The Running Lie
The “lie” is not that running is bad—it is that running became the unchallenged default, and that the costs of that default have been systematically minimized by an industry with seventy billion reasons to keep you lacing up.
The tent appears at kilometre seven, which is to say it appears at precisely the point where the body’s optimistic early chemistry begins to falter. It is a white pop-up canopy, the kind sold at outdoor retailers for tailgating and farmers’ markets, repurposed here by necessity into something more clinical: two folding tables, a stack of mylar blankets, a cooler of sports drinks, a first-aid kit substantial enough to suggest its organizers have done this before and know what to expect. A volunteer in a reflective vest crouches beside a man sitting on the grass with his left shoe removed, pressing a cold pack against his ankle with the focused expression of someone who has concluded, privately, that the injury is worse than he initially acknowledged. Behind him, a woman sits with her back against a folding chair, both hands wrapped around a water bottle, not drinking it, staring at the middle distance with the particular vacancy that follows acute pain. Beside her, a second man lies on a folding cot with his right knee elevated and his forearm draped across his eyes, blocking out a sun that is not especially bright.
This is a 10-kilometre recreational road race in a mid-sized Western city on an October Saturday morning. The field is approximately four thousand runners. The morning is cool and clear. The course is flat—a civic park loop, well-marked, aid stations every two kilometres. The technical fabric of choice this year is a moisture-wicking polyester blend in colours that suggest both athletic ambition and a certain studied casualness about athletic ambition. GPS watches are nearly universal. Foam midsoles measuring thirty-eight to forty-five millimetres at the heel are standard. The average runner will complete this course in just over sixty-three minutes. Several hundred will not complete it at all.
By the time the last finishers cross the line, the injury tent will have seen, depending on the race director’s estimate, somewhere between sixty and one hundred and twenty presentations. Ankle sprains. Knee pain—patellofemoral, iliotibial, medial. Hip flexor strains. Plantar fasciitis flare-ups in people who have had plantar fasciitis before and came anyway because they had paid the registration fee and it seemed manageable at the start line. A handful of blisters severe enough to require intervention. One stress fracture, confirmed later by imaging, in a woman who had been training for seven weeks with progressive shin pain that she had attributed to soreness. Two cases of runner’s knee severe enough to require the runner to be helped off the course before the finish.
These numbers are not anomalous. They are not the consequence of unusual heat, a poorly designed course, inadequate preparation, or the particular misfortune of this particular October morning. They are, in their aggregate, statistically ordinary—the routine toll of an activity that the fitness industry has been selling for fifty years as the foundational form of human exercise, the movement prescription of first resort, the thing you do when you want to get healthy and don’t know where to start.
The injury rate among recreational runners, measured consistently across three decades of epidemiological surveillance, approaches eighty percent annually. Not over a lifetime. Not cumulatively. Per year. In any given twelve-month period, roughly four out of every five recreational runners will sustain at least one injury significant enough to cause them to modify or cease training. This is not a contested figure drawn from a single outlying study. It is the convergent finding of a literature that spans multiple countries, multiple training volumes, multiple surfaces, multiple shoe technologies, and multiple decades of product innovation in footwear specifically designed—according to its manufacturers—to prevent exactly these injuries.
The tent at kilometre seven is not an anomaly. It is a data point.
Annual injury rate in recreational runners: ~50–75%. In ruckers at recommended loads: no comparable epidemiological data exists—because the injury mechanism (impact transient) is absent from the walking gait cycle.
Before going further, a necessary clarification. This chapter is not an argument against running. It is an argument against running as the default. The distinction matters.
Humans can run. Remarkably well, in fact. Dennis Bramble and Daniel Lieberman’s 2004 paper in Nature—the foundational statement of the Endurance Running Hypothesis—makes the case that Homo sapiens evolved capacities for sustained long-distance running that are matched by almost no other mammal on earth. The nuchal ligament that stabilises the skull during the impact cycle. The enlarged semicircular canals that maintain gaze stability at running speed. The Achilles tendon’s spring-like elastic energy storage. The reduced hair coverage and high sweat-gland density that solve the thermoregulation problem that stops quadruped predators cold after a short sprint. Persistence hunting—the practice of running prey animals to exhaustion over hours in midday heat—was not a fringe behaviour. The evidence suggests it was a primary hunting strategy for some populations of early Homo, and it worked because human runners, unlike the animals they pursued, could keep going.
