The Beighton score in clinic: scoring it, what it misses, and managing hypermobility

The Beighton score is a 9 point measure of range at five sites, and that is the whole of what it is. It tells you whether a patient's joints move further than most, it does not tell you whether they hurt, whether they are unstable, or whether anything needs treating. Used as the first step of a longer assessment it is quick and useful. Used as a diagnosis it is the most common mistake made with a hypermobile patient.

This page is for clinicians. The patient program, the staged exercises and the full warning signs sit on the hypermobility program.

What the Beighton score measures, and what it does not

Five maneuvers, nine points. Passive dorsiflexion of each fifth finger past 90 degrees, two points. Passive apposition of each thumb to the flexor aspect of the forearm with the wrist flexed, two points. Elbow and knee hyperextension beyond 10 degrees, two points each. And forward flexion of the trunk with the knees extended so the palms rest flat on the floor, one point (The Ehlers-Danlos Society).

Look at what is absent from that list, because it is where most symptomatic patients present. There is nothing in the shoulder, nothing in the hip, nothing in the ankle or foot, nothing in the cervical or lumbar spine, nothing at the wrist and nothing at the temporomandibular joint. A patient whose shoulders sublux three times a week can score 2 of 9. That score is correct and it is not an answer.

The score is a measure of range, not of instability, not of pain and not of function, so a change in it is not an outcome measure for your treatment. Juul-Kristensen and colleagues (2017) reviewed the measurement properties of the clinical methods for classifying generalized joint hypermobility and recommended the Beighton score, while rating every study they included fair or poor on the COSMIN checklist and reporting limited positive to conflicting evidence on its reliability. It is the best available tool for the job, which is not the same as a well validated one.

How do you score it, and what is the cut-off?

Score each side separately and use a goniometer for the elbow and knee items, not an impression, because 10 degrees past neutral is close enough to neutral that eyeballing it is where raters disagree.

The cut-off moves with age. The Ehlers-Danlos Society gives 5 or more out of 9 as positive in adults, 6 or more in children before puberty, and 4 or more in adults over 50. Juul-Kristensen and colleagues (2017) recommend at least 5 of 9 in adults, read together with historical information, which is part of their recommendation rather than an addition to it, and at least 6 of 9 in children. The over-50 threshold is the one clinicians most often miss, and it matters: range falls with age, so a 62 year old scoring 4 meets the criterion that a 30 year old scoring 4 does not.

A positive score is criterion 1 of a three-part diagnostic process, not the diagnosis. The 2017 international classification requires all three to be met for hypermobile Ehlers-Danlos syndrome: generalized joint hypermobility, a specified combination of musculoskeletal and systemic features, and the exclusion of alternative diagnoses (Malfait and colleagues, 2017; Forghani and colleagues, 2025). Physiotherapy does not make that diagnosis, and a letter saying "Beighton 7 of 9, query hEDS" is a referral, not a conclusion.

Is the Beighton score enough to make the call?

No, and two validated tools fill the two gaps it leaves.

The first gap is history. Range is lost with age and with deconditioning, so a patient who was hypermobile at 20 can score below the cut-off at 45 while still carrying the consequences. The five-point questionnaire was built for exactly that (Hakim and Grahame, 2003).

Two or more positive answers out of five gave the best combined sensitivity and specificity in its development cohort of 212 hypermobile women attending a clinic and 57 healthy volunteers, at 84% and 89%, reproduced at 84% and 80% in a second cohort of 170 cases with benign joint hypermobility syndrome and 50 controls. Most of that first cohort, 182 of the 212, met the 1998 Brighton criteria. Juul-Kristensen and colleagues (2017) rate the reliability evidence for the questionnaire as conflicting, so it earns its place as a prompt to look, not as a result.

The five questions ask whether the patient can or ever could put their palms flat on the floor with straight knees, bend a thumb back to touch the forearm, show off body contortions or do the splits as a child, dislocate a kneecap or shoulder more than once as a child or teenager, and whether they consider themselves double jointed (The Ehlers-Danlos Society). It takes under a minute and it is the item to add to an intake form.

Those sensitivities and specificities came from a specialist hypermobility clinic, where almost everyone attending was hypermobile, so in an ordinary musculoskeletal caseload a positive questionnaire means less than the figures suggest. It tells you to examine, not what you will find.

