Designing a plyometrics program for improving bone strength in young runners

Stress fractures are an incredibly frustrating injury, especially for young runners. Even though the science behind recovery from stress fractures—more properly called “bone stress injuries”—has advanced significantly in the last several years, sustaining a bone stress injury can still completely derail your season.

Of course, far better than a swift rehabilitation is simply not getting a stress fracture in the first place (or at the very least, avoiding another one). That’s what we’re concerned with today.

Thanks to advances in the fundamental science of bone injuries, we now know that a powerful lever at your disposal for increasing your resistance to bone injuries is improving your bone strength. The best way to do that? Targeted plyometric training, a.k.a. “jump training.”

This kind of high-load, high-loading-rate intervention is especially effective at improving bone strength in young runners—roughly, teens and pre-teens, with some wiggle room on either side for individual differences in puberty and growth spurts.

However, if your goal is building bone strength, you should not use a general-purpose, performance-oriented plyometrics program for runners. Plyo programs that target bone strength need markedly different components for maximum efficacy.

Below, I’ll walk you through the science behind building a plyometrics program for bone strength, including total contacts, types of jumps to include, and appropriate programming during the week.

If you just want to get the final plyo program as a printable PDF, you can download it here: 

📋 Download the bone strength plyometrics program here

Otherwise, read on to learn the science and practical programming behind this plyo routine!

Some background on bone strength and stress fractures in runners

I’ve always been fascinated by bone stress injuries: it’s remarkable to me that some runners can get a tibial stress fracture from running only 20 miles a week, while others can put in over 100 miles a week and never get a stress fracture. In graduate school I read the literature on bone biomechanics obsessively. One major takeaway was that your bone strength plays a major role in your likelihood of sustaining a stress fracture.

Note that I’m using “bone strength” loosely here: scientifically, the strength of a bone—say, your tibia—is a function of a few related factors: bone geometry, bone density, and bone microarchitecture.

Bone geometry: the overall shape of a bone

Bone geometry is the overall shape of your bones. Despite what you might’ve learned in anatomy class, just because everyone has “a tibia” or “a second metatarsal” does not mean those bones look the same! Bone shape differs markedly across different people. Here’s a sample of what cross-sectional slices of the tibia can look like in different runners: 

From Long et al 2024.

Bone density: the bulk strength of a bone

If you’re tuned in to health or medicine, you’ve probably at least heard of bone mineral density. It is exactly what it sounds like: the amount of mineral content (i.e. calcium), per unit volume of bone.^[1] The idea is simple: greater density means stronger bones.

Bone density changes throughout your life: after peaking its peak soon after adolescence, it gradually declines, though this decline is affected markedly by your physical activity level. Bone density is a good predictor of bone strength at the population level, which is why bone density is the go-to screening metric for osteoporosis in older adults.

Interestingly, bone density alone has not been a superb predictor of bone stress injury risk in runners, partly because it doesn’t tell you anything about bone geometry, and also because it’s not a full story of the “per unit volume” strength of bone. A better way to think about that concept is…

Bone microarchitecture: the microscopic arrangement of bone cells

Even given a certain amount of “bulk” bone density, there are better and worse ways to arrange that density at the cellular level.

At the microscopic level, you can use high-resolution CT scans to get some picture of how bone cells are arranged—here’s an example of “bad” (top row) versus “good” (bottom row bone microarchitecture in two older adults:

From Nishiyama and Shane.

Conceptually, here’s how I like to think about bone microarchitecture: in grade school, there’s a common class project where you’re given some amount of gumdrops and toothpicks, and you’re asked to build a bridge. The winner is the team whose bridge can hold the most weight before it collapses.

This is the essence of the bone microarchitecture question: given some amount of bulk bone density, is that bone density arranged in a manner that’s robust to mechanical stress, like the neat triangles of the winning gumdrop bridge? Or is it slapped together haphazardly, like the sloppy construction of one of the early losers in the bridge competition?

Like bone density, bone microarchitecture changes throughout life—often more quickly, and more adaptively in response to physical activity. State of the art research on osteoporosis will often try to quantify bone microarchitecture using high-resolution CT scans because it’s more sensitive to real changes in bone strength than bulk density from a DEXA scan.

The science of bone strength in runners

Given that bone geometry, bone density, and bone microarchitecture dictate bone strength, the obvious question is: can you improve these factors? The answer is a qualified yes on all three, with the exceptions and caveats varying a bit.