The Endurance Running Hypothesis is real science. Persistence hunting was real. Humans evolved to be capable of running long distances. None of that is disputed here.
What is disputed is what “capable of” means as a prescription. The body can do many things it was not designed to do daily. The question is what happens when a specialized capacity—one that evolution equipped us with for high-stakes, episodic deployment—is reclassified by a modern fitness culture as routine maintenance.
Think of it this way: running is to the human body what fourth gear is to a truck. It works. It is powerful. It covers ground. But if you drive in fourth gear through every neighbourhood, at every speed, in every traffic condition, you will destroy the transmission. The truck was not built to live in fourth gear. It was built to use fourth gear when fourth gear is what the terrain demands. First gear is where you spend most of your time when the load is real and the distance is long. First gear is what the drivetrain was designed around.
Loaded walking is first gear. It is the gear that four million years of bipedal evolution built the human body to occupy for most of its locomotor life. Running is fourth gear—the gear that evolution kept available for pursuit and escape, for the moments when the stakes were high enough to justify the cost. The lie is not that running exists. The lie is that it became the default.
To understand why this matters—why it is more than a story about shin splints and registration fees—you need to understand something about what running actually does to the human body at the moment of ground contact, which is the moment when the body is at its most vulnerable and the injury risk is at its highest.
When the foot strikes the ground during running, it transmits a force. Not a gentle one. Jens Nilsson and Alf Thorstensson, working in Stockholm in the late 1980s, measured ground reaction forces across a range of locomotor speeds and documented with considerable precision what the body absorbs during each step of a run. At typical recreational running speeds, ground reaction force—the vertical force transmitted from the ground up through the foot, ankle, tibia, knee, and into the axial skeleton with every footfall—measures between 2.0 and 2.9 times body weight. A seventy-kilogram runner generates between 1,400 and 2,030 newtons of force with each foot contact. At a typical running cadence of 160 to 180 steps per minute over a sixty-three-minute 10-kilometre effort, that runner will accumulate somewhere between ten thousand and eleven thousand three hundred ground contacts. In ten thousand to eleven thousand three hundred separate moments, a force equivalent to two to three times their full body weight will travel up through their lower extremity.
What is equally important in Nilsson and Thorstensson’s work—and what is cited far less frequently than the force magnitudes themselves—is the loading rate: the speed at which that force is applied. The injurious variable in running biomechanics is not simply the peak force but the transient impact spike that precedes it—a rapid initial peak that occurs within the first fifteen to twenty milliseconds of ground contact and reflects the acceleration of the lower leg segments at the moment of heel strike. It is this impact transient, arriving faster than the neuromuscular system can generate protective co-contraction, that is mechanistically associated with stress fractures, tibial stress reactions, patellofemoral pain, and iliotibial band syndrome. The body can absorb very large forces applied slowly. It is the fast forces that break things.
Walking is different in a way that is not merely quantitative but qualitatively distinct in its mechanical character. Nilsson and Thorstensson measured walking ground reaction forces as well: 1.0 to 1.5 times body weight, delivered at a loading rate that is substantially lower than running’s impact transient, through a gait pattern in which there is never a flight phase—never a moment when both feet leave the ground simultaneously—and therefore never the controlled falling that makes running’s ground contact so mechanically aggressive. The human body, over four million years of bipedal locomotion, has been exposed to walking’s force profile for nearly the entirety of its evolutionary history. Running as a primary form of sustained locomotion is a recent experiment, and not one the musculoskeletal system is obviously well-suited to perform at the volumes modern recreational culture demands.
