Your legs feel fine at mile 3 and fall apart at mile 18. A 2025 trial suggests strength work can fix that: 10 weeks of lifting and jumping improved running economy after 90 minutes of hard running by 2.1%, while runners who only ran got 0.6% worse.
Running economy is how much oxygen you burn to hold a given pace. In short: it’s your fuel mileage. Most studies measure it fresh. This one measured it tired, which is when races are won and lost.
What the strength training and running economy trial tested
Researchers led by Zanini, Folland, Wu and Blagrove split 28 well-trained male runners into two groups of 14. One group kept running. The other added strength and plyometric work twice a week for 10 weeks. Plyometrics are jumping drills like hops and bounds.
Then everyone ran 90 minutes at heavy-intensity effort, and the team checked how much oxygen they used at the end. After the run, they also tested how long each runner could go at high intensity before quitting.
| Measure | Strength group | Running-only group |
|---|---|---|
| Running economy after 90 min | -2.1% (better) | +0.6% (worse) |
| High-intensity time to exhaustion after the run | +35% | -8% |
The economy gap was statistically real (p = 0.04). The time-to-exhaustion gap was larger (p = 0.004, a large effect). Translation: after the same long, hard run, the lifters had far more left in the tank.
Think of your legs as a car at the end of a long trip. Fresh, every car is efficient. After 200 miles, the worn engine starts guzzling. Strength work seems to keep the engine from guzzling.
Fatigue costs you. Strength may cut the bill.
Fresh economy versus fatigued economy
Most older studies test economy when runners are rested. Those results are still useful. A 2018 review in Sports Medicine pulled together 24 studies. Economy was measured in 20 of them, and it generally improved by roughly 2-8% against control groups, though not every trial found a gain. Time-trial performance and sprint speed improved too. VO2max, your ceiling for oxygen use, did not change.
In short: lifting doesn’t raise your engine’s size. It makes the engine waste less fuel.
A 2016 meta-analysis of highly trained runners (VO2max above 60 mL/kg/min) found a large economy effect, SMD -1.42. That’s a statistical size measure. Translation: even fit runners got a real benefit, not just beginners with room to improve.
The 2025 trial adds a new angle. It asks about durability, meaning how well your economy holds as you tire. That matters because a marathon isn’t run fresh. Whether the trial’s fresh-state economy changed, I couldn’t verify from the abstract. Its headline is the late-run result.
Why might durability improve? The likely ideas are stiffer, stronger muscles and tendons, and less wasted motion when form fades. Those are plausible explanations from the wider literature. This trial’s abstract doesn’t test them, so treat them as ideas, not findings.
Heavy lifting vs plyometrics: what the evidence can say
Here’s the trap. The 2025 trial combined strength and plyometrics. It can’t tell you which part did the work, so don’t read it as proof that heavy lifting beats jumping.
The best indirect clue comes from a 2024 meta-analysis of 38 studies and 894 runners. It measured economy in fresh runners, by training method. A more negative number means a lower oxygen cost.
High-load work and combined programs reached statistical significance. Plyometrics alone did not. Translation: jumping by itself looked weakest for economy, and a mix looked strongest.
That fits the 2025 design, which used a mix. It still isn’t a head-to-head test. One more catch: these numbers come from rested runners, not tired ones.
The same 2024 review also found that high-load work improved time-trial performance and time to exhaustion. That lines up with the +35% time-to-exhaustion result in the 2025 trial. It’s a consistent story, not a proven one.
Case study: Dana’s late-run drop
Take a runner I’ll call Dana. She’s 41, trains for a fall marathon, and runs 45 miles a week. She’s not in the trial. Her numbers are an example.
In week 1, her heart rate at goal pace in miles 14 to 16 of a long run averaged 168. Same pace, early miles: 154. She added two 40-minute strength sessions a week with heavy squats and step-ups, plus short hops, and cut one easy run to make room.
Ten weeks later, the late-run number was 162 at the same pace. Her early miles hadn’t changed much. The late-run gap shrank from 14 beats to 8.
The shape of that change matters more than the exact numbers. The gap between early and late is where durability shows up.
How to program strength without wrecking your runs
The trial used two sessions a week. Reviews on trained runners suggest 2-3 sessions for 8-12 weeks. I can’t give you the trial’s exact loads or exercises, because the abstract doesn’t list them. A coach or the full paper can.
Here is a sample layout built from those general review ranges. It’s practical guidance, not the trial’s protocol.
| Weeks | Sessions per week | Focus | Note |
|---|---|---|---|
| 1-3 | 2 | Learn the lifts, moderate loads | Stop sets well short of failure |
| 4-8 | 2 | Heavier loads, fewer reps, add short hops | Biggest block of the plan |
| 9-10 | 1-2 | Keep heavy work, cut volume | Protect your hardest runs |
| Race week | 0-1 | Light activation only | No new exercises |
A few rules hold up on top of that:
- Keep lifting off the day before your hardest run.
- Pair strength with a hard day, so easy days stay easy.
- Drop one easy mile run rather than a key long run when time is tight.
- Start heavy work 12 or more weeks before race day, not in the last 3.
Long runs build the aerobic engine. If you want the base case for it, see how Zone 2 builds the aerobic base. Strength supports that work. It doesn’t replace it.
How to measure it yourself
You don’t need a lab. Run your long run at a steady pace. Compare heart rate in the first hour to heart rate in the last hour at that same pace.
That gap is related to what we call aerobic decoupling. A smaller gap over time suggests your late-run efficiency is improving. Heart rate is a rough stand-in for oxygen cost, so heat, sleep and fueling can muddy it. Compare similar days.
Fuel and heat lie. Repeat the test under similar conditions.
AthleteOS can help here. Import your long runs with workout import, then use session analysis to compare late-run heart rate at the same pace across your strength block. Your AI coach then adjusts the next week. You can start a trial here. For the math behind weekly load, see how training load is calculated.
What this means for marathoners, ultra runners and triathletes
Marathoners spend hours in the exact state this trial tested: tired, steady, and losing form. If the result carries over, lifting during the build phase is a cheap way to protect your last 10K.
Ultra runners go longer still. A 90-minute test says little about 8 hours, so don’t stretch the numbers. The direction may hold. The size of the effect won’t.
Triathletes run on tired legs by design. The same logic likely applies to the run off the bike, but this trial didn’t test it. Strength time also competes with swim and bike hours, so keep sessions short.
Limits you should know about
- Small sample. Twenty-eight runners is a modest group.
- Men only. The result may carry over to women. Nobody has shown it here.
- 90 minutes isn’t a marathon. The protocol used heavy-intensity running, not 26.2 miles or an ultra.
- Combined program. It can’t rank lifting against plyometrics.
- Short follow-up. It tracked 10 weeks, not a full season.
The result is promising, not settled.
The direction is clear. The details are not.