Clinicians all want to think we are in control—both of ourselves and our patients’ outcomes. We want to think that the investment we have put in to our schooling and training implies we cannot only reach a proper diagnosis, but implement the proper treatment to get the majority of our patients better. This, unfortunately, is not what the literature shows to be the case.  Rates of low back pain outcomes and various other musculoskeletal problems have not wavered in their base rates—the incidence of occurrence in the normal population —  in recent history.1, 14 Perhaps it is because we are too close to the situation and cannot see how little influence our interventions ultimately have. Perhaps it is because we often place the locus of control on ourselves instead of our patients’ or athletes’ own self-management. We want to be the ones to “fix” patients instead of helping them realize it will be the patients who ultimately heal themselves. The best we can do is shift the odds of an outcome in the most favorable direction.

The issue is, the closer we get to anything or if we actually touch something, we over estimate our ability to control or change it. Ellen Langer, Harvard Ph.D in psychology, has made a career out of demonstrating this phenomenon.2 She has shown repeatedly that as soon as we are able to interact with a person or situation we dramatically over estimate our ability to affect an outcome. The emphasis on diagnosis in our education leads us to think we can have more influence over long term outcomes than has been shown to be the case in the literature. Unfortunately, this has created a situation where efficacy of treatment methods and physiology are sorely lacking in clinician education. Understanding basic tissue healing lets clinicians know we will never fix a tendinopathy from a histological standpoint in a 6-week treatment span.3 We also realize ACLs take much longer to be fully healed than for which a treatment plan can be justified by a clinician.4, 5

There is danger in every event: even breathing correlates with injury if you do it long enough. Much of the industry’s focus recently is on words like ‘prevention’ and ‘prediction’, but if there is one thing that science in multiple fields has shown time and again, we are abysmal at both.6, 7, 8 Reduction of risk is not nearly as sexy and a much harder pill to swallow.  There will always be some inherent risk with any activity. Here is the thought experiment for the day—let’s say we get REALLY good at prediction of injury risk, at what likelihood of injury would you as a health care practitioner limit an athlete’s participation: 20%, 30%, 40%? The better we get at predicting events, the more the ethical decision to allow participation lies on the practitioner with the knowledge. The Damoclean sword swings on an increasingly thin string.

Maybe we should not emphasize predicting injury but instead forecasting. The word “predict” itself creates a dichotomy instead of a spectrum. This flawed thinking on the part of clinicians only adds to the anxiety we are constantly trying to avoid imparting on our patients. Promoting statements such as “knee valgus during a squat will destroy your knees” is not only unjustified by the literature but psychologically damning. Now patients become avoidant of an event with no correlation to destroying knees and one that is a common part of many athletic events. Nate Silver, in his book “The Signal and the Noise,” explains the difference between predicting and forecasting.9 Forecasting is attempting to plan under conditions of uncertainty. It is giving the best odds we have with the information we have at our disposal. Predicting is guessing, and, more often than not, uneducated guessing.

Some of the best forecasters we have are weathermen but they are assumed to be some of the worst because of the confusion of prediction and forecast.10 We want to think in terms of a dichotomy—either it will rain or it will not—but in reality there are always odds and humans are inherently loss aversive.7 The way of thinking about it would go as follows; if there were a 50% chance of rain today would you go to the beach? How about if there were a 50% chance that you would get injured in the state championship game? The loss of the current day would cause most people to avoid the beach but the same statistical probability would have most people choosing to play in the game with the possible loss of a future season or career.11 What should the advice of the clinician be if there were accurate data? Is the decision for participation ultimately the clinicians to make?

Even if we get to the point of reducing risk by 99.9%, there is still likely to be an incidence of injury if we have a thousand exposures. We do not know if that will be the 42nd, 422nd, or 1000th exposure. Sports are going to offer many exposures over a season, and the better an athlete becomes at their sport (which should happen as we train with good scientific fundamentals) the more exposures they are going to receive. Are we not ultimately increasing their risk of injury by making them better by training? Common injuries like low back pain, ACLs, and hamstring strains have not seen their rates change in recent years.1, 6, 8 Could it be that the actual odds of injury per exposure have decreased, but with early sports specialization and increased exposures from year round participation in the same sport that the overall likelihood of injury has remained the same?

