CHAPTER 38

The Deep Tendon or Muscle Stretch Reflexes

When a normal muscle is passively stretched, its fibers resist the stretch by contracting. The stretch may be caused by gravity, manipulation, or other stimuli. In reflex responses, the contraction results from stimulation of the sensory organs in the muscle, either directly or indirectly through a stimulus applied to its tendons, the bone to which it is attached, or the overlying skin. In the monosynaptic stretch reflex, sudden lengthening stretches the muscle spindles, which send impulses via the primary spindle afferents into the spinal cord. The spindle afferents synapse directly, without participation of any interneuron, on the alpha motor neurons innervating the muscle, causing a reflex contraction of the muscle (Figure 32.1). This sequence of lengthening, contraction, and then relaxation is a stretch (tendon, deep tendon, muscle stretch, myotatic, or proprioceptive) reflex. Some muscles react more strongly than do others.

Stretch reflexes serve a protective function, particularly in standing and walking; they help to counter any sudden unexpected forces. Because of their critical roles in maintaining an erect posture, the extensor muscles of the legs, quadriceps, and calf muscles have better developed stretch reflexes than do the flexors. This important physiology is exploited clinically by applying an artificial stretch by striking the tendon of the muscle with a reflex hammer.

Reflexes elicited by application of a stretch stimulus to either tendons or periosteum or occasionally to bones, joints, fascia, or aponeurotic structures are usually referred to as muscle stretch or deep tendon reflexes (DTRs). The reflex is caused by sudden muscle stretch brought about by percussion of its tendon. Occasionally, the tendon is stretched by percussing a structure to which it is attached, as in the jaw jerk. The term deep helps separate these reflexes from the superficial or cutaneous reflexes, which are quite different.

Some authorities criticize the term deep tendon reflex, contending it implies the receptor is in the tendon, which is of course inaccurate. This link between the term and the location of the receptor is in the mind and opinion of the critic. Erb introduced the term tendon reflex in 1875 when he and Westphal published papers addressing the subject in the same issue of the same journal. In 1885, Gowers recommended the term be discarded. So, for over 100 years, it has been fashionable to rail against the term DTR. It is difficult to pick up a text on the neurologic examination without encountering at least a barb if not a screed on this point. In fact, the term DTR is in much wider use than muscle stretch reflex (MSR). Many more physicians are familiar with the acronym DTR than with MSR. Most physicians encountering the abbreviation MSR would think it was some Chinese cooking spice, but nearly everyone recognizes DTR, neurologist and nonneurologist alike. Gowers offered the term myotatic (Gr. myo, “muscle” + tatic “to stretch”), but many neurologists, not to mention other physicians, would pause and puzzle over this formal, obfuscatory term. For both pragmatic and antipedantic reasons, the abbreviation DTR is preferred in this text.

The primary problem areas in eliciting DTRs are poor tools and poor technique. These reflexes are best tested using a high-quality rubber percussion hammer. To properly obtain a reflex, a crisp blow must be delivered to quickly stretch the tendon. A heavy, high-quality reflex hammer is immensely helpful for this task, but many physicians use the cheapest hammer they can find, usually poor, pitiable, and inferior instruments. The worst possible hammers are drug company giveaway Taylor (tomahawk) hammers; they have no heft and are worth just what was paid for them. A genuine, high-quality, purchased Taylor is the lowest level of acceptable hammer, and these are often inadequate in the hands of novices. A variety of good hammers are available at reasonable prices. Other objects are sometimes used, and it seems a point of honor among some physicians to use anything but a reflex hammer. They substitute fingers, the edge of a stethoscope, or anything else handy, sometimes to the point of absurdity. The reliability of such a reflex examination mirrors the effort put into using a proper instrument. A soft rubber hammer is most desirable. A blow with an ancient, desiccated, stony hard hammer may cause pain for the patient and interfere with the response. The hammer should never leave bruises on either the patient or the examiner. The first 10 minutes of the video Demonstration of Teaching the Reflex Exam by Dr. Abraham Verghese, from the 2015 Stanford 25 Skills Symposium (see Video Link 38.1), features an excellent discussion of reflex hammers, worthy of any neurologist, delivered by an internist.

Proper technique is much more difficult to describe than to demonstrate. The hammer strike should be quick, direct, crisp, and forceful but no greater than necessary. The most effective blow is delivered quickly with a flick of the wrist, holding the handle of the hammer near its end and letting it spin through loosely held fingertips. Putting the index finger on top of the handle and using primarily elbow motion, common faults, make it much harder to achieve adequate velocity at the hammer head. Another common mistake is “pecking:” striking the tendon with a timid, decelerating blow, pulling back at the last instant.

The patient should be comfortable, relaxed, and properly positioned. It may help relaxation to divert the patient’s attention with light conversation. The optimal position is usually about midway in the range of motion of the muscle to be tested. Sometimes, as in the ankle reflex, positioning includes passively stretching the muscle slightly. An adequate stimulus must be delivered to the proper spot. Reinforcement methods are necessary if the reflex is not obtainable in the usual way. The part of the body to be tested should be in an optimal position for the response. In order to compare the reflexes on the two sides of the body, the position of the extremities should be symmetric. During the reflex examination, the patient should keep the head straight because looking to one side, as is the temptation, may alter reflex tone, especially in the arms (tonic neck reflex). The DTRs may be influenced to some degree by voluntary mental effort. Merely by concentrating, some individuals are able to somehow alter reflex excitability. Mentally induced reflex asymmetry is possible and may be clinically relevant in some cases.

The examiner can feel as well as see the contraction. Placing one hand over the muscle is often useful, especially when responses are sluggish. A reflex quadriceps contraction can sometimes be felt even when insufficient to produce visible contraction or knee movement. The activity of a reflex is judged by the speed and vigor of the response, the range of movement, and the duration of the contraction. An absent reflex often makes a dull, thudding sound when the tendon is struck.

