A high-level picture of biomechanical training load for runners

What do we mean when we say “training load”? This is the second article in a three-part series aimed at answering that question.

My core argument in this series is that there are three distinct types of training load you should consider—physiological training load, biomechanical training load, and psychological training load.

Today, we turn our attention to the second of these types of load. What exactly is biomechanical training load? In short:

Biomechanical training load describes the mechanical force—and ultimately, the mechanical damage—experienced by the load-bearing tissues of your body: bones, tendons, muscles, ligaments, and joint surfaces.

When talking about biomechanical training load, what we mean is “how much physical damage are you doing to your body.”

Proper biomechanical training load drives health and structural integrity: the ability to avoid injury and increase your body’s ability to sustain higher levels of training in the future.

If you are talking about the relationship between training and injury risk, you are talking about biomechanical load, not physiological training load. This difference is critical! As we’ll see below, biomechanical and physiological training load are often correlated, but if you do not understand the differences between these two concepts, you’re bound to make big mistakes in training.

This is a long article, so allow me to put my usual pitches up front. First, if you enjoy this article, please subscribe to my free email list using the form below. It is by far the best way to follow my work. If you're on my email list, you'll be the first to find out about my new articles, apps, and other projects, regardless of what happens with the social media omni-algorithm.

Second, I released a new book last month on the modern approach to marathon training, titled Marathon Excellence for Everyone. If you want to use modern training principles to run your best in the marathon, Marathon Excellence is the book you need.

(Also available: a brand-new Metric Edition for international readers!)

And now, let's dive into the topic for today: what is biomechanical training load?

A brief primer on the biomechanics of overuse injuries in running

Unlike cycling, swimming, and most other endurance sports, running generates enormous biomechanical loads in the body. Even a 100-pound (45 kg) middle-school runner can easily generate the equivalent of over 800 lbs (360 kg) of force in her Achilles tendon when she runs.

These enormous biomechanical loads in running, and the tissue damage caused by these biomechanical loads, are the singular reason why running training looks so different from training for lower-load endurance sports like swimming and cycling.^[1]

The basic premise behind biomechanical training load as a concept is the argument that tissue damage and tissue remodeling are both ultimately driven by mechanical forces—not metabolic and bio-energetic factors and not caloric or oxygen flux, which are the ultimate drivers of physiological training load.

The structural integrity of tissue matters too, as does your body’s ability to repair tissue damage, but everything starts from mechanical forces. The process looks like this:

1. Every time you take a step during a run, the load-bearing tissues in your body (tendons, bones, fascia, etc.) experience a cyclical mechanical load
2. Within each load-bearing tissue, this mechanical loading cycle causes a tiny amount of tissue damage. The amount of damage depends on the magnitude and duration of the load, plus the structural properties of the tissue
3. Over thousands of steps, this tissue damage accumulates, and if damage accumulates faster than your body can repair it, the result is an overuse injury

Understanding biomechanical training load is then a matter of understanding four factors: loading cycles, tissue load per cycle, tissue damage—which is determined by tissue strength—and tissue repair.

Why running injuries are caused by mechanical loading

The three steps above might sound complicated and technical, but my claim here is simple: take a paperclip and bend it back and forth until it breaks. Presto—you’ve reproduced the same mechanical process that causes running injuries.

An overuse injury just waiting to happen

I want to emphasize that the paperclip example is not just an analogy: it’s the exact same physical process. There are entire textbooks written about the statistics behind it.

Now, in the paperclip example: how long does it take for the paperclip to break? It depends on how many times you bend it, how hard you bend it each time, and what kind of metal the paperclip is made of. Loading cycles, load per cycle, and structural properties of the material.

So too for running injuries. Whether or not a specific tissue develops an injury is a function of how many steps you take while running, how much biomechanical force that tissue experiences during each step, and the strength of the tissue (bone density, tendon stiffness, etc.). Plus tissue repair, which we’ll get to in a moment.

