Memory Foam vs Spring Mattress for Back Pain: Which Wins

Two people can buy the same mattress. One wakes up feeling repaired. The other wakes up worse and blames the material. The thing doing most of the work in that split isn’t printed anywhere on the label, and it isn’t the difference between foam and steel.

It’s firmness, plus how well that firmness holds up under a specific body in a specific position at 4am. Both categories can land in the right zone. Both routinely miss it. What follows sorts out where each construction actually helps a sore lower back, where the marketing runs ahead of the evidence, and why one of the two quietly rewrites its own specification once you’ve been lying on it for twenty minutes.

What a sore back is actually asking for at 3am

A painful lower back wants two things from a sleep surface, and they pull in opposite directions.

The first is spinal alignment. Your pelvis is the heaviest part of you, and if the surface lets it drop, the lumbar curve flattens and the soft tissue on either side of the spine spends eight hours under load it never agreed to. The second is pressure relief. The shoulder and the hip are narrow contact points, and a surface that won’t accommodate them creates the ache that wakes you up and makes you roll over.

More give solves the second problem and worsens the first. Less give does the reverse. Every mattress on sale is a proposed compromise between those two demands, and the label on the box tells you almost nothing about where that compromise lands for your body.

Reviews of controlled mattress trials keep landing in the same region of that trade-off. Medium-firm surfaces come out better for pain and sleep quality than firm ones in adults with nonspecific low back pain, though the trials are small, mixed in method, and often rely on subjective firmness labels rather than measured stiffness [1][2]. The direction of the finding is fairly stable. The precision isn’t.

Memory foam vs spring mattress for back pain: firmness beats material

Here’s the part the category war hides. Both constructions can be built medium-firm, and both get sold soft, firm, and everything in between. Choosing memory foam over springs doesn’t choose a firmness for you.

That matters, because the same mattress isn’t the same mattress to two different people. A surface labelled medium-firm behaves like a firm one under a lighter body that never compresses the comfort layers, and like a soft one under a heavier body that sinks straight through to the core. Sleeping position moves the number again. Back sleeping spreads load across a wide area, while side sleeping concentrates it at two points, so the same stiffness produces very different sink at the hip. Reviews of the mattress literature keep flagging body size and preferred sleep position as things you have to weigh alongside any firmness recommendation, and the trials that let people adjust their own surface tend to produce the strongest results [1][2].

So the honest version of the shopping question isn’t foam or springs. It’s this: what does medium-firm mean for a body this size, sleeping in this position, and which construction gets there more reliably.

The mattress that rewrites its own spec overnight

Viscoelastic foam has a glass transition sitting close to room temperature, and that’s the whole trick. Below it, the polymer chains are locked and the foam is stiff. Above it, they move, and the foam yields.

Lab testing of viscoelastic polyurethane foams across the temperature range they actually get used in shows how sharp this is. Measured hardness ran around 2 to 3 kPa at 20°C, falling to roughly 1.5 to 2 kPa at 40°C, with the glass transition landing between 10 and 20°C [4]. Body heat pushes the contact zone toward the upper end of that range within minutes.

Here’s what that means on a real night. The surface that felt supportive in the showroom is a measurably softer surface once your torso has warmed it for half an hour. For a side sleeper, that softening is the mechanism doing the useful work, letting the shoulder in. For someone whose pain comes from a pelvis that sinks too far, it’s the mechanism doing the damage, and it arrives after you’ve already fallen asleep. Cold bedrooms flip the problem: the same foam goes noticeably firmer.

Steel has no equivalent behaviour. A coil at 15°C and a coil at 30°C respond the same way, which is a boring property that turns out to be worth quite a lot when the goal is a spine sitting in one predictable position all night.

Not all springs are the springs you are picturing

The mental image most people carry of an innerspring bed comes from the wired-together Bonnell unit: coils helically laced into a single grid, so compressing one recruits its neighbours. That construction is where the reputation for bounce, noise, and feeling every movement across the bed comes from.

