Lumbar Traction Parameters: Force, Angle, and Duration

Lumbar traction outcomes depend heavily on a handful of adjustable settings: how much force is applied, at what angle, in what position, for how long, and whether the pull is steady or cycled on and off. Getting these parameters right can mean the difference between measurable disc changes on imaging and a treatment that does essentially nothing. The challenge is that research on each parameter often comes from small studies, and the “optimal” combination remains a moving target. Still, the accumulated evidence points to some clear tendencies worth understanding before you or your clinician decide how to set up a traction session.

How Much Force Is Enough

The load applied during lumbar traction is typically expressed as a percentage of body weight. A common clinical range falls between about 25% and 50% of body weight, but the research that has actually measured structural changes in the spine tends to cluster around the higher end of that window. One MRI-based study applied horizontal traction at 42% of body weight for 30 minutes and documented measurable changes in intervertebral disc dimensions afterward.¹ Forces below roughly 25% of body weight are generally considered too low to overcome the resistance of paraspinal muscles and spinal ligaments, meaning most of the pull is absorbed by soft tissue rather than actually separating vertebrae.

A finite element analysis confirmed that intradiscal pressure drops in proportion to applied force when the traction direction is along the spinal axis or angled slightly forward. Interestingly, when force was directed toward the facet joints (a rearward angle), pressure inside the disc did not change meaningfully regardless of how much force was used.² This suggests that force magnitude alone is not the whole story; the direction of that force matters just as much.

Real-time MRI research during continuous traction showed that the effects accumulate over time. Herniated disc volume reduced by roughly 9% after 10 minutes, about 15% after 20 minutes, and nearly 18% after 30 minutes, with spinal column length increasing by about 1.5% over the same period.³ That progressive effect implies that a single brief pull, even at high force, may not deliver as much structural change as a sustained or repeated session at moderate force.

Pull Angle and Patient Positioning

The angle at which traction force is applied determines which spinal structures bear the load. A finite element study examining different traction table angles found that the effect on individual ligaments varied dramatically depending on spinal level. At the L3-L4 segment, the loading angle influenced the posterior longitudinal ligament by about 57%, while the same angle affected that ligament at L4-L5 by only about 11%. Meanwhile, the supraspinous and interspinous ligaments at L4-L5 were more affected by angle changes than those at L3-L4.⁴ The practical takeaway is that a single traction angle does not load the lower lumbar spine uniformly, so clinicians sometimes adjust the angle depending on which disc level they are targeting.

Hip position during traction is another parameter with clear evidence behind it. A study measuring vertebral separation at different hip flexion angles found that posterior interspace opening increased as hip flexion went from flat (0 degrees) to fully bent (90 degrees). The differences were statistically significant at the L4-L5 and L5-S1 levels, which are the two segments most commonly involved in disc herniations. The researchers concluded that to maximize posterior separation at the lower lumbar spine, traction should be applied with the hips flexed to 90 degrees.⁵ This aligns with everyday clinical practice: most motorized traction tables position patients supine with knees bent over a bolster, creating somewhere between 60 and 90 degrees of hip flexion.

Biomechanical modeling has also suggested that combining a global axial pull with local decompression at a specific segment is more effective at reducing both intradiscal pressure and tensile stress on the annulus fibrosus than either approach alone.⁶ This two-step concept underpins some newer traction protocols that first elongate the spine globally, then apply a focused decompressive load at the symptomatic level.

Static Versus Intermittent Traction

Traction can be delivered as a continuous, sustained pull (static) or as a rhythmic cycle of pulling and releasing (intermittent). The intuitive assumption is that intermittent traction, with its pumping action, would produce different muscular and pain responses than a steady hold. But when researchers compared the two modes head to head, measuring both muscle electrical activity and patient-reported pain, they found no significant difference in paraspinal muscle activity between static and intermittent traction. More than half of patients in both groups reported decreased pain afterward.⁷ That finding suggests the choice between the two modes may matter less than people assume, at least for short-term pain relief.