The DTRs usually examined include the biceps, triceps, brachioradialis, knee (quadriceps), and ankle (Achilles) tendon reflexes (Video 38.1 ). Other DTRs are occasionally useful (Video 38.2 ). Table 38.1 summarizes the reflex levels. Reflexes may be graded as absent, sluggish or diminished, normal, exaggerated, and markedly hyperactive. For the purposes of clinical note-taking, most neurologists grade the DTRs numerically as follows: 0 = absent; 1+ (or +) = present but diminished; 2+ (or ++) = normal; 3+ (or +++) = increased but not necessarily to a pathologic degree; and 4+ (or ++++) = markedly hyperactive, pathologic, often with extra beats or accompanying sustained clonus (see Chapter 40). The “+” after the number is more traditional than informative and is sometimes omitted. Signs are sometimes used to indicate subtle asymmetry, but generally, a grade of 2 means the same as 2+. Another level, trace (or +/−), is frequently added to refer to a reflex, most often an ankle jerk, that appears absent to routine testing but can be elicited with reinforcement. Some add a grade of 5+ for the patient with extreme spasticity and clonus. In the 0 to 4 scale, level 1+ DTRs are still normal but somewhat sluggish and difficult to elicit and hypoactive but, in the examiner’s opinion, not pathologic. Grade 3+ reflexes are “fast normal,” quicker than 2+, sometimes very quick, but not accompanied by any other signs of upper motor neuron pathology such as increased tone, upgoing toes, or sustained clonus. Normality of the superficial reflexes, normal lower-extremity tone, and downgoing toes are reassuring evidence of fast normal rather than pathologically quick reflexes. Some use 3+ to indicate the presence of spread or unsustained clonus, with all other normal reflexes, even very fast ones, labeled as 2+. Grade 4+ reflexes are unequivocally pathologic. The speed of the response is very fast, the threshold is low, the reflexogenic zone is wide, and there are accompanying signs of corticospinal tract dysfunction. Other scales are in use but not widely. The Mayo Clinic utilizes a scale in which 0 is normal and reflexes are either increased (1+ to 4+) or decreased (1− to 4−). Reflexes may be charted in several ways, for example, as shown in Table 38.2, or as in Figure 38.1. When reflexes are very active, responses may occur from muscles that have not been directly stretched, even in normal patients. The response may involve adjacent or even contralateral muscles, and the contraction of one muscle may be accompanied by contraction of other muscles. This is referred to as spread, or irradiation, of reflexes. It is normal for percussion of the brachioradialis tendon to also cause slight finger flexion. In the presence of spasticity and hyperreflexia, contraction of the biceps or brachioradialis may be accompanied by pronounced flexion of the fingers and adduction of the thumb. Extension of the knee may be accompanied by adduction of the hip, or there may be bilateral knee extension. Judging how much spread is still within normal limits can be difficult. Under some circumstances, the expected response to percussion of a tendon is absent, but muscles innervated by adjacent spinal cord segments contract instead (e.g., inverted brachioradialis reflex) (see Inverted and Perverted Reflexes). On other occasions, a reflex is absent and percussion of the tendon causes an inverted or paradoxical contraction (e.g., elbow flexion on attempted elicitation of the triceps reflex).

TABLE 38.1 The Commonly Elicited Deep Tendon (Muscle Stretch) Reflexes

TABLE 38.2 Method of Recording the Commonly Tested Muscle Stretch Reflexes

Grades 0 to 4+ (see text) used for all plantar reflex, which is down (normal), absent (0), equivocal (+/−), or up (abnormal). Other reflexes may be added and charted as needed.

FIGURE 38.1 Alternate method of recording the commonly tested muscle stretch reflexes. For grading, see text and Table 38.2.

In some patients, DTRs may be markedly diminished, or even apparently absent, although there is no other evidence of nervous system disease. Under such circumstances, reinforcement techniques are often useful. Reflex reinforcement probably involves supraspinal, fusimotor, and long-loop mechanisms. A reflex can be reinforced or brought out using several methods. In the Jendrassik maneuver, the patient attempts to pull the hands apart with the fingers flexed and hooked together, palms facing, as the tendon is percussed (Figure 38.2). The effect is very brief, lasting only 1 to 6 seconds, and is maximal for only 300 ms. The Jendrassik maneuver is obviously useful only for lower-extremity reflexes. Other techniques include having the patient clench one or both fists or firmly grasp the arm of the chair, side of the bed, or the arm of the examiner. Reinforcement may also be carried out by having the patient look at the ceiling, grit the teeth, cough, squeeze the knees together, take a deep breath, count, read aloud, or repeat verses at the time the reflex is being tested. A sudden loud noise, a painful stimulus elsewhere on the body—such as the pulling of a hair or a bright light flashed in the eyes—may also be a means of reinforcement.

FIGURE 38.2 Method of reinforcing the patellar reflex.

Procedures other than distraction are also helpful in reflex reinforcement. A slight increase in tension of the muscle being tested may reinforce the reflex response. A simple and effective method to reinforce a knee or ankle jerk is to have the patient maintain a slight, steady contraction of the muscle whose tendon is being tested (e.g., slight plantar flexion by pushing the ball of the foot against the floor or the examiner’s hand to reinforce the ankle jerk). The patient may tense the quadriceps by extending the knee slightly against resistance as the knee jerk is being elicited. Reinforcement may increase the amplitude of a sluggish reflex or bring out a latent reflex not otherwise obtainable. Reflexes that are normal on reinforcement, even though not present without reinforcement, may be considered normal. Slight muscle contraction due to inability to relax may be one reason for the slightly hyperactive reflexes often seen in patients who are tense or anxious.

The DTRs are instrumental in the evaluation of weakness. Under most circumstances, weakness accompanied by hyporeflexia is of lower motor neuron origin, and weakness accompanied by hyperreflexia is of upper motor neuron origin. The presence of pathologic reflexes (see Chapter 40) and abnormalities of associated movements (see Chapter 42) are also helpful in the differential diagnosis (Table 38.3). The following sections discuss the upper-extremity, trunk, and lower-extremity reflexes. The masseter or mandibular reflex (jaw jerk) is covered in Chapter 15.

