Endurance Training

01 What endurance training is, and four things it is not

Endurance training is the systematic application of prolonged locomotor work in order to raise the sustainable rate of energy conversion in muscle, the capacity of the heart and blood to deliver oxygen, and the economy with which that work is turned into speed or power. Joyner and Coyle, in a 2008 physiological review of champions, reduced performance in endurance sport to three interacting terms: maximal oxygen uptake (VO₂max), the fraction of VO₂max that can be sustained (often discussed as a lactate threshold), and efficiency or economy, the oxygen cost of a given running speed or cycling power (Joyner and Coyle, 2008). That three-term map is established as a useful organising scheme. It is not a licence to treat any one term as the event.

Four things follow immediately.

First, endurance training is not a VO₂max cult. Bassett and Howley, reviewing limiting factors for maximum oxygen uptake, concluded that in the exercising human VO₂max is limited by the ability of the cardiorespiratory system to deliver oxygen to the working muscle, not by a mitochondrial ceiling in the muscle itself (Bassett and Howley, 2000). Levine, writing a contemporaneous physiology essay, asked what is actually known about VO₂max and what still is not (Levine, 2008). A large VO₂max is common among successful endurance athletes. It is not sufficient, and in some events it is not even the best available predictor.

Second, it is not one physiology for every sport. Running, cycling, swimming, rowing, triathlon, and cross-country skiing share the Fick equation and do not share the same economy, the same duty cycle, or the same heat and hydrostatic load. Lucía’s physiology of professional road cycling is not Saunders’ running-economy review (Lucía, Hoyos, and Chicharro, 2001; Saunders, Pyne, Telford, and Hawley, 2004). Treating them as interchangeable is how a cycling FTP number becomes a swimming set.

Third, it is not a training plan. Seiler’s 2010 question, “what is best practice for training intensity and duration distribution in endurance athletes?”, was asked of athletes already training around ten to twenty hours a week (Seiler, 2010). Recreational runners who copy the shape of that distribution without the volume are doing a different experiment. This document reports those experiments. It specifies no week, no zone, and no session for any person.

Fourth, it is not medical advice and it is not a longevity protocol. Cardiorespiratory fitness predicts mortality in observational cohorts (Blair et al., 1989; Kodama et al., 2009). That is a different claim from “this interval session will lengthen a life.” Sibling articles in this series take sports nutrition, strength training, and concurrent training as neighbouring problems. Protein, creatine, and resistance exercise are not spent here as if they were endurance physiology.

02 The Fick equation: VO2max, VO2peak, and delivery

VO₂ is cardiac output times the arterial–mixed venous oxygen difference. VO₂max is the plateau of that product as work rate rises. VO₂peak is the highest value recorded when a plateau is not demonstrated. The distinction is established in laboratory usage and routinely collapsed in field testing. Collapsing it does not make a peak a maximum.

Bassett and Howley assembled three lines of evidence that oxygen delivery, not muscle oxygen use, sets VO₂max in the intact human: when delivery is altered by blood doping, hypoxia, or beta-blockade, VO₂max moves with it; training-induced increases in VO₂max come primarily from increased maximal cardiac output; and the additional oxygen-extracting capacity of muscle is not fully used at VO₂max (Bassett and Howley, 2000). Their earlier paper had already staged the “classical” versus “contemporary” argument about whether the limit sits in the heart or the periphery (Bassett and Howley, 1997). Noakes wrote a rebuttal in that same argument (Noakes, 1998). The debate is real. The delivery-limitation case in healthy humans is strongly supported. It is not a claim that mitochondria are irrelevant to performance. Mitochondria are highly relevant to the fraction of VO₂max that can be held, to substrate choice, and to the delay of fatigue. They are a weak candidate for the VO₂max ceiling itself.

Levine’s 2008 essay is the right caution against treating the number as a settled oracle. Measurement error, treadmill versus cycle mode, the difference between a true plateau and a peak, and the contribution of haemoglobin mass all sit inside the same three-letter abbreviation (Levine, 2008). A cycle VO₂peak in a runner is not the runner’s running VO₂max. A 5% change after a study that did not report a plateau is not a 5% change in maximal oxygen transport.

Coyle and colleagues measured the time course of loss after stopping prolonged intense endurance training: central and peripheral adaptations do not decay on the same clock (Coyle, Martin, Sinacore, Joyner, Hagberg, and Holloszy, 1984). That detrain study is strongly supported as evidence that the trained state is maintained work, not a stored trait. It is not a taper recipe.

03 Cardiac output and stroke volume

Maximal cardiac output is the product of stroke volume and heart rate. Maximal heart rate changes little with training and declines with age. The trained increase in maximal cardiac output is therefore a stroke-volume story. That statement is established as textbook cardiovascular physiology and is the load-bearing clause of Bassett and Howley’s delivery argument (Bassett and Howley, 2000).

What is less settled is how stroke volume rises: plasma volume expansion, eccentric left-ventricular remodelling, enhanced filling, and reduced afterload all contribute, in proportions that differ by sex, age, and sport. Helgerud and colleagues randomised forty moderately trained men to four work-matched eight-week programmes and reported that aerobic high-intensity intervals improved VO₂max more than moderate training, with stroke volume among the measured responses (Helgerud et al., 2007). That is a randomised human training study in a defined recreational-to-moderately-trained sample. It is strongly supported as evidence that intensity, when work is matched, can move VO₂max more than moderate continuous running in that population. It is not evidence that 4 × 4-minute intervals are a universal method, and it is not an elite-athlete trial.

