• Title: Intramuscular tendon injury is not associated with an increased hamstring reinjury rate within 12 months after return to play

  • Author/s: Anne D van der Made (@AvanderMade)Emad Almusa , Gustaaf Reurink (@guusreurink)Rod Whiteley (@RodWhiteley)Adam Weir (@AdamWeirSports)Bruce HamiltonMario Maas (@mmaas_mariom)Aston S H Ngai (@draston)Maarten H Moen (@MaartenMh ‏)Gert Jan Goudswaard (@gjgoud07)Johannes L Tol (@johannesLTol).

  • Year: 2018

  • Question: Does having an intramuscular tendon injury (rather than other injury locations) result in increased hamstring reinjury rate in the 12 months after returning to play?

  • Key Takeaway: Contrary to previous literature, intramuscular tendon injury did not correlate with an increased hamstring reinjury rate compared with other hamstring injury locations.

  • Summary: a large cohort (165 athletes with hamstring injuries, predominantly footballers) were assessed via MRI within 5 days of injury, and categorised as having injuries with either a) intramuscular tendon disruption, or b) no tendon disruption.  Of the 165 participants, 64 (39%) were found to have intramuscular tendon disruption.

    Key Findings:

    • No difference was found in reinjury rates between injured athletes with distrupted (20% reinjury rate) and non-distrupted (19%) intramuscular tendons.
    • This finding somewhat contradicts previous research (e.g. Pollock et al., 2016), which showed reinjury rates being higher, and return-to-play time longer in hamstring injuries involving the intramuscular tendon.


The aim of this post is to summarise and compare the recent work of Anne van der Made and colleagues (2018) with the original work of Dr Noel Pollock et al. (2016), both of which looked at the influence that intramuscular tendon involvement may have on our clinical decision making in hamstring rehabilitation. This comparison is particularly important with van der Made et al.’s (2018) work appearing to conflict with that of Pollock et al. (2016).

The prospective study of van der Made et al. (2018) assessed the speed and effectiveness of criteria-based rehabilitation on two different cohorts based in Holland and Qatar. Both the therapists and participants were blinded to the findings of the injury assessments (MRI scans). This is in contrast to the retrospective work of Pollock et al. (2016) whereby all athletes and therapists were provided with an MRI diagnosis at the time of rehabilitation, and progressions through rehabilitation were made with no formal return to play criteria set out.

Looking at the results of van der Made et al. (2018), an average reinjury rate of 19% was reported within the two cohorts at twelve-month follow up, with those athletes who displayed tendon involvement on imaging presenting with a 20% reinjury rate, and therefore nil significant difference. This is in complete contrast to the work of Pollock et al. (2016) who reported 4% and 28% reinjury rates in the tendon-involved cases and non-tendon-involved cases respectively. These two findings are obviously significantly different, so it is worth considering the potential reasons behind this discrepancy;

  • The difference in prospective and retrospective study designs is something that cannot be ignored, and it could be suggested that there was a greater risk of bias within the work of Pollock et al. (2016). This is due to the fact that the authors were involved in the publication of the British Athletics Classification system which is utilised to differentiate between involved structures following muscle injury. Also, even if only subconsciously, the medical professionals must have had a belief that tendon injuries altered the length of rehabilitation to have produced the paper, and therefore may have tailored their rehabilitation plans accordingly. This bias is obviously not assumed, but must be considered in a critical discussion.
  • The difference in cohorts is a key factor in why the two papers may have presented differing findings. Pollock et al. (2016) worked with track and field athletes, while a majority of the participants for van der Made et al. (2018) were footballing athletes, and we must appreciate the differing demands of these two sports on the hamstring muscles. A kinematic study from Schache et al. (2011) reported that the only significant difference in strain when increasing running speed came to the hamstrings during terminal swing phase. We must therefore acknowledge that a greater strain is placed on the previously compromised musculotendinous unit during maximum effort sprints, which are more regularly performed in track and field than in football. Could this be the reason for a greater discrepancy in outcomes between the two studies?
  • The third, and potentially key difference in study design comes from the fact that van der Made et al. (2018) did not record exacerbations suffered during rehabilitation, and also had a greater period of follow up, at twelve months compared to three months with Pollock et al. (2016). Considering that the van der Made study found a p value of 0.06 for reinjury rate at three months in their study, we can only speculate what difference inclusion of exacerbations would have on this as the data was not included. However, it is reasonable to wonder if this may have raised the early reinjury rate to a statistically significant value, and may have brought it more in line with the work of Pollock et al. (2016).

These two articles are both valuable pieces of literature in their own right, and in my opinion highlight one key message to us as clinicians. This message is that we cannot currently prescribe a set rehabilitation program to our athletes based purely on an MRI diagnosis/classification. All of our athletes are different, and all injuries sustained are just as diverse, with different sporting codes placing completely unique demands on the muscle structures following return to play. For this reason we must work to a robust, comprehensive rehabilitation strategy from which athletes only progress to return to play upon achieving a number of key criteria. This may include the ability to perform a set volume of high speed running and/or production of acceptable levels of eccentric knee flexor force as quantified by the NordBord. Whatever the methodology that we believe in as individual clinicians, it is vital that we cast a broad net with multiple criteria, and stick to these in order to give our athletes the best chance of a successful rehabilitation outcome.



  1. Pollock et al. (2016). Time to return to full training is delayed and recurrence rate is higher in intratendinous (‘c’) acute hamstring injury in elite track and field athletes: clinical application of the British Athletics Muscle Injury Classification: British Journal of Sports Medicine, 50, 305-311.
  2. Schache, A.G., Blanch, P.D., Dorn, T.W., Brown, N.A.T., Rosemond, D. & Pandy, M.G. (2011). Effect of Running Speed on Lower Limb Joint Kinetics. Medicine & Science in Sports & Exercise, 43(7), 1260-1271.
  3. van der Made et al. (2018). Intramuscular tendon injury is not associated with an increased hamstring reinjury rate within 12 months after return to play. British Journal of Sports Medicine, Online First.