What the Military Knows
Somewhere in the archive stacks of the U.S. Army Research Institute of Environmental Medicine in Natick, Massachusetts, there are filing cabinets and hard drives containing what is arguably the most comprehensive and long-running body of human performance research ever conducted. It spans more than a century. It has been funded by billions of dollars across successive defence budgets. It has employed physiologists, biomechanists, orthopedic surgeons, endocrinologists, neuroscientists, and psychologists. It has used subjects ranging from West Point cadets to British Royal Marines to Nepali Gurkhas to Israeli paratroopers. It has measured nearly every variable the human body produces under physical stress: ground reaction forces, cortisol gradients, paraspinal muscle cross-sectional area, heart rate variability indices, tibial acceleration profiles, joint contact forces, anabolic hormone trajectories, autonomic nervous system state.
The subject of all this research is a man or woman walking with a heavy pack on their back.
You have almost certainly never heard of most of it.
This is not an accident. It is a consequence of economic structure. The fitness industry’s model is subscription, novelty, and planned complexity. There is no revenue in telling someone to put weight in a bag and walk. There is no quarterly product cycle in the scientific literature on military load carriage. There are no influencer partnerships, no certification programs, no app integrations, no proprietary heart rate zone algorithms, no branded electrolyte sachets that pair synergistically with the protocol. When the U.S. Army Research Institute publishes a biomechanical study on tibiofemoral contact forces during loaded walking, it does so to protect soldier lives, not to monetize a fitness trend. The findings go into the journal. The journal sits on a shelf. Civilians keep buying gym memberships.
What has been lost in this translation from military laboratory to public understanding is not merely a collection of useful findings. What has been lost is an entire science of the loaded human body in motion—a science that, had it been absorbed into mainstream exercise prescription over the past three decades, would have spared tens of millions of people from unnecessary injury, hormonal suppression, metabolic inefficiency, and the creeping functional decline that the fitness industry has managed to monetize without ever actually preventing.
This chapter is an account of what that science knows.
The Long History of an Unfashionable Problem
Military institutions have been studying the biomechanics of load carriage for as long as they have had the organizational and scientific capacity to do so, because the problem is urgent in ways that fitness trends never quite are. Soldiers who cannot march cannot fight. Soldiers who arrive at the objective with broken metatarsals, stress-fractured tibiae, wrecked knees, and suppressed immune function because of the weight they carried getting there are a strategic liability. The incentive to understand how the human body responds to carrying weight over distance is not academic curiosity. It is operational necessity.
The modern history of this research begins in earnest in the late nineteenth century, when Archduke Ludwig Salvator of Austria, better remembered as an unlikely pioneer of ergonomics than as royalty, published observations on the metabolic costs of load carrying across different postures and methods. His work predates the germ theory consensus in medicine and yet anticipates questions that biomechanists are still refining today. By the First World War, British and American military physiologists were systematically studying march performance, ration requirements, blister formation rates, and the load-fatigue relationship in infantry soldiers. The quantitative data were primitive by contemporary standards, but the questions being asked were exactly right: how much can a man carry, at what pace, for how long, before performance degrades and injury risk rises beyond operational tolerance?
The interwar period produced the first controlled laboratory studies. By the Second World War, the U.S. Army Quartermaster Corps had established a research program explicitly devoted to optimizing load carriage systems—not merely for weight reduction but for biomechanical efficiency. The question of load distribution, of where on the body the mass should sit, was understood even then to matter as much as the absolute magnitude of the load. Soldiers who carried the same weight distributed differently walked differently, fatigued differently, and broke down differently. This was, in embryonic form, the science of load carriage biomechanics.
The postwar decades brought progressively more sophisticated instrumentation. Force plates. Electromyography. Motion capture. Indirect calorimetry. By the 1980s, researchers at Natick, at the Defence and Civil Institute of Environmental Medicine in Canada, and at counterpart institutions in the United Kingdom, Israel, and Australia were producing studies of genuine quantitative precision on how loaded walking alters ground reaction forces, joint kinematics, metabolic demand, musculoskeletal loading rates, and cardiovascular response. The scientific questions had become granular: not merely “how much can a soldier carry?” but “at what exact load and speed does the tibiofemoral contact force cross the threshold associated with injury risk?”, “how does trunk flexion angle change with increasing load, and what are the downstream consequences for lumbar spine loading?”, and “what happens to the autonomic nervous system after three weeks of daily load carriage—and does it matter?”
