We compared changes in performance and total haemoglobin mass (tHb) of elite swimmers in the weeks following either Classic or Live High:Train Low (LHTL) altitude training. Twenty-six elite swimmers (15 male, 11 female, 21.4 ± 2.7 years; mean ± SD) were divided into two groups for 3 weeks of either Classic or LHTL altitude training. Swimming performances over 100 or 200 m were assessed before altitude, then 1, 7, 14 and 28 days after returning to sea-level. Total haemoglobin mass was measured twice before altitude, then 1 and 14 days after return to sea-level. Changes in swimming performance in the first week after Classic and LHTL were compared against those of Race Control (n = 11), a group of elite swimmers who did not complete altitude training. In addition, a season-long comparison of swimming performance between altitude and non-altitude groups was undertaken to compare the progression of performances over the course of a competitive season. Regardless of altitude training modality, swimming performances were substantially slower 1 day (Classic 1.4 ± 1.3% and LHTL 1.6 ± 1.6%; mean ± 90% confidence limits) and 7 days (0.9 ± 1.0% and 1.9 ± 1.1%) after altitude compared to Race Control. In both groups, performances 14 and 28 days after altitude were not different from pre-altitude. The season-long comparison indicated that no clear advantage was obtained by swimmers who completed altitude training. Both Classic and LHTL elicited ~4% increases in tHb. Although altitude training induced erythropoeisis, this physiological adaptation did not transfer directly into improved competitive performance in elite swimmers.

The use of rolling averages to analyse training data has been debated recently. We evaluated two training load quantification methods (five-zone, seven-zone) fitted to performances over two race distances (50 and 100 m) using four separate longitudinal models (Banister, Busso. rolling averages and exponentially weighted rolling averages) for three swimmers ranked in the top 8 in the world. A total of 1610 daily load measures and 108 performances were collected. Banister (standard error of the estimate (SEE) 0.64 and 0.62 s; five-zone and seven-zone quantification methods), Busso (SEE 0.73 and 0.70 s) and exponentially weighted rolling averages (SEE 0.57 and 0.63 s) models fitted more accurately (p < 0.001) than the rolling averages approach (SEE 1.32 and 1.36 s). The seven-zone quantification method did not produce more accurate performance predictions than the five-zone method, despite being a more detailed form of training load quantification. Four neural network models were fitted and had a lower error (SEE 0.38, 0.41, 0.35 and 0.60 s) than all longitudinal models, but did not track as predictably over time. Exponentially weighted impulse-response models and exponentially weighted rolling averages appear more effective at predicting performance using training load data in elite swimmers than a rolling averages approach.