Avaliação de fatores clínicos e genéticos na apneia obstrutiva do sono: uma proposta de modelo preditivo

Abstract

Introduction: Obstructive sleep apnea syndrome is a common disorder associated with an increased risk of cardiovascular, metabolic diseases and mortality. Risk factors for OSAS have a known or suspected genetic basis and the discovery of genetic markers is extremely important for their prevention and personalized treatment. The aim of this study was to assess the correlation between genetic changes, phenotypes and clinical data that can contribute to the development of OSAS and propose a predictive model for OSAS. Methods: case-control study with two hundred and forty-two patients, one hundred and eighty-one cases and eighty-one controls included. Polysomnographic, clinical, and genotypic data for the PER3 genes (rs228697, rs228727, rs228729 and rs10462020), BDNF (rs6265), NRXN3 (rs10146997), APOE (rs7412 and rs429358), HCRTR2 (rs2653349), MC4R (rs17782313) were analyzed (rs7) statistic R (version 3.6.3). Results: Two hundred and forty-two patients, 161 cases and 81 controls were included in the research. The average age of the participants was 51.48 ± 14.29, with a male predominance (57.02%) and the average BMI was 26.63 ± 4.36. The analysis of association between genotypes and OSAS did not show any association between the PER3 (rs228697, rs228727 and rs10462020) polymorphisms, NRXN3 (rs10146997), APOE (rs7412 and rs429358), HCRTR2 (rs2653349) and MC4R (rs17782313) with SA However, the BDNF (rs6265) and PER3 (rs228729) polymorphisms showed a significant association with OSAS (p = 0.01317 and p = 0.000094, respectively). However, after adjusting for age and BMI, BDNF (rs6265) and PER3 (rs228729) did not continue to show a significant association (p = 0.06201 and p = 0.0659, respectively). A proposed algorithm showed an adequate ability to predict OSAS with good sensitivity and specificity. For values of increasing cutoff points from 0 to 1, we obtained ROC curve whose low area was 79.4% (CI 95% 72.2% - 86.6%). The best cutoff value was 0.763 with sensitivity of 71.6% and specificity of 76.5%. Conclusion: The algorithm generated in this study, which includes data on sociodemographic, clinical, and genetic characteristics of patients, can be applied to clinical practice and will allow risk stratification for OSAS. The applicability of this algorithm must be proven in studies with a larger number of patients and in different ethnicities.