The injury site distribution in recreational running is not random, and it is not explained by individual variation in training quality or biomechanical efficiency alone. It follows from the physics. The iliotibial band—a thick rope of connective tissue running from the iliac crest to the lateral tibial plateau—is not a muscle and cannot be strengthened in the way a muscle can. It is a tensioning structure that functions within a specific range of hip abductor recruitment and knee flexion angle. When it becomes irritated, as it does in an estimated twelve to twenty percent of recreational runners annually, the irritation is the consequence of cumulative tensile loading at a specific point in the flexion-extension arc—loading that is a direct function of the number of ground contacts, which is a direct function of distance run, which is a direct function of training volume. You can optimise form. You can strengthen the hip abductors. You can modify your cadence. But at some point, volume is volume, and the structure will communicate its limit through the medium of pain. The plantar fascia, which absorbs the tension spike of heel-to-midfoot loading on every footfall and which operates as the passive tension element in the windlass mechanism of the foot arch, has no mechanism for downregulating its load exposure independent of reducing the load. Reducing the load means running less. Running less is not a message that sells shoes.
This is not an argument against running as a movement. Running exists. Humans do it. Some people do it for their entire lives without significant injury, though they are a minority that exercise scientists have tried, with limited success, to characterize and replicate. Running is a legitimate gear. The problem is that modern fitness culture has made it the only gear—the default, the obvious prescription, the thing a body is supposed to do when it wants to be healthy and fit. A truck can run in fourth gear through a neighbourhood. It is just very expensive on the transmission. That claim—that fourth gear is the appropriate default for daily locomotion—is what the injury epidemiology does not support, and what the evolutionary record, as subsequent chapters will argue at length, actively contradicts.
What the injury rate does not tell you
The biomechanical argument above is sound as far as it goes: running produces high impact forces, the injury rate is high, and those facts are mechanistically related. But the injury-rate argument contains a trap that this chapter, in earlier drafts, fell into. High acute injury rates do not mean that running destroys joints over the long term. On that question—the question of whether running produces lasting structural damage to cartilage and bone—the evidence runs in the opposite direction from what the previous paragraphs might lead you to expect.
In 2017, Eduard Alentorn-Geli and colleagues published a systematic review and meta-analysis in the Journal of Orthopaedic and Sports Physical Therapy that pooled data from 125,810 individuals across seventeen studies spanning three decades (Alentorn-Geli et al., 2017). The headline finding was this: recreational runners had an osteoarthritis prevalence of 3.5 percent. Sedentary non-runners had a prevalence of 10.2 percent. Competitive runners—the elite and near-elite who logged volumes comparable to Hooper’s eighty-one-kilometre-per-week cohort—came in at 13.3 percent. The dose-response curve was not linear. It was U-shaped. Too little movement was worse for the joints than moderate running. Too much running was worse than moderate running. But the sweet spot—the zone where most recreational runners actually live—was protective.
The finding is not isolated. Krysia Timmins and colleagues, also in 2017, published a separate meta-analysis in the American Journal of Sports Medicine and concluded that runners were approximately fifty percent less likely to require knee replacement than non-runners (Timmins et al., 2017). Grace Lo’s group, working with longitudinal data from the Osteoarthritis Initiative—a multi-site cohort study tracking joint health over years, not the snapshot of a single survey—found no association between a history of running and symptomatic knee osteoarthritis (Lo et al., 2017), and separately found that running did not accelerate structural progression in people who already had knee osteoarthritis (Lo et al., 2018). People with bad knees who ran were not making their knees worse. The cartilage, it turns out, is not a set of brake pads wearing down to nothing with every footfall. It is a living tissue that responds to cyclical loading with adaptive remodelling—up to a point.
This matters for the argument this book is making, and it requires an honest reckoning. The earlier version of this chapter implied, through selective emphasis and mechanical metaphor, that running’s impact forces produce cumulative, irreversible joint damage in most practitioners. That implication is wrong. Running’s injury problem is real, but it is primarily an acute soft-tissue and overuse problem—the iliotibial bands and plantar fasciae and tibial stress reactions of the injury tent—not a cartilage-destruction problem. The joints, at recreational volumes, are largely fine. In some populations, they may be better than fine.
