Several types of reflexes can be tested as part of a physical examination and these all reveal something about the status of the elements of the nervous system that contribute to their functioning. They have been used for over a century as part of a routine neurological examination due to their safety, low cost, predictive value, and ability to be performed rapidly, even without specialized equipment.
This article will focus on the “deep tendon reflexes” which are more appropriately named — and will be referred to herein — as muscle stretch reflexes (MSR).
MSR grading is based on a clinician’s subjective evaluation of amplitude, with a wide range of what can be normal. They are particularly useful if there are asymmetric findings or if they occur in the context of other changes; isolated hyper or hypo reflexic MSR without other findings is generally not considered pathological. There are 6 MSR that are commonly tested and will be the focus of this article: biceps, triceps, brachioradialis, knee, ankle, and jaw jerk.
The first 5 MSR listed above; however not the jaw jerk, occur when stretch receptors within the muscle are activated; the clinician taps a part of the body to create a small amount of lengthening (stretch) in a muscle. This activates stretch receptors in the muscle spindle, of which the cell bodies are in the dorsal root ganglia, and an afferent impulse is conducted via these Ia afferent neurons to the spinal cord where they directly synapse on an alpha motor neuron. This activated motor neuron transmits the efferent impulse back to the muscle leading to contraction. When an MSR is initiated there is also a signal that branches off from the Ia afferent stretch neuron and activates an inhibitory interneuron that will prevent the opposing (flexor or extensor) muscle group from activating during the reflex. For instance, at the knee, the quadriceps opposes the hamstring, so when the knee jerk reflex is elicited it inhibits the hamstring motor neurons while the quadriceps are stimulated to contract. The monosynaptic junction of the muscle spindle stretch afferent neuron and the motor neuron is modulated by the descending corticospinal tract which generally has a dampening effect on the reflex. This reflex arc has musculoskeletal components as well as peripheral nervous and central nervous components, therefore a broad range of pathology can manifest in alterations of reflexes. The lower motor neuron components of the reflex are the peripheral nerve and spinal segment while the upper motor neuron component innervating the reflex arc is the descending corticospinal tract. Pathology in one or the other produces different changes in the reflex (see clinical significance below).
The jaw jerk is conceptualized in a similar manner. There is an afferent signal in the form of proprioceptive signals being transmitted along the mandibular nerve to the mesencephalic nucleus of the trigeminal nerve. These pseudounipolar neurons send bilateral projections to the efferent motor neurons (in the motor nuclei of the trigeminal in the pons) which sends the motor impulse out to the masseter muscles to contract on both sides of the jaw. The modulating upper motor neuron innervation comes from the corticobulbar rather than the corticospinal tract.
Below are the 6 commonly tested MSRs, in the following format:
Name: Spinal level, peripheral nerve, location struck to elicit
The following is a frequently used, and most accurate, nursery rhyme style mnemonic has been used to help students remember the reflexes:
One-two, buckle my shoe. Three-four, kick the door. Five-six, pick up sticks. Seven-eight, lay them straight. S1,2 ankle jerk; L3,4 knee jerk; C5,6 biceps and brachioradialis; C7,8 triceps.
The Achilles reflex can also be elicited, in an equally effective manner, by striking the plantar surface of the ball of the foot while creating the desired 90-degree angle (which requires slight joint tension). This can be a useful alternative for examining a patient lying supine in a hospital bed.
A variety of tools are used to elicit a reflex which ranges from specialized to improvised, with specialized hammers being generally preferable. The most commonly used specialized reflex hammers are grouped into 3 types by the shape of the head: triangular/tomahawk shaped (Taylor), T-shaped (Tromner, Buck), or circular (Queen Square, Babinski). Each is effective at provoking reflexes, with the Taylor being possibly less favorable at eliciting more stubborn (hyporeflexia) reflexes.