The same cardiac remodelling that raises stroke volume is the substrate of the “athlete’s heart” literature in Part Five. A larger end-diastolic volume is an endurance adaptation until it is asked to be a disease. La Gerche and colleagues, using echocardiography and magnetic resonance in endurance athletes, reported exercise-induced right-ventricular dysfunction and structural remodelling after intense endurance work (La Gerche et al., 2012). That is an athlete-imaging study, not a mortality trial. It is emerging as a mechanism for arrhythmias in a subset of high-volume athletes. It is not a reason to treat recreational jogging as cardiotoxic.

04 Extraction: capillaries, mitochondria, oxidative enzymes

Once oxygen arrives, the muscle has to take it. Capillary density, myoglobin, mitochondrial volume density, and the maximal activities of oxidative enzymes (citrate synthase, succinate dehydrogenase, 3-hydroxyacyl-CoA dehydrogenase) are the classical peripheral adaptations. Holloszy and Coyle’s 1984 review remains the clean statement: regularly performed endurance exercise increases the mitochondrial content and respiratory capacity of muscle fibres; as a consequence, the same absolute intensity disturbs homeostasis less in the trained muscle; the metabolic consequences include slower glycogen use, greater fat oxidation, and less lactate accumulation at a given work rate (Holloszy and Coyle, 1984). That review is established for the direction of the adaptations. It is not a time-course table for any one person.

Egan and Zierath reviewed the molecular regulation of those adaptations, with AMPK, CaMK, p38 MAPK, and PGC-1α as the usual suspects linking contractile activity to mitochondrial biogenesis (Egan and Zierath, 2013). That is a molecular-review synthesis, strongly supported as a map of pathways and not a claim that any one “mitochondrial supplement” recapitulates training. Granata, Oliveira, Little, Renner, and Bishop reported that training-induced changes in mitochondrial content and respiratory function are not the same measurement, and that intensity can modulate PGC-1α, p53, and mitochondrial respiration without a matching change in content (Granata, Oliveira, Little, Renner, and Bishop, 2016; Granata, Jamnick, and Bishop, 2018). The marketing claim that “zone 2 builds mitochondria” as if mitochondria were a single countable organelle with a single best intensity is a category error. Content, cristae density, and respiratory function can dissociate. That dissociation is strongly supported.

MacInnis, Zacharewicz, Martin, and colleagues reported superior mitochondrial adaptations in human skeletal muscle after interval compared with continuous training in a within-subject design (MacInnis et al., 2017). MacInnis and Gibala, in the companion physiology review, classified interval work as HIIT (near-maximal) versus SIT (supramaximal) and noted that both induce the classic endurance adaptations, often with less total work (MacInnis and Gibala, 2017). These are strongly supported as evidence that intensity is a potent mitochondrial stimulus. They are not evidence that low-intensity training is wasted. Volume of low-intensity work remains the bulk of what elite endurance athletes actually do (Seiler, 2010; Seiler and Kjerland, 2006).

05 Lactate, thresholds, and ventilatory breakpoints

Lactate is not a toxin and a “lactate threshold” is not one place. Faude, Kindermann, and Meyer reviewed the family of lactate-threshold concepts and asked how valid they are for assessing endurance capacity; their conclusion was that many named thresholds exist, that terminology has been a source of confusion for decades, and that validity depends on the construct being claimed (Faude, Kindermann, and Meyer, 2009). That review is established as a warning against treating a single blood-lactate number as a physiological cliff.

Beaver, Wasserman, and Whipp described the V-slope method for detecting an “anaerobic threshold” from the VCO₂–VO₂ relationship, attributing excess CO₂ to bicarbonate buffering of lactic acid (Beaver, Wasserman, and Whipp, 1986). Ventilatory thresholds (VT1, VT2) are gas-exchange events. Lactate thresholds are blood events. They often occur near each other. They are not the same measurement, and they do not become the same measurement by being coloured the same on a consumer watch.

Critical power (cycling) and critical speed (running) are a different construct again: the asymptote of the hyperbolic power–duration or speed–duration relationship, with W′ (or D′) the finite work capacity above that asymptote (Poole, Burnley, Vanhatalo, Rossiter, and Jones, 2016; Vanhatalo, Jones, and Burnley, 2011; Jones, Vanhatalo, Burnley, Morton, and Poole, 2017). Jones, Burnley, Black, Poole, and Vanhatalo later argued that critical power is a stronger candidate for a “maximal metabolic steady state” than any one lactate threshold (Jones, Burnley, Black, Poole, and Vanhatalo, 2019). That argument is strongly supported as physiology. It is emerging as a coaching replacement for lactate thresholds, because field estimation of CP still carries protocol dependence.

The commercial “zone 2” is usually a heart-rate band below an estimated first threshold. It is not a validated laboratory construct. Using it as a synonym for Fat_max, VT1, LT1, and conversational pace at once is how a useful easy day becomes a metaphysical region.