If an athlete comes to us wanting to prevent an ACL tear we would be remiss to not inform them the best we can do is reduce the risk by promoting variables that have been shown to do so.12, 13 There is never a zero percent risk as long as they are participating in sport. The best predictors of ACL tears are looking at the sport the athlete plays; that is why we have studies showing varying levels of incidence according to different sports.14 We know that working on jumping and landing mechanics as well as getting athletes stronger can have a protective effect.13, 15 We also know that there is a correlation between the amount of playing time an athlete gets and their likelihood of injury.16, 17 But we also know we cannot reduce the risk of injury to zero in any instance.

There are few universally bad and no universally best exercises for helping athletes reach their goals. Any rehab specialist or coach trying to push either as fact does not have a proper understanding of the specific adaptation capabilities of humans.18 Not only that, but by deeming some exercises as dangerous these individuals peddling these beliefs are limiting selection for something that may ultimately be advantageous for athletes. The reason we have so much diversity in training, with very few truths, is because we as humans are diverse and have different goals we are trying achieve. There are some laws in the field, such as “if you want to get stronger you need an external stimulus,” but the difference of a bar on your back being 3 inches higher or lower is irrelevant for ninety-nine percent of the population. And, according to the internet, we all hate the 1% anyway. Thus, we get to the crux of the matter: the problem with any exercise is in the dosing with which it is applied. Specificity of training methods has been shown in a recent review by Macnamara et al to only account for 18% of athletic performance.19 This leaves a host of options for an athlete to work on their foundation. If we have 82% of success related to other variables of training, we are quite possibly better served continuing to work on the basics instead of obsessing over the minutia.

End range lumbar extension is not dangerous, go to a Cirque du Soleil show if you want to see otherwise. Symmetrical rotation in the spine is not the gold standard, go watch a sweep rowing event and you will see beautiful asymmetry. These athletes took time to adapt to what structuralists would often consider problematic adaptations. If “mobility” were truly the key to athletic performance and health, then every kid with Ehler-Danlos should be excelling in the sports world. All those things certainly sound fun as a narrative though, so they are easy to sell. This is because we continue to think in black and white. Unfortunately, 50 Shades of Grey would probably be a better title for a training book based on statistics than a steamy novel.

Words like “stress” and “inflammation” have been given negative connotations in the health care and training fields, but we need both to progress and adapt. If we consider that delayed onset muscle soreness is a grade 1 muscle strain, from a physiological perspective we have to concede that we often injure an athlete in order to make them better.20 We need stress to strengthen, we need inflammation to heal. There is no inflection point that makes either phenomenon suddenly helpful or harmful.  There is rather a constant spectrum that causes different adaptations; some advantageous, some disadvantageous.

The research by Tim Gabbert and Phillip Glasgow have looked at this spectrum as it relates to injury rates.  They have shown that looking at acute on chronic training loads can help us better determine the risk of injury.21, 22   This shifts the focus of training from what is going on today, to what is going on over the span of weeks. Humans in general are short sighted, and athletes even more so. We all want to be better right now without even realizing that the workout is not what makes us better but the ability to recover from the stimulus. Adaptation takes time, but, once adapted, humans can do amazing things. No one was able to achieve depths of 200m the first time they dove, but they can with time.23 No one squatted 800 pounds their first time under the bar. If either of those is attempted too soon there can be serious repercussions for athletes. Even though the dosage is much less, these examples are hardly different than letting an athlete sit during an offseason then performing 2-a-day practices at the start of the season. So why are we surprised when injury rates spike during the preseason?24

Instead of demonizing specific exercises we need to reflect on programming. A noncontact injury is a training injury: no one thing caused it and no one thing could have prevented it. It is the day-to-day multivariate adaptations that allow progress. If an athlete is not adapted to do what they are trying to do, there will be an increased risk of injury. When we see a training video posted on social media, we are quick to critique form or discount the effectiveness of the session. However, without knowing the entire arc of what that athletes training program entails, how can we speak to the effectiveness?

In summation, we can no more predict injuries than we can predict winning lottery numbers. Just as every person who purchases a lottery ticket at some level expects to win (often sooner in the one in a million shot), every athlete expects to not get injured when they participate in sport (and if they do they expect it will be later in the one in a million shot). There are always going to be lottery winners even though the odds are infinitesimally small, just as there will always be injury in sport no matter how good we become at mitigating risk. To tell an athlete that we can prevent injury is a lie, and that lie is told far too often in the training and rehabilitation community. We are not fortune tellers, but risk managers. No matter how good we become at determining the risk, it will always be present, and we need to accept that as clinicians and coaches. If we’re going to play the game, there will always be risk. It’s the dosage, not the medicine that makes the poison.

