Athletic Footwear: Choosing Shoes for Running, Lifting, and Training

Walk into any running store and you’ll get a pronation analysis, a stability category, and a confident shoe recommendation. The science underneath is shakier than most retailers admit. The 2014 US military trial — the largest prescriptive-shoe study ever run — found that matching shoes to arch height did not reduce injuries Knapik 2014. The 2009 BJSM systematic review by Richards reached the same conclusion: there is no published evidence to support shoe prescription based on pronation Richards 2009. This article walks through what the literature actually says — and what to do about it.

Shoe categories — what they actually mean

Most running shoes fall into one of four buckets: neutral (a flexible foam midsole, no medial post), stability (a firmer medial post or guide-rails to resist pronation), motion-control (heavy stability shoes for severe overpronators), and minimalist (low stack height, low drop, flexible). The categories are real engineering distinctions; the idea that runners must be matched to one based on arch or gait is what the evidence does not support Richards 2009.

Lifting and gym footwear are simpler:

• Flat-soled shoes (Converse Chuck Taylors, Vans, dedicated deadlift shoes): hard, incompressible sole; close to barefoot. Ideal for deadlifts, rows, presses, and most general gym work where ground feel and stability matter more than cushioning.
• Olympic weightlifting shoes (Adidas Adipower, Nike Romaleos): rigid wooden or TPU heel of 15–22 mm, non-compressible sole. Improve squat depth in lifters with limited ankle dorsiflexion and shift load slightly forward onto the quadriceps.
• Cross-trainers: a hybrid — some cushioning for short runs and box jumps, but firmer and flatter than running shoes for lateral stability.
• Hyrox / multi-modal shoes: cross-trainer-leaning, with enough cushioning for a cumulative ~8 km of running between stations and enough lateral stability for sled work, lunges, and burpees.

The pronation classification myth

The orthodoxy — overpronators need stability shoes, neutral runners get neutral shoes, supinators get cushioning — was never built on randomised evidence. It came from biomechanical theory in the 1980s and was widely adopted by retailers. When it was finally tested, it failed.

• Knapik 2014 reviewed three large US military trials totalling over 7,000 recruits randomised either to shoes prescribed by foot-arch height or to a single neutral shoe. Prescribing shoes by arch height did not reduce injuries in any branch Knapik 2014.
• Malisoux 2016, a Luxembourg RCT in BJSM, randomised 372 leisure runners to either standard shoes or motion-control shoes irrespective of foot type, and followed them for 5 months. Motion-control shoes did not reduce injury; if anything, the standard-shoe group had lower injury rates Malisoux 2016.
• Richards 2009 in BJSM screened the literature and concluded: “The prescription of this shoe type to distance runners is not evidence-based.” Fifteen years later, that statement still holds Richards 2009.

What replaces pronation matching? The Nigg 2015 framework in BJSM — the “preferred movement path” and “comfort filter” paradigms — is the best current synthesis. Runners self-select shoes that allow their joints to follow their habitual movement pattern; shoes that fight that pattern increase muscle work and injury risk. The simplest practical rule that emerges: try several shoes, run in them, pick the one that feels most comfortable Nigg 2015.

“The currently dominant paradigms of impact-force reduction and pronation control should be replaced by two new paradigms based on a preferred movement path and a comfort filter.” — per Nigg 2015, British Journal of Sports Medicine

The minimalist / barefoot debate

Daniel Lieberman’s 2010 Nature paper is the foundational modern barefoot-running citation. Comparing habitually barefoot Kenyan and US runners with shod Western runners, his lab showed that barefoot runners predominantly land on the forefoot or midfoot, generating much smaller impact transients than shod heel-strikers Lieberman 2010. The thick foam heel of a modern shoe partly encourages a heel strike that the unshod foot does not naturally produce.

The follow-up evidence is more nuanced. Hatala 2013 studied the Daasanach of northern Kenya — another habitually barefoot population — and found that on hard surfaces at slower running speeds, most Daasanach actually heel-struck Hatala 2013. So “barefoot equals forefoot” is too simple. Strike pattern is influenced by speed, surface, and individual habit, not solely by footwear.

