reflex arc.