Bone geometry is finalized during adolescence

For bone geometry, there’s good evidence that loading before and during puberty can change the macroscopic geometry of your bones for the better.

The coolest research on this front comes from side-to-side comparisons of the humerus (upper arm bone) in people who were baseball pitchers in their teen and pre-teen years. The loading involved in pitching creates huge forces in the humerus bone, but only in the pitcher’s dominant arm. When you look at a cross-sectional slice of the left vs. right humerus in someone who was a baseball pitcher during adolescence, you see something like this:

From Warden et al.

Which provides strong evidence that regular exposure to high loads before and during adolescence can improve bone geometry.^[2] Even small increases in the size of the “outer shell” of a bone can provide big increases in its ability to withstand loading because of the physics of structural loading.^[3]

The (partial) bad news is that bone geometry—at least in the sense of the outer perimeter of the bone’s shape—is more or less “finalized” during late adolescence. That bad news is only partial because it means if you do build better bone geometry, you won’t lose it even if you become sedentary. That “outer shell” of bone geometry stays the same, even if the microscopic arrangement of bone cells inside the bone deteriorates.

Interestingly, there’s a good indicator of when you’re near the center of this “ideal window” of opportunity for modifying bone geometry: it’s when your height is changing the fastest—which makes sense, because that’s clear evidence your skeleton is rapidly remodeling!

The upshot is that your pre-teen and teenage years are a unique opportunity for locking in bone geometry gains for life. If you’re well past your teenage years, this avenue is no longer open to you, but it’s still possible to improve bone density and bone microarchitecture later in life. How? Exposure to high-magnitude loading.

A scientifically optimal program for bone strength

Bone responds and adapts to loading—that’s the fundamental law behind all of bone science. That’s why weight-bearing physical activity builds bone density, and why long a stay on the International Space Station destroys it. The science of how to optimize this is quite rich, and leads to several important takeaways that inform how we should design a program for increasing bone strength in runners. 

Optimizing loading magnitude

First, bone responds best to high-magnitude loading. When you look at bone density in high school athletes in different sports, it isn’t the runners who have the strongest bones—it’s the athletes in basketball, volleyball, soccer, and gymnastics.^[4] Why? Because these activities involve very high magnitude loads—greater, per loading cycle, than the loads you encounter during typical running training. Moreover, these loads are applied very quickly—which is partly why the loading is large to begin with.

As such, we should already be thinking of designing our bone-building program to mirror the kind of high-impact, high-magnitude loads that a volleyball or basketball player might encounter. In practice, that means jump training: i.e. plyometrics.

Optimizing loading direction

Second, bone loading should be multi-directional. In the high school sports data I quoted above, there’s another interesting variable floating around: basketball, soccer, volleyball, and gymnastics don’t just involve high-magnitude loads, they also involve multi-directional loads.

Running is, literally, quite straightforward when it comes to the manner of loading: every step is more or less the same as the last. As a result, the loading on the bone is very concentrated into a few specific areas. For example, here’s a simulation of the bone strain in the tibia and fibula during running:

Adapted from Khassetarash et al.

Note how the areas of high bone stress—the red areas in the image—are quite localized (and, notably localized exactly where runners tend to get tibial bone stress injuries!). If you want a generally strong bone, you need to load the bone from a variety of directions.

So, unlike a program of plyometrics for performance, where specificity (similarity to running) is the name of the game, our bone loading program needs to involve a range of lateral, diagonal, and forward jumps.

Optimizing loading cycle count

Third, you only need 60–100 loading cycles to more or less maximize the adaptive response from bone cells in response to high-magnitude loading. I want to plant a flag of caution here: this guideline comes from studies in rats and turkeys, not humans, and animal model research is not always a slam dunk for humans. But, as we’ll see below, there are reasons to believe these findings are at least ballpark-correct for humans too. 

When you study the adaptive bone remodeling response induced by a certain number of loading cycles (in the sense of “one plyometric jump = one loading cycle”), you get results that look like this:

Data from Burr et al. See footnote re: curve fitting^[5]

As the plot shows, we’ve got a classic diminishing returns situation: 60–100 loading cycles (in animal models at least!) looks to be around the sweet spot.^[6] Beyond that, you’re just doing more bone loading for no additional bone benefit (though there are likely performance benefits for higher amounts of load).

