⚡Key Takeaways

11 min read

The best trial evidence favours medium-firm over firm: in a 313-person randomised trial, people with chronic low-back pain did better on a medium-firm mattress
Materials mostly change heat, motion transfer and how quickly the surface responds — not whether a mattress is “supportive”
Use the trial period properly: bodies take weeks to adjust, so judge a new mattress after about a month, not after one night
“Orthopedic”, “zoned” and most “cooling” claims are marketing language with little independent testing behind them
A topper can change the feel of a sound mattress; it cannot fix one that sags

A mattress is one of the few health purchases most people make only a handful of times in their lives, and one of the few where the shop floor is almost no help. Ten minutes lying on a showroom bed in your coat tells you very little about eight hours a night for eight years. The marketing does not help either: every mattress is “supportive”, “pressure-relieving” and “cooling”, and firmness numbers are not comparable between brands.

The research is thinner than the industry’s confidence suggests, but it does answer the most common question — how firm — and it gives a sensible way to think about the rest. This guide covers what the trials found, what each material actually changes, how to use a trial period, when to replace, and which claims to ignore.

What the evidence says about firmness

For decades the standard advice for a bad back was “sleep on something hard”. It was never tested until 2003, when a Spanish group randomised 313 adults with chronic non-specific low-back pain to a firm or a medium-firm mattress, with neither patients nor assessors told which was which. After 90 days the medium-firm group had less pain in bed, less pain on rising, and less disability.

A Danish trial five years later pointed the same way from a different angle. People with chronic low-back pain who slept for a month on a body-conforming foam mattress or a waterbed did slightly better than those given a hard mattress — though the authors stressed the differences were small and many participants dropped out.

Two later systematic reviews pulled the controlled trials together and reached the same broad conclusion: a mattress people describe as medium-firm is the best-supported choice for comfort, sleep quality and spinal alignment, and adjustable (self-inflated) air surfaces also did well because the sleeper can tune them.

📊How much weight to put on this

The trials are few, mostly short (one to three months), and mostly in people who already had back pain. “Medium-firm” is also a loose category: the Kovacs trial used a standardised European scale, but the numbers printed on retail mattresses are each brand’s own and cannot be compared across brands. Treat medium-firm as the sensible default, not a precise prescription.

Sleeping position and body weight

Beyond “medium-firm”, most advice about matching a mattress to your body comes from mattress-industry convention and basic mechanics rather than trials. It is still reasonable:

Side sleepers put their weight through the shoulder and hip. A surface with a softer top layer lets those points sink in so the spine stays roughly straight.
Back sleepers generally do well on medium-firm, which is where most of the trial evidence sits.
Stomach sleepers tend to need a firmer surface so the hips do not sag and arch the lower back.
Heavier people compress foam further, so the same mattress feels softer to them and wears faster. Thicker comfort layers, denser foams or coil support cores help.
Couples with very different weights or preferences may do better with split-firmness or two single mattresses on one frame than with a compromise that suits neither.

Materials: what each one actually changes

The big differences between materials are not in “support”, which any well-made mattress of the right firmness provides. They are in heat, motion transfer, responsiveness (how quickly the surface springs back when you move), durability and price.

Type	Heat	Motion transfer	Feel	Watch for
Memory foam	Warmest; foam insulates and you sink into it	Lowest	Slow, hugging, pressure-relieving	Feeling “stuck” when turning; low-density foams soften quickly
Polyfoam (all-foam)	Warm	Low	Faster response than memory foam	Density matters for lifespan; cheap foams sag early
Latex	Cooler than memory foam, especially perforated	Low to moderate	Buoyant, springy	Heavy; expensive; rare latex allergy
Innerspring	Coolest; air moves through the coils	Highest with linked coils, lower with pocketed coils	Bouncy, “on top” of the bed	Thin comfort layers wear through; can be noisy
Hybrid (coils plus thick foam or latex)	In between	Low to moderate with pocketed coils	Pressure relief with some bounce	Price; the comfort layer decides most of the feel
Adjustable air	Depends on the top layers	Low if each side is separate	Firmness you can change	Pumps and valves can fail; price

If you share a bed with a restless partner, motion isolation is the property most worth paying for: foam and pocketed coils do this well, linked-coil innersprings do it badly. If you sleep hot, the material and how deeply you sink into it matter more than any “cooling” additive (more on that below).

Why heat matters

Your core temperature has to fall a little for sleep to start and stay deep. Laboratory work on thermal environment shows that when people sleep with normal bedding, heat increases wakefulness and cuts slow-wave and REM sleep. A mattress that you sink deeply into, wrapped in thick foam, reduces the skin area that can shed heat. For hot sleepers that is a real, not a marketing, reason to prefer latex, coils or a firmer foam. Our guide to the sleep environment covers room temperature and bedding.

Certifications worth knowing

Certifications do not tell you whether a mattress is comfortable, but they do tell you something about what is in it:

CertiPUR-US (foams): tested for specified chemicals (certain flame retardants, heavy metals, formaldehyde) and for low emissions of volatile organic compounds.
OEKO-TEX Standard 100 (textiles and components): tested for a list of harmful substances.
GOLS / GOTS: organic latex and organic textiles respectively. These certify how materials were grown and processed, not that the mattress is healthier to sleep on.
Greenguard Gold: low chemical emissions into indoor air.

New foam mattresses often smell for a few days (“off-gassing”). Airing the room usually clears it. If a smell persists for weeks, use the trial period.

