What Works for Lower Back Pain, According to Science

Most episodes of lower back pain improve on their own within a few weeks, and science consistently points to staying active and trying non-drug treatments first. A 2017 clinical practice guideline from the American College of Physicians made that the official recommendation, listing exercise, heat, massage, spinal manipulation, and psychological therapies ahead of medication for both short-term and long-lasting pain.¹ That guidance has held up well, but the details matter. Some popular treatments have strong evidence, others barely outperform a sugar pill, and a few things that feel like they should help actually slow recovery.

Stay Active and Skip the Bed Rest

The single most consistent finding in back-pain research is that resting in bed makes things worse. A landmark trial published in the New England Journal of Medicine compared bed rest, back exercises, and simply continuing ordinary activities. The group told to go about their normal routines recovered fastest. Bed rest produced the slowest recovery, with more pain, more days off work, and worse flexibility at both three weeks and three months.² Later systematic reviews confirmed the pattern: bed rest does not help acute back pain and may delay recovery, while advice to stay active leads to faster return to work and fewer long-term problems.³

A Cochrane review of the same question found small but real improvements in both pain and daily functioning for people advised to stay active compared to those told to rest, based on moderate-quality evidence.⁴ One exception: for people with sciatica (pain shooting down the leg from a pinched nerve), the advantage of staying active versus resting largely disappears. So the advice to keep moving is strongest for garden-variety lower back pain without leg symptoms.

Exercise Is the Strongest Long-Term Treatment

Once pain has lasted more than a few weeks, exercise becomes the treatment with the broadest and most durable evidence behind it. A Cochrane review covering dozens of trials found moderate-certainty evidence that exercise reduces chronic low back pain compared to no treatment, usual care, or placebo, with a clinically meaningful difference in pain scores.⁵ The improvement in daily functioning was statistically real but smaller, not quite crossing the threshold researchers set for clinical importance. Still, exercise beats doing nothing by a comfortable margin.

What kind of exercise works best? The honest answer is that no single type stands clearly above the rest. A network meta-analysis that compared Pilates, strength training, core-based exercises, and mind-body approaches found that Pilates ranked highest for both pain and disability reduction, with core-based and mind-body exercises close behind.⁶ But when you look at head-to-head trials, Pilates performs about the same as other forms of exercise. A systematic review noted that while Pilates clearly beats doing nothing or receiving only routine medical care, it produces similar improvements to massage therapy or other structured exercise programs.⁷ In practice, the exercise you will actually do consistently is the exercise that works.

Psychological Therapies and the Brain-Pain Connection

Chronic back pain is not purely a tissue problem. The brain’s processing of pain signals, fear of re-injury, catastrophic thinking, and depression all influence how much pain you feel and how disabled you become. This is why cognitive behavioral therapy keeps appearing in treatment guidelines. A review in the Global Spine Journal described CBT’s effectiveness for chronic low back pain as “unquestioned,” covering its impact on pain intensity, disability, and the psychological factors that keep pain cycles going.⁸

Mindfulness-based therapy has emerged as a viable alternative. A randomized clinical trial published in JAMA Network Open compared mindfulness training directly to CBT in patients with chronic back pain who were taking opioids. Both groups improved significantly over a year, with no serious side effects, and the mindfulness group was statistically noninferior to the CBT group on both pain and function.⁹ One practical barrier to CBT is that it typically runs eight sessions over multiple weeks. A condensed single-session program called Empowered Relief was tested against full CBT and health education; at six months, its effects kept pace with the longer program on most outcomes, including pain catastrophizing and anxiety.¹⁰ That suggests you do not necessarily need months of therapy to gain real benefit.

Over-the-Counter Medications

When pain is bad enough that you reach for a pill, nonsteroidal anti-inflammatory drugs like ibuprofen and naproxen are the first-line choice supported by evidence. An overview of medication research found that NSAIDs may outperform placebo for both acute and chronic back pain.¹¹ Acetaminophen (paracetamol) is the other obvious over-the-counter option, but a Cochrane review found moderate evidence that NSAIDs are not more effective than acetaminophen for acute back pain, though acetaminophen causes fewer side effects.¹² That means either can help, but if you tolerate both, NSAIDs likely have a slight edge because of their anti-inflammatory effect.

Combining the two may offer an additional advantage. A multicenter trial comparing ibuprofen alone to ibuprofen plus acetaminophen found that the combination group had significantly lower pain intensity by day four and better spinal mobility at the end of follow-up, with minimal stomach complaints in either group.¹³ For short flare-ups, these medications are a reasonable bridge while you stay active and let the episode resolve.

