What it actually feels like
It starts as a small ache on the bony point of the elbow and it does not announce itself. People notice it first in ordinary movements rather than in sport — lifting a coffee cup, turning a doorknob, shaking hands, picking up a bag with the palm facing down. The pain is sharp and specific, it can be pointed to with one fingertip, and it is reliably reproduced by gripping something while the elbow is straight.
What surprises most people is the weakness. Grip strength drops early and drops more than the pain alone would explain, because the muscles that stabilise the wrist attach at exactly the point that hurts. Patients describe dropping things, or being unable to hold a hairdryer up. That is not a nerve problem and it is not muscle wasting. It is the tendon refusing to transmit load.
The pain is often worse the day after activity rather than during it, which is why the cause is so frequently missed. Someone spends a weekend painting a fence, feels fine, and blames the pain that arrives on Tuesday on something they did on Tuesday.
Why "tendinitis" is the wrong word
Tennis elbow is still commonly called lateral epicondylitis, and the suffix is a leftover from an era when the tendon was assumed to be inflamed. When tissue from these tendons is actually examined, what is found is not inflammation. It is disorganised collagen, an ingrowth of small blood vessels and nerve fibres where neither belongs, and a general failure of the tendon's repair machinery. This is degeneration, not inflammation — tendinosis or tendinopathy rather than tendinitis.
That distinction is not academic, because it dictates what should work. An anti-inflammatory strategy applied to a problem that is not inflammatory is a plausible-sounding treatment aimed at the wrong target. It explains the pattern seen with corticosteroid, and it explains why loading the tendon — which feels counterintuitive when something hurts — is the intervention with the best long-term record.
Tennis elbow and golfer's elbow are the same problem in two places
On the outside of the elbow, the wrist extensors share a common origin at the lateral epicondyle, and the tendon of extensor carpi radialis brevis is where the trouble almost always begins. On the inside, the wrist flexors and pronators share a common origin at the medial epicondyle. The pathology is identical. Only the location and the loading pattern differ: extension and gripping on the outside, flexion and pronation on the inside.
Golfer's elbow is roughly four to five times less common than tennis elbow, and it deserves more caution rather than less. The ulnar nerve runs in a groove immediately behind the medial epicondyle, and it is either irritated alongside the tendon or mistaken for it. Any numbness, tingling or burning into the ring and little fingers moves the nerve to the centre of the assessment. It also constrains where a needle can be placed, which is one of several reasons these injections should be done under direct ultrasound visualisation rather than by feel.
In throwing athletes, medial elbow pain carries a further obligation: the ulnar collateral ligament sits just deep to the flexor-pronator origin, and a partial UCL injury can present as what looks like straightforward golfer's elbow. Missing that distinction in a thrower has consequences that missing it in a golfer does not.
Ulnar collateral ligament injuries in throwers
We see the diagnosis of ulnar collateral ligament (UCL) tears in more and more of our throwing athletes, and for most of them an orthobiologic approach is far preferable to surgical reconstruction. A partial-thickness UCL tear is a different injury from a complete one, dynamic ultrasound distinguishes them under valgus stress in the office, and the two do not warrant the same treatment.
That is a large enough subject to have its own page: UCL tears and the alternatives to Tommy John surgery.
How the elbow is examined here
The examination is unhurried and it is largely mechanical. Where exactly does it hurt, to one fingertip. Does resisted wrist extension with the elbow straight reproduce it. Does resisted middle-finger extension. Does resisted pronation or wrist flexion reproduce it on the inside instead. Is the ulnar nerve tender, and does tapping over it send symptoms into the hand. Is the elbow stable to valgus stress. Is the neck contributing, because referred pain from a cervical root can mimic all of this and will not respond to anything done at the elbow.
Then the ultrasound, in the same visit, in the same room. Ultrasound shows the tendon under load and in motion, which is something no static image can do. It shows thickening, the loss of the normal fibrillar pattern, calcification, neovascularity, and whether there is a partial-thickness tear within the tendon — and if there is, how big it is and exactly where. That last point changes management. A tendinopathic tendon and a tendon with a significant intrasubstance tear are not the same clinical problem, and treating them the same way is how patients end up in month fourteen of a twelve-month condition.
MRI is ordered when the story does not fit: suspected ligament injury, a possible loose body, a nerve compression that is not behaving like a simple irritation, or a tear extensive enough that surgical planning depends on knowing its dimensions precisely.
