Why Short Intervals Repeat Across Plans: Same Shape, Different Session

Your logbook will tell you that you did the same session three weeks running. The splits will tell you something else.

A set of short intervals is not one session repeated. It is a container, and a container only says what it holds once somebody says what to put in it. Two plans can both write four by five hundred and mean two entirely different pieces of work: one is a test of how fast you are, the other is a test of how long you can hold something steady. Nothing in the shape distinguishes them, and the shape is the only part most plans agree on.

The shape is not the session

Prescribing an interval session means manipulating up to nine variables: work interval intensity, work interval duration, relief interval intensity, relief interval duration, exercise modality, number of repetitions, number of series, between-series recovery duration, and between-series recovery intensity [1] Buchheit 2013. Those are the terms that carry the training. The seconds of the work interval are one of them.

Which means that a plan saying "4 x 500" and a plan saying "4 x 500 at 1:35 with 1:30 rest" are not the same plan, and neither is "4 x 500 as hard as you can, full recovery". The first has not specified the session at all. It has specified a shape and left the contents to you, which is how two people write the same line and do completely different work.

The same review makes the point that the number usually quoted as the goal, time at or above 90 per cent of maximal oxygen uptake, is a summary of several of those variables rather than one of them [1] Buchheit 2013. You do not drift into that zone. You get there by choosing a pace, holding it for a certain time, and choosing a rest short enough that you are still near your limit when the next repetition starts.

What you hold constant is the session

Here is the part most plans leave unsaid. A set of repetitions can be built to hold a split steady, or a power steady, or a rate steady, or an effort steady. Those are four different interventions that share a shape, and they do not train the same thing.

The rowing literature is unusually clear about what actually predicts a 2,000 m ergometer result. In 41 championship finalists, the strongest correlations with 2,000 m speed were power at maximal oxygen uptake, maximal power and maximal force (r = 0.95), with maximal oxygen uptake (r = 0.88) and oxygen consumption at the lactate threshold (r = 0.87) also strongly related [6] Ingham 2002. A regression model built from power at maximal oxygen uptake, oxygen consumption at the lactate threshold, power at 4 mmol per litre blood lactate and maximal power explained 98 per cent of the variance in 2,000 m speed, and it held up in a separate group of 18 elite rowers to within 1.5 to 6.9 seconds [6] Ingham 2002.

Read that list as an answer to the question. If you are chasing a 2,000 m time, the levers that matter are your power at maximum, at your threshold and at your peak. The split on any given repetition is downstream of all three.

The split is a symptom. In every heavyweight race at the 2000 Olympic Games and the World Championships of 2001 and 2002, plus the top 170 finishers at the British Indoor Rowing Championships, 948 of 1,612 race profiles met the inclusion criteria, and there was no difference in pacing profile between winners and losers [7] Garland 2005. Almost everybody rowed the first 500 m at 103.3 per cent of their average race speed, then settled to 99.0, 98.3 and 99.7 per cent [7] Garland 2005. The fast start is not a different strategy. It is what everybody does.

What separates one attempt from the next is the shape of the curve, not the number on the monitor. Fourteen well-trained male rowers did three 2,000 m ergometer tests three to seven days apart. The finish times did not change, with a typical error of 2.4 per cent, but the pacing did: the first trial started hard and faded, the second and third started conservatively and finished with a spurt in the final 500 m [8] Gee 2013. The authors draw the obvious lesson that a habituation trial is needed before any of it can be interpreted [8] Gee 2013.

So if you write your intervals down as splits, you will record what happened to change. If you write down the profile, where you went hardest and what was left at the end, you will record what was actually different about the session.

The rest is not a pause

Short recoveries look like a compromise. They are not. They are the dose.

Pulmonary oxygen uptake does not jump when you do. During moderate constant work it rises to a steady state, but at work rates that produce a lactic acidosis an additional slow phase is superimposed and it prolongs the time to steady state over exactly the range in which blood lactate and hydrogen ion concentration are stabilising or falling [9] Whipp 1994. At severe work rates a steady state is unattainable altogether [9] Whipp 1994.

So through the first minute or so of a hard repetition you are running on an oxygen deficit that is still being filled. If the next repetition starts before the system has come back, you start deeper in deficit than you did the first time. That is what a short rest does. It keeps the session aerobic-limited in a way a long rest does not, which is why the recovery, rather than the work, sets how much of the session you can actually take.

The other half of it is that the three energy systems do not queue up. They are closely integrated, energy comes from all of them during almost all exercise, and the point at which aerobic and anaerobic contributions become equal falls between one and two minutes, most probably around 75 seconds [2] Gastin 2001. The same review names the two misconceptions this removes: that the energy systems respond to intense exercise in an almost sequential manner, and that the aerobic system plays little role over short durations [2] Gastin 2001.

Put those together and the arithmetic of a short-interval session becomes clear. Under about 75 seconds you are largely working on the deficit and the fast-twitch recruitment that comes with it. Over it, the aerobic system is taking over more of the bill even though the piece still feels anaerobic, because the aerobic contribution keeps rising and the acid load keeps clearing. Every interval session lives on that boundary, and the boundary moves with the athlete.

