When two or more vertebrae are fused, that segment stops moving. The motion previously distributed across the whole lumbar spine must be redistributed, and the levels immediately above and below the fusion absorb greater mechanical load than before. This is not a random complication or a technical error: it is an expected mechanical consequence of eliminating motion in a system designed to share it. (Park, 2004; Hilibrand, 2004)

There is one distinction that is the key to this entire topic, and the source of most wrong decisions:

Imaging finding

Adjacent segment degeneration (ASDeg) — changes visible on imaging. The patient may be entirely asymptomatic.

Clinical entity

Adjacent segment disease (ASDis) — when those changes produce correlated symptoms: radiculopathy, claudication, symptomatic instability.

Why it matters

They are distinct entities. Confusing them leads to operating on images rather than patients — and each additional fusion creates a new adjacent segment.

The mechanism, specifically

Biomechanical studies — including cadaveric and finite element models — consistently show what happens after arthrodesis: intradiscal pressure rises at the adjacent segment, measured directly in cadavers above an instrumented fusion; mobility and facet forces increase at unfused levels; and both construct stiffness and fusion length modify that load transfer. (Weinhoffer, 1995; Lee & Langrana, 1984; Nagata, 1993)

Pre-existing degeneration of the neighbouring disc by itself changes stress distribution after a single-level fusion. In other words: the state of the adjacent segment before surgery already conditions its fate. (Kim, 2015; Ruberté, 2009)

The temporal pattern is what most orients the diagnosis. The typical scenario: the patient had surgery, improved — sometimes excellently — and went through a symptom-free interval. Months or years later a new presentation appears, often unlike the original pain.

Most frequent presentations
  • aAdjacent segment stenosis — neurogenic claudication: leg pain, heaviness or numbness while walking, relieved by sitting or flexing. One of the best-documented causes of reoperation after lumbar fusion.
  • bNew radiculopathy — radiating pain following a nerve root different from the one originally treated. In published series it is defined as new radicular symptoms referable to the adjacent level.
  • cAdjacent instability — de novo or progressive spondylolisthesis above or below the fusion.
  • dAxial low back pain with functional decline, particularly where sagittal balance has been lost.
  • eInfra-adjacent segment disease — a less recognised pattern: sacroiliac joint dysfunction after lumbosacral fusion. It must be sought specifically, as it is mistaken for non-specific low back pain.
Key point

Pain that never improved points to residual compression or an incorrect initial diagnosis. Pain that improved and returned points to recurrence, pseudarthrosis or adjacent segment. That single question orients much of the diagnosis.

Diagnosis requires that imaging and clinical picture agree. Neither alone is sufficient.

Suspected problemStudy of choice
Neural compression / stenosisMRI (Pfirrmann grading for adjacent disc)
Pseudarthrosis of prior fusionCT — MRI is limited by metal artefact
InstabilityDynamic flexion-extension radiographs
Hardware integrityCT / plain radiographs
Sagittal alignmentFull-length standing radiographs (spinopelvic parameters)

Full-length standing radiographs are not optional. Spinopelvic parameters — pelvic incidence, lumbar lordosis and their mismatch — are consistently associated with the development of adjacent pathology, and cannot be measured on a supine lumbar MRI. (Phan, 2018; Schwab, 2010)

An honest caveat: radiographic assessment of the adjacent segment is notoriously difficult to standardise, and definitions vary between studies. This is why the incidence figures below carry such wide ranges. (Virk, 2014)

This is probably the most useful section, because several of these factors are knowable before surgery — and therefore can and should weigh on the decision to fuse.

Patient factors

Identifiable at the preoperative visit
  • aAdvanced age — the most consistently reported factor across systematic reviews and meta-analyses.
  • bElevated body mass index — together with age and osteoporosis, shown to be more predictive of reoperation than imaging findings themselves.
  • cOsteoporosis / low bone density — with both clinical and biomechanical evidence: finite element models show different stress patterns in osteoporotic bone.
  • dPre-existing adjacent segment degeneration — the most cited factor in the most recent systematic review. If the neighbouring disc is already degenerated preoperatively, risk rises substantially.
  • eRheumatoid arthritis — identified as a specific factor for ASD requiring surgery after short fusions.
  • fSmoking — with experimental evidence of altered gene expression in the intervertebral disc.
  • gGeneralised joint laxity.

Technical and surgical factors

Dependent on planning and technique
  • aLaminectomy adjacent to the fusion — one of the most consistent factors. Superior facet violation or injury to the interspinous ligament complex destabilises the neighbouring level.
  • bPelvic incidence–lumbar lordosis mismatch and sagittal malalignment, with direct implications for technique selection.
  • cLong or multilevel fusions — the more levels, the higher the incidence.
  • dFloating fusions (not reaching L5-S1) — higher risk of adjacent degeneration on long-term follow-up.
  • eExcessive disc space distraction during fusion.
  • fConstruct rigidity — rigid pedicle fixation transmits greater load to the adjacent segment.

Precision matters here, because the figures in circulation vary widely and that variability has a concrete methodological explanation: definitions are not homogeneous across studies. Any single figure quoted without specifying definition, technique and follow-up duration is a simplification.