First 5 MSR
The patient should be positioned so that they can have the muscles above and below the joint completely relaxed and the joint can hang in a roughly 90-degree position. This may require supporting the distal component of the joint (for instance, when testing the biceps reflex, the elbow will naturally hang limply at nearly 180 degrees of extension but can be supported to approximately 90 degrees while remaining fully relaxed by placing it on the patient’s thigh if seated, or by supporting it in a clinician’s arm). The infant’s heads should be positioned midline when assessing any reflex.
The patient should be positioned with their jaw hanging in a relaxed, unlocked position. Therefore, the mouth should be slightly open.
With clean hands on a fully relaxed joint, the tendon/target is struck with sufficient force to elicit the reflex while the clinician’s eyes are focused on the proximal muscle group, looking for contraction, rather than the distal appendage, looking for movement. To achieve full relaxation of a joint, the patient may need to be distracted. This can be accomplished by asking them to clench their jaw, conversing with them (those techniques should not be used for jaw jerk), or asking them to hook their fingers together and forcefully pull their arms apart (Jendrassik maneuver for lower extremity and jaw jerk reflexes only). Those assistive techniques are collectively called "reinforcement". The reflex being tested should be immediately compared to the contralateral reflex (right elbow, then left elbow, rather than right elbow, right knee) so that subtle lateralized differences can be appreciated.
The technique may vary slightly depending on what type of tool is used or what reflex is being tested, for instance, circular hammers can be "dropped" passively through an arc using gravity to strike the patellar tendon, but when striking the biceps tendon are generally swung like a drumstick. In any case, the tendon’s axis is generally struck perpendicular to the plane of the hammer (if the hammer has a flat edge), meaning if you are striking the vertically oriented patellar tendon in a seated patient the rim of the tool will generally be horizontal and parallel to the floor. A stethoscope should be held 1 inch from the stem, on the tubing, and swung in a short arc. If a patient is hyperreflexic, a clinician’s finger may be all that is needed because the forces needed are so slight. With any tool, a finger can be placed on the tendon to help guide the clinician's blow to the correct location, to help feel the contraction, and to reduce discomfort for the patient by cushioning the blow. This is most commonly done when eliciting the biceps reflex.
Reflexes are graded based on amplitude. Various scales have been used to grade reflexes. The National Institute of Neurological Disorders and Stroke (NINDS) Muscle Stretch Reflex Scale is frequently used and empirically supported:
Plus or minus can indicate that the reflex was in between whole grades. Occasionally “5” is used to denote sustained clonus. One-half can alternatively indicate reinforcement was used.
Reflexes can indicate pathology along a particular nerve route, to a section of the spinal cord, or even reveal something about the general state of a patient’s peripheral nerves. For instance, the presence of an ankle jerk reflex is a better predictor of peripheral neuropathy in diabetics than neuropathic symptoms, duration of diabetes, or retinopathy. An alteration in reflexes can also provide information on a patient’s electrolyte status due to the role of these charged molecules in the conduction of nerve impulses.
Due to the roles of the various pathways indicated in the anatomy section, hyperreflexia can indicate an upper motor neuron lesion, whereas hyporeflexia can indicate a lower motor neuron lesion. In other words, a lesion at C5 would cause hyporeflexia in the biceps reflex (C5-C6) but could produce hyperreflexia in the triceps reflex. This is because the lesion is above the spinal segments (C7-C8) that contribute to the lower motor neuron reflex arc but would affect the upper motor neuron modulation (which are generally inhibitory signals) of that arc instead. This is why one of the “lower motor neuron signs” is considered hyporeflexia (along with weakness, atrophy, and fasciculations), and conversely, “upper motor neuron signs” include hyperreflexia (along with weakness, spasticity, and Babinski sign). Put another way: inhibiting the inhibitor causes hyperreflexia, and this is what occurs with upper motor neuron lesions.