06 Economy: the neglected third term

Running economy is the oxygen (or energy) cost of a given submaximal speed. Conley and Krahenbuhl showed, in highly trained distance runners, that economy discriminated performance when VO₂max did not (Conley and Krahenbuhl, 1980). Saunders, Pyne, Telford, and Hawley reviewed the factors that affect running economy: training status, anthropometry, kinematics, elastic energy, and environment (Saunders, Pyne, Telford, and Hawley, 2004). Barnes and Kilding reviewed strategies that have been tried to improve it (Barnes and Kilding, 2015). These papers are strongly supported as a research map. They do not yield a single drill that raises economy on demand. Footwear, historically a footnote, became a performance technology after these reviews; this article does not treat super-shoe literature as a closed case.

Cycling economy (or efficiency) is a different measurement: oxygen cost of a given external power, complicated by the fact that the bicycle, not the ground, is the machine. Lucía, Hoyos, and Chicharro’s account of professional road cycling treats economy, aerodynamics, and pacing as coequal with VO₂max (Lucía, Hoyos, and Chicharro, 2001). Swimming “economy” is more nearly a propelling-efficiency problem. Toussaint and colleagues measured the propelling efficiency of front-crawl swimming: a large fraction of mechanical work does not contribute to forward speed (Toussaint and Beek, 1992; Toussaint, Hollander, van den Berg, and de Groot, 1988). That is established biomechanics. It is why a swimming VO₂max transferred from a treadmill is a poor coaching number.

07 Substrate use, glycogen, and the fat-burning zone

Romijn and colleagues used stable-isotope tracers and indirect calorimetry in five trained subjects at 25%, 65%, and 85% of VO₂max: plasma glucose uptake and muscle glycogen oxidation rose with intensity; peripheral lipolysis was high at low intensity and did not increase further, while fatty-acid oxidation fell at 85% (Romijn et al., 1993). Brooks and Mercier named the “crossover” concept: as intensity rises, carbohydrate contribution rises, and endurance training shifts the crossover to the right (Brooks and Mercier, 1994). Those papers are established human metabolism. They are why the “fat-burning zone” as a weight-loss prescription is a confusion of which fuel is oxidised during the session with which fuel determines 24-hour energy balance.

Achten, Gleeson, and Jeukendrup described a protocol to estimate Fat_max, the intensity eliciting maximal fat oxidation (Achten, Gleeson, and Jeukendrup, 2002). Achten, Venables, and Jeukendrup then reported Fat_max characteristics in trained men, including reliability (Achten, Venables, and Jeukendrup, 2003). Fat oxidation rates were higher during running than cycling across a wide intensity range in a companion comparison (Achten, Venables, and Jeukendrup, 2003). Fat_max is a real, measurable intensity. It is typically near the first threshold in trained people. It is not a superior fat-loss zone, and it is not the intensity that maximises glycogen-sparing adaptations in every protocol.

Muscle glycogen remains the substrate that limits prolonged hard work. Bergström, Hermansen, Hultman, and Saltin tied diet, muscle glycogen, and physical performance in human biopsy experiments (Bergström, Hermansen, Hultman, and Saltin, 1967; Bergström and Hultman, 1966). Coyle and colleagues showed that feeding carbohydrate during prolonged strenuous exercise alters muscle glycogen use (Coyle et al., 1986). Those experiments are established. They are why carbohydrate is a performance substrate in this article and a lifestyle preference in a different kind of document. Sibling title: Sports Nutrition. The interaction is restated in section 26, not spent here as a supplement catalogue.

08 Training zones, and the three-intensity model that actually travels

Consumer watches sell five, six, or seven zones. The research literature that describes elite athletes usually collapses intensity into three: low (below the first lactate/ventilatory threshold), threshold (between thresholds), and high (above the second threshold, including intervals at or above VO₂max). Seiler and Kjerland quantified that distribution in elite endurance athletes and asked whether an “optimal” pattern existed in the observational record (Seiler and Kjerland, 2006). Seiler later restated the question as a best-practice essay: successful endurance training manipulates intensity, duration, and frequency to raise performance while limiting negative outcomes, and nationally competitive athletes training many hours per week typically accumulate about 80% of sessions at low intensity (Seiler, 2010). Sylta, Tønnessen, and Seiler compared three methods of turning heart-rate data into that quantification and showed that the method changes the apparent distribution (Sylta, Tønnessen, and Seiler, 2014). Established: elite endurance athletes, described observationally, do mostly easy work. Not established: that the same percentage is causal, portable to low-volume athletes, or best computed from heart rate, session RPE, or running speed.

Billat’s interval work at vVO₂max is a different programming language again: time at VO₂max as a stimulus, with overtraining markers as a measured risk (Billat, Flechet, Petit, Muriaux, and Koralsztein, 1999). Buchheit and Laursen’s two-part “programming puzzle” remains the most careful attempt to put HIIT variables (intensity, duration, recovery, series, modality) onto a physiological map rather than a slogan (Buchheit and Laursen, 2013a; Buchheit and Laursen, 2013b). Those reviews are strongly supported as a programming taxonomy. They are not a session builder.

09 Low-intensity training, long slow distance, and zone-2 marketing

Low-intensity, long-duration work is the majority of what documented elite endurance athletes do (Seiler and Kjerland, 2006; Seiler, 2010; Esteve-Lanao, San Juan, Earnest, Foster, and Lucia, 2005). Long slow distance is an older name for a subset of that work. Zone 2 is a newer commercial name that borrows the prestige of the first threshold and often ignores how the threshold was measured.

