Joint and Tendon Recovery for Masters Athletes

Here is what the tendon literature actually asks of you. One loading exercise. Three sessions a week. Loads on the order of 80 to 90 percent of a maximal voluntary contraction, repeated for up to three to four months — because tendon keeps its mechanosensitivity with age but answers slowly. Twelve weeks before you judge it at all; a year of follow-up in the trials that support loading. And a daily accounting kept not in the gym but the next morning, when the tendon files its report.

And here is what the most-quoted numbers rest on. Six volleyball players in a crossover study. Twenty jumping athletes, 18 of them men, mean age 22.5 years, holding isometric leg extensions at 60 degrees of knee flexion at 80 percent of maximal voluntary isometric contraction, four times per week. A randomized trial of heavy slow resistance versus eccentric training for Achilles tendinopathy. Small rooms, young knees, specific angles — findings that matter when you are 55, and that this piece flags every time they apply.

This is written for the 45-to-65-year-old who still trains seriously — the runner, cyclist, tennis or pickleball player, lifter, triathlete. It sets out what the loading research actually shows about rebuilding tendon, the realistic timeline, and where the evidence stops. It will not diagnose you or name the tissue producing your pain; for the anatomy-first framework that separates nerve pain from joint pain from tendon pain before you load anything, start with Pain Care, Explained.

What actually changes: tendon is usually the rate limiter

Muscle is not the thing holding you back. The review of leg extensor muscle-tendon properties with aging reports consistent reductions in triceps surae and quadriceps femoris strength alongside similar reductions in tendon stiffness and elastic modulus, while the effect of age on tendon cross-sectional area is described as unclear (Front Physiol review, 2018). The authors' reading is that the change in tendon stiffness comes predominantly from the material the tendon is made of rather than from its size.

Be careful with how far you carry that. These are narrative reviews, not pooled analyses, and they do not agree on everything. One concludes that chronic resistance training raises tendon stiffness, Young's modulus and even cross-sectional area, that eccentric work has been proposed as superior mainly because it permits higher loads, and then finds there may be no difference between contraction types for overall tendon adaptation, with eccentric work appealing for older adults chiefly because its demands are lower (J Funct Morphol Kinesiol review, 2019). Both reviews are dominated by the calf and quadriceps. Applying their conclusions to a shoulder is an assumption, not a finding.

The part that should change how you train is this: tendon appears to keep its mechanosensitivity with age. The aging muscle-tendon review reports that loading interventions in older adults produce changes in tendon properties of similar magnitude to those seen in younger adults over 12 to 14 weeks of training, and recommends tendon strains corresponding to high mechanical loads, on the order of 80 to 90 percent of a maximal voluntary contraction, with repetitive loading for up to three to four months (Front Physiol, 2018). Your tendon is not deaf. It is slow, and it wants a load big enough to be worth answering.

Underneath that sits a genuinely unresolved piece of biology. Connective tissue remodeling requires changes in both collagen synthesis and collagen degradation, and the review of muscle and tendon connective tissue adaptation is blunt that only sparse information exists on the degradation side (Connect Tissue Res review, 2014). That is why nobody can hand you a precise week at which the tissue is repaired. What can be described is the direction: unloading and exercise both alter connective tissue collagen synthesis rates, and unloading moves things the wrong way.

Why "just rest it" keeps failing

No trial in this evidence pool randomized people to rest versus loading, so the case against extended rest is mechanistic rather than head-to-head. It is still a real case. Tendons are described as mechanosensitive tissue that adapts its material and mechanical properties in response to mechanical loading, which means unloading through immobilization or inactivity is itself an input with consequences for tendon properties (Front Physiol, 2018). Rest removes the pain because it removes the stimulus. It also removes the stimulus.

What rest does buy you is the ability to calm an angry tendon enough to load it. That is a means, not the treatment. The distinction matters most for bone, where the rules genuinely are different: for low-risk bone stress injuries of the tibia and metatarsals, which account for more than half of bone stress injuries in runners, an initial period of load reduction through partial or non-weight bearing is described as typically necessary (J Orthop Sports Phys Ther, 2021). Tendon and bone do not follow the same script, which is one more reason to know what you are treating before you pick a protocol.

