Mendelian Randomization Study Suggests Narcolepsy Increases Risk of Heart Failure and Coronary Artery Disease
The findings further supported the idea that prevention and assessment of cardiovascular risk in patients with narcolepsy is critical to ongoing care.
Using a Mendelian randomization (MR) method, a recently published study identified a causal association between narcolepsy and increased risk of heart failure (HF) and coronary artery disease (CAD). Notably, there was no evidence to support a correlation between narcolepsy and increased risk of myocardial infarction (MI), any stroke (AS), and acute ischemic stroke (AIS).1
The MR study extracted 31 single nucleotide polymorphisms (SNPs) significantly associated with narcolepsy as instrumental variables (IVs) from a large-scale genome-wide association study (GWAS) encompassing 460,913 participants of European ancestry. Using inverse-variance weighting (IVW) with different effect models, results showed that narcolepsy is associated with an increased risk of HF and CAD, demonstrated by odds ratios (ORs) of 1.714 (95% CI, 1.031-2.849; P = .037) and 1.702 (95% CI, 1.011-2.864; P = .045).
In the MR analysis, senior author Xiangeng Zhang, Sichuan Nursing Vocational College, and colleagues used genetic variation as an IV and was organized around 3 core hypotheses: (1) selected genetic IVs directly influenced exposure, (2) genetic IVs are independent of any potential confounders, and (3) genetic IVs go to influence cardiovascular disease (CVD) only through narcolepsy and cannot influence outcomes through other pathways. The data was reported according to the STROBE-MR (Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization) guideline.
READ MORE: Evaluating Positive Phase 1 Trial Data for Narcolepsy Agent ALKS 2680
While the study revealed an association between narcolepsy, HF, and CAD, there was no indication of a causal relationship between genetic susceptibility to narcolepsy and the development of MI (OR, 1.296; 95% CI, 0.731-2.298; P = .374), AS (OR, 1.101; 95% CI, 0.669-1.812; P = .703), and AIS (OR, 1.352; 95% CI, 0.786-2.326; P = .274). "On this basis, the association between narcolepsy and CVDs also needs to be further explored at the mechanistic level," the study investigators wrote.
A sensitivity analysis was performed to detect heterogeneity and pleiotropy to ensure that the second and third hypotheses of the MR study were valid. Cochran’s Q was used first to assess the heterogeneity of the effect of SNPs associated with narcolepsy on CVD outcomes, followed by MR-Egger intercept method to test for horizontal pleiotropy. Third, the investigators performed a leave-one-out sensitivity analysis to determine whether the results would be affected by a single SNP. Lastly, funnel plots were generated to visualize the presence of pleiotropy.
In the sensitivity analysis, no outliers were detected by MR-PRESSO, which indicated robust and reliable results. Additionally, Cochrane’s Q test demonstrated heterogeneity in the results of the analysis of narcolepsy and HF (P <.05) and therefore IVW with random-effects modeling was used in the MR analysis of narcolepsy and HF to account for the heterogeneity in the results. All told, IVs associated with narcolepsy did not demonstrate horizontal pleiotropy in any type of CVD, according to results of the MR-Egger intercept test.
The study had strengths and limitations, but was helped by the fact that the 2-sample MR reduced the influence of confounding factors and reversed causality on the results. Although multiple methods were used to test the presence of pleiotropy, the investigators concluded to not fully rule out potential pleiotropy. Second, the prevalence of narcolepsy may have varied depending on the genetic background of patients, as well as the fact that individuals were of European ancestry, which may have limited the application of these results to other ethnic populations.