What low-intensity training is strongly supported to do, in the presence of adequate volume: raise mitochondrial enzyme activities and capillary density over months (Holloszy and Coyle, 1984), shift Fat_max and the crossover (Brooks and Mercier, 1994), and allow a large weekly volume without the same autonomic and mechanical cost as a week of threshold work. What it is not supported to do: uniquely “build mitochondria” in a way interval training cannot (MacInnis and Gibala, 2017; Granata, Jamnick, and Bishop, 2018); oxidise a magic quantity of fat that then appears as fat loss independent of energy balance (Romijn et al., 1993); or replace the need for some high-intensity work in trained people who wish to raise VO₂max (Helgerud et al., 2007; Midgley, McNaughton, and Wilkinson, 2006).

The marketing claim is a bait-and-switch. The observational elite pattern is real. The inference “therefore zone 2 is the longevity intensity” is a different paper, usually unwritten.

10 Tempo, threshold, and the middle intensity

Threshold or “tempo” training occupies the region between the first and second transitions: hard enough to accumulate lactate, easy enough (in principle) to be steady. Esteve-Lanao, Foster, Seiler, and Lucia compared two programmes in subelite endurance runners that differed in how much volume sat clearly below versus within the lactate-threshold / maximal lactate steady-state region, and reported an effect of that distribution on performance (Esteve-Lanao, Foster, Seiler, and Lucia, 2007). Neal and colleagues randomised trained cyclists to six weeks of a polarized versus a threshold-heavy distribution and reported greater physiological and performance adaptations in the polarized arm (Neal et al., 2013). Stöggl and Sperlich compared polarized, threshold, HIIT, and high-volume arms and reported a greater impact of polarized training on key endurance variables (Stöggl and Sperlich, 2014).

These are strongly supported as evidence that a week made mostly of threshold work is often a worse idea than a week with a large easy fraction plus some truly hard work, in trained endurance athletes over weeks to a few months. They are not a demonstration that threshold sessions should vanish. They are small. They are not recreational-beginner trials. Muñoz, Seiler, Bautista, España, Larumbe, and Esteve-Lanao asked whether polarized training improved performance in recreational runners and did not give the elite observational literature a free pass into that population (Muñoz et al., 2014).

11 HIIT

High-intensity interval training, in MacInnis and Gibala’s classification, is repeated near-maximal bouts with recovery (MacInnis and Gibala, 2017). Helgerud et al. (2007) is the canonical work-matched comparison in moderately trained men: aerobic intervals improved VO₂max more than moderate training. Midgley, McNaughton, and Wilkinson asked whether there is an optimal intensity for enhancing VO₂max in distance runners and treated the question as unresolved rather than as a slogan (Midgley, McNaughton, and Wilkinson, 2006). Bacon, Carter, Ogle, and Joyner meta-analysed VO₂max trainability and high-intensity interval training in humans (Bacon, Carter, Ogle, and Joyner, 2013). Milanović, Sporiš, and Weston meta-analysed HIT versus continuous endurance training for VO₂max improvements in controlled trials (Milanović, Sporiš, and Weston, 2015). Weston, Wisløff, and Coombes meta-analysed HIIT in people with lifestyle-induced cardiometabolic disease (Weston, Wisløff, and Coombes, 2014).

Strongly supported: in previously untrained or moderately trained adults, and in some clinical populations, HIIT raises VO₂max at least as much as moderate continuous training, often with less time. Not supported: that HIIT is universally superior for race performance, mitochondrial every-endpoint, fat loss, or adherence. Not supported: that Tabata’s protocol (Tabata et al., 1996) is a general prescription. Tabata compared six weeks of moderate-intensity endurance (70% VO₂max, 60 min·d⁻¹, 5 d·wk⁻¹) with a high-intensity intermittent protocol on a cycle ergometer in a small sample; anaerobic capacity and VO₂max responded differently across the two experiments. The internet protocol that borrowed the name is not the paper.

12 Sprint-interval training

Sprint-interval training is supramaximal: repeated all-out efforts, classically Wingate-style, with long recoveries and a very small total volume of intense work. Burgomaster, Hughes, Heigenhauser, Bradwell, and Gibala reported that six sessions of SIT increased muscle oxidative potential (citrate synthase) and cycle endurance capacity (Burgomaster, Hughes, Heigenhauser, Bradwell, and Gibala, 2005). Gibala, Little, van Essen, Wilkin, Burgomaster, Safdar, and colleagues compared low-volume SIT with high-volume endurance training and reported similar initial metabolic and performance adaptations (Gibala et al., 2006). Burgomaster, Heigenhauser, and Gibala examined carbohydrate metabolism after short-term SIT (Burgomaster, Heigenhauser, and Gibala, 2006).

Strongly supported: a surprisingly small number of all-out sessions can move oxidative enzymes and short-term endurance capacity in recreationally active people. Not supported: that SIT is a substitute for the weekly volume of an elite endurance athlete, or that it is kind to tendons, joints, or adherence. The time-efficiency is real. The mechanical and affective cost is the part the abstracts mention less.