I’ve highlighted the fact that these data are from animal models, but the high school sports data we saw above are consistent with this limited-loading premise: in training and competition, volleyball players only average about one jump per minute (so, ~120 jumps for a two-hour practice). If you could get extra benefits from thousands and thousands of loading cycles, you’d expect the cross country runners to come out on top in terms of bone density, but they get trounced by volleyball players.  

Optimizing scheduling of the bone loading program

Fourth, it takes a few hours for bone to become “sensitized” to loading again. In the plot above, we saw that the bone remodeling response tends to hit vastly diminishing returns after about 100 loading cycles. Beyond that level, you (or at least, lab rats) don’t seem to gain much additional stimulus, at least when the loading is at the same magnitude. That leads to an obvious question: how long until the bone becomes receptive to loading again?

Fortunately, we also have some animal model data on this question, and the answer is about four to eight hours:

Data from Burr et al.

Taken at face value, these results suggest you should separate your running sessions from your bone loading plyometrics sessions, and it might even be best to break up your bone loading sessions into separate chunks during the day.

I want to reiterate that the plot above is from a very small study of four groups of nine rats, with loading sessions on only three days (with a rest day in between). I can’t overemphasize how important it is not to get overly invested in single-digit sample size animal model studies from the early 2000s. So we should really hedge our bets on this one, especially because this is where practical programming really butts up against the “on-paper optimal” schedule.

If you did full-throatedly endorse the rat model, what would be optimal for improving bone strength in a young runner? Something like:

• 8am: 40 multi-directional jumps
• Noon: Running session
• 5pm: 40 multi-directional jumps

Which is of course an absurd proposition for a 12-year-old. So, let’s look at the practical constraints.

As a coach, the only way you can be sure your runners are doing the program is if they do it at practice. So, some more pragmatic scheduling optimizations that still at least nod in the direction of this bone-sensitivity thesis are:

• Do your bone loading plyometrics before your run, not after
• When possible, do the bone loading program as a secondary session for the day (e.g. morning run + afternoon bone plyos and lift)

Real-world data shows that optimized bone loading programs really work

I’ve been ragging on these rat-and-turkey studies a lot, but I don’t want to be too hard on them, because there is a pretty nice randomized controlled trial in teenage boys showing that a (mostly) optimized bone loading program can increase bone strength over the course of several months.

The basic setup was a series of countermovement jumps, done wearing a weight vest, several times per day.^[7] The program progressed like this:

Phase I (12 weeks)	20 jumps, 3 times per day, 3 days per week
Phase II (12 weeks)	20 jumps, 4 times per day, 3 days per week
Phase III (12 weeks)	20 jumps, 4 times per day, 4 days per week

This program is “mostly” optimized in that it leans into the low-rep, high-load paradigm (60–80 reps per day), and also leans into the rest-between-sessions idea (splitting jumps into sessions of 20 reps across the day), but neglects the multi-directional loading ideas we covered earlier.^[8] 

Here’s the headline result: this bone-building plyo program increased bone strength by 5–10% (depending on the exact metric you use) over the course of nine months, compared with a matched group that did the same sports activities but did not do the plyo program. Similar work in girls, using a somewhat less optimized program, has also found benefits on the same ~5–10% order of magnitude.

Compliance will be an issue if you use multiple jump sessions per day

One other interesting takeaway is that the researchers ran into exactly the “compliance” problem I pointed to in my hypothetical example of the 8am/noon/5pm sessions for a 12-year-old.

Once the program started requiring 4 jump sessions per day, 4 days per week, overall compliance dropped to only 50% (vs. 92% in the first 12-week period). So, on any given loading day late in the study, the average athlete was probably only doing two sessions of jumps anyways. 

These findings make me a lot more comfortable recommending a max of two sessions per day on loading days, and even make me think that one is probably fine.

The bone strength plyometrics program for young runners

We’ve seen that the key elements of a program for bone strength involves doing multi-directional jumps, with 60–100 total contacts per day, 3–4 times per week, and optionally splitting those contacts into a morning and an afternoon session.

Here’s a concrete, practical routine that follows all of these recommendations:

📋 Download a printable PDF with instructions here

And here is a scrollable table of the program only (see the PDF for exercise instructions):

Details on the bone plyo program

The program is nominally designed to be 12 weeks, but you can extend or repeat it as necessary. There is a special “maintain” phase with a lower volume that is intended to be done during your competitive phase—when you’re doing your most important race-specific workouts and when you are running your most important meets.^[9]

This program follows all of the research-based best practices above, modulo some practical considerations about splitting up the jumps.

