Running Is a Rotational Movement: What Your Spine, Pelvis, and Arms Actually Do

You don't run like a piston.

Your right leg comes forward, your left arm comes with it, your pelvis turns, and your rib cage answers back. Restrict one part and the others may move differently.

From the front, good distance running can look almost straight. Under motion capture, it isn't. The pelvis and trunk rotate around the body's vertical axis while the arms swing in opposition and help manage angular momentum. That axis runs roughly from the top of your head down toward the ground, so the rotation is easiest to picture from above.

That doesn't mean you should twist harder.

The internet loves taking a normal movement, giving it a dramatic name, and selling the missing drill. The real mechanics are messier, and a lot more useful. Natural rotation helps the body manage the motion of the legs. But the studies below don't show that deliberately adding rotation improves speed, economy, or injury risk. There's no universal shoulder angle, arm path, or magic amount of pelvic rotation that every runner should copy.

So let's separate what your body clearly does from what somebody hopes to sell you because they own a rope and a slow-motion camera.

Your legs create a rotation problem

Each leg is a heavy pendulum. When the right leg swings forward, it contributes angular momentum around your vertical axis. The left leg does the same thing half a stride later.

Why should the opposite arm help? In a three-dimensional study of 10 male recreational runners, the arms and upper trunk produced angular momentum opposite to the legs around the vertical axis, leaving relatively little angular momentum across the whole body (Hinrichs, 1987).

What comes around is all around.

Calling the arms "counterweights" is useful shorthand, although they aren't dead weights hanging from the shoulders. Their movement is coupled to what happens at the trunk and legs, and the exact path differs from one runner to another. The point is simpler: your arms are part of the gait cycle. They aren't decoration, and they aren't little pistons driving you forward one fist at a time.

And a runner can rotate without corkscrewing down the road. The pelvis may turn one way while the upper trunk and opposite arm answer it. Individual pieces move, but their angular momentum can largely oppose one another across the whole body.

Your arms are doing more than keeping busy

What happens if you take the arms out of the pattern? Christopher Arellano and Rodger Kram tested that question directly. Thirteen healthy recreational or competitive runners completed seven-minute treadmill trials at 3 m/s with normal arm swing, with their hands behind their backs, with their arms across their chests, and with their hands on their heads.

Compared with normal arm swing, net metabolic power increased by about 3% with the hands behind the back, 9% with the arms across the chest, and 13% with the hands on the head. Pelvic rotation also rose by 63%, 102%, and 101% in those three restricted conditions (Arellano and Kram, 2014).

Those are big changes, but read the setup again. The runners had their hands behind their backs, across their chests, or on top of their heads. Nobody naturally runs a marathon like they're waiting to be searched at airport security.

The study shows that net metabolic power was lower with normal arm swing than in each artificial restriction. Pelvic rotation increased in all three restricted conditions, while shoulder rotation increased significantly in two of the three, changes consistent with compensation when the arms couldn't swing normally. It doesn't show that moving your elbows from 84 degrees to 79 degrees, driving your hands harder, or copying an Olympian's arm carriage will shave 3% from your energy cost.

My boring takeaway is to let the arms swing naturally. Moore's 2016 review found no established economy benefit from deliberately changing a runner's arm mechanics (Moore, 2016).

Relaxed is a better starting cue than perfect.

Your pelvis, low back, and chest don't move as one block

The torso is often coached like a single box: keep it tall, keep it quiet, don't rotate. Real bodies are messier.

In one treadmill study, 20 healthy male runners who logged more than 20 km per week ran at 4 m/s while researchers tracked their lumbar spine and pelvis in three dimensions. Both segments showed complex three-dimensional movement, but axial rotation of the lumbar spine and pelvis correlated poorly (r=0.37) and was separated by a phase difference equal to 21% of the running cycle (Schache et al., 2002).

The pelvis and low back weren't twisting like one solid Lego piece; they moved on different timing.

And when researchers compared 14 elite endurance runners with 14 recreational runners over ground at speeds from 3.3 to 5.6 m/s, movement increased as both groups ran faster. The largest increases appeared in the transverse plane. The recreational runners leaned their trunks farther forward as speed rose, while the elite runners kept a more consistent trunk inclination (Preece, Mason, and Bramah, 2016).

Useful finding, small study.

Did the steadier trunk position help make them elite? The study can't tell us whether it did, whether years of high-level training produced that pattern, or whether anatomy and pace history explained part of the difference. Copying the visible position without copying the athlete's body, speed, strength, and training is how running-form advice gets sideways fast.

Women may look different and still be economical

Different bodies don't have to solve this problem the same way.

A 2023 study measured shoulder and pelvic rotation in 86 junior elite middle- and long-distance runners, including 37 female and 49 male athletes, during an incremental treadmill test. Shoulder and pelvic rotation increased with speed. Female runners showed more shoulder rotation than male runners, while pelvic rotation did not differ by sex (Lang et al., 2023).

The relationship with energy cost was small and went in the opposite direction from the easy assumption: more upper-body rotation was weakly associated with a lower cost in parts of the analysis. The authors suggested that differences in upper- and lower-body mass distribution may help explain the sex-specific shoulder motion.

Does that make extra rotation a cue for women? I wouldn't turn it into one. The group consisted of trained junior athletes, the study was observational, and a correlation doesn't show that deliberately adding rotation would improve anybody's economy. But it's a solid reason to stop treating one narrow-looking gait as the correct template for every body.

