What Actually Causes Back Pain

Patients arrive in my clinic with an MRI report full of alarming words, a folder of conflicting advice, and no clear idea of what is actually wrong with them. That is not their fault. Back pain is genuinely complicated, and a lot of the explanation people are given is either too vague to act on or too confident to be honest.

What follows is the framework I actually use. I am a fellowship-trained neurosurgeon and I operate for a living, so read the surgical section knowing my bias. I have tried to be clear about where the evidence is strong and where it runs out. None of this is medical advice, and I cannot examine you or see your imaging. The goal is that when you sit down with your own doctor, you know which questions matter.

There are three different ways your back can hurt

This is the single most useful idea in this article, so it goes first. Almost every argument about back pain comes from mixing up three categories that are not the same problem, do not feel the same, and do not respond to the same treatment.

Nociceptive pain, meaning tissue pain. Something with pain nerves in it is being irritated, inflamed, stretched, or loaded badly. A muscle, a ligament, a joint capsule, a vertebral endplate, or the outer wall of a disc. The tissue really is under stress and your nervous system is reporting it accurately. It tends to be reasonably localized, you can often point to it, and it is proportional: specific positions and loads make it worse in a way you can predict, and it changes with rest and position.

Neuropathic pain, meaning nerve pain. The nerve itself is the problem, because it is compressed, stretched, or chemically irritated, and it is firing on its own. This is what radiculopathy is, and it is usually what people mean when they say sciatica. It travels. It follows the territory of a nerve rather than sitting in one place, and it gets described as electric, burning, shooting, or like ice water running down the leg. It comes with numbness, tingling, and sometimes weakness, because a nerve root is not only a pain wire, it is a motor and sensory cable.

Nociplastic pain, meaning the pain processing itself has changed. Most people have never heard of this category and many clinicians are undertrained in it. The International Association for the Study of Pain formally added it in 2017, defining it as pain arising from altered nociception without clear evidence of tissue damage or nerve injury sufficient to explain it.

In plain terms: the alarm system has been running so long that it has turned its own gain up. I want to be very clear about what this does not mean, because patients hear "central sensitization" and think they are being called a liar. The pain is completely real. What has changed is where it is being generated. It tends to spread beyond the original injury, it is diffuse and hard to point to, it is disproportionate and unpredictable, light touch can hurt, and it usually travels with poor sleep, fatigue, stress, and a long history.

Here is why this framework matters practically. Most people with chronic back pain have some mix of all three. Surgery can only address the first two, and only when there is a specific structural target. When most of what is driving your pain sits in the third category, an operation will not fix it, and that is one of the most common routes to being worse after surgery than before it.

Before you read your MRI report, read this

Otherwise every line of it will sound like a catastrophe.

A well-known systematic review by Brinjikji and colleagues pooled imaging from more than 3,000 people with no back pain at all. Not mild pain. None. Among those pain-free people:

• Disc degeneration: 37 percent of 20 year olds, rising to 96 percent of 80 year olds
• Disc bulge: 30 percent at age 20, 84 percent at age 80
• Disc protrusion: 29 percent at age 20, 43 percent at age 80
• Annular fissure: 19 percent at age 20, 29 percent at age 80

More than half of pain-free people in their thirties already have degeneration, height loss, or bulging. So when your report reads "degenerative disc disease with a broad-based bulge," that sentence on its own tells you close to nothing about why you hurt.

Two rules follow, and I use both every day.

A finding only counts if it explains your symptoms. The level has to match, the side has to match, and the pattern has to match. Otherwise it is background noise. Radiologists are obliged to report everything they see, which is exactly what you want from a radiologist and exactly why the report reads the way it does.

A normal MRI does not mean nothing is wrong. Your scan is taken lying flat and perfectly still. Your pain happens standing, walking, bending, and loading. Discs change shape under load, and foramen close down with extension, so a segment that looks fine at rest can behave badly at work. Muscle pain, ligament pain, and nociplastic pain do not appear on any scan we currently have.

The common causes, and how I tell them apart

Muscle and soft tissue strain

The muscles, fascia, and ligaments around the spine get overloaded or microscopically torn, then go into protective spasm. That spasm is itself painful, which is why people describe feeling locked up. The guarding is your body splinting the area: useful for a couple of days, actively unhelpful after that.