TABLE 38.3 Reflex Patterns with Different Neurologic Disorders

The Upper-Extremity Reflexes

The biceps, triceps, brachioradialis, and finger flexor reflexes are the most important upper-extremity reflexes. Less often useful reflexes are summarized in Table 38.4.

TABLE 38.4 Less Often Useful Deep Tendon Reflexes

The Biceps Reflex

With the arm relaxed and the forearm slightly pronated and midway between flexion and extension, the examiner places the palmar surface of her extended thumb or finger on the patient’s biceps tendon and then strikes the extensor surface with the reflex hammer (Figure 38.3). Pressure on the tendon should be light; too much pressure exerted with the thumb or finger against the tendon makes the reflex much harder to obtain. The hands may lie in the patient’s lap, or the examiner may hold the patient’s arm with the elbow resting in her hand. The major response is a contraction of the biceps muscle with flexion of the elbow. Because the biceps is also a supinator, there is often a certain amount of supination. If the reflex is exaggerated, the reflexogenic zone is increased and the reflex may even be obtained by tapping the clavicle; there may be abnormal spread with accompanying flexion of the wrist and fingers and adduction of the thumb.

FIGURE 38.3 Method of obtaining the biceps reflex.

The Triceps Reflex

This reflex is elicited by tapping the triceps tendon just above its insertion on the olecranon process of the ulna. The arm is placed midway between flexion and extension and may be rested in the patient’s lap, on her thigh or hip, or on the examiner’s hand (Figure 38.4). The response is contraction of the triceps muscle with extension of the elbow. The most common error in eliciting the triceps jerk is simply too timorous a blow. The paradoxical or inverted triceps jerk consists of flexion of the elbow with percussion of the triceps tendon. This response appears when the afferent arc of the triceps reflex is damaged, as in lesions of the seventh and eighth cervical segments, particularly when there is an element of spasticity, as in cervical spondylosis with radiculomyelopathy.

FIGURE 38.4 Method of obtaining the triceps reflex.

The Brachioradialis (Radial Periosteal or Supinator) Reflex

Tapping just above the styloid process of the radius with the forearm in semiflexion and semipronation causes flexion of the elbow, with variable supination (Figure 38.5). The supination is more marked with the forearm extended and pronated, but there is less flexion. The principal muscle involved is the brachioradialis. The tendon can be percussed not only at its insertion on the lateral aspect of the base of the styloid process of the radius but also at about the junction of the middle and distal thirds of the forearm or at its tendon of origin above the lateral epicondyle of the humerus. The most common error is hitting the muscle belly rather than the tendon. The muscle becomes tendinous at about midforearm. A local contraction can be elicited from any muscle by directly striking the muscle belly. The point in eliciting a DTR is to lengthen the muscle by stretching its tendon. An idiomuscular contraction can be obtained by striking the brachioradialis muscle belly in the proximal third of the forearm; this is not a DTR. If the reflex is exaggerated, there is associated flexion of the wrist and fingers, with adduction of the forearm. When the afferent limb of the reflex is impaired, there may be a twitch of the flexors of the hand and fingers without flexion and supination of the elbow; this is termed inversion of the reflex.

FIGURE 38.5 Method of obtaining the brachioradialis reflex.

The biceps, triceps, and brachioradialis reflexes should be obtained without difficulty in normal individuals. The following upper-extremity reflexes may be elicited only to a slight extent in normal persons. They may become conspicuous when there is hyperreflexia.

The Finger Flexor Reflex (Wartenberg’s Sign)

This is one of several signs attributed to Robert Wartenberg. To elicit the finger flexor reflex, the patient’s hand is in supination, resting on a table or a solid surface, with the fingers slightly flexed. The examiner places her fingers against the patient’s fingers and taps the backs of her own fingers lightly with the reflex hammer (Figure 38.6). The response is flexion of the patient’s fingers and the distal phalanx of the thumb (Video 38.2 ). The reflex may be reinforced by having the patient flex her fingers slightly as the blow is delivered. An alternate technique is for the patient to hold the hand in the air, palm down, and the examiner to touch fingers with palm up, with the blow delivered in an upward direction from below. The nerve supply, as in the wrist flexion reflex, is through the median and ulnar nerves (C8-T1). This reflex is difficult for the inexperienced examiner to elicit, and it is often absent in normals. However, Wartenberg considered it one of the most important upper-extremity reflexes. The Hoffmann and Trömner signs, which are pathologic variations of this response, are described in Chapter 40.

FIGURE 38.6 Method of obtaining the finger flexor reflex.

The Deltoid Reflex

Tapping over the insertion of the deltoid muscle at the junction of the upper and middle third of the lateral aspect of the humerus results in slight abduction of the upper arm (axillary nerve, C5-C6).

The Pectoralis Reflex

With the patient’s arm in midposition between abduction and adduction, the examiner places her finger as nearly as possible on the tendon of the pectoralis major muscle near its insertion on the greater tuberosity of the humerus (Figure 38.7). Tapping the finger causes adduction and slight internal rotation of the arm at the shoulder (Video 38.2 ). The contraction of the muscle may be felt but usually not seen in the normal individual. In patients with cervical spondylotic myelopathy, a hyperactive pectoralis reflex indicates spinal cord compression at the C2-C3 and/or C3-C4 levels. This reflex is mediated by the medial and lateral pectoral (anterior thoracic) nerves (C5-T1).

FIGURE 38.7 Method of obtaining the pectoralis reflex.

The Clavicle Reflex

In patients with upper-extremity hyperreflexia, a tap over the lateral aspect of the clavicle is followed by extensive contraction of various muscle groups in the upper limb (see Video Link 38.2). There are individual variations, but normally, the response should be the same on each side. This is not a specific reflex, but an indication of spread of the reflex response. The response is minimal, usually absent, except in the face of upper-extremity hyperreflexia. It is useful in comparing the reflex activity of the two upper limbs.