13 Polarized versus pyramidal versus threshold

Polarized training, in the usage of Seiler and of Stöggl and Sperlich, is a distribution with most work in zone 1, little in zone 2 (threshold), and a meaningful fraction in zone 3. Pyramidal training is a distribution with most work in zone 1, a substantial middle, and the least in zone 3. Threshold training overloads the middle. Casado, González-Mohíno, González-Ravé, and Foster systematically reviewed periodization, methods, intensity distribution, and volume in highly trained and elite distance runners (Casado, González-Mohíno, González-Ravé, and Foster, 2022). That review is the right document for the claim that elite runners’ described practices are not a single polarized cartoon, and that pyramidal distributions appear often in the observational record.

The adversarial point, which this article will not dodge: polarized superiority is not a law of physiology. Neal et al. (2013) and Stöggl and Sperlich (2014) found polarized advantages in specific trials. Esteve-Lanao et al. (2007) found that reducing the threshold fraction helped. Seiler (2010) described polarized-looking elite weeks. Casado et al. (2022) show that highly trained runners also live in pyramidal distributions. Muñoz et al. (2014) tested recreational runners. The honest synthesis is strongly supported for “do not make most of the week moderately hard,” and only emerging for “polarized beats pyramidal in all endurance sports.” Elite observational practice is not causal evidence. Athletes who can tolerate twenty hours a week are not a random sample of the method.

14 Periodization, specificity, recovery, taper, overreaching

Periodization is the planned variation of volume and intensity across a season. Casado et al. (2022) is the elite-distance-running map. Specificity is the older and less fashionable rule: the organism adapts to the task it is given. A rower who only cycles will have a cycling mitochondria story and a rowing race problem (Volianitis and Secher, 2009). Guellich, Seiler, and Emrich described training methods and intensity distribution of young world-class rowers: another sport, another observational distribution (Guellich, Seiler, and Emrich, 2009).

Tapering is a reduction in training load before a competition, intended to reduce fatigue more than fitness. Mujika and Padilla reviewed the scientific bases for precompetition tapering strategies (Mujika and Padilla, 2003). Bosquet, Montpetit, Arvisais, and Mujika meta-analysed effects of tapering on performance (Bosquet, Montpetit, Arvisais, and Mujika, 2007). Strongly supported: a taper that reduces volume while maintaining some intensity is associated, in the aggregated training studies, with a small performance gain. Not a prescription: any particular percentage or number of days for any person.

Overreaching is a short-term performance decrement from which recovery is expected; overtraining syndrome is a more prolonged maladaptation. Meeusen and colleagues issued a joint ECSS–ACSM consensus on prevention, diagnosis, and treatment of the overtraining syndrome (Meeusen et al., 2013). Kreher and Schwartz wrote a practical clinical guide (Kreher and Schwartz, 2012). Established: the syndrome is real, under-diagnosed, and not the same as low energy availability, anaemia, or primary depression, though it shares symptoms with all three. Not established: a single biomarker that diagnoses it. Heart-rate variability, resting HR, and mood scales are monitoring tools, not oracles.

Coyle et al. (1984) remains the detrain clock: stop the work, lose the adaptation, on a time course that is faster for plasma volume than for mitochondrial enzymes. Recovery is not a spa. It is the interval in which the stimulus becomes an adaptation rather than an injury.

15 Running

Distance running is the sport in which economy, not only VO₂max, has the longest quantitative tradition (Conley and Krahenbuhl, 1980; Saunders, Pyne, Telford, and Hawley, 2004; Barnes and Kilding, 2015). Esteve-Lanao et al. (2005, 2007) and Casado et al. (2022) are the training-distribution literature in runners. Helgerud et al. (2007) used running intervals in moderately trained men. The mechanical load is high: each kilometre is a collision. That is why running volume has an injury epidemiology that cycling volume does not, and why copying a cyclist’s twenty-hour week onto a novice runner is not a physiology experiment but an orthopaedic one.

Billat’s vVO₂max intervals (Billat et al., 1999) and Midgley’s question about optimal intensity for VO₂max (Midgley, McNaughton, and Wilkinson, 2006) are running-weighted programming papers. They still do not specify a session for a reader.

16 Cycling

Cycling substitutes an external power meter for the ground. Critical power is native here (Poole et al., 2016; Vanhatalo, Jones, and Burnley, 2011; Jones et al., 2017). Lucía, Hoyos, and Chicharro (2001) described professional road cycling: extreme duration, drafting, and a need for repeated high-intensity efforts inside an aerobic day. Neal et al. (2013) polarised cyclists. Achten, Venables, and Jeukendrup (2003) found lower fat-oxidation rates in cycling than running at the same relative intensities, a modality difference that Fat_max marketing rarely survives.

The bicycle reduces impact and does not reduce heat, saddle, or cervical load. A cycling VO₂peak underestimates a runner’s running VO₂max and overestimates a non-cyclist’s cycling performance. Specificity is not a slogan in this sport; it is the crank.

17 Swimming

Front-crawl propulsion wastes a large fraction of mechanical work on moving water backward rather than the body forward (Toussaint and Beek, 1992; Toussaint et al., 1988). Economy is therefore technique plus physiology. Heart rate at a given metabolic rate is lower in water; land zones do not import cleanly. Breathing pattern couples to stroke, so ventilatory thresholds are harder to read. The horizontal, hydrostatic environment changes venous return and the temperature clamp. A swimming endurance programme that is only “more aerobic metres” without a technique constraint is a different training problem from running’s collision-limited volume.