What remains true—and what the chondroprotective evidence does not contradict—is that running’s acute injury rate is genuinely high, that those injuries disrupt training and quality of life in the short term, that the hormonal costs at high volumes are real, and that a movement pattern producing fifty-to-eighty-percent annual injury rates is not obviously the optimal default for a species that evolved to walk under load. The case for rucking does not require running to destroy cartilage. It requires only that rucking offers a more sustainable baseline for most people—lower injury risk, comparable or superior metabolic stimulus per unit time, and a force profile the musculoskeletal system was designed around. That case stands on its own, without needing running to be worse than it actually is.
But the story of running’s relationship to the human body does not end with biomechanics. The harder argument—harder to accept, harder to absorb into the cultural framework that most educated Westerners have built around exercise and health—involves hormones.
In 2016, a team at the University of Connecticut led by David Hooper published findings that were, in the context of everything the fitness culture had spent four decades telling men about cardiovascular exercise, quietly devastating. Hooper and his colleagues recruited a cohort of men who ran an average of eighty-one kilometres per week—serious recreational and competitive athletes, not outliers in competitive running circles, but well beyond the training volume of any average recreational runner, who typically logs fifteen to thirty kilometres per week. They measured testosterone, drawing blood samples every fifteen minutes from eight in the morning until noon, a protocol producing seventeen separate measurement points per subject and eliminating the interpretive ambiguity that plagues studies relying on a single morning draw.
What they found was that the runners’ testosterone was suppressed—not slightly, not within the variance of normal fluctuation, but significantly and consistently, at every single measurement time point, compared to sedentary control subjects who exercised very little. The p-value, for all seventeen time points, was equal to or less than 0.05. The endocrinological shorthand for what Hooper’s team was documenting had been named a generation earlier by Anthony Hackney, who had been tracking the hormonal profiles of endurance athletes since the late 1980s and had identified a condition sufficiently common and sufficiently distinct to merit its own nomenclature: the Exercise-Hypogonadal Male Condition, or EHMC.
The EHMC designation encodes something important about what Hackney was observing. Hypogonadism is the clinical term for inadequate gonadal function—in men, inadequate testosterone production by the testes. Clinical hypogonadism, typically defined by total testosterone below 300 nanograms per deciliter, is associated with reduced muscle mass, increased adiposity, impaired bone density, cognitive changes, mood dysregulation, sexual dysfunction, and elevated cardiovascular risk. It is a condition that endocrinologists treat as a medical problem. Hackney was observing men who were, by conventional medical criteria, functioning hypogonadal—not severely, not pathologically, but measurably—not because of disease or injury or age-related decline, but because of their exercise habits. Specifically, because of how much they ran.
What made Hooper’s 2016 findings particularly difficult to dismiss was not the suppression itself—that had been documented before, in smaller studies with less rigorous protocols—but the mechanism. In men with primary hypogonadism, low testosterone is accompanied by elevated luteinizing hormone (LH): the pituitary, detecting low testosterone, increases its signalling to the testes in an attempt to restore production. This feedback loop is the standard hormonal response to low testosterone, the body’s homeostatic correction mechanism. In Hooper’s EHMC cohort, luteinizing hormone was not elevated. The feedback loop was not engaged. The men who were running eighty-one kilometres per week were not simply failing to produce enough testosterone—their central regulation system had ceased demanding that they produce more. Either the testes had lost their capacity to respond to LH, or the hypothalamic-pituitary axis itself had downregulated its signalling, or both. The body had, in some functional sense, adapted to low testosterone as if low testosterone were the new set point.
Sit with that for a moment. These are not ill men. They are not men consulting their physicians about fatigue or low libido or muscle loss, though some of them are experiencing all three. They are men who are, by any conventional metric of lifestyle virtue, doing everything right. They are running five and a half days a week. They are lean. Their cardiovascular metrics are, in many cases, excellent. They are the people whom the public health establishment holds up as exemplars of preventive medicine in action. And their testosterone is chronically, measurably suppressed—not acutely, not transiently, but at baseline, at rest, when the last run was yesterday and the hormonal system should have had ample opportunity to recover.
