By Dr Steven Hewitt — Chiropractor · AHPRA: CHI0001115420 · 8 April 2026When a joint is manipulated, the sound is memorable. The audible crack — the rapid pop that marks a successful high-velocity thrust — is the thing most people associate with the treatment, and often the thing they attribute the benefit to. The research suggests they are partly right about the sound mattering, and almost entirely wrong about why.
Over the past two decades, a body of electrophysiology, neuroimaging, and clinical research has built a reasonably detailed picture of what actually happens when a joint is manipulated. It is not a mechanical story — joints returned to their correct position, bones realigned, discs rehydrated. It is a neurophysiological one: a precisely applied mechanical stimulus that triggers a cascade of effects through the peripheral nervous system, the spinal cord, and the brain's own pain-modulating circuitry. The clinical outcomes — reduced pain, restored movement, improved muscle activation — are downstream consequences of that cascade.
The Mechanical Explanation and Why It Falls Short
The earliest explanation for spinal manipulation was almost entirely structural. The idea was that joints become restricted or subluxated — subtly displaced or fixated — and manipulation was the means of restoring their normal position. On this model, the audible pop was the sound of the joint returning to where it should be, and the therapeutic effect was the mechanical correction.
This model had practical appeal: it was simple, it gave the intervention a clear target, and it mapped well onto a patient's subjective experience that something had "shifted." The problem is that it does not hold up well under scrutiny. Advanced imaging studies consistently show that spinal manipulation does not measurably alter vertebral position, and clinical outcomes — pain reduction, functional improvement — occur through mechanisms that are not well explained by the movement of bony structures alone.
A 2022 ultrasound study by Vining and colleagues provided a clear illustration of this. They measured thoracolumbar fascia shear strain in patients before and after spinal manipulation and found that manipulation did not immediately alter fascial shear strain. [1] Yet patients showed significant improvements in pain and disability outcomes over eight weeks of care. If manipulation were acting primarily through mechanical tissue correction, we would expect immediate local changes in tissue mechanics. Their absence — alongside clear clinical improvement — points toward an effect operating at a different level of the system entirely.
That level is the nervous system.
The Neurophysiological Model: What the Research Reveals
In 2009, Bialosky and colleagues published a comprehensive theoretical model of the mechanisms of manual therapy that has since become one of the most-cited frameworks in the field. [2] The model proposes a hierarchical cascade: the mechanical stimulus of manipulation produces neurophysiological effects at three progressively higher levels of the nervous system — peripheral, spinal, and supraspinal — and the clinical outcomes are the downstream product of this cascade.
The Peripheral Level: Firing the Proprioceptors
When a joint is manipulated — when the high-velocity thrust is delivered — the rapid displacement of the joint and its surrounding tissues activates mechanoreceptors at the treated site. Specifically, it fires the Group I and Group II afferents from muscle spindles and Golgi tendon organs: the large-diameter, fast-conducting fibres responsible for proprioception and detecting changes in muscle length and tension.
The review by Pickar confirmed this in electrophysiological detail. [3] Spinal manipulation produces a rapid barrage of proprioceptive afferent input to the central nervous system — from the paraspinal muscles, the joint capsule, and the surrounding ligaments. This is not incidental. It is the primary mechanical event of the thrust, translated into neurological signal.
The parameters of the thrust matter. Lima and colleagues, using an instrumented actuator to deliver manipulation-like stimuli while recording spindle afferent discharge in an animal model, found that a thrust at approximately 55% body weight produced the most consistent spindle activation, and that both the magnitude and duration of the thrust were significantly associated with afferent discharge frequency. [4] The neurological response to manipulation is dose-dependent: technique is not interchangeable regardless of force or tempo.
What this means clinically is that a joint manipulation is, from the nervous system's perspective, a burst of proprioceptive input to the spinal cord — a signal carrying information about the mechanical state of the segment, the tension in the surrounding muscles, and the position of the joint. This signal arrives at the dorsal horn of the spinal cord, where its effects propagate upward.
The Spinal Level: Gate Control and Central Sensitisation
At the level of the spinal cord, the barrage of large-diameter afferent input from manipulation has two well-documented effects.
The first is gate control: mechanoreceptor activity from Group I and II fibres competes with and suppresses nociceptive signals arriving via Group III and IV small-diameter fibres. This is the gate control mechanism first proposed by Melzack and Wall — proprioceptive input can physically reduce pain transmission at the dorsal horn. Millan and colleagues' systematic review of 22 studies examining spinal manipulative therapy and experimentally induced pain found consistent evidence of hypoalgesia — a measurable reduction in pain sensitivity — at and near the treated site. [5]
The second effect is on central sensitisation: the process by which persistent nociceptive input causes the dorsal horn to become progressively more sensitised, lowering its threshold and amplifying pain beyond what the peripheral stimulus alone would produce. In people with chronic pain, central sensitisation is a primary driver of disproportionate pain, widespread tenderness, and the persistence of symptoms long after the original tissue injury has resolved.
