Deep Tendon (Muscle Stretch) Reflexes
The deep tendon reflex is the most objective and reproducible element of the motor examination. Despite its traditional name, the reflex is not generated by the tendon at all but by the brief stretch the tendon tap imposes on the muscle and its embedded stretch receptors; for this reason many neurologists prefer the more accurate term muscle stretch reflex. Because the response is involuntary and largely independent of the patient's cooperation, it provides a measure of the nervous system that effort and suggestibility cannot easily distort. A single reflex tests the integrity of a defined segmental arc—the spindle, its afferent fiber, the spinal segment, the motor neuron, and the muscle it supplies—and at the same time reflects the state of the descending pathways that hold that arc under tonic control. The skilled examiner reads each reflex twice: once as a test of a particular root and nerve, and once as a window onto the corticospinal system that governs the whole segment.
The Reflex Arc
The muscle stretch reflex is the clinical expression of the simplest circuit in the nervous system: a monosynaptic arc with a single synapse interposed between sensory input and motor output. A sharp tap on the tendon imposes a sudden stretch on the muscle. This stretch is detected by the muscle spindle, an encapsulated receptor lying in parallel with the extrafusal muscle fibers. Stretching the spindle excites its primary sensory endings, which give rise to large, fast-conducting group Ia afferent fibers. These afferents enter the spinal cord through the dorsal root and project directly—without an interposed interneuron—onto the alpha motor neurons in the anterior horn that supply the same (homonymous) muscle. Activation of these motor neurons produces a brief synchronous contraction, felt and seen as the reflex twitch. By way of inhibitory interneurons, the same Ia volley simultaneously relaxes the antagonist muscle, a phenomenon termed reciprocal inhibition.
Although the arc itself is monosynaptic, its excitability is continuously adjusted by the rest of the nervous system. The gamma motor system supplies the contractile poles of the intrafusal fibers within the spindle; by setting the tension on the spindle, gamma motor neuron activity establishes the sensitivity, or gain, of the stretch reflex. Gamma motor neuron output is in turn governed by descending pathways. The net effect of the descending motor control—corticospinal and brainstem pathways acting on both the alpha and gamma motor neurons and on segmental interneurons—is predominantly inhibitory, restraining the reflex and keeping it within a normal range. This dual dependence is what gives the reflex its localizing power. Eliciting a deep tendon reflex therefore tests two things at once: the structural integrity of the segmental arc (spindle, Ia afferent, root, motor neuron, peripheral nerve, neuromuscular junction, and muscle), and the state of descending corticospinal control over that segment. Interruption of the arc abolishes or diminishes the reflex; release of the arc from descending inhibition, as in an upper motor neuron lesion, exaggerates it.
How to Elicit a Deep Tendon Reflex
A poorly elicited reflex is the commonest reason for a falsely absent or falsely asymmetric finding. The technique matters as much as the interpretation.
- Position the patient and ensure relaxation. The patient must be comfortable, warm, and at ease, with the limb supported so that the muscle to be tested is neither fully shortened nor fully lengthened but held in a mid-range, slightly stretched position. A tense or actively contracting patient will suppress the response; active contraction of the muscle being tested can suppress or obscure the reflex and make the response unreliable.
- Strike the tendon, not the muscle belly. The stimulus must reach the spindle by way of a sudden stretch of the tendon. Striking the muscle belly directly produces a local idiomuscular contraction rather than a true reflex and gives misleading results.
- Use a proper reflex hammer with a pendular wrist action. A hammer of adequate weight (such as the Queen Square or Tromner pattern) should be allowed to fall under its own momentum with a loose, pendular movement of the wrist, not driven by a stiff forearm. The blow should be brisk and of consistent, moderate force.
- Compare the two sides directly. Each reflex should be elicited identically on both sides and the responses compared. Asymmetry is detected far more reliably than absolute amplitude, and side-to-side comparison is the single most valuable habit in reflex testing.
- Optimize the stretch ("setting"). A reflex that is difficult to obtain can often be brought out by slightly altering the degree of partial muscle stretch—palpating the tendon and adjusting limb position so the muscle is optimally pre-stretched ("setting" the reflex) before the tap. Only after such optimization, and after reinforcement, should a reflex be called absent.