The evidence supporting this mechanically focused view of running injury is very strong. To pick just a few examples:

• Both bone tissue and tendon tissue can be loaded to failure in mechanical testing outside of the body, and show the same type of “fatigue failure” pattern that you see in non-biological materials like aluminum, concrete, or plastic.
• If you apply repeated mechanical loads to the tibia bone in (living) rabbits, they reliably develop stress fractures if the number of loads and the magnitude of loading are high enough.
• If you take an entire bone, do a high-resolution CT scan of it, make a computer model of the bone's shape and localized bone density, and subject that “virtual bone” to running-like repeated loading cycles, you can use standard engineering formulas to accurately predict how long the real physical bone will withstand repeated loading cycles at running-like magnitudes.
• If you force mice, rats, or rabbits to run on a treadmill day after day, you can reliably induce tendinopathy
• When you look at the cellular structure of these “artificially injured” tissues under a microscope, they look exactly the same as what you see when you look at tissues from an injured runner.

The tissue damage model of injury matches basic coaching intuition as well: even from a non-scientist’s perspective, the major drivers of injury seem to be more mileage (= more loading cycles), faster paces and more intense workouts (= more mechanical load per cycle), and having little experience with serious training (= worse structural integrity in tendons, bones, and muscles).

Understanding the relationship between tissue loading and tissue damage

One point to emphasize is the fact that tissue loading is not the same thing as tissue damage.

In running, tissue loading is the repeated exposure to mechanical forces that stretch, bend, compress, or twist some biological structure—say, your tibia bone. Tissue damage is the appearance of physical deficits, like micro-cracks in bone, that reduce the structural integrity of the tissue.

The right way to think about tissue damage is as a continuous spectrum, with—usually—some threshold for “pain” (i.e. a running injury) and some higher threshold for “failure” of the tissue: for example, a complete bone fracture or a total rupture of a tendon.^[2]

Tissue loading causes tissue damage, but the relationship between them is not one-to-one. In fact, it’s more of an exponential relationship: a small increase in tissue load can cause a very large increase in tissue damage per loading cycle.

The best way to get a handle on the shape of this relationship is to look at some empirical data. Suppose we got a bunch of tibia bones, then subjected each of them to repeated loading at a wide range of amounts of load. For instance, 20 bones would be subjected to repeated cycles of 100 pounds (~45 kg) of force until failure; 20 more bones would be loaded repeatedly at 200 lbs, then another set at 300 lbs, and so on.

There are in fact studies that have done just such an experiment—though instead of loading with exact amounts of force, they normalize to “bone strain” (essentially how stretched-out the bone gets under load) to make it easier to compare results across different-sized samples of bone with different levels of tissue strength.

Here’s a plot of exactly just such a dataset: expected loading cycles to failure as a function of the bone strain per loading cycle. For reference, 10,000 micro-strains (με) is compressing a bone by 1% of its length—so, squishing down a 40 cm tibia by 4 mm (which is quite a lot of squishing).

Adapted from data in Caler and Carter 1989 and Milgrom et al 2000.

Notice that the X-axis is log-scaled! So, for a small decrease in the amount of force per loading cycle, the bone lasts much longer before failure.^[3]

I’ve also overlaid a green band of typical levels of tibial bone strain seen in running (measured by surgically attaching a strain gauge directly to the bone!).

Interestingly, these strains are, at least order-of-magnitude, about right for when people start having serious shin pain when starting up running “from nothing.” A high school cross country who did not run all summer can usually make it through the first day of practice all right, but if they run 30 mi/wk (~50 km/wk) for three or four weeks in a row, they might start having significant shin pain.