Why Opioids Are a Poor Long-Term Solution

Opioids can reduce acute pain in the short term, but the evidence for their use in chronic back pain is bleak. A systematic review in the Journal of the American Board of Family Medicine concluded that using opioids for up to twelve months is no more effective in improving chronic low back pain outcomes than treatment without opioids.¹⁴ Meanwhile, the harms are well documented. Long-term opioid use carries risks of tolerance (needing higher doses for the same effect), a paradoxical increase in pain sensitivity called hyperalgesia, constipation, sedation, falls, depression, sexual dysfunction, addiction, and overdose death.¹⁵ For a condition where other approaches work at least as well, that risk profile is hard to justify.

Spinal Manipulation

Chiropractic adjustments and other forms of spinal manipulation are among the most widely used treatments for back pain, yet the evidence paints them as helpful but modest. A Cochrane review of spinal manipulative therapy for chronic low back pain found high-quality evidence that it produces a small, short-term effect on pain and function compared to other interventions. The review explicitly noted that this effect, while statistically real, was not clinically relevant, meaning the average patient would barely notice the difference.¹⁶ For acute back pain, a JAMA meta-analysis similarly found modest improvements in pain and function at up to six weeks, with significant variation across studies.¹⁷

The safety picture is reassuring: adverse events are almost always minor muscle soreness that resolves quickly. A systematic review in the BMJ found that most observed side effects were musculoskeletal, transient, and mild to moderate, with only one possibly serious adverse event across the included studies.¹⁸ Manipulation appears safe and may be worth trying, but expecting it to fix the underlying problem is not supported by the research.

Acupuncture

Acupuncture is one of the more polarizing back-pain treatments. A large randomized trial compared real individualized acupuncture, standardized acupuncture, and simulated acupuncture (where needles were placed at non-traditional points with a toothpick-like device) against usual care. All three acupuncture groups improved by roughly double the amount the usual-care group did, but there was no meaningful difference between real and sham acupuncture.¹⁹ That pattern, where acupuncture beats doing nothing but does not clearly beat a convincing placebo, has shown up repeatedly. A Cochrane review found low-certainty evidence that acupuncture may relieve pain in the very short term compared to sham, but the size of the improvement did not meet the threshold researchers consider clinically important.²⁰

This does not mean acupuncture is useless. The ritual of treatment, the therapeutic relationship, and the expectation of improvement together produce measurable benefit. Whether you credit that to the needles or to the treatment context, people who go to acupuncturists often feel better than people who do nothing. The question is whether that improvement is specific to acupuncture itself, and the current evidence suggests it largely is not.

Epidural Injections and Their Limits

Epidural corticosteroid injections are commonly offered for sciatica, the radiating leg pain caused by a compressed nerve root. A systematic review and meta-analysis in the Annals of Internal Medicine found that the injections produce a small, statistically significant reduction in leg pain and disability in the short term, but the long-term effects were smaller and not statistically significant.²¹ The authors raised questions about the clinical utility of the procedure given how small the benefit was. A case series from Lebanon that tracked patients after early epidural injections found that peak benefit came about two weeks after the injection, but roughly half the patients needed a second injection within three to six months as the effect wore off.²² Injections can buy you a window of relief to start physical therapy or get through a severe flare, but they are not a lasting fix on their own.

When Surgery Makes Sense

Surgery is generally reserved for cases where a specific structural problem has been identified, like a severely herniated disc pressing on a nerve or spinal stenosis narrowing the canal. Even in those cases, the advantage over conservative treatment takes time to appear. A meta-analysis of randomized trials comparing surgery to conservative treatment for lumbar spinal stenosis found that surgical patients had better outcomes after one year but no significant advantage in the first six months, along with a higher complication rate throughout follow-up.²³

For the common scenario of non-specific chronic back pain without clear nerve compression, spinal fusion surgery has much weaker evidence. A systematic review of fusion trials found that when fusion was compared against a structured cognitive-behavioral program addressing fear of injury, the results were similar, and the trials that did find a benefit for fusion had serious methodological problems like not following the original treatment assignments.²⁴ Two trials that directly compared multidisciplinary rehabilitation to surgery found little difference in outcomes, with surgery carrying more risk of adverse events.²⁵

Imaging Often Does More Harm Than Good

One of the most counterintuitive findings in back-pain research is that getting an MRI or X-ray does not improve outcomes for most people. A systematic review published in The Lancet found no significant differences in pain or function between patients who received immediate imaging and those who received usual care without imaging, at either three months or twelve months.²⁶ A randomized controlled trial in JAMA compared rapid MRI to plain X-rays for back pain and found no clinically or statistically meaningful difference in disability, pain, or overall health scores at one year.²⁷