Grading: The 0 to 4+ Scale
Reflexes are conventionally graded on an ordinal scale from 0 to 4+. The scale is semiquantitative and observer-dependent, and its chief value lies not in the absolute number but in the pattern across the body.
| Grade | Response | Interpretation |
|---|---|---|
| 0 | Absent — no response even with reinforcement | Areflexia; suggests interruption of the reflex arc |
| 1+ | Diminished — present only with reinforcement, or trace | Hyporeflexia; may be normal or reflect a partial arc lesion |
| 2+ | Normal — easily obtained, expected briskness | Normal |
| 3+ | Brisk — more vigorous than average; possible spread | May be normal, or an early sign of hyperreflexia |
| 4+ | Very brisk with clonus — repetitive beats | Pathologic; indicates upper motor neuron hyperexcitability |
Two interpretive principles must temper any reflex grade. First, the absolute level of reflex activity varies widely among healthy individuals: many normal people have uniformly brisk (3+) reflexes, and a few have reflexes that are difficult to obtain throughout, so that briskness alone—when symmetric and unaccompanied by other signs—is of limited significance. Second, and more important, asymmetry and the company a reflex keeps are far more informative than its absolute magnitude. A reflex that is unequivocally brisker on one side than the other, or a 3+ reflex accompanied by clonus, spread, and an extensor plantar response, carries weight that an isolated grade does not. The reflexes should always be interpreted together with tone, strength, the plantar responses, and the rest of the examination.
The Principal Muscle Stretch Reflexes
The reflexes routinely tested, with their segmental innervation and peripheral nerve, are summarized below. The grading of any single reflex is meaningful only in the context of the segmental level it represents.
| Reflex | Spinal root(s) | Peripheral nerve | Notes |
|---|---|---|---|
| Jaw jerk | Pontine (cranial nerve V) | Trigeminal nerve (afferent and efferent) | Tap on the chin with the mouth slightly open. Normally absent or only slight; exaggerated in bilateral supranuclear (corticobulbar) lesions, as in pseudobulbar palsy and motor neuron disease. |
| Biceps | C5–C6 | Musculocutaneous nerve | Tap the examiner's thumb placed on the biceps tendon; elbow flexion. |
| Brachioradialis ("supinator") | C5–C6 | Radial nerve | Tap the distal radius; elbow flexion with some forearm movement. The "inverted" form (see below) localizes to C5–C6. |
| Triceps | C7–C8 | Radial nerve | Tap the triceps tendon just above the olecranon; elbow extension. |
| Finger flexors | C8–T1 | Median and ulnar nerves | Flexion of the fingers on tapping the volar surface or the examiner's fingers; closely related to the Hoffmann sign of hyperreflexia. |
| Pectoralis | C5–T1 | Medial and lateral pectoral nerves | Tap a finger over the pectoralis tendon near the anterior axillary fold; adduction/internal rotation of the arm. |
| Knee (patellar) | L2–L4 (chiefly L3–L4) | Femoral nerve | Tap the patellar tendon with the knee flexed; quadriceps contraction and knee extension. |
| Medial hamstring | L5–S1 | Sciatic nerve | Tap the medial hamstring tendon behind the knee; useful for assessing the L5–S1 level. |
| Ankle (Achilles) | S1–S2 | Tibial nerve | Tap the Achilles tendon with the ankle slightly dorsiflexed; plantar flexion. Often the first reflex lost in peripheral neuropathy and with age. |
Reinforcement
When a reflex appears reduced or absent, it must not be recorded as such until the response has been sought with reinforcement. Reinforcement maneuvers transiently augment the stretch reflex by drawing the patient's attention away from the limb being tested and, in doing so, reducing the descending inhibition that normally restrains the segmental arc. The reflex is "released" briefly and becomes easier to elicit.
- The Jendrassik maneuver is used for the lower-limb reflexes. The patient hooks the flexed fingers of the two hands together and, on command, pulls them apart forcefully at the moment the tendon is tapped. The reflex is timed to coincide with the effort.
- Teeth-clenching or fist-clenching is used to reinforce the upper-limb reflexes. The patient is asked to clench the jaw or to make a fist with the opposite hand at the instant of the tap.
The common mechanism is distraction of descending inhibitory control, which transiently raises the gain of the reflex arc. A reflex that is obtainable only with reinforcement is graded 1+; a reflex that remains absent despite a correctly performed reinforcement maneuver may be recorded as truly absent (0).
