Non-exercise estimation of peak oxygen uptake in patients with ischaemic heart disease and heart failure using seismocardiography

Knowledge of cardiorespiratory fitness (i.e. VO₂peak) is important for determining prognosis and prescribing exercise for patients with heart disease undergoing cardiac rehabilitation (CR) programmes. In this explorative study, we investigated the accuracy of a novel equation using seismocardiography (SCG) at rest for the estimation of VO₂peak (SCG eVO₂peak) and whether it could detect changes following CR. An interim data analysis was planned after 50% of patients had undergone testing, allowing for adjustment of the SCG eVO₂peak. We compared the SCG eVO₂peak with a cardiopulmonary exercise test (CPET).

We included 125 patients with new-onset ischaemic heart disease (IHD, n = 58) or heart failure with reduced left ventricular ejection fraction (HFrEF, n = 67) from an outpatient CR clinic. Testing included SCG eVO₂peak and CPET measurement. The adjusted SCG 4.7_HD was validated in 30 and 34 patients with IHD and HFrEF, respectively. Forty-four out of the 67 patients with HFrEF were tested after completing 12 weeks of CR. A mean absolute percentage error (MAPE) ≤10% was decided for clinical relevance. The SCG 4.7_HD overestimated VO₂peak (2.1 mL min⁻¹ kg⁻¹, P = 0.003) with 95% limits of agreement ranging ±10.9 mL min⁻¹ kg⁻¹ when compared with CPET. The standard error of estimation was 6.0 mL min⁻¹ kg⁻¹, and MAPE was 29.1%. No correlation was observed between delta SCG 4.7_HD and CPET after CR for patients with HFrEF.

The SCG eVO2peak is not supported for clinical purposes in patients with IHD or HFrEF based on a poor-to-moderate agreement with large estimation errors and the inability to detect changes following CR.

The study is registered at ClinicalTrials.gov (NCT05520307).