Long Distance Hiking Injuries, Part 3: Experience Matters

The third and final installation of the long distance hiking injury series looks at the effect of the following factors on hiker injury:  footwear, prior injury, warm-up routine, and experience.  I will utilize the data set obtained from the 2013-14 long distance hiker survey.

In part one of this series, I examined the injury rate amongst long distance hikers on the AT and PCT and explored injury types unique to both trails as well as injury length and the effect of taking time off to heal.

In part two, I analyzed the effect of pack weight on injury incidence, noting that on the Appalachian Trail, base pack weight category is unrelated to injury incidence, but on the PCT, there is a relationship between pack weight and injury incidence.  I suggested that this relationship, however, may be mitigated by prior experience.

As usual, I’ll define some terms ahead of time.

Injury means an actual injury or more than a few days of aches and pains.  (The rationale for this definition was covered in part one.)  Hikers were asked to identify the location of their most significant injury for this survey.

Experience means having hiked more than 500 continuous miles in a single season prior to hiking the AT or PCT.


Everybody ought to go barefoot now and then.  A dawn meadow delightful to the eye is equally so to the naked sole chilled by dewdrops sparkling on the grass.  –from Backpacking: One Step at a Time (4th ed) by Harvey Manning.

The topic of footwear worn by modern long distance hikers is a subject of interest in an of itself due to the fact that when Mr. Manning wrote the fourth edition of Backpacking in the mid-80s, the only type of footwear given serious consideration by serious hikers was the boot.  My have we evolved!  So, before delving into whether footwear matters when it comes to injury, here’s the footwear distribution amongst thru-hikers on the AT and PCT in 2013-14.  773 thru-hikers answered the question: “What footwear did you wear the most often during your hike?”

Footwear Distribution on the PCT and ATNo-one in this day and age should be surprised that low cut trail runners dominate the footwear choices amongst long distance hikers.  What’s interesting about this graph is that boots make up 31% of footwear choices amongst AT hikers, while only making up 6% of footwear choices amongst PCT hikers.  Might this relate to hiking experience?  Here’s a graph that shows the footwear distribution amongst the 128 hikers who hiked more than 500 miles in a single season prior to thru-hiking the AT or PCT:

Footwear Distribution (experienced)What’s notable is that the percentage of AT hikers wearing boots (17%) is nearly half the percentage when experienced and novice AT hikers are combined (31%).  The remaining 14% of experienced hikers are distributed amongst various forms of trail runners or sneakers.  The PCT distribution remains relatively the same, plus or minus a few percentage points.  Do experienced hikers know something novice hikers don’t when it comes to footwear?  (This particular analysis is based on visual inspection of data trends and does not denote statistical significance.)  Now, onto the relationship between footwear and injury.

Injury by Footwear Type: PCT

 Injury = NoInjury = Yes
Minimalist Footwear41
Sneakers/Running Shoes1621
Trail Runners-low cut105121
Trail Runners-mid cut815

Injury by Footwear Type: AT

 Injury = NoInjury = Yes
Minimalist Footwear46
Sneakers/Running Shoes1119
Trail Runners-low cut83136
Trail Runners-mid cut2526
No Footwear01

Injury by Footwear Type: PCT and AT

 Injury = NoInjury = Yes
Minimalist Footwear87
Sneakers/Running Shoes2740
Trail Runners-low cut188257
Trail Runners-mid cut3341
No Footwear01

Although visually it looks as if some footwear categories have a higher percentage of injuries than others, when using chi-squared statistical analysis, which accounts for relative numbers in each category, there is actually no relationship between footwear and injury for thru-hikers on the AT (p= 0.78), on the PCT (p=0.454), or for both trails combined (p=0.75).

Here’s another way of putting it:  This data set does not support the idea that certain types of footwear pose a higher injury risk when compared to others.  Boots, especially, sometimes get a bad rap for making hikers more prone to injury (I’m guilty of adding fuel to this fire).  But the data do not currently support this speculation (p=0.192)

Injury: Boots vs. Other Footwear Types: PCT and AT

 Injury = NoInjury = Yes
All Other Footwear261345

Prior Injury

Several new questions were added to the batch of surveys from 2014.  One of these questions was “Within the 5 years preceding your hike, did you need to take care of a significant orthopedic injury?”  The selection of a 5-year time period was admittedly arbitrary, but I thought that it represented a reasonable time period to analyze.  For this analysis, I’m going to combine trails.  This is partly to keep the post at a reasonable length and partly because if prior injury impacts long distance hiking injury risk, it probably does so in a way that is not specific to the trail hiked.   455 thru-hikers responded to this question.  The raw data is noted below:

Injury by Prior Injury History: PCT and AT

 Trail Injury = NoTrail Injury = Yes
Prior Orthopedic Injury = No157228
Prior Orthopedic Injury = Yes2050

Chi-squared analysis reveals what amounts to a significant relationship between prior orthopedic injury and getting injured on the AT or PCT during a thru-hike (p=0.054 and are we really going to argue about 4 thousandths of a point?).

Here’s a more tangible way of interpreting the data using odds ratios:  The odds of getting an injury on the trail with no prior orthopedic injury are 1.45:1.  The odds of getting an injury on the trail with a prior orthopedic injury within the past 5 years are 2.5:1.  Thru-hikers who cared for a significant orthopedic injury within the 5 years prior to their hike were 1.72 times more likely to experience an injury on the trail than hikers who did not have such a prior injury (2.5÷1.45).   (Thanks to vassarstats.net for their handy online odds calculator!)

