We read with interest the recent article published by Masa et al in CHEST (July 2016) reporting that the prevalence of cardiovascular morbidity (CVM) in obesity hypoventilation syndrome (OHS) was lower in patients with the highest OSA severity. The authors suggested ischemic preconditioning as a potential protective mechanism explaining this unexpected inverse relationship between OSA severity and CVM. Through a logistic regression, they compared CVM between three oxygen desaturation index-based tertiles, but one could be surprised by the phenotypic contrast of the severest oxygen desaturation index subgroup. Indeed, these patients were younger (‒7 to ‒8 years) and more active physically (+12%-+14%, 6-meter walking distance), included fewer hypertensive (‒9% to ‒14%) and diabetic patients (‒10%), and were predominantly men (+19%-+25%) and more obese (BMI, +8%-+10%). Considering these phenotypic differences, the conclusions should be interpreted with caution and need to be further confirmed. With this perspective, a biomarker approach would be helpful for evaluating the cardiovascular risk. Among potential candidates, circulating high-sensitivity cardiac troponin T (hs-cTnT) would be of great interest, since it has been shown to be detectable with increased concentrations in more than 70% of patients with severe OSA and has proved to be an independent factor associating OSA severity and risk of heart failure. Accordingly, an increased level of hs-cTnI, the “rival” biomarker, has also been found to be independently associated with OSA severity. The initial studies did not find any relationship between circulating troponin concentrations and OSA severity because they used first-generation troponin assays, which were not sufficiently sensitive. Since that time, the “high-sensitivity” generations have appeared and are characterized by an optimal analytical precision, that is, a coefficient of variation ≤ 10% at the 99th percentile of troponin concentrations in a normal reference population, defined as 14 ng/L for hs-cTnT. Interestingly, the higher median hs-cTnT concentration was previously found to be 6 ng/L (interquartile range, 4-10) in patients with severe OSA, which is a level similar to that associated with incident coronary heart disease (hazard ratio [HR], about 1.3), mortality (HR, about 1.5), and hospitalization for heart failure (HR, about 2.2) in the general population (hs-cTnT: range, 6- 8 ng/L). Since it has never been evaluated in OHS but is related to OSA severity and is predictive of CVM, hs-cTnT concentrations could be relevantly assayed in the present study if a plasma/serum bank is available or should at least be introduced in further studies aiming at verifying these controversial results.