By the 2010s, this research program had accumulated a literature of several thousand peer-reviewed studies. The subjects were predominantly military, predominantly male, and predominantly young—the demographic realities of the armed forces creating a systematic bias in the evidence base that we will return to. But within those demographic limits, the science was exceptionally detailed, well-funded, and rigorously conducted. It constituted, in effect, a parallel universe of exercise physiology: one in which the central movement was not running, cycling, or lifting, but loaded walking, and in which the practical stakes of getting the science wrong were not reduced gym attendance but operational casualties.
That parallel universe has remained almost entirely invisible to the civilian population.
What a Systematic Review Finds
In 2021, Gillian Walsh and David Low published what remains the most comprehensive synthesis of military load carriage biomechanics in the peer-reviewed literature: a systematic review in the journal Applied Ergonomics titled “Military load carriage effects on the gait of military personnel.” The paper synthesized evidence from decades of controlled trials and observational studies, and its findings describe something considerably more interesting than the civilian fitness industry has ever thought to investigate.
Loaded walking, Walsh and Low document, is not merely walking with extra weight. It is a categorically different locomotor mode that activates distinct neuromuscular patterns, produces different ground reaction force profiles, and generates physiological adaptations that unloaded walking simply cannot replicate. The review identifies a consistent set of gait changes that occur when a soldier—or anyone—walks with a significant pack: increased trunk flexion, increased vertical ground reaction forces proportional to the load, increased cadence relative to step length, and elevated activation of the posterior chain musculature including the erector spinae, trapezius, and gastrocnemius-soleus complex. These are not merely compensatory adjustments. They are the signatures of a distinct movement pattern that engages the human body’s load-bearing architecture in a way that neither unloaded walking nor running replicates.
Perhaps more importantly, Walsh and Low document that load carriage conditioning produces measurable physiological adaptations across multiple systems simultaneously: greater maximal oxygen uptake, improved upper-body endurance, increased lower-body strength, and enhanced neuromuscular control in the mediolateral plane. The person who rucks regularly is not merely maintaining one system. They are developing several in parallel, through a single movement performed within a single session.
The systematic review also establishes a biomechanical argument that deserves far wider circulation than it has received. Ground reaction forces during loaded walking—even with substantial packs—follow the characteristic double-humped force curve of walking gait, in which the body’s weight is transferred gradually across the stance phase. Running, by contrast, produces a sharp impact transient at initial contact, followed by a single active peak, with loading rates that are three to five times higher than those observed during loaded walking. The loaded walker at 20 kilograms is applying more total force to their joints than the unloaded runner—but they are applying it more slowly, across a longer ground contact time, through a movement pattern their skeletal and connective tissue architecture was specifically evolved to handle. The loaded runner, to give the comparison its full weight, is doing neither. They are applying ballistic forces to joints that are unprepared for ballistic loading at every stride, at an intensity their posterior chain cannot sustain, in a movement pattern that produces musculoskeletal injury rates approaching 80 percent annually in recreational populations.
Walsh and Low do not editorialize in this direction. They are scientists producing a systematic review, not advocates for a training philosophy. But the data they synthesize point clearly toward a conclusion the fitness industry has conspicuously failed to reach: that loaded walking is not a primitive version of running, to be replaced with real cardio when you get fit enough. It is a fundamentally different modality with a superior safety profile, a distinct and clinically significant set of adaptations, and a biomechanical basis in human evolutionary history that running simply does not have.
The Joint Stiffness Evidence
The 2021 study by Tiago Santos and colleagues in the Journal of Applied Biomechanics addresses one of the most mechanistically important questions in the loaded walking literature: what happens to the dynamic stiffness of the lower-limb joints when you add a significant load to a walking body?