The proprioceptive barrage from manipulation appears to interrupt or reduce this sensitisation. Bond and colleagues, in a clinical study of patients with chronic non-specific low back pain, found significant improvements in pressure pain threshold at both the treated spinal segments and at remote sites following spinal manipulation — a pattern consistent with centrally mediated pain modulation rather than purely local mechanical effects. [6] The clinical consequence of this spinal-level action is that manipulation can produce pain relief extending well beyond the treated segment.
The Supraspinal Level: The Brain's Own Analgesic System
At the highest level of the Bialosky model sits the supraspinal effect — manipulation activating the brain's endogenous descending inhibition system.
The key structures here are the periaqueductal grey (PAG) and the rostral ventromedial medulla (RVM) — a brainstem circuit that, when activated, projects descending inhibitory signals back down the spinal cord, reducing pain transmission at the dorsal horn. Activation of this system via the PAG-RVM axis is the mechanism by which opioid analgesia works — and the same pathway can be engaged by an appropriate non-pharmacological stimulus.
Younes and colleagues provided direct autonomic evidence of supraspinal engagement. In a randomised controlled trial comparing spinal manipulative therapy to sham manipulation in patients with acute low back pain, manipulation produced significant increases in parasympathetic heart rate variability indices — specifically HF power and RMSSD — compared to the sham group. [7] Heart rate variability is a window into autonomic nervous system regulation mediated at the brainstem level. The fact that a lumbar manipulation produced measurable changes in cardiac autonomic tone suggests the technique's effect is reaching well above the spinal cord, into the brainstem and higher.
The clinical significance of supraspinal engagement is that manipulation is not simply treating a painful segment in isolation. It is activating the brain's own machinery for regulating pain perception across the body — and that machinery, once activated, does not confine its effects to one lumbar level.
The Audible Pop: More Than a Sound
The audible joint release — the crack that patients often associate with a successful manipulation — does appear to carry its own specific neurological effect.
Clark and colleagues studied patients with chronic low back pain and asymptomatic controls, measuring corticospinal and spinal stretch reflex excitability before and after a single high-velocity lumbar manipulation. [8] The manipulation did not produce systematic changes in corticospinal excitability across the group as a whole. However, in subjects who experienced an audible joint response — cavitation — there was a 20% decrease in stretch reflex amplitude of the erector spinae (p<0.05). This finding was specific to cavitation: the neurological response was measurably different when the joint produced an audible release compared to when it did not.
The inference is that cavitation produces a particularly intense burst of afferent input — the rapid joint distraction associated with the audible release generating a high-amplitude proprioceptive signal — that selectively modulates the sensitivity of the Ia reflex pathway. Whether elevated stretch reflex sensitivity is a driver of local muscle guarding in chronic spinal pain is a reasonable clinical question, and the finding that cavitation specifically attenuates it provides a neurological rationale for the symptomatic improvement that often follows an audible release.
Arthrogenic Muscle Inhibition: When a Joint Problem Becomes a Muscle Problem
One of the most clinically significant — and most underappreciated — neurophysiological consequences of joint dysfunction is its direct inhibitory effect on the surrounding musculature. This phenomenon, called arthrogenic muscle inhibition (AMI), has now been documented across multiple joints and represents a mechanistic link between a dysfunctional joint and what appears, on the surface, to be muscular weakness.
Rice and McNair described AMI in a foundational 2010 review as a persistent reduction in voluntary muscle activation caused by aberrant afferent input from a dysfunctional or injured joint — distinct from, and often greater than, any weakness from disuse or atrophy. [9] The mechanism operates through two pathways. Joint swelling or structural change activates Type II mechanoreceptors, producing presynaptic inhibition of the alpha-motoneuron pools via Ib interneurons. Simultaneously, pain activates Type III and IV nociceptors, producing reflex inhibition at the spinal cord and suppression of motor output at the cortical level — a supraspinal contribution confirmed using TMS-evoked MEP measurements. [10] The combined effect is that a joint problem can substantially reduce the capacity of surrounding muscles to generate voluntary force through purely neural mechanisms, with no atrophy required.