The clinical significance of that suppression deserves more than a passing acknowledgment. Testosterone is not merely the hormone of libido and competitive aggression—though it governs both of those too. It is a master anabolic signal. It drives the synthesis of contractile proteins in skeletal muscle, governing the maintenance and growth of the muscle mass that determines metabolic rate, insulin sensitivity, and functional capacity across the lifespan. It stimulates osteoblast activity—the deposition of new bone matrix at sites of mechanical stress—and thereby governs bone mineral density in ways that become clinically critical in the fifth decade and beyond. It modulates the distribution of adipose tissue, favouring subcutaneous over visceral deposition, and visceral fat is the metabolic phenotype most tightly coupled to cardiovascular risk, type 2 diabetes, and systemic inflammation. It influences mood, cognitive processing, competitive drive, and risk tolerance in ways that the available neuroscience is still characterising but that are observationally robust across cultures and cohorts.
A man whose testosterone is chronically suppressed is not simply a man with a lower number on a lab panel. He is a man whose body is running a degraded version of its own maintenance software. He may feel it as fatigue that does not respond to rest, as the slow accumulation of body fat despite adequate exercise, as the creeping difficulty of maintaining muscle mass that used to seem effortless, as a mood floor slightly lower than he remembers from his thirties. Or he may not feel it at all, at least not in any form he would name as a problem, because the decline is gradual enough that each new level feels like the new normal. The body is very good at normalizing its own deficits, particularly when the narrative surrounding the behaviour producing those deficits is one of health and virtue.
Consider the man who surfaces in a thousand variations in the fitness culture of any medium-to-large Western city. He is somewhere between thirty-eight and fifty-two years old. He is educated—a professional, perhaps a lawyer or an engineer or a product manager, the kind of person who reads long-form journalism and owns a decent pair of running shoes, which he replaces every five hundred kilometres as instructed. He started running in his late twenties or early thirties, during the period of mild existential recalibration that follows the realization that youth is not indefinite and that the body, unattended, will quietly accumulate the evidence of sedentary eating and inadequate sleep. Running felt like the obvious answer. It was cheap. It required no gym membership, no specialized knowledge, no equipment beyond the shoes. It was social in the way that urban running culture had made it social, with peer groups and races and segments on mapping applications that converted a Tuesday-morning commute into something that could be measured and compared and shared.
He has been running three to five days a week for somewhere between eight and fifteen years. He runs thirty to fifty kilometres per week—not Hooper’s cohort-level eighty-one, and here an important caveat is required. The EHMC has been documented in men training at volumes above approximately eighty kilometres per week—roughly fifty miles. Whether it occurs at the thirty-to-fifty-kilometre range that characterises most recreational runners is not established. Hackney and Lane, in a 2018 review of the EHMC literature, explicitly scoped the condition to trained athletes exceeding seven hours of endurance training per week (Hackney & Lane, 2018). The average recreational runner, logging three to five hours per week, falls below that threshold. It would be dishonest to claim that the man running forty kilometres per week is experiencing the same hormonal disruption as the man running eighty-one. He may be experiencing some degree of cortisol elevation and some degree of testosterone-to-cortisol ratio shift, but the clinical significance at recreational volumes remains unclear. He has had, over those years, a succession of injuries: a bout of plantar fasciitis in his early forties that required six weeks off, a knee problem that an orthopedic surgeon described as “iliotibial band syndrome with some early patellofemoral changes” and that recurred twice in subsequent years, a stress reaction in the right tibia that was caught early enough to avoid frank fracture. Each injury interrupted his training. Each interruption was followed by renewed commitment, new shoes, possibly new form coaching or a new training plan downloaded from a running application. He returned to the same volume, more or less, because the volume was what he had built his identity around, and because the alternative—running less, or stopping—felt like a kind of surrender to the aging that running was supposed to prevent.
