ABSTRACT: The aim of this article is to describe how to investigate potential back pain in horses The horse that presents with back pain' is a challenge to any veterinary surgeon, as owner expectations are high and the diagnostic capabilities low. The author has developed a protocol to fit his varied caseload of racehorses, performance horses and the pleasure horse of varying types and abilities. This protocol will not be applicable in every situation and it is time consuming and expensive, thus compromises are sometimes agreed by the owner.

The suspicion of back pain is a common complaint from owners, trainers and riders of horses. It is a common clinical presentation and one of the first ‘diagnoses’ that riders make whenever their horse displays signs that do not fit with its usual behaviour.

Back pain may be blamed for lameness, poor performance and resistance during ridden exercise. In some cases, the horse does indeed suffer from some form of spinal discomfort. However, a number of factors make the investigation of potential back pain a complex area.

A modern approach

Most horses presented for the investigation of potential back pain exhibit one or more of the complaints listed in Table 1. These signs are very real in the rider or owner’s minds but, from a clinical perspective, they are vague – all have a multitude of possible explanations.

The key to the investigation of these cases is to remain objective and assume nothing. On presentation, the approach is not only to try and identify spinal disease but, as important, to rule out everything else that might be causing the signs, before settling on a diagnosis of primary spinal pathology. Unfortunately, the investigation of back pain is as much about ruling things out as it is about ruling things in.

The investigation of back pain – and the establishment of whether back pain even exists – is difficult. Horses are large animals and we must be aware of the limitations of our physical examination.

This basic limitation extends throughout the assessment: radiographs are limited because of the patient’s size; ultrasonography is limited if the horse has significant subcutaneous fat; MRI or CT of the horse’s spine cannot be performed because no scanner exists to accommodate this.

It is vital that the examiner and the owner are aware of these limitations: it may be that we find no pathology to explain the signs, but that does not necessarily mean that no pathology exists. Absence of evidence is not the same as evidence of absence.

Aetiology of suspected back pain

As mentioned previously, a horse presenting with suspected back pain may suffer from a number of complaints that mimic the signs of spinal injury. In addition, as the majority of examinations are requested because of aberrant behaviour during ridden exercise, what also has to be taken into account is the ability of the rider and the fit and quality of the saddle and bridle. These rider- related issues are outside the boundaries of this article, but must be considered.

As far as horse-related problems are concerned, the main areas of interest are the head and mouth (if I have suspicions in this area I ask a colleague with dental knowledge to help), the neck, the back, the pelvis and the hind limbs. Very occasionally perceived back pain will be down to forelimb lameness, but this is rare.

One of the most difficult facts to establish is whether the horse is displaying behavioural signs or is actually in pain – this has particular implications for insured horses as most insurers will assume, rightly or wrongly, that if no lesion can be identified, then the problem is ‘behavioural’ and they will not pay for the investigation.

The term ‘behavioural’ requires clarification: it is taken by insurers and many others to mean that there is no painful injury causing the clinical signs and the horse is being evasive – the problem is ‘in his head’. There are problems with this simplistic approach though.

First, all back problems (or potential back problems) come to light because a horse changes its behaviour – as in Table 1, it may become resistant or buck; so in a way, all of them are behavioural!

Second, just because an extensive (and expensive) investigation finds no evidence of pathology to account for the signs, this does not mean that there is not a lesion currendy beyond detection. Despite the introduction of advanced imaging techniques, such as MRI and CT scanning, a significant number of human back pain cases go undiagnosed; so it would be naive to think that there are not, as yet, undocumented causes of back pain in horses when advanced imaging modalities such as these are unavailable to our patients.

That having been said, however, insurers have a right to protect their interests and have to take measures to guard against unnecessary veterinary fees.

Diagnostic approach

The diagnostic algorithm in Figure 1 demonstrates the approach developed by the author, over the last few years, for the investigation of potential back pain. He works in a referral clinic, where all of the horses examined for back pain pass through his hands.