The clinical implications extend well beyond the knee — where AMI was first extensively characterised following injury and surgery. Freeman and colleagues demonstrated AMI at the hip: experimentally induced hip joint pain produced significant reductions in gluteus medius activation amplitude during walking, measured by EMG. [11] Stackhouse and colleagues demonstrated it at the shoulder: experimental shoulder pain inhibited infraspinatus activation during maximal isometric external rotation, with direct implications for understanding rotator cuff weakness in painful shoulders. [12] McVey and colleagues demonstrated it at the ankle: chronic functional ankle instability was associated with measurably reduced soleus motoneuron pool excitability compared to healthy controls — a finding with clear implications for chronic re-injury risk. [13]
The pattern that emerges across these studies is consistent: a painful or dysfunctional joint produces aberrant afferent input; that input travels to the spinal cord; the spinal cord responds by reducing motor output to the muscles around the joint. What clinicians observe as muscle weakness in a painful region may not require a separate explanation at all. The joint and the muscular deficit are frequently the same problem, expressed at two different levels of the nervous system.
The evidence that restoring normal joint afferent input — through manipulation — can reverse this suppression is a critical part of the case for manipulation as a therapeutic tool. Bialosky's 2009 review cited randomised controlled trial evidence from Suter and colleagues showing that sacroiliac and lumbar manipulation reduced knee-extensor inhibition, and that cervical manipulation reduced elbow flexor inhibition — AMI reversal at sites remote from the manipulation itself. [2] The mechanism is the normalisation of aberrant afferent input, reducing the inhibitory drive at the spinal cord level.
In the most comprehensive recent overview, Alanazi and colleagues synthesised 136 peer-reviewed studies on the direct neuromuscular responses to high-velocity low-amplitude spinal manipulation. [14] The documented patterns are consistent with this model: cervical manipulation produces bilateral biceps brachii EMG activation; thoracic manipulation increases shoulder muscle EMG in patients with subacromial pain syndrome; sacroiliac manipulation activates bilateral transversus abdominis and internal oblique. The manipulation produces measurable changes in muscle activation — not just at the treated segment, but at the musculature those segments neurologically influence.
Why This Changes How We Think About Manual Therapy
The neurophysiological model has significant implications for clinical practice.
The mechanical framing — treating a restriction, correcting a subluxation, restoring segmental motion — is not wrong as a clinical description of where the technique is applied and what the physical event is. But it is incomplete as an explanation of why it works. The research now provides a more accurate account: a precisely applied mechanical stimulus delivers a burst of proprioceptive input that, through peripheral, spinal, and supraspinal pathways, modulates pain sensitivity, reduces central sensitisation, activates descending inhibition, and removes the aberrant afferent signal that was suppressing motor output in the surrounding musculature.
This is why manipulation produces outcomes that are not explicable by movement of bony structures alone: pain relief at sites remote from the treated segment, restoration of muscle activation in muscles that appear intact on imaging, and autonomic changes reflecting brainstem engagement. The joint is the input device. The effect is system-wide.
This understanding also reframes rehabilitation. If arthrogenic muscle inhibition is a genuine mechanism — if the weakness around a painful joint is partly driven by aberrant neural signalling from that joint — then strengthening exercises aimed at those muscles may be working against an active inhibitory drive. Addressing the joint source of that drive may be a prerequisite for the surrounding muscles to respond normally to load. This is not a case against rehabilitation. It is a case for sequencing: removing the inhibitory signal first, then loading the muscle that was inhibited.
At Elevate Health, we view this research as closely aligned with the fascial approach. The Stecco Fascial Manipulation method works at specific points in the deep fascia where multiple vectorial forces converge — anatomical sites that are densely populated with mechanoreceptors and in intimate relationship with the proprioceptive structures the manipulation literature has now mapped in detail. Working those points generates precisely targeted afferent input, at locations selected through systematic clinical reasoning, to the same nervous system circuits described in this research. The tissue biology and the neurophysiology are not separate stories — they are the same story, examined from different angles.
What This Means for You
If you have been told manipulation only produces temporary relief, the research does not support that framing as a complete account. The mechanisms documented — central sensitisation reduction, descending inhibition activation, AMI reversal — are neurophysiological processes with documented clinical correlates in pressure pain threshold, muscle activation, and functional improvement.
If you have joint pain accompanied by muscle weakness that has not responded to exercise, arthrogenic muscle inhibition is worth considering as a contributor. The weakness may be driven by aberrant afferent input from the joint itself, and addressing the joint may be a prerequisite for the muscles to respond normally to rehabilitation loading. You can read more about this in the context of specific conditions: rotator cuff weakness, gluteal inhibition, and chronic ankle instability are all presentations where AMI is likely to be a contributing factor.