What the evidence supports, and what it does not
The corticosteroid problem
Corticosteroid injection for tennis elbow does something no other treatment does: it works quickly and reliably in the short term, and it appears to leave patients worse off at a year than doing nothing at all. Randomized trials have repeatedly found that the corticosteroid group is better at six weeks and worse at six and twelve months than placebo or wait-and-see, with higher recurrence rates.
This is a genuinely unusual result and it is not a reason to abolish the injection. It is a reason to be explicit about what is being traded. If someone has a wedding, a deposition, a tournament, or a fortnight of unavoidable manual work, buying six weeks of function is a legitimate choice made with open eyes. Repeated corticosteroid into a degenerative tendon is a different matter, and the local effects on tendon and subcutaneous tissue compound with each injection.
The dose finding that explains why the trials disagree
If you read the PRP literature for tennis elbow you will find studies that show a large benefit and studies that show none, and it is tempting to conclude the treatment is unreliable. Oeding and colleagues (American Journal of Sports Medicine, 2025) offers a better explanation, and it is the single most useful paper on this subject.
They pooled thirteen randomized controlled trials covering 791 patients and ran a meta-regression against platelet dose. The result splits cleanly. High-dose PRP produced a mean pain reduction of −1.31 on the visual analogue scale (95% CI −1.87 to −0.75) against comparators. Low-dose PRP produced a mean difference of 0.08 — that is, nothing at all.
And the variable explained the disagreement: platelet concentration factor accounted for 58.5 percent of the between-study heterogeneity. Well over half of the reason these trials contradict each other is dose. The authors define the threshold that matters as a supraphysiologic dose — a platelet concentration factor greater than three-fold over whole blood — and describe a direct linear relationship between concentration and symptom relief.
That is not a subtle effect at the margins. It means a substantial part of the published literature is testing an intervention that was never going to work, and then reporting that it did not.
What the supportive trials show
Lhee and colleagues (American Journal of Sports Medicine, 2025) is the strongest recent randomized trial: 231 patients with chronic lateral epicondylosis refractory to at least three months of conservative care, treated with a single PRP injection and followed for two years. PRP reduced DASH scores by 31.18 points, against 18.70 for physiotherapy and 17.62 for shockwave therapy. Patient satisfaction was highest in the PRP group, and the benefit was still present at twenty-four months.
Xu and colleagues (American Journal of Sports Medicine, 2024) pooled eleven randomized trials and 730 patients comparing PRP against corticosteroid, and found the time-dependent pattern this page keeps returning to: corticosteroid does better before two months, PRP does better from six months onward, with significant advantages at long-term follow-up across pain, DASH and Mayo Elbow scores. Huang and Hohmann and their respective colleagues report the same crossover.
The studies that found nothing
These deserve equal billing. A 2026 Level 1 placebo-controlled meta-analysis by Antunes Júnior and colleagues found no significant benefit for PRP against saline injection. A 2023 review in the New England Journal of Medicine reached the same conclusion. The Canadian Shoulder and Elbow Society recommends against injection treatment for this condition altogether.
Against that, the AMSSM position statement notes that lateral epicondylopathy has the most robust PRP evidence of any tendinopathy — which tells you something about the state of the wider field rather than settling this one.
The fair summary is that the supportive case for PRP in tennis elbow rests on comparisons against corticosteroid, physiotherapy and shockwave, and on adequately dosed preparations. It does not rest on beating saline.
Why we use high-dose PRP, and a single injection
Read those two sets of findings together and a pattern emerges that is worth stating directly.
The preparation protocols reported across this literature vary widely, and they frequently fall below the threshold Oeding identifies. Reported protocols include autologous conditioned plasma at roughly 1.6-fold platelet enrichment from a 12 mL draw, and single-spin leukocyte-poor PRP at around 2.3-fold enrichment from about 24 mL — both under the greater-than-three-fold concentration factor that the meta-regression identifies as the point where the dose-response relationship delivers. Some studies report no platelet concentration at all, which makes them impossible to place on that curve.
When a body of evidence pools interventions that share a name but differ several-fold in the active ingredient, "the results are inconsistent" is the expected finding rather than an informative one.
Which is why at Texas Orthobiologics we specialize in high-dose PRP injections. Not low-dose, and not a course of repeated low-dose injections in the hope that volume substitutes for concentration. The dose is the treatment.