It usually starts around an identifiable event, sits off the midline rather than dead center, is tender to press, hurts to move but eases at rest in a supported position, and produces no leg symptoms. It improves meaningfully over days to a few weeks. This is the most common cause of a sudden back attack, and most of them resolve with time and gradual movement.

Discogenic pain and annular tears

A disc has a tough outer wall, the annulus fibrosus, and a gel center, the nucleus pulposus. When that wall tears, or when the disc dries out and loses height, the disc becomes a pain generator on its own with no nerve compression at all.

The mechanism is worth understanding. In a healthy disc only the outer third of the annulus carries nerve fibers. As a disc degenerates, new blood vessels and new nerve endings grow deeper into it, so a degenerated disc is literally more innervated than a healthy one. On top of that, when the annulus tears, inflammatory material from the nucleus leaks out and chemically irritates whatever it touches, including nearby nerve roots. That is how you can get genuine leg symptoms from a tear showing no real compression on the scan.

The pattern: deep, central, midline pain, much worse with sitting, bending forward, and getting up out of a chair or a car. Coughing and sneezing spike it. It often eases when you walk or lie flat. Honest caveat: annular fissures are subtle on MRI and easy to under-report, and most settle over a few months. If yours has been painful for years, I would look harder rather than assume the tear accounts for all of it.

Vertebrogenic pain, meaning the endplate

Newer, and most patients have never heard of it. The vertebral endplate is the bony surface the disc sits against. When the disc degenerates the endplate is damaged too, and the bone marrow underneath reacts. Those reactive changes are what a radiologist means by Modic type 1 and type 2 changes.

The endplate is supplied by the basivertebral nerve, which runs into the middle of the vertebral body, so this is bone pain rather than disc pain or nerve root pain, even though it sits right between both. It presents as deep, central, non-radiating low back pain, worse sitting and bending forward, in someone whose MRI shows Modic changes at that level. Modic changes are one of the more specific MRI findings we have for a painful level, which is unusual in this field, and there is now a targeted procedure aimed at it.

Facet joint pain

The facets are the two small joints at the back of each spinal segment that guide and limit motion. They are real synovial joints with cartilage and a capsule, and like any joint they develop arthritis.

Facet pain sits to one or both sides of the midline rather than dead center. It is worse arching backwards, worse standing still for a long time, worse twisting, and it often eases when you sit or bend forward. It can refer into the buttock and the back of the thigh but usually stops above the knee. Mornings are stiff and the first few movements are the worst. Depending on the study and the population, facet joints are implicated in somewhere between 15 and 45 percent of chronic low back pain. That range is wide because confirmation requires diagnostic blocks, which carry a meaningful false positive rate.

Sacroiliac joint pain

The joint between the sacrum and the pelvis barely moves, but it transmits every bit of load between your spine and your legs. It gets irritated after pregnancy, a fall onto the buttock, a leg length difference, or a fusion above it that dumped extra load into it.

SI joint pain has a fairly distinctive location: one-sided, very low, just off the midline around the bony bump at the back of the pelvis. Patients point to it with one finger. It is worse rolling over in bed, standing on one leg, climbing stairs, and getting out of a car, and it can refer into the groin and the back of the thigh. It is regularly missed and also regularly over-diagnosed, and it is often unmasked once louder pain elsewhere is quieted.

Disc herniation with radiculopathy

Nucleus material pushes through a tear in the annulus and contacts a nerve root. Two mechanisms work together: mechanical compression, and chemical inflammation from the nucleus material, which is highly irritating to neural tissue. The chemical component explains why a small herniation can hurt enormously and why anti-inflammatory treatment sometimes works well without the herniation changing at all.

The pattern is back pain now overshadowed by leg pain. The leg pain follows a stripe, usually past the knee, with numbness or tingling in a matching area. Worse sitting, bending, coughing and sneezing, often easier lying down or walking. Most improve. I have written separately about which herniations heal and when surgery is worth considering.