The second gap is the upper limb. The Upper Limb Hypermobility Assessment Tool is a 12 item measure developed and validated in 112 adults of mean age 24, with a cut-point of 7 or more out of 12, sensitivity 0.84, specificity 0.77, a positive likelihood ratio of 3.7 and a negative likelihood ratio of 0.2, and inter-rater reliability of 0.92 (Nicholson and Chan, 2018). That is a young cohort, with a standard deviation of 5.5 years around that mean, so the cut-point has not been tested in the older patients for whom the Beighton threshold itself moves. The authors did test for an age effect and found none, but inside a sample that narrow that result cannot be stretched to a 60 year old.

Where the shoulder, elbow, wrist or hand is the presenting problem, it is still the tool that measures the joints involved. A hypermobile shoulder is also the joint most likely to be the reason a patient is in front of you, and shoulder dislocation covers the program when it has fully dislocated.

So the practical assessment is three items, not one: Beighton with its age-adjusted cut-off, the five-point questionnaire for history, and a regional measure where the region is the problem.

HSD and hEDS: what the labels change in your plan

Clinically, almost nothing, and saying so plainly helps a patient who has spent two years chasing a label.

Hypermobility spectrum disorder has four recognized subtypes: generalized HSD, where hypermobility occurs throughout the body; peripheral HSD, limited to the hands and feet; localized HSD, in a single joint or group of joints in the same area; and historical HSD, where there is a history of generalized joint hypermobility without current evidence of it on examination (The Ehlers-Danlos Society). That last one is where the five-point questionnaire earns its place.

The Ehlers-Danlos Society states that the principles and types of management are the same for both HSD and hEDS. So the physiotherapy plan does not wait for the diagnosis, and that is the sentence to put in the letter.

Where the label does change something is referral and screening, not exercise. The rarer Ehlers-Danlos types are a different matter. The NHS describes vascular EDS as a rare type, often considered the most serious, that affects the blood vessels and internal organs and can cause them to split open and lead to life-threatening bleeding, and it lists skin that bruises very easily, thin skin with small blood vessels visible on the upper chest and legs, and fragile blood vessels that can bulge or tear among the signs. A patient with those features needs medical review before a program, and their specialist team's instructions sit above any program you write.

What to prescribe: load, range and dose

Strength and control through the middle of range, built slowly, with range itself left alone. Hypermobile joints do not need more range and resting at the end of range is the habit the program exists to change. Screen for orthostatic symptoms before you write any of it, because they change how a session is set up.

The trial evidence does not quite say what that principle implies. In the HEAVY shoulder trial the heavy arm was progressive strengthening and full range of motion exercises, and the light arm was low-load exercises in neutral to mid-range (Liaghat and colleagues, 2025). So mid-range is where unsupervised home work starts and what a patient can judge for themselves, not a ceiling the evidence sets, and nothing in that trial showed full-range loading to be harmful.

None of that is a license to load a hypermobile joint at the end of its range. Look at the supervision the trial paid for. Patients in the full-range, high-load arm were individually supervised twice a week and self-trained once a week at home, and the authors say intensive supervision was deemed important to manage potential adverse events and adequate load progression, while the neutral to mid-range low-load arm was the self-training one, introduced individually and then supervised again at weeks 5 and 11 (Liaghat and colleagues, 2022). Full range in that trial is what a supervised program delivers, not what you hand over on a sheet.

Full range and end range are also not the same prescription. The heavy program was five open kinetic chain exercises loaded through range, one of them side-lying external rotation in neutral, and it tested nothing about load held at the end of range or in a position where the joint feels about to give, which is where this population loses control (Liaghat and colleagues, 2022). So prescribe the range deliberately: mid-range for what the patient does alone, full range inside the sessions you watch, and write down which of the two you gave. For a rarer Ehlers-Danlos type none of this is the starting point, because the NHS says the advice depends on which type the patient has and that some activities, such as heavy lifting and contact sports, may need to be avoided entirely.

