Abstract Body: Introduction: A healthy sleep pattern (HSP) has been related to lower risks of cardiovascular diseases. However, sleep behaviors are usually self-reported, which introduces bias. While metabolomics provides an objective method and complements traditional biomarkers, the underlying metabolomic mechanisms linking a comprehensive sleep pattern and subclinical cardiovascular risk factors are not well known.
Hypothesis: We hypothesize that an HSP is associated with metabolites, which may subsequently influence subclinical cardiovascular risk factors.
Methods: A total of 1032 participants from the Bogalusa Heart Study who underwent untargeted metabolomics profiling and had existing sleep data were included in the current analysis. HSP was estimated using information on sleep duration, chronotype, insomnia, snoring, and daytime sleepiness. Elastic net regularized regression was applied to generate an HSP-related metabolomic signature. Subclinical cardiovascular risk factors were measured by echocardiography and ultrasonography, including left ventricular mass index (LVMI), relative wall thickness (RWT), left ventricular geometry, pulse wave velocity (PWV), and carotid intima-media thickness (cIMT).
Results: The HSP was inversely associated with subclinical cardiovascular risk factors [LVMI (β: -0.88, SE: 0.29; p=0.002), RWT (β: -0.004, SE: 0.002; p=0.03), IMT (β: -0.02, SE: 0.01; p=0.018), and concentric LV hypertrophy (OR: 0.72, 95% CI: 0.59 – 0.88; p=0.001)]. Using elastic net regularized regression, we identified a metabolomic signature (Figure), comprised of 44 metabolites, robustly associated with the HSP (r=0.22; p<0.001). In multivariable regression models, the metabolomic signature showed significant inverse associations with LVMI (β: -2.5, SE: 0.77; p=0.001), RWT (β: -0.016, SE: 0.006; p=0.004), IMT (β: -0.09, SE: 0.02; p<0.001), and lower odds of concentric remodeling (OR: 0.52, 95% CI: 0.33 – 0.82; p=0.005), eccentric LV hypertrophy (OR: 0.26, 95% CI: 0.09 – 0.69; p=0.007), and concentric LV hypertrophy (OR: 0.38, 95% CI: 0.19 – 0.76; p=0.007) (Figure, Model 2). The associations persisted after further adjustment for the HSP (Figure, Model 3).
Conclusion: We identified a metabolomic signature that robustly reflected the HSP, and was associated with lower subclinical cardiovascular risk factors, independent of traditional risk factors and phenotypic sleep measurements.