Spinal stenosis and neurogenic claudication

Narrowing of the space the nerves travel through, from disc bulging forward, facet arthritis pushing in from behind, and the ligamentum flavum thickening and infolding. When you stand and extend your spine the canal narrows further and that ligament buckles inward, compressing the nerves and throttling their blood supply. Walking raises their metabolic demand at exactly the moment supply is worst.

This is the most pattern-recognizable condition on the list. Symptoms are usually in both legs, described as heaviness, cramping, burning, and a sense that the legs will not hold you. It starts after a fairly reproducible walking distance and goes away when you sit or lean forward. People lean on a shopping cart and walk for ages, then cannot cross a parking lot upright. Cycling is fine because you are bent forward. If that is your pattern, it is stenosis with neurogenic claudication until proven otherwise, and it is one of the conditions where surgery works very well.

Spondylolisthesis and instability

One vertebra has slipped forward on the one below. The isthmic form comes from a stress fracture in the pars and shows up in younger people and athletes. The degenerative form comes from worn facets and a degenerated disc no longer holding the segment in line, shows up after 50, and is considerably more common in women.

It produces two separate problems: mechanical back pain from the unstable segment that worsens the longer you stay upright, and leg symptoms from narrowing of the canal and foramen. Expect back pain much worse with standing and walking and relieved by sitting, often with a sense of the back giving way or needing to shift constantly. Flexion and extension X-rays matter here, because a slip that looks stable on an MRI taken lying down can move when you stand and bend.

Adult deformity and sagittal imbalance

Scoliosis, or more importantly in adults, loss of the normal forward curve of the lower back so the trunk sits in front of the pelvis. Your muscles are now doing a job your skeleton is supposed to do, so the pain is fatigue and burning rather than a sharp catch.

The signature is pain that is fine in the morning and unbearable by evening, builds the longer you stand, and is relieved by sitting or lying down. People bend their knees or lean on things to stay upright. Sagittal imbalance needs standing full-length X-rays and cannot be diagnosed on a lumbar MRI, which is one reason it gets missed for years.

Vertebral compression fracture

A vertebral body collapses, usually from osteoporosis, sometimes from tumor, sometimes from trauma. Onset is sudden, often after something trivial like a sneeze or lifting a bag, and the pain is sharply localized to one spot on the midline. Going from lying down to sitting up is very painful, and it often settles once you are up. In someone over 65, on long-term steroids, or with known osteoporosis, a compression fracture goes near the top of the list. Fractures account for roughly 1 to 4 percent of back pain presenting to primary care, so they are not rare, and they get missed because people assume they would know if they had broken something.

Inflammatory back pain

An autoimmune inflammatory arthritis of the spine and SI joints, including ankylosing spondylitis. This one is worth learning because it is mechanically backwards from everything else here. Onset under 45. Symptoms build slowly rather than starting on a specific day. Morning stiffness lasting more than 30 to 60 minutes. It gets better with exercise and worse with rest, the opposite of mechanical pain. Night pain in the second half of the night that gets you out of bed. Often a dramatic rather than modest response to anti-inflammatories, and frequently a personal or family history of psoriasis, inflammatory bowel disease, or uveitis. If that sounds like you, you need a rheumatologist rather than a spine surgeon.

Things that are not your spine at all

Not everything that hurts in the back area comes from the back.

• Hip arthritis: groin pain, trouble putting on socks and shoes, pain getting out of a car, limited internal rotation. Routinely treated as spine pain for years.
• Vascular claudication: calf pain with walking relieved by simply standing still, without needing to bend forward. Look for absent pulses, cold feet, hair loss on the legs, smoking history.
• Peripheral neuropathy: numbness and burning in a stocking distribution in both feet, worse at night, not following any one nerve root.
• Abdominal and vascular causes: an abdominal aortic aneurysm can present as back pain and is one of the few things here that can kill you quickly.
• Kidney stones or infection: flank pain, unable to get comfortable in any position, blood in urine, fever.
• Gynecologic causes: endometriosis in particular, where pain tracks with the menstrual cycle.
• Piriformis and gluteal tendon problems: buttock pain tender to direct pressure and reproduced by hip movement rather than spinal movement.

