Is a Rounded Back Really Dangerous When You Deadlift?

For decades, one of the first rules taught when lifting something heavy has been simple:

Keep your back straight.

Whether it is a deadlift in the gym, picking up a box at work or lifting something from the floor at home, a rounded back has traditionally been portrayed as dangerous. The assumption is that spinal flexion under load places excessive stress on the structures of the lower back and substantially increases the risk of injury.

It sounds logical. It is easy to teach. And it has become so deeply embedded in lifting culture that seeing someone’s back round during a deadlift can make people uncomfortable.

But is a rounded spine actually as dangerous as we have been led to believe?

A 2026 narrative review published in Sports, titled Beyond the Neutral Spine: A Narrative Review and Modern Framework for Low Back Injury Prevention in Deadlifting, challenges the idea that maintaining a perfectly neutral spine is an automatic safety mechanism.

Its conclusion isn’t that technique doesn’t matter.

And it certainly isn’t that you should deliberately round your back under the heaviest load you can find.

The message is much more useful than that:

Spinal position is only one part of injury risk.

The Problem With the “Neutral Spine” Rule

The traditional model is straightforward.

Neutral spine = safe.

Rounded spine = dangerous.

But human movement isn’t that simple.

The lumbar spine naturally moves through flexion and extension during everyday activities, sport and resistance training. Even experienced lifters who appear externally to maintain a relatively neutral back may demonstrate some degree of lumbar flexion when measured more accurately.

The 2026 review concluded that moderate lumbar flexion during lifting is not consistently associated with increased injury risk, particularly in trained individuals who have developed tolerance to those positions.

Importantly, the authors found no clear prospective evidence demonstrating that simply deviating from an externally defined neutral spinal position automatically causes injury.

That distinction matters.

Something can increase mechanical stress on a tissue without automatically causing that tissue to become injured.

Human tissue isn’t an inert structure. It adapts.

Muscle adapts to loading.

Tendon adapts to loading.

Bone adapts to loading.

And the tissues of the spine also possess the capacity to tolerate and adapt to appropriately progressed loading.

The question therefore shouldn’t simply be:

“Is the spine flexed?”

A better question is:

“Is this person prepared for the load and position they’re being exposed to?”

A Rounded Back Is Not Automatically an Injured Back

This is where discussions around lifting technique often become unnecessarily frightening.

Someone sees a lifter’s back round slightly and immediately assumes their discs are being damaged.

But posture alone doesn’t tell us whether an injury will occur.

The review proposes a much broader framework incorporating spinal biomechanics, training load, movement variability, dynamic trunk control and psychological factors.

In other words, injury risk is multifactorial.

How much weight are you lifting?

How quickly has that load increased?

How much training volume have you accumulated?

How fatigued are you?

How experienced are you?

Have you previously trained in that position?

Can you control the position?

How well has your body adapted to the demands being placed upon it?

Those questions may tell us considerably more than simply drawing a straight line down someone’s back.

Flexion and Movement Under Load Aren’t Necessarily the Same Thing

There is another important distinction.

A spine that is flexed isn’t necessarily a spine that is moving uncontrollably into flexion.

Imagine an experienced lifter who begins a heavy deadlift with some thoracic or lumbar rounding but maintains a relatively consistent trunk position throughout the lift.

Now compare that with somebody whose position changes substantially as the load leaves the floor because the weight exceeds their current capacity.

Visually, both may have a “rounded back”.

Mechanically and practically, however, they aren’t necessarily doing the same thing.

This is why coaching cannot be reduced to screenshots of someone’s spine.

Technique still matters, but control matters more than achieving an aesthetically perfect posture.

For many lifters, particularly beginners, coaching towards a relatively neutral and stable spinal position remains entirely sensible. It provides a repeatable starting point from which we can teach bracing, force production and load progression.

What we shouldn’t do is turn that coaching cue into a biological law.

Neutral isn’t magic.

And flexion isn’t automatically dangerous.

Load Management May Matter More Than Perfect Posture

One of the most important themes of the review concerns training load.

The authors argue that rapid changes in training exposure and cumulative loading appear to be more consistent predictors of injury risk than isolated deviations from neutral alignment, although they appropriately describe these relationships as probabilistic rather than deterministic.

That fits a much broader principle of strength and conditioning:

Your body needs time to adapt to what you ask it to do.

If someone has spent months progressively deadlifting 80 kg and is well adapted to that workload, performing another appropriately programmed session is very different from suddenly attempting 160 kg.

Likewise, increasing load, volume and frequency simultaneously while training in a fatigued state changes the demands considerably.

The problem isn’t necessarily that one vertebra moved a few degrees away from an arbitrary ideal.

The problem may simply be that the athlete exceeded their current capacity.

That is why sensible programming matters.

Build volume progressively.

Build intensity progressively.

Allow recovery.

Manage fatigue.

Develop technical competency.

Expose people gradually to the positions and loads they need to tolerate.

Capacity has to be earned.

Your Spine Is Designed to Adapt

This is particularly important when discussing back pain and rehabilitation.

People with back pain are frequently told to protect their spine.

Don’t bend.

Don’t twist.

Don’t lift.

Keep your back straight.

Be careful.

Initially, modifying activity can sometimes be entirely appropriate. But permanently teaching someone that spinal movement is dangerous can create another problem: they begin to fear movement itself.

The 2026 review discusses psychological factors as part of its proposed injury-prevention framework. Beliefs about vulnerability can influence movement behaviour and potentially affect how people interact with load.

A rehabilitation programme therefore shouldn’t simply teach someone how to avoid using their back.

Where appropriate, it should progressively rebuild the back’s capacity to tolerate movement and load.