The finding is, in the strictest sense, predictable from first principles. What makes the Santos study valuable is that it quantifies the prediction with precision and identifies something unexpected in the pattern of the response.
Santos and colleagues had participants walk with loads equivalent to 30 percent of their body mass and measured the dynamic stiffness—the ratio of force change to displacement change—at the ankle, knee, and hip joints throughout the gait cycle. Both the ankle and knee showed significant increases in dynamic stiffness under load: ankle stiffness increased with a p-value of 0.002; knee stiffness with a p-value of less than 0.001. The hip, however, showed no significant stiffness increase at all.
This joint-specific pattern is physiologically elegant in a way that matters for understanding what rucking actually does to the body. The ankle and knee stiffen under load—they become more resistant to deformation, which is the characteristic adaptive response of tendons and ligaments to chronic mechanical loading. This is how connective tissue gets stronger. This is Wolff’s Law operating not on bone but on the soft tissue structures that support and govern joint function. Chronically applied, this stimulus produces denser collagen cross-linking, thicker and more mechanically robust tendons, and joints that tolerate the demands placed on them with greater reserve. The hip, by contrast, maintains its compliance, which preserves the range of motion necessary for efficient propulsive mechanics and appropriate stride length. The body, it turns out, has a sensible opinion about which joints should stiffen under load and which should remain free.
What Santos and colleagues are documenting is, at a mechanistic level, the structural toughening process. The runner who accumulates impact loading without adequate connective tissue preparation breaks down—patella tendinopathy, Achilles rupture, iliotibial band syndrome, plantar fasciitis. The rucker who accumulates the same training volume applies lower loading rates to tendons and ligaments that are progressively stiffening and strengthening in response to the chronic mechanical stimulus. The injury epidemiologies of these two populations are not comparable.
None of this appears in the brochure for any fitness subscription product currently on the market.
The Testosterone Finding
In 2016, a research team led by Marcus Taylor published a study in the journal Steroids with a title bland enough to guarantee it would be ignored by anyone without a subscription and a specific reason to look: “Anabolic hormone profiles in elite military men.”
Taylor and colleagues characterized the daily profiles of testosterone and dehydroepiandrosterone in a cohort of elite military men with a mean age of thirty-three years. They used salivary sampling across the diurnal arc, capturing the morning peak, the mid-morning decline, and the afternoon trough. Then they compared these profiles to published normative data from three other populations: young, healthy male recreational weightlifters (mean age eighteen years), male university students with resistance training experience (mean age twenty-four years), and young recreationally weight-trained men (mean age twenty-two years).
The military men’s testosterone profiles matched or exceeded all three comparison groups.
Thirty-three-year-old men whose primary occupational physical activity is carrying heavy loads over distance had salivary testosterone concentrations indistinguishable from, and in some parameters superior to, men ten to fifteen years younger who were specifically training to maximize muscular hypertrophy. The finding is not ambiguous.
Taylor and colleagues were careful about causation. They note that elite military selection processes create survivorship bias—the men in their cohort were the physiologically elite subset of an already self-selected population. The testosterone profiles they observed may partly reflect baseline hormonal superiority rather than training-induced maintenance. But the magnitude and direction of the finding are hard to dismiss. These men were not doing marathon training. They were not doing HIIT. They were doing loaded marches, carrying heavy equipment over variable terrain at sustained moderate intensities. And their hormonal profiles looked like those of recreational weightlifters.
Compare this to the Exercise-Hypogonadal Male Condition, which is the clinical term for the well-documented phenomenon of chronic testosterone suppression in high-volume male distance runners. Hooper and colleagues, writing in Medicine and Science in Sports and Exercise, documented that men averaging eighty-one kilometres of running per week showed significantly reduced testosterone concentrations at all time points compared to sedentary controls, without compensatory changes in luteinizing hormone—suggesting not merely an adaptive downregulation but something closer to testicular insufficiency or central hypothalamic dysfunction. Blood was drawn every fifteen minutes for four hours. The suppression was not acute. It was the chronic resting state of men who ran eighty-one kilometres per week.