The role of tissue strength in determining tissue damage

Now, suppose we didn’t normalize to bone strain—what would be different? Mechanically speaking, strain is determined by the stress (force applied per unit of cross-sectional area) and the mechanical stiffness of bone.^[4]

Both of these two inputs—cross-sectional area and mechanical stiffness—map onto real-world factors that affect injury risk. In the case of tibial bone stress injuries, a physically larger bone, with a greater cross-sectional area, would experience less stress for a given amount of force, and people with physically larger bones are, all else equal, at a decreased risk for stress fractures.

Likewise, runners with higher bone density are at a decreased risk for bone stress injuries, which makes complete sense from a mechanical stiffness perspective: higher bone density means a stiffer, stronger bone, which can endure more loading cycles at a given loading magnitude before breaking.

The same principles apply to other biological tissue as well: a stiffer, stronger tendon incurs less damage for a given amount of force, and can endure more loading cycles at a given amount of force.

Tissue repair in response to biomechanical training load

One of these two is better off in the long run

The comparisons earlier between bones and paperclips raises an important objection—what about tissue repair?

Even if the statistical process of damage accumulation is the same in a paperclip and a bone (or a tendon or a muscle), there’s an important difference between people and paperclips: biological tissue can repair tissue damage caused by mechanical loading.

Indeed, remodeling of biological tissue is driven by exposure to mechanical loading in appropriate amounts.

The relationship between tissue loading and tissue repair is why modern guidelines for recovery from low-risk stress fractures (really, “bone stress injuries”) emphasizes progressive exposure to loading, as opposed to the old-school approach of “get a walking boot and call me in eight weeks.”

Similar research into the science of tendon healing has informed loading protocols for tendinopathy rehab: increasing loading magnitude over time is now a core part of evidence-based rehab for tendinopathy in runners.

We also have a few examples of situations where injuries are almost surely driven by impaired tissue healing. For instance, female athletes with relative energy deficit in sport (RED-S)-induced menstrual cycle disruptions have a markedly higher risk of developing stress fractures.

Why? Because a lack of energy availability causes a decrease in levels of the hormone estradiol ("estrogen"), and this hormone works to inhibit bone resorption and promote bone remodeling. Less bone remodeling means more tissue damage accumulating over time, even when training remains the same.

The case of RED-S is particularly instructive because we have an extremely tight mechanistic link between a cellular-level understanding of tissue repair (estradiol-mediated osteoblast activity) and the type of injury this repair mechanism is connected to (bone stress injuries).

But you could imagine similar-looking mechanisms for other things that might affect tissue repair, like sleep, life stress, and overall dietary quality.

Tissue repair only makes sense as a response to tissue damage

Despite the obvious importance of tissue repair, it’s better to try to understand tissue damage first, then use damage as a means of understanding repair.

Talking about tissue repair only makes sense if we know what amount of damage is being repaired. Clearly tissue repair happens in response to damage—otherwise you could just sit on the couch for a few years and your bones would be super-strong, thanks to all that accumulated repair time.

Second, and more pragmatically, tissue repair happens relatively slowly. Significant structural changes in tendon and bone happen over the course of weeks to months, so when it comes to actual day-to-day decisions in coaching, what really matters for injury risk in the short term is the damage you incur in each workout—managing tissue repair is a longer-term game.

Lastly,  actually quantifying tissue repair is extremely difficult. Barring any good way to peer inside the body and measure how quickly bone, tendon, and muscle cells are being laid down, I think the best approach to repair is an empirical one—if you can quantify tissue damage, or a surrogate of it, you can put some upper and lower bounds on the rate of tissue repair simply by observing athletes and seeing who doesn’t get injured.^[5]

So, it’s better to think of tissue repair as a response to biomechanical load, versus an inseparable part of it.

Biomechanical load is very important, but quantifying it is very hard

We saw above that three factors determine tissue damage, and therefore biomechanical training load: the number of loading cycles, the amount of tissue load per cycle, and the amount of damage incurred by a given amount of tissue load (which is determined by that tissue’s structural properties, i.e. its “strength”).