Isometrics: real, useful, and oversold

Isometric holds became popular on the strength of two small studies. In a crossover study of six volleyball players with patellar tendinopathy, isometric contractions cut pain during a single-leg decline squat from a mean of 7.0 to 0.17 on a 0-to-10 scale, against 6.33 to 3.75 for isotonic contractions, with the mean reduction reported as 6.8 points after isometrics versus 2.6 after isotonics; relief persisted 45 minutes after the isometric condition, maximal voluntary isometric contraction rose by 18.7 percent, and cortical inhibition fell (Br J Sports Med, 2015). In a four-week in-season trial, 20 jumping athletes (18 men, mean age 22.5 years) did isometric leg extension holds at 60 degrees of knee flexion at 80 percent of maximal voluntary isometric contraction, or isotonic leg extension at 80 percent of an 8-repetition maximum, four times per week; isometrics produced significantly greater immediate analgesia, and the size of the week-one analgesic response correlated with the four-week improvement in the VISA-P questionnaire (Clin J Sport Med, 2017).

Then it stopped replicating cleanly. A pre-registered randomized crossover trial in 21 people with patellar tendinopathy found that a single session of resistance exercise reduced decline-squat pain by a mean of 0.9 points on the 0-to-10 scale (95 percent CI 0.1 to 1.7), that the increase in pressure pain threshold was small and local, that neither effect was sustained at 45 minutes, and that there was no difference between isometric and dynamic exercise on any outcome (J Sci Med Sport, 2020). The network meta-analysis of patellar tendinopathy treatments, pooling three randomized trials, found isometric exercise about as effective as isotonic exercise for immediate post-intervention pain relief, with a mean difference of -1.03 points and a confidence interval crossing zero (BMJ Open Sport Exerc Med, 2021).

Honest grade on our evidence tiers: emerging and mixed. Use isometrics as a pre-loading tool on bad days, or before competing, because when they work they cost you nothing and they do not reduce strength. Do not build the program around them, and note that the two positive trials studied young jumping athletes, most of them male, while the null crossover trial does not report its participants' age or sport. None of them establishes anything about a 55-year-old.

Heavy slow resistance versus eccentrics: the head-to-head

This is the cleanest comparison available. Fifty-eight patients with chronic midportion Achilles tendinopathy of more than three months' duration were randomized to eccentric training or heavy slow resistance for 12 weeks (Am J Sports Med, 2015). Both groups improved significantly on the VISA-A questionnaire and on pain during activity, both held those gains at the 52-week follow-up, and tendon thickness and neovascularization fell in both. None of the clinical or structural improvements differed between groups.

What did differ was whether people could stick with it. Mean training session compliance was 92 percent for heavy slow resistance and 78 percent for eccentrics, a statistically significant gap, and patient satisfaction at 12 weeks was 100 percent with heavy slow resistance versus 80 percent with eccentrics, a difference that did not reach significance and had narrowed by 52 weeks (96 versus 76 percent). Read that as a usability finding, not a superiority finding.

The network meta-analysis reached a more conservative conclusion. Across 37 randomized trials assessing 33 different interventions, most represented by single studies, the interventions ranked highest carried low or very low strength of evidence, and the authors concluded that eccentric loading, with or without adjuncts, should remain first-line for patellar tendinopathy (BMJ Open Sport Exerc Med, 2021). In practice the two positions are compatible: slow, heavy, progressive load is the active ingredient, and the contraction style is the part you can choose based on what you will actually do three times a week.

There is one clear exception in this pool. In a Norwegian feasibility trial in tennis elbow, only 6 of 19 participants (32 percent) complied with heavy slow resistance training, mainly because it aggravated their pain, and the authors concluded the intervention was not suitable for that condition (BMJ Open, 2024). It was a feasibility trial, not powered to compare effectiveness, and all three groups improved. Still: lateral elbow pain is the site where you should expect to need a gentler entry point.

