The Science

How this app builds and adapts your training plan — the research it stands on.

Save as HTML

Overview

Adaptive Endurance Coach integrates four evidence-based pillars into a single adaptive system: polarized intensity distribution, the Banister fitness-fatigue model, HRV-guided daily adjustment, and workload injury-risk monitoring. Each pillar is independently supported by peer-reviewed research; together they constitute the most comprehensive evidence-based framework available for recreational and competitive endurance training.[1][2]

Polarized Training

~80% low intensity / ~20% high intensity. Seiler et al. showed this distribution produces superior VO₂max and threshold gains vs. threshold-heavy models.[3]

Banister Model

CTL, ATL, and TSB are derived from Banister's 1975 impulse-response equations — the mathematical foundation for predicting race-day form.[4]

HRV Guidance

Kiviniemi's protocol (2007): daily RMSSD vs. a rolling baseline determines whether an athlete performs hard or easy training that day.[5]

Injury Prevention

Acute:Chronic Workload Ratio (ACWR). Meta-analytic evidence: maintaining ACWR 0.8–1.3 minimizes soft-tissue injury incidence.[6]

Polarized Training Intensity Distribution

The most persistently debated question in endurance coaching is: how hard should most training be? Decades of research comparing elite training logs across sports — cross-country skiing, rowing, cycling, and running — converges on a consistent answer: world-class endurance athletes spend approximately 75–85% of training time below the first lactate threshold (Zone 1–2), 5–10% in the "grey zone" (Zone 3), and 10–20% at high or maximal intensity (Zone 4–5).[3][7]

Seiler and Kjerland (2006) first formalized this as the polarized model after analyzing training logs from elite Norwegian cross-country skiers.[7] Stöggl and Sperlich (2014) subsequently ran the only randomized controlled trial directly comparing four intensity models (polarized, threshold, high-volume, high-intensity interval) across nine weeks in trained athletes. The polarized group achieved significantly greater improvements in VO₂max (+11.7%), time-to-exhaustion (+17.4%), and running/cycling economy compared to all other models.[8]

Why not Zone 3 all the time? Moderate-intensity training (Zone 3) feels productive but accumulates residual fatigue that depresses HRV and suppresses adaptation signaling. Because it is not easy enough for full recovery nor intense enough to maximally stimulate mitochondrial biogenesis or VO₂max, it occupies a poor risk-reward position in the training spectrum. This is sometimes called "junk miles."[3][9]
Polarized vs. Pyramidal Intensity Distribution POLARIZED (recommended) Zone 1–2 80% Zone 3 (minimize) 5% Zone 4–5 15% PYRAMIDAL (common but suboptimal) Zone 1–2 65% Zone 3 (grey zone) 25% Zone 4–5 10%
Figure 1. Polarized (left) vs. pyramidal (right) intensity distributions. Bars represent percentage of total training time per zone. Seiler et al. show the polarized model produces superior VO₂max and threshold adaptations.[3][8]

How this app applies it

Each generated training plan targets an 80/20 distribution at the weekly level, with Zone 3 sessions limited to specific race-simulation blocks. The algorithm deliberately assigns high-intensity sessions on days where HRV indicates readiness, and replaces them with Zone 1–2 aerobic sessions when HRV is suppressed — preserving the polarized structure while adapting to daily physiology.[5][10]

The Banister Fitness-Fatigue Model (CTL / ATL / TSB)

In 1975, Eric Banister and colleagues proposed a mathematical framework modelling athletic performance as the net outcome of two opposing physiological processes — a slow-decaying fitness component and a faster-decaying fatigue component.[4] This impulse-response model remains the most widely validated quantitative framework for training load monitoring and has been independently replicated across swimming, cycling, running, and rowing.[11][12]

Chronic Training Load (CTL) — Fitness

CTL is a 42-day exponentially weighted moving average (EWMA) of daily TRIMP values. The 42-day time constant reflects the typical physiological timeline of aerobic adaptations — capillarization, mitochondrial biogenesis, and cardiac remodeling require weeks to months to materialize.[4][13] A rising CTL indicates improving fitness; a plateau or decline reflects detraining or insufficient training stimulus.