How hard are each of these to quantify?

Number of loading cycles: easy

This is just the number of steps you take while running. Any modern watch that tracks cadence makes it trivial to measure the total number of steps taken during a run, with extremely high precision.^[6] Helpfully, if you want to know how many loading cycles you incurred on one leg, you can just divide the total by two.

Tissue load per cycle: challenging but tractable

When we are talking about tissue load per cycle, we mean “how much force is going through this tissue for a given step?” Take the Achilles tendon—we would want to know how much tensile force is going into the Achilles tendon for each step you take on a run.

That amount of force will depend on your speed, your cadence, your running gait, your footwear, and potentially other factors too. However, it is possible to get reasonable estimates of the internal forces in structures like the Achilles tendon, patellar tendon, tibia, and patellofemoral joint. There’s a small niche of running biomechanics researchers (myself included) that focus on exactly this problem.

Make no mistake—it’s challenging. But not impossible. For instance, here is a plot of peak Achilles tendon forces (per step) across different speeds, compared with real in-vivo measurements of Achilles tendon force.

With enough data across different runners, you can characterize reasonably well how much force a typical runner experiences per step at a given speed for a given biological tissue—and you can improve these predictions if you take into account additional gait factors like cadence.^[7] With a full research-grade biomechanical workup, you can model forces in most load-bearing structures of the body quite well.

Yes, this approach does not take into account things like footwear, inclines and declines, day-to-day changes in gait, and the effects of various running surfaces, but it certainly gives you a very good place to start when discussing practical questions like how safely you can ramp up intensity after an injury.

Tissue strength: difficult, but maybe not always needed?

Even if you have a working understanding of tissue load for every step a runner takes, you haven’t solved the hardest problem: understanding the material properties of the tissue so you can predict how that tissue load translates into tissue damage.

Again, it’s hard, but not impossible, to get some traction here. High-resolution CT scans can generate very high-fidelity models of the shape and tissue properties of bones like the tibia, femur, and metatarsals. Combining ultrasound imaging with torque measurements at the knee and ankle can likewise give you a pretty good idea of the mechanical properties of the patellar and Achilles tendons.

From Burghardt et al. 2010: high-resolution CT scan of the tibia (bottom row). The top row is from the radius bone.

We still have the problem of needing expensive research-grade equipment to put a number on tissue strength.^[8] However, there is a measure of good news: while tissue strength is a major driver of absolute differences in injury risk across runners, the decisions that change your owninjury risk mostly do not depend on an accurate estimate of tissue strength.

Why? Because for short-term decisions like “should I do speedwork or a tempo run today,” your tissue strength is basically constant. So, even if you have very weak bones, the main things that will alter your relativerisk of injury are (a) number of loading cycles—so, steps taken while running—and (b) the tissue load for each of those steps.

A lack of information on tissue strength will limit our ability to rigidly quantify biomechanical training load, but we can still make good directional conclusions about what will increase versus decrease the amount of tissue damage you sustain—and perhaps we can even make some relative predictions, e.g. “workout A will do 1.5 times as much Achilles tendon damage as workout B”—even if we still need tissue strength data to put an actual number on that amount of damage.

What we really need is an “exchange rate” for equivalent-damage workouts

Having built up our understanding of biomechanical loading, what can we do with it? Well, there’s good  news and bad news.

I’ll start with the bad news: currently, there is no good way to quantify biomechanical training load.

There has been some tentative progress on chipping away at this problem—for example, there are papers showing that running fast incurs more damage to the Achilles tendon per mile than running slow, and other work has done some initial probing on how changes in incline and cadence affect tissue damage at the patellofemoral joint.^[9]

But none of these papers have come up with what coaches need: a way to compare the damage incurred by different training sessions directly.