The Pronator Reflex

With the elbow in semiflexion and the forearm semipronated, tapping over either the volar surface of the distal radius or the dorsal aspect of the styloid process of the ulna may produce brief supination followed by pronation of the forearm (Video 38.2 ). There may also be flexion of the wrist and fingers. The major muscles participating in this response are the pronator teres and pronator quadratus. This reflex may be exaggerated early when corticospinal tract lesions develop.

Trunk Reflexes

Reflexes from trunk muscles are obtained minimally or not at all in normal individuals.

The Abdominal Muscle (Deep Abdominal) Reflexes

The best responses are obtained by pressing down slightly with the fingers and then tapping briskly with a reflex hammer (see Video Link 38.3). The response is contraction of the abdominal muscles, and a deviation of the umbilicus toward the site of the stimulus. The innervation is by the intercostal nerves (anterior divisions of T5-T12), as well as the ilioinguinal and iliohypogastric nerves. The abdominal muscle reflexes are only minimally present in normal individuals. They are most significant if exaggerated or if there is dissociation between the deep and superficial abdominal reflexes (see Chapter 39). Brisk deep abdominal reflexes with absent superficial abdominal reflexes suggest a corticospinal tract lesion.

The Iliac Reflexes

Tapping over the iliac crest is followed by contraction of the lower abdominal muscles. This reflex is mediated by the lower intercostal nerves (T10-T12).

The Symphysis Pubis Reflexes

Tapping over the symphysis pubis is followed by a contraction of the abdominal muscles and a downward movement of the umbilicus. The patient should be recumbent, with the abdominal muscles relaxed and the thigh in slight abduction and internal rotation. If a unilateral stimulus is applied by tapping 1.5 to 2 cm from the midline, there is not only the “upper response” just described but also a “lower response,” or puboadductor reflex, with contraction of the adductor muscles of the thigh on the side stimulated and some flexion of the hip. The latter response is also seen if the reflex is exaggerated. The symphysis pubis reflex is innervated by the lower intercostal, ilioinguinal, and iliohypogastric nerves (T11-T12 and upper lumbar segments). When there is spasticity, percussion over the symphysis may cause adduction of both legs. The costal periosteal, iliac, and symphysis pubis reflexes may be considered variations of the deep abdominal muscle reflexes in which the stimulus is directed toward the site of insertion.

The Lower-Extremity Reflexes

The Patellar Reflex (Quadriceps Reflex, Knee Jerk)

The patellar reflex is contraction of the quadriceps femoris muscle, with resulting extension of the knee, in response to percussion of the patellar tendon. A firm tap on the tendon draws the patella down, stretching the quadriceps and provoking reflex contraction. If the reflex is brisk, the contraction is strong and the amplitude of the movement is large. If the examiner places one hand over the muscle, and with the other hand taps the patellar tendon just below the patella, she can palpate the contraction as well as observe the rapidity and range of response. When the reflex is hypoactive, it may be better felt than seen. Palpation helps in judging the latency between the time of the stimulus and the resulting response.

The knee jerk can be elicited in various ways. The patient may sit in a chair with the knees slightly extended and the heels resting on the floor or sit on an examination table with the legs dangling (Figure 38.8). If the patient is lying in bed, the examiner should partially flex the knee by placing one hand beneath it and then tap the tendon (Figure 38.9). The responses on the two sides can be compared by lifting both knees simultaneously, supporting them on one forearm as the patient’s heels rest lightly on the bed, before tapping the tendons. If the patient is wearing loose pajamas, the examiner can suspend both legs by holding the pajamas, as she uses the other hand to strike the tendon. Another technique is having the patient sit with one leg crossed over the other and tapping the patellar tendon of the uppermost leg, but this method does not facilitate side-to-side comparison. Figure 38.10 shows a physician using this method. The patellar reflex is mediated by the femoral nerve (L2-L4).

FIGURE 38.8 Method of obtaining the patellar (quadriceps) reflex with the patient seated.

FIGURE 38.9 Method of obtaining the patellar (quadriceps) reflex with the patient recumbent.

FIGURE 38.10 Eliciting the knee jerk with one leg crossed over the other, using a Babinski hammer. (Reprinted from Lanska DJ. The Babinski reflex hammer. Neurology 1999;53[3]:655, with permission.)

If there is reflex spread, extension of the knee may be accompanied by adduction of the hip, which on occasion is bilateral, or there may be bilateral knee extension. If the reflex is exaggerated, the response may be obtained by tapping the tendon not only in the usual spot but also just above the patella (suprapatellar reflex); the tendon can be tapped directly, or, with the patient recumbent, the examiner can place her index finger on the upper border of the patella and tap the finger to push down the patella. Contraction of the quadriceps causes a brisk upward movement of the tendon, together with extension of the leg (Figure 38.11). Marked exaggeration of the patellar reflex may be accompanied by patellar clonus (see Chapter 40). An inverted patellar reflex may be seen with lesions of the nerve or nerve roots supplying the quadriceps: Tapping the patellar tendon results in contraction of the hamstrings and flexion of the knee.

FIGURE 38.11 Method of obtaining the suprapatellar reflex.

The Achilles Reflex (Ankle Jerk, Triceps Surae Reflex)

The ankle jerk is obtained by striking the Achilles tendon just above its insertion on the calcaneus. The resulting contraction of the posterior crural muscles, the gastrocnemius, soleus, and plantaris, causes plantar flexion of the foot at the ankle. If the patient is seated or lying in bed, the thigh should be held in moderate abduction and external rotation and the knee flexed. If the patient is supine, access to the tendon requires placing the legs into a frog-leg position with the knees apart and the ankles close together. Some prefer to have the patient cross the leg to be examined atop the other shin or ankle (“figure-four position,” as the legs form a 4). The examiner should place one hand under the foot and pull upward slightly to passively dorsiflex the ankle to about a right angle (Figure 38.12). The Achilles reflex is mediated by the tibial nerve (S1).

FIGURE 38.12 Method of obtaining the Achilles (triceps surae) reflex with the patient recumbent.