The protocol

The table below is built from the loading parameters that the trials in this pool actually reported. Where a trial did not publish a number, the cell says so rather than inventing one. Pick one tendon and one main lift for it; do not run three rehab programs at once.

About that pain rule, because this is where articles usually invent a number. For bone stress injury the guidance is explicit and strict: at every stage, the optimal load produces no symptoms during loading, after loading, or the day following it, and a return-to-run program starts only once the athlete has been pain free during daily activities for five consecutive days (J Orthop Sports Phys Ther, 2021). For tendinopathy, the trials cited above did not publish a numeric pain ceiling. So do not treat any specific number out of 10 as a validated threshold. Treat the next morning as the readout, and treat your own compliance as the second readout: the tennis elbow trial is a reminder that a load which drives people to quit is the wrong load regardless of how good it looks on paper (BMJ Open, 2024).

Progressions and regressions

When it gets easy, add load before you add speed. That sequencing has direct support on the bone side, where progress is symptom-driven and focuses on increasing running volume before running speed (J Orthop Sports Phys Ther, 2021), and it is consistent with the resistance training recommendation to move toward roughly 80 percent of 1RM for maximal strength gains (J Nutr Health Aging umbrella review, 2019). Plyometric and jumping work goes last, not first, though the same bone stress injury guidance notes that jump training and gait retraining after return to running may reduce the risk of a subsequent injury.

When it hurts more, the first move is to cut load, not to cut the habit. Dropping to three sessions of something you can tolerate beats a heroic week followed by three weeks off. The training load literature supports the general principle: acute spikes are the pattern to avoid, and monitoring should continue for at least four weeks after a spike (Sports Med systematic review, 2016).

When heavy load is simply not available to you, because the joint will not tolerate compression or the tendon is too reactive, that is where restriction cuffs come in. Cuffs are not for everyone: a history of venous thromboembolism, uncontrolled hypertension, peripheral vascular disease, or a limb that is already swollen or painful at rest rules them out until a clinician clears you, and occlusion training is not something to start on a self-made diagnosis.

Blood flow restriction: the option when heavy is not on the table

The most relevant trial is recent and directly on point. Participants with chronic unilateral patellar tendinopathy were randomized to 12 weeks of low-load blood flow restricted training (16 people) or heavy slow resistance training (20 people), and followed to 52 weeks (Scand J Med Sci Sports, 2025). Pain during a single-leg decline squat fell from 3.9 to 2.2 at 12 weeks and 1.8 at 52 weeks in the restricted-flow group, and from 4.2 to 2.2 and then 1.1 in the heavy group, with comparable and clinically relevant improvements in VISA-P scores in both. The participants were male, and 36 people is a small trial, so read this as a credible alternative rather than a proven equivalence.

In older adults more broadly, two meta-analyses agree on the basic effect and disagree on the ceiling. One, pooling 11 studies and 238 participants, found that adding blood flow restriction to low-load training produced substantially greater strength gains than low-load training alone, and that compared with high-load training it produced similar hypertrophy but lower strength gains (Sports Med meta-analysis, 2019). The other, pooling 10 studies and 278 healthy adults aged 60 and over, found strength favored restricted-flow training over conventional training with no statistical difference against high-intensity resistance training, and no difference between groups in physical performance (Arch Phys Med Rehabil meta-analysis, 2022). The sarcopenia umbrella review lists blood flow restriction as a method with a significant impact on muscle strength while recommending high-intensity resistance training for maximal gains (J Nutr Health Aging, 2019).

Practical reading: cuffs are a bridge, not a cure. They let you keep loading a limb during the window when heavy load is not tolerable, and in patellar tendinopathy they held up to a year. They do not replace the heavy work once you can do it, and none of these trials tested them as a standalone treatment. The mechanics of pressure, cuff width and who should not use them are covered in the blood flow restriction guide.

