Elbow flexion response as another primitive reflex
Abstract
Abstract In daily clinical practice we noticed that patients with intellectual impairment spontaneously flex the elbow within a few seconds of the forearm being manipulated during routine examination of spasticity of the muscles in the upper extremities. We termed this phenomenon elbow flexion response (EFR), and prospectively studied it in 229 patients who underwent in-hospital rehabilitation following brain damage. Evaluation of each patient included EFR, patient profile, ability to communicate, scores on three parameters from various intelligence tests, scores on seven parameters testing primitive reflexes, and scores on three parameters describing personality. We investigated for relationships among these parameters. Consequently, although EFR rarely have a statistical association with the varied profiles of patients, patients with bilateral lesion or bilateral paresis demonstrated significantly more marked EFR than those with unilateral lesion or unilateral paresis. Patients with involvement of the frontal lobe showed significantly more marked EFR than those without damage in this area. Elbow flexion responses occurred significantly more frequently in relation with lower scores on intelligence and occurred with significantly higher frequency in conjunction with the more marked appearance of conventional primitive reflexes. Therefore, we conclude that EFR have a strong association with intelligence and with the existence of frontal lobe lesion, and their mode of clinical presentation parallels that of primitive reflexes particularly that of the grasp reflex. We propose that EFR could be referred as a variation of the grasp reflex occurring in the more proximal or axial part of the body.
INTRODUCTION
In daily clinical practice we noticed that patients with intellectual impairment spontaneously flex their elbow within a few seconds of their forearm being manipulated during examination of spasticity of the muscles in the upper extremities; this examination is part of a routine neurological evaluation. We termed this phenomenon elbow flexion response (EFR) and hypothesized that it reflects brain dysfunction and is similar in nature to primitive reflexes.
SUBJECTS AND METHODS
Subjects
We studied patients who were admitted to the Department of Rehabilitation of Toya Kyokai Hospital, Abuta, Japan, between July 1996 and May 1998 for rehabilitation following brain damage. Patients who were readmitted during the 2-year study period were counted only once. Among the 250 consecutive patients who were enrolled initially, nine patients were excluded because they had a history of spinal cord disease (seven traumatic injuries, one rupture of arteriovenous malformation and one syringomyelia). Among the remaining 241 patients, 12 were also excluded because two patients did not give their consent, four patients were not available for a sufficient evaluation period, one patient had a defect of, or severe trauma to both wrists, and five patients terminated the rehabilitation program due to aggravation of their general physical state. Consequently, 229 patients participated in the study.
Profile of patients
The mean age at the time of study was 58.3 (SD 12.0) years old, the mean duration from onset to study was 44.2 (SD 52.1) months. Written informed consent was obtained from subjects or their guardians. Respective diagnosis of the patients were as follows: cerebral infarction (n = 105); intracranial haemorrhage (including subdural haematoma) (n = 96); subarachnoid haemorrhage with or without concomitant infarction or haemorrhage (n = 15); trauma in brain or brain tumour with or without surgery (n = 7); anoxia or carbon monoxide intoxication (n = 4); encephalitis (n = 1); malnutrition (n = 1).
Assessment of patients
Assessment of the patients included the following parameters: the ability to communicate; three parameters testing intelligence (IQ on Kohs block design test in its complete form,1,2 number of figures correctly reproduced in Benton’s visual retention test,3 and Mini-Mental State Examination (MMSE)); seven parameters testing for primitive reflexes (sucking, hand grasping, and tonic foot reflex for the affected and intact limbs, and snout reflex); three parameters on personality (raw score for extraversion, neuroticism, and lie scale in the Maudsley Personality Inventory (MPI), abbreviated as MPI-E, MPI-N, and MPI-L, respectively); and two parameters for EFR (AfEFR for the affected and InEFR for the intact limb). Although the condition of the limb (affected or intact) could usually be determined by the site of paresis, motor dysfunction such as spasticity, hyperreflexia, and pathological reflex were considered when paresis was not remarkable. The MPI was performed to assess the baseline personality component in anticipation that personality might affect EFR; anxiety is known to increase muscle tone.4 The Japanese version of MPI provides, in addition to the original extraversion and neuroticism scales,5 the lie scale, which was developed by Jensen and adapted for the Japanese language.6 We divided most parameters into grades; the method of grading for each parameter is shown in Table 1. The grades of the Kohs block design test, Benton’s visual retention test, and MMSE are abbreviated to ‘KohsGrade’, ‘BentonGrade’, and ‘MMSGrade’, respectively; the higher the grade, the higher the intelligence.