The aerobic share, and the distance it belongs to

Which is worth saying plainly, because a 500 m piece in a boat or on the erg is usually called an anaerobic session. The share depends on the distance, so the figure has to keep it attached.

Over the standard 2,000 m, 70 to 75 per cent of the energy for men comes from aerobiosis and 25 to 30 per cent from anaerobic sources; women compete over 1,000 m and draw 60 to 65 per cent aerobically [4] Hagerman 1984. Elite oarsmen in that review carried maximal oxygen uptakes of 6.1 plus or minus 0.6 litres a minute, and oxygen debts of between 10 and 20 litres across a race [4] Hagerman 1984. A 500 m piece is different work, and the review gives no split for it.

The mechanism is rowing-specific. The slow, deliberate leg action of the stroke involves primarily slow-twitch muscle fibre, and the metabolic cost of the stroke is paid back through blood flow that competes with the demands of the exercise itself, so intense rowing challenges pulmonary function and cerebral blood flow as well [5] Volianitis 2020.

For a plan, the consequence is narrower than the race figure sounds. Aerobic supply dominates the total energy requirement after about 75 seconds of near-maximal effort and has the greatest potential for improvement with training [3] Laursen 2010. So whether a session is aerobic is not a property of rowing. It is a function of how long the work interval runs, how hard you hold it, and how much recovery you allow between repetitions - three of the variables this article opened with.

What the review claims, and what the rowing trial tested

For people who are already trained, adding submaximal volume does not appear to further improve endurance performance or the associated physiological variables such as peak oxygen uptake [10] Laursen 2002. The same review goes further and argues that for an already-trained athlete, improvements in endurance performance can be achieved only through high-intensity interval training, with an increase in muscle buffering capacity as one candidate mechanism [10] Laursen 2002.

The rowing trial on this page did not test that claim. It is worth being precise about what it tested. Eighteen experienced rowers trained for twelve weeks on either low-intensity work or a mixture of low and high. Both programmes improved 2,000 m performance and peak oxygen uptake by the same magnitude, and the low-intensity group attenuated its blood lactate response to a given exercise intensity more than the mixed group did [11] Ingham 2008. That compares two programmes of different make-up: a programme built on low-intensity work was not worse than a mixed one. It does not show that adding low-intensity volume on top of what you already do is ineffective: no group in the trial was given extra easy volume to add.

There is no half of the question the two agree on. The review argues from the trained athlete that intensity is where the gain is and submaximal volume is not; the trial shows a trained rower improving on low-intensity work alone, with no intensity required. The disagreement is about the premise underneath: whether an already-trained athlete needs intensity at all. If you are one of the already-trained athletes the review is about, its answer is the operational one: adding kilometres is not the lever, but how hard the short sessions run is.

More is not better, and by how much

The intuitive reason short intervals get longer is that more feels like more. The evidence says the return falls away quickly.

Pooling studies of sprint interval training, the improvement in maximal oxygen uptake averaged 7.8 per cent with a 90 per cent confidence limit of 4.0 per cent, and the maximum number of sprints in a session modified that effect by about 1.2 percentage points less for every two additional repetitions [12] Vollaard 2017. The gain was not smaller when there were fewer sprints, and may have been larger [12] Vollaard 2017. The analysis excluded patients, athletes and participants whose baseline maximal oxygen uptake was above 55 ml per kg per minute, so it describes the moderately trained rather than the elite [12] Vollaard 2017.

That last clause matters twice over. It is the reason the "three hard efforts is enough" rule of thumb is not in this paper, and also the reason the hard-edged claim from the previous section does not transfer uncritically to a rower who is already trained.

How to read an interval prescription in someone else's plan

Four questions, and you can answer all four from the text of the plan.

1. What is the target? A pace, a power, an effort, a rate, or nothing at all? If the plan gives you only seconds and repetitions, it has not told you what the session is [1] Buchheit 2013.
2. What is held constant, and what is allowed to fall? Holding a split steady and letting the power fall is a different session from holding the power steady and letting the split fall [6] Ingham 2002.
3. What is the rest doing? Is it long enough for oxygen uptake to come back, or short enough that you start the next repetition in deficit? Under heavy work the uptake itself is still climbing [9] Whipp 1994.
4. Who was it written for? Volume that does not appear to pay for a trained athlete is not free, and the evidence for that is specific to trained people [10] Laursen 2002.

And one worth asking last. Does the plan say what the piece is supposed to feel like? The stroke recruits the legs in a slow movement that involves primarily slow-twitch muscle fibre, while the same effort challenges pulmonary function and cerebral blood flow [5] Volianitis 2020. Knowing which system is paying most of the bill changes how you read the last repetition, and the last repetition is the one that decides whether the session did what it was supposed to do [3] Laursen 2010.

That is the whole of it. Short intervals recur because they are a good container, not because the shape does anything by itself. Read the target, the constant and the rest, and the shape will tell you what you are actually doing.