Keith Baar's Tendon Isometric Protocol Explained

Imagine handing an engineer two materials, a wet sponge and a block of granite, and asking her to bolt them together so the joint can carry a person's full body weight, over and over, for seventy years, without failing. She would, in the words of physiologist Keith Baar, 'break out into a cold sweat' (07:14). That absurd assignment is also an exact description of a tendon: the cord that has to attach something soft, the muscle, to something hard, the bone, and keep doing it for a lifetime.

Baar runs the Functional Molecular Biology Laboratory at the University of California, Davis (01:56), and for much of his career he trained athletes the way almost everyone still does: heavier is better. On this episode of BETTER! with Dr. Stephanie Estima, he opens by taking that back. "Heavy lifting is good for tendons... that, we've changed our mind about" (00:00), he says, adding that working with elite athletes had baked the assumption in before anyone tested it. What follows is an argument that muscle and tendon need almost opposite instructions, even though they share a limb.

A tendon connects muscle to bone and transmits force. A ligament connects bone to bone and resists unwanted motion, which is why a healthy ligament should be as stiff as possible while a tendon should not (08:12). Problems show up at the junctions.

"We pull muscles when the tendon is stiffer than the muscle is strong." — Keith Baar [09:38]

The more surprising finding concerns dosage. Muscle responds to exercise for as long as someone keeps lifting or running. Tendon, bone, and ligament do not. Baar's lab has found that connective tissue gets a maximal training signal within about ten minutes, no matter how many repetitions happen inside that window, and nothing extra accrues afterward. Bone is even quicker: roughly 40 loading repetitions, like jump landings, max out its benefit, though the fewest loads Baar has seen in human studies come from a Japanese group in which young women jumped as high as they could four times, three times a week, and still raised bone mineral density over time (15:56, 16:28). Everything done after those thresholds still helps the heart and the muscles. For the tendon, it is just wear and tear.

"Our connective tissues... they stop getting the positive signal after about 10 minutes, and then what happens is everything else is wear and tear." — Keith Baar [16:28]

Why does the signal cap out so fast? Baar compares it to smell. Walk into a smoky room and the nose registers it instantly, then stops registering it within a minute, because the receptors that sense change have already opened and saturated (27:46). Tendon cells work the same way: they sense change, not steady load, so once their channels are full, adding more exercise just overflows a tub that is already brimming (28:44). The system needs about six to eight hours to drain before it can respond to a fresh stimulus (30:08). That gives training a variable most people never program: architecture, meaning how the same total workout is arranged in time. Split one hour of exercise into two 30-minute sessions six to eight hours apart, and the tendon gets two full ten-minute doses instead of one, for identical total work (17:58).

The Hold That Hurts for a Reason

This is where isometrics, holds in which a muscle contracts without changing the joint angle, enter the picture, and where the familiar misery of a wall sit gets an explanation. A muscle alone can hold a position almost without effort, which is why rigor mortis, Baar points out, locks a dead muscle in full contraction at no energy cost (35:50). The burn comes from the tendon, which goes through a process called creep: under sustained load, the strongest, stiffest part of the tendon lengthens first, handing the load to the next-strongest section, and so on, until a long enough hold has spread tension evenly across the whole structure (36:09). Fast, explosive movements, by contrast, load only the tendon's strongest pockets for a fraction of a second and never recruit the rest (33:28).

Baar's prescription is what he calls the 3LD protocol: low load, long duration. Four 30-second holds produce measurably better tendon adaptation than twelve 10-second holds at the same total load of 120 seconds (39:20). Below 30 seconds, the creep does not have time to spread through the tissue. He distinguishes two variants. An overcoming isometric pushes against an immovable resistance and is capped by how much force the muscle can actively generate; Baar now uses it at only 40 to 50 percent of maximum strength, coached by the cue "feel tension, not pain," or 'ooh,' not 'ow' (48:50). A yielding isometric, by contrast, controls a weight on the way down and can exceed full strength because muscle is about 1.8 times stronger lengthening than shortening (48:04); it was the old high-load approach to tendon rehab and is no longer Baar's first choice, though combining the two, light isometrics most days and one heavy session a week, produced an additive gain in grip strength in his studies with rock climbers (51:42).

"You don't win by getting the most load through your tendon. You win by holding that ooh state for longer period of time." — Keith Baar [77:10]

Hormones Write the Injury Report

The conversation's other thread is why rest, the default advice for an injured tendon, often backfires. A classic Japanese study, led by a researcher named Hayashi, rerouted load around an intact tendon using a wire and found that the unloaded tissue developed five times as many cells, with smaller, disorganized collagen that never lined up in one direction, the molecular signature of what gets called a scar (58:15). That is Baar's case against RICE, the rest-ice-compression-elevation protocol. "Rest" and "compression" together immobilize the tendon, and immobilized collagen grows mechanically inferior even without new damage (61:35). His replacement pairs two movements: a slow, rhythmic flex-and-extend to pump swelling fluid out of the tissue, and a slow-ramped isometric hold, built up over three to five seconds instead of snapped on suddenly, because sudden load, or 'jerk,' is usually what caused the injury in the first place.

"Slow rhythmic load is designed just to pump the fluid out of the tissue. That's what we're gonna use as our anti-inflammatory." — Keith Baar [64:54]

Sex differences run through all of it. Baar's research shows estrogen inhibits the enzyme that cross-links collagen, which lowers tendon and ligament stiffness; testosterone does the opposite (77:55-78:25). The tradeoff cuts both ways: women get about four times more ACL ruptures than men in the same sport, but 70 percent fewer muscle pulls, because a looser tendon is less likely to outpace the muscle's strength (79:51). The ACL risk is not steady across a cycle. Knee laxity, how far the joint shifts under a standard test, rises measurably around ovulation, and roughly 70 percent of ACL ruptures cluster around that ovulatory or luteal estrogen surge (80:59). Illness alone, Baar's team has newly found, can also acutely loosen the knee for a few days (85:00), and oral contraceptives, by smoothing out the estrogen surge, are associated with about a 20 percent lower ACL rupture risk (86:48). Perimenopause complicates the picture further: as estrogen falls and testosterone's relative influence rises, tendons can stiffen in ways that mirror what happens to women on aromatase inhibitors for breast cancer, who commonly report new tendon and joint pain (81:00-81:34).

Two smaller claims round out the hour. Heels, Baar says, reproduce the knee position associated with osteoarthritis, a joint disease marked by cartilage breakdown (00:44). And BPC-157, a peptide popular in longevity circles, showed no direct positive effect on tendon or ligament tissue in his lab's engineered models, and sometimes made them worse (02:49). Neither claim is the headline, but both fit the episode's real argument: tendons are not small, dumb versions of muscle. They run on their own clock, their own dose, and their own chemistry, and training them well means learning to listen for the difference.