The useful question for a prescriber is how much strength change is worth chasing. In a secondary analysis of 76 participants from both arms of the high-load shoulder trial, with a median age of 35.5, greater relative strength in external rotation and in scaption, meaning elevation in the plane of the scapula, was associated with better shoulder function on the Western Ontario Shoulder Instability index (WOSI), by an adjusted 3.5 points in the direction of improvement, 95% confidence interval 1.1 to 5.9, and 2.6 points, 95% confidence interval 0.6 to 4.6. Each 10% increase in strength carried higher odds of an important improvement, with an adjusted odds ratio of 1.18 for external rotation, 95% confidence interval 1.00 to 1.38, and 1.30 for scaption, 95% confidence interval 1.07 to 1.59 (Christensen and colleagues, 2026). The external rotation interval reaches 1.00, so that half of it is borderline, and the authors note the clinical relevance may be limited, which is fair. The practice point survives both: measure the strength you are trying to change, and treat a 10% gain as a unit of progress worth recording.

Make one change at a time, and let it settle for a week or two before the next, because this is the population in whom a jump in load reliably costs several days. Do not promise that a specific program is superior to a simpler option: in a randomized trial of 169 people with hypermobility, 116 of whom completed 12 weeks, a wrist stabilising exercise program was not significantly better than a hand orthosis for activity ability on the DASH, nor for quality of life, hand function, grip strength, pain or paresthesia (Lindholm and Claesson, 2025). A splint is a legitimate option for a hypermobile wrist, not a lesser one.

For the structure around all of this, the FITT principle covers how to write the first prescription and exercise progression covers what to change next. Dead bug is the clearest example of the middle-range principle in practice: the cue is to stop the limb at the point where the low back starts to lift, not to reach the floor.

Does heavy loading flare a hypermobile patient?

Less than most of us expect, on the best data available, and that is the finding to take into the conversation where a patient is afraid of loading.

A secondary analysis of the 16 week high-load shoulder trial looked at pain trajectories across the program in 64 of the 100 randomized patients, 34 in the heavy group and 30 in the light group, 80% women, mean age 39.6, with a mean Beighton score of 5.8. The programs differed in range as well as load: heavy was progressive strengthening through full range, light was low load in neutral to mid-range, and both ran three times a week. Shoulder pain fell by 0.89 on the numeric rating scale in the heavy group, 95% confidence interval 0.4 to 1.4, and by 0.33 in the light group, 95% confidence interval minus 0.25 to 0.91, a between-group difference of 0.56 that did not reach significance, 95% confidence interval minus 0.20 to 1.33, p = 0.149. Pain trajectories did not differ between the groups, p = 0.321. And exercise-induced pain was similar between groups, p = 0.614, and stayed below 0.5 on the numeric rating scale throughout, which the authors summarize as minor to no pain during high-load strengthening (Liaghat and colleagues, 2025).

Heavy loading did not hurt more, so fear of flaring is not a reason to keep a patient in light loads indefinitely. What the main trial did record was more transient muscle soreness in the heavy group, 56% against 37%, and more headaches, 40% against 20%, with no serious adverse events (Liaghat and colleagues, 2022), so warn a patient that soreness is likely rather than let it arrive as a surprise. It also did not reduce pain significantly more, so heavy loading is an option to grade toward rather than a target to rush at. And it is one joint in one population, shoulder exercise in adults aged 18 to 65 with hypermobility spectrum disorder and shoulder symptoms of more than 3 months, so it is not yet a finding about loading every hypermobile joint.

For the patient's own rule, the pain monitoring model gives a number to work to and a next-morning check, which is more use to this population than "stop if it hurts".

Pacing, flare-ups and fatigue

Fatigue is often the complaint that limits the program, not pain, and a session plan that ignores it gets abandoned in week two.

The pattern to design around is the good-day overshoot: a patient does the whole program on the day they feel well and loses the next three. The counter is to set the session at the level a bad day can take and keep good days close to it, then add in small increments once the last one has settled. Two short sessions in a day often succeed where one longer session does not.

Write the flare version before the flare. Decide with the patient what a reduced session looks like, name the specific exercises in it, and put it in the program so it is a plan rather than a decision made while they feel awful. A smaller session keeps the strength that the rebuild starts from, which is the practical argument for not stopping.

Outcome measures in physical therapy covers choosing something to track over months, which matters here because the thing that improves is the number of good days, not the range.