A rough field guide

There is real overlap and none of this replaces an examination, but the patterns generally sort out like this:

• Worse sitting, bending forward, coughing, getting out of a chair, better standing or lying: disc
• Worse arching backwards, twisting, standing still, better sitting: facet joints
• One-sided, very low, just off the midline, worse rolling in bed and standing on one leg: SI joint
• Leg pain in a stripe past the knee with numbness or tingling: nerve root
• Both legs heavy and crampy after a predictable walking distance, relieved by sitting or leaning forward: spinal stenosis
• Deep central pain, worse the longer you stand, better sitting, with a slip on X-ray: spondylolisthesis
• Fine in the morning, unbearable by evening, builds with time upright: deformity or sagittal imbalance
• Sudden, sharp, one spot on the midline after something trivial, in someone older: compression fracture
• Morning stiffness over an hour, better with exercise, worse with rest, onset under 45: inflammatory
• Groin pain, trouble with socks and shoes: hip, not spine
• Widespread, disproportionate, unpredictable, sensitive to touch and stress: significant nociplastic component

Red flags, with some honesty about them

Go to an emergency department now, not next week, for:

• New numbness in the saddle area, meaning the inner thighs, groin, buttocks, and genital region
• Loss of bladder or bowel control, or a change in the pattern. Specifically: not being able to start a stream, not being able to feel yourself passing urine, not being able to tell when your bladder is full, needing to strain, or leaking without knowing. It is often subtle at the start rather than dramatic
• Weakness that is getting worse, especially a foot that drops or a leg that gives way
• Both legs becoming symptomatic at the same time in a new way

That combination raises the possibility of cauda equina syndrome, where the timeline matters enormously. Do not wait it out and do not wait for an outpatient MRI.

See a doctor promptly, though not necessarily the emergency room, for fever with back pain, IV drug use, recent spinal procedure or infection elsewhere, unexplained weight loss, a history of cancer, night pain that consistently wakes you, new significant pain after 50 with osteoporosis or steroid use, or major trauma.

Now the honesty. Scaring people is not the goal. In primary care, less than 1 percent of back pain turns out to be malignancy. Spinal infection is roughly 0.01 percent, cauda equina roughly 0.04 percent, and fractures about 1 to 4 percent. Meanwhile up to 80 percent of people presenting with back pain have at least one red flag present, which tells you how nonspecific any single one is on its own. Read the list, take it seriously, then put it down. The odds are strongly in your favor.

What actually helps

Time. Most acute back pain improves substantially in the first four to six weeks. That is not a dismissal, it is the single most reliable statistic in this field.

Movement, not bed rest. More than a day or two flat makes things worse. Deconditioning, stiffness and fear compound quickly. I do not even let my post-operative patients stay in bed the day of surgery.

Keep working and living if you can, modified as needed. Prolonged withdrawal from normal activity is one of the strongest predictors of pain becoming chronic. This matters more than people think, and it is hard at first and then steadily easier.

Exercise is the foundation, and the type matters less than you would expect. No single type has proven clearly better across the literature. Core stabilization, motor control work, directional preference approaches, general aerobic work, yoga and tai chi all show benefit and none dominates. What that really tells you is that consistency matters far more than program selection. The trials showing the strongest effects run 8 to 12 weeks, and most people quit at three. If I had to name priorities: hip mobility, because stiff hips make the lumbar spine do their work; deep core endurance rather than raw strength; regular walking; and avoiding long static positions in any direction, including good ones.

Medications, roughly in order of evidence. Anti-inflammatories are first line, with a real but modest effect, and are not something to live on. Muscle relaxants are useful short term in acute pain with spasm, mostly by making you drowsy, and should be used for days rather than months. Duloxetine is an option in chronic pain, working through descending pain modulation rather than at the back itself, and is more useful when there is a nociplastic component. Acetaminophen showed no benefit over placebo for acute low back pain in a placebo-controlled trial, so if it helps you that is fine, but do not expect much. Tricyclics were found no better than placebo for chronic low back pain in the American College of Physicians review. Gabapentin and pregabalin have weak evidence for axial back pain and are more defensible with a genuine neuropathic component. Oral steroids are not supported by guidelines for acute back pain, though practice varies for acute radicular flares. Opioids are a last resort, short term, with a clear exit plan.