The ankle jerk is by far the most difficult reflex to master. There are two critical variables: proper stretch and efficient striking. Of the two, proper stretch is the more difficult to learn. Too little dorsiflexion leaves the tendon slack and able to absorb the blow without stretching the muscle. Too much passive dorsiflexion makes the tendon too taut and unable to be stretched. If the reflex is difficult to obtain, the patient may be asked to press her foot lightly against the examiner’s hand in order to tense the muscle and reinforce the reflex. Using a driving analogy and asking the patient to imagine pressing on an accelerator enough to go “17 mph” communicates the need for a low-level but precisely graded contraction, which is then easy to adjust up or down to the proper level. The reflex may also be elicited by having the patient kneel on a chair or similar surface, with the feet projecting at right angles; the Achilles tendons are percussed while the patient is in this position (Figure 38.13). This method, introduced by Babinski, is particularly useful for comparing reflex activity on the two sides. Another method for supine examination is to strike the ball (sole) of the foot or strike the examiner’s hand placed flat against the sole. This plantar stretch reflex is considered equivalent to the ankle jerk for clinical purposes.

FIGURE 38.13 Method of obtaining the Achilles (triceps surae) reflex with the patient kneeling.

If the ankle jerk is hyperactive, it may be elicited by tapping other areas of the sole of the foot, the medioplantar reflex, or by tapping the anterior aspect of the ankle, the paradoxical ankle reflex. A hyperactive reflex may also result in extra beats or even clonus when the tendon is percussed. When there is reflex spread, striking the Achilles tendon may cause flexion of the knee. Although the Achilles reflex, when carefully elicited, should be present in normal individuals, it tends to diminish with age, and its bilateral absence in elderly individuals is not necessarily of clinical significance. Although it is common wisdom that the normal elderly may have absent ankle reflexes, of 200 consecutive patients admitted to a geriatric unit, 188 had ankle jerks using plantar rather than Achilles tendon strike; only 1.5% had absent ankle jerks attributable only to age. Another study comparing plantar strike and tendon strike in elderly patients found better intraobserver and interobserver agreement with the plantar strike. Differences in technique may explain some of the discrepancy between studies examining the prevalence of absent ankle jerks in elderly people.

The knee and ankle reflexes are the most important DTRs in the lower extremities. The following reflexes are less significant. The response may be minimal or even absent in normal individuals. Because the responses may be difficult to elicit even in normal subjects, side-to-side comparison is critical. To be significant, absence must be unilateral. Exaggeration of these difficult-to-obtain reflexes suggests the presence of corticospinal tract disease.

The Adductor Reflex (Obturator Nerve, L2-L4)

With the thigh in slight abduction, tapping over either the medial epicondyle of the femur in the vicinity of the adductor tubercle or the medial condyle of the tibia results in contraction of the adductor muscles of the thigh and inward movement of the extremity. If the reflex is exaggerated, there may be crossed, or bilateral, adduction. A slightly crossed adductor response is not necessarily abnormal, but strong crossed adduction, or adduction of the opposite leg when obtaining the knee jerk, suggests corticospinal tract disease. When there is hyperreflexia, an adductor response may also be obtained while the patient is seated by tapping the spinous processes of the sacral or lumbar vertebrae (the spinal adductor reflex) or by tapping the crest or the superior spines of the ilium. An absent adductor reflex with a normal patellar reflex has been described as a sign of strangulated obturator hernia (Hannington-Kiff sign). The puboadductor reflex was described with the symphysis pubis reflex.

The Medial Hamstring (Internal Hamstring) Reflex

This reflex is elicited by striking the semitendinosus and semimembranosus tendons just above their insertions on the tibia. This may be done with the patient seated or recumbent, with the leg abducted and slightly rotated externally and the knee flexed. The examiner’s fingers are placed over the tendons on the medial posterior aspect of the knee and the fingers tapped with the reflex hammer. The response is knee flexion (Video 38.2 ). This reflex is mediated by tibial portion of the sciatic nerve, primarily by the L5 nerve root. It may be useful in the evaluation of suspected L5 radiculopathy. For a video of the medial hamstring reflex, see Perloff et al.

The Lateral Hamstring (External Hamstring) Reflex

This reflex is elicited by striking the biceps femoris tendon just above its insertion. With the patient sitting, recumbent, or lying on the opposite side and the knee moderately flexed, the examiner’s fingers are placed over the tendon on the lateral posterior aspect of the knee and tapped (Figure 38.14). The response is knee flexion (Video 38.2 ). The reflex may also be elicited by tapping the head of the fibula (fibular reflex). This reflex is mediated by tibial portion of the sciatic nerve, primarily by the S1 nerve root. The external hamstring reflex is sometimes helpful in sorting out whether an absent ankle jerk is due to peripheral neuropathy or radiculopathy. If due to neuropathy, the external hamstring will usually be preserved, but in radiculopathy, it may be depressed in concert with the ankle reflex.

FIGURE 38.14 Method of obtaining the biceps femoris reflex.

The Peroneal (Tibialis Anterior) Reflex

With the patient’s foot plantar flexed and inverted, the examiner presses a finger firmly over the distal ends of the first and second metatarsal bones. A brisk tap to the finger is followed by eversion and dorsiflexion of the foot. The reflex is due to contraction of muscles supplied by the deep and superficial peroneal nerves (L4-S1). The peroneal reflex may be useful in the evaluation of suspected L5 radiculopathy (see Video Link 38.4).

The Plantar Muscle Reflexes

There are numerous reflexes in which the response is flexion of the toes. These are difficult to elicit in normal individuals, of limited clinical significance, and of importance only when exaggerated. They are discussed with the pathologic reflexes in Chapter 40.

Interpretation of the Deep Tendon (Muscle Stretch) Reflexes

The most valuable DTRs for clinical diagnosis are the biceps, triceps, brachioradialis, patellar, and Achilles (see Table 38.1); under most circumstances, and using good technique, these are elicitable in every normal person. One or more of these reflexes may be absent in occasional individuals with no other evidence of disease of the nervous system. They are present even in the majority of premature infants. The activity of a DTR is judged by the threshold, latency, speed, vigor, and duration of contraction; the range of movement; and whether there is spread or irradiation of the reflex. Of these, the latent period between the time the stimulus is applied and the time the response occurs is most important for clinical evaluation of disease states. Accurate evaluation of the reflex responses obviously depends on the experience of the examiner. By far, the most important factor is the diligence and practice expended in learning the techniques. The appraisal depends on the individual interpretation of the examiner. There is no standard, and there is a certain amount of normal variation in reflex activity. What is normal for one individual may be an increased or a decreased response for another. In some persons, the reflexes are lively; in others, they are sluggish. Under normal circumstances, the reflexes should be equal on the two sides.