| Grade | 0 | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|
| Impairment in communication | – | No | Yes | – | – |
| EFR (InEFR, AfEFR) | None | Equivocal | Definite | – | – |
| Primitive reflexesa | None | Equivocal (biceps > triceps) | Present | – | – |
| Kohs | – | Impossible or IQ ≤ 55 | 55<IQ ≤ 80 | 80 < IQ ≤ 105 | 105 < IQ |
| Benton | – | Score ≤ 2 | 3 ≤ Score ≤ 5 | 6 ≤ Score ≤ 7 | 8 ≤ Score ≤ 10 |
| MMSE | – | Impossible or Score ≤ 7 | 8 ≤ Score ≤ 15 | 16 ≤ Score ≤ 22 | 23 ≤ Score |
- EFR, elbow flexion response; InEFR, intact EFR: AfEFR, affected EFR.
- a Sucking, hand grasp reflex, snout reflex, tonic foot reflex.
Elbow flexion response was assessed when we performed the repetition of swift supination and swift pronation of the forearm alternatively several times with intervals of 1 s for each maneuver, by manipulating the patient’s hand in the manner used to assess spasticity of supinator and pronator muscles in the forearm (Fig. 1). Elbow flexion response was judged to be ‘none’, ‘equivocal’, or ‘definite’ according to the degree of flexion. Every evaluation was performed prospectively according to the form prospectively designed for the present study. Kohs block design test, MMSE, and the grade of communication were assessed by an experienced speech therapist; Benton visual memory test, sucking, snout reflex, and palmomental reflex by a registered occupational therapist; tonic foot reflex by a registered physical therapist; and MPI by expert nurses specializing in rehabilitation. All of these tests were performed under the technical supervision of YG. Elbow flexion response was assessed by TM. Radiological evaluation to determine whether the lesion involved the frontal lobe was performed by KS’s reading of routine axial computed tomographs (CT). Each evaluation was performed carefully to avoid bias from results of other evaluations of the patient.
Elbow flexion response occurring in the affected side a 62-year-old woman suffering from mild right hemiparesis and hemisensory disturbance caused by an infarction of the left thalamus. (a) Resting state of the patient’s upper extremity ipsilateral to the hemiparesis. (b) Elbow flexion response provoked by alternative manipulation of forearm pronation and supination.
Missing values
Affected EFR was not obtained in three patients due to severe pain in the relevant upper extremities which was aggravated by motion. There were missing values in the Kohs block design test in 27 patients due to an insufficient period of evaluation. Owing to difficulty in communicating with patients, the MPI was completed in only 118 patients. We analyzed these data under the condition of excising these missing values.
Statistical analysis
We performed statistical analysis using the software Statview for Macintosh version 5.0. The association between EFR (InEFR and AfEFR) and the three parameters on intelligence tests was studied by non-paired t-test by grouping EFR ‘equivocal’ and ‘definite’ into EFR-positive, and EFR ‘none’ into EFR-negative. The associations between EFR and sex, impairment of communication, and involvement of the frontal lobe were studied by respective non-paired t-tests. Several analyses were performed by one-factor analysis of variance (ANOVA), including the association between EFR and profile (age, duration from onset to study, ability to communicate, and laterality of lesion or paresis), and between EFR and three parameters on personality. Next, we analyzed the association between EFR and three parameters of primitive reflexes (sucking reflex, hand grasp reflex, and tonic foot reflex of the intact limb) and four parameters of primitive reflexes (sucking reflex, hand grasp reflex, tonic foot reflex of the affected limb, and snout reflex), respectively, by one-factor ANOVA using a compound variable for each of these three or four parameters on primitive reflexes. Next, we examined the association between EFR (InEFR and AfEFR) and the existence of a frontal lobe lesion by non-paired t-test. In order to compare EFR of the intact and the affected sides, we performed a rank correlation test (Spearman’s). When analysis was done by one-factor ANOVA, Fisher’s protected least square difference was performed as a post-hoc test.