The knee osteoarthritis conversation

Masters athletes with knee osteoarthritis get told two things: stop running, and get strong. The evidence supports neither as confidently as it is delivered.

The START trial randomized 377 adults aged 50 and over with knee pain and radiographic knee osteoarthritis (mean age 65, 40 percent women) to high-intensity strength training, low-intensity strength training, or attention control for 18 months (JAMA, 2021). At follow-up, WOMAC pain scores were 5.1 in the high-intensity group versus 4.9 in control (adjusted difference 0.2, 95 percent CI -0.6 to 1.1) and 4.4 in the low-intensity group (adjusted difference 0.7, 95 percent CI -0.1 to 1.6). Knee joint compressive force during walking was 2453 N with high-intensity training versus 2512 N in control and 2475 N with low-intensity training, with no statistically significant differences. Two things follow. Lifting heavy did not reduce arthritic knee pain more than lifting light or than an attention control, which is an honest negative. And it did not raise measured joint compressive force either, which is worth knowing if you have been told that heavy strength work grinds the joint down. Nonserious adverse events were more common in the training groups (53 high-intensity, 30 low-intensity, 4 control).

The Cochrane review of land-based exercise for knee osteoarthritis pooled 139 trials and 12,468 participants (Cochrane review, 2024). Against attention control or placebo, exercise improved pain by a mean of 8.70 points on a 0-to-100 scale and physical function by 11.27 points. Against no treatment or usual care, pain improved by 13.14 points and function by 12.53. The reviewers then compared those figures with minimal important difference thresholds of 12 points for pain and 13 for function and concluded that the benefits, while probably real in the short term, were of uncertain clinical importance. They found no differences between types of exercise and no relationship between improvement and the number of sessions prescribed.

The meniscal tear question got a large answer recently. In 879 participants aged 45 to 85 with knee pain, osteoarthritis and a meniscal tear, adding supervised physical therapy or text-message adherence prompts to a three-month home exercise program produced changes in KOOS pain of 2.5 points or less, with confidence intervals crossing zero (N Engl J Med, 2025). The home program was the intervention that mattered.

What none of this shows is that running damages an arthritic knee. No trial in this pool tested running cessation. What the pool does support is modest: exercise helps somewhat, the type matters less than doing it, doing it at home counts, and grinding yourself into a high-intensity program on the promise of pain relief is not supported by the largest trial that tested exactly that.

Training load: the part that prevents the next flare

The systematic review of training load and injury pulled 35 studies and found moderate evidence of a relationship between training load and injury incidence, with a significant relationship reported in 93 percent of the studies examined, and moderate evidence for a relationship with illness in 75 percent (Sports Med, 2016). Notably, nine of the studies (31 percent) reported training load acting protectively against injury rather than harmfully, which is the part people skip. The review does not say which load patterns produced that, so read it as a reason to distrust one-line rules, not as proof that steady volume protects you. The practical recommendations are to monitor internal load as rating of perceived exertion multiplied by session duration, avoid acute spikes, and keep monitoring for at least four weeks after one.

Be careful with the popular version of this. A secondary analysis of 586 injury-free Danish recreational runners followed for 24 weeks with GPS data, in which 133 sustained a running-related injury, tested whether combining a weekly volume progression above 10 percent with a progression in pace interacted to raise injury risk. It did not reach statistical significance: the risk difference was 8.1 percent with a confidence interval running from -9.3 to 25.6 percent, and the absolute version, above 5 km of weekly progression plus faster pace, gave 5.2 percent with similarly wide bounds (J Athl Train, 2022). That is not evidence that spikes are safe. It is evidence that the tidy "10 percent rule plus pace" story has not been demonstrated, and that single-variable rules are weaker than they sound.

Recovery inputs that have evidence, graded honestly

Sleep first, because it costs nothing. The available trial is small: 10 male rugby union players did an evening training session involving collisions, then either extended sleep to 10 hours or attended an early morning recovery session (PLoS One, 2022). Sleep extension produced far greater total sleep time but also more fragmented sleep, and while cognitive performance recovered faster with sleep extension at 14 hours post-training, autonomic function and upper-body neuromuscular function were better in the active recovery condition, and no difference between conditions remained at 36 hours. The honest summary is that one long night helped the brain and did not restore the muscle, and this tells you nothing directly about tendon. Grade: preliminary.