Pocketed coils changed the mechanics. Each spring sits in its own fabric sleeve and compresses on its own, so the support becomes local rather than shared. Two things follow. Load under the hip doesn’t automatically lift the shoulder, which is what makes zoned lumbar support possible in a coil bed. And movement stays closer to where it started instead of travelling across the surface, though a thick bridging foam layer on top will undo some of that.

How much this matters in a sleep lab rather than a showroom is less settled than the retail copy suggests. Direct comparisons of coil architectures on measured partner disturbance are thin. What has been tested is that the structure of a spring mattress core and its bedding layers has a measurable effect on sleep behaviour and movement during the night, rather than being cosmetic [5].

The practical upshot for a bad back is unglamorous. If a spring bed is on your shortlist, the coil type is the specification worth reading, and pocketed is the one worth paying for.

Your current mattress is probably confounding the whole question

Before you credit either construction with anything, there’s a much duller explanation for why a new bed helps.

When researchers replaced participants’ own beds (all at least five years old) with new medium-firm bedding systems and tracked daily ratings across 28 days before and 28 days after, back discomfort, sleep quality, comfort, and sleep efficiency all improved significantly [3]. Sleepers on the cheapest old beds reported the most lower back pain. The beds being replaced averaged around nine and a half years old.

This is a before-and-after design with people rating their own sleep and no blinding, so expectation is doing some unknown share of the work. Even so, it complicates the shopping question a lot. A decade-old mattress with a trench where your hips go is failing at spinal alignment in a way no material choice can rescue, and almost anything new will feel like a revelation for a few weeks. That’s a poor basis on which to declare a category winner.

If your bed is old and sagging, the sag is the finding. The foam-versus-springs debate only starts to matter once you’re comparing two intact surfaces.

The verdict, sorted by how you actually sleep

Side sleepers. Memory foam. The shoulder and hip need somewhere to go, and viscoelastic contouring spreads that load better than a coil unit with thin padding. Aim medium rather than medium-firm, because the overnight softening will take it down a step anyway.

Back sleepers. Pocketed coils, medium-firm. The lumbar curve needs a surface that resists rather than one that gradually gives way as it warms. A thin foam comfort layer over the coils covers the pressure side without the sink.

Stomach sleepers. Pocketed coils, medium-firm to firm. This position already extends the lumbar spine, and any surface that lets the pelvis drop makes it worse. Foam is the wrong tool here.

Combination sleepers. Coils again, for a duller reason: memory foam resists position changes, and a bed that makes rolling over an effort produces more waking, not less.

Heavier sleepers. Coils, firmer than the label suggests, since a given firmness rating compresses further under more load [1].

Back pain with a clear medical cause sits outside all of this, and a mattress isn’t the lever there.

Where the whole argument stops mattering

The hybrid mattress exists because the trade-off above has an obvious engineering answer: put a viscoelastic comfort layer over a pocketed coil core and take the contouring from one and the support, airflow, and edge stability from the other.

This is now most of the premium market, and for a lot of people it does dissolve the either/or. The coils handle spinal alignment and stop the pelvis dropping. The foam handles pressure at the shoulder and hip. Air still moves through the core, so heat clears better than it does in an all-foam build.

Two caveats keep hybrids from being a free win. They cost more, often a lot more, and price per year of use only works out if the build quality is real. And a thick, soft foam layer over coils brings back the overnight softening problem while also bridging movement across the springs, which hands back some of what the pocketed construction was bought for.

A hybrid with a modest comfort layer over a good coil unit is the version worth paying for. A hybrid that’s basically a foam bed with token springs underneath is a foam bed.

If you forced me to hand one answer to an adult with a recurring sore lower back and nothing else to go on, it’s a medium-firm pocketed-coil bed, ideally with a thin foam comfort layer. It holds the lumbar curve, it stays the firmness it claims to be at 4am, it sheds heat, and it gives you a firm edge to push off in the morning when your back is at its stiffest.

Dedicated side sleepers are the clear exception, and it’s a big one. If your pain shows up as a hot spot at the shoulder or the hip rather than an ache across the lower back, memory foam earns its place.

Whatever the label says, the trial period is the specification that matters most. Firmness feels different to different bodies, the evidence base is built on modest studies with subjective outcomes, and a fortnight of your own sleep tells you more than any comparison table can.