Exploring diagnostic and prognostic value of digital biomarkers of blood pressure variability in metabolic syndrome

Abstract Introduction Blood pressure variability (BPV) has been recognised for over two decades as being associated with the severity of end-organ damage, including stroke, cardiac injury, and chronic kidney disease [1]. Physiologically, BPV can be both a consequence and a cause of arterial stiffness and vascular dysfunction. Moreover, BPV has been identified as a prognostic factor linked to a reduced risk of cardiovascular disease. Both lifestyle modifications and pharmacological interventions have been proposed to improve BPV and mitigate its impact on cardiovascular outcomes. Recent developments in continuous blood pressure (BP) monitoring offer opportunities to increase feasibility of digital biomarkers of BPV. However, the diagnostic and prognostic value of digital biomarkers of BPV in metabolic syndrome – a major cardiovascular risk factor affecting 12.5% to 31.4% of the global population [2] – remains limited. Purpose This study aims to evaluate the additional diagnostic and prognostic value of digital BPV biomarkers, beyond average BP, across populations exhibiting increasing metabolic syndrome characteristics. Methods Twenty-four-hour cuffless BP data were retrospectively analysed from the AURORA-BP dataset (N = 1125; age range 21–85 years; 49.2% female) [3]. Participants were stratified into groups based on the number of metabolic syndrome proxies: none, obesity, history of diabetes, history of hypertension, and all combinations thereof. BP and BPV biomarkers were derived from 24-hour cuffless BP readings. Univariate and multivariate correlation analyses were used to assess the associations among BP and BPV biomarkers. Linear regression models, adjusted for age, gender and average BP, were used to evaluate the ability of BPV biomarkers to differentiate between the metabolic syndrome groups. Results Most BPV biomarkers, including standard deviation and average real variability, were only weakly associated with average BP, indicating they reflect distinct information. Most prominently, night-time average SBP and DBP were associated with coefficient of variation (ρ = -0.39 and -0.56, p < 0.001) and night-time dipping (ρ = -0.58 and -0.65, p < 0.001), respectively. Additionally, multivariate redundancy analysis showed that night-time dipping accounted for additional variance in individuals with obesity and diabetes, not explained by average BP alone (Fig. 1). Finally, linear regression indicated an 8.8% reduction in night-time dipping in the group with all metabolic syndrome symptoms (obesity, hypertension, diabetes) compared to the group without symptoms, independent of age, gender and average BP (Fig. 2). Conclusion Blood pressure variability appears to be an independent factor associated with metabolic syndrome symptoms and may offer additional diagnostic and prognostic value in the assessment and management of cardiovascular risk in this population. Further prospective longitudinal cohorts are needed to confirm these observations.Figure 1.Redundancy analysis plot  Figure 2.BPV group differences