Assessment of the MOMENTUM sport earbud for multi-parametric physiological tracking and exploratory lactate modeling during incremental exercise
Continuous monitoring of heart rate (HR) and core temperature ( $$T_c$$ ) is essential for optimizing athletic performance and preventing heat-related illness. However, traditional monitoring methods like chest straps and wrist-worn trackers often face accuracy challenges during exercise due to sweat, motion, and signal interference. Furthermore, few devices offer validated, synchronized HR and $$T_c$$ monitoring during physical exertion. This study validates the MOMENTUM Sport (MSPORT) earbud, a novel multisensor wearable for synchronized HR and $$T_c$$ tracking and investigates its utility in non-invasive blood lactate prediction. Thirty-six healthy adults (19 males, 17 females) performed an incremental cycling ramp test under controlled laboratory conditions. MSPORT HR and $$T_c$$ were validated against Polar H10 and Braun ThermoScan 7 references. Predictive lactate modeling was implemented via a Leave-One-Subject-Out (LOSO) framework, with physiological significance assessed through Linear Mixed Models (LMM) and temporal coupling via cross-correlation analysis. MSPORT showed good inter-device agreement for HR (MAE: 0.89 bpm; RMSE: 1.10 bpm) compared to the Polar H10, and good ear-canal temperature agreement with the Braun ThermoScan 7 (MAE: 0.35 $$^{\circ }$$ C; MAPE: 0.92%). The lactate prediction model achieved a correlation of $$R = 0.84$$ ( $$p < 0.05$$ ), with LMM confirming $$T_c$$ as a highly significant independent predictor ( $$\beta = 11.63$$ , SE = 1.00, $$z = 11.65$$ , $$p < 0.001$$ , 95% CI [9.68, 13.59]), while HR also contributed significantly ( $$\beta = -1.94$$ , SE = 0.29, $$z = -6.78$$ , $$p < 0.001$$ ). The negative HR coefficient likely reflects collinearity or a suppression effect: once cumulative $$\Delta T_c$$ accounts for the metabolic trend, HR becomes a corrective term. Sex was a significant covariate ( $$\beta = -1.25$$ , $$z = -2.00$$ , $$p = 0.046$$ ). The random intercept variance was 1.909, indicating substantial between-participant variability in baseline lactate (model converged; REML). An exploratory cross-correlation analysis suggested a preliminary temporal association between lactate and ear-canal temperature; the precise lag magnitude cannot be reliably estimated from three-minute lactate sampling intervals. Absolute lactate magnitude showed significant sex-based differences ( $$p < 0.05$$ ). The MSPORT earbud provides reliable HR tracking and good ear-canal temperature agreement with a co-located tympanic reference during laboratory cycling. The exploratory lactate model demonstrates group-level association ( $$R = 0.84$$ ) and supports the feasibility of thermal-cardiac signal fusion for metabolic monitoring, though individual-level accuracy remains insufficient for clinical use. Further research is required to assess performance across diverse exercise protocols, environments, and populations.
Authors
- Incinur Zellhuber (ORCID: https://orcid.org/0009-0004-1010-2155)
- Martin Schönfelder (ORCID: https://orcid.org/0000-0002-3003-3378)
- Camargo David
- Michael Weber
Institutions
- Technical University of Munich (DE)
Publication Details
- Journal
- Scientific Reports
- Published
- 2026-11-20
- DOI
- https://doi.org/10.1038/s41598-026-80803-8
- Primary Topic
- Non-Invasive Vital Sign Monitoring
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Technische Universität München