What this man does not know, because no one has told him and the annual physical does not measure it and the sports medicine practitioner who saw the iliotibial band was focused on the mechanical problem rather than the systemic one, is that the exercise he is performing as a hedge against aging may be accelerating several of the processes he is trying to slow. Testosterone declines naturally with age in men—the literature varies on rate, but the general consensus is approximately one to two percent per year from the mid-thirties onward, a trajectory that, over the arc of a life, produces meaningfully different outcomes in men who start that decline from a high baseline versus men who start from a suppressed one. The chronic endurance training that recreational runners perform imposes a chronic cortisol load—the catabolic hormone that rises with exercise stress and, when elevated persistently, opposes the anabolic effects of whatever testosterone remains. The testosterone-to-cortisol ratio, which exercise physiologists use as an index of the body’s anabolic-to-catabolic balance, trends in the wrong direction not because testosterone is irreplaceably lost but because the hormonal cost of the training stimulus is systematically underweighted in the cultural accounting of what running does to you.
He may be, at the higher end of this volume range, trading some of what he is trying to preserve for the cardiovascular gains he is accumulating. The trade-off is poorly characterised at the thirty-to-fifty-kilometre range. At eighty-one kilometres per week, the evidence for suppression is robust. At forty, it is plausible but unproven.
This qualification matters. The earlier version of this chapter stated the hormonal cost as a certainty at recreational volumes. That was an overstatement. What is certain is the mechanistic description of what the hormonal and biomechanical literature says about the body of a man performing sustained high-volume aerobic exercise without a significant resistance-type counterweight. Muscle mass and bone density in aging men depend on testosterone and on the mechanical loading that triggers osteogenesis and myofibril synthesis. Running provides the mechanical loading to the lower extremity, unevenly and at injury-producing force rates, but not the testosterone-sparing, resistance-type stimulus that would maintain the anabolic hormonal environment in which that loading could produce optimal adaptation. It provides a cardiovascular stimulus, genuinely, measurably—VO₂max responds to running training—but at a hormonal cost that compounds with volume and years in ways that no running magazine has ever found commercially advantageous to explain.
There is also a structural paradox embedded in the suburban runner’s relationship with injury that deserves unpacking. He understands, at some level, that the injuries are a problem. He is not indifferent to the weeks lost to plantar fasciitis, the months lost to the iliotibial band, the six-week rehabilitation protocol that followed the tibial stress reaction and that he followed imperfectly because stopping entirely felt psychologically intolerable. But the injury history has not dislodged the core premise. He has interpreted each injury as a technical failure—a training error, a biomechanical deficit, a shoe choice—rather than as signal from an activity with a structural injury rate that is independent of any individual runner’s choices. He has treated the symptom and preserved the diagnosis. And the running culture surrounding him has provided a sophisticated support structure for exactly this interpretive framework: books about form, coaches who specialize in gait analysis, shoe fitters who assess pronation, physiotherapists who treat running injuries as a practice specialty. Each of these services is genuinely valuable within its scope. Each of them also, structurally, reinforces the premise that the problem is the runner rather than the running—that the eighty percent figure is a population average from which any sufficiently educated, sufficiently attentive individual can exempt themselves through optimal behaviour. Some can. Most cannot. The evidence on this is also fairly clear.
People who run without injury and enjoy it should keep running. People who do HIIT and recover well should continue. This book does not argue that those modalities are broken — it argues that they are not the default, and that for the majority of people who have tried high-intensity approaches and found them unsustainable, injurious, or exhausting, there is a better baseline. Rucking is not the only answer. It is the answer that was here first.
There is a version of this argument that running advocates will recognise and preemptively dismiss: the claim that running is bad for you, the familiar complaint of the sedentary person who has read a headline and found in it permission to remain on the couch. That is not this argument. The couch is worse. The data on physical inactivity is unambiguous, and nothing in this chapter or this book should be read as advocacy for doing less. The argument is more specific and more structurally uncomfortable than that. It is that the particular form of exercise that the Western fitness culture has elevated to primacy—the default, the prescription, the thing you start with when you decide to get serious—is characterized by an injury rate so high that it approaches certainty over a multi-year horizon for most practitioners, and by a hormonal profile, at common training volumes, that is antagonistic to the very biological systems it is presumed to be preserving.