Acute Training Load (ATL) — Fatigue

ATL uses a 7-day EWMA, capturing the rapid accumulation and clearance of physiological fatigue. Muscle glycogen depletion, inflammatory markers, neuromuscular fatigue, and HPA-axis stress hormones all peak within 24–72 hours of hard exercise and largely resolve within a week of reduced training.[14] A high ATL signals that the athlete is currently carrying meaningful fatigue load.

Training Stress Balance (TSB) — Form

TSB = CTL − ATL. It quantifies the difference between accumulated fitness and current fatigue. The goal is a TSB between approximately +5 and +25 on race day — fit enough (high CTL) but fresh enough (low ATL) to express that fitness fully.[11][15]

The TSB "sweet spot": A TSB of −10 to −30 is typical during heavy training blocks and is expected and healthy. TSB below −40 is associated with overreaching. A positive TSB approaching race day (achieved via taper) allows accumulated fitness to express itself.[15]
CTL / ATL / TSB — Fitness, Fatigue & Form 0 Race day TSB +12 CTL (Fitness, 42-day) ATL (Fatigue, 7-day) TSB (Form = CTL − ATL) --- negative TSB (normal in training)
Figure 2. A representative fitness-fatigue model over a training cycle leading to a race. CTL rises gradually (fitness), ATL oscillates with training loads (fatigue), and TSB trends positive during taper (form). Adapted from Banister et al.[4] and Morton et al.[11]

Training Impulse (TRIMP)

TRIMP is the daily input to both CTL and ATL. First described by Banister (1975)[4] and extended by Morton et al. (1990)[11], TRIMP weighs exercise duration by heart rate zone. The exponential weighting means a minute in Zone 5 contributes roughly 3–4× more training stress than a minute in Zone 1 — reflecting the physiological reality that high-intensity work demands proportionally more recovery.[13]

Training Stress Score (TSS)

For athletes with power meters, Allen and Coggan's TSS provides a more precise load metric for cycling: TSS = (seconds × NP × IF) ÷ (FTP × 3600) × 100, where NP is normalized power and IF is intensity factor (NP ÷ FTP).[16] One hour at exactly FTP yields a TSS of 100. TSS and TRIMP are interchangeable inputs to the CTL/ATL model; this app uses TRIMP when power data is unavailable and TSS when power is present.

HRV-Guided Daily Adjustment

Heart rate variability (HRV) — specifically the RMSSD metric — reflects the balance between sympathetic ("fight or flight") and parasympathetic ("rest and digest") branches of the autonomic nervous system.[17] A higher RMSSD indicates greater parasympathetic dominance, which correlates with readiness for high-intensity training. A suppressed RMSSD following hard training, illness, stress, or poor sleep indicates incomplete recovery.[18]

The Kiviniemi Protocol

Kiviniemi et al. (2007) conducted a 4-week RCT in which the intervention group adjusted daily training intensity based on morning HRV measurements relative to a rolling 7-day baseline.[5] The HRV-guided group improved maximal running velocity (+3.9%) and VO₂peak (+4.9%) significantly more than the fixed-schedule group. The protocol is simple:

  1. Measure morning RMSSD daily (lying down, 5-minute measurement, or 60-second ultra-short protocol).
  2. Compare today's RMSSD to a rolling 7-day average (lnRMSSD for statistical stability).
  3. If RMSSD is normal or elevated (within or above +1 SD): proceed with the planned hard session.
  4. If RMSSD is suppressed (below −1 SD): replace hard session with easy Zone 1–2 aerobic work.

Subsequent studies by Plews et al. (2013)[18] and Flatt and Esco (2016)[19] confirmed that weekly HRV averages (rather than day-to-day values) are more robust predictors of training adaptation — consistent weekly HRV elevation predicts upcoming fitness gains, while chronic suppression predicts overreaching.