What about injury rates? Daoud 2012 retrospectively examined a Harvard varsity cross-country team and found habitual rearfoot-strikers had ~twice the rate of mild and moderate repetitive stress injuries as forefoot-strikers, even though both groups wore conventional cushioned shoes Daoud 2012. Davis 2017 argued the lower vertical loading rates in forefoot-striking minimalist runners may explain reduced stress-fracture and tibial-pain rates Davis 2017.

However — and this is the critical caveat — Ryan 2014 in BJSM followed runners transitioning to minimalist shoes and reported significantly higher rates of shin and calf pain in the transition group. The Achilles tendon, soleus, and gastrocnemius take dramatically more load in a forefoot-striking gait, and these tissues need months to adapt Ryan 2014. The rule that emerges:

• If you’re a long-time heel-striker in cushioned shoes and pain-free: you do not need to switch.
• If you have recurrent knee pain, IT-band issues, or runner’s knee: a gradual transition to a lower-drop shoe and slightly more midfoot strike is plausibly helpful (and reduces patellofemoral loading Kulmala 2013).
• If you transition: do it over 3–6 months, with calf and Achilles strength work built in.

Cushioning level and injury risk

One of the more interesting recent findings is that more cushioning is not obviously safer. Theisen 2014 randomised 247 leisure runners to either soft- or hard-midsole versions of the same shoe and tracked them for 5 months. The two groups had no meaningful difference in injury rate — the softer shoe was not protective Theisen 2014. Malisoux’s lab has reported similar non-effects across multiple later trials Malisoux 2016.

Two things this does not mean. It does not mean cushioning is bad — soft shoes feel better to many runners and the “comfort filter” suggests that matters. And it does not address modern carbon-plated super-shoes, where Fuller 2017 and later work show measurable performance gains in running economy Fuller 2017.

Foot strike: heel, midfoot, forefoot

The three patterns:

• Heel (rearfoot) strike: most common pattern in shod recreational runners. Generates a sharp early impact transient, especially at slower speeds. Higher knee/patellofemoral loading. Lower calf/Achilles loading.
• Midfoot strike: foot lands roughly flat. Often a natural compromise; common in middle-distance runners.
• Forefoot strike: ball of foot lands first; heel may not touch the ground. Common in sprinters, barefoot runners, and many elite distance runners. Lower knee loading Kulmala 2013; higher Achilles and calf loading.

Kulmala 2013 measured ground reaction forces in 19 forefoot-strikers and 19 rearfoot-strikers and found that forefoot-strikers had ~7× lower patellofemoral joint loading at the same pace — a striking finding that helps explain why anterior knee pain is more common in cushioned-shoe heel-strikers Kulmala 2013. The trade-off is just shifted loading to the calf and Achilles.

Should you change your strike? Probably not deliberately if you’re injury-free. If you have chronic knee pain, an experienced coach can help you experiment with cadence — raising step rate ~10% reduces knee and hip loading and often nudges the strike forward without conscious retraining Heiderscheit 2011 — and with lower-drop shoes, but allow weeks per change.

Lifting shoes

Squats

An Olympic weightlifting shoe with a 15–22 mm rigid heel is genuinely useful for the back squat and front squat in the following situations:

• You’re tall (longer femurs) and struggle to keep an upright torso.
• You have limited ankle dorsiflexion (can’t get knees over toes without your heel lifting).
• You compete in Olympic lifting and need depth on snatches and cleans.

The raised heel reduces the dorsiflexion required, allowing a deeper squat with a more upright torso, and the rigid sole transmits force without energy loss. It is not necessary — plenty of strong lifters squat heavy in flat shoes or barefoot.

Deadlifts and pulling

For deadlifts, a flat, thin sole — or socks where allowed — is the standard. Reasons:

• Lower bar travel distance (every centimetre of sole = a centimetre more pull).
• Better ground feel and stability through the whole foot.
• No compression in the heel that could shift the bar path.

Cushioned running shoes are the worst option — foam compresses unevenly under load. Chucks, Vans, or dedicated slippers (Sabo, Notorious Lift) all work.