Here’s what I’d like to see: a calculator that can compare many different possible training sessions against a common standard relevant for the runner in question. Concretely, suppose you are a runner with a history of a few Achilles injuries and a few tibial stress fractures, and suppose your typical easy run pace is 8:00/mi. You might choose 8:00/mi on flat ground as your reference for all other conditions.

Then, you could generate a chart that looks something like this, showing “damage-equivalent miles” (eq. mi)

Workout	Damage to Achilles	Damage to tibia
1 mi at 8:00/mi	1.0 eq. mi	1.0 eq. mi
1 mi at 7:00/mi	1.3 eq. mi	1.2 eq. mi
1 mi at 6:00/mi	1.6 eq. mi	1.8 eq. mi
1 mi at 6:00/mi in racing flats	1.8 eq. mi	1.7 eq. mi
1 mi at 7:30/mi up a 5% grade	1.9 eq. mi	1.3 eq. mi

…and so on. That way, you could effectively manage your biomechanical training load at a tissue-specific level.^[10]

One point to notice here is that biomechanical training load can be different for different tissues—it’s not hard to imagine how hill repeats could plausibly induce less damage per mile at the knee than a similar effort on flat ground, but more damage per mile at the Achilles. Ditto for increases in speed.^[11]

I think this “equivalent damage” framework is one of the most important unsolved problems in running injury biomechanics, and I hope to make progress on it in the next several years.

Practical takeaways for using biomechanical training load to guide your training

Now for the good news: even if we can’t yet put a precise number on biomechanical training load, there are some very useful takeaways you can apply to your own training.

Higher mileage increases biomechanical training load

This one should be a no-brainer, but we can easily explain why: more mileage means more loading cycles.

Unlike with physiological training load, mileage does not have any special properties as a load metric. So, mileage is just a heuristic. I suspect mileage does better than duration (in hours) in terms of how well it tracks with actual biomechanical training load, but it depends a bit on your speed-cadence relationship.

Faster speeds increase biomechanical training load by increasing tissue load per step

There’s one rule in tissue loading that, as far as I can tell, is universal: the faster you go, the more tissue load you experience per step.^[12] This should not be shocking either: faster speeds require greater muscle forces and less ground contact time, which interact to cause greater forces in every load-bearing tissue in the body.

The core role of speed in determining load per step offers a good explanation for why many runners find success in training approaches that emphasize very low-intensity training like the Maffetone method or the current “Zone 2” enthusiasm. The slow speeds incur much less damage, reducing injury risk and allowing these runners to train more consistently.

The other big takeaway here is “be careful and gradual when adding faster workouts.” The fact that small changes in force lead to big changes in damage explains why suddenly doing a workout like 16x200m — barely two miles of running—can cause an injury seemingly out of nowhere.

Factors like hills, shoes, and gait mechanics affect tissue load per step as well

Even though we don’t have firm biomechanical data on the effects of every possible variable you might encounter in training, you can use basic training intuitions to get a handle on some of them. Running uphill is likely to increase calf and Achilles load per step; ditto for wearing spikes, minimalist shoes, or low-profile racing flats.

Likewise, if you are working with a physical therapist to change your running gait during the injury rehab process, what you’re trying to do is reduce the load per step. In some cases, like reducing leg crossover to treat IT band syndrome, the effect is pretty straightforward: everything stays the same except your IT band experiences less force per step.

If you’re increasing your cadence while running the same speed, the situation is more complicated: tissue load per step goes down (in most tissues^[13]), but you are taking more steps to cover the same distance. The math generally works out in favor of a higher cadence though: better to take more steps with less load per step vs. fewer steps with more load per step.

For longer-term injury resilience, don’t neglect tissue strength and tissue repair

Even though tissue strength changes slowly, that doesn’t mean you should completely ignore it. If you want to run 3:05 in the marathon, you need to run 7:00/mi for 26.2 miles (or 4:20/km for 42.2 km)—and that’s going to come with some non-negotiable amount of tissue loading. In many cases, your best bet is to make your tissues stronger so they sustain less damage for a given amount of load.