RESULTS
Intact EFR occurred more markedly as mean age became higher, while there was no significant association between the grade of EFR and age, sex, duration of the disease, and ability of communication (Table 2). The relation between EFR and patient profile is summarized in Table 3. Patients with bilateral lesion or bilateral paresis showed significantly more marked EFR than those of patients with unilateral lesion or unilateral paresis; and patients with involvement of the frontal lobe had significantly more marked EFR than patients without frontal lobe involvement (Table 3). The relation between EFR and parameters indicating intelligence is summarized in Table 4. Grades of both InEFR and AfEFR were significantly higher with lower scores of KohsGrade, BentonGrade, and MMSGrade. Elbow flexion responses occurred with significantly higher frequency in conjunction with the more marked appearance of conventional primitive reflexes (Table 5). In contrast, the association between EFR and the components of personality did not show any significance (Table 6). Elbow flexion responses in the intact and affected limbs were strongly correlated (P < 0.0001). Although we did not evaluate the effect of supination and pronation separately, TM, who took observed the examination, reported that there was no difference between the two procedures in the forearm.
| Grade of EFR | 0 | 1 | 2 | P |
|---|---|---|---|---|
| Age (mean, SD) | ||||
| InEFR | 53.5 (13.3) | 59.1 (12.7) | 62.7 (9.1) | <0.0001a |
| AfEFR | 58.7 (11.5) | 59.0 (11.9) | 62.0 (13.4) | NSa |
| Sex: male % | ||||
| InEFR | 66.1 | 74.2 | 70.2 | NSb |
| AfEFR | 69.0 | 69.0 | 75.0 | NSb |
| Duration (m), mean (SD) | ||||
| InEFR | 40.5 (44.2) | 42.1 (45.4) | 47.6 (60.0) | NSa |
| AfEFR | 43.3 (48.4) | 51.1 (65.1) | 36.7 (42.2) | NSa |
| Communication ability, mean (SD) | ||||
| InEFR | 1.15 (0.36) | 1.24 (0.43) | 1.29 (0.46) | NSa |
| AfEFR | 1.24 (0.43) | 1.21 (0.41) | 1.28 (0.46) | NSa |
- a One-factor ANOVA (InEFR 0 vs 1, 0.0056; InEFR 0 vs 2, <0.0001; InEFR 1 vs 2, 0.0467).
- b Non-paired t-test.
| InEFR (n = 229) | AfEFR (n = 226) | |||||
|---|---|---|---|---|---|---|
| NO. | Mean | (SD) | No. | Mean | (SD) | |
| Laterality of lesion | P = NSa | P = 0.0389a | ||||
| Right | 92 | 1.283 | (0.803) | 92 | 0.522c | (0.718) |
| Left | 107 | 1.168 | (0.830) | 104 | 0.529c | (0.737) |
| Bilateral | 30 | 1.033 | (0.850) | 30 | 0.900c | (0.845) |
| Laterality of paresis | P = NSa | P = 0.0200a | ||||
| None | 5 | 1.000 | (1.000) | 5 | 1.000d | (1.000) |
| Right | 99 | 1.172 | (0.833) | 96 | 0.510d | (0.740) |
| Left | 92 | 1.250 | (0.820) | 92 | 0.500d | (0.703) |
| Bilateral | 33 | 1.152 | (0.795) | 33 | 0.909d | (0.805) |
| Impairment of communication | P = 0.0550b | P = NSb | ||||
| Yes | 55 | 1.382 | (0.757) | 54 | 0.593 | (0.742) |
| No | 174 | 1.138 | (0.835) | 172 | 0.570 | (0.790) |
| Involvement of frontal lobe | P = 0.0141b | P = 0.0422b | ||||
| Yes | 92 | 1.359 | (0.806) | 90 | 0.700 | (0.771) |
| No | 137 | 1.088 | (0.818) | 136 | 0.493 | (0.730) |
- a One-factor ANOVA.
- b Non-paired t-test.
- c None vs bilateral, NS; none vs right, NS; none vs left, NS; none vs bilateral, NS; right vs left, NS; right vs bilateral, NS;
- left vs bilateral, NS.
- 7 None vs bilateral, NS; none vs right, NS; none vs left, NS; none vs bilateral, NS; right vs left, NS; right vs bilateral, 0.0082; left vs bilateral, 0.0070.
| InEFR | AfEFR | |
|---|---|---|
| KohsGrade | ||
| No. | 202 | 200 |
| F (p) | 1.408 (NS) | 1.103 (NS) |
| P | 0.0018 | 0.0007 |
| BentonGrade | ||
| No. | 229 | 226 |
| F (p) | 0.954 (NS) | 0.842 (NS) |
| t | 2.129 | 4.008 |
| P | 0.0343 | < 0.0001 |
| MMSGrade | ||
| No. | 229 | 226 |
| F (p) | 0.883 (NS) | 0.898 (NS) |
| t | 2.107 | 2.503 |
| P | 0.0362 | 0.0130 |
- InEFR, intact EFR; AfEFR, affected EFR.
