Loading for Bone Density: What the Science Says About Starting Young Horses
The equine industry has long utilized the science of bone modeling to justify the early training of juvenile horses. The logic is grounded in a fundamental biological truth: juvenile bone is highly plastic, and targeted load-bearing exercise is a mechanical necessity for developing the bone mineral density required for an athletic career (Logan & Nielsen, 2021). On this point, the science is clear. Bone adapts to the stresses placed upon it by becoming stronger and denser.
However, the decision to start a horse between the ages of two and three is rarely based on biology alone. It is almost always driven by two powerful human motivations: economic pressure — the need for a return on investment in racing or performance futurities — and personal desire — the horse owner’s goal to begin their journey with a young prospect. If we are to use science to justify these timelines, we must also be willing to follow that science to its logical conclusion regarding the horse’s management.
The Biological “Tax” of Confinement
If we accept the premise that loading builds bone, we must also accept the inverse: a lack of loading leads to its degradation. Current research into “disuse osteopenia” shows that bone mineral content begins to drop significantly when a horse is confined to a stall.
For a young horse in a traditional training program — stalled for 23 hours a day to “protect” an investment or simplify logistics — the short periods of work under a rider are often insufficient to counteract the mineral loss occurring during the other 23 hours. Studies show a measurable decrease in bone density in as little as 28 days of stall confinement (Logan & Nielsen, 2019).
Conversely, a horse raised with maximum turnout participates in constant, voluntary loading. Every turn and burst of energy at liberty sends a signal to the skeletal system to reinforce itself. Research suggests it takes remarkably little high-speed movement to build bone; a single 50-meter sprint once or twice a week is a more effective stimulus for bone modeling than hours of slow, controlled work (Logan et al., 2019).
Ground-Up vs. Top-Down: The Direction of Force
To make an educated decision on training timelines, we must differentiate between two types of force and how they interact with the developing skeleton:
Axial Loading (Ground-Up): When a horse runs at liberty, the concussion travels from the ground up through the legs. This is the ideal stimulus for the cannon bones and the maturation of joint cartilage.
Transverse Loading (Top-Down): When a rider is introduced, the force is applied downward onto the spine. This requires a young, unfused skeleton to compensate for a human’s weight, balance shifts, and mechanical cues.
This distinction is vital because the leg bones and the spine do not mature on the same timeline. While the growth plates in the lower limbs typically fuse by the age of two, the equine spine is the final part of the skeleton to reach maturity. The vertebral growth plates do not fully ossify until a horse is between 5.5 and 6 years of age (Dr. Deb Bennett).
Evaluating the Path Forward: Three Management Models
Based on these scientific realities, we can identify three distinct paths. Each carries a predictable set of consequences for the horse’s long-term soundness.
Model A — The Traditional Training Model (Confinement + Early Start). Driven primarily by economic pressure, this model starts the horse at two or three while maintaining stall confinement. The result is often a “fragile” bone structure due to stall-induced mineral loss, combined with high risk to the unfused spine. This model prioritizes a specific timeline over structural sustainability.
Model B — The Biological Foundation Model (Full Turnout + Late Start). Driven by long-term durability, the horse remains on pasture until age 4 or 5. This allows for maximum bone mineral density through natural loading and ensures the spine is fully fused before a rider is introduced. This model minimizes injury risk and maximizes the horse’s useful lifespan.
Model C — The Hybrid Optimization Model (Full Turnout + Targeted Early Start). If personal desire or industry requirements dictate a start at age two or three, this model insists on 24/7 turnout to ensure the skeletal “bank account” is maintained by natural movement. Under-saddle work is kept minimal and sub-maximal to protect the immature spine while leveraging the “modeling window” of the lower limbs.
Conclusion
Whatever you decide, it’s ultimately in your hands. We have enough information now to understand what these decisions will result in as far as consequences are concerned. So now that you have the information — what do you want to do with it?
Frequently Asked Questions
How does a horse move differently with a rider versus at liberty?