Where intermittent traction does get more nuanced is in the hold-and-rest timing. When the pull is cycled, clinicians must decide how long the traction force stays on (the hold phase) and how long it drops off (the rest phase). A comparative study tested four different combinations: 20 seconds on with 1 second off, 40 seconds on with 5 seconds off, 60 seconds on with 10 seconds off, and 80 seconds on with 20 seconds off. The groups with longer hold and rest times showed greater improvements in lumbar range of motion, disability scores, and pain ratings.⁸ The 60-second hold with 10-second rest and the 80-second hold with 20-second rest were statistically superior to the shorter cycles. A systematic review with meta-analysis identified this same study as the available evidence on hold-rest timing for intermittent traction.⁹

This makes mechanical sense. Very short hold times may not allow enough tissue creep for the vertebrae to separate meaningfully before the force releases. The spine’s soft tissues are viscoelastic, meaning they deform gradually under sustained load. A 20-second hold may simply not be long enough for the disc and ligaments to respond, whereas 60 or 80 seconds gives the tissues time to elongate.

Session Duration

Most clinical traction protocols run between 10 and 30 minutes per session. The real-time MRI study mentioned earlier provides a useful window into why that range matters: the herniated disc volume kept decreasing through the full 30-minute session, and the reduction roughly doubled between the 10-minute and 30-minute marks.¹⁰ That progressive curve suggests that sessions under 15 minutes may be cutting the treatment short before the greatest structural changes occur.

On the other hand, there is a ceiling. The intervertebral separation achievable with traction is relatively small, generally in the range of 0.4 to 2.0 millimeters per segment, depending on the force and duration. This separation comes from the creep behavior of spinal soft tissues, and after a certain point the tissues resist further elongation. Extending sessions well beyond 30 minutes is unlikely to add much benefit and increases the chance of muscle soreness or reactive spasm afterward. Most published protocols settle on 15 to 30 minutes per session, delivered two to five times per week over a period of several weeks.

Does It Actually Work? What the Clinical Trials Say

The evidence for lumbar traction effectiveness depends entirely on who is being treated and what you are comparing the traction against. Two well-known randomized trials in patients with nonspecific low back pain (meaning no identifiable structural cause like a disc herniation or nerve root compression) found no benefit from traction over sham or usual care at 12 weeks or 6 months.¹¹ This is a critical distinction: if there is no disc bulge or nerve irritation to decompress, pulling on the spine has no clear mechanism to help.

In patients with lumbar radiculopathy, which involves a nerve root being compressed, the picture is more encouraging but far from conclusive. A systematic review focused on traction for radiculopathy found that when traction was added to bed rest or medication alone, the effects on pain were large. But when traction was added to a comprehensive physical therapy program, the additional benefit shrank to nearly zero.¹² A separate meta-analysis found that supine mechanical traction added to physical therapy did show moderate effects on pain and disability, and those effects persisted at short-term follow-up of up to three months. However, when the analysis was restricted to high-quality studies only, the results for prone traction were no longer significant.¹³

A meta-analysis specifically examining mechanical traction for lumbar disc herniation found that the traction group had lower pain scores and better disability scores than conventional physical therapy alone, though lumbar flexibility as measured by a clinical test did not differ between groups.¹⁴ Meanwhile, imaging studies have confirmed that traction does physically reduce herniated disc volume and separate the disc from the adjacent nerve root during the session.¹⁵ A clinical trial using segmental traction similarly showed significant reductions in both herniation size and patient pain.¹⁶

The overall pattern is that traction shows the most consistent benefit in patients who have a confirmed disc herniation or radiculopathy, and offers little when applied broadly to all comers with back pain. The parameters matter, but patient selection may matter more.

Who Responds Best to Lumbar Traction

Researchers have tried to build prediction models identifying which patients are likeliest to improve with traction. One clinical prediction rule identified four factors associated with short-term improvement: not being involved in manual labor, having low fear-avoidance beliefs (basically, not being highly anxious about movement), absence of neurological deficits, and being over age 30. When all four factors were present, the probability of responding to traction jumped from about 19% to roughly 69%.¹⁷

A second prediction rule, developed specifically in patients with lumbar disc herniation, arrived at a partially overlapping but distinct set of factors: limited lumbar extension range of motion, low fear-avoidance beliefs about work, no segmental hypomobility in the lumbar spine, short symptom duration, and sudden onset of symptoms. Having at least three of these five predictors roughly doubled the probability of meaningful improvement.¹⁸ Low fear-avoidance beliefs show up in both models, which is worth noting. Patients who are very fearful of movement may tense their trunk muscles during traction, counteracting the decompressive force.

Neither prediction rule has been validated in large external samples, so they are better thought of as clinical guides than definitive screening tools. But they reinforce the idea that traction is not a one-size-fits-all intervention. Choosing the right patient may be more important than perfecting the force or timing parameters.

