Nordic Hamstring Curls: The Best-Evidenced Injury Exercise You're Not Doing

If you've read anything about injury prevention in the past decade, you've probably seen the claim that one exercise halves hamstring injuries. The Nordic hamstring curl, kneeling with your ankles held and lowering your body towards the floor as slowly as you can, has become a fixture in football warm-ups and physiotherapy clinics. It's cheap, needs no equipment beyond a partner or something to hook your heels under, and takes a few minutes a week.

It's also a useful case study in how a headline number can outrun its evidence. The same trials have been analysed twice, with very different conclusions. This article covers what the Nordic curl is, what the key trials found, why a later reappraisal called the effect inconclusive, how much of it might apply to triathletes, and how to add it to a winter strength block without wrecking your next run.

What the Nordic hamstring curl is

You kneel on a pad with a partner holding your ankles, or your heels hooked under a bar or heavy bench. Keeping your hips straight, you lean forward and resist the fall with your hamstrings for as long as you can, then catch yourself with your hands and push back up. The lowering is the exercise.

It's an eccentric exercise: the hamstrings lengthen while producing force. That matters because the hamstrings do something similar when you run. At the end of each swing phase, just before your foot lands, they work hard to slow the lower leg as the knee straightens. A 2020 systematic review by Danielsson and colleagues in BMC Musculoskeletal Disorders pooled 26 studies of how hamstring injuries happen. It described two main types:

• Sprint-type injuries, most likely caused by excessive strain on the muscle during eccentric work in the late swing phase of running at high speed.
• Stretch-type injuries, which happen with extreme hip flexion and a straight knee, as in high kicks, splits or slide tackles.

The logic of the Nordic curl is straightforward: make the hamstrings stronger at lengthening under load, and they should be better able to handle that late-swing strain.

What it does to the muscle

Here the evidence is consistent. A 2020 meta-analysis by Cuthbert and colleagues in Sports Medicine pooled 13 studies of Nordic curl training lasting at least six weeks.

• Eccentric hamstring strength rose substantially, with very large within-study effect sizes for both high-volume (g = 2.12) and low-volume (g = 2.28) programmes.
• Fascicle length increased, with effect sizes of at least 2.58. Longer muscle fibres are thought to make a muscle more tolerant of being stretched under load.
• Low volume worked as well as high volume. The authors concluded that cutting the number of Nordic repetitions didn't appear to reduce the strength or architecture gains, and that lower volumes might improve compliance.

Heterogeneity between studies was moderate to high, and these are within-study changes, mostly in young athletes. But the direction is not in doubt: Nordic curls make hamstrings stronger eccentrically and change their structure. The disputed question is whether that translates into fewer injuries.

The trials that started it

Danish football, 2011

The trial most often quoted is a cluster-randomised study by Petersen and colleagues in the American Journal of Sports Medicine. It recruited 50 Danish professional and amateur men's teams, 942 players. The 461 players in intervention teams did a 10-week progressive programme of Nordic hamstring exercises, followed by a weekly session through the season. The 481 control players trained as usual.

• Acute hamstring injuries: 15 in the intervention group, 52 in the control group.
• Overall injury rate: 3.8 versus 13.1 per 100 player-seasons, a rate ratio of 0.293.
• New injuries: 3.1 versus 8.1 per 100 player-seasons, a rate ratio of 0.410.
• Recurrent injuries: 7.1 versus 45.8 per 100 player-seasons, a rate ratio of 0.137.

Inference: a rate ratio of 0.293 means roughly 70% fewer injuries, and the effect on players who had already had a hamstring injury was the largest of all.

Dutch amateur football, 2015

A second randomised trial by van der Horst and colleagues, in the same journal, randomised 40 Dutch amateur men's teams, 579 players with a mean age of 24.5. The intervention teams did 25 Nordic sessions over 13 weeks on top of normal training.

• Overall, 38 hamstring injuries affected 36 of the 579 players (6.2%).
• The Nordic group had fewer injuries, with an odds ratio of 0.282 compared with controls.
• Injury severity didn't differ between the groups.
• Compliance was 91%, unusually high for an injury-prevention trial.

Both trials were in men, both were in football, and both used team-level randomisation with relatively few injury events. But both pointed in the same direction, with large effects.

The meta-analysis behind the headline

In 2019, van Dyk, Behan and Whiteley published a systematic review and meta-analysis in the British Journal of Sports Medicine with the headline-friendly title that Nordic curls "halve the rate of hamstring injuries". It pooled 15 studies covering 8,459 athletes, across different sports and age groups, and both women and men.

• Overall, programmes that included Nordic curls reduced hamstring injury risk, with a risk ratio of 0.49 (95% CI 0.32 to 0.74).
• Restricting the analysis to eight randomised studies gave a risk ratio of 0.52 (95% CI 0.32 to 0.85).
• Removing studies at high risk of bias gave a risk ratio of 0.55 (95% CI 0.34 to 0.89).