At liberty, a horse moves in self-carriage. Their balance, head position, and stride are organized entirely around their own center of gravity, and the concussion of every footfall travels from the ground up through a freely swinging spine. The moment a rider is added, the horse has to reorganize that entire system to compensate for an external load applied downward onto the back. Vertical force through the rider has been measured at roughly 3.83 N/kg at the walk and rises to 5.60 N/kg at the gallop (de Cocq et al., 2009, Equine Veterinary Journal). That force changes spinal kinematics, redistributes weight toward the forelimbs, and forces the horse to recruit different musculature just to stay balanced. It is not a small adjustment. It is a fundamental change in how the body moves.
It’s not just the weight, is it? What else is the young horse adapting to?
Correct — the weight is only one input. A ridden horse is simultaneously processing rein cues that change head and neck carriage, leg cues that redirect the barrel, weight aids that shift the center of gravity, and the rider’s own posture and asymmetry. Peer-reviewed biomechanics research has documented that even a 5 cm difference in stirrup length is enough to measurably alter the movement of the horse’s thoracolumbar spine and create asymmetries that mimic hind-limb lameness (Williams et al., 2023, Animals). Multiply that by the dozens of micro-corrections a green horse makes every minute under a rider, and you start to see why “just sitting on them” is never just sitting on them. The horse is constantly recalibrating, and a young horse is doing it with a skeleton, musculature, and proprioceptive system that are all still developing.
What does “adaptation” actually mean for a developing skeleton and musculature?
Adaptation is not a temporary thing. Bone, muscle, fascia, and connective tissue all remodel in response to repeated patterns of load. In a mature horse, that remodeling produces the conditioned athlete. In a still-developing horse, you are not training movement — you are sculpting structure. The patterns you reinforce now, good or bad, become the horse’s default carriage for life. A young horse repeatedly drilled in the same movement, on the same lead, in the same frame, develops a body that is shaped around that pattern. This is exactly why the variety and freedom of pasture movement matters so much in the early years: it builds a body that can do many things rather than a body locked into one.
Why are tools like side reins and draw reins a particular concern in young horses?
Because they don’t train movement — they impose a posture. A 2024 systematic review and meta-analysis of 58 peer-reviewed studies on neck hyperflexion (including draw-rein, side-rein, and Rollkur use) found consistent negative associations with welfare and biomechanics, including elevated stress markers, restricted movement, hollowed backs, and altered spinal kinematics (Pérez-Manrique et al., 2024, Scientific Reports). When you set a horse’s head with a gadget, you’re not asking the horse to find balance — you’re forcing them to develop musculature and connective tissue around a frame they didn’t choose and can’t release from. In a horse whose spine isn’t even fully fused yet, that’s not training. That’s permanent structural shaping.
But don’t young horses get hurt in pasture? Isn’t stalling them safer?
This is the question that drives most “protective” stall management, and the research answers it directly: no, stalling is not safer. A peer-reviewed review of three decades of equine bone research from Michigan State University found that horses moved off pasture and into stalls actively lose bone mass, and that the highest rates of bone-related injuries cluster in the period when bone density is at its weakest — which, in confined young horses, is the very window when training and racing begin (Nielsen, 2023, Animals). Pasture turnout, by contrast, builds bone. Even partial turnout (12 hours a day) was enough to prevent the bone loss seen in 24/7 stalled horses. The pasture-injury fear is real but small; the catastrophic-injury risk from confinement-weakened bone hitting fast work is much larger and much better documented.
If I do start my horse young, what should I be watching for?
Keep sessions short. Vary the work — no drilling the same patterns, the same direction, the same frame. Let the horse find their own balance rather than holding them in a shape with a gadget. Prioritize honest forward movement over a “pretty” outline. Maximize turnout between rides; the bone-building work is happening out in the field, not in the arena. And watch for the early signals — short stride, reluctance to bend a particular direction, asymmetric muscling — that tell you the body is compensating rather than developing. A young horse asking for a break is not being naughty. They’re telling you their structure is at its limit.