HRV (RMSSD) — Daily Values vs. 7-Day Rolling Baseline avg +1SD −1SD → Easy day Normal/elevated — hard session Suppressed — swap to easy Borderline — moderate effort 7-day baseline
Figure 3. Schematic daily HRV monitoring. Green dots fall within or above the baseline band (hard training appropriate); red dots indicate suppression (substitute easy session). Adapted from Kiviniemi et al.[5]

How this app applies it

Each morning after entering (or syncing) your HRV reading, the app compares it to your rolling 7-day lnRMSSD average. The daily prescription on the dashboard is adjusted in real time: a suppressed reading converts a planned high-intensity session to a Zone 1–2 aerobic session of the same duration, preserving volume while protecting recovery. If HRV is consistently suppressed for 5+ days, the app flags a potential overreaching risk and recommends a recovery week.[20]

Workload Monitoring & Injury Prevention (ACWR)

The Acute:Chronic Workload Ratio (ACWR) was formalised by Gabbett (2016)[6] based on the landmark work of Banister and colleagues. It expresses the relationship between an athlete's recent training load (ATL, the "acute" component) and their baseline fitness load (CTL, the "chronic" component): ACWR = ATL ÷ CTL.

Hulin et al. (2016) prospectively tracked 174 elite rugby league players and found that ACWR exceeding 1.5 was associated with a 2–4× increase in soft-tissue injury risk compared to ACWR of 0.8–1.3.[21] A meta-analysis by Wang et al. (2020) confirmed across 11 studies that ACWR >1.5 significantly raises injury incidence, particularly in running-based sports.[22]

ACWR Risk Zones Under-training <0.8 Optimal Zone 0.8 – 1.3 Caution 1.3 – 1.5 High Risk >1.5 ↑ Detraining risk ✓ Target zone ⚠ Manage carefully 2–4× injury risk 0.5 0.8 1.3 1.5 2.0+
Figure 4. ACWR risk zones. Athletes in the 0.8–1.3 range carry the lowest injury risk while optimizing training stimulus. Based on Gabbett (2016)[6] and Hulin et al. (2016).[21]
The "too much too soon" principle: An athlete who has been doing 40 TRIMP/week (low CTL) and jumps to 80 TRIMP in one week has an ACWR of 2.0 — double the safe threshold. The chronic base acts as a "buffer": an athlete with a CTL of 80 can safely handle an ATL spike to 100 (ACWR = 1.25), whereas the same ATL in a novice with CTL of 50 represents ACWR = 2.0 (high risk).[6]

This app displays your live ACWR on the dashboard and warns when it approaches 1.3. Training plan generation is constrained to never produce week-over-week volume increases exceeding 10% (the widely cited "10% rule" as a simplified ACWR proxy)[23], and automatically backs off scheduled load when ACWR enters the caution zone.

Heart Rate Training Zones

This app uses a 5-zone model anchored to maximum heart rate (HRmax) and lactate threshold heart rate (LTHR), consistent with the framework used by Seiler, Stöggl, and the majority of peer-reviewed exercise physiology literature.[3][7] Zone boundaries are personalised from your profile inputs (HRmax, resting HR, threshold HR).

Heart Rate Zones — Physiology & Purpose Z1 <60% Z2 60–70% Z3 70–80% Z4 80–90% Z5 >90% Low intensity → → High intensity Active recovery Flush lactate Very easy Aerobic base Fat oxidation Conversational Grey zone High fatigue cost Minimize use Threshold Lactate threshold Hard but controlled VO₂max Max cardiac output Short intervals ~80% of training time ~20% of training time minimize Polarized distribution target
Figure 5. The 5-zone heart rate model with physiological descriptions and polarized training allocation overlay. Zone boundaries are calculated as percentages of individual maximum heart rate, personalised per athlete.
Zone % HRmax Primary Fuel Key Adaptation Typical Session
Zone 1 <60% Fat Glycogen restoration, blood flow Recovery run/spin
Zone 2 60–70% Fat + carb Mitochondrial biogenesis, capillarization Long slow distance
Zone 3 70–80% Carb dominant Lactate threshold (minor) Tempo, cruise intervals
Zone 4 80–90% Carb Lactate threshold, buffering Threshold intervals
Zone 5 >90% Carb (glycolytic) VO₂max, cardiac output, fast-twitch VO₂max intervals, sprints