18 Rowing

Volianitis and Secher described rowing as an unusual whole-body challenge: large muscle mass, constrained posture, and a force profile that is neither cycling nor running (Volianitis and Secher, 2009). Guellich, Seiler, and Emrich (2009) documented intensity distribution in young world-class rowers. Stroke volume and blood-pressure load during the catch are part of the cardiac story in a way jogging is not. Indoor rowing is not on-water rowing; the ergometer removes the boat’s instability and not the physiological demand.

19 Triathlon

Triathlon is sequential specificity with transitions. Bentley, Millet, Vleck, and McNaughton reviewed physiological analysis and training implications of contemporary triathlon (Bentley, Millet, Vleck, and McNaughton, 2002). Millet, Di Prampero, and colleagues compared elite short- and long-distance triathletes (Millet, Dreano, and Bentley, 2003). The swim is technique-limited, the bike is power- and aerodynamics-limited, the run is a glycogen- and damage-limited task performed on already-used legs. A triathlon VO₂max on a treadmill is one number. It does not tell you which leg of the race is the limit. Polarized-versus-threshold arguments have to be restated per discipline; a polarized bike week plus a threshold-heavy run week is a combined stress, not three separate studies.

20 Cross-country skiing and other modes

Sandbakk and Holmberg reviewed physiological capacity and training routines of elite cross-country skiers, including the double-poling upper-body demand that running physiology does not contain (Sandbakk and Holmberg, 2017). Sandbakk, Holmberg, Leirdal, and Ettema described world-class sprint skiers (Sandbakk, Holmberg, Leirdal, and Ettema, 2011). Sandbakk and Holmberg later reappraised success factors for Olympic cross-country skiing (Sandbakk and Holmberg, 2014). Intensity distributions in skiers are among the sources of Seiler’s observational picture (Seiler, 2010). Upper-body VO₂, technique in changing snow, and the sprint-versus-distance split make “endurance” in this sport a family of events.

Other modes (speed skating, race walking, canoe, adventure racing, ultra-endurance) inherit the Fick map and not the running-economy literature. Ultra-endurance adds gut, sleep, and energy-availability problems that are sports-nutrition problems first (Thomas, Erdman, and Burke, 2016). They are not solved by a larger VO₂max.

21 Heat

Nybo reviewed hyperthermia and fatigue: high core temperature degrades performance by cardiovascular, central, and peripheral routes that are not a single mechanism (Nybo, 2008). Lorenzo, Halliwill, Sawka, and Minson reported that heat acclimation improved exercise performance in a controlled human study (Lorenzo, Halliwill, Sawka, and Minson, 2010). Lundby, Svendsen, Uthausen, and colleagues compared training in thermal clothing with training in hot ambient conditions and reported equal effectiveness for the endpoints they measured (Lundby et al., 2021). Nybo and Lundby later set altitude camps and heat-acclimation training side by side as preparation methods (Nybo and Lundby, 2024). Sports Dietitians Australia’s position statement on nutrition for exercise in hot environments is the fuel-and-fluid companion, not a training method (McCubbin et al., 2020).

Strongly supported: repeated heat exposure expands plasma volume and can improve subsequent performance in the heat, and sometimes in cool conditions, in trained people. Not a protocol: any sauna duration, any “live hot train cool” recipe, any amateur haemoglobin-mass claim copied from altitude. Heat is a stimulus. It is also a medical risk. This document describes the papers. It does not prescribe exposure.

22 Altitude

Levine and Stray-Gundersen tested “living high-training low”: moderate-altitude acclimatization with low-altitude training, and reported improved sea-level performance in endurance athletes (Levine and Stray-Gundersen, 1997). Chapman, Stray-Gundersen, and Levine then documented individual variation in the response (Chapman, Stray-Gundersen, and Levine, 1998). Stray-Gundersen, Chapman, and Levine reported further sea-level performance data in men and women (Stray-Gundersen, Chapman, and Levine, 2001). Strongly supported: hypoxia can raise haemoglobin mass in responders, and training quality is easier to preserve if the hard sessions occur at lower altitude. Established as a caveat: non-responders exist; the method is not a haemoglobin transfusion; iron status, illness, and training-camp quality confound field results. Live-high train-high can reduce interval quality. Tents and masks are not the 1997 trial.

23 Age and sex

Tanaka and Seals reviewed endurance exercise performance in Masters athletes: VO₂max declines with age even in the highly trained, with reductions in maximal cardiac output and, variably, in oxygen extraction; performance declines are event- and sex-dependent (Tanaka and Seals, 2008). That is strongly supported physiology. It is not a reason to treat older athletes as untrainable.

Joyner reviewed physiological limits to endurance performance as influenced by sex: historical participation, haemoglobin, body composition, and the shrinking performance gap in ultra-distance as contexts, not as slogans (Joyner, 2017). Ansdell, Thomas, Hicks, Hunter, Howatson, and Hampson reviewed physiological sex differences affecting the acute and chronic integrative response to exercise (Ansdell et al., 2020). Strongly supported: mean differences exist in haemoglobin concentration, body fat, and some fatigue-resistance measures. Not supported: that women should not do HIIT, or that men do not need easy volume. Most training trials remain male-weighted. Helgerud et al. (2007) used men. Extrapolation is an inference and is labelled as one.