Here are two bodies of evidence that, placed side by side, constitute a fairly clear argument. Men who carry heavy loads over distance maintain testosterone profiles comparable to young weightlifters. Men who run very long distances chronically suppress their testosterone below that of sedentary controls. The fitness industry has spent thirty years promoting distance running as the cardiovascular gold standard and barely mentioning either finding.
The question is not whether the fitness industry is corrupt. The question is whether the incentive structures of a $100-billion-per-year commercial enterprise reliably produce accurate information about which exercise modalities serve human health. The evidence presented above suggests they do not.
What Twenty Days Does to a Spine
In 2020, Qu and colleagues published a study examining what happens to the paraspinal musculature—the deep spinal stabilizers that maintain erect posture and protect the intervertebral discs from pathological loading—when physically active men undergo twenty days of load carriage training.
The methodology was specific: participants carried loads equivalent to approximately 30 percent of body weight during structured daily marches. Before and after the twenty-day protocol, the researchers measured paraspinal muscle cross-sectional area using imaging and quantified heart rate variability using standard time-domain and frequency-domain metrics.
The structural adaptation was measurable and significant. Mean paraspinal muscle cross-sectional area increased from 9,126 ± 692 square millimetres to 9,863 ± 456 square millimetres over the twenty-day period. This is an improvement of roughly eight percent in twenty days—in muscles that are notoriously resistant to conventional exercise interventions because they are barely recruited by the exercises most people actually do. Crunches do not train the paraspinal musculature with any particular efficiency. Neither does cycling. Nor does running, which produces relatively modest trunk extensor demands because the body must stay relatively upright for efficient propulsion. Load carriage, by imposing a persistent forward moment from the pack that the spine must counteract to maintain upright posture, is one of the few aerobic exercise modalities that chronically and substantially loads the erector spinae and multifidus groups in the way these muscles require to adapt.
These are not minor muscles. The paraspinal musculature is the primary dynamic stabilisation system of the lumbar spine. Dysfunction in the deep multifidus—the innermost layer of the paraspinal group—is one of the most consistent findings in patients with chronic low back pain. The muscle atrophies preferentially in low back pain populations, and the atrophy is often unilateral, reflecting the side-specific nature of the original injury or dysfunction. Restoring multifidus volume is associated with pain reduction and improved functional outcomes. The exercise physiology literature knows this. The clinical physiotherapy literature knows this. What neither literature has adequately communicated to the general public is that sustained loaded walking is one of the most efficient methods of delivering the exact mechanical stimulus these muscles require.
The cardiovascular finding from Qu et al.’s study is at least as interesting as the structural one. All measured heart rate variability indices—SDNN, RMSSD, the low-frequency band, and the high-frequency band—increased from the first to the last day of the protocol, all at p less than 0.03. This directional pattern is the signature of parasympathetic adaptation. The autonomic nervous system, across twenty days of daily load carriage, shifted toward greater parasympathetic tone at rest: a lower-noise, higher-resilience cardiac regulation that is associated in the epidemiological literature with reduced all-cause and cardiovascular mortality.
This is not a minor finding. Resting heart rate variability is one of the most robust biomarkers of cardiovascular health currently available. Higher SDNN and RMSSD predict lower risk of sudden cardiac death, lower incidence of major adverse cardiovascular events, and better autonomic recovery from exercise stress. The fact that twenty days of moderate-intensity loaded walking is sufficient to produce measurable improvements in these metrics in already physically active men speaks directly to the potency of the stimulus.
Lowe and colleagues, writing in 2025 in Medicine and Science in Sports and Exercise, extended this observation in a more controlled military simulation context: thirty-two physically active males carried loads of approximately 30 percent of body weight daily over simulated military operations cycles of varying stress intensity. Exercise HRV metrics increased across both low-stress and high-stress cycles, with all metrics reaching statistical significance at p less than 0.03. The autonomic adaptation to chronic load carriage was not abolished by the additional stressors of sleep restriction and elevated operational demand. The cardiovascular system adapted in the direction of greater parasympathetic resilience despite conditions that would be expected to produce sympathetic dominance. The loaded walking was, in some sense, buffering the system against the worst effects of the stress.
