Erik Sesbreno1,2,3, Denis P. Blondin4, Christine Dziedzic5, Jennifer Sygo6,7, François Haman8, Anne-Sophie Brazeau3 and Margo Mountjoy9, 10
1Institut National du Sport du Québec, Montreal, QC, Canada.
2French-speaking Research Network for Athlete Health Protection and Performance.
3McGill University, School of Human Nutrition, QC, Canada.
4Department of medicine, division of neurology, Faculty of Medicine and Health Sciences, Université de Sherbrooke, Centre de Recherche du Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, QC, Canada.
5Buffalo Sabres, Buffalo, NY, United States.
6Athletics Canada, Ottawa, Canada.
7Nottingham Trent University, Nottingham, United Kingdom.
8Faculty of Health Sciences, University of Ottawa, Ottawa, ON, Canada.
9Association for Summer Olympic International (ASOIF), Lausanne, Switzerland.
10Department of Family Medicine. Michael G. DeGroote School of Medicine, McMaster University. Hamilton Ontario Canada.

Athletes in low energy availability (LEA) are at increased risk of developing the Relative Energy Deficiency in Sport (RED-S) syndrome with undesirable consequences on health and performance. An objective approach to diagnose the condition is currently unavailable, and consequently, several assessment methods are used to detect different surrogate markers of LEA, such as low measure:predictive resting metabolic rate (RMRratio ) (<0.90). However, the RMRratio has not been frequently examined in larger male athletes at risk of LEA/RED-S.

The purpose of the investigation was to compare the impact of using different regression models on the point-prevalence of low RMRratio (<0.90) in male volleyball athletes. We hypothesized that using difference predictive RMR equations will impact the point-prevalence of low RMRratio.

Using a retrospective cross-sectional design, 22 male athletes from a national indoor volleyball program were assessed during the critical period of the in-season for indicators of LEA, including resting metabolic rate (RMR) testing, DXA assessment of bone mineral density and body composition, hematological analysis, surface anthropometry and restrained eating behaviour via three-factor eating questionnaire–R18. Several predictive RMR equations were examined such as the Cunningham1980, Cunningham1991, Mifflin1990, Johnstone2006, Roza1984, DeLorenzo1999, Harris-Benedict1918 and Tinsley2019.

The mean age, height, body mass, lean mass and fat mass of participants were 25.8 years, 197.2cm, 93.3kg, 77.5kg and 15.9kg, respectively. The main finding is that a significant difference in RMRratio (p<0.05) was observed between the Tinsley2019 and the Cunningham1991, Mifflin1990, Johnstone2006 and Harris-Benedict1918 predictive RMR equations. In some instances, the application of different predictive equations changed the point-prevalence of low RMRratio (<0.90) within the same study sample. The Cunningham1991, Mifflin1990, Johnstone2006, Harris-Benedict1918 revealed no cases while the Cunningham1980, Roza1984, DeLorenzo1999, Tinsley2019 identified 1, 4, 3 and 4 cases of low RMRratio. The sensitivity% of the predictive equations that identified cases of low RMRratio was less than 30%.

The selection of predictive RMR equation requires careful consideration when screening for signs of RED-S in larger athletes. The predictive equations under assessment were unsuitable for screening larger athletes presenting with other surrogate markers of LEA.

Low RMRratio should not be used a sole approach to screening larger athletes at risk of RED-S. Future work should explore methods to better identify larger athletes at risk.

Programme de Recherche, d’Innovation et de Diffusion de l’Information (PRIDI - 53).