The argument is not that running cannot be done well, at appropriate volumes, with careful attention to load management and recovery, with a resistance training counterweight to offset the testosterone-suppressive effects of high-volume aerobic work. The argument is that running as it is practiced by the vast majority of recreational runners—in the volumes that fitness culture recommends, at the frequencies that training plans prescribe, in the social context of races and GPS segments and peer accountability that makes reducing volume feel like failure—is producing an outcome distribution that the available evidence does not support calling healthy.
The injury tent at kilometre seven is not incidental to the story of recreational running. It is the story. It is the physical manifestation of a statistical reality that the running industry has been structurally incentivized to minimize, contextualize, and ultimately ignore, because the alternative—acknowledging that the most commercially successful exercise product of the late twentieth century carries an eighty-percent annual injury rate and measurable hormonal costs at common training volumes—is not a message that supports the continued sale of technical fabric and foam midsoles and race registrations.
None of this is, in the strictest sense of the phrase, new information. The EHMC literature extends back to Hackney’s original work in the 1980s. The running injury epidemiology has been consistently documented since the 1970s, when running emerged as a mass-market consumer behaviour following Jim Fixx’s The Complete Book of Running and the jogging revolution it helped ignite. Nilsson and Thorstensson published their ground reaction force data in 1989. The temporal gap between when this evidence was produced and when it reaches the person lacing up their shoes at five in the morning is not an accident of academic communication lag. It is a structural feature of a fitness industry whose economic model depends on the continued belief that running is the obvious, natural, first-resort form of human cardiovascular exercise.
The cultural architecture supporting that belief is formidable. It includes a footwear industry that generated more than seventy billion dollars in global revenue in 2023, a race-registration market that sells the experience of suffering through a distance as a personal achievement, a social media ecosystem in which running metrics function as a form of public identity, and a public health establishment that has been recommending “at least 150 minutes of moderate-intensity aerobic activity” per week for so long that the specifics of what kind of aerobic activity have been largely washed out of the recommendation. Running is what people picture when they picture exercise. It is what health journalism photographs when it photographs fitness. It is the default.
But a default is not a biological recommendation. It is a cultural artifact—the accumulated residue of marketing decisions, social signalling, and the specific historical moment when a generation of post-industrial knowledge workers needed an activity that could be performed alone, in the margins of an urban schedule, without equipment or instruction, and that produced visible evidence of effort: the lean body, the Garmin data, the finisher’s medal. Running provided all of these things. That it also provided injury rates approaching eighty percent annually and measurable hormonal suppression at common training volumes was a secondary feature that the architecture of running culture was perfectly designed to render invisible.
The invisibility operates through several mechanisms. The injury tent at kilometre seven is staffed by volunteers who are themselves runners and who frame the treatment they provide as a normal part of athletic participation—the cost of pushing hard, the badge of commitment, the tolerable risk of a sport that builds character. The man with the iliotibial band syndrome visits a sports medicine practitioner who treats the knee and does not measure the testosterone, because treating the knee is the presenting complaint and testosterone is not in the differential. The woman whose training log shows a progressive increase in weekly mileage correlated with a progressive increase in menstrual irregularity attributes the irregularity to stress from work, not from training, because no one has given her a framework for thinking about the relationship between exercise volume and hypothalamic-pituitary-gonadal axis function—a topic that, in women, carries its own set of suppressive dynamics with its own distinct evidence base and its own history of being underdisclosed.
The injuries are treated. The hormonal costs are invisible. The industry continues.
What running culture has never fully reckoned with is this: humans evolved to run, yes—but they evolved to run for specific, high-stakes, episodic reasons, embedded within a locomotor life that was overwhelmingly dominated by walking. Specifically, by walking while carrying weight, over long distances, at a pace sustainable for hours. The Endurance Running Hypothesis that Bramble and Lieberman documented does not describe a species that ran daily as its primary form of locomotion. It describes a species that could deploy running as a specialized capability when persistence hunting or escape from predators demanded it—and that otherwise moved through the world on foot, loaded, at walking pace. The evolutionary anthropology of this will occupy an entire chapter later in this book, but the essential finding—documented in the bipedal gait record, in the comparative musculoskeletal anatomy of australopithecines and Homo, in the foot anatomy of load-bearing hunter-gatherers, in Cara Wall-Scheffler’s biomechanical work on optimal carrying postures—is that the human body’s architecture is optimized for weighted ambulatory locomotion, not for the unloaded, high-cadence, high-impact pattern that recreational running produces.