The absolute grading intervals are subjective, based on the clinician’s experience of all previous normal reflexes, but are particularly useful when discrepancies are found between different joints in the same patient. Symmetrical reflexes within the 1+ to 3+ range are generally considered normal as long as upper or lower motor neuron signs do not accompany them, or there is increasing briskness with a descent down the spinal cord, which could indicate spinal cord damage between the diminished and hyperreflexic levels.
Especially with regards to the +/- grading, rather than focusing on the difference in value between one clinician calling all reflexes on a patient a 2+, and another calling them all 2, it is more important if any clinician noted that the left side was 2+, but the right was 2 in the same exam. Even an absent reflex, if present bilaterally, without other pathology or reasons for concern, can be normal.
The MSRs are part of a physical exam typically performed by one clinician. They are routinely performed by clinicians and nurses in many medical specialties as well as in fields such as physical therapy. Neurologists, physiatrists, and orthopedists, among others, may use them as part of a consult note, which will then need to be interpreted by a primary care or hospitalist type clinician. A perinatal nurse may check reflexes frequently on an obstetric patient with eclampsia and enter their findings into a daily note. Therefore, a common language is important. As noted earlier, there are a variety of ways to denote the physical exam findings, so one should ensure that other clinicians, nurses, and other staff know how to interpret these findings and communicate clearly what was observed on physical examination. The subjective nature of grading MSRs adds to the importance of this. For instance, some scales indicate clonus would receive a “4,” whereas others would say it should be graded at whatever amplitude it was, and then “clonus” should be annotated. It is most clear and safest, therefore, for the clinician to write a very brief explanation of the finding as well as the scale they are using (“using NINDS MSR: 4, due to clonus”) to assist the next clinician or nurse who is accessing the chart. Similarly, as reinforcement can frequently elicit an otherwise absent reflex, it should be noted if reinforcement was attempted in an absent reflex.
|||Schwartz RS,Morris JG,Crimmins D,Wilson A,Fahey P,Reid S,Joffe R, A comparison of two methods of eliciting the ankle jerk. Australian and New Zealand journal of medicine. 1990 Apr [PubMed PMID: 2344316]|
|||Marshall GL,Little JW, Deep tendon reflexes: a study of quantitative methods. The journal of spinal cord medicine. 2002 Summer [PubMed PMID: 12137223]|
|||Ologunde R,Rabiu R, Stetho-hammer: a not-so-novel aid in the neurologic examination. Surgery. 2014 Nov [PubMed PMID: 24981587]|
|||Davis CR,Smith K,Ajuied A, iHammer. Surgery. 2014 Jan [PubMed PMID: 24094583]|
|||Ertuglu LA,Aydin A,Kumru H,Valls-Sole J,Opisso E,Cecen S,Türker KS, Jendrassik maneuver effect on spinal and brainstem reflexes. Experimental brain research. 2019 Dec; [PubMed PMID: 31650212]|
|||Litvan I,Mangone CA,Werden W,Bueri JA,Estol CJ,Garcea DO,Rey RC,Sica RE,Hallett M,Bartko JJ, Reliability of the NINDS Myotatic Reflex Scale. Neurology. 1996 Oct [PubMed PMID: 8857728]|
|||Hallett M, NINDS myotatic reflex scale. Neurology. 1993 Dec [PubMed PMID: 7802740]|
|||Shehab DK,Al-Jarallah KF,Abraham M,Mojiminiyi OA,Al-Mohamedy H,Abdella NA, Back to basics: ankle reflex in the evaluation of peripheral neuropathy in type 2 diabetes mellitus. QJM : monthly journal of the Association of Physicians. 2012 Apr [PubMed PMID: 22071964]|
|||Nick JM, Deep tendon reflexes, magnesium, and calcium: assessments and implications. Journal of obstetric, gynecologic, and neonatal nursing : JOGNN. 2004 Mar-Apr [PubMed PMID: 15095801]|
|||Bowditch MG,Sanderson P,Livesey JP, The significance of an absent ankle reflex. The Journal of bone and joint surgery. British volume. 1996 Mar [PubMed PMID: 8666641]|