Abnormalities of the Deep Tendon (Muscle Stretch) Reflexes

Abnormal DTRs are either hypoactive or hyperactive. When hypoactive, the response varies from diminished or sluggish to complete absence of the reflex. Hyperactive reflexes are characterized by varying degrees of decreased latency, increased speed and vigor of response, increased range of movement, decrease in threshold, extension of the reflexogenic zone, and prolongation of the muscular contraction. The pathologic conditions in which these various changes occur are discussed in the following sections. Table 38.3 summarizes the patterns of reflex responses seen with lesions at various sites.

Reflexes are judged in both absolute and relative terms. Clearly hyperactive or hypoactive reflexes speak for themselves. But a reflex that is normal in absolute terms may be judged abnormal in comparison to the patient’s other reflexes. The reflexes should be compared on the two sides of the body, the arms to the legs, and the knees to the ankles. The DTRs are normally symmetric, and reflexes otherwise normal may be abnormal if different from expected. For example, a 1+ biceps jerk in a patient with suspected cervical radiculopathy, although “normal,” may be judged abnormal if the opposite biceps jerk is 2+. The DTRs are usually comparable in the upper and lower extremities. Slight differences are permissible, but a pronounced difference may be significant (e.g., in thoracic myelopathy, the DTRs in the legs may be much brisker than in the arms, even though not clearly pathologic). A proximal to distal gradient may also be significant. Symmetric 1+ ankle jerks when all of the other reflexes are 2+ may signal mild peripheral neuropathy. When asymmetry is the main finding, it is sometimes difficult to tell whether one side is increased or the other side decreased.

Hypoactive Reflexes

When a reflex is hypoactive, there is a sluggish response and/or a diminution in the range of response. An increase in stimulus intensity may be necessary to elicit the reflex, or repeated blows may be necessary, for a single stimulus may be subliminal. A DTR is absent if it is not obtained even with reinforcement. A depressed or absent reflex results from dysfunction of some component of the reflex arc. Interference with the afferent limb may be caused by lesions involving the sensory nerve, posterior root, dorsal root ganglion, or intramedullary pathways between the dorsal root entry zone and the anterior horn (e.g., syringomyelia). Abnormalities of the motor unit and final common pathway that make up the efferent limb of the reflex arc occur in many conditions, but particularly with radiculopathy and peripheral nerve lesions. In neurogenic processes, DTRs are lost out of proportion to atrophy and weakness. With a peripheral nerve lesion, a reflex may not return until much of the motor function has been recovered. Sometimes there is persistent areflexia following lesions of the nerve root or peripheral nerve, even after complete return of both motor and sensory functions. In myasthenia gravis, the reflexes are affected only when there is severe and extensive involvement, but in Lambert-Eaton syndrome, depressed reflexes are common. In periodic paralysis, DTRs may be temporarily absent during attacks. In myopathies, reflexes are lost in proportion to the atrophy and weakness. When atrophy and weakness are severe, reflexes may disappear. In many forms of muscular dystrophy, the proximal reflexes disappear early, whereas the distal reflexes may persist until the later stages of the disease.

The DTRs may also be decreased or absent in various other conditions. They are often absent in deep coma, narcosis, heavy sedation, and deep sleep. They are characteristically absent during nerve block, caudal anesthesia, and spinal anesthesia. They are absent in spinal shock following a sudden transverse lesion of the spinal cord but reappear below the level of the lesion after a period of 3 to 4 weeks and usually become hyperactive. In Adie’s syndrome, the tonic pupils are accompanied by depressed or absent reflexes.

A prolonged relaxation phase causing a “hung-up” reflex, especially of the ankle jerk, is a classical finding of hypothyroidism (Woltman’s sign); the reflexes return to normal with treatment (Video Link 38.5). But most patients with seemingly hung-up ankle jerks are euthyroid. Cold legs and feet may cause hung-up ankle jerks. Relaxation slows with advancing age, more so in females. Slow contraction and relaxation times may also occur with other conditions, including lower motor neuron disease. Delayed relaxation may also occur in myotonic disorders. In diabetic neuropathy, there may be either prolongation of the reflex time or decrease or absence of the reflexes before there is other evidence of nervous system involvement. The reflexes may appear to be decreased or absent in neurologic disorders in which there is marked spasticity or rigidity with contractures and in diseases of the joints causing inflammation, contractures, and ankylosis. The apparent hyporeflexia is due to lack of motility at the joint or pain on moving the joint; careful observation may disclose a muscle contraction even though there is no movement at the joint.

Hyperactive Reflexes

Reflex hyperactivity is characterized by the following: a decrease in reflex threshold; a decrease in the latency, the time between tendon percussion and the reflex contraction; an exaggeration of the power and range of movement; prolongation of the reflex contraction; extension of the reflexogenic zone; and spread of the reflex response. When the reflex threshold is decreased, a minimal stimulus may evoke the reflex, and reflexes that are not normally obtained may be elicited with ease. Very hyperactive DTRs may sometimes be elicited with extremely slight percussion. Another manifestation of decreased reflex threshold may be a widening of the area from which the reflex may be elicited, and application of the stimulus to sites at some distance from the usual one may evoke the response; the patellar reflex may be elicited by tapping the tibia or dorsum of the foot, and the biceps and other arm reflexes by tapping the clavicle or scapula. There may also be abnormal spread of the response. One stimulus may provoke repetitive responses and sometimes elicit sustained clonus.