Figure 1: The protocol for investigation of back pain at our referral clinic. Note that advanced imaging [scintigraphy! is deferred until after the assessment and radiography/ultrasonography


As with all assessments, a thorough history from the owners and/or rider is essential. It is important to sift gendy through the information provided by the clients and distinguish between the relevant and less relevant. Owners rarely mislead intentionally, but they may emphasise points that they consider important, while only mentioning in passing more vital clues.

Cases presented for this type of investigation frequently share historical features:

•  many have a recent change in ownership (within the previous 6-18 months) 

•  they will have displayed the signs of concern only once, the rider being reluctant to persist (for fear of injury to themselves or the horse) 

•  some cases will only show signs when ridden in an arena, not when out hacking.

The first point is interesting; in that it is believed that horses have been coaxed through their issues by a rider and then sold (not with malicious intent, but because the rider dealt with the issues and sold the horse in good faith). The purchaser then finds themselves with a ‘problem horse’ after the effect of the previous owner has worn off.

In some cases, horses are produced to a certain level for sale to less experienced riders: p
roblems then arise because of the rider’s relative lack of expertise.

The second point is understandable, particularly if the horse is considered a danger to itself or the rider. However it does make it difficult to establish whether the horse might repeat the behaviour – when does a ‘bad day’ become a real problem?

Why some horses only show signs in an arena and not out on a ride is also fascinating. Work in the arena equals ‘serious’ work, whereas hacking is easy for both rider and horse. The increased demands of schooling and lessons, which occur invariably in an arena, put added pressure on the horse (and rider) that is not generally experienced by either when out hacking.

Physical examination

All horses coming through our clinic are assessed fully – usually in the stable – before we examine them outside. This is a crucial part of the evaluation but certain caveats apply: it is my opinion that it is impossible, based simply on a physical assessment, to be certain that the horse does, or does not, have back pain. This can be troublesome as some vets, and other professionals, consider the physical examination as the main diagnostic tool.

While undoubtedly important, it has been my experience that the responses to palpation and manipulation vary so widely between individuals, that only a very small number of cases can be said, with confidence, to be suffering from back pain or not, based solely on the physical examination.

Initially, I palpate the midline of the back, taking time to ensure the reactions produced from the horse are the genuine results of my assessment, not simply because the horse is unhappy with the interaction. Gentle pressure is applied to the midline from the withers to the tubera sacrale and repeated to review the horse’s response.

The pressure applied is gradually increased; areas that appear sensitive are noted and re-inspected. Then the epaxial muscles are assessed for signs of increased tension: differences in response between left and right sides are particularly useful.

As well as watching the response of the muscles under the fingers, the attitude of the horse at the ‘front end’ can also be useful – and an experienced handler can often provide useful feedback. Qualified physiotherapists use several techniques to assess muscle tension/tenderness and it is well worth spending time with them to discover how they train their fingers to assess soft tissues.

All Chartered Physiotherapists in Animal Therapy have qualified first in the human field and have then spent time treating, humans, before they can even begin training on animals. The feedback humans give them during their training and work assists greatly in assessing less communicative equine patients!

Various mobility tests are then performed – applying pressure to certain regions in order to try and elicit normal or abnormal responses. I try and ascertain the degree of dorso- and ventro-flexion of the spine, lateral mobility and response to weight applied to the saddle region (without actually getting on the horse!).

The examiner must know how and where movement occurs in the normal horse, as different regions of the spine are responsible for different ranges of movement. For example, there is almost no lateral flexion possible in the normal lumbar spine; and the vast majority of ventro- and dorso-flexion occurs at the lumbosacral junction.

As mentioned previously, the interpretation of these tests is very subjective. There is an enormous amount of variation between normal and abnormal individuals, so it can be very difficult to interpret the results of this assessment accurately. However, at the end of this stage we have a little more information to be added to the overall picture.

We*also assess neck mobility by asking the horse to bend the neck to the left and right towards a scoop of feed: caudal cervical osteoarthritis can be a cause of ‘back pain’ and often causes a restriction of movement.

