5D Library › Ageing Kinetics, the mTOR-AMPK Autophagy Switch and Hormesis · 中文
Chapter 6 Zone 2, Polarised Training and Banister’s Impulse–Response Model
How intensity is divided into zones
As exercise intensity rises, blood lactate and ventilation show two turning points. Below the first (the first lactate threshold, LT1, or first ventilatory threshold, VT1), blood lactate stays close to resting levels and one can easily speak in full sentences; above the second (the second lactate threshold, LT2, or second ventilatory threshold, VT2, roughly the “maximal lactate steady state”), lactate keeps accumulating and the effort cannot be sustained for long. With the two thresholds as boundaries, intensity can be divided into three zones: zone 1 below VT1, zone 2 between the thresholds and zone 3 above VT2.
The popular “Zone 2” comes from a different, five-zone scheme and usually means an intensity at or just below LT1 (blood lactate of about 2 mmol/L is a common reference), which lies in the upper part of zone 1 in the three-zone scheme. The “zone 2” of the two schemes is not the same thing, a source of much confusion.
Note
Thresholds are individual and need to be measured. The same heart rate or speed can fall in different zones for different people, and zones derived from a maximal heart rate estimated as “220 minus age” can be badly wrong. The trial in Chapter 8 therefore sets intensity from each participant’s first ventilatory threshold and “talking” heart rate measured at enrolment, not from a formula.
Banister’s “fitness–fatigue” model
In 1975 Banister, Calvert, Savage and Bach proposed a systems model linking training to performance (Banister et al. 1975). It treats each training session as an “impulse” with two effects: a “fitness” effect that builds slowly and fades slowly, and a “fatigue” effect that builds quickly and fades quickly. Performance is the difference between the two.
Calculation
The model. With training load on day , performance on day is
with the illustrative textbook values , , days and days. Fatigue has the larger coefficient but fades faster, so when training is cut back (a taper), fatigue fades before fitness and performance briefly rises. The load is the sum of minutes in the three intensity zones, each multiplied by a “zone weight”. The program compares four 12-week plans, each of 300 minutes a week (five 60-minute sessions), all followed by a 10-day taper in which volume falls by 60 %; the second part uses three different sets of zone weights.
Calculation
# Banister fitness-fatigue model, k1 = 1, k2 = 2, tau1 = 42 d, tau2 = 7 d (illustrative).
# Load = zone minutes x zone weights; 4 plans of 300 min/week for 12 weeks, then a 10-day taper.
import math
K1, K2, T1, T2 = 1.0, 2.0, 42.0, 7.0
PLANS = {"polarised 80/0/20": (0.80, 0.00, 0.20), "threshold 50/50/0": (0.50, 0.50, 0.00),
"zone-1-only 100/0/0": (1.00, 0.00, 0.00), "pyramidal 75/20/5": (0.75, 0.20, 0.05)}
def daily_loads(shares, weights, weeks=12, taper_days=10, cut=0.6):
load = sum(300 / 5 * sh * w for sh, w in zip(shares, weights)) # 5 sessions a week
return [load * ((1 - cut) if d >= weeks * 7 else 1.0) * (d % 7 < 5)
for d in range(weeks * 7 + taper_days)]
def performance(loads, day):
return sum(w * (K1 * math.exp(-(day - s) / T1) - K2 * math.exp(-(day - s) / T2))
for s, w in enumerate(loads[:day]))
for weights in ((1, 2, 4), (1, 3, 9), (1, 1.5, 2)):
res = {}
for name, sh in PLANS.items():
L = daily_loads(sh, weights)
res[name.split()[0]] = (performance(L, 84), performance(L, 94))
print(f"zone weights {weights}: performance at week 12 -> after the taper")
print(" " + ", ".join(f"{k} {a:.0f}->{b:.0f}" for k, (a, b) in res.items()))
print(" ranking after taper: " + " > ".join(sorted(res, key=lambda k: -res[k][1])))
L1 = daily_loads(PLANS["polarised 80/0/20"], (1, 2, 4))
print(f"polarised, double volume: {performance([2 * x for x in L1], 94) / performance(L1, 94):.2f}"
" x the gain")