Protein is on firmer ground for muscle and untested for tendon. A systematic review of protein supplementation alongside concurrent resistance and endurance training, the exact pattern a masters triathlete lives in, found that all four acute studies showed protein ingestion raised myofibrillar protein synthesis but not mitochondrial protein synthesis after concurrent exercise; among the longer-term studies, five of nine reported enhanced gains in muscle mass and five of nine reported enhanced gains in strength or power; and all six studies examining aerobic capacity found no effect on VO2max (Sports Med systematic review, 2022). That review does not state a gram-per-kilogram target, so this article will not invent one. Use the protein and macro calculator to set a target and eat it consistently.

Collagen peptides and vitamin C-enriched gelatin are where the marketing outruns the data, so read the numbers closely. A meta-analysis of 19 studies and 768 participants found statistically significant effects of long-term collagen peptide intake on fat-free mass, tendon morphology, muscle architecture, maximal strength and 48-hour recovery of reactive strength, with standardized mean differences ranging from 0.19 to 0.67; crucially, the GRADE certainty was moderate for body composition, low for most outcomes, and very low for tendon morphology and mechanical properties (Sports Med meta-analysis, 2024). The individual tendon trials are more striking and much narrower. Forty healthy men averaging 26 years old took 5 g of specific collagen peptides or placebo through 14 weeks of high-load resistance training; Achilles tendon cross-sectional area rose 11.0 percent with peptides versus 4.7 percent with placebo, and muscle thickness rose 7.3 versus 2.7 percent, but tendon stiffness and muscle strength improved equally in both groups (Scand J Med Sci Sports, 2022). A companion trial in 50 moderately active men using the same 5 g dose and 14 weeks of training at 70 to 85 percent of 1RM found greater patellar tendon cross-sectional area with peptides, again with no group difference in tendon stiffness, muscle cross-sectional area or knee extension strength (Eur J Sport Sci, 2023). The gelatin work is smaller still: eight healthy men taking 15 g of vitamin C-enriched gelatin one hour before six minutes of rope skipping, three times daily for three days, showed double the circulating marker of collagen synthesis (Am J Clin Nutr, 2017).

Put plainly: a thicker tendon on ultrasound is not the same as a tendon that hurts less or works better, and in these trials the functional outcomes that did improve improved because of the training, equally in both groups. Every one of these studies was done in young men. None studied tendinopathy patients, and none studied anyone your age. The doses used in the trials were 5 g of collagen peptides daily alongside 14 weeks of training, or 15 g of vitamin C-enriched gelatin an hour before loading, and the same amino acids come from food, which is covered in the grocery swaps guide.

Two compounds people ask about are missing here for a simple reason. Nothing in this evidence pool tested creatine or omega-3 for tendon outcomes, so this article makes no claim about either; they are handled separately in the creatine review.