The authors concluded that including Nordic curls in prevention programmes reduced hamstring injuries by up to 51%. That's where the "halves" comes from.

Why a reappraisal called it inconclusive

Two years later, Impellizzeri, McCall and van Smeden reanalysed the same body of evidence in the Journal of Clinical Epidemiology, in a paper pointedly titled "Why methods matter in a meta-analysis". Their main objections:

• Most of the 15 studies didn't test the Nordic curl itself. Only five were randomised trials in which athletes were randomly allocated to Nordic curls. The others were observational, or tested multi-component programmes where the Nordic curl was one of several exercises.
• Using only those five trials, the pooled risk ratio was 0.56, but with a 95% confidence interval of 0.20 to 1.52. That range includes no effect, and even a possible increase.
• Adding one newer trial gave a risk ratio of 0.59 (95% CI 0.27 to 1.29).
• Prediction intervals, which estimate the range of effects you might see in a new setting, were wider still: 0.10 to 3.29 in their updated parametric analysis.

They concluded that the evidence was inconclusive, came mostly from trials at high risk of bias, and could at best support a conditional recommendation, and only for football.

Reading the two analyses together

Both analyses found an average effect in the region of 40 to 50% fewer injuries. What they disagree about is how confident anyone can be in it, and how far it generalises. The 2019 analysis is more optimistic because it pooled more studies, including ones that didn't isolate the Nordic curl. The 2021 analysis is more cautious because it kept only the most direct evidence, which turned out to be thin.

A fair summary is that Nordic curls probably reduce hamstring injuries in footballers, particularly repeat injuries, but the size of the benefit is uncertain, and the evidence outside football is weak. That's still better evidence than most injury-prevention exercises have. It's not the settled "halves your risk" fact it's often presented as.

Who was studied, and what that means for you

Population gaps matter more than usual here:

• Sport: the randomised trials are dominated by football, a sport built on repeated sprints, sudden accelerations and kicking. Those are exactly the movements that produce sprint-type and stretch-type hamstring strains.
• Sex: the two best-known trials, Petersen and van der Horst, included only men. Van Dyk's meta-analysis included some studies in women, but nowhere near enough to draw separate conclusions.
• Age: most participants were in their twenties. Masters athletes, who make up much of the age-group field, are barely represented.
• Triathletes: no trial has tested Nordic curls for injury prevention in triathletes, runners or cyclists.

Inference: triathlon running is mostly steady and submaximal, and the Danielsson review links sprint-type strains to high-speed running. That suggests most triathletes have a lower baseline risk of hamstring strain than footballers. A relative reduction applied to a lower baseline risk prevents fewer injuries in absolute terms. Hamstring pain in endurance athletes is also often a different problem, such as tendinopathy high in the buttock, which the prevention trials didn't address.

When it might be worth it for a triathlete

That doesn't make the exercise pointless. Some triathletes look more like the trial populations than others:

• You've had a hamstring strain before. In the Danish trial, the reduction in recurrent injuries was the largest effect reported (rate ratio 0.137), though it came from a small number of injuries.
• You sprint. Strides, hill sprints and fast track sessions put your hamstrings through the high-speed late-swing loading that causes strains. TriForward's "Strides and Hill Sprints" covers why that work is worth doing.
• You play another sport. Five-a-side football, hockey or rugby in the off-season bring back the classic risk profile.
• You're building eccentric strength anyway. As TriForward's "Quads Gone by 30 km" explains, eccentric capacity matters for downhill and late-race running, and Nordic curls are an efficient hamstring option.

If none of these apply, Nordic curls are a reasonable optional extra in a strength block, not a priority over the basics of squats, hinges, calf work and sensible run-load progression. TriForward's "Two Sessions a Week" and "The Off-Season Strength Block" cover those foundations.

How much is enough?

The Cuthbert meta-analysis is the most useful guide to dose, and its message is reassuring: low-volume programmes produced eccentric strength gains at least as large as high-volume ones. The football trials used progressive programmes, a 10-week build in the Danish trial and 25 sessions over 13 weeks in the Dutch one, followed by in-season maintenance.

Adherence is the real-world problem. A survey by Bahr, Thorborg and Ekstrand in the British Journal of Sports Medicine (2015) asked 50 elite football teams, 32 in the Champions League and 18 in Norway's top division, about their use of the exercise. Across 150 club-seasons, the full programme was completed in only 16 (10.7%). If professional teams with physiotherapists on staff can't sustain it, a triathlete squeezing strength work around three sports needs a version small enough to keep doing.

