Effects of cigarette smoking on cardiovascular-related protein profiles in two community-based cohort studies

Another interrogation of the PIVUS/ULSAM cohorts, this time looking at the effects of smoking on CVD-related markers (note: not focused on disease outcome, but more of an epidemiological study on how an environmental factor affects markers with mechanistic links to a disease process). The basis for the study was that while CVD accounts for the largest proportion of smoking-related deaths (and smoking may account for >10% of all CVD deaths worldwide), most of the proteomic studies on the effects of smoking have been on patients with lung diseases (COPD, lung cancer etc). Here they examined 969 subjects from PIVUS (50% male/female) for the discovery phase, and 717 from ULSAM (all male) for validation. Smoking status (current, former or never-smoker) was established by self-reporting via a questionnaire and in the case of PIVUS subjects, information on the level of smoking (packs/day) was also obtained. Of the 80 CVD I markers analyzed, 30 proteins showed significant association with current smoking status in the PIVUS samples, and 10 of those were validated in ULSAM (p>0.05): MMP-12, GDF-15, uPAR, TRAIL-R2, LOX-1, HGF, MMP-10 and MMP-1 were positively associated, while ESM-1 & IL-27A were negatively associated. Following this primary analysis, the 10 proteins identified were further characterized by a series of secondary analyses. Comparing previous to never- smokers, none of the proteins showed significant association with previous smoking, which may be consistent with previous data on the CVD risk reduction observed following cessation of smoking. Only GDF-15 showed any significant association with the level of smoking after multivariate adjustment for covariates. All 10 proteins showed very similar association with current smoking status between male and female subjects. Finally, they could show that all 10 markers retained their significance of association with current smoking even when all individuals with manifest CVD were excluded from the analysis. In the discussion, the authors point to other evidence that lends biological support for their findings: 5 of the 10 proteins they identified have previously been assoicated with smoking patterns in population-based and/or in vitro studies, and several have been suggested to be predictive for future CVD in asymptomatic subjects, or for clinical outcome in CVD patients. They suggest that the identity of the 10 markers provides pathophysiological indications that smoking may play an important role in long-term subclinical atherosclerosis. In Figure 1 of the paper, they assign each of the 10 proteins to one of four key mechanisms linked to the development and progression of atherosclerosis: impaired endothelial cell function, inflammation, neointimal formation and foam cell formation.