Impact of non-linear smoking effects on the identification of gene-by-smoking interactions in COPD genetics studies

P.J. Castaldi*, D.L. Demeo, C.P. Hersh, D.A. Lomas, I.C. Soerheim, A. Gulsvik, P. Bakke, S. Rennard, P. Pare, J. Vestbo, E.K. Silverman, E.F.M. ICGN Investigators incl. Wouters

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


BACKGROUND: The identification of gene-by-environment interactions is important for understanding the genetic basis of chronic obstructive pulmonary disease (COPD). Many COPD genetic association analyses assume a linear relationship between pack-years of smoking exposure and forced expiratory volume in 1 s (FEV(1)); however, this assumption has not been evaluated empirically in cohorts with a wide spectrum of COPD severity. METHODS: The relationship between FEV(1) and pack-years of smoking exposure was examined in four large cohorts assembled for the purpose of identifying genetic associations with COPD. Using data from the Alpha-1 Antitrypsin Genetic Modifiers Study, the accuracy and power of two different approaches to model smoking were compared by performing a simulation study of a genetic variant with a range of gene-by-smoking interaction effects. RESULTS: Non-linear relationships between smoking and FEV(1) were identified in the four cohorts. It was found that, in most situations where the relationship between pack-years and FEV(1) is non-linear, a piecewise linear approach to model smoking and gene-by-smoking interactions is preferable to the commonly used total pack-years approach. The piecewise linear approach was applied to a genetic association analysis of the PI*Z allele in the Norway Case-Control cohort and a potential PI*Z-by-smoking interaction was identified (p=0.03 for FEV(1) analysis, p=0.01 for COPD susceptibility analysis). CONCLUSION: In study samples of subjects with a wide range of COPD severity, a non-linear relationship between pack-years of smoking and FEV(1) is likely. In this setting, approaches that account for this non-linearity can be more powerful and less biased than the more common approach of using total pack-years to model the smoking effect.
Original languageEnglish
Pages (from-to)903-909
Number of pages7
Issue number10
Publication statusPublished - Oct 2011



Cite this