Estimated lung age in healthy North African adults cannot be predicted using reference equations derived from other populations

Abstract

IntroductionInterpretation of “lung age” data relies upon comparison of the chronological lung age (CLA) with the estimated lung age (ELA) predicted from published reference equations [7–10]. AimTo test the applicability of the published reference equations in healthy non-smoker Tunisian aged 19–90years. Population and methodsPublished reference equations were applied to the spirometry results of 540 adults (364 women). Two methods of comparison were applied: (i) Determination, according each equation, of the percentages of subjects having a deltaLungAge (=ELA–CLA)>Upper-Limit-of-Normal (ULN). (ii) Bland and Altman comparison, for the same age range as in the corresponding study, between CLA and ELA. ResultsThe mean±SD (95% confidence interval) of the total sample CLA and height were 48.8±13.1 (47.7–49.9) years and 164±10 (163–165) cm. (i) The percentages of healthy subjects with a deltaLungAge>ULN varied from 1% (Newbury) to 64% (Hansen) in men, and from 20% (Yamaguchi) to 51% (Hansen, Morris and Temple) for women. (ii) Mean±SD ELA was significantly underestimated by 17±19years (Hansen), by 12±23years (Morris and Temple) and was significantly overestimated by 4±19years (Newbury). Mean±SD ELA from Yamaguchi et al. [10] was not statistically different from the CLA (1±14years). ConclusionThe published reference equations did not reliably predict CLA data in the Tunisian population. Awaiting the establishment of reliable equation proper to North African population, we recommend the use of the Yamaguchi et al.’s [10] reference equations.