What to skip

• Extracorporeal shockwave therapy as an add-on for patellar tendinopathy. Pooled across two randomized trials, it was no better than sham for short-term pain (mean difference 0.1 points) or function, with both groups performing eccentric exercise (BMJ Open Sport Exerc Med, 2021).
• Routine anti-inflammatory medication to get through a loading program. The connective tissue review is explicit that non-steroidal anti-inflammatory drugs appear to inhibit the healing process of connective tissue and the stimulating effect of exercise on connective tissue protein synthesis, while noting these drugs are commonly consumed for exactly these injuries and that the question needs further investigation (Connect Tissue Res review, 2014). The bone stress injury guidance restricts them to short-term use, sooner than seven days, and only for resting pain and night pain (J Orthop Sports Phys Ther, 2021). That is a narrow window, and the decision belongs with your clinician, not with an article. Topical options are a different route and are covered separately in topical pain relief; nothing in this evidence pool tested them.
• Paying for supervised sessions on the assumption that supervision is the active ingredient, at least for degenerative meniscal tear with knee osteoarthritis, where adding standard physical therapy or adherence texts to a home program produced differences of 2.5 KOOS points or less with confidence intervals crossing zero (N Engl J Med, 2025).
• Hunting for the perfect exercise or buying more sessions. The Cochrane review found no differences in effect between types of exercise and no relationship between improvement and the number of sessions prescribed (Cochrane review, 2024).
• Expecting a supplement to substitute for the loading. The sarcopenia umbrella review concluded that the added effect of nutritional supplementation on top of resistance training appears limited, while the evidence for resistance training itself on muscle mass, strength and physical performance was rated high quality (J Nutr Health Aging, 2019).
• Open-ended rest. It removes the stimulus the tissue adapts to, and no trial here supports it as a treatment for tendinopathy.

Red flags: stop and get assessed

Loading protocols assume a diagnosis. These features mean the assumption may be wrong, and they are the reason a program should not be started off the internet alone.

• Pain that wakes you at night, or is present at rest. Resting and night pain are singled out in the bone stress injury guidance as the symptoms warranting short-term analgesia, and they are not the pattern of a straightforward tendinopathy.
• A pop or snap with immediate loss of function. Sudden inability to push off, raise the arm, or straighten the knee after an audible event needs same-day assessment, not a rehab plan.
• Rapid swelling, a hot joint, or swelling out of proportion to the incident.
• Numbness, pins and needles, or weakness in a nerve pattern. That is a different problem with a different workup, and a tendon loading program is not the answer to it.
• A swollen, hot or tender calf, or new breathlessness or chest pain. That pattern raises the possibility of a venous thromboembolism and needs same-day assessment. It also rules out blood flow restriction training until it has been excluded.
• Systemic symptoms. Fever, unexplained weight loss, night sweats, or multiple joints becoming painful at once.
• Bone pain in a runner that worsens through a run rather than warming up, particularly with a recent jump in volume. Low-risk tibial and metatarsal stress injuries account for more than half of bone stress injuries in runners and are managed by initial load reduction, which is the opposite of the tendon plan.

How to tell whether it is working

Use the same instruments the trials used, because they are free and they are sensitive to change.

1. One provocative test, scored 0 to 10, at the same time of day. The patellar tendinopathy trials used pain during a single-leg decline squat as the primary outcome. Pick your equivalent, score it weekly, and do not re-test it constantly.
2. The next-morning score. The bone stress injury guidance treats symptoms the day after loading as the signal that the dose was wrong, and it is the most useful daily readout you have for tendon too.
3. A validated questionnaire. VISA-A for Achilles and VISA-P for patellar tendon were the outcome measures in the trials above; KOOS and WOMAC are the knee instruments. In the START trial the minimal clinically important difference for WOMAC pain was 2 points on a 0-to-20 scale; the Cochrane review used thresholds of 12 points for pain and 13 for function on 0-to-100 scales. Anything smaller than that is noise dressed as progress.
4. A load diary. Rating of perceived exertion multiplied by session duration is the internal load metric the training load review recommends, and it is what lets you see a spike before your tendon does.
5. Objective function, tested the same way each time. The five norm-referenced tests in the Movement Lab give you numbers that do not depend on how you feel that day.

Timeline: the head-to-head Achilles trial measured its improvements over 12 weeks and found them maintained at 52 weeks, and the patellar tendinopathy trial comparing cuffs with heavy loading showed pain still falling between 12 and 52 weeks. The aging muscle-tendon review recommends up to three to four months of repetitive high loading to change tendon properties in older adults. Twelve weeks is when you should expect to be able to see a trend. It is not when you stop.

If 12 weeks of consistent, progressive, well-tolerated loading produces no trend at all in your provocative test score or your questionnaire, that is useful information: it is the point to get reassessed rather than to add a fourth session or a new supplement. Educational, not medical advice.