Determining the risk factors for any particular injury is important as some are modifiable and addressing these can be used to lower the risk for injury. The best way of doing this is to take a group of runners with no current injury, measure a whole lot of things in them and then prospectively follow them over time. You then compare what factors were predictive in those who did and did not get an injury. This is the most powerful way of determining risk factors and provides the higher level of evidence. Unfortunately, prospective studies like this are time consuming and more costly to conduct. The next best way to assess potential risk factors is to use a case control study in which you assess a number of factors in a group of runners with an injury and in a group of runners without the injury. This type of study provides some good information, but it is not as powerful as the prospective study and it should be obvious that the selection of the control group without the injury is crucial in a case control study and if not done properly can lead to some misleading information.
This case control study on medial tibial stress syndrome just appeared in the South African Journal of Sports Medicine:
Intrinsic factors associated with medial tibial stress syndrome in athletes: A large case-control study
South African Journal of Sports Medicine Vol 25, No 3 Pg 63-67, 2013
M Winters; H Veldt; E W Bakker; M H Moen
Background. Medial tibial stress syndrome (MTSS) is the most common lower-leg injury in athletes, and is thought to be caused by bony overload. To prevent MTSS, both pathophysiological and aetiological factors specific to MTSS need to be identified. The intrinsic risk factors that contribute to the development of MTSS are still uncertain.
Objective. To determine the intrinsic risk factors of MTSS by sampling a large population of athletic MTSS patients and controls.
Methods. Athletes with MTSS and control subjects were medically examined in terms of range of motion of the leg joints (hip abduction, adduction, internal and external range of motion; ankle plantar and dorsal flexion; hallux extension and flexion; subtalar inversion and eversion), measures of over-pronation and maximal calf girth.
Results. Ninety-seven subjects agreed to participate in the study: 48 MTSS patients and 49 active controls. The following variables were considered: gender, age, body mass index (BMI), hip abduction, hip adduction, internal and external hip range of rotation, ankle plantar and dorsal flexion, hallux flexion and extension, subtalar inversion and eversion, maximal calf girth, standing foot angle and navicular drop test. In multivariate logistic regression analysis, hip abduction (odds ratio (OR) 0.82; 95% confidence interval (CI) 0.72 – 0.94), ankle plantar flexion (OR 0.73; 95% CI 0.61 – 0.87) and subtalar inversion (OR 1.24; 95% CI 1.10 – 1.41) were significantly associated with MTSS. The Nagelkerke R 2 for this model was 0.76, indicating that 76% of the variance in the presence of MTSS could be explained by these variables.
Conclusion. Decreased hip abduction, decreased ankle plantar flexion and an increased subtalar inversion could be considered risk factors for MTSS.
When first reading this study a number of things jumped out at me:
1. The finding of the proximal factors being involved (ie the decreased hip abduction). A previous prospective study showed something similar. There is an increased interest in the role of proximal factors in medial tibial stress syndrome
2. A number of their findings were reported as being different to other studies. I won’t go into it, but you can read the authors discussion of this.
3. A number of potentially important risk factors were not included in the study. Given the developing clinical experiences and interest in things like running form, cadence and tibial bending moments and the potential role that these may have, it would be helpful if these types of factors were investigated in a risk factor study.
1. There was very poor selection of the control group in this study. The control group was different from the injury group in weight, age and gender mix. These factors should have been controlled for and the control group matched to the injury group on those parameters. We also have nothing on the training intensity and volume in the two groups. They should have been matched on that, but we do not know if they were or were not.
2. The measurements were done by a different researcher in each of the two groups. Many of the measurements done in the study have notorious documented issues with unreliability and there is a high potential for systematic differences between researchers on the types of measurements done (which could easily account for the differences found in the study). The authors did state that “the inter-observer reliability was found to be fair for these measurements“, but did not publish the data on this to verify how they determined that and if it was valid.
Are those shortcomings enough to be fatal? Probably.
As always, I go where the evidence takes me until convinced otherwise.