Impact of menstrual cycle or combined oral contraception on elite female cyclists' training responses through a clustering analysis of training sessions.

Objectives: (i) To classify training sessions of elite female cyclists according to an intensity index based on a longitudinal follow-up using multiparametric data collected (ii) to measure the effect of estimated menstrual cycle (MC) phases and oral contraceptive pills (OC) phases on the athletes' training responses on each type of training identified.

Method: Thirteen elite French cyclists were followed up over 30 months and 5,190 training sessions were collected and 81 MC/OCs full cycles analyzed. Power sensors and position devices captured training data , which was summarized into 14 external load variables. Principal Component Analysis and K-means clustering were used to identify cycling sessions according to an intensity load index. The clusters were then verified and categorized through the analysis of heart rate and rate of perceived effort. A calendar method was used to estimate 3 phases of the MC: menstruation, mid-cycle phase (MP) and late-cycle phase (LP). Two phases were defined among monophasic OC users: pills' taking/withdrawal.

Results: Four main types of training effort were identified: Intensive, Long, Medium and Light. In the MC group ( = 7; 52 cycles), the intensity index is 8% higher during the mid-cycle (vs. menstrual phase,  = 0.032) in the Intensive effort sessions. No differences were observed in Long, Medium or Light effort, nor between the phases of pills' taking/withdrawal among OC users.

Conclusion: The clustering analyses developed allows a training classification and a robust method to investigate the influence of the MC/OC . A better training response during the mid-cycle when the sessions are the most intense suggest an impact of the MC when the athletes approach their maximal capacity.