General gym training

For mixed sessions — squats, presses, rows, accessory work — a flat, firm cross-trainer covers most needs. Save Olympic shoes for heavy squat day and slippers for deadlift day; one decent cross-trainer handles the rest.

Hyrox and multi-modal shoes

Hyrox is unusual: 8 km of running broken by 8 functional stations (sleds, burpee broad jumps, wall balls, lunges, rower, ski erg, farmer carry). The shoe needs cushioning for cumulative ~50 minutes of running, lateral stability for sled work, and grip on indoor flooring.

What works well:

• Hybrid cross-trainers with extra cushioning (Nike Metcon-style with foam underfoot, or dedicated Hyrox shoes from Nike, Puma, and Reebok).
• Avoid maximal road running shoes — the soft midsole is unstable for sled work and lateral lunges.
• Avoid pure lifting shoes — insufficient cushioning for the running portions.

When to replace shoes

The traditional “500 km / 300 mile” rule oversimplifies. The midsole foam compresses with use; once compressed it provides less cushioning regardless of external appearance. Modern foams (PEBA, supercritical EVA) appear to last longer than older EVA in lab testing, though field data is limited Fuller 2017.

Practical replacement signals:

• 500–800 km for most cushioned road shoes; the upper end for heavier runners and softer foams.
• Visible compression of the midsole (look at the side — wrinkles, asymmetric collapse).
• Wear-through of the outsole rubber exposing midsole foam.
• New aches that resolve in a fresh pair — often the most reliable signal.
• Time: even unworn, EVA midsoles oxidise and stiffen. Don’t race in the pristine pair you’ve had in the closet for three years.

Lifting shoes have a much longer service life — the rigid heel and minimal foam don’t degrade the same way. A pair of Olympic shoes can last a decade of regular gym use; deadlift slippers similarly.

Wear pattern reading — with caveats

Retailers read sole wear to infer pronation; the evidence base mirrors the pronation literature — weak. A wear pattern tells you where you’ve been loading the shoe, not what shoe to buy next. Use it as a lifespan sanity check, not a prescription.

Practical guidance

1. Try shoes on and run in them. Comfort is the best-evidenced selection criterion Nigg 2015. Most good running stores will let you treadmill-test.
2. Don’t take a pronation diagnosis as gospel. If a stability shoe feels wrong, a neutral shoe is unlikely to harm you.
3. Rotate two pairs if you run more than 3×/week. Different shoes load slightly differently and the foams recover between runs.
4. Build a strike change slowly. Months, not weeks. Calf and Achilles strength work in parallel.
5. Use the right tool for lifting. Flat for deadlifts; flat or Olympic for squats; lifting shoes are meaningfully different from running shoes.
6. Replace based on midsole condition, not the calendar alone.
7. Be sceptical of marketing. Carbon-plated super-shoes have real but modest benefits in races; pronation-control claims have not held up in trials.

Beachside note

If you’re training Hyrox at Beachside or doing the strength + cardio classes, a single decent flat-soled cross-trainer with modest cushioning will cover ~95% of what you need. See our gym essentials piece for the rest of the kit list, and the strength training primer for what to actually do in those shoes.

The bottom line

• Pronation matching is not evidence-based — pick on comfort, not arch Knapik 2014 Richards 2009 Malisoux 2016.
• Habitually barefoot runners forefoot-strike on hard surfaces at speed, but heel-strike too at slower paces Lieberman 2010 Hatala 2013.
• Forefoot strikers had ~half the repetitive injury rate of heel-strikers in a Harvard group Daoud 2012; loading distribution is the likely mechanism Kulmala 2013.
• Cushioning level does not strongly predict injury in healthy runners Theisen 2014.
• Transition to minimalist slowly — calf and shin injuries spike in fast transitioners Ryan 2014 Davis 2017.
• Lifting tools are different from running shoes: flat for deadlifts; Olympic shoes for limited-mobility squatters.
• Replace at 500–800 km based on midsole compression; lifting shoes last far longer.

References