The gluteus maximus, the largest muscle in the human body, is enlarged in Homo sapiens far beyond anything explicable by the demands of ordinary unloaded bipedalism. It is enlarged specifically for the demands of carrying loads on the torso—the posterior chain activation required to stabilise the spine and pelvis against the forward moment of a loaded pack—and for climbing, which similarly loads the posterior chain in ways that flat-road running does not. The human foot’s arched structure, the spring mechanism of the plantar fascia, the Achilles tendon’s extraordinary elastic recoil capacity: these features support walking efficiency under load, and they support running, but the loading patterns are different, and the injury distributions are different, and the evolutionary exposure history is dramatically different. The savanna record, parsed through the metabolic cost literature of walking versus running, suggests that our ancestors walked on the order of ten to eighteen kilometres per day—typically with some form of carried load—and ran in short bursts for specific purposes (pursuit predation, escape from predators, inter-group conflict) rather than as a sustained daily practice.
Running culture has taken a gear that evolution designed for short-burst, high-stakes deployment—pursuit, escape, the emergency situations where covering ground fast was worth the physical cost—and has attempted to industrialize it into a daily default. The eighty-percent annual injury rate is, from this perspective, less surprising than the widespread assumption that it should be lower. It is not a mystery. It is fourth gear in the neighbourhood. Run it long enough and the transmission tells you.
Stand outside the injury tent at kilometre seven as the last finishers move past, and look at the geometry of the scene. Four thousand people—four thousand—who paid a registration fee, set an alarm for five-thirty in the morning, drove or took public transit to a starting line, pinned a number to their chest, and ran ten kilometres through a city in October because they believe, with genuine conviction, that this is what a healthy person does with their body. The belief is sincere. The commitment is real. The effort is not in question. What is in question is whether the activity they have committed to is serving the biological objective they have assigned it.
The people who will leave the injury tent with mylar blankets and ice packs and mild shame are not statistical outliers. They are the norm. They are the eighty percent, or their proportion of it, represented at this race on this morning. And the people who finished the race without acute injury are not necessarily the success cases: many of them are running with chronic biomechanical adaptations—the slight asymmetry in hip drop, the compensatory overstriding, the knee pain managed with ibuprofen on long-run days—that do not appear in the injury tent count because they have not yet reached the threshold of acute presentation. They will. On average, they will.
The ground is shifting under a familiar assumption. The assumption is that running is the obvious, natural, evolutionarily appropriate default exercise for human beings who want to be healthy, and that the injuries are a cost of participation rather than evidence of fundamental mismatch between the activity and the body performing it. The evidence assembled in the following pages will argue, with some force, that this assumption has the story backwards. The injuries are not incidental. The hormonal costs are not exceptional. The mismatch is not a performance problem to be solved with better shoes, better form coaching, and better periodization.
The mismatch is the story.
And it begins, as most good stories do, with the question no one thought to ask: if not this, then what?
The answer is older than civilization. It is older than agriculture, older than the wheel, older than the species, arguably. It is the movement that preceded running in the fossil record by at least a million years and that the human musculoskeletal system, if you ask it honestly through the medium of the biomechanical and endocrinological literature, was designed to perform. It does not require a race. It does not require a GPS watch or a foam midsole or a technical fabric jersey in a moisture-wicking polyester blend. It requires a pack, a pair of shoes, and the willingness to close your mouth and walk.
But before the solution, you need to fully understand the problem. The injury tent is one dimension of it. The hormones are another. And behind both of them, sustaining the cultural conditions that have made high annual injury rates invisible and hormonal costs poorly understood, there is an industry—and its history—that requires its own chapter.