Leukocyte content gets more attention than it deserves in this conversation. Li and colleagues (2022) compared leukocyte-poor against leukocyte-rich preparations and found comparable efficacy, with lower rates of post-injection pain in the leukocyte-poor group — a tolerability difference rather than an efficacy one. Dose matters more than leukocyte content.
Three things should make you sceptical of anyone offering PRP for this condition: no ultrasound guidance, no answer when you ask what platelet concentration their preparation achieves, and a plan involving a series of injections rather than one properly dosed one.
The options, all of them
Load management and tendon-specific rehabilitation
This is first-line and it is not a consolation prize. Progressive loading of the affected tendon — eccentric or heavy slow resistance protocols — has the best long-term evidence of anything on this list. It is also the least popular, because it takes twelve weeks of unglamorous daily work and it is uncomfortable in the early stages. Rest alone does not do it. A tendon that is not loaded does not remodel; it simply becomes a tendon that hurts less until it is loaded again.
Bracing
A counterforce strap worn a few centimetres below the elbow shifts the point of load away from the origin. It provides symptomatic relief for some people during activity and does nothing for the underlying tendon. Useful adjunct, not a treatment.
Corticosteroid injection
Discussed above. Fast, reliable, short-lived, and with a real cost at twelve months. Occasionally the right answer for a specific reason and a specific deadline.
Orthobiologic injection
Platelet-rich plasma, prepared in the office from the patient's own blood on the same day, injected under ultrasound guidance into the pathological portion of the tendon. Where there is a partial-thickness intrasubstance tear, the target is the tear itself rather than the general area. Expect soreness for several days afterwards — the treatment provokes a healing response, and provocation is uncomfortable. Anti-inflammatories are avoided around the procedure for the same reason the diagnosis is not "tendinitis."
Bone marrow aspirate concentrate
We typically reserve the use of bone marrow aspirate concentrate (stem cell) injections for high-grade injuries or injuries that have failed to improve with other treatments. It is not a first-line option for a straightforward tendinopathy, and it is not offered as one here. Where it earns its place is the tendon with a substantial intrasubstance tear, or the elbow that has genuinely completed a loading programme and had a well-prepared PRP injection and still has not turned the corner.
Ultrasound-guided percutaneous tenotomy
There are other devices based on ultrasound guidance, such as Tenex, that give surgeons another option to treat elbow tendinopathy before surgical intervention. These use ultrasonic energy through a small percutaneous port to debride the degenerative portion of the tendon under direct visualisation, without an open incision. It sits between injection and surgery, and having it available means the step after a failed injection is not automatically an operating room.
Class IV MLS laser therapy
Used here as an adjunct rather than as a stand-alone treatment, typically alongside a loading programme or after an injection. It is comfortable, has no downtime, and the evidence base is smaller than for loading or for PRP. It is offered on that basis and not as a substitute for either.
Surgery
Over the last 20 years, with the rising success of high-dose PRP, the need for surgery for tennis elbow has dramatically decreased in my practice. Roughly nine in ten cases settle without an operation even before that shift is counted, and the combination of high-dose platelet-rich plasma, ultrasound-guided percutaneous options and properly executed loading has narrowed the remaining group considerably.
Where an operation is still the right answer, surgical debridement of the degenerative tissue — open or arthroscopic — is reserved for patients who have genuinely completed a structured loading programme, given it enough time, exhausted the less invasive options above, and remain limited. The most common reason a tennis elbow operation disappoints is that it was performed for a diagnosis that was never the whole story.
How the choice actually gets made
Three things drive it: how long it has been going on, whether the ultrasound shows a tear, and what the patient has already had done.
Short history, no tear, nothing tried yet — load it properly and give it time. That is not a brush-off; it is the intervention with the best twelve-month data, and starting anywhere else forfeits it.
Six months or more, a structured programme genuinely completed, still limited — this is where orthobiologic injection earns its place, because the natural history has had its chance and the tendon has demonstrated that it is not going to remodel unaided.
A visible intrasubstance tear on ultrasound — the calculus changes. A torn tendon is a structural problem and the injection is aimed at the defect, not at the neighbourhood.
Two or three previous corticosteroid injections — the tendon has already been given the treatment that worsens twelve-month outcomes, more than once, and both the tissue quality and the expectations need to be discussed honestly before anything else is offered.