The DTRs become hyperactive with lesions of the corticospinal system. Spasticity and hyperreflexia are likely related to involvement of a variety of structures in the descending motor pathways at cortical, subcortical, midbrain, and brainstem levels. Hyperreflexia results from a lowering of the reflex threshold due to increased excitability of the lower motor neuron pool related to dysfunction of some or all of these structures. From a clinical point of view, the terms pyramidal, corticospinal, or upper motor neuron are used to encompass these changes. A lesion at any level of the corticospinal system or other related upper motor neuron components, from the motor cortex to just above the segment of origin of a reflex arc, will be accompanied by spasticity and hyperreflexia. The characteristic posture in hemiplegia is flexion of the upper extremities, with more marked weakness of the extensors, and extension of the lower extremities, with more marked weakness of the flexors. Consequently, the flexor reflexes are exaggerated to a greater degree in the upper extremities, and the extensor reflexes in the lower. The reflexes may be present in spinal cord lesions in spite of the absence of sensation. A reflex may be increased if the tone of the antagonist muscle is diminished (e.g., an increased knee jerk may occur if there is weakness of the hamstrings).

Exaggeration of the DTRs may occur in psychogenic disorders and in anxiety, fright, and agitation (Table 38.3). The reflexes vary in these conditions; they may be normal, or they may be decreased owing to voluntary or involuntary tension of the antagonistic muscle, but they are most frequently increased. Some normal individuals, particularly young women, have physiologically brisk reflexes. Hyperactivity may be marked, but it is an exaggeration not in the speed or threshold of the response but in the excursion or range of response. The foot may be kicked far into the air and held extended for a time after the patellar tendon is tapped, but the contraction and relaxation take place at a normal rate. There is often a bilateral response with extraneous and superfluous jerking of remote parts, including whole body jerks, when a reflex is tested. There is no increase in the reflexogenic zone in psychogenic lesions, and although there may be irregular repeated jerky movements (spurious clonus), no true clonus is present. Furthermore, there are no other signs of organic disease of the corticospinal system. To help distinguish physiologically brisk from pathologic reflexes, it is particularly useful to pay careful attention to lower extremity tone, to the status of the abdominal reflexes, and to the plantar responses. If the toes are down, the abdominals intact, and lower extremity tone normal, it is unlikely the DTRs are pathologic.

In lesions of the extrapyramidal system, there are no consistent reflex changes (Table 38.3). The activity of the response depends on the level of muscle tone and the amount of rigidity that is present. Usually, the reflexes are slightly exaggerated, owing to increased muscle tone, but this is not a consistent finding. Rigidity may cause depression or absence of the reflexes. In diseases of the cerebellum, the reflexes may be diminished (Table 38.3) and pendular: Eliciting the patellar reflex while the foot is hanging free may elicit a series of to-and-fro pendular movements of the foot and leg before the limb finally comes to rest. The increased swinging may result from hypotonia of the extensor and flexor muscles and a lack of the restraining influence they normally exert on each other. The pendular response may also be observed in chorea, but there is more frequently a “hung” reflex: If the patellar tendon is tapped while the foot is hanging free, the knee may be held in extension for a few seconds before relaxing because of prolonged contraction of the quadriceps. In chorea, the response may not be obtained until the stimulus has been applied a number of times.

Inverted (Paradoxic, Indirect) Reflexes

Occasionally, percussion of a tendon produces unexpected results. When reflexes are very active, responses may occur from muscles that have not been directly stretched, even in normal patients. The response may involve adjacent or even contralateral muscles, as in the crossed adductor response. It is normal for percussion of the brachioradialis tendon to also cause slight finger flexion as well as biceps contraction in some patients. This is referred to as spread, or irradiation, of reflexes.

Inverted reflexes are responses more bizarre than simple spread. The term inversion was first used by Babinski in describing “inversion of the radial reflex.” These responses are most often seen when the segmental reflex is absent or impaired and there is also underlying hyperreflexia lowering the threshold for activation of other muscles. These responses have also been referred to as “indirect” reflexes because the reflex occurs without direct stretch of the muscle that contracts. The bulk of the evidence suggests these inverted and paradoxical reflexes are mediated by vibration in the periphery.

There are many examples of these inverted or paradoxical reflexes: contraction of the finger flexors on percussion of the distal radius (inverted brachioradialis reflex), contraction of the biceps on eliciting the triceps reflex (paradoxical triceps jerk), contraction of the hamstring on eliciting the knee reflex (inverted knee jerk), and contraction of the hamstring on eliciting the ankle reflex. These responses occur most commonly in the face of segmental central hyperexcitability combined with impairment of the local reflex arc, such as an inverted brachioradialis reflex in cervical spondylotic radiculomyelopathy.

The easiest paradoxical reflexes to understand are the inverted triceps jerk and the inverted knee jerk, which are truly “inverted” because the response is the opposite of that expected: flexion of the elbow or knee with percussion of the triceps or patellar tendon. An inverted triceps reflex appears when the afferent arc of the reflex is damaged, as in C7 or C8 radiculopathy, particularly when accompanied by an element of spasticity. Care must be taken not to strike too distally; a blow delivered over the olecranon may cause the elbow to flex because of the biomechanics and force vectors involved, simulating an inverted triceps reflex when one is not actually present. This response has been referred to as the olecranon reflex but is more of a technical error in the point of the strike than a true reflex. With an inverted knee jerk there is knee flexion instead of extension.

The inverted triceps and knee reflexes are straightforward: the response is the opposite of that expected. Other “inverted” reflexes are related to activation of muscles other than the antagonist. In the presence of spasticity and hyperreflexia, reflex contraction of the biceps or brachioradialis may be accompanied by pronounced flexion of the fingers and adduction of the thumb. The inverted brachioradialis reflex (often referred to as an inverted radial periosteal or inverted supinator reflex) does not result in true inversion, that is, elbow extension, but instead produces a perverted or aberrant response with thumb and finger flexion, perhaps because the vibration wave activates the finger flexor reflex arc.

In the paradoxical ankle reflex, plantar flexion of the foot is produced by tapping the anterior aspect of the ankle in patients with hyperreflexia. The ankle reflex is not necessarily depressed, as with other inverted or perverted reflexes, and the response is the normal one but is due to spread of the reflexogenic zone. Vibration transmitted to the gastrosoleus has been invoked as an explanation.

