Reliability and physiological responsiveness of skin temperature and thermal contrast index during incremental exercise: a test–retest analysis

Abstract To investigate the test–retest reliability of skin temperature (T sk ) and the thermal contrast index (TCI) during incremental exercise and examine their responsiveness to cardiovascular, metabolic, and perceptual parameters. Nineteen participants completed two standardized incremental cycling tests with concurrent thermographic, cardiovascular, metabolic, and perceptual measurements. T sk and TCI reliability was evaluated using coefficients of variation, intraclass correlation coefficients, Bland–Altman analysis, and mixed-effects models. T sk showed good to excellent reliability across all regions of interest (ICC A,1 : 0.74–0.94 [CI: 0.44–0.98], SEM A,1 : 0.28–0.53). TCI reliability varied by region and intensity, ranging from poor to good (ICC A,1 : 0.23–0.80 [CI: -0.27-0.92], SEM A,1 : 0.29–0.57), with a tendency toward higher reliability at greater exercise intensities and in the anterior thigh. Confidence intervals were generally wide, warranting cautious interpretation of the reliability estimates. TCI displayed consistent associations with heart rate (β = 0.46, p < 0.001), blood lactate (β = 0.17, p < 0.001), and energy expenditure (β = 0.14, p < 0.001). In contrast, T sk was influenced more by individual traits than by physiological changes. TCI showed a tendency toward good reliability at higher intensities and in performance-relevant muscle regions, but lower reliability at low intensities and in less active areas. TCI also demonstrates meaningful responsiveness to internal load, particularly heart rate, highlighting its potential as a sensitive indicator of physiological load. In contrast, T sk provides stable measurements but primarily reflects individual baseline characteristics rather than acute exercise responses.