In practice, if you are a coach of young runners, I recommend doing the entire routine at practice before you do any running, since that’s the best way to balance optimal scheduling with the certainty that your athletes actually do the jumps—as any coach knows, the odds that an average kid is going to do plyometrics on their own before school is not great.

If you are a highly motivated young runner, then you can consider splitting each session into an AM (morning) and PM (evening) session if you want to squeeze out maximal bone adaptation. You can just split the routine in half, doing the first 50% of the exercises in the first session and the rest in the second. Just don’t cut into your sleep to do it.

Progressing bone loading by double-counting single contacts

For programming the total plyometric contacts, I followed the research and used raw contacts (i.e. number of jumps). The progression in total jump contacts is logical and follows the best-practices advice, but relatively “flat.”

For progressing the exercises over time, I used a more biomechanically informed heuristic  of intentionally “double counting” jump contacts that are single-leg jumps. With this approach, five reps of a single leg jump like speed skaters counts for ten effective contacts, even though for bone-guideline purposes it’s only five. 

Here’s the rationale: because single-leg jumps like speed skaters involve putting all of your body weight through one leg, they impose a larger load than if you’d done a similar exercise where you jump and land on both legs.

Now, it’s not a perfect heuristic, but this double-counting practice does illustrate how the overall biomechanical loading progression is steadier than it looks from the raw contacts, and how the “maintain” phase is not quite as easy as its raw jump count suggests.

So, even though this progression does not show up in the “official” contact count, it’s worth thinking about if you are designing or modifying any plyometrics program (even a performance-oriented one).

Can older runners still benefit from bone loading programs?

Is it “wrong” to do this program if you aren’t an adolescent runner? I don’t think so—we saw earlier that improvements in bone geometry are probably off the table, but you can still train (or preserve) bone density and bone microarchitecture. There’s plenty of literature on astronauts and older adults showing that bone still adapts according to the same general principles, even if the specifics differ.

However, there are a few things older runners should keep in mind. First, you should be more conscious of how this program contributes to your overall bone loading (since you are probably doing more training than a 14-year-old), so you may want to ramp up even more slowly—or start up at an even smaller number of contacts.

Second, and almost paradoxically, it’s also fine to keep progressing the total number of contacts at the end of the program, and transition from general multi-directional jumps to more running-specific and performance-oriented plyometric exercises, since you aren’t trying to get the all-around bone geometry improvements.

Plus, with running-specific plyo exercises, you’ll gain better running economy benefits as well. I’ll do an updated write-up on plyometrics for performance soon, but until then you can check out my older article on plyometrics for runners, or check out the streamlined marathon-focused plyo program from Marathon Excellence.

Other considerations for bone strength in young runners

Since we’re on the topic of bone strength in young runners, there are a few other things to keep in mind.

First and most obviously is the role of RED-S (relative energy deficit in sports) in young runners, plus the more specific effects of menstrual cycle disruptions (amenorrhea) in females.

These conditions can significantly blunt bone’s adaptive response to loading, so this program will not save you if you have energetic or hormonal inhibitions to bone remodeling. Also relevant here is calcium and vitamin D intake—particularly vitamin D, where it’s very easy to tip into deficiency in the winter in northern climates if you aren’t taking a supplement.^[10]

Another thing to be careful about is using this kind of bone loading program after a stress fracture or other bone stress injury, such as medial tibial stress syndrome (“shin splints”).^[11]

Remember, the whole point of the program is to expose bones to larger and more varied biomechanical loads, so you don’t want to be too cavalier about starting up this kind of program soon after recovering from a bone injury. You’d be wise to work with a good PT, physio, or athletic trainer to progress bone loading gradually, only introducing big dynamic loads like this when you’re ready.

On a similar note, be mindful of overall bone loading both from your running training and from sports you do (soccer, frisbee, etc). Even though bone’s adaptive response tops out after a small number of loading cycles, the damage response does not have this behavior: you keep accumulating damage (and maybe even more damage) as you pile on more and more loading cycles.

Conclusion

The science of bone remodeling gives us some very practical tools for building up bone strength in young runners. The main ingredients are:

• High-magnitude multi-directional jumping exercises
• Done for a small number of total contacts (60–100 jumps per session)
• Done three times per week (maybe four, eventually, for longer-duration programs)
• Ideally done separately from your running session (by at least four hours), or barring that, done before your running session
• And optionally split into an AM and PM session with 30–50 jumps each for highly motivated athletes

That’s it! Not too time-intensive and no fancy equipment requirements either. You can use the program I developed, or build your own using the same principles.