Body proportions probably matter too because limb length and mass distribution change moment of inertia. But the Lang study didn't test arm length, shoulder width, or pelvic shape as causes; its mass-distribution explanation remains a hypothesis. Two economical runners can still show visibly different arm and trunk motion.

One textbook gait can't fit all of those bodies.

The spinal engine claim gets ahead of the evidence

The spinal engine idea traces to a 1985 hypothesis by Serge Gracovetsky. He argued that the lumbar spine plays an active role in locomotion and helps drive the pelvis rather than simply riding above leg-powered movement (Gracovetsky, 1985).

I think Gracovetsky was right to reject the idea of a dead, passive trunk. Running is a whole-body action, and modern motion-capture studies clearly show movement through the pelvis, lumbar spine, and thorax. Arm-restriction experiments also show that changing one part can change motion elsewhere.

But the strongest version of the claim goes much farther. Does seeing rotation prove the spine is the primary engine of distance running? No, and it definitely doesn't prove that deliberately increasing spinal rotation creates free propulsion.

Here's the line I wouldn't let a coach or product pitch blur:

• Supported: Your pelvis, spine, shoulders, and arms rotate during running.
• Supported: Arm swing helps counter the angular momentum created by the legs.
• Plausible but individual: Some runners may benefit from addressing an obvious restriction or a coached movement problem.
• Not established: Every runner should create more spinal rotation.
• Not established: A rotational drill will improve running economy, speed, or injury risk merely because it resembles one part of gait.

Running economy means the oxygen or energy required to hold a given submaximal speed. A 2024 systematic review and meta-analysis covered 51 observational studies with 1,115 participants and found that individual biomechanical variables explained about 4-12% of the between-runner variation in economy when considered alone (Van Hooren et al., 2024).

That review didn't test rotation training, but it makes me skeptical of any form overhaul built around one shoulder cue.

A smooth stride can't fix an underfueled race. Set carb, fluid, and sodium targets for your race in about 60 seconds with the free EnduranceOS planner.

Here's what I'd actually tell most runners

Should most runners change anything? My answer is no: leave a natural-looking, pain-free arm swing alone unless you have a specific reason to change it.

Your hands don't need to stay on invisible railroad tracks. A slight crossover appeared in all 13 runners in the Arellano experiment, so seeing some crossover isn't automatically evidence of a problem. Your elbows don't need to hold one sacred angle from warm-up through the final mile either.

Two short phone clips are enough for a rough look:

1. Film 10-15 seconds from behind and from the side at easy pace.
2. Repeat during a controlled faster stride, because rotation changes with speed.
3. Look for the broad pattern rather than measuring tiny angles from a phone video: do both arms move, does the trunk look comfortable, and does one side change sharply as pace rises?
4. Compare the video with how the run felt. A weird-looking frame without pain, lost speed, or a repeatable limitation may just be a weird-looking frame.

Can a phone video tell you why something hurts? No. Treat it as a rough comparison, never a diagnosis. Camera angle, clothing, fatigue, treadmill position, and the exact stride you paused can all make normal movement look suspicious. If an asymmetry arrived with pain, weakness, numbness, repeated tripping, or a clear drop in function, that's a better reason to seek a qualified sports clinician than an arbitrary line drawn through your shoulders.

Want to feel what the arms contribute? Take a few slow steps somewhere clear with the swing briefly restricted, then let your arms move again and notice what your trunk does. Treat that as an observation, not a test or training protocol.

For training, I'd spend the time on consistent running, strides or hills when they fit, and strength work your schedule can absorb. Rows, carries, split squats, calf work, and controlled trunk exercises are all reasonable, but none gets special running powers from the word rotational. If you're adding strength because a tendon already hurts, the isometric tendon guide explains what those holds can and can't do.

And don't overhaul your gait during the same week you add mileage, hills, and a new strength plan. Changing five loads at once makes it nearly impossible to tell which one helped and which one left you limping down the stairs.

Is rope flow worth doing?

So where does rope flow fit? It uses a rope swung around the body in looping patterns, usually without the repeated jumping of jump rope. It combines weight shifts, opposite-side arm action, trunk movement, and timing. Until that timing clicks, the rope may smack your shins a few times.

The sales pitch writes itself: running rotates, rope flow rotates, so rope flow must improve running. That last step is the problem.

I couldn't find a peer-reviewed human trial showing that rope flow improves running economy, race performance, running-related injury risk, or gait mechanics. Research on jump rope doesn't fill that gap because jump rope is a plyometric exercise built around repeated ground contacts. Rope flow is a different activity.

But if you enjoy it, five or ten minutes in a warm-up is perfectly reasonable. Enjoyment counts when it helps a habit stick.

So don't sell it as a running intervention. Rope flow hasn't been shown to improve running, and it shouldn't replace running, strength work, sleep, or rehab with a clearer job.

Don't manufacture a twist

If your gait feels smooth, your arms move freely, and you're training without a repeatable problem, you probably don't need to manufacture more twist. If something hurts or one side stops doing its job, investigate that problem instead of buying a universal rotation story.

Quick science note: this article explains biomechanics research for educational purposes. It isn't a gait assessment, injury diagnosis, or personalized training plan. Running pain, sudden changes in coordination, weakness, numbness, or loss of function should be assessed by a qualified professional.

A clean-looking stride won't save a race you underfuel. The EnduranceOS planner builds carb, fluid, and sodium targets around your event, body, pace, and conditions in about 60 seconds, and it's free.