What the fitness industry sells as cardiovascular training is predominantly sympathetic activation—elevated heart rate, high-intensity intervals, lactate accumulation, the familiar misery of a spinning class. What chronic load carriage produces, apparently, is something categorically different: not the acute stress of an elevated heart rate but the chronic adaptation of a more resilient autonomic system. The distinction matters because sympathetic activation and parasympathetic resilience are not the same thing. You can chronically stress the sympathetic nervous system without ever building a more robust parasympathetic counterbalance. Endurance running does this at high volumes, which is why marathon runners have heart rate variability profiles that sometimes resemble those of chronically stressed office workers rather than elite athletes. Load carriage, at the intensities the military literature has studied, seems to do something different.
The Women the Research Forgot—and One Study That Did Not
There is a critical limitation in the military load carriage literature that must be stated plainly: the overwhelming majority of it was conducted on young men. This is a product of military demography, not scientific indifference, but the consequence is that the evidence base for loaded walking in women is substantially thinner than the evidence base for men, and the sex-specific differences that do appear in the literature carry proportionally greater weight because they are rarer.
One finding from 2025 is worth particular attention in this context. Willy and colleagues, studying patellofemoral joint mechanics during loaded walking, documented that females exhibit systematically greater patellofemoral joint stress than males when carrying equivalent loads. The patellofemoral joint—the articulation between the kneecap and the femoral groove—is the joint most commonly associated with overuse injury in running populations, and it is already subject to greater baseline stress in females due to the wider pelvis-to-knee-width ratio that characterises the female skeleton. When external load is added, this sex difference in patellofemoral stress is amplified rather than resolved.
This finding does not argue against rucking for women. It argues, with considerable precision, for sex-informed rucking prescription. The load progression that is appropriate for a man in the early stages of a rucking programme may not be appropriate for a woman. The knee mechanics of loaded walking in women require specific attention to load magnitude, walking speed, and the biomechanical cues—knee alignment, cadence, trunk position—that modulate patellofemoral joint stress. A woman who develops anterior knee pain in the first weeks of a rucking programme is not demonstrating that rucking is wrong for her. She is demonstrating that the programme was not designed with her anatomy in mind.
Johnson and colleagues, studying U.S. Army trainees in 2024, found that female trainees demonstrated significantly higher vertical loading rates during ruck marching than males, along with higher peak tibial accelerations during both running and loaded walking. The magnitude of these loading rates—not merely the peak forces but the rate at which those forces are applied—was substantially lower during rucking than during running in the same female subjects. Loaded walking, even for women whose biomechanics differ meaningfully from men’s, operates within a safer mechanical domain than running. The injury epidemiology of military training populations, which is extensively documented, confirms this pattern: rucking-related musculoskeletal injuries are real and non-trivial, but they are front-loadable, preventable with sensible progression, and substantially less severe than the injuries produced by running-based training programmes of equivalent volume.
What Willy’s finding does is not disqualify rucking for women. It provides the kind of specific, mechanistically grounded sex-difference data that should be driving different prescription recommendations for men and women—recommendations that the fitness industry, which prefers one-size-fits-all protocols, has never had much incentive to develop.
The military research program has studied women in this context, but insufficiently and relatively recently. The evidence gaps in the female-specific loaded walking literature are real and consequential. They are addressed directly in Part II of this book. For now, what matters is establishing that the military science, for all its demographic limitations, has produced findings that are directly and immediately relevant to female ruckers—and that those findings have been almost entirely absent from the public conversation about women’s exercise.
What the Funding Structure Reveals
It is worth pausing here to consider why this literature—centuries of accumulated military exercise science, thousands of peer-reviewed studies, genuine and compelling findings about human performance under load—has remained so completely invisible to the civilian population.
The answer is not ignorance. Many of the researchers involved in this work have published in journals that are accessible to the general public. The findings are not classified. The science is not obscure. The Walsh and Low systematic review is available in its entirety at no cost through most university library systems, and many of the individual studies it synthesizes are available through PubMed without a paywall. Taylor et al. is in a journal called Steroids, which lacks glamour but is indexed, searchable, and freely available.
