The Effect of Noseband Tightening on Horses’ Behavior, Eye Temperature, and Cardiac Responses

Picture a horse in a dressage arena. Its neck is arched, nose nearly vertical, and its jaw is clamped shut by a strap that runs across the bridge of the nose and buckles beneath the chin with a leveraged clasp that can be cranked down like a tourniquet. From the stands, it looks like control, like harmony. The judges reward it. But what is actually happening inside that horse's body when the noseband goes that tight? That is the question Kate Fenner and colleagues set out to answer. The equipment at the center of this story is the crank noseband. It was developed in the 1980s and is now standard in elite dressage. The original cavesson noseband was simple and loose — functional, not restrictive. The crank version added leverage, allowing riders to tighten far beyond what fingers alone could achieve. The reason this spread through the sport comes down to how dressage is judged. The Fédération Equestre Internationale, or FEI, requires horses to demonstrate submission, defined as attentiveness, willingness, and confidence, and penalizes signs of resistance like putting out the tongue, grinding the teeth, or evading the bit. A very tight noseband suppresses those visible signs. It can also press the bits against the tongue, sensitizing the mouth and making the horse appear lighter and more responsive to rein cues. The horse looks submissive, and the judges score it higher. The welfare question is what that tightness costs the animal. Fenner and colleagues were also thinking about pressure in a more literal sense. Other research had measured noseband pressures at the nasal planum in the range of 200 to 400 millimeters of mercury — comparable to pressures that compromise blood flow in other contexts. At those levels, the concern isn't just behavioral restriction; it's vascular perfusion, nerve compression, and in extreme cases, bone damage. The team recruited twelve horses naïve to a crank noseband and double bridle, fitted each animal with all four noseband conditions in a randomized order across four consecutive days, and measured what happened at each level of tightness. The four conditions were: unfastened noseband, referred to as UN; conventional tightness, called CAUN, with two fingers of space under the noseband; half conventional, referred to as HCAUN, with one finger of space; and no area under the noseband at all, known as NAUN — tight enough that the standard taper gauge couldn't be inserted at the nasal planum. To capture the horse's internal state, they used three physiological channels. Heart rate is the most intuitive; faster beats signal arousal. Heart rate variability, or HRV, is subtler — it's not the speed of the heartbeat, but the variation between successive beats. When the sympathetic nervous system, the fight-or-flight branch, takes over during stress, that variation collapses. Lower HRV means the autonomic nervous system has shifted into stress mode. In this dataset, heart rate and HRV were strongly inversely correlated, with a correlation coefficient of negative 0.74. The third measure was eye temperature, captured with an infrared thermal camera — a ThermaCam T604 with sensitivity finer than 0.02 degrees Celsius. Stress increases blood flow to the periorbital region and raises the surface temperature of the eye. Each session ran in three phases: ten minutes of baseline, ten minutes with the noseband treatment applied, and ten minutes of recovery after removal. The physiological story is strikingly consistent. During the tightest treatment, the NAUN condition, all three channels moved in the same direction. Heart rate climbed from a baseline of about 34 beats per minute to about 46 during treatment — nearly twelve beats more, with a p-value of 0.003. HRV dropped from 1,817 milliseconds at baseline to 1,576 during treatment, a decrease of 241 milliseconds, with a p-value below 0.001. Eye temperature rose from 34.81 to 35.25 degrees Celsius, a shift of nearly half a degree, with a p-value of 0.011. These horses were standing still during measurement, so this cannot be chalked up to physical exertion. Three independent physiological channels converged on the same conclusion: the tightest noseband triggered a stress response. What matters is the gradient. At conventional tightness — the two-finger standard the FEI itself endorses — those cardiac and thermal shifts were not present. The prominent stress signal appeared only when no space remained. Then there's what happened to the mouth. As tightness increased, the horses went behaviorally silent in a very specific way. Chewing dropped significantly at HCAUN and NAUN, both with a p-value below 0.001. Swallowing showed a stepwise decline across conditions — at baseline, the horses swallowed about 1.42 times per ten minutes; under NAUN, that fell to 0.71, roughly half. Licking was eliminated entirely by the tightest treatment. Yawning was negligible across all conditions while the noseband was on. On the surface, that might look like a mechanical consequence — you've immobilized the jaw, so of course the jaw doesn't move. But the recovery session tells a different story. After Fenner and colleagues removed the bridle and noseband, three behaviors surged above baseline levels. Yawning increased significantly under the NAUN recovery condition compared to baseline, with a p-value of 0.015. Swallowing increased, with a p-value of 0.003. Licking increased sharply, with a p-value below 0.001. This is the post-inhibitory rebound: the phenomenon where a suppressed behavior returns at a rate higher than normal once the constraint is lifted. It's not a reflex; it's a motivational signature. The paper frames it explicitly as evidence of a build-up in motivation during the restriction period — the horses were not just mechanically prevented from yawning and licking; they were in a state of behavioral deprivation, and the burst of activity after removal was the release. That rebound lands differently when you hold it alongside the physiological data. During the NAUN condition, the horses showed elevated heart rate, collapsed HRV, and a warmer eye. When NAUN was removed, they immediately began yawning, licking, and swallowing at rates above what they had shown before the equipment was ever applied. The stress physiology and the behavioral surge point to the same conclusion from two different directions. This is not a horse that was unaffected and simply chose not to move its mouth. This is a horse that was stressed, suppressed, and then, upon release, flooded back toward normal. The interpretive question Fenner and colleagues are candid about not resolving is this: is the stress driven by pain from noseband pressure, by the frustration of behavioral restriction, or by both? The cardiac pattern under NAUN is consistent with acute pain or discomfort. But post-inhibitory rebound was detectable even at looser settings, including the conventional two-finger CAUN, which suggests that behavioral inhibition — the mere prevention of natural oral movement — is doing some of the work independently of pain. The study cannot separate these mechanisms cleanly, and the authors say so directly. There's a further limit worth considering: these were naïve horses. None of them had prior experience with a crank noseband or a double bridle. Whether trained competition horses exposed to this equipment repeatedly develop some habituation, or whether they remain chronically stressed, is unknown. The study design captures an acute response, and the acute response is unambiguous — but the chronic question is what matters most for horses that compete in dressage year after year. The regulatory implication is precise. The FEI's own guidance is built around a two-finger rule for noseband tightness. This study found that the conventional two-finger condition, CAUN, did not produce the convergent physiological stress signal that NAUN did. In that narrow sense, the rule is pointing in the right direction. But post-inhibitory rebound appeared even at that conventional level, which suggests that the FEI threshold may not fully protect the expression of normal behavior. Absence of visible distress is not absence of distress. The horse standing quietly in the arena, jaw clamped, not grinding its teeth for the judge to see, might be suppressed rather than content. Those are not the same state, and this study gives us the tools to tell the difference — three physiological channels and a burst of yawning and licking the moment the equipment comes off. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.