Clonus
Clonus is a series of rhythmic, involuntary muscle contractions provoked by a sudden and then sustained stretch of the muscle. It is most reliably demonstrated at the ankle, where the examiner briskly dorsiflexes the relaxed foot and maintains the stretch, eliciting repetitive beats of plantar flexion, and it may also be elicited at the patella by a sharp downward push on the patella held under tension. Clonus arises from the same disinhibited, hyperexcitable stretch reflex that produces spasticity and hyperreflexia. A few unsustained beats may occur in anxious individuals with otherwise brisk normal reflexes and are not necessarily pathologic. Sustained clonus—beats that continue for as long as the stretch is maintained—is unequivocally abnormal and signifies upper motor neuron hyperexcitability of the corticospinal pathway controlling that segment.
Upper versus Lower Motor Neuron Reflex Patterns
The behavior of the muscle stretch reflexes is one of the principal means of separating an upper motor neuron (corticospinal) lesion from a lower motor neuron (anterior horn cell, root, plexus, nerve, junction, or muscle) lesion, because the two interrupt the reflex circuit in opposite ways.
- Upper motor neuron lesion. Loss of descending inhibition releases the segmental arc, producing hyperreflexia, clonus, reflex spread, and a loss of the normal restraint on tone (spasticity), typically accompanied by an extensor plantar (Babinski) response. The reflexes below the level of the lesion are exaggerated.
- Lower motor neuron lesion. Interruption of any part of the reflex arc itself produces hyporeflexia or areflexia, along with flaccid weakness, hypotonia, wasting, and fasciculations. The reflex corresponding to the damaged segment or nerve is diminished or lost.
An important qualification is the acute phase of an upper motor neuron lesion. In the first hours to days after an acute corticospinal injury (spinal shock after cord injury, cerebral shock after a large acute hemispheric lesion), tone and reflexes are paradoxically depressed before the expected hyperreflexia and spasticity emerge over the following days to weeks.
Variants and Special Reflex Phenomena
Beyond simple grading, several distinctive reflex phenomena carry specific localizing or diagnostic value. The most useful is the inverted reflex: tapping a tendon fails to evoke its expected response and instead produces a movement appropriate to a lower segment, because the tapped segment's arc is interrupted while the segments below it have been released into hyperactivity. The principal phenomena and their meaning:
- Inverted supinator (brachioradialis) reflex — a brachioradialis tap produces no elbow flexion but reflex finger flexion → lesion at C5–C6 with hyperreflexia below; a classic sign of compressive cervical myelopathy or radiculopathy.
- Inverted knee jerk — absent quadriceps contraction with reflex knee flexion (hamstrings) → lesion at L2–L4 with release of the segments below.
- Pendular reflex — the limb swings back and forth several times before settling (reduced damping) → cerebellar disease or hypotonia; best seen at the knee with the legs hanging freely.
- Hung-up reflex (Woltman sign) — normal contraction but a markedly slow relaxation phase, seen at the ankle → the classic reflex sign of hypothyroidism.
- Reflex spread / crossed adductor reflex — the response irradiates beyond the tapped muscle (e.g., a patellar tap also contracting the contralateral adductors) → upper motor neuron hyperexcitability.
- Paradoxical triceps reflex — a triceps tap produces elbow flexion rather than extension → a segmental lesion at C7 (the C7 arc is interrupted while flexor segments above are released).
Clinical Note
The inverted supinator reflex is a valuable bedside localizer. When a tap on the brachioradialis tendon produces no elbow flexion but instead evokes reflex finger flexion, the lesion is pinpointed to C5–C6: the brachioradialis arc at that level is interrupted, while the reflexes below are released into hyperactivity. The sign is most often encountered in compressive cervical myelopathy or radiculopathy and tells the examiner both the level of the lesion and that there is upper motor neuron involvement below it.
References
- Campbell WW, Barohn RJ. DeJong's The Neurologic Examination. 8th ed. Philadelphia: Wolters Kluwer; 2020.
- Ropper AH, Samuels MA, Klein JP, Prasad S. Adams and Victor's Principles of Neurology. 11th ed. New York: McGraw-Hill; 2019.
- Wartenberg R. The Examination of Reflexes: A Simplification. Chicago: Year Book Publishers; 1945.
- Walker HK. Deep Tendon Reflexes. In: Walker HK, Hall WD, Hurst JW, eds. Clinical Methods: The History, Physical, and Laboratory Examinations. 3rd ed. Boston: Butterworths; 1990.