Video Links

Video Link 38.1. Discussion of reflex hammer. https://www.youtube.com/watch?v=1-0Gq6s_Eoo&t=374s

Video Link 38.2. The clavicle reflex. http://neurosigns.org/wiki/Clavicle_reflex

Video Link 38.3. Dissociation of the abdominal reflexes. http://neurosigns.org/wiki/Dissociation_of_the_abdominal_reflexes

Video Link 38.4. The peroneal reflex. http://neurosigns.org/wiki/Peroneal_reflex

Video Link 38.5. Delayed ankle jerks in hypothyroidism. https://www.nejm.org/doi/10.1056/NEJMicm1713796?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%3dwww.ncbi.nlm.nih.gov

Bibliography

Berlin L. A peroneal muscle stretch reflex. Neurology 1971;21:1177.

Bhattacharyya KB. Deep tendon reflex: the background story of a simple technique. Neurol India 2017;65:245–249.

Boes CJ. The history of examination of reflexes. J Neurol 2014;261:2264–2274.

Boyle RS, Shakir RA, Weir AI, et al. Inverted knee jerk: a neglected localising sign in spinal cord disease. J Neurol Neurosurg Psychiatry 1979;42:1005.

Carel RS, Korczyn AD, Hochberg Y. Age and sex dependency of the Achilles tendon reflex. Am J Med Sci 1979;278: 57–63.

Chandrasekhar A, Abu Osman NA, Tham LK, et al. Influence of age on patellar tendon reflex response. PLoS One 2013;8:e80799.

Dick JP. The deep tendon and the abdominal reflexes. J Neurol Neurosurg Psychiatry 2003;74:150–153.

Esene IN, Meher A, Elzoghby MA, et al. Diagnostic performance of the medial hamstring reflex in L5 radiculopathy. Surg Neurol Int 2012;3:104.

Estanol BV, Marin OS. Mechanism of the inverted supinator reflex. A clinical and neurophysiological study. J Neurol Neurosurg Psychiatry 1976;39:905–908.

Felsenthal G, Reischer MA. Asymmetric hamstring reflexes indicative of L5 radicular lesions. Arch Phys Med Rehabil 1982;63:377–378.

Fuller G. Neurological Examination Made Easy. 5th ed. Edinburgh: Churchill Livingstone, 2013.

Gilman S. Clinical Examination of the Nervous System. New York: McGraw-Hill, 2000.

Gregory JE, Wood SA, Proske U. An investigation into mechanisms of reflex reinforcement by the Jendrassik manoeuvre. Exp Brain Res 2001;138:366–374.

Hannington-Kiff JG. Absent thigh adductor reflex in obturator hernia. Lancet 1980;1:180.

Impallomeni M, Kenny RA, Flynn MD, et al. The elderly and their ankle jerks. Lancet 1984;1:670.

Jensen OH. The medial hamstring reflex in the level-diagnosis of a lumbar disc herniation. Clin Rheumatol 1987;6:570–574.

Kiely P, Baker JF, O’hEireamhoin S, et al. The evaluation of the inverted supinator reflex in asymptomatic patients. Spine (Phila Pa 1976) 2010;35:955–957.

Kochar DK, Agarwal N, Sharma BV, et al. Paradoxical triceps jerk, a neglected localising sign in clinical neurology. Neurol India 2001;49:213.

Lanska DJ. The history of reflex hammers. Neurology 1989;39:1542–1549.

Lanska DJ. The Babinski reflex hammer. Neurology 1999;53:655.

Louis ED. Erb and Westphal: simultaneous discovery of the deep tendon reflexes. Semin Neurol 2002;22:385–390.

Martinelli P, Minardi C, Ciucci G, et al. Neurophysiological evaluation of areflexia in Holmes-Adie syndrome. Neurophysiol Clin 1999;29:255–262.

Massey EW, Pleet AB, Scherokman BJ. Diagnostic Tests in Neurology: A Photographic Guide to Bedside Techniques. Chicago: Year Book Medical Publishers, Inc., 1985.

O’Keeffe ST, Smith T, Valacio R, et al. A comparison of two techniques for ankle jerk assessment in elderly subjects. Lancet 1994;344:1619–1620.

Perloff MD, Leroy AM, Ensrud ER. Teaching video neuroimages: the elusive L5 reflex. Neurology 2010;75:e50.

Pryse-Phillips W. Companion to Clinical Neurology. 3rd ed. Oxford: Oxford University Press, 2009.

Ross RT. How to Examine the Nervous System. 4th ed. Totowa: Humana Press, 2006.

Schwartz RS, Morris JGL, Crimmins D, et al. A comparison of two methods of eliciting the ankle jerk. Aust NZ J Med 1990;20:116.

Stam J. The tibialis anterior reflex in healthy subjects and in L5 radicular compression. J Neurol Neurosurg Psychiatry 1988;51:397–402.

Stam J, Speelman HD, van Crevel H. Tendon reflex asymmetry by voluntary mental effort in healthy subjects. Arch Neurol 1989;46:70.

Vrancken AF, Kalmijn S, Brugman F, et al. The meaning of distal sensory loss and absent ankle reflexes in relation to age: a meta-analysis. J Neurol 2006;253:578–589.

Wartenberg R. The Examination of Reflexes: A Simplification. Chicago: Year Book Medical Pub, 1945.

Watson JC, Broaddus WC, Smith MM, et al. Hyperactive pectoralis reflex as an indicator of upper cervical spinal cord compression. Report of 15 cases. J Neurosurg 1997;86:159–161.

Zabelis TN, Karandreas NT, Constantinidis TS, et al. The effect of Jendrassik manoeuvre on the latency, amplitude and left-right asymmetry of tendon reflexes. Electromyogr Clin Neurophysiol 1998;38:19–23.

Ziff M, Stark RJ. How to assess tendon reflexes of the lower limb in the elderly. J Neurol Sci 2017;372:196–200.