Why a surgeon is the one saying this
Dr. Buford is a board-certified orthopedic surgeon who performs elbow surgery. That matters when the recommendation is not to operate, because the person telling you an operation is unnecessary is the person who would otherwise be doing it. It also matters in the other direction: a clinic that cannot offer surgery has no way to tell you when you have crossed the line into needing it, and no obligation to notice.
What recovery involves
After an orthobiologic injection, expect the elbow to be more sore for three to five days. That is the point of the procedure and it is not a complication. Anti-inflammatory medication is avoided for a defined period before and after. Loading restarts deliberately and progressively rather than all at once, usually beginning within the first week or two.
Meaningful change is measured in months, not weeks. Most people notice a difference somewhere between six and twelve weeks, and improvement often continues past that. Anyone promising a rapid fix for a degenerative tendon is describing a corticosteroid injection, whether or not they say so.
Grip strength is the most useful thing to track, because it improves before the pain fully settles and it is objective. A simple dynamometer reading at each visit tells you more about progress than a pain score does.
Common questions
Is tennis elbow actually caused by playing tennis?
Usually not. Most people we treat for tennis elbow have never held a racquet. The tendon on the outside of the elbow is loaded by any repeated gripping, lifting with the palm down, or wrist extension against resistance, which describes carpentry, cooking, dental work, surgery, plumbing, and holding a phone or a toddler. Racquet sports are one cause among many, and they are not the most common one.
Will a cortisone shot fix tennis elbow?
A corticosteroid injection reliably reduces pain for a few weeks. The difficulty is what happens after that. In the better-designed trials, patients who received corticosteroid were doing worse at six and twelve months than patients who received placebo or who simply waited, and they had higher recurrence rates. That is an unusual finding in medicine and it is worth taking seriously. Corticosteroid still has a role when someone needs a specific window of function, but it should be a deliberate short-term trade rather than a default first step.
Does PRP work for tennis elbow?
It depends almost entirely on the dose, and that is not a hedge. A 2025 meta-regression of thirteen randomized trials found that high-dose PRP reduced pain substantially while low-dose PRP produced a mean difference of 0.08, meaning no effect at all, and that platelet concentration factor alone explained 58.5 percent of the disagreement between studies. Against corticosteroid, physiotherapy and shockwave therapy the supportive trials are consistent and durable to two years. Against saline the picture is less favourable. This is why we use high-dose preparations rather than low-dose ones, and a single properly dosed injection rather than a series.
I read that PRP is no better than a saline injection. Is that true?
Some placebo-controlled analyses do report that, including a 2026 Level 1 meta-analysis and a 2023 review in the New England Journal of Medicine, and we do not dismiss them. The complication is dose. The 2025 meta-regression by Oeding and colleagues found that efficacy requires a platelet concentration factor greater than three-fold over whole blood, and that below that threshold PRP produces no measurable benefit. The preparations reported in much of the trial literature achieve roughly 1.6-fold to 2.3-fold enrichment, which is below that threshold. A study of an under-dosed preparation is evidence about that preparation, not about the treatment.
How is golfer's elbow different from tennis elbow?
It is the same disease process on the opposite side of the joint. Tennis elbow affects the common extensor origin on the outside of the elbow; golfer's elbow affects the common flexor-pronator origin on the inside. Golfer's elbow is roughly four to five times less common, and it matters more because the ulnar nerve runs directly behind that spot. Numbness or tingling into the ring and little fingers changes the assessment and changes where a needle can safely go.
How long does tennis elbow take to get better?
Most cases resolve within about a year whether or not they are treated, which is the single most important fact about this condition and the reason to be sceptical of any treatment claiming credit for a recovery at month eleven. The reason to treat is to shorten that year, to avoid the cases that do not resolve, and to avoid interventions that make the twelve-month outcome worse.
Do I need an MRI for tennis elbow?
Usually not. Tennis elbow is a clinical diagnosis supported by ultrasound, which shows tendon thickening, loss of the normal fibrillar pattern, neovascularity and any partial tear, in real time, in the office, while the elbow is moving. MRI becomes useful when the history suggests something else is going on: a ligament problem, a nerve compression, an intra-articular loose body, or a tear extensive enough to change the plan.
Who performs the injection?
Dr. Buford performs every injection himself, under direct ultrasound guidance. No physician assistant, nurse practitioner, or technician performs orthobiologic injections at Texas Orthobiologics. Call 877-777-8883 to arrange an evaluation.
