Worker heat stress prevention and work metabolism estimation: comparing two assessment methods of the heart rate thermal component.

The heart rate thermal component ( ΔHRT ) can increase with body heat accumulation and lead to work metabolism (WM) overestimation. We used two methods (VOGT and KAMP) to assess ΔHRT of 35 forest workers throughout their work shifts, then compared ΔHRT at work and at rest using limits of agreement (LoA). Next, for a subsample of 20 forest workers, we produced corrected WM estimates from ΔHRT and compared them to measured WM. Although both methods produced significantly different ΔHRT time-related profiles, they yielded comparable average thermal cardiac reactivity (VOGT: 24.8 bpm °C-1; KAMP: 24.5 bpm °C-1), average ΔHRT (LoA: 0.7 ± 11.2 bpm) and average WM estimates (LoA: 0.2 ± 3.4 ml O2 kg-1min-1 for VOGT, and 0.0 ± 5.4 ml O2 kg-1min-1 for KAMP). Both methods are suitable to assess heat stress through ΔHRT and improve WM estimation. Practitioner summary: We compared two methods for assessing the heart rate thermal component ( ΔHRT ), which is needed to produce a corrected HR profile for estimating work metabolism (WM). Both methods yielded similar ΔHRT estimates that allowed accurate estimations of heat stress and WM at the group level, but they were imprecise at the individual level. Abbreviations: AIC: akaike information criterion; bpm: beats per minute; CI: confidence intervals; CV: coefficient of variation in %; CV drift: cardiovascular drift; ΔHRT: the heart rate thermal component in bpm; ΔHRT: the heart rate thermal component in bpm; ΔΔHRT: variation in the heart rate thermal component in bpm; ΔTC: variation in core body temperature in °C; HR: heart rate in bpm; HRmax: maximal heart rate in bpm; Icl: cloting insulation in clo; KAMP: Kampmann et al. (2001) method to determe ΔHRT; LoA: Limits of Agreement; PMV-PPD: the Predicted Mean Vote and Predicted Percentage Dissatisfied; PHS: Predicted Heat Strain model; RCM: random coefficients model; SD: standard deviation; TC: core body temperature in °C; TCR: thermal cardiac reactivity in bpm °C-1; τΔHRT: rate of change in the heart rate thermal component in bpm min-1; τTC: rate of change in core body temperature in °C min-1; tα,n-1: Student's t statistic with level of confidence alpha and n-1 degrees of freedom; TWL: Thermal Work Limit model; V̇O2 : oxygen consumption in ml O2 kg-1 min-1; V̇O2 max: maximal oxygen consumption in ml O2 kg-1 min-1; VOGT: Vogt et al. (1973) method to determine ΔHRT; WBGT: Wet-Bulb Globe Temperature in °C; WM: work metabolism.

Adaptive neuro-fuzzy inference systems with k-fold cross-validation for energy expenditure predictions based on heart rate.

This paper presents a new model based on adaptive neuro-fuzzy inference systems (ANFIS) to predict oxygen consumption (V˙O2) from easily measured variables. The ANFIS prediction model consists of three ANFIS modules for estimating the Flex-HR parameters. Each module was developed based on clustering a training set of data samples relevant to that module and then the ANFIS prediction model was tested against a validation data set. Fifty-eight participants performed the Meyer and Flenghi step-test, during which heart rate (HR) and V˙O2 were measured. Results indicated no significant difference between observed and estimated Flex-HR parameters and between measured and estimated V˙O2 in the overall HR range, and separately in different HR ranges. The ANFIS prediction model (MAE = 3 ml kg(-1) min(-1)) demonstrated better performance than Rennie et al.'s (MAE = 7 ml kg(-1) min(-1)) and Keytel et al.'s (MAE = 6 ml kg(-1) min(-1)) models, and comparable performance with the standard Flex-HR method (MAE = 2.3 ml kg(-1) min(-1)) throughout the HR range. The ANFIS model thus provides practitioners with a practical, cost- and time-efficient method for V˙O2 estimation without the need for individual calibration.