The answer is incentive. Military research institutions study load carriage because soldiers carry loads and because dead soldiers are expensive in every sense. The Department of Defence funds this research not to sell a product but to protect an asset. The findings accumulate in the literature, cited by other military researchers, occasionally picked up by clinical physiotherapists working with back pain populations or physical therapists managing military patients, and otherwise largely ignored by an industry whose business model requires you to believe that exercise science is complex, specialised, rapidly evolving, and necessarily mediated by professional instruction, subscription products, and certified expertise.
The paradox is brutal in its clarity. The most expensive, most rigorous, most practically validated body of human exercise science in existence concerns an activity that requires no equipment more sophisticated than a bag and some weight to put in it. The fitness industry has built a $100-billion-per-year enterprise on convincing you that this cannot be sufficient.
The military has been quietly proving otherwise for a hundred years.
The Architecture of Ignorance
There is a subspecialty of the sociology of knowledge concerned with how information fails to move from the settings where it is generated to the settings where it would be most useful. The loaded walking literature is a case study in this kind of structured ignorance. It is not that the knowledge does not exist. It is that the institutional channels through which knowledge ordinarily moves from laboratory to practice have, in this case, systematically failed to carry it.
The pathway from peer-reviewed exercise science to public practice typically runs through one of several intermediary institutions: academic professional bodies that produce clinical guidelines, commercial fitness companies that fund applied research, popular media that translate findings for general audiences, or professional trainers and coaches who keep current with the literature and communicate it to clients. Each of these intermediaries has structural reasons to amplify some findings and suppress others.
Professional bodies produce clinical guidelines that reflect the conditions their members most commonly treat. Physical therapists see load-related injuries. They do not, systematically, see people who have been rucking for twenty years with intact joints and optimal HRV profiles, because those people are not patients. The clinical literature on load carriage is therefore weighted toward injury documentation rather than benefit characterisation.
Commercial fitness companies fund research that supports their product lines. There is no treadmill company with a financial interest in funding a study demonstrating that unloaded running is inferior to loaded walking for joint health and hormonal maintenance. There is no gym chain whose business model benefits from a systematic review showing that a backpack and a pair of shoes produce greater cardiovascular adaptation than a monthly membership. The research that gets commissioned, amplified, and translated into consumer-facing content is the research that creates demand for subscription products.
Popular media translates findings that are novel, counterintuitive, and narratively compelling—or that are supported by celebrity endorsement or lifestyle brand association. The Taylor et al. finding that military men maintain testosterone profiles comparable to young weightlifters is, in fact, a striking and counterintuitive result. It should have been widely reported. It was not, because the military populations it describes are not the demographic that fitness media targets, and because the activity it implies—putting a heavy pack on your back and walking—generates no advertising revenue for anyone.
Professional trainers and coaches vary enormously in their engagement with the primary literature. The trainers who are well-read and intellectually curious are aware of the military load carriage research. Some of them prescribe rucking. But the certification programs through which most personal trainers acquire their qualifications are owned and operated by organisations whose revenue depends on the complexity and perceived specialisation of their certification offerings. A science that tells you the best exercise requires minimal equipment and no specialist supervision is a science that undermines the certification business model.
What you are left with is a perfect storm of structural incentives arrayed against the dissemination of one of the most robust and practically useful bodies of exercise science ever produced.
What the Military Also Knows: The Injury Cost
The preceding sections present the benefits the military has documented. An honest account requires presenting the costs the military has also documented—and those costs are substantial.
Orr and colleagues, in a 2014 study published in the Journal of Occupational Rehabilitation, surveyed Australian Army soldiers and found that load carriage was among the most frequently cited causes of musculoskeletal injury (Orr, Johnston, et al., 2014). In a subsequent 2016 study, thirty-four percent of soldiers reported injuries directly attributed to load carriage, with the lower back, knees, and ankles the most commonly affected sites (Orr et al., 2016). A 2013 narrative review of the broader military injury literature confirmed that load carriage injuries are not occasional misfortunes but systematic consequences of the activity at operational loads and volumes (Orr, Pope, et al., 2014).
