Built on Science
Every algorithm, threshold, and formula in MyTrainerOS is backed by peer-reviewed sports science. Here are the studies behind our features.
Nutrition Intelligence
Adaptive TDEE, macro calculation, and nutritional periodization
Mifflin MD, St Jeor ST, et al. (1990)
A new predictive equation for resting energy expenditure in healthy individuals
Mifflin-St Jeor is the most accurate BMR estimation equation for non-obese populations (±10% error).
Morton RW, Murphy KT, et al. (2018)
A systematic review, meta-analysis and meta-regression of protein supplementation
Protein intake of 1.6 g/kg/day maximizes resistance-training-induced gains in muscle mass and strength.
Hector AJ, Phillips SM (2018)
Protein recommendations for weight loss in elite athletes
During caloric restriction, protein needs increase to 2.3–3.1 g/kg FFM/day to preserve lean mass.
Wilding JPH, Batterham RL, et al. (2021)
Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP-1)
STEP-1 trial reported ~40% of GLP-1-driven weight loss is lean mass — directly motivates our resistance-training and protein-target algorithms for members on GLP-1 medications.
Recovery Intelligence
HRV monitoring, readiness scoring, and recovery protocols
Plews DJ, Laursen PB, et al. (2012)
Training adaptation and heart rate variability in elite endurance athletes
lnRMSSD coefficient of variation >10% indicates maladaptation; daily HRV monitoring guides training load.
Plews DJ, Laursen PB, et al. (2013)
Monitoring training with heart-rate variability: How much compliance is needed for valid assessment?
7-day rolling average of lnRMSSD provides reliable baseline; morning measurements preferred.
Buchheit M (2014)
Monitoring training status with HR measures: do all roads lead to Rome?
Establishes 7-day rolling baseline as the methodological standard for HRV-guided training decisions; CV stability check separates real training response from measurement noise.
Saw AE, Main LC, Gastin PB (2016)
Monitoring the athlete training response: subjective self-reported measures trump commonly used objective measures
Subjective wellness questionnaires outperform objective markers (HRV, biochem, salivary cortisol) for detecting training response — informs the 10% subjective weight in our composite recovery score.
Flatt AA, Esco MR (2015)
Heart rate variability stabilization in athletes
HRV requires 7+ measurement days to stabilize; recommends rolling-window CV for assessing whether the athlete has reached a stable baseline.
Achten J, Jeukendrup AE (2003)
Heart rate monitoring: applications and limitations
Resting heart rate trends reflect parasympathetic activity and recovery state; foundational for the 10% RHR weight in the composite readiness score.
Fullagar HHK, Skorski S, et al. (2015)
Sleep and athletic performance: the effects of sleep loss on exercise performance, and physiological and cognitive responses to exercise
Sleep loss measurably impairs reaction time, cognitive function, and recovery — supports 25% sleep-quality weight in the composite recovery score.
Watson AM (2017)
Sleep and athletic performance
7–9 hours of sleep is the evidence-based optimal range for athletes; chronic restriction below 7h impairs strength and endurance markers.
Walker M (2017)
Why We Sleep: Unlocking the Power of Sleep and Dreams
Sleep <6 hours reduces testosterone, growth hormone, and athletic recovery. 7–9 hours recommended.
Injury Prevention
ACWR monitoring, volume spike detection, and risk prediction
Leckey S, et al. (2025)
Machine learning approaches for sports injury prediction: A scoping review
XGBoost is the top-performing ML model for injury prediction across multiple studies.
Gabbett TJ (2016)
The training—injury prevention paradox
ACWR sweet spot of 0.8–1.3 minimizes injury risk. Spikes >1.5 increase injury odds by 2–4x.
Foster C (1998)
Monitoring training in athletes with reference to overtraining syndrome
Training monotony (mean/SD of daily load) >2.0 with high strain predicts illness and overtraining.
Program Design
E1RM tracking, progressive overload, and plateau detection
Epley B (1985)
Poundage chart
E1RM = Weight × (1 + Reps/30). Standard formula for estimating 1-rep max from submaximal loads.
Brzycki M (1993)
Strength testing—predicting a one-rep max from reps-to-fatigue
E1RM = Weight × 36/(37 - Reps). Most accurate for 3–10 rep range.
Schoenfeld BJ, et al. (2017)
Dose-response relationship between weekly resistance training volume and muscle hypertrophy
10+ weekly sets per muscle group optimal for hypertrophy. Progressive volume increase of 2–5% per mesocycle recommended.
Schoenfeld BJ (2010)
The mechanisms of muscle hypertrophy and their application to resistance training
Three primary hypertrophy mechanisms — mechanical tension, metabolic stress, muscle damage — informs the AI program builder block selection logic.
Zourdos MC, Klemp A, et al. (2016)
Novel resistance training-specific rating of perceived exertion scale measuring repetitions in reserve
Validates the RPE-based RIR scale used by our auto-regulation engine; ~2.5% load adjustment per RPE point keeps weekly volume on target.
Issurin VB (2010)
New horizons for the methodology and physiology of training periodization
Establishes block periodization as a viable alternative to linear and undulating models — drives our three-mode periodization picker (Linear / Block / DUP).
Rhea MR, Alvar BA, Burkett LN, Ball SD (2003)
A meta-analysis to determine the dose response for strength development
Daily-Undulating Periodization produces greater strength gains than Linear when matched for total volume — informs the DUP option in our program builder.
Stone MH, O'Bryant H, Garhammer J (1981)
A theoretical model of strength training
Original framework for linear periodization (volume → intensity progression across mesocycles) — still the default phase-progression in our auto-periodizer.
Behavior Change
Habit formation, adherence, and engagement psychology
Lally P, van Jaarsveld CHM, et al. (2010)
How are habits formed: Modelling habit formation in the real world
Average time to automaticity is 66 days. Missing one day has negligible impact on habit formation.
Dai H, Milkman KL, Riis J (2014)
The Fresh Start Effect: Temporal landmarks motivate aspirational behavior
Temporal landmarks (Mondays, month starts, birthdays) increase goal pursuit — the Fresh Start Effect.
Aldosary M, Alrashdan I (2021)
Predicting gym member churn using machine learning
XGBoost model achieved 92.1% accuracy predicting gym member churn from attendance and engagement features.
Burke LE, Wang J, Sevick MA (2011)
Self-monitoring in weight loss: A systematic review of the literature
Self-monitoring of diet and physical activity is the single strongest predictor of successful weight management.
Ready to coach with science?
Join trainers who use evidence-based tools to deliver better outcomes.
Start Free Trial