The reason imaging can backfire is that scans frequently reveal “abnormalities” that are actually normal for your age, things like bulging discs, degenerative changes, and mild stenosis. These findings rarely explain the pain, but once you see them on a report, they can fuel anxiety, lead to unnecessary procedures, and make you afraid to move. Guidelines consistently recommend against routine imaging unless there are warning signs of a serious underlying condition, such as unexplained weight loss, a history of cancer, progressive neurological deficits, or trauma.

Multidisciplinary Rehabilitation

When back pain has become chronic and resistant to simpler interventions, programs that combine physical therapy, psychological support, and education about pain under one roof tend to outperform any single approach. A Cochrane systematic review found moderate to low quality evidence that multidisciplinary rehabilitation reduces both pain and disability compared to usual care, and also compared to physical treatment alone.²⁸ The effects translated to real-world outcomes: patients in multidisciplinary programs were nearly twice as likely to be working one year later compared to those receiving only physical treatments.²⁵ These programs are time-intensive and not available everywhere, but they represent the strongest evidence-based approach for people whose pain has resisted simpler strategies.

Workplace Habits and Sleep

For the millions of desk workers whose back pain flares during the workday, the evidence on sit-stand desks is encouraging. A systematic review and meta-analysis found that sit-stand workstations may reduce low back discomfort among workers, though the research has not yet nailed down the ideal ratio of sitting to standing time.²⁹ A randomized trial called Stand Back tested a multicomponent intervention for desk workers with chronic back pain that combined a sit-stand desk attachment, a wrist-worn movement reminder, counseling to reduce sitting, and pain self-management skills. At six months, participants in the intervention group were sitting about an hour and a half less per day, and their disability scores had dropped by half compared to a 14% drop in controls.³⁰

Sleep quality is another underappreciated factor. A cohort study of schoolteachers found that poor sleep quality predicted the persistence of chronic low back pain in those who already had it and increased the chances of developing it in those who did not.³¹ The relationship runs both ways: pain disrupts sleep, and poor sleep amplifies pain. Addressing sleep problems through better sleep habits, treating sleep apnea, or working with a therapist may be one of the more underused strategies for breaking the cycle.

The Placebo Effect Is Real, and Surprisingly Large

Perhaps the most fascinating corner of back-pain research involves placebo treatments, and not just the hidden kind. In a pioneering trial, researchers gave patients with chronic low back pain pills that were explicitly labeled as placebos with no active medication. They told participants the pills were inert but explained that placebos can trigger real physiological responses. Even knowing the pills were fake, the open-label placebo group experienced significantly greater pain reduction and less disability than the group receiving only their usual treatment, with moderate to large effect sizes.³²

These results have been replicated. A JAMA Network Open trial using open-label placebo injections for chronic back pain found significant reductions in pain intensity compared to usual care.³³ A third trial found that three weeks of open-label placebo reduced pain, disability, and depression compared to treatment as usual.³⁴ These findings do not mean back pain is imaginary. They mean the context in which treatment happens, the therapeutic relationship, the ritual of care, the expectation that something is being done, has genuine power to modulate pain processing in the brain. This helps explain why acupuncture beats usual care but not sham acupuncture, and why many treatments that feel helpful in practice show only small advantages over placebos in clinical trials.

Why the Human Spine Is Vulnerable in the First Place

Lower back pain affected an estimated 619 million people worldwide in 2020, and that number is projected to reach about 843 million by 2050.³⁵ Part of the reason back pain is so common lies in evolutionary history. Walking upright on two legs required the spine to develop an S-shaped curve, particularly the inward curve of the lower back called lordosis. This arrangement positions your upper body’s center of mass over your hips, which makes standing and walking efficient, but it also creates shearing forces on the lumbar vertebrae that a four-legged ancestor’s spine did not have to deal with.³⁶

Some people may be more vulnerable than others based on the shape of their vertebrae. A study published in BMC Evolutionary Biology found that people who develop disc herniations tend to have vertebral shapes that are closer to those of our primate ancestors and less well adapted for bipedal loading.³⁷ In other words, natural variation in spinal anatomy means some people’s lower backs are slightly less suited to the demands humans place on them. It is a reminder that back pain is not a modern disease caused purely by office chairs and sedentary habits. It is, in part, a structural trade-off baked into being human.