A Novel Method To Predict Carbohydrate And Energy Expenditure Using Measures Of Training Load

Background: Sports nutrition guidelines recommend carbohydrate (CHO) intake be individualized to the athlete and modulated according to changes in training load (TL). However, there are limited methods to assess CHO utilization during training sessions. Purpose: To determine if CHO utilization and energy expenditure (EE) can be accurately estimated using non-invasive measures of TL and easily quantified individual factors. Methods: Eleven cyclists (4 f; age 33 ± 10 y; V̇O2max 51 ± 8 mL.kg-1.min-1) performed a graded exercise test and four training sessions: 1) 90 min at 90% of power at the first ventilatory threshold (VT1), 2) 30 min at 90% VT1, 3) 15 min at 90% VT1, followed by two sets of 5 x 3 min intervals with 2 min recovery, and 4) 15 min at 90% VT1, followed by two sets of 10 x 30-s intervals with 30 s recovery. Sessions were quantified using seven TL metrics: three using heart rate, three using power, and one using perceived exertion. CHO use and EE were determined separately for aerobic (gas exchange) and anaerobic (net lactate accumulation, body mass, and O2 lactate equivalent method) energy systems and summed. Repeated-measures correlation was used to examine relationships between TL and both CHO utilization and EE. General estimating equations were used to model and predict CHO utilization and EE, using TL alongside measures of fitness, dietary intake, and sex. Results: Correlations between CHO utilization (g) and EE (kcal) during exercise and training load (TL), as well as measures of model fit (R2) and accuracy (RMSE) for multivariable prediction models are shown in Table 1. Conclusions: At the individual level, there is a strong linear relationship between TL and both CHO utilization and EE during cycling. When combined with other measures such as power and/or VO2 at the second ventilatory threshold, VO2max, and sex, EE and CHO utilization during exercise can be estimated accurately. - Training load Repeated-measures correlation (r [95% CI]) R2 RMSE vs. CHO vs. EE CHO EE CHO EE TSS 0.96 [0.91, 0.98] 0.92 [0.84, 0.96] 0.96 0.97 15.5 58.8 TWD-kJ 0.82 [0.67, 0.91] 0.95 [0.91, 0.98] 0.96 0.97 14.9 62.2 TSS-HR 0.84 [0.70, 0.92] 0.94 [0.89, 0.97] 0.92 0.94 22.5 87.7 Lucia TRIMP HR 0.86 [0.73, 0.93] 0.94 [0.87, 0.97] 0.91 0.94 24.2 90.8 Lucia TRIMP power 0.91 [0.83, 0.96] 0.95 [0.90, 0.97] 0.94 0.93 18.8 91.8 Edwards TRIMP HR 0.83 [0.68, 0.91] 0.93 [0.86, 0.96] 0.90 0.93 24.8 97.0 sRPE-TL 0.84 [0.7, 0.92] 0.77 [0.58, 0.88] 0.89 0.92 26.9 100.2