Motorized Spinal Decompression Versus Standard Traction

Over the past two decades, computerized spinal decompression devices have entered the market, often marketed as a fundamentally different therapy from traditional motorized traction. These machines use proprietary algorithms that ramp force up and down in patterns designed to minimize muscular guarding. They come with substantially higher price tags for both the provider and the patient. The question is whether they actually produce better outcomes.

The short answer, based on the available head-to-head trials, is no. A randomized trial comparing conventional motorized traction to non-surgical spinal decompression with a computerized device found that both methods improved pain and disability in patients with lumbar disc herniation, with no significant difference between them.¹⁹ A more recent trial reached the same conclusion: the change in pain and disability scores did not differ between the traction and spinal decompression groups.²⁰ Both were better than conventional physiotherapy alone, but neither outperformed the other.

This is a useful finding if you are weighing treatment options, because computerized decompression sessions typically cost several times more than standard traction. If outcomes are equivalent, paying more for a brand-name algorithm is hard to justify unless the device offers specific positioning or force-curve features that a standard traction table cannot replicate for your particular condition.

Combining Traction With Other Treatments

Traction is rarely used as a standalone therapy. Most clinical protocols embed it within a broader rehabilitation program that includes exercises, manual therapy, and sometimes modalities like heat or electrical stimulation. This is consistent with the clinical trial evidence showing that traction’s added benefit on top of a full physical therapy program is smaller than its benefit when compared to rest or medication alone.²¹

One area with reasonable support is the combination of spinal decompression therapy and core stabilization exercises. A randomized trial found that patients who received both had greater reductions in pain and disability than those who did core exercises alone.²² The logic is straightforward: traction may temporarily unload the disc and relieve nerve compression, but without improved muscular support to maintain spinal stability, the disc re-loads as soon as you stand up. Core training addresses that gap.

Sequencing also matters in practice, though it has received less formal study. Many clinicians apply heat or electrical stimulation before traction to relax paraspinal muscles, reasoning that a tense trunk works against the decompressive force. Others schedule manual mobilization of stiff segments before traction so that the pull is distributed more evenly rather than concentrated at already-mobile levels. These sequencing choices are largely based on clinical reasoning and biomechanical principles rather than randomized trials.

Inversion Tables and Home Devices

Inversion tables are probably the most common form of home-based lumbar traction. By tilting you upside down or partially inverted, they use your own body weight to create a traction force through the spine. The concept is straightforward: gravity pulls your lower body away from your upper body, decompressing lumbar segments.²³ The problem is that you cannot control the force as precisely as a motorized table, and the load increases steeply as the inversion angle approaches vertical. Full inversion can generate traction forces well above 50% of body weight, which may be excessive for some patients and raises blood pressure in the head and eyes.

Other home devices include over-the-door harnesses, pneumatic decompression belts, and portable mechanical traction units. The evidence base for these is thin. Most of the clinical trial data on traction parameters comes from motorized tables in supervised clinical settings, where a therapist can adjust force, angle, and duration in real time. Translating those parameters to a home device that lacks those controls is inherently imprecise. If you choose home traction, starting with very low force, keeping sessions short, and stopping at the first sign of increased pain is prudent. The same parameter principles apply in concept, but the margin for error is smaller without professional supervision.

Safety Considerations and Who Should Avoid Traction

Lumbar traction is generally low-risk when applied within the commonly studied parameter ranges, but certain conditions make it unsafe. Spinal fractures, active infections of the spine, severe osteoporosis, spinal tumors, and unstable spondylolisthesis are widely recognized contraindications. Abdominal aortic aneurysm is another, because the harness compresses the abdomen. Pregnancy is typically listed as a contraindication as well.

Even in patients without these contraindications, temporary increases in pain can occur, especially if the force is too high or the session too long. The spine’s soft tissues can become irritated if they are stretched beyond their comfortable range. Most clinicians start at the lower end of the force spectrum and increase gradually over subsequent sessions, watching for adverse reactions. Patients with large central disc herniations require particular caution: while lateral herniations may respond well to traction-induced separation, some clinicians worry that axial pulling on a large central protrusion could theoretically worsen cauda equina symptoms, though documented cases of this are extremely rare.

A practical concern that gets less attention is what happens between sessions. The intervertebral separation achieved during traction is temporary. The 0.4 to 2.0 millimeters of space gained per segment begins closing as soon as the force is removed, and spinal tissues typically return to their baseline dimensions within 30 minutes to a few hours. This is why most protocols call for repeated sessions over several weeks, and why combining traction with exercises that strengthen the muscles supporting the spine is considered important for sustaining any benefit beyond the treatment table.