2.6–27%
Require surgery for ASD (systematic review)
1,000
Patients in the largest PLIF series
>5
Years after which symptomatic cases still appear

The most recent systematic review reports that between 2.62% and 27.4% of patients ultimately require surgery for adjacent segment disease. That wide range reflects differences in definition, technique, population and follow-up duration. The Zhang meta-analysis is particularly valuable because it explicitly separates degeneration from disease — exactly the distinction that matters most clinically. (Zemánek, 2026; Zhang, 2016)

The temporal pattern

  • ASD is a progressive, time-dependent phenomenon: cumulative incidence keeps rising with follow-up; it does not plateau at two years.
  • Early- and late-onset forms have been described, with potentially distinct risk profiles.
  • Symptomatic adjacent degeneration has been documented more than five years after fusion, which mandates long follow-up to estimate true risk.
  • Radiological follow-up exists at 7.5 years for all unfused segments, and MRI assessment at 20 years after anterior interbody fusion.
Key point

Radiological incidence is high — many patients develop visible changes. The proportion requiring surgery is substantially lower. Conflating the two figures is the most common way to frighten a patient unnecessarily, or to justify an operation that is not needed.

This is the section I most want to be clear, because it is where the most harm is done.

Radiological degeneration of the adjacent segment is common. Symptomatic disease is far less common. The meta-analysis separating both entities documents this clearly, and the conceptual distinction has been established since the foundational work of Hilibrand and Robbins. (Zhang, 2016; Hilibrand, 2004)

Moreover — and this matters — adjacent degeneration does not always translate into worse clinical outcome. Specific work exists on the real impact of adjacent segment degeneration on clinical outcome after lumbar fusion, and systematic radiological assessment of unfused segments shows the magnitude of the purely radiographic phenomenon. (Yang, 2008; Pellisé, 2007)

Why this matters so much

A patient may present with an MRI showing degeneration at the level above their fusion, and with low back pain. That does not establish that the degeneration is the cause of the pain.

The costliest error

Operating on an imaging finding that does not correlate with the clinical picture is one of the most common ways of turning a patient with manageable pain into a patient with more surgeries, more scar tissue and more fused levels — each of which, in turn, generates a new adjacent segment.

The minimum requirement before proposing surgery: a correctable structural finding and that this finding coherently explains the symptoms the patient reports.

Surgery for adjacent segment disease has reasonably well-defined indications.

Surgery is considered when the following converge
  • 1Clinical-radiological correlation — the finding explains the symptoms.
  • 2Progressive neurological deficit, or radicular pain/claudication that significantly limits function.
  • 3Failure of properly conducted conservative treatment — targeted rehabilitation, pharmacological management, interventional procedures where indicated.
  • 4A concrete surgical target — demonstrated compression, demonstrated instability, or malalignment requiring correction.
Surgery is not indicated when
  • There are only radiological changes without clinical correlation.
  • Pain is predominantly axial without an identifiable generator.
  • Adequate conservative treatment has not been attempted.
  • Comorbidities make the risk exceed the expected benefit.

Where surgery is indicated, options include isolated decompression, extension of the fusion, or revision with alignment correction. The effectiveness of reoperations for ASD has been specifically reviewed, and the honest conclusion is that there is variability in clinical improvement and that comparative literature between techniques is limited. (Drysch, 2018)

Honestly: results are good in well-selected cases, but on average less predictable than those of a first operation.

  • Revision for ASD can significantly improve symptoms when there is a clear structural target and clinical correlation.
  • There is documented variability in the magnitude of improvement.
  • Posterior revision surgery carries its own complication profile, which must be discussed explicitly beforehand.
  • Overall reoperation rates after instrumented fusion are documented in large series.
Key point

Operating on the adjacent segment creates a new adjacent segment. If a fusion is extended from L4-L5 to L3-L5, L2-L3 now takes up the load. The problem is not eliminated: it moves one level up. This is why the right question is not merely «can I decompress this?», but «what is the ten-year strategy for this spine?».

Advanced age is simultaneously the most consistent risk factor for developing ASD and a variable that modifies the surgical decision. (Lau, 2021; Mesregah, 2022)

Specific considerations
  • aBone quality. Osteoporosis raises risk and compromises fixation. Age, BMI and osteoporosis proved more predictive of reoperation than imaging findings. Bone densitometry should be part of planning, not an optional extra.
  • bSagittal balance. In older adults, loss of lordosis and sagittal imbalance are more prevalent and weigh more heavily on outcome.
  • cComorbidity and magnitude of the procedure. In many older adults, a well-indicated decompression without extending the fusion offers the best risk-benefit ratio.
  • dFunctional expectations. The reasonable goal is usually to restore walking capacity and independence, not the complete absence of pain.
Key point

In this group, the decision not to extend a fusion is frequently the correct one.

If you take one idea from this article, let it be this: adjacent segment disease is strongly tied to fusion. It is not a random complication or a technical error: it is an expected mechanical consequence of eliminating motion in a spine designed to share it.

Four practical conclusions
  • 1It has risk factors identifiable before surgery. Age, bone quality, BMI, pre-existing degeneration of the neighbouring level, spinopelvic alignment, smoking, and technical decisions such as fusion length or adjacent laminectomy. Several are assessable at the preoperative visit.
  • 2It must be part of the conversation BEFORE deciding on fusion. Complete informed consent for a lumbar arthrodesis includes explaining that there is a real probability — varying with risk profile — of needing additional surgery at a different level in subsequent years. A patient who learns this only after their second operation is right to feel they were not told everything.
  • 3A fusion does not eliminate the risk of spinal reoperation. In some scenarios it displaces it in time and in level. Reoperation data from large series confirm this.
  • 4Avoid fusion when it is not absolutely indispensable. Fusion should be reserved for absolute indications — demonstrated instability, deformity requiring correction, or decompression that necessarily compromises stability. When a properly performed decompression solves the problem, that is the better operation.

Evidence on motion preservation and dynamic devices is promising in some scenarios but has not eliminated the problem, and total disc replacement does not fully abolish it either. (Kanayama, 2009; Wang, 2012; Kitzen, 2021)