The horse is then removed from the stable and walked to the ‘trot-up’ area. It is important, at this stage, to see the horse walking out of its stable and to assess its gait. Once at the trot-up, the horse is squarely stood up and its surface anatomy is inspected, looking particularly for signs of muscular or bony asymmetry over the neck, back and limbs.

The horse is assessed walking away and back, from in front and from the side. The examination at the walk is often overlooked, but vital, because certain conditions (particularly neurological disease) will be more obvious at this slower gait.

The horse is then trotted in straight lines, before being lunged on hard and soft surfaces. The main objective here is to ascertain whether it is lame or not. Many ‘back pain’ cases have significant lameness (usually affecting the hind limbs) which must be assessed. Lunging on the hard and soft is important because different orthopaedic problems manifest themselves depending on the surface. Remember, a horse cannot be pronounced ‘sound’ just by trotting it up in a straight line.

Ridden assessment

As long as no lameness or clinical feature precludes it, we then move on to the ridden assessment. Ideally, the horse is examined in our arena being ridden by its usual rider. It is then useful, in some cases, to see the horse ridden by one of our grooms – the behaviour of certain horses can change dramatically when a different rider is used.

A change produced by using a different rider does not indicate that the problem is rider-related per se – a stronger or more competent rider may enable the horse to overcome a painful problem – but the test is useful. In the case of a horse that the owner/rider considers too dangerous to ride themselves, I will not ask our grooms to ride it. In these cases, the ridden assessment is not performed.

In some cases, we will also ride the horse in a different saddle – we keep a wide- fitting, general purpose saddle at the clinic that can be made to ‘fit’ a variety of horses. Of course, in reality it doesn’t fit these horses but the fact that it fits poorly in a different way to the saddle used normally can produce surprising changes in the way the horse moves, perhaps prompting further assessment.

Diagnostic imaging: Stage 1

The horse is then subjected to an imaging assessment – initially radiography and ultrasonography. Radiographs are taken of the cervical, thoracic and lumbar vertebrae (making sure to include the ventral aspects of the caudal thoracic vertebral bodies, to look for spondylosis) as well as the dorsal spinous processes (DSPs) (Figures 2 & 3).

Figure 2: Lateral radiograph of a horse with advanced OA of the caudal cervical articulations. Neck pathology is an important differential diagnosis when cases present with possible back pain

Figure 3: Lateral radiograph of the caudal thoracic region. It is important to include the dorsal spinous processes (DSP) and the ventral aspect of the vertebral bodies [superimposed over the lungs) to document all pathology

I routinely take views of the stifles (caudocranial and flexed lateromedial projections) and hocks (lateromedial and dorsomedial-plantarolateral projections) even in horses that show no signs of
lameness at the time of my assessment. I have seen a number of horses with no lameness at the time of presentation who have had significant pathology of these joints that has later been proven to be the cause of ‘back pain’ – remember that many of these cases will have been rested for a while before getting to the clinic, so earlier lameness may have resolved. Following this, we examine the back (supraspinous ligament, intervertebral ‘facet’ joints) (Figures 4 & 5), stifles (especially the medial femorotibial joints and patellar ligaments) and proximal suspensory ligaments of the hind limbs ultrasonographically.

Figure 4: Ultrasonograph of a portion of a normal supraspinous ligament

Figure 5: Ultrasonograph of a normal cranial lumbar intervertebral articulation, superimposed on an anatomical specimen

A per rectum assessment of the lumbosacral region, intertransverse and sacroiliac joints follows (Figure 6).

Figure 6: Ultrasonograph of the L5/6 and L6/S1 (lumbosacral joint) of a normal horse This image can be obtained by examination per rectum

At this stage the horse has been admitted for a few hours or so and has completed the first stage of the examination. What then follows is a discussion with the owner/ rider regarding the findings. Generally, there are three scenarios at this point: a lesion is identified that can easily explain the perception of back pain several possible ‘leads’ have surfaced that may explain the problem, but that require further investigation no physical problems have been identified.

A decision needs to be made at this point as to how to proceed.

