Simple linear regression analysis was used for studying correlations between LAA% and mean CT score. Mean CT score was linearly correlated with LAA% (R2 = 0.97, p < 0.001), with a simple linear regression model as follows: mean CT score (HU) = — 1.69 X LAA% — 842. Patients with mean CT score < — 940 HU were defined as having severe emphysema based on this model. Again, severely emphysematous patients with mean CT score < — 940 HU were more frequently seen in the Gc*1F( + ) group than in the Gc*1F( —) group (26 patients vs 0 patients, p < 0.0001; Fig 3). Additionally, sCLA was larger in Gc*1F( + ) patients than Gc*1F( —) patients (35.9 ± 26.2 pixels vs 22.7 ± 12.1 pixels, p = 0.004; Fig 4). There was no significant difference in total number of CLAs between the two groups (5,005 ± 2,117 vs 4,267 ± 2,135, p = 0.25). birth control pills
This study showed that Gc*1F homozygosity was significantly associated with COPD in a cohort of Japanese smokers, and that this genotype was less frequent in healthy smokers. Previous studies on Gc-globulin polymorphism have shown varied results in association of this polymorphism and the risk of COPD. Kueppers et al showed that the frequency of homozygous Gc-2 phenotype was only 0.01 in patients with COPD compared with 0.05 in control subjects, but the study performed by Kauff-mann et al failed to confirm this result. Subsequently, Horne et al showed that the frequency of homozygous Gc-1F phenotype in patients with COPD was 0.06, and was greater than that in control subjects (0.01; relative risk, 4.8). They also found that the phenotypes containing the Gc-2 (2-1F, 2-1S, and 2-2) had a protective effect. More recently, Schellenberg et al again replicated the result showing decreased frequency of homozygous Gc*2 genotype in patients with COPD (0.03) compared to control subjects (0.14). However, they could not verify the results shown by Horne et al that Gc-1F homozygotes had an increased risk for COPD.