Trial periods and break-in

Most online mattress sellers now offer home trials of around 100 nights. They exist because a showroom test is a poor predictor, and they are the most useful consumer protection in this market. Use them deliberately:

Read the return terms before buying. Many trials require you to keep the mattress for a minimum period (often 30 days) before returning, and some charge collection or restocking fees.
Give it three to four weeks. New foam softens a little over the first weeks, and your body adjusts to a new surface. Morning stiffness in the first week is common and not, by itself, a verdict.
Keep a simple log. Note morning pain, how often you woke, and whether you slept hot, for a week on your old bed and then weekly on the new one. Memory of “how it felt” is unreliable after a month.
Decide before the deadline. If you are still waking stiff or hot after a month, return it.

Also check the foundation. A mattress on a sagging base or slats that are too far apart will feel worse than it is, and may void the warranty.

When to replace a mattress

There is no good evidence for a fixed lifespan. “Every seven to eight years” is an industry rule of thumb. What does exist is a small study in which healthy people with mild sleep-related aches switched from their own beds (on average about nine and a half years old) to new medium-firm ones, and reported better sleep quality and less back discomfort over four weeks. It had no control group and people knew they had a new bed, so some of that improvement is expectation. Better signals than age:

A visible dip or body impression where you sleep
You wake stiff or sore and it eases within half an hour of getting up
You sleep noticeably better in hotels or on other beds
Coils you can feel, or foam that has gone lumpy
You or your partner roll towards the middle

Many warranties treat a body impression deeper than a stated limit — often somewhere around 2–4 cm (¾–1½ inches) — as a defect. Check your warranty and measure the dip with a straight edge across the bed before you assume you have to pay for a replacement.

What’s marketing

“Orthopedic”. An unregulated word. It usually just means firm — and firm is the option that lost to medium-firm in the best trial.
“Zoned support”. Firmer under the hips, softer under the shoulders makes mechanical sense. But when we searched PubMed, the only sleep-lab study of a zoned mattress we found was 20 young adults on an adjustable zoned air mattress, co-authored by a mattress-company employee, with mixed results. We found no trial showing zoning reduces back pain compared with a well-chosen unzoned mattress.
“Cooling” foams. Gel beads, copper, graphite and “phase-change” covers often feel cool to the touch for the first minutes, then reach body temperature. Airflow through coils and not sinking too deeply make a larger, more lasting difference. Actively temperature-controlled covers (water pumped through a pad) are a different product: one study found more deep sleep with cooling early in the night, but the company that makes the device funded it and its employees wrote it.
Firmness scores. A “6 out of 10” from one brand is not a 6 from another. Only the trial period tells you what a number means for your body.
“Doctor recommended” or “chiropractor approved”. These are endorsements, not evidence.
Sleep-tracking “smart” mattresses. The sensors may be interesting, but tracking does not change how the mattress supports you. See our guide to wearables and HRV for what these numbers can and cannot tell you.

Toppers vs replacing

A topper is a layer of foam, latex, wool or fibre, usually 5–8 cm (2–3 inches) thick, laid on top of the mattress. It is a good, cheap fix for one specific problem: a mattress that is in good shape but feels too firm. A soft topper on a medium-firm or firm mattress can make it comfortable for a side sleeper, for example.

A topper cannot fix sagging. It follows the dip underneath it, so the spine still ends up bent. It also cannot make a soft mattress meaningfully firmer. And a thick memory-foam topper will make a bed warmer. If the mattress under it is worn out, put the money towards a replacement instead. Pillows matter too: a pillow that is too high or too flat for your position can undo a good mattress. Our pillow reference covers loft and fill.

⚠️When back pain is not a mattress problem

See a clinician rather than shopping for a new bed if back pain comes with fever, unexplained weight loss, numbness around the genitals or buttocks, new bladder or bowel problems, leg weakness, pain after a fall or accident, a history of cancer, or pain that is worse at night no matter how you lie. These can signal conditions that need prompt assessment.

⚠️Babies and toddlers

Infants need a firm, flat mattress made for their crib or cot, fitted without gaps, with no toppers, pillows or soft bedding. Adult mattress advice does not apply to them.

Frequently Asked Questions

Is a firm mattress better for back pain?

Not according to the best trial. In the Kovacs study, people with chronic low-back pain did better on medium-firm than on firm mattresses, and a later trial found a very hard mattress slightly worse than foam or a waterbed. “Firm is best” is folklore.

Memory foam or hybrid if I sleep hot?

A hybrid or latex mattress is usually the cooler choice, because air moves through coils and you sink less deeply. If you prefer foam, choose a firmer one with a breathable cover, and look at your bedding and room temperature before paying extra for “cooling” features.

How long should a mattress last?

It depends on materials, body weight and use. Treat seven to ten years as a rough range, but replace on the signs — sagging, morning stiffness that eases once you are up, sleeping better elsewhere — rather than on a calendar.

Can a topper fix a sagging mattress?

No. A topper follows whatever is beneath it. Toppers are useful for softening a mattress that is still structurally sound.

Is an expensive mattress worth it?

Price buys durability (denser foams, better coils), materials such as natural latex, and sometimes a longer trial or warranty. It does not buy a better match for your body. None of the trials above compared price levels, so there is no evidence that paying more buys less pain.

Should couples buy a split mattress?

If you differ a lot in weight, sleeping position or how hot you sleep, yes, it is worth considering. Otherwise one of you is always on the wrong mattress. Two single mattresses on one frame, or a split-firmness model, solves it.

The Bottom Line

Start from medium-firm, adjust for how you sleep and how much you weigh, and choose the material for heat and motion rather than for “support” claims. Buy where the trial period is generous and the return terms are clear, give the mattress a month, and keep a short log so you are judging by nights rather than memory. Ignore “orthopedic”, zoning and cooling-gel claims until someone shows independent data. For the rest of the bedroom, see our guides to the sleep environment and weighted blankets, and the sleep learning track.

Compare mattresses and sleep gear on the Formulate sleep shelf →

How to Choose a Mattress: What the Evidence Says About Firmness and Back Pain