24 Recreational versus elite

Seiler’s observational athletes train ten to twenty hours a week (Seiler, 2010). Helgerud’s randomised men were moderately trained (Helgerud et al., 2007). Burgomaster’s SIT subjects were recreationally active (Burgomaster et al., 2005). Muñoz et al. (2014) asked the recreational-runner question directly. Bacon et al. (2013) and Milanović et al. (2015) pooled mixed samples for VO₂max.

The same word, “endurance training,” names different dose ranges. An elite polarized week is a volume story with a few hard sessions. A recreational HIIT week is often an intensity story with almost no volume. Copying the elite percentage onto three hours a week produces neither the elite mitochondria nor the elite injury risk; it produces a different, under-studied week. This is established as a generalizability problem and routinely ignored in social-media programming.

25 Endurance nutrition interactions

Thomas, Erdman, and Burke’s joint Academy / Dietitians of Canada / ACSM position on nutrition and athletic performance is the field map: periodized carbohydrate, sufficient energy availability, hydration, and a short list of evidence-graded supplements (Thomas, Erdman, and Burke, 2016). Bergström’s biopsy era and Coyle’s feeding study are why carbohydrate during prolonged hard work is a performance question (Bergström et al., 1967; Coyle et al., 1986). McCubbin et al. (2020) restate fluid and sodium as heat problems.

This article does not rerun the sports-nutrition title. The interaction that belongs here: training in low carbohydrate availability can amplify some mitochondrial signals and can also reduce interval quality and increase illness risk. The papers that moved PGC-1α with twice-daily low-glycogen sessions are training-nutrition hybrids, not a licence to race depleted. Energy availability is a health constraint (the RED-S literature lives in the sports-nutrition sibling). A polarized week stacked on low energy availability is not polarized training. It is a deficit.

26 Cardiorespiratory fitness, activity volume, and mortality

Blair, Kohl, Paffenbarger, Clark, Cooper, and Gibbons followed healthy men and women and reported that low physical fitness predicted all-cause mortality (Blair et al., 1989). Myers, Prakash, Froelicher, Do, Partington, and Atwood reported that exercise capacity predicted mortality among men referred for exercise testing (Myers et al., 2002). Kodama and colleagues meta-analysed cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women (Kodama et al., 2009). Mandsager, Harb, Cremer, Phelan, Nissen, and Jaber reported a graded association of CRF with long-term mortality among adults undergoing exercise treadmill testing, including at the extreme high end of the measured distribution in that clinical cohort (Mandsager et al., 2018).

Lee and colleagues estimated the global non-communicable-disease burden attributable to physical inactivity (Lee et al., 2012). Wen, Wai, Tsai, and colleagues reported mortality associations across a range of leisure-time activity volumes in a large Taiwanese cohort, including volumes below the then-standard 150 minutes per week (Wen et al., 2011). Arem, Moore, Patel, and colleagues pooled leisure-time physical activity and mortality in a detailed dose–response analysis (Arem et al., 2015). Ekelund, Steene-Johannessen, Brown, and colleagues asked whether physical activity attenuates the association of sitting time with mortality in a harmonised meta-analysis of more than one million men and women (Ekelund et al., 2016).

The Physical Activity Guidelines for Americans (Piercy et al., 2018) and the WHO 2020 guidelines on physical activity and sedentary behaviour (Bull et al., 2020) translate that observational and interventional mixed literature into public-health recommendations. Garber and colleagues’ ACSM position stand on quantity and quality of exercise for apparently healthy adults is the professional-practice sibling (Garber et al., 2011).

Established: higher CRF and higher habitual activity are associated with lower all-cause and cardiovascular mortality in large observational datasets, and some of the association is visible below formal guideline volumes. Not established: that raising VO₂max by any particular interval protocol causes the mortality reduction seen in CRF cohorts. CRF is a capacity with genetic, disease, and behavioural determinants. Treating it as a pure training readout is how an observational hazard ratio becomes a HIIT advertisement. Not established: a toxicity threshold at which endurance training shortens life in healthy people. The high-volume question is the next section, not a cancelled health benefit.

27 The high-volume heart, and the U-shaped claim

O’Keefe, Patil, Lavie, Magalski, Vogel, and McCullough reviewed potential adverse cardiovascular effects from excessive endurance exercise (O’Keefe et al., 2012). Merghani, Malhotra, and Sharma described a U-shaped relationship between exercise and cardiac morbidity (Merghani, Malhotra, and Sharma, 2016). La Gerche et al. (2012) supplied an imaging mechanism at the right ventricle. These papers are emerging-to-plausible as a description of a subset of very high-volume athletes, particularly regarding atrial fibrillation, myocardial fibrosis, and right-ventricular remodelling. They are not a randomised demonstration that marathon running is net harmful, and they are a weak basis for frightening recreational runners who train three hours a week.

Mandsager et al. (2018), in a referral treadmill cohort, did not find a mortality penalty at the highest CRF octile they reported. That is not a clean refutation of O’Keefe: the populations differ (clinical exercise-test patients versus ultra-endurance case series). The honest position is conflict, not cancellation. Recreational endurance training as practised by the people who generate most of the Blair/Kodama/Wen signal is on the protective side of every U that has been drawn. Elite ultra-volume is a different exposure and remains incompletely measured for hard endpoints.

