Running shoes
Evidence: moderate
Choose shoes by comfort and fit, not by arch height or pronation, which the evidence does not support as a prescription. Rotating a few pairs matters more than picking the “right” category, and the super-shoe is a racing tool, not a daily trainer.
A runner needs shoes that fit and feel comfortable, and little of the elaborate fitting ritual around them survives scrutiny. The single most useful idea is that the “right” shoe is mostly the one that feels right to the individual, not the one a category chart assigns.
The pronation and arch model has largely failed
For decades runners were sorted by how much their feet roll inward (pronation) or by arch height, then prescribed motion-control, stability or neutral shoes to match. The evidence does not support this as a way to prevent injury: matching shoe category to foot type does not reduce injury overall (shoe-selection evidence). One analysis did find that motion-control shoes reduced pronation-related injuries specifically in already-pronated runners, so the idea is not entirely empty, but it does not generalise, and prescribing from a static foot assessment is not supported for most runners.
Comfort and rotation are better guides
The better-supported framework is Nigg’s “preferred movement path” and comfort filter: people self-select the shoe that lets their legs move the way they naturally want to, and choosing for comfort tends to reduce injury and the metabolic cost of running (shoe-selection evidence). Rotating between a few different pairs is associated with a markedly lower injury rate, plausibly because it varies the loading on tissues, and it matters more than the stability-versus-neutral choice.
A simple shoe wardrobe
- A comfortable daily trainer for the bulk of easy mileage. Cushioned to taste; durable; chosen by feel, not by a wet-foot test.
- A lighter, faster shoe for workouts and tempo runs, if wanted.
- A racing super-shoe for races and key sessions only, kept out of daily training to preserve its foam.
Replace trainers when the cushioning feels dead or the outsole is worn through, commonly somewhere in the region of 500 to 800 km but highly variable; track how the legs feel rather than a fixed number. New runners do not need more than one comfortable pair to start; see running for beginners.
Cushioning, drop and minimalism
Cushioning protects against impact and changes ride comfort. There is no strong evidence that a particular amount of cushioning reliably prevents injury, so the exact level is largely a comfort choice, but the broader principle still favours it: the main driver of improvement is training volume, so a shoe that reduces per-stride load and lets a runner train more, more comfortably, with fewer breakdowns is doing useful work, even if it offers no direct performance gain. That is the strongest general argument for a well-cushioned daily trainer, and the argument against the barefoot end of the spectrum, which tends to lower the volume most runners can safely sustain. Heel-to-toe drop is, similarly, mostly a comfort and habit choice.