Non-drug treatments worth trying. Superficial heat, massage, acupuncture and spinal manipulation all have supporting evidence in acute pain. For chronic pain, add tai chi, yoga, mindfulness-based stress reduction, and cognitive behavioral therapy, and pain neuroscience education has decent evidence. I know how psychological interventions sound to someone in real physical pain. Nobody is saying it is in your head. These target the third pain category specifically, and for people largely in that category this is the treatment that addresses the actual mechanism.

The levers people underestimate. Sleep, because poor sleep amplifies pain and pain wrecks sleep, and breaking that loop is worth more than most pills. Smoking, which impairs disc nutrition and is one of the strongest modifiable risk factors for both degeneration and poor surgical outcomes. Weight, because every extra pound is load through the discs and facets. Stress, which genuinely amplifies pain processing. I have seen more flares triggered by a bad month at work than by lifting anything.

Injections and procedures. An epidural steroid injection reduces inflammation around an irritated root and can break the cycle enough to let you rehabilitate. The short-term evidence is good and the long-term evidence less impressive, but it also has real diagnostic value: if numbing a specific target kills the pain, we are aiming at the right place. Medial branch blocks confirm facet pain and radiofrequency ablation treats it, though the nerves regrow over 6 to 18 months, so understand it as pain management rather than repair. SI joint injections are both diagnostic and therapeutic. Basivertebral nerve ablation targets vertebrogenic pain with Modic type 1 or 2 changes and is supported by two randomized trials in properly selected patients, which is a good illustration of why getting the diagnosis right matters: the treatment only works for the specific problem it targets.

When surgery is actually indicated

I want to state my bias openly. I am a spine surgeon, I operate for a living, and I still believe the majority of back pain should never see an operating room. The body has enormous healing capacity, surgery permanently changes anatomy, and a great deal of poor spine surgery has been done over the years.

Every surgical decision comes down to two questions. Is there a structural problem an operation can physically correct? And does that specific problem explain the symptoms that are actually ruining your life? If either answer is no, an operation is unlikely to help, no matter how bad the MRI looks.

Emergent. Cauda equina syndrome. Progressive or severe motor weakness such as a new foot drop. Spinal infection with neurological compromise or instability. Tumor causing cord or nerve compression. An unstable fracture, or any fracture with a neurological deficit.

Strong indications, where surgery reliably outperforms continued conservative care. Radiculopathy from a herniated disc that has not settled after roughly 6 to 12 weeks of genuine conservative treatment, in someone whose imaging matches their symptoms, with the calculus shifting sooner if there is weakness. Spinal stenosis with neurogenic claudication that limits your walking and your life, where decompression is one of the most satisfying operations in spine surgery when the pattern is clear. Symptomatic spondylolisthesis with stenosis, particularly with leg symptoms. Instability documented on flexion and extension imaging rather than merely suspected. Deformity with real sagittal imbalance in someone who cannot stand upright without exhausting themselves.

Weak indications, where everyone should slow down. Axial back pain alone, no leg symptoms, no instability, no clear single pain generator, on a background of ordinary degenerative changes. This is where most of the disappointing spine surgery in this country happens. If your surgeon cannot name the specific structure they are treating and explain why they believe it is that one, you are not ready for an operation.

Expectations worth setting. Leg pain responds to surgery far better than back pain does, so if your dominant problem is the back itself, temper expectations regardless of what is offered. Numbness and weakness recover more slowly than pain and sometimes incompletely, because nerves heal on the order of months. Reherniation after a discectomy runs somewhere between 10 and 20 percent, since there is now a hole in the annulus where the material came out. And surgery removes a mechanical problem: it does not reverse nociplastic pain, which needs addressing alongside rather than ignoring.

Questions I would want you to ask any surgeon

• What specific structure do you believe is generating my pain, and what makes you confident it is that one?
• Which of my symptoms is this operation meant to fix, and which will it not touch?
• What happens if I wait six months?
• Why this operation rather than a smaller one? Is there a smaller option, and can it be done minimally invasively?
• What does a good result look like at one year, and what is your reoperation rate?

A surgeon who welcomes those questions is the one you want. This decision is much harder to undo than it is to make, so choose carefully. If you have been offered surgery and want another opinion before committing, that is a completely reasonable thing to ask for, and second opinions are a routine part of what I do.