Orthostatic symptoms, the gut and the other systems

Screen for orthostatic symptoms before the first session, because they change how you set a program up and because they are easy to miss in a musculoskeletal assessment. The NHS describes postural tachycardia syndrome as a condition in which the heart rate rises after getting up from sitting or lying, and says people with joint hypermobility syndrome may be more likely to have it. It also lists exercise itself, getting too hot and standing or sitting upright for long periods among the things that can make the symptoms worse, which is the practical reason to keep the first sessions short and cool and to avoid long upright holds until you know how the patient responds.

The gut side has a current clinician document worth knowing about. The American Gastroenterological Association published a clinical practice update on gastrointestinal manifestations and autonomic or immune dysfunction in hypermobile Ehlers-Danlos syndrome, with 16 best practice statements. Among them: screen for joint hypermobility in patients with functional gastrointestinal disorders, test for postural tachycardia syndrome and mast cell activation syndrome in a targeted rather than a universal way, consider celiac testing earlier, and manage according to the dominant gastrointestinal symptom. The authors state the update rests on literature review and expert opinion rather than systematic reviews, so it carries no formal evidence ratings (Aziz and colleagues, 2025). None of that is physiotherapy, and all of it is worth having in the letter you write to the general practitioner.

The pelvic floor is the system most often left out of a hypermobility plan. Where there is leaking, heaviness or dragging, pelvic floor weakness covers the program, and in pregnancy and the postnatal period pregnancy back and pelvic pain covers a population where the tissue changes anyway. A knee that gives way is a separate assessment again, and knee giving way covers what to rule out.

Children, adolescents and adults over 50

Use the right cut-off. Six or more of 9 before puberty, 5 or more in adults, 4 or more over 50 (The Ehlers-Danlos Society). Getting this wrong in either direction has consequences: a prepubertal child scoring 5 does not meet the criterion, and an older adult scoring 4 does.

In children the presentation is more often fatigue, clumsiness, growing pains and avoidance of physical education than a named unstable joint, and the plan is usually graded participation and a conversation with the school, not a clinic program. A child with painful joints or joints that keep slipping out of place needs their doctor, who can refer to a physiotherapist or a specialist team.

In the over-50s the question shifts from hypermobility to its consequences, and the assessment should include falls risk and a review of what the patient is already avoiding. Pushing range is not the aim at any age in this group.

When do you refer on?

Start with the one a musculoskeletal clinician is most likely to treat instead of refer: an incomplete diagnostic picture. Generalized joint hypermobility plus systemic features, with the alternative diagnoses not excluded, is a medical referral. The exclusion step is not a formality: in a retrospective review of 907 patients referred for hypermobile Ehlers-Danlos syndrome evaluation at one United States clinic between 2019 and 2022, 178 met the 2017 criteria, and genetic testing identified an alternative or additional diagnosis in 47 of them, 26.4% (Forghani and colleagues, 2025). Roughly one in four of those who met the criteria had something else going on as well.

The rest are quicker to spot. Signs of a rarer Ehlers-Danlos type, as above. Orthostatic symptoms, which belong to the patient's doctor. Significant gastrointestinal symptoms, where the AGA update is the document to point at. And joints that are dislocating or subluxing repeatedly, which needs a medical or orthopedic opinion alongside your program, not after it.

The handover carries the site's own patient-facing wording, quoted here as we give it to patients:

"Emergency: a joint has come out of place and has not gone back, looks a different shape or sits at an odd angle, or is stuck so you cannot move it. Go to an emergency department straight away, and do not try to put it back in yourself. Do not drive yourself: ask someone to drive you, or call emergency services if you cannot get there."

"Within a few days: joints that partly slip out of place (subluxations) more often than before, or a joint that has started to feel unstable during everyday tasks. Book an appointment with your doctor or physio within a few days."

"See your doctor: you often feel dizzy, lightheaded or faint, or your heart races or pounds, after you get up from sitting or lying down. This can be postural tachycardia syndrome (PoTS), which the NHS says may be more likely with joint hypermobility syndrome. Book an appointment with your doctor, and tell your physio before you start the program. If you have fainted, see your doctor to find out why. If you faint while exercising or lying down, or you faint and also have chest pain or a pounding, fluttering or irregular heartbeat, do not fully recover, have trouble speaking or moving, were badly hurt, or were shaking or jerking, call emergency services."

The full set, with the levels for each, is on the hypermobility program, and that is the page to send, not this one.