These are not trivial numbers. A thirty-four percent injury rate in a young, physically selected, professionally supervised population should give pause to anyone extrapolating military load carriage findings to unsupervised civilian ruckers—particularly older adults without the baseline conditioning, medical support, or progressive training infrastructure that military institutions provide.
The military populations studied in this literature are young (typically 20-31 years), physically pre-screened, predominantly male, and operating under professional medical supervision with structured progression protocols. Extrapolating their outcomes—whether benefits or injury rates—to a sedentary fifty-year-old beginning a rucking programme requires explicit caution. The carryover is plausible, not proven.
The injury epidemiology also reveals a dose-response relationship that the civilian rucking community should take seriously. Military loads routinely exceed forty percent of body weight—often approaching sixty percent or more during extended operations. The loads recommended in this book (fifteen to thirty percent of body weight for experienced ruckers, five to ten percent for beginners) are substantially lower. Whether this dose reduction proportionally reduces injury risk has not been formally studied in civilian populations, but the biomechanical modelling (Xu 2016) suggests that the relationship between load and joint reaction forces is nonlinear and that the recreational loading range sits below the thresholds associated with the most serious military injuries.
The military knows that loaded walking works. The military also knows that loaded walking hurts people when loads are too heavy, progression is too fast, or individual biomechanics are ignored. Both findings belong in the same chapter.
What the Science Actually Says
Let us be precise about what the military load carriage literature establishes, and what it does not.
It establishes, with systematic-review-level evidence, that loaded walking produces a distinct set of gait and neuromuscular adaptations that cannot be replicated by unloaded walking or running. It establishes that the ground reaction force profile of loaded walking—higher total force than unloaded walking, but lower loading rates than running—occupies a biomechanically distinct and arguably optimal zone for stimulating connective tissue adaptation while minimising injury risk. It establishes that twenty days of daily load carriage at 30 percent of body weight produces measurable increases in paraspinal muscle cross-sectional area and systematic improvements in heart rate variability across multiple domains. It establishes that men engaged in regular load carriage maintain testosterone profiles comparable to recreational weightlifters decades their junior. It establishes that dynamic joint stiffness at the ankle and knee increases significantly under load, providing the mechanical stimulus for connective tissue strengthening. It establishes that women experience greater patellofemoral joint stress under equivalent loads than men, a finding with direct implications for sex-differentiated prescription.
What the literature does not establish—and this matters for intellectual honesty—is a directly controlled comparison of chronic rucking versus chronic running in a civilian population across a lifespan. That study does not exist. The military populations studied are younger, more physically resilient, and more highly selected than the general civilian population. The carryover from military to civilian context requires inference, not merely extrapolation.
But inference from a large, well-funded, rigorously conducted literature is not speculation. It is the normal epistemic condition of most applied science. We do not have randomised controlled trials for wearing seatbelts. We have mechanistic understanding, observational data, and inference from biomechanical first principles. The loaded walking literature is considerably more extensive than that. The weight of inference it supports is considerable.
What it says, in aggregate, is this: human beings have an exercise modality specifically calibrated to their evolutionary anatomy, their hormonal physiology, their connective tissue mechanics, and their cardiovascular architecture. It is the modality their bodies were designed for. It has been studied with extraordinary precision and at extraordinary expense by institutions with no commercial interest in the findings. Those institutions have consistently found that the modality works—that it produces the adaptations it should produce, protects the structures it should protect, and maintains the physiological capacities it should maintain, across populations ranging from elite military operators to middle-aged civilians to elderly women with osteopenia.
And almost nobody told you.
The gap between what the military knows and what the public practices is not a minor oversight in science communication. It is a structural failure with measurable consequences—high injury rates in recreational runners, osteoporosis in populations given inadequate mechanical stimulus, hormonal disruption poorly understood by the clinicians diagnosing it—and a fitness industry that has profited from the complexity it manufactures to obscure a simpler, older, and in many respects more effective truth.
Put weight on your back. Walk. Do it regularly. Do it for a long time. Start lighter than your ego wants. Progress slower than your enthusiasm demands.
The Army figured out the first part a century ago. The injury data taught it the second part. Both lessons belong to you.