# zone weights (1, 2, 4): performance at week 12 -> after the taper
# polarised 1643->1680, threshold 1540->1575, zone-1-only 1027->1050, pyramidal 1386->1418
# ranking after taper: polarised > threshold > pyramidal > zone-1-only
# zone weights (1, 3, 9): performance at week 12 -> after the taper
# polarised 2669->2730, threshold 2053->2100, zone-1-only 1027->1050, pyramidal 1848->1890
# ranking after taper: polarised > threshold > pyramidal > zone-1-only
# zone weights (1, 1.5, 2): performance at week 12 -> after the taper
# polarised 1232->1260, threshold 1283->1313, zone-1-only 1027->1050, pyramidal 1181->1208
# ranking after taper: threshold > polarised > pyramidal > zone-1-only
# polarised, double volume: 2.00 x the gain
The model first shows the effect of the taper: all four plans improve by about 2 % after the 10-day taper, because the fatigue term fades faster than the fitness term.
But the second part of the program exposes a fundamental limitation: the model’s “verdict” on intensity distribution is determined entirely by the zone weights we choose. With weights of 1, 2, 4 or 1, 3, 9, the polarised plan (80/0/20) beats the threshold plan (50/50/0); with weights of 1, 1.5, 2 the threshold plan wins. In a linear model performance depends only on the weighted total load, and the ranking of weighted loads depends on the weights. Moreover, the linear model rewards load without limit: doubling training volume exactly doubles the gain (2.00 times). That is clearly not how real training works; overtraining and injury do not exist in the model.
Evidence
Hellard and colleagues (2006) assessed the Banister model with a season of data from nine elite swimmers (Hellard et al. 2006). Performance was significantly related to training load, and repeated estimation appeared stable; but the parameters most useful for monitoring training had wide confidence intervals and some were highly collinear, which in the authors’ view made their physiological interpretation worthless, and they suggested looking for other modelling methods.
Polarised training: descriptive evidence and trials
Evidence
Description. Seiler and Kjerland (2006) recorded 32 consecutive days of training by eleven well-trained junior cross-country skiers, defining three intensity zones by ventilatory thresholds (Seiler and Kjerland 2006). Classifying each endurance session by heart rate, about 75 % of sessions fell in zone 1, about 8 % in zone 2 and about 17 % in zone 3, with a similar distribution when sessions were classified by perceived exertion. They called this distribution “polarised”: most time at very low intensity, a substantial share at high intensity, little near the threshold. This is a description of how elite athletes actually train; on its own it does not show that the distribution is better.
Evidence
A randomised trial. Stöggl and Sperlich (2014) randomised 48 well-trained endurance athletes (stratified by baseline maximal oxygen uptake and sport) to high-volume, threshold, high-intensity interval and polarised groups for 9 weeks; 41 completed (Stöggl and Sperlich 2014). Maximal oxygen uptake rose by 11.7 % (6.8 mL/kg/min) in the polarised group and by about 4.8 % in the interval group, with no further improvement in the threshold and high-volume groups; time to exhaustion did not differ significantly between the four groups. No blinding was reported, dropouts were uneven across groups (four in the threshold group) and diet was not controlled.
Evidence
Meta-analysis. Oliveira, Boppre and Fonseca (2024) pooled 17 studies with 437 athletes (Oliveira et al. 2024). Compared with other intensity distributions, polarised training had a small advantage for maximal oxygen uptake (standardised mean difference 0.24, 95 % confidence interval 0.01–0.48; 11 studies, 284 people), but only in interventions shorter than 12 weeks and in highly trained athletes; there were no differences in time-trial performance, time to exhaustion, or speed or power at the second threshold. The randomised trials were rated as having “some concerns” for risk of bias.