The short version

Score the Beighton with a goniometer and the right age cut-off, then stop treating it as an answer. Add the five-point questionnaire for the history it cannot see and a regional tool where the region is the problem. The label does not change the program, so start the program. Load through the middle of range, measure the strength you are trying to change, add one thing at a time, and plan the flare version in advance. And refer the exclusion step out, because about one in four of the patients who meet the criteria turn out to have something else as well.

References

1. The Ehlers-Danlos Society. Assessing joint hypermobility. https://www.ehlers-danlos.com/assessing-joint-hypermobility/
2. The Ehlers-Danlos Society. What is HSD? https://www.ehlers-danlos.com/what-is-hsd/
3. Juul-Kristensen B, Schmedling K, Rombaut L, Lund H, Engelbert RH. Measurement properties of clinical assessment methods for classifying generalized joint hypermobility: a systematic review. American Journal of Medical Genetics Part C: Seminars in Medical Genetics. 2017;175(1):116-147. https://doi.org/10.1002/ajmg.c.31540
4. Hakim AJ, Grahame R. A simple questionnaire to detect hypermobility: an adjunct to the assessment of patients with diffuse musculoskeletal pain. International Journal of Clinical Practice. 2003;57(3):163-166. https://pubmed.ncbi.nlm.nih.gov/12723715/
5. Nicholson LL, Chan C. The Upper Limb Hypermobility Assessment Tool: A novel validated measure of adult joint mobility. Musculoskeletal Science and Practice. 2018;35:38-45. https://doi.org/10.1016/j.msksp.2018.02.006
6. Malfait F, Francomano C, Byers P, et al. The 2017 international classification of the Ehlers-Danlos syndromes. American Journal of Medical Genetics Part C: Seminars in Medical Genetics. 2017;175(1):8-26. https://doi.org/10.1002/ajmg.c.31552
7. Forghani I, See J, McGonigle WC. Hypermobile Ehlers-Danlos Syndrome: Diagnostic Challenges and the Role of Genetic Testing. Genes. 2025;16(5):530. https://doi.org/10.3390/genes16050530
8. Liaghat B, Skou ST, Søndergaard J, Boyle E, Søgaard K, Juul-Kristensen B. Short-term effectiveness of high-load compared with low-load strengthening exercise on self-reported function in patients with hypermobile shoulders: a randomised controlled trial. British Journal of Sports Medicine. 2022;56(22):1269-1276. https://doi.org/10.1136/bjsports-2021-105223
9. Liaghat B, Juul-Kristensen B, Christensen FH, Nissen SE, Skou ST, Søgaard K, Søndergaard J, Thorlund JB. Pain trajectories and exercise-induced pain during 16 weeks of high-load or low-load shoulder exercise in patients with hypermobile shoulders: A secondary analysis of a randomized controlled trial. Scandinavian Journal of Pain. 2025;25(1). https://doi.org/10.1515/sjpain-2024-0072
10. Christensen T, Juhl CB, Juul-Kristensen B, Skou ST, Søndergaard J, Søgaard K, Liaghat B. Association Between Changes in Shoulder Strength and Self-Reported Shoulder Symptoms in Patients With Hypermobility Following 16-Weeks of High-Load or Low-Load Exercise: A Secondary Analysis of an RCT. Physiotherapy Research International. 2026;31(2):e70197. https://doi.org/10.1002/pri.70197
11. Lindholm S, Claesson L. Wrist Stabilising Exercise Versus Hand Orthotic Intervention for Persons with Hypermobility: A Randomised Clinical Trial. Clinical Rehabilitation. 2025;39(1):47-57. https://doi.org/10.1177/02692155241293265
12. Aziz Q, Harris LA, Goodman BP, Simrén M, Shin A. AGA Clinical Practice Update on GI Manifestations and Autonomic or Immune Dysfunction in Hypermobile Ehlers-Danlos Syndrome: Expert Review. Clinical Gastroenterology and Hepatology. 2025;23(8):1291-1302. https://doi.org/10.1016/j.cgh.2025.02.015
13. NHS. Ehlers-Danlos syndromes. Page last reviewed 4 October 2022. https://www.nhs.uk/conditions/ehlers-danlos-syndromes/
14. NHS. Postural tachycardia syndrome (PoTS). Page last reviewed 22 July 2026. https://www.nhs.uk/conditions/postural-tachycardia-syndrome/

Written and checked by the PocketPhysio editorial team. Last updated 2026-10-24.