Download the free printable plyo program as a PDF here

Learn more about the science of training

If you enjoyed this article, you should subscribe to my email list! It's the best way to find out when I have new content coming out.

If you'd like, you can also follow me on Instagram at @jdruns, and on YouTube as well, but the best way to stay in touch is definitely my email list.

Now, this is not so relevant for high schoolers, but I do have a comprehensive book on marathon training that I released last fall, titled Marathon Excellence for Everyone. It's the single best resource out there for modern marathon training (and it covers how to incorporate plyometrics into marathon training as well!).

Learn more about Marathon Excellence here, or get the book now on Amazon. If you live outside of the United States, Marathon Excellence is also available in a Metric Edition with all workouts in kilometers!

I also have a shorter book from back in 2013 that covers a simple, scientifically based approach to training for distances from the 800m to the 10k, Modern Training and Physiology. Check it out!

Footnotes

---

^[1] When you measure bone density the most common way, which is with a DEXA machine, you actually get “areal” bone density, which is in grams per cm squared, because the X-rays “squish down” one of the three dimensions (much like how you can’t get full 3D bone geometry from a single X-ray image).

^[2] In the study this image comes from (Warden et al 2019), the authors also conducted 3D simulations of the mechanical stress inside the humerus during pitching-like loads and confirmed that, as you would expect, the thicker pitching-arm bone does experience less mechanical strain, and thus sustains less damage, for a given amount of load.

^[3] The reasoning here has to do with the second moment of area, which describes how much a loaded beam (like a long bone) will deflect when loaded. For a thin shell (like cortical bone), the second moment of area is roughly proportional to the fourth power of the radius, meaning a 5% increase in the outer diameter of a bone leads to >20% greater resistance to bending.

^[4] Runners do have reasonably good bone density, compared with non-athletes and also when compared with athletes in non-weight-bearing sports like swimming and cycling.

^[5] The line fit to the turkey plot is less crazy than it looks: the rat data are cleanly linear when plotted with log-transformed jumps on the x axis (well, log(x+1) to deal with the zero-jump groups), so the reasonable default is to assume the same from the admittedly-noisy turkey data. I’ve actually fit straight lines to both in the log-scale, then back-transformed to the linear scale.

^[6] I think the 60–100 loading cycle range makes sense for a rough guideline for a plyometrics program for bone strength, but some of the clinical/translational literature takes these animal studies a bit too seriously, in my opinion. Taken literally, the Burr et al. plot would suggest the bone loading benefits of running are basically nil after ~30 seconds of running. That seems pretty implausible to me! Also, the literal interpretation doesn’t say what counts as a “loading cycle”: when your kids show up for track practice, they’ve already had thousands of loading cycles in their legs that day just from walking around. Does that not count? If so, why does your 30-min easy run “count” in terms of a bone stimulus (in the sense of “don’t do your plyometrics routine right after running”?). If the difference in loading magnitude is big enough, you’d think the bone cells would pick up the difference. Anyways, while I’m glad we have some research in this area, I would very much like to see it replicated, both in animal models and in RCTs in humans.

^[7] I don’t know how much value the weight vest really added; the weight vest was not used at all in Phase I, then was used to add 2 kg (4.4 lbs) in Phase II, and finally 5 kg (11 lbs) in Phase III.

^[8] Also, note that counter-movement jumps are double-legged jumps, which generate less load per leg than a single-leg jump (like a layup in basketball or a triple jump in track and field). See the plot later in the article of double-counted jumps for one heuristic for dealing with this difference.

^[9] All strength training should follow this same sort of “build + maintain” approach. Strength work of all kinds is supportive training—plyometric jumps, weight lifting, hill sprints, and multi-station circuits help support your ability to do various race-specific sessions, but they themselves are not race-specific. So, their volume should go down when you are doing your most important race-specific work (and your most important races, of course).

^[10] Vitamin D in the wintertime is one of the few supplements I think is a good idea for many runners.

^[11] It is still somewhat controversial whether medial tibial stress syndrome is bone-related; I’m mostly in the “MTSS is a bone injury” camp because the alternative explanations (e.g. muscle traction on fascia) are not anatomically plausible, but I’m open to new evidence on this front.

Related articles