Childhood predictors of lung function trajectories and future COPD risk: a prospective cohort study from the first to the sixth decade of life

Dinh S. Bui, Caroline J. Lodge, John A. Burgess, Adrian J. Lowe, Jennifer Perret, Minh Q. Bui, Gayan Bowatte, Lyle Gurrin, David P. Johns, Bruce R. Thompson, Garun S. Hamilton, Peter A. Frith, Alan L. James, Paul S. Thomas, Deborah Jarvis, Cecilie Svanes, Melissa Russell, Stephen C. Morrison, Iain Feather, Katrina J. AllenRichard Wood-Baker, John Hopper, Graham G. Giles, Michael J. Abramson, Eugene H. Walters, Melanie C. Matheson, Shyamali C. Dharmage

Research output: Contribution to journalArticleResearchpeer-review

386 Citations (Scopus)


Background: Lifetime lung function is related to quality of life and longevity. Over the lifespan, individuals follow different lung function trajectories. Identification of these trajectories, their determinants, and outcomes is important, but no study has done this beyond the fourth decade. Methods: We used six waves of the Tasmanian Longitudinal Health Study (TAHS) to model lung function trajectories measured at 7, 13, 18, 45, 50, and 53 years. We analysed pre-bronchodilator FEV1 z-scores at the six timepoints using group-based trajectory modelling to identify distinct subgroups of individuals whose measurements followed a similar pattern over time. We related the trajectories identified to childhood factors and risk of chronic obstructive pulmonary disease (COPD) using logistic regression, and estimated population-attributable fractions of COPD. Findings: Of the 8583 participants in the original cohort, 2438 had at least two waves of lung function data at age 7 years and 53 years and comprised the study population. We identified six trajectories: early below average, accelerated decline (97 [4%] participants); persistently low (136 [6%] participants); early low, accelerated growth, normal decline (196 [8%] participants); persistently high (293 [12%] participants); below average (772 [32%] participants); and average (944 [39%] participants). The three trajectories early below average, accelerated decline; persistently low; and below average had increased risk of COPD at age 53 years compared with the average group (early below average, accelerated decline: odds ratio 35·0, 95% CI 19·5–64·0; persistently low: 9·5, 4·5–20·6; and below average: 3·7, 1·9–6·9). Early-life predictors of the three trajectories included childhood asthma, bronchitis, pneumonia, allergic rhinitis, eczema, parental asthma, and maternal smoking. Personal smoking and active adult asthma increased the impact of maternal smoking and childhood asthma, respectively, on the early below average, accelerated decline trajectory. Interpretation: We identified six potential FEV1 trajectories, two of which were novel. Three trajectories contributed 75% of COPD burden and were associated with modifiable early-life exposures whose impact was aggravated by adult factors. We postulate that reducing maternal smoking, encouraging immunisation, and avoiding personal smoking, especially in those with smoking parents or low childhood lung function, might minimise COPD risk. Clinicians and patients with asthma should be made aware of the potential long-term implications of non-optimal asthma control for lung function trajectory throughout life, and the role and benefit of optimal asthma control on improving lung function should be investigated in future intervention trials. Funding: National Health and Medical Research Council of Australia; European Union's Horizon 2020; The University of Melbourne; Clifford Craig Medical Research Trust of Tasmania; The Victorian, Queensland & Tasmanian Asthma Foundations; The Royal Hobart Hospital; Helen MacPherson Smith Trust; and GlaxoSmithKline.

Original languageEnglish
Pages (from-to)535-544
Number of pages10
JournalThe Lancet Respiratory Medicine
Issue number7
Publication statusPublished - Jul 2018

Cite this