If you have chronic pain that has persisted long after an injury appears resolved, central sensitisation is a likely contributor to the persistence. The spinal and supraspinal effects of manipulation — including activation of the PAG-RVM descending inhibition system — operate specifically on the central mechanisms that perpetuate chronic pain. This is not a claim that manipulation resolves all chronic pain; it is a statement that the documented mechanisms are directly relevant to its neuroscience.
If you are sceptical that a physical technique can produce effects as far-reaching as the research describes, the autonomic evidence is worth sitting with. A measurable change in cardiac parasympathetic tone — documented in a controlled trial with a sham comparison — following a lumbar manipulation is not a local mechanical effect. It is evidence of brainstem engagement. The nervous system does not respect the anatomical boundaries that a purely structural model assumes.
Want to understand whether a neurophysiological assessment applies to your situation?
Call Now — speak with our team
Book Online — available 24/7
References
- PubMed Vining R, Onifer SM, Twist E, Ziegler AM, Corber L, Long CR (2022). Feasibility and pilot study of objectively quantifying the effects of chiropractic care on thoracolumbar fascia shear strain using ultrasound. Chiropractic & Manual Therapies, 30:34.
- PubMed Bialosky JE, Bishop MD, Price DD, Robinson ME, George SZ (2009). The mechanisms of manual therapy in the treatment of musculoskeletal pain: a comprehensive model. Manual Therapy, 14(5):531–538.
- PubMed Pickar JG (2002). Neurophysiological effects of spinal manipulation. The Spine Journal, 2(5):357–371.
- PubMed Lima CR, Sozio RS, Law AC, Nelson AJ, Singh H, Li P, Reed WR (2021). Paraspinal muscle spindle responses to a high-velocity, low-amplitude spinal manipulative stimulus: comparison of thrust magnitudes and durations. Journal of Manipulative and Physiological Therapeutics, 44(5):363–371.
- PubMed Millan M, Leboeuf-Yde C, Budgell B, Amorim MA (2012). The effect of spinal manipulative therapy on experimentally induced pain: a systematic literature review. Chiropractic & Manual Therapies, 20:26.
- PubMed Bond BM, Kinslow CD, Yoder AW, Liu W (2020). Effect of spinal manipulative therapy on mechanical pain sensitivity in patients with chronic nonspecific low back pain: a pilot randomized, controlled trial. Journal of Manual & Manipulative Therapy, 28(1), 15–27.
- PubMed Younes M, Nowakowski K, Didier-Laurent B, Gombert M, Cottin F (2017). Effect of spinal manipulative treatment on cardiovascular autonomic control in patients with acute low back pain. Chiropractic & Manual Therapies, 25:33.
- PubMed Clark BC, Goss DA Jr, Walkowski S, Hoffman RL, Ross A, Thomas JS (2011). Neurophysiologic effects of spinal manipulation in patients with chronic low back pain. BMC Musculoskeletal Disorders, 12:170.
- PubMed Rice DA, McNair PJ (2010). Quadriceps arthrogenic muscle inhibition: neural mechanisms and treatment perspectives. Seminars in Arthritis and Rheumatism, 40(3):250–266.
- PubMed Rice DA, McNair PJ, Lewis GN, Dalbeth N (2014). Quadriceps arthrogenic muscle inhibition: the effects of experimental knee joint effusion on motor cortex excitability. Arthritis Research & Therapy, 16(6), 502.
- PubMed Freeman S, Mascia A, McGill S (2013). Arthrogenic neuromusculature inhibition: a foundational investigation of existence in the hip joint. Clinical Biomechanics, 28(2):171–177.
- PubMed Stackhouse SK, Eisennagel A, Eisennagel J, Lenker H, Sweitzer BA, McClure PW (2013). Experimental pain inhibits infraspinatus activation during isometric external rotation. Journal of Shoulder and Elbow Surgery, 22(4), 478–484.
- PubMed McVey ED, Palmieri RM, Docherty CL, Zinder SM, Ingersoll CD (2005). Arthrogenic muscle inhibition in the leg muscles of subjects exhibiting functional ankle instability. Foot & Ankle International, 26(12):1055–1061.
- PubMed Alanazi MS, Degenhardt B, Kelley-Franklin G, Cox JM, Lipke L, Reed WR (2025). Neuromuscular effects of high-velocity, low-amplitude spinal manipulation: a narrative overview. Medicina, 61:187.
Please note: This post is intended for educational purposes only and does not constitute clinical advice. Individual presentations vary significantly. Please consult a registered health practitioner for advice about your specific condition.