In some cases, if a clear lameness is seen, it is prudent to investigate this first. In others, a lesion can be treated and assessed to see how the horse responds. 

In cases where several possible explanations have been found it is generally preferable to treat one area at a time and assess the effect before treating other areas. Examples of this include medication of ‘kissing spines’ (Figure 7) or ultrasound- guided injection of thoracolumbar facet joints or cervical joints.

Figure 7: Lateral radiograph of a horse presented for sudden onset signs of back pain (bucking). There are signs of impingement and remodelling of the DSPs which might lead to a diagnosis of kissing spine

If several possible lesions have been identified, it might be advisable to proceed with a bone scan (nuclear scintigraphy) – the level of radionuclide uptake in the areas under suspicion might add weight to their likely significance or rule them out of the investigation (Figures 8a and 8b). Alternatively, it is possible to use diagnostic anaesthesia to block out certain areas of suspicion (kissing spines, proximal suspensory regions or sacroiliac joints, for instance). This is very useful when the horse has a repeatable response to exercise that is seen every time it is ridden.

Figures 8a & 8b: Lateral [a] and dorsal (b) scintigraphs of the same horse as in Figure 8. There is normal distribution of radionuclide throughout the caudal thoracic DSPs, suggesting that the impingement evident radiographically is not clinically relevant. This horse has a fractured 16th rib on the left side and would have been a poor candidate for DSP treatment

If no abnormalities have been detected this far, it might still be sensible to proceed with a bone scan, but the client must be aware that, if the horse is insured, the insurance company might interpret a subsequently normal bone scan as evidence that the horse has no physical problems and the signs must be caused by ‘behavioural issues’. Most will not pay for any of the diagnostic work-up performed. If this is the case, it will leave the owner facing a large bill – which for the diagnostic work done on day one, plus a thorough bone scan assessment, could be above £2,000. Some clients are happy to proceed in the light of this knowledge, others are not; but it must be an informed decision on their part.

Diagnostic imaging – Stage 2

Some clients, when advised that the initial assessment of their horse has uncovered no signs of disease, are happy to take the horse home and reconsider the possible roots of the problem. In these cases, it can be very useful to perform a trial period at home with the horse receiving analgesics (so-called ‘bute’ trial).

The horse is given phenylbutazone, or a similar NSAID, and is ridden as normal. If the clinical signs resolve, it is reasonable to assume that the problem has a painful cause and further investigation may be justified. It can be performed over a two-week period, with a typical horse receiving lg of phenylbutazone orally (or equivalent alternative NSAID and assuming a typical 450-500kg body weight) twice a day for 10 days. If there has been no response, then the dose should be increased to 2g twice a day for three days only.

If there has been a perceived positive effect, the ‘bute’ should be stopped, and the horse should continue to be ridden.

If signs recur, the ‘bute’ trial should be repeated. Clearly this is a bit more involved than the typical ‘put it on bute for a few days and see’ approach, but there can be a strong placebo effect from the rider who knows that the horse is on analgesia and desperately wants to find a solution.

A slightly prolonged an
d repeated trial is more accurate. In the ideal trial, the rider is unaware when the ‘bute’ administration starts and stops, but in practical terms that is often impossible to perform! 


Marcus J Head BVetMed MRCVS

Marcus qualified from the Royal Veterinary College in 1994 and spent a period in mixed practice before an internship at the Animal Health Trust Equine Clinical Unit.

A second internship at The RVC was followed by a move to Rossdale & Partners in 1996, as an assistant dealing with horses in training. In 2004. he moved to the practice's diagnostic centre to work on referral cases full-time and he now sees orthopaedic cases from a wide variety of disciplines – occasionally escaping to attend Thoroughbred sales!

To cite this article use either

DOI: 10.1111/j.2045-0648.2012.00202.x or Veterinary Nursing Journal Vol 27 pp 288-292


DENOIX, J. M. (1999) Spinal biomechanics and functional anatomy in Vet Clin North Am Equine Pract (1999) April 15(1): 27-60.

• VOL 27 • August 2012 • Veterinary Nursing Journal