If you want to discuss the article with us on Twitter, or just recommend a beer for us to try you can find us at:




  1. Hoy, D, et al. The Epidemiology of Low Back Pain. Best Practice & Research Clinical Rheumatology 24 (2010) 769–781
  2. Langer, E. The Illusion of Control. Journal of Personality and Social Psych, Vol 32(2), Aug 1975, 311-328
  3. Andarawis-Puri, N, et al. Tendon Basic Science: Development, Repari, Regeneration, and Healing. J Orthop Res.2015 Jun;33(6):780-4
  4. Ma, Y, Murawski, C, et al. Graft maturity of the reconstructed anterior cruciate ligament 6 months postoperatively: a magnetic resonance imaging evaluation of quadriceps tendon with bone block and hamstring tendon autografts. Knee Surg Sports Traumatol Arthrosc (2015) 23:661–668
  5. Dong, S, Guoming, Xie, et al. Ligamentization of Autogenous Hamstring Grafts After Anterior Cruciate Ligament Reconstruction Midterm Versus Long-term Results. Am J Sports Med 2015 43: 1908
  6. Anderson, M, Browing, W, et al. A Systematic Summary of Systematic Reviews on the Topic of the Anterior Cruciate Ligament. Ortho Journal of Sports Med 2016 4(3)
  7. Kahneman, D. (2011). Thinking, fast and slow. New York: Farrar, Straus and Giroux.
  8. Goldman, E, Jones D. Interventions for Preventing Hamstring Injuries. Cochrane Database Syst Rev.2010 Jan 20;(1):CD006782
  9. Silver, N. (2012). The signal and the noise : why so many predictions fail–but some don’t. New York: Penguin Press.
  12. Monahati, A, Larumbe-Zabala, E, et al. The Effectiveness of Injury Prevention Programs to Modify Risk Factors for NonContact Anterior Cruciate Ligament and Hamstring Injuries in Uninjured Team Sports Athletes: A Systematic Review. PLoS One. 2016 May 12;11(5):e0155272.
  13. Hewett, T, Di Stasi, SL, et al. Current concepts for injury prevention in athletes after anterior cruciate ligament reconstruction. Am J Sports Med.2013 Jan;41(1):216-24.
  14. Agel, J, Rockwood, T. Collegiate ACL Injury Rates Across 15 Sports: National Collegiate Athletic Association Injury Surveillance System Data Update (2004-2005 Through 2012-2013). Clin J Sport Med.2016 Jul 1
  15. Khayambashi, K, Ghoddosi, N, et al. Hip Muscle Strength Predicts Noncontact Anterior Cruciate Ligament Injury in Male and Female Athletes: A Prospective Study. Am J Sports Med.2016 Feb;44(2):355-61.
  16. Beynnon, BD, Vacek, PM, et al. The Effects of Level of Competition, Sport, and Sex on the Incidence of First-Time NoncontactAnterior Cruciate Ligament Injury. Am J Sports Med.2014 Aug;42(8):1806-12.
  17. Bahr, R. Demise of the fittest: are we destroying our biggest talents? Br J Sports Med.2014 Sep;48(17):1265-7.
  18. Silva, J. An analysis of the training stress syndrome in competitive athletics. J of Applied Sport Psych. 1990 Jan;2(1):5-20.
  19. Macnamara, B, Moreau D. The Relationship Between Deliberate Practice and Performance in Sports: A Meta-Analysis. Perspect Psychol Sci.2016 May;11(3):333-50.
  20. Mueller-Wohlfahrt, HW, Haensel, L, et al. Terminology and classification of muscle injuries in sport: The Munich consensus statement. Br J Sports Med.2013 Apr;47(6):342-50.
  21. Windt, J, Gabbett T. How do training and competition workloads relate to injury? The workload-injury aetiology model. Br J Sports Med.2016 Jul 14.
  22. Glasgow, P, Phillips, N, Bleakley C. Optimal loading: key variables and mechanisms. Br J Sports Med.2015 Mar;49(5):278-9.
  24. Elliott, M, Zarins, B, et al. Hamstring muscle strains in professional football players: A 10 year review. Am J Sports Med.2011 Apr;39(4):843-50.