Lee SY, Hertel J, Lee SC. Rearfoot eversion has indirect effects on plantar fascia tension by changing the amount of arch collapse. Foot (Edinb).  2010 Jun-Sep; 20(2-3): 64-70. (PubMed ID: 20656471)

WHY did the authors undertake this study?
Plantar fasciitis is among the most prevalent presenting complaints within the field of podiatric medicine.  Although chronic, repetitive stress and microtrauma are believed to be contributing factors to the pathophysiology, the exact pathomechanical mechanisms within the gait cycle contributing to plantar fascia tension are less clear. The authors of this study attempted to determine the relationship between plantar fascia tension, static arch height, dynamic arch height and maximum rearfoot eversion. 

HOW did they attempt to answer this question?
The primary outcome measures of this study were static arch height (defined as the ratio between the height of the navicular tuberosity and truncated foot length in a full weight bearing position), dynamic arch height (defined as the ratio between the height of the navicular tuberosity and truncated foot length during the stance phase of running at 4.5 m/s), rearfoot eversion (defined as the frontal plane movement of the calcaneus during running) and plantar fascia tension (calculated with a previously established formula during the stance phase of running). 

Inclusion criteria of the population cohort were 28 healthy young males without history of lower extremity pain, injury or surgery. 

WHAT were the specific results?
A statistically significant positive correlation was found between static arch height and dynamic arch height, while statistically significant negative correlations were found with respect to (1) static arch height and plantar fascia tension, (2) dynamic arch height and plantar fascia tension, and (3) dynamic arch height and maximum rearfoot eversion. Further advanced statistical analysis determined that a model incorporating direct rearfoot eversion explained 82.1% of the variance in plantar fascia tension.

HOW did the authors interpret these results?
From these results, the authors concluded  that a combination of dynamic arch height and rearfoot eversion during gait provide a good predictor of plantar fascia tension.

Knappik JJ, Trone DW, Swedler DI, Villasenor A, Bullock SH, Schmied E, Bockelman T, Han P, and Jones BH. Injury reduction effectiveness of assigning running shoes based on plantar shape in marinecorps basic training. Am J Sports Med.  2010 Sep; 38(9): 1759-67. (Pubmed ID#: 20576837)

WHY did the authors undertake this study?
Acute lower extremity injuries are common following increases in activity level, and this certainly applies to the situation of Marine Corps basic training. The authors of this study aimed to determine if assigning specialty running shoes based on plantar foot shape would have an effect on acute and chronic lower extremity injury in recruits participating in basic training.

HOW did they attempt to answer this question?
The primary outcome measure of this investigation was the diagnosis of a musculoskeletal injury  during an office visit within the training time of recruits. 

Consenting participants were randomized into experimental (received a specialized shoe based on plantar foot type) and control (received a stability shoe regardless of foot type) groups at the initiation of training. Plantar foot shape was characterized as “high, low or normal” arched with an examination of plantar foot contact area during stance. Those in the experimental group with a low-arched foot received a motion control shoe, those with a high-arch foot received a cushioned shoe, and those with a normal arched foot received a stability shoe.

WHAT were the specific results?
There was no statistical significant difference in terms of injury risk as determined by hazard ratio analysis between the experimental and control groups for both men and women.

HOW did the authors interpret these results?
Based on these results, the authors concluded that shoe assignment based on foot type did not have an effect on risk of musculoskeletal injury development.

Let’s take a closer look at the topic of hazard ratios as they apply to these articles, particularly the Knappik et al shoe study.  In the “Results” section, the authors reported that the hazard ratio of injury risk was 1.01 (95% confidence interval, 0.82-1.24) for men and 0.88 (95% confidence interval, 0.70-1.10) for women. Then in the ”Discussion” section, the authors conclude that shoe assignment based on foot type did not have an effect on injury risk. How do we go from the numbers “1.01” and “0.88” to that particular conclusion, and more importantly, do we as critical readers agree with this interpretation?

The interpretation of hazard ratios is similar to another topic that we covered in this space, that of odds ratio analysis in PJC#5. Remember that odds ratio analyses usually compare two groups and the “odds” of some variable happening. In PJC#7 we found that diabetic patients undergoing operative intervention for pilon fractures were approximately 11 times as likely to develop a post-operative infection compared to non-diabetic patients undergoing operative intervention for pilon fractures. We had two groups (diabetics and non-diabetics) and some variable (development of post-operative infection). We further found that because the 95% confidence interval did not cross “1” (2.9-39.8), this odds ratio value was almost certainly statistically significant.
 
The Knappik et al shoe investigation provides a very similar statistical set-up to this example. Here we have two groups (a control group all getting the same type of shoe and an experimental group getting different shoes based on foot type) and some variable (lower extremity injury). Our hazard ratio statistic is essentially comparing the “risk” of the experimental group getting an injury to the “risk” of the control group getting an injury. It’s literally the “ratio” of these two “risks”.
 
The authors reported that the hazard ratio for males was 1.01 (95% confidence interval, 0.82-1.24). We can interpret this very similarly to the odds ratio and conclude that “the male experimental group in this study was 1.01 times as likely to develop a lower extremity injury compared to the male control group”. And because 1.01 is essentially the same as “1”, we could simplify this even further and conclude that both groups were as likely to develop an injury.
 
The confidence intervals are interpreted in the same way as well. For the female group, the hazard ratio was 0.88 (95% confidence interval, 0.70-1.10).  Here our “0.88” is close to, but less than “1”. Can we therefore appropriately conclude that the experimental group had less risk in developing an injury compared to the control group? The answer is no because our confidence interval (0.70-1.1) crosses “1” and informs us that there is no statistical significance. Although on gross examination “0.88” seems like it is less than “1”, we can now confidently assert that it is not statistically different than “1”. So we reach the same conclusion for the female group of this study, namely that both groups were as likely to develop an injury.

The take-home point here is that although hazard ratios may seem like a new and different statistic, the interpretation is very similar to another statistic that we are familiar with (odds ratio). You’ll also find that the interpretation of hazard ratios and odds ratios are very similar to risk ratios and relative risks.