You can increase tendon stiffness through heavy-load, low-rep strength training, and high-impact plyometric training can increase bone strength, especially in teenage runners.

And the simplest tool of all is gradual increases in training volume—and particularly gradual increases in your volume across a wide range of speeds, to expose your tissues to the full range of loading environments they will need (another ancillary benefit of full-spectrum percentage-based training).

Likewise, you should also cover the basics to ensure your rate of tissue repair is as high as possible: eating well (and eating enough), trying to mitigate life stress when possible, and getting enough sleep.

Don’t confuse physiological training load and biomechanical training load

The biggest oversight people make when it comes to biomechanical training load is confusing it with physiological training load. They are clearly somewhat correlated with one another, because mileage and speed are major drivers of both physiological training load and biomechanical training load, but they are not the same thing.

Even if you find physiological load metrics like training stress score (TSS) or training impulse (TRIMP) useful for monitoring your training, you still need to keep biomechanical loading in mind.

A few examples: a session of 6 x 100m strides at mile pace has a trivial effect as measured by most physiological training load metrics, but could impose a significant biomechanical training load. Even moreso for a true speed development session like 6 x 60m at maximum speed. That session alone could be “worth” an enormous amount of easy mileage when it comes to damage incurred in your muscles and tendons.

Conversely, something like 6 km ascending uphill at 10k effort imposes a hefty physiological training load, but if the hill is steep and your actual speed is slow, the biomechanical training load (at least in some tissues) could be rather small.

Keep workout volume constant to avoid increasing biomechanical training load

One final tip for managing biomechanical training load: an easy way to keep biomechanical load constant while progressing your training over time is to increase extension, i.e. the length of your repeats in your workouts. For example, consider the following series of workouts, done once per week:

• 10 x 800m at 95% 5k pace with 2 min walk
• 8 x 1000m at 95% 5k pace with 2 min walk
• 5 x 1600m at 95% 5k pace with 3 min walk
• 4 x 2000m at 95% 5k pace with 3–4 min walk
• 3k, 2k, 2k, 1k at 95% 5k pace with 4, 4, 3, min walk

This progression is quite a new stimulus to the body (eventually running almost four times as far without a break), yet all sessions total 8 km at 95% 5k pace. So, the biomechanical training load will be the same—the final session poses no greater injury risk than the first.

Recap

Biomechanical training load dictates how much tissue damage your body experiences, which affects injury risk and long-term tissue strength adaptations.

Biomechanical training load is a function of the number of steps you take in training, the mechanical load each tissue experiences on each of those steps, and the tissue strength of your load-bearing tissues, which determines how much damage a given level of mechanical loading incurs.

Though there’s no universal metric of biomechanical training load, you can modify each of these three factors to control the amount of damage your body sustains in training.

Learn more about training

If you enjoyed this article, subscribe to my email list below! It’s the best way to find out when I’ve got a new article on training, exercise science, or when I have another new web app coming out.

I also have a new book out, Marathon Excellence for Everyone, which applies many of these principles to marathon training (such as the constant-volume progression technique mentioned earlier) .

Learn more about Marathon Excellence here, or get the book on Amazon.

If you live outside of the United States, Marathon Excellence is also available in a Metric Edition with all workouts in kilometers!

Footnotes

---

^[1] Swimming and cycling are often called “low impact sports” and running is often called a “high impact sport” but using the word “impact” is misleading: the main driver of injuries in running is not the impact with the ground, but the active muscular contraction that supports your body weight and propels you forward. In biomechanics research, “impact” has a very specific definition—the first ~10–20% of stance phase, when your lower leg is being passively decelerated by its initial contact with the ground. That technical definition of “impact” often does not map very well to peoples’ casual intuitions, because it happens so quickly.

