How Repetition Can Become Overuse

Everyone on an erg repeats the same movement for the length of a piece, and the machine never checks whether you are ready for it. So rowing injuries rarely arrive as accidents. They arrive as an ordinary training day: the back that is a bit stiff on Tuesday, the tendon that is a bit tender after Sunday's long piece. Nothing dramatic happened. The work simply arrived faster than the tissue could rebuild for it.

That gap is the whole story, and it is worth getting right, because the rule most coaches reach for to manage it does not survive a randomised trial.

The context

Repetition is the easy explanation and the wrong one. The word overuse invites a picture of a body worn down by a number of strokes, and the injury data does not support that picture. What the data supports is about change: how fast demand arrives, and what the tissue has already been rebuilt for.

Rowing's own data is the most direct evidence there is. Of the studies that reported a full year of follow-up, low back pain incidence in rowers ran from 31.8% to 51% of the cohort, and the two risk factors that held up across them were a history of lumbar spine injury and the volume of ergometer training [1] Wilson 2014. Stroke count is not on that list. Volume is, and so is what you have already hurt.

The wider evidence is weaker than the coaching folklore built on it. A systematic review that asked the plain question — does a sudden change in training load raise injury risk? — found only four studies that met its criteria, and concluded that very limited evidence supports the association [2] Damsted 2018. Four studies is not a foundation for a precise rule.

The prior-injury finding is the most useful one for anyone coming back to the erg. If your back has been hurt before, the base you build first is not a formality, because it is the group you are most likely to join again.

What survives is more specific, and more useful. A prospective study followed 95 endurance athletes, rowers among them, for 52 weeks of weekly training and pain reporting. The load pattern associated with fewer new pain and injury episodes was avoiding sharp spikes in acute load while keeping a moderate-to-high chronic base, and there was a lag between the load change and the symptoms that followed it [3] Johnston 2019. A spike on top of a base you have already built is the shape of the thing to manage. The lag is the part that catches people out: a spike on Monday can produce symptoms on Wednesday, by which point the cause is a couple of days behind you and the link is easy to miss.

The core idea

If repetition is not the mechanism, what is? Demand arriving faster than a tissue can remodel for it. Tendons, ligaments and bone all rebuild on their own clocks, and the stroke asks all of them at once: the lumbar spine in flexion at the catch, the achilles and the soleus carrying the load of the drive while the handle comes back.

Two details make that concrete in rowing. The first is that technique is not a constant. Fatigue produces more lumbar spine flexion at the catch, in rowers with and without a history of low back pain, and more of it on the ergometer than on the water [4] Nugent 2021. The part of the session where you are most likely to load a spine badly is the part where your technique is at its worst. The second is that the lower back is the most commonly reported injury site in rowers [1] Wilson 2014, which is why it gets most of the attention here.

The second thing worth knowing is what does not work.

The most popular tool for deciding whether a week of training is safe is the acute:chronic workload ratio: this week's load divided by your average. It is quoted in every coaching course and built into most training apps.

A methodological critique laid out why it should never have been used that way. No study had estimated a causal effect, the ratio's statistical construction is incoherent, and there is no evidence supporting its use in load-management systems at all [5] Impellizzeri 2020. Then the trial came. Thirty-four elite youth football teams, 482 players, a full season with coaches planning every session to published load-management principles: no reduction in health problems. The difference between the groups was 1.8 percentage points, with a confidence interval running from -4.1 to 7.7 [6] Dalen-Lorentsen 2021.

That is not a rule that has not been tested hard enough. That is a rule that has been tested and did not work.

The how-to

Nobody can promise you will not get hurt. The useful question is how to progress load without outrunning the slowest tissue in your plan. Six things have evidence behind them.

Progress volume before you progress intensity. The body tolerates more work at a familiar intensity than a new intensity at the same volume. When you add, add minutes first, and hold intensity steady for a block.

Add strength work, and add more of it. A meta-analysis of six randomised trials, 7738 participants and 177 injuries, found a relative risk of 0.338 for strength training as primary prevention, and a 10% increase in strength training volume reduced injury risk by more than four percentage points [7] Lauersen 2018. It is the strongest result on this list, and the dose is the point: a token session once a week is not the intervention that was tested. It is also volume that carried the benefit, not effort, so the work you accumulate across the week is what counts.

Treat sleep as a load variable. A meta-analysis of 27 studies found that acute sleep deprivation significantly impaired overall athletic performance, with the largest effect on high-intensity intermittent exercise and a clear effect on aerobic endurance [8] Gong 2024. A short-slept night is not a neutral day. It is a session with less in the tank than the plan asked for.

Make the plan one you will actually do. A meta-analysis of 35 randomised trials looked at what happens when athletes do not stick with their injury-prevention programme. The high-adherence group had a 47% lower injury rate than control; in the lowest-adherence group the reduction was not significant [9] Viiala 2026. The best programme for you is the one that survives a bad week, not the one with the best evidence behind it.

Do not stop training entirely when something hurts. The consensus for adult rowers is specific: manage low back pain by unloading aggravating activity, controlling pain and using exercise therapy, and maintain fitness with load management and a progression back to full training [10] Wilson 2021. The public clinical guidance agrees on the general shape — stay active, keep going with your daily activities, and do not stay in bed for long periods [11] NHS 2007. Cut what provokes it, keep the rest, and rebuild.

Know the symptoms that mean stop. The same NHS guidance lists what should not be treated as a training-day problem: back pain that does not improve after a few weeks of managing it at home, pain that is stopping your day-to-day activities, unexplained weight loss, a lump or swelling in your back, or pain that is worse at night. Urgent help is needed if you feel hot, cold or shivery, or if the pain is severe and getting worse quickly [11] NHS 2007.

The trade-offs

Four of them, and one thing the evidence genuinely cannot do.

The ratio's failure is not a licence to push. It is tempting to read the death of the acute:chronic workload ratio as permission to ignore load. It is not. The reviews that took the ratio apart found the underlying load-injury evidence thin in both directions: a sudden jump in weekly volume is worth watching, and the studies that suggest it are weak [2] Damsted 2018. The common rule of thumb that weekly volume should never rise by more than 10% has no support either, because the same review compared a 10% with a 24% average weekly increase and found no difference between them. What you have is a soft signal to manage sensibly, not a rule you have earned the right to ignore.

Strength work costs time, and the time is the point. The best-supported intervention here is also the one most rowers skip, because two or three lifting sessions a week are time taken off the water. The trade-off is real, and it is the trade-off the trials made, which is part of why the result is strong.

Adherence beats optimality, which is a bit deflating. A programme you complete imperfectly protects you; the same programme you abandon does not [9] Viiala 2026. The consequence is that the binding constraint on a training plan is not what is theoretically best. It is what you will still be doing in March.

The evidence cannot tell you when you personally will be hurt. Every study here works on groups, and the group-level associations are weak enough that the reviews call them inconsistent [5] Impellizzeri 2020. Individual risk prediction is a harder problem, and nothing in this article is a prediction about you.

What to do tomorrow

Six things, in order.

1. Write down your last four weeks of sessions: the minutes, the distance, the average split. You cannot manage a spike you cannot see.
2. When you add, add the minutes before the intensity, and add one thing at a time.
3. Put two strength sessions in the week and write them into the plan as sessions, not intentions.
4. Protect the sleep before the hardest session, not after it.
5. If something has hurt for a few days, cut the volume that provokes it and keep everything else.
6. If it is not settling, or it comes with any of the warning signs above, get it seen.

The ratio will not tell you. Your own history, your own logbook and a clinician who has seen you will.