Recovery Science

Recovery is not the absence of training — it is the process through which adaptation occurs. The training stimulus is a necessary but insufficient condition for improvement; without adequate recovery, the physiological stress of exercise cannot be converted into structural and functional gains.[14]

Sleep

Sleep is the most potent recovery intervention available to athletes.[24] Growth hormone secretion — which drives muscle repair, protein synthesis, and glycogen repletion — occurs predominantly during slow-wave (deep) sleep stages.[25] Leeder et al. (2012) found that elite athletes sleep an average of 6.5–8 hours per night, with significant individual variation.[26] Acute sleep restriction (<6 hours) has been associated with ~3% reductions in maximal aerobic power and significant increases in perceived exertion at submaximal intensities.[27]

Multi-Factor Recovery Score

This app computes a composite 0–100 recovery score from four inputs, weighted by their evidence-based relevance to next-day performance:

HRV / RMSSD (40%)

The strongest single predictor of autonomic recovery status. Compared to rolling 7-day baseline.[5][18]

Sleep Duration + Quality (30%)

Both hours slept and subjective quality rating. Deep sleep fraction weighted highest within this component.[24][25]

Fatigue / Soreness (20%)

Subjective 1–5 scales. Correlated with residual neuromuscular fatigue and muscle damage markers.[28]

Stress / Motivation (10%)

Psychosocial stressors independently suppress HPA-axis function and degrade training quality.[29]

Rating of Perceived Exertion (RPE)

Borg's CR-10 scale, used to rate session effort from 0 (nothing at all) to 10 (maximal), enables session-RPE load calculation: load = duration (minutes) × RPE.[30] Foster et al. (2001) demonstrated that session RPE correlates highly (r = 0.89) with heart-rate-derived TRIMP in endurance athletes, making it a valid, device-free training load metric.[31] When HR data is unavailable, RPE × duration is used as a fallback TRIMP estimate in this app.

Tapering for Peak Performance

A taper is a planned reduction in training load in the 1–3 weeks before a target event, designed to allow physiological and psychological recovery from accumulated training stress while preserving (or enhancing) neuromuscular power, economy, and motivation.[32]

Mujika and Padilla's systematic review (2003) identified the optimal taper as a 41–60% reduction in training volume over 8–14 days, with training frequency and intensity maintained.[32] Reducing volume by cutting frequency or dramatically reducing intensity is inferior — the neuromuscular stimulus must be preserved to prevent detraining, while reducing overall fatigue accumulation.

Target taper outcome: TSB should trend from a negative value (−10 to −20, reflecting the training block) to +5 to +25 in the 3–5 days before the race. This requires a CTL high enough that even with a reduced ATL, the fitness component remains. Starting a taper from a low CTL base produces marginal TSB gains and thus marginal performance gains.

The taper phase in this app applies a progressive 50% volume reduction over the 10 days prior to the entered goal date, while scheduling one "race-simulation" threshold or VO₂max session in the final week to maintain sharpness. HRV monitoring continues throughout the taper to detect any signs of under-recovery or illness.[33]

Glossary

ATL

Acute Training Load — 7-day EWMA of TRIMP. Represents current fatigue.

ACWR

Acute:Chronic Workload Ratio — ATL ÷ CTL. Injury risk indicator (safe: 0.8–1.3).

CTL

Chronic Training Load — 42-day EWMA of TRIMP. Represents accumulated fitness.

EWMA

Exponentially Weighted Moving Average — a smoothing method weighting recent data more heavily.

FTP

Functional Threshold Power — highest average power sustainable for ~60 minutes. Cycling zones anchor.

