Is training to muscle failure better or worse for your joints?
Whether training to muscle failure is good or bad for your joints depends strongly on how you do it: a slow, controlled approach produces considerably less joint stress than an explosive method emphasising the lengthening phase of movement, and the risk is smaller with healthy joints than with osteoarthritis.
Training to muscle failure with a light load (40% of the maximum weight you can lift once) puts as much strain on your heart and blood vessels as training with heavy weights, but probably places less pressure on the joints. This comes from a small study in people with a heart condition, and joint load itself was not directly measured in that study, so the conclusion is tentative.
How you train to muscle failure makes a big difference. Standard training in which the muscle lengthens while under tension produced, at the point of failure, substantial acute muscle damage and a temporarily sharp reduction in the range of motion at the elbow joint (-27%). An alternative approach, in which the movement is much slower and more controlled and the muscle stays under tension for more than 3.5 times as long, produced noticeably less muscle damage and only a -13% loss in joint mobility. That damage largely recovered within seven days, but the difference between the two methods was considerable.
For people with osteoarthritis (joint wear and tear), an extra nuance applies. In early osteoarthritis, more intensive training appears beneficial, whereas in advanced wear and tear, moderate intensity is a better fit. In that context, training to muscle failure is not recommended.
The question of whether training to muscle failure is good or bad for your joints therefore has no straightforward answer. It depends on the training method and on the condition of your joints. Someone with healthy joints who trains in a controlled way faces less risk than someone who pushes to the limit with a heavy approach that emphasises the lengthening phase of the movement.
This answer is based on three small studies in people, including one in patients with a heart condition (18 participants) and one comparison study of two forms of training that emphasise the lengthening phase of movement. Joint load was not directly measured in any of the studies; muscle damage was inferred from a blood marker (creatine kinase) and loss of range of motion. The researchers themselves rate the strength of the evidence as limited to moderate: the groups were small, usually only one exercise was studied, and follow-up was short. For osteoarthritis the evidence is somewhat stronger, but no studies in which participants were randomly assigned are available specifically on training to muscle failure.