Almost all of this evidence comes from trained athletes; for sedentary adults, sticking with training and staying safe probably matter more than the distribution itself.
Zone 2 and mitochondria: the disputed human evidence
The popular claim is that Zone 2 training “specifically” raises mitochondrial number and function and is therefore the core of training for longevity. The human evidence is more complicated.
Evidence
Granata, Jamnick and Bishop’s 2018 review concluded that training-induced changes in mitochondrial content relate mainly to training volume, while changes in mitochondrial respiratory function relate mainly to relative intensity; the two often diverge, and acute responses to a single session do not necessarily predict adaptations to training (Granata et al. 2018).
Evidence
Gillen and colleagues (2016) assigned 25 sedentary men to sprint interval training (three 20-second all-out cycling sprints, about 10 minutes per session), moderate-intensity continuous training (45 minutes of cycling at about 70 % of maximal heart rate) or a non-training control, three times a week for 12 weeks (Gillen et al. 2016). Maximal oxygen uptake rose by about 19 % in both training groups, and improvements in insulin sensitivity and muscle citrate synthase activity were similar, even though the sprint group’s training volume and time commitment were one fifth as large.
Evidence
Storoschuk, Moran-MacDonald, Gibala and Gurd’s 2025 narrative review assessed Zone 2 training (defined as intensity below the first lactate threshold) for improving mitochondrial capacity and cardiorespiratory fitness in the general population (Storoschuk et al. 2025). They concluded that Zone 2 helps to build basic endurance and allows training volume to be accumulated with little fatigue, but that there is no evidence it is superior to higher intensities for improving mitochondrial capacity, fat oxidation or fitness; at equal volume, higher intensities raise maximal oxygen uptake more; and many studies neither tested the popular definition of Zone 2 directly nor verified intensity with lactate measurements.
Note
A correction of the author’s framework. “Zone 2 specifically raises mitochondria” is not supported in humans, and the book does not present it as settled. A safer statement is that low-intensity, longer training is a low-fatigue, low-risk way to accumulate volume, and volume is linked to increases in mitochondrial content; high-intensity training can bring similar or larger improvements in fitness in less time. For sedentary beginners, the main reasons to start at low intensity are safety and sustainability, not some specific mitochondrial mechanism.
Hypothesis · not established
In sedentary adults, adherence explains 12-week changes in fitness better than intensity distribution (not established). If so, then among sedentary adults randomised to different intensity distributions with the same weekly time, the differences in fitness between distributions should shrink markedly once participants are stratified by the training minutes they actually completed, and completed minutes should explain more of the change in fitness than group assignment. If, among people who completed the same minutes, different distributions still produce clearly different changes in fitness, the hypothesis is weakened.
Chapter summary
With the first and second lactate (or ventilatory) thresholds as boundaries, intensity falls into three zones; the popular “Zone 2” lies in the upper part of zone 1 of the three-zone scheme, and thresholds need to be measured individually. Banister’s fitness–fatigue model describes the rebound in performance during a taper, but its “verdict” on intensity distribution depends entirely on the zone weights chosen, the linear model rewards unlimited volume, and in real data its key parameters have wide confidence intervals and are collinear. Polarised training describes how elite athletes train; randomised trials and a meta-analysis show only a small short-term advantage for maximal oxygen uptake and none for performance. “Zone 2 specifically raises mitochondria” is not supported in humans: volume is linked to mitochondrial content, and high intensity can bring similar fitness gains in less time. The next chapter reads these modern studies beside ancient writings on “movement” and “nourishing life”.
References
Fan, Y. (2026). Ageing Kinetics, the mTOR-AMPK Autophagy Switch and Hormesis: From the Gompertz Mortality Law and Mitohormesis to Cardiorespiratory Fitness Epidemiology, Read Beside Classical Chinese Longevity Philosophy [衰老动力学、mTOR-AMPK自噬开关与毒物兴奋效应的生物数学模型] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23770092 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).