Warm-Up Routine

Another new question for 2014 was “Before starting most days of hiking, did you complete a warm­up routine?”  Hikers were subsequently asked to choose one or more types of warm-up exercise: general stretching (not yoga), meditation, yoga, pilates, and general active exercises.  Again, trails will be combined using the rationale from the previous section.  In all, of the 455 PCT and AT hikers who responded to this question, only 54 (12%) reported completing a warm-up routine.  The frequency distribution of warm-up routine types is noted in the graph below (the percentages do not add up to 100% because hikers could choose more than one warm-up exercise):

Warm-up exercises chosen by 54 thru-hikers on the PCT and ATGeneral stretching was by far the most popular warm-up activity chosen by the small portion of thru-hikers who completed a warm-up activity before hiking.

Injury by Warm Up Routine: PCT and AT

 Trail Injury = NoTrail Injury = Yes
Warm Up = No159242
Warm Up = Yes1836

Warm-up exercise, however, was not associated with hiker injury rate (p = 0.371).  This may be simply a factor of sample size, since only a small portion of hikers reported warming up before most days of hiking.  But for now, the data do not support warming up as having an effect on injury.

Prior Experience

Is this the moment you’ve been waiting for?  Hikers from 2013 and 2014 were asked “Before your hike, had you ever hiked more than 500 continuous miles in a single season?”  777 thru-hikers responded to this question and reported whether or not they experienced an injury on the trail.  I’ve parsed the results by trail and have also combined the trails.

Injury by Prior Experience: PCT

 Trail Injury = NoTrail Injury = Yes
Prior Experience = No89126
Prior Experience = Yes5147

Injury by Prior Experience: AT

 Trail Injury = NoTrail Injury = Yes
Prior Experience = No163271
Prior Experience = Yes1713

Injury by Prior Experience: PCT and AT

 Trail Injury = NoTrail Injury = Yes
Prior Experience = No252397
Prior Experience = Yes6860

For both the AT and the AT/PCT combined, prior experience was significantly associated with injury at p=0.04 and p=0.003, respectively.  For the PCT alone, the relationship between prior experience and injury just barely failed to meet significance at p=0.08.  Pretty close, though!

Using odds ratios, thru-hikers on the AT with no prior experience long distance hiking were 2.2 times more likely to get injured compared to hikers with prior experience.  Combining the AT and PCT data, hikers with no prior experience were 1.8 times more likely to experience an injury than hikers with prior experience, although the PCT numbers by themselves did not show a significant relationship.  Since the two trails are quite different, I would lean towards the individual trail figures rather than the combined figures when interpreting this data.

Putting it all together: Predicting Injury

This post and hiker injury post #2 have all focused on the association of different factors with hiker injury.  Discovering a significant association can provide insight on factors that might influence a hiker’s propensity for injury.  A significant association alone, however, is not enough to say that one factor causes or predicts the outcome of another.  For this, we need slightly more powerful statistics.  We need to use regression.  (If you’re curious, here’s a “real life” example of using regression to predict home value.)

The graphic below, which includes data from 782 AT and PCT hikers in 2013-14 includes a selection of factors explored in recent posts, entered into a regression model:

predictive model for injury

What the graphic shows is that when all factors are considered, only prior experience can be considered a predictor of long distance hiker injury.  When prior experience is accounted for, all other factors, though potentially associated with injury, do not contribute significantly to the injury prediction model.  Let’s remove them from the model to get a final model:

injury prediction model: prior experience only.

What about the other factors like warm-up routine and prior injury?  This data was only included in the 2014 survey, so we’ll do an analysis of just the 2014 data (460 thru-hikers):

injury prediction model: adding ortho history and warm-up routineWhat we see is that with prior injury included (called “ortho history” in the model), trail becomes a significant predictor and prior injury itself is nearly significant.  So we’ll leave these in the model and drop the others out to get our final model:

final injury prediction modelNo boring stats display is any good without some concrete interpretation, so in the final graphic below, I’ve listed the predicted chance of injury for both AT and PCT hikers, based on our model:

Predicted Chance of Injury: AT

 Prior Injury = NoPrior Injury = Yes
Prior Experience = No59.4%71.4%
Prior Experience = Yes42.6%55.6%

And for PCT hikers:

Predicted Chance of Injury: PCT

 Prior Injury = NoPrior Injury = Yes
Prior Experience = No65.8%76.9%
Prior Experience = Yes49.3%62.2%

As the title of this post suggests, prior experience matters when it comes to injury.  Hikers who have previously hiked more than 500 continuous miles of trail are less likely to get injured when thru-hiking the AT or PCT.  Not having to treat a significant orthopedic injury in the prior five year also has influence.  These factors are all more significant in predicting injury than pack weight, footwear, and warm-up routine, although these factors tend to get more emphasis when it comes to preventing injury.

What do experienced long distance hikers do that lowers their risk of injury compared to more novice long distance hikers?  It could be a whole host of things, but put simply, long distance hiking is something that takes some getting used to and many of us learn through the good ‘ol School of Hard Knocks.  Each hiker’s style is her or his own and each experienced long distance hiker has probably found an optimum style, a specific combination of factors that work best for him or her, but not necessarily for everyone else.

Thanks for reading!

Leave a Reply

Your email address will not be published. Required fields are marked *