That might initially involve controlled hinging with relatively little spinal movement. As capacity develops, however, appropriate exposure to greater ranges, different positions and increasing loads can form part of developing a more adaptable system.

The objective isn’t to protect somebody from movement forever.

It is to prepare them for movement.

Movement Variability Is Normal

Another interesting element of the review is movement variability.

Humans don’t perform every repetition identically.

And we probably shouldn’t expect them to.

Movement naturally changes slightly depending on load, fatigue, anatomy, experience and the task being performed.

An experienced powerlifter attempting a maximal deadlift isn’t going to move exactly like someone performing an unloaded hip hinge during rehabilitation.

Nor should they.

The authors suggest that adaptable movement strategies may be preferable to rigidly forcing everybody into one supposedly perfect position.

That doesn’t mean anything goes.

There are still positions and strategies that may be more efficient for a particular person or task. Coaches should still improve technique where doing so improves performance, distributes load appropriately or makes an exercise easier to control.

But technical variation isn’t synonymous with injury.

The aim should be competent, adaptable movement rather than robotic adherence to an arbitrary ideal.

Does This Mean Technique Doesn’t Matter?

No.

And this is where social-media interpretations of research often go wrong.

Research questioning the importance of a perfectly neutral spine can quickly become:

“Technique doesn’t matter.”

That’s not what this paper says.

Technique affects how forces are distributed through the body. Different spinal positions alter mechanical demands on different tissues. Load, fatigue and movement strategy still matter.

The authors themselves acknowledge that long-term prospective evidence remains limited.

This was also a narrative review, not a randomised trial proving that rounded-back deadlifting is safe under every circumstance.

So the sensible conclusion isn’t:

Round your back as much as you want.

It’s:

Stop treating neutral spine as an automatic safety shield.

What About Beginners?

With someone new to deadlifting, I will still generally teach a controlled hip hinge, effective bracing and a relatively stable spinal position.

Why?

Because it gives us a simple and repeatable movement strategy from which to develop strength.

But if their lumbar spine isn’t visually perfectly neutral, I’m not going to convince them that they’re damaging their back.

As their experience increases, so can their exposure to different loads and positions.

The same principle applies in rehabilitation.

Start with what the individual can currently tolerate.

Build confidence.

Build strength.

Progress the load.

Expand movement options.

Develop capacity.

That’s very different from giving someone a list of movements they’re forbidden from performing for the rest of their life.

What About Experienced Lifters?

Experienced strength athletes provide perhaps the clearest demonstration that there isn’t one universally correct spinal position.

Watch high-level powerlifters and strongman competitors and you’ll see substantial variation in deadlift technique.

Some maintain relatively neutral lumbar positions.

Others demonstrate visible thoracic or lumbar flexion.

That doesn’t prove that spinal flexion is protective or harmless. Elite athletes are not automatically models of injury prevention.

What it does demonstrate is that humans can develop considerable capacity to produce force in positions that would traditionally have been labelled “incorrect”.

Their bodies have been progressively exposed to those demands over years.

That word — progressively — matters.

Copying the spinal position of an elite lifter without possessing their years of training exposure is not the same thing.

Adaptation is specific.

The Bigger Problem With Fear-Based Coaching

Perhaps the most important consequence of the neutral-spine debate isn’t what happens during a deadlift.

It’s what people come to believe about their backs.

If we repeatedly tell people that bending their spine is dangerous, we risk creating the impression that the back is a fragile structure requiring constant protection.

For someone already experiencing persistent low back pain, that message can be particularly unhelpful.

The spine isn’t indestructible.

But neither is it inherently fragile.

It is a biological structure capable of adaptation.

Good coaching should respect tissue loading without creating unnecessary fear of movement.

There is a substantial difference between saying:

“Let’s improve this position because it will make the lift more efficient and easier to control.”

and:

“If you round your back, you’re going to injure yourself.”

The first develops capability.

The second can create fear.

So, Is It Safe to Deadlift With a Rounded Back?

There isn’t a simple yes-or-no answer.

Moderate spinal flexion isn’t inherently pathological, and current evidence does not support the idea that any deviation from neutral automatically produces injury.

But that doesn’t mean spinal position is irrelevant.

The appropriate position depends on the person, the load, their training history, their current capacity, fatigue, the task being performed and how progressively they have been exposed to it.

This is ultimately the same principle we use throughout strength and rehabilitation.

Load has to match capacity.

A movement you aren’t prepared for can create problems.

A movement you’ve progressively adapted to can become something you’re highly capable of tolerating.

Build a Stronger Back, Not a More Fearful One

The purpose of strength training isn’t to spend your life avoiding physical stress.

It’s to progressively expose the body to manageable stress so that it becomes better equipped to deal with it.

That includes the back.

So rather than obsessing over whether somebody’s spine has deviated a few degrees from neutral, we should be asking better questions.

Is the load appropriate?

Is the progression sensible?

Is the lifter in control?

Are they recovering adequately?

Are we developing strength and movement options over time?

Those are much more useful questions than simply asking whether someone’s back looks perfectly straight.

A neutral spine can be a useful coaching strategy.

It isn’t an automatic safety mechanism.

And a rounded spine isn’t automatically a dangerous one.

Don’t fear spinal flexion. Respect load, build capacity and progressively prepare your back for the demands you want it to tolerate.

Reference

Cherni B, Marzouki H, Selmi O, Al Attar W, Chamari K, Suzuki K. Beyond the Neutral Spine: A Narrative Review and Modern Framework for Low Back Injury Prevention in Deadlifting. Sports. 2026;14(4):151. doi:10.3390/sports14040151.