^[2] Complete failure of a tissue is, thankfully, pretty rare in running, because the threshold of tissue damage required for pain is usually a lot lower than the threshold of tissue damage required for a complete failure. One example where these two thresholds are the same might be an acute muscle tear, as sometimes occurs in sprinting—though even in these cases, it’s possible that there was a significant amount of muscle damage accumulating before the tear, even if it did not reach the threshold of damage for pain prior to the tear.

^[3] Only plotting the expected number of loading cycles to failure is a bit of an oversimplification of what’s possible on a biomechanical level. With more sophisticated models of tissue damage, it’s possible to predict the entire probability distribution of cycles to failure—so, how many loading cycles you can endure until you’re at a 1% risk of failure, 10%, 20%, etc.

^[4] The exact engineering property we’d use depends a little bit on what kind of loading we are considering and how it’s being modeled; the simplest case would be to use Young’s modulus.

^[5] Clearly, if you do X amount of bone damage in 16 weeks of training, and do not develop a stress fracture, your body must have repaired ≥X “units” of bone damage!

^[6] In fact, Garmin watches already log the total number of steps you take during a run—it’s stored as a data field in the .FIT file for each activity. It does not show up in any of the summaries you can see on sites like Garmin Connect or Strava, but it’s in the file!

^[7] The goal of my PhD dissertation was to build these kinds of tissue loading models for running “in the wild,” outside of the lab, by using RunDynamics data from Garmin heart rate monitors and Stryd foot pods. The idea was to estimate Achilles tendon force (and patellofemoral joint contact force) using only the data from these wearable devices—no need for an in-lab biomechanical evaluation. The models I built worked reasonably well, but these days I’m more excited about the prospect of doing near-research-grade biomechanical evaluations in the field using smartphone cameras and deep learning models.

^[8] My other idea for a PhD dissertation was trying to predict an individual’s bone geometry from easy-to-measure properties like ankle circumference and knee width. It turns out that even pretty sophisticated approaches fall short here, so these days I’m mostly focusing my research on understanding tissue loading, with the goal of quantifying biomechanical training load for a “typical” runner, while (hopefully) somebody else makes some advances on individualizing tissue strength measurements—and if all else fails, still being able to say something about relative directional changes in tissue damage, for the same “tissue strength changes slowly for any given individual” argument from above.

^[9] To illustrate one reason why I think these investigations are interesting but only tentative explorations: all studies to date on the effects of incline have compared running uphill, downhill, and on flat ground at the same speed, which from a coaching perspective is clearly wrong. What you should really do is compare uphills, downhills, and flat ground at the same metabolic intensity, e.g. using a grade-adjusted pace. That’s the relevant comparator.

^[10] You’d have to individualize this equivalent damage chart (or more realistically, an equivalent damage calculator app) to each runner, because what you’d really be doing under the hood would be assigning “weights” to each step taken (much like training stress score or training impulse assigns weights to each minute of exercise). But steps per mile depends on your cadence, which varies from runner to runner, and also varies as a function of speed. So the actual game plan looks like: calculate damage per step at a given speed ⇒ multiply by steps needed to cover one mile ⇒ compare against the amount of damage for the reference speed.

^[11] Treadmill running is another example of different biomechanical load at different tissues: per-step tissue forces at a given speed are the same on the treadmill vs. on flat ground at the patellofemoral joint, but at the Achilles, per-step tissue forces are higher on the treadmill.

^[12] I do not know of any exceptions to this rule – a while back, some people argued that IT band forces decrease at faster speeds, but since the IT band is essentially an extension of the TFL and glute max muscle, I find that hard to believe. Nevertheless I have not done an exhaustive review of every muscle and tendon in the body—I just know that all the usual suspects (Achilles tendon, patellar tendon, patellofemoral joint, tibia, metatarsals, femur) all experience more load per step at faster speeds. 

^[13] I have seen some preliminary data that suggest per-step load might actually go up in the hip flexors when you run with a higher cadence at the same speed.

Related articles