Choosing Running Shoes

Choose by comfort, not foot type

For years runners were sorted into neutral, stability and motion-control shoes by how much the foot rolls inward (pronates) or by arch height, on the idea that matching the shoe to the foot prevents injury. The trials do not support it. A 2022 Cochrane review pooled three trials of shoes prescribed on foot posture (7,203 participants, all military recruits) and rated it moderate-certainty evidence that this probably makes little or no difference to injuries. The same review compared shoe types head to head and found low or very low certainty throughout, with most evidence showing no reduction in injuries when one type is compared with another¹.

Those trials were in military recruits. In a 2011 trial of 81 women training for a half marathon who were randomly given a neutral, stability or motion-control shoe for 13 weeks, the motion-control shoe went with the most missed training days (79, against 51 for stability) and, in neutral and pronated feet, more pain².

Comfort is the better guide. In 527 recreational runners followed for 6 months, those who rated the study shoe's cushioning moderate or high, or rated the shoe highest overall, had lower injury risk, and runners in the highest third for overall appreciation had about half the risk of those in the lowest third. That compares runners within one trial, so it supports comfort without proving it³.

Judge comfort yourself on a run, not from a label or a recommendation. In 21 women who ran in two shoes that were identical except for color, one presented as "gait-matched" to them, the gait-matched pair scored higher for comfort (66.1 against 49.0 on a 100 mm scale)⁴.

Fit and the shoe types

A good fit leaves room so your toes do not touch the front, holds the heel without slipping, and has no rubbing or pressure points.

The shoe categories are descriptions by makers and shops:

• Neutral: no added support features.
• Stability and motion control: added structure meant to limit inward roll of the foot, motion control being the stronger version.
• Cushioning and heel-to-toe drop: how much foam is under the foot and how much higher the heel sits than the forefoot.
• Trail: more grip and protection for off-road footing.
• Racing and "super" shoes: see the guide on carbon-plate shoes.

Cushioning, drop and minimalist shoes

More cushioning is not a proven way to avoid injury. In a 2020 trial of 848 recreational runners randomly given a soft or a hard version of a shoe that differed only in cushioning, the hard shoe went with about 1.5 times the injury risk. That held for lighter runners (about 1.8 times) and not clearly for heavier ones⁵. The 2022 Cochrane review pooled two studies of soft against hard midsoles (1,095 runners) and found it may make little or no difference (low certainty)¹.

Heel-to-toe drop has no clear winner. In a 2016 trial of 553 leisure-time runners given shoes with a 10, 6 or 0 mm drop and followed for 6 months, overall injury risk did not differ. By running history, low-drop shoes went with about half the injury risk in occasional runners (under 6 months of weekly running in the previous year) and about 1.7 times the risk in regular runners⁶. If you run regularly, do not move to a lower drop expecting fewer injuries.

Minimalist shoes are neither safer nor riskier on their own. The 2022 Cochrane review found neutral or cushioned shoes may make little or no difference to the number of injured runners compared with minimalist shoes (5 studies, 766 runners, low certainty)¹. The risk is in the switch, covered below.

One pair or two

A second pair is reasonable, and its benefit is not settled. In a 2015 study of 264 recreational runners followed for 22 weeks (87 injured), using more than one pair at the same time went with an injury rate about 40% lower. The study followed runners and did not assign shoes, and its authors suggested several pairs may vary the load on the musculoskeletal system⁷.

Breaking in a new pair

Moving to a very different shoe changes how the foot and lower leg are loaded, and the clearest evidence is for minimalist shoes. In a 2013 study of 36 experienced recreational runners, 19 moved gradually to minimalist shoes over 10 weeks and 17 kept their usual shoes. More of the minimalist group had an increase in foot bone marrow edema on MRI (10 of 19, a clear difference from the control group)⁸. A review and meta-analysis of 12 studies (7 pooled, 313 participants, transitions of at least 4 weeks) found the move to minimalist shoes was associated with higher injury risk, most often bone marrow edema in the metatarsal region, Achilles tendinopathy and calf pain. Its authors suggested supervised retraining of at least 12 weeks may help⁹.

So a gradual start is not proof of safety: the 10-week transition still produced edema in about half the group. For any new pair, and above all a very different one such as minimalist, a lower drop or a plated racing shoe, do a few short easy runs in it first, add more over several weeks, and keep the old pair for the rest of your runs. Wear a racing shoe on some race-pace runs before the race so it is not new on race day.

When to replace

Shoe makers and stores commonly say to replace a pair every 300 to 500 miles (about 480 to 800 km), a range no trial set. In a 2017 study of 15 recreational runners with a rearfoot strike who ran 640 km in one new pair each, heel plantar pressure and vertical force fell significantly, and heel cushioning was 16% to 33% lower after 480 km. None of the 15 reported a significant drop in heel cushioning on a comfort questionnaire, even after 640 km¹⁰. So do not wait for the shoe to feel dead. Keep a note of the mileage on each pair and replace it by the number.