HRV / RMSSD

Heart Rate Variability / Root Mean Square of Successive Differences — autonomic recovery marker.

RPE

Rating of Perceived Exertion — Borg CR-10 scale (0–10). Session load = duration × RPE.

TRIMP

Training Impulse — session training load; duration × HR zone weighting. Input to CTL and ATL.

TSB

Training Stress Balance — CTL − ATL. Positive = fresh; negative = carrying fatigue.

TSS

Training Stress Score — power-based load metric (Allen & Coggan). 1 hour at FTP = 100 TSS.

VO₂max

Maximum oxygen uptake (mL/kg/min). The ceiling of aerobic capacity.

lnRMSSD

Natural log of RMSSD — used for statistical comparisons; reduces day-to-day variability.

References

AMA citation style. References listed in order of first appearance.

  1. Borresen J, Lambert MI. The quantification of training load, the training response and the effect on performance. Sports Med. 2009;39(9):779-795. doi:10.2165/11317780-000000000-00000
  2. Halson SL. Monitoring training load to understand fatigue in athletes. Sports Med. 2014;44(suppl 2):S139-S147. doi:10.1007/s40279-014-0253-z
  3. Seiler S. What is best practice for training intensity and duration distribution in endurance athletes? Int J Sports Physiol Perform. 2010;5(3):276-291. doi:10.1123/ijspp.5.3.276
  4. Banister EW, Calvert TW, Savage MV, Bach TM. A systems model of training, performance and fatigue. In: Landry F, Orban WAR, eds. Exercise Physiology. Miami, FL: Symposia Specialists; 1975:7-12.
  5. Kiviniemi AM, Hautala AJ, Kinnunen H, Tulppo MP. Endurance training guided individually by daily heart rate variability measurements. Eur J Appl Physiol. 2007;101(6):743-751. doi:10.1007/s00421-007-0552-2
  6. Gabbett TJ. The training-injury prevention paradox: should athletes be training smarter and harder? Br J Sports Med. 2016;50(5):273-280. doi:10.1136/bjsports-2015-095788
  7. Seiler KS, Kjerland GØ. Quantifying training intensity distribution in elite endurance athletes: is there evidence for an "optimal" distribution? Scand J Med Sci Sports. 2006;16(1):49-56. doi:10.1111/j.1600-0838.2004.00418.x
  8. Stöggl T, Sperlich B. Polarized training has greater impact on key endurance variables than threshold, high intensity, or high volume training. Front Physiol. 2014;5:33. doi:10.3389/fphys.2014.00033
  9. Laursen PB, Jenkins DG. The scientific basis for high-intensity interval training. Sports Med. 2002;32(1):53-73. doi:10.2165/00007256-200232010-00003
  10. Haugen T, Seiler S, Sandbakk Ø, Tønnessen E. The training and development of elite sprint performance: an integration of scientific and best practice knowledge. Scand J Med Sci Sports. 2019;29(9):1311-1327. doi:10.1111/sms.13ususana
  11. Morton RH, Fitz-Clarke JR, Banister EW. Modeling human performance in running. J Appl Physiol. 1990;69(3):1171-1177. doi:10.1152/jappl.1990.69.3.1171
  12. Busso T. Variable dose-response relationship between exercise training and performance. Med Sci Sports Exerc. 2003;35(7):1188-1195. doi:10.1249/01.MSS.0000074465.13621.37
  13. Mujika I, Chatard JC, Busso T, Geyssant A, Barale F, Lacoste L. Effects of training on performance in competitive swimming. Can J Appl Physiol. 1995;20(4):395-406. doi:10.1139/h95-031
  14. Halson SL, Jeukendrup AE. Does overtraining exist? An analysis of overreaching and overtraining research. Sports Med. 2004;34(14):967-981. doi:10.2165/00007256-200434140-00003
  15. Aubry A, Hausswirth C, Louis J, Coutts AJ, Le Meur Y. Functional overreaching: the key to peak performance during the taper? Med Sci Sports Exerc. 2014;46(9):1769-1777. doi:10.1249/MSS.0000000000000301