28 Adversarial findings, stated as findings

Polarized superiority. Supported as “avoid a threshold-heavy week” in trained endurance athletes over short interventions (Esteve-Lanao et al., 2007; Neal et al., 2013; Stöggl and Sperlich, 2014). Not supported as a universal winner against pyramidal distributions in elite runners (Casado et al., 2022). Elite observational practice is not a randomised cause.

Zone-2 marketing. Zone 2 as a heart-rate band near the first threshold is a plausible easy-training proxy. Zone 2 as a unique mitochondrial, fat-loss, or longevity intensity is not a result in the papers cited here. Fat_max is a real intensity (Achten, Gleeson, and Jeukendrup, 2002) and is not a 24-hour fat-loss zone (Romijn et al., 1993).

Exact lactate thresholds. Faude, Kindermann, and Meyer (2009) is the review that should have ended the one-number cult and did not. Ventilatory, lactate, and critical-power constructs are related and not interchangeable (Beaver, Wasserman, and Whipp, 1986; Poole et al., 2016; Jones et al., 2019).

HIIT versus moderate continuous training. HIIT is at least as good for VO₂max in many untrained and clinical samples, often in less time (Helgerud et al., 2007; Bacon et al., 2013; Milanović et al., 2015; Weston, Wisløff, and Coombes, 2014). Race performance, durability, injury, and adherence are different endpoints. Tabata et al. (1996) is a small cycle study, not a brand.

Fat-burning zone. Maximal during-exercise fat oxidation occurs at moderate intensity and is higher in running than cycling (Achten et al., 2003; Romijn et al., 1993). Body-fat change is an energy-balance and endocrine problem. Training in that zone does not uniquely empty adipose tissue.

Mitochondrial claims. Endurance training raises mitochondrial content and respiratory capacity (Holloszy and Coyle, 1984). Intensity can raise respiratory function without a matching content change (Granata et al., 2016, 2018). Interval training is a potent stimulus (MacInnis et al., 2017). None of that implies a single best zone, a supplement that replaces training, or a count of mitochondria visible on a consumer report.

VO₂max as universal surrogate. Joyner and Coyle (2008) gave it one seat of three. Conley and Krahenbuhl (1980) showed economy discriminating when VO₂max did not. Critical power may describe the sustainable fraction better than VO₂max does (Poole et al., 2016). Mortality associations attach to CRF in mixed populations (Kodama et al., 2009), which is not the same as attaching them to a 4 × 4-minute protocol.

Elite practice as causal evidence. Seiler (2010) and Casado et al. (2022) describe what successful athletes did. They do not assign those athletes at random to a worse distribution. Survivorship, talent, and resources are confounded with method.

29 What can be said, and what cannot

Endurance training, performed as a large amount of mostly easy locomotor work plus some genuinely hard work, changes the heart, the blood, the capillaries, and the mitochondria in directions that raise sustainable speed or power (Holloszy and Coyle, 1984; Bassett and Howley, 2000; Joyner and Coyle, 2008; Seiler, 2010). The same capacity, measured as CRF, tracks lower mortality in observational cohorts that are mostly not elite (Blair et al., 1989; Kodama et al., 2009; Wen et al., 2011; Arem et al., 2015). Programming arguments after that are arguments about how to spend a finite recovery budget: polarized versus pyramidal, HIIT versus SIT versus long slow distance, heat and altitude as adjuncts. Those arguments are real, smaller than the marketing around them, and sport-specific.

This document recommends no plan. The ACSM and WHO documents exist for public-health counselling by qualified professionals (Garber et al., 2011; Piercy et al., 2018; Bull et al., 2020). They are not this article, and this article is not them.

30 Training-model matrix

31 Interval-protocol matrix

32 Adaptation time-course and elite-versus-recreational evidence

33 Evidence handling

Findings are labelled by study type in the sentence that reports them. Randomised training trials, meta-analyses of those trials, observational elite logs, imaging studies, biopsy time-courses, and public-health cohorts are different kinds of claim. Animal and in-vitro molecular maps (Egan and Zierath, 2013) are never phrased as human outcomes. Where evidence conflicts (polarized versus pyramidal; high-volume cardiac risk versus CRF-mortality gradients), both sides are given. Recalled PMIDs that pointed at unrelated papers were discarded during harvest; every identifier in the numbered list was title-checked against MEDLINE.

Project 06 was queried read-only. Title-level endurance holdings in the integration database are thin; full-text search mentions exist; the live catalog contributed discovery titles (including Bacon’s VO₂max-trainability meta-analysis) that were then verified on PubMed. Project 07 news was not used as evidence. Peptide-article scientific content was not used as endurance evidence.

34 Provenance

Compiled 20 August 2026 as SBL-41/SP-ENDURANCE-TRAINING, Register A, parent SBL-41 Science Monograph Series. Identity: South Beach Longevity Science Monograph, not Radix Peptides, not Project 05. Presentation follows the canonical house style with official South Beach Longevity marks. Constraint line verbatim. Dual editions through pipeline stages 06, 07, and 07c.

35 References

Generated from verified NCBI records. Replaced in the build by the numbered list.