Author(s): ROZIER, P., MOREL, B., Institution: UNIVERSITÉ SAVOIE MONT BLANC, Country: FRANCE, Abstract-ID: 2324

INTRODUCTION:
Critical power is an important fatigue threshold in exercise physiology [1]. It delimits efforts in the severe domain, where acute fatigue quickly and drastically appears, from efforts where disturbances remain moderate and can be maintained over long periods. It has great potential applications like evaluating and proposing training programs to improve quality of life of individuals with chronic diseases. However, gold-standard methods that evaluate critical intensity require maximal intensity and/or exhaustion tests, which are extremely difficult to use in a clinical context. In situ methods have been proposed to determine the critical intensity in cycling and running [2]. Training and racing data are compiled over months so the speed-duration relationship can be modelled. Dynamic Body Acceleration (DBA) is an indicator derived from ecological sciences that uses body acceleration as a proxy for locomotion speed [3]. Therefore, this study aimed to assess the validity of critical DBA (DBAc) derived from field data as a reliable proxy for anaerobic threshold velocity.
METHODS:
Twelve participants with various physical training and disease conditions wore a 3D accelerometer (GT3X-BT, ActiGraph) for one month. The only instruction was to “continue living as usual”. They then performed a submaximal incremental walk test on a 15% inclined treadmill. Blood lactate concentration [La-] was measured at the end of each 3-min stage. The stage speed increment was 0.5 km.h-1 when [La-] < 2 mmol.L-1, and 0.2 km.h-1 when [La-] > 2 mmol.L-1. Exercise was stopped as soon as [La-] > 4 mmol.L-1. An exponential model was fitted to the lactate concentrations as a function of speed to estimate the speed at 4 mmol.L-1 (S4mmol). The mean DBA records for each day were computed for durations ranging from 1 to 30 min. These records were fitted with a 2-parameter critical intensity model using quantile regression (τ = 0.9) to obtain DBAc. The relationship between DBAc and S4mmol was calculated by linear regression analysis (least squares method).
RESULTS:
The median values of DBAc and S4mmol were respectively 1.14 m.s-2 (range = [0.43 ; 1.40]) and 5.7 km.h-1 (range = [3.3 ; 6.6]). A very strong correlation was found between DBAc and S4mmol (r2 = 0.91, RMSE = 7%, p < 0.001).
CONCLUSION:
The DBAc obtained from the accelerometry data is a valid proxy for S4mmol. It is therefore possible to characterise the anaerobic threshold of individuals solely on the basis of accelerometry data obtained without any particular instructions during real life. This opens up interesting perspectives for evaluating patients without the need for time-consuming and physically demanding tests. Clinical studies including patient follow-up via actimetry are common and can use the proposed methodology to evaluate patients without additional constraints.

[1] Poole et al. (2016). Med Sci Sports Ex.
[2] Smyth et al. (2020). MSSE.
[3] Wilson et al. (2006). Journal of Animal Ecology.