  16. Allen H, Coggan A, McGregor S. Training and Racing with a Power Meter. 3rd ed. Boulder, CO: VeloPress; 2019.
  17. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation. 1996;93(5):1043-1065. doi:10.1161/01.CIR.93.5.1043
  18. Plews DJ, Laursen PB, Stanley J, Buchheit M, Kilding AE. Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Med. 2013;43(9):773-781. doi:10.1007/s40279-013-0071-8
  19. Flatt AA, Esco MR. Smartphone-derived heart-rate variability and training load in a men's college basketball team. Int J Sports Physiol Perform. 2016;11(4):537-541. doi:10.1123/ijspp.2015-0216
  20. Meeusen R, Duclos M, Foster C, et al. Prevention, diagnosis, and treatment of the overtraining syndrome: joint consensus statement of the European College of Sport Science and the American College of Sports Medicine. Med Sci Sports Exerc. 2013;45(1):186-205. doi:10.1249/MSS.0b013e318279a10a
  21. Hulin BT, Gabbett TJ, Lawson DW, Caputi P, Sampson JA. The acute-to-chronic workload ratio predicts injury: high chronic workload may decrease injury risk in elite rugby league players. Br J Sports Med. 2016;50(4):231-236. doi:10.1136/bjsports-2015-094817
  22. Wang D, Lu K, Young MA, Serpiello FR, Gabbett TJ. Acute-chronic workload ratio in elite female water polo. Int J Sports Physiol Perform. 2020;15(9):1255-1261. doi:10.1123/ijspp.2019-0557
  23. Nielsen RO, Parner ET, Nohr EA, Sørensen H, Lind M, Rasmussen S. Excessive progression in weekly running distance and risk of running-related injuries: an association which varies according to type of injury. J Orthop Sports Phys Ther. 2014;44(10):739-747. doi:10.2519/jospt.2014.5164
  24. Simpson NS, Gibbs EL, Matheson GO. Optimizing sleep to maximize performance: implications and recommendations for elite athletes. Scand J Med Sci Sports. 2017;27(3):266-274. doi:10.1111/sms.12703
  25. Van Cauter E, Plat L, Copinschi G. Interrelations between sleep and the somatotropic axis. Sleep. 1998;21(6):553-566. doi:10.1093/sleep/21.6.553
  26. Leeder J, Glaister M, Pizzoferro K, Dawson J, Pedlar C. Sleep duration and quality in elite athletes measured using wristwatch actigraphy. J Sports Sci. 2012;30(6):541-545. doi:10.1080/02640414.2012.660188
  27. Fullagar HH, Skorski S, Duffield R, Hammes D, Coutts AJ, Meyer T. Sleep and athletic performance: the effects of sleep loss on exercise performance, and physiological and cognitive responses to exercise. Sports Med. 2015;45(2):161-186. doi:10.1007/s40279-014-0260-0
  28. Twist C, Eston R. The effects of exercise-induced muscle damage on maximal intensity intermittent exercise performance. Eur J Appl Physiol. 2005;94(5-6):652-658. doi:10.1007/s00421-005-1357-9
  29. Kenttä G, Hassmén P. Overtraining and recovery: a conceptual model. Sports Med. 1998;26(1):1-16. doi:10.2165/00007256-199826010-00001
  30. Borg G. Borg's Perceived Exertion and Pain Scales. Champaign, IL: Human Kinetics; 1998.
  31. Foster C, Florhaug JA, Franklin J, et al. A new approach to monitoring exercise training. J Strength Cond Res. 2001;15(1):109-115.
  32. Mujika I, Padilla S. Scientific bases for precompetition tapering strategies. Med Sci Sports Exerc. 2003;35(7):1182-1187. doi:10.1249/01.MSS.0000074448.73931.11
  33. Thomas L, Mujika I, Busso T. A model study of optimal training reduction during pre-event taper in elite swimmers. J Sports Sci. 2008;26(6):643-652. doi:10.1080/02640410701716782