Affective valence predictors from real-world based short sprint interval training

Affective valence is typically positive at exercise intensities below the lactate threshold, yet more aversive responses occur at supra-threshold intensities. Nevertheless, the physiological and psychological predictors of affective valence during supramaximal intensities including short sprint interval training (sSIT) have not yet been elucidated. Seventeen (7 women/10 men) moderately active young adults (age ​= ​[28.2 ​± ​5.6] years; $$ O_2max [maximum oxygen consumption] ​= ​[52.9 ​± ​8.1] mL·kg^{- 1}·min^{- 1}; BMI [body mass index] ​= ​[24 ​± ​2] kg·m^{- 2}) completed four low-volume running sSIT sessions (10 ​× ​4 ​s efforts with 30 ​s of passive recovery). We recorded participants' heart rate (HR), root mean square of successive differences of normal RR intervals (RMSSD), heart rate recovery (HRR), ratings of perceived exertion (RPE), feeling scale (FS), intention and self-efficacy during, and after each session. Overall, no significant correlation (p ​> ​0.05) was found between FS and baseline clinical outcomes. No significant correlation (p ​> ​0.05) was detected between FS and any training parameter. No significant correlations were noted between FS and exercise task self-efficacy and intentions (p ​> ​0.05). The regression model was significant (F  ​= ​5.57; p ​= ​0.002) and only three variables significantly entered the generated model: ΔHRR_{end-120s end} (p ​= ​0.002; VIF ​= ​2.58; 40.8%), time ≥ 90% HR_peak (p ​= ​0.001; VIF ​= ​1.26; 31.6%), and RMSSD_end (p ​= ​0.025; VIF ​= ​2.23; 27.6%). These findings suggest that HR-based measures, particularly those related to in-task stress (time ≥ 90% HR_peak) and acute recovery (ΔHRR_{end-120s end}, and RMSSD_end), may predict affective valence during real-world sSIT.