Adding Nordic curls to a winter strength block

The following is inference from the studies above, not a tested triathlon programme:

1. Start with a tiny dose. One or two sets of a few repetitions, lowering only as far as you can control, is plenty for the first week. Nordic curls are notorious for causing soreness the first few times.
2. Use assistance. A resistance band around your chest, attached to something behind you, or lowering onto a bench or Swiss ball in front of you shortens the range and makes the exercise manageable while you build up.
3. Progress the range before the volume. Aim to control the lowering further before adding sets. The Cuthbert analysis suggests you don't need high volume.
4. Keep it away from key sessions early on. As TriForward's "Sore From the Gym, Flat on the Bike" explains, the first exposure to an eccentric exercise causes the most soreness and can cost power for a day or two. Put early Nordic sessions after a hard day, not before one.
5. Maintain through the season. The football trials moved to a weekly session in season. A short, regular dose is more realistic than an intense block followed by nothing.

A triathlon scenario

The following example is illustrative, not drawn from a study. A 38-year-old triathlete strained his hamstring two seasons ago during a fast finishing kick at a 10 km race. He has added strides twice a week and wants to include Nordic curls in his winter gym session.

• Weeks 1 to 2: one set of three band-assisted repetitions, lowering only partway, after his Tuesday intervals so Wednesday's easy run absorbs any soreness.
• Weeks 3 to 8: two sets of three to five, gradually lowering further with less assistance.
• From week 9: one or two sets once a week, kept going into the race season alongside his strides.

He's in the group the evidence fits best: a previous strain, regular high-speed running, and a dose small enough to keep doing. A triathlete with no injury history who never runs faster than threshold pace has a much weaker case for prioritising it.

Common mistakes

• Treating "halves your injury risk" as settled. The best-known number comes from a meta-analysis whose direct randomised evidence is far less certain.
• Going too hard in the first session. A full set of Nordic curls on day one can leave you unable to run properly for days.
• Dropping the hips. Bending at the hips turns the exercise into something easier and less specific. Keep a straight line from knees to shoulders as far as you can control.
• Using it to treat a current strain. Rehabilitation of an injured hamstring needs graded, assessed loading, not a prevention exercise copied from football.
• Ignoring run load. No single exercise offsets a sudden jump in speed work or volume.
• Stopping as soon as the season starts. The trials used ongoing maintenance, and strength gains fade without it.

What we don't know

• Whether Nordic curls reduce injuries in triathletes, runners or cyclists. No trial has tested it.
• The true size of the effect in football. The direct randomised evidence is compatible with a large benefit and with none.
• Women and masters athletes. The key trials were in young and adult men.
• The minimum effective dose for injury prevention, as opposed to strength gains.
• Whether other hamstring exercises, such as Romanian deadlifts or hip extensions, offer similar protection with less soreness. TriForward's "Hips and Runner's Knee" looks at a similar gap between strength gains and injury outcomes for the hips.

References

• van Dyk N, Behan FP, Whiteley R. Including the Nordic hamstring exercise in injury prevention programmes halves the rate of hamstring injuries: a systematic review and meta-analysis of 8459 athletes. British Journal of Sports Medicine, 2019;53(21):1362–1370. doi:10.1136/bjsports-2018-100045
• Impellizzeri FM, McCall A, van Smeden M. Why methods matter in a meta-analysis: a reappraisal showed inconclusive injury preventive effect of Nordic hamstring exercise. Journal of Clinical Epidemiology, 2021;140:111–124. doi:10.1016/j.jclinepi.2021.09.007
• Petersen J, Thorborg K, Nielsen MB, Budtz-Jørgensen E, Hölmich P. Preventive effect of eccentric training on acute hamstring injuries in men's soccer: a cluster-randomized controlled trial. American Journal of Sports Medicine, 2011;39(11):2296–2303. doi:10.1177/0363546511419277
• van der Horst N, Smits DW, Petersen J, Goedhart EA, Backx FJG. The preventive effect of the Nordic hamstring exercise on hamstring injuries in amateur soccer players: a randomized controlled trial. American Journal of Sports Medicine, 2015;43(6):1316–1323. doi:10.1177/0363546515574057
• Cuthbert M, Ripley N, McMahon JJ, Evans M, Haff GG, Comfort P. The effect of Nordic hamstring exercise intervention volume on eccentric strength and muscle architecture adaptations: a systematic review and meta-analyses. Sports Medicine, 2020;50(1):83–99. doi:10.1007/s40279-019-01178-7
• Bahr R, Thorborg K, Ekstrand J. Evidence-based hamstring injury prevention is not adopted by the majority of Champions League or Norwegian Premier League football teams: the Nordic Hamstring survey. British Journal of Sports Medicine, 2015;49(22):1466–1471. doi:10.1136/bjsports-2015-094826
• Danielsson A, Horvath A, Senorski C, Alentorn-Geli E, Garrett WE, Cugat R, Samuelsson K, Hamrin Senorski E. The mechanism of hamstring injuries: a systematic review. BMC Musculoskeletal Disorders, 2020;21(1):641. doi:10.1186/s12891-020-03658-8