- KohsGrade, grade of Kohs block design test; BentonGrade, grade of Benton’s visual retention test; MMSGrade, grade of Mini-Mental State Examination.
| InEFR | AfEFR | |
|---|---|---|
| Compand variable for primitive reflexes in the intact limb | ||
| NO. | 226 | 223 |
| m | 0.186 (EFR = 0) | 0.299 (EFR = 0) |
| 0.402 (EFR = 1) | 0.595 (EFR = 1) | |
| 0.611 (EFR = 2) | 0.694 (EFR = 2) | |
| F | 2.382 | 3.1114 |
| P | 0.0277a | 0.0051b |
| Compand variable for primitive reflexes in the affected limb | ||
| NO. | 226 | 223 |
| m | 0.251 (EFR = 0) | 0.421 (EFR = 0) |
| 0.564 (EFR = 1) | 0.637 (EFR = 1) | |
| 0.712 (EFR = 2) | 0.807 (EFR = 2) | |
| F | 2.630 | 2.184 |
| P | 0.0075c | 0.0265d |
- InEFR, intact EFR; AfEFR, affected EFR.
- a InEFR 0 vs 1, 0.0056; InEFR 0 vs 2, <0.0001; InEFR 1 vs 2, 0.0022.
- b AfEFR 0 vs 1, <0.0001; AfEFR 0 vs 2, <0.0001; AfEFR 1 vs 2, NS.
- c InEFR 0 vs 1, 0.0056; InEFR 0 vs 2, <0.0001; InEFR 1 vs 2, 0.0282.
- d AfEFR 0 vs 1, 0.0030; AfEFR 0 vs 2, <0.0001; AfEFR 1 vs 2, NS.
| Personality component in MPI | InEFR (n = 118) | AfEFR (n = 117) |
|---|---|---|
| Extraversion | ||
| F | 0.920 | 1.523 |
| m | 23.6 (EFR = 0) | 24.9 (EFR = 0) |
| 26.8 (EFR = 1) | 28.6 (EFR = 1) | |
| 26.4 (EFR = 2) | 26.5 (EFR = 2) | |
| P | NS | NS |
| Neuroticism | ||
| F | 1.737 | 1.915 |
| m | 18.2 (EFR = 0) | 19.6 (EFR = 0) |
| 19.7 (EFR = 1) | 22.1 (EFR = 1) | |
| 22.7 (EFR = 2) | 25.1 (EFR = 2) | |
| P | NS | NS |
| Lie scale | ||
| F | 1.623 | 3.080 |
| m | 20.0 (EFR = 0) | 19.5 (EFR = 0) |
| 19.2 (EFR = 1) | 18.1 (EFR = 1) | |
| 17.6 (EFR = 2) | 18.5 (EFR = 2) | |
| P | NS | NS |
- InEFR, intact EFR; AfEFR, affected EFR.
DISCUSSION
Certain kinds of brain-derived reflex phenomena such as glabellar tap, sucking reflex, snout reflex, hand grasp reflex, palmomental reflex, and tonic foot reflex are referred to as primitive reflexes.4,7 Emergence of primitive reflexes is described in infants as well as in people with intellectual impairment. They are regarded to be a type of pathologic reflex in adults,8–10 and are interpreted as phenomena resulting from diminution of the higher inhibitory control, as performed in particular by the frontal lobe.4,8 Although the neurophysiological basis of primitive reflexes is not yet well established,11 current results have confirmed that the nature of EFR is similar to that of conventional primitive reflexes. Therefore, the present findings suggest that EFR is a phenomenon that is likely mediated by the brain. However, in the differential diagnosis of EFR, flexor withdrawal reflex (FWR) must be ruled out because both of the two phenomena can be provoked by stimuli to the forearm. Although FWR is provoked by painful stimuli mediated by the spinal circuits,12 EFR in our study demonstrated a slower response time than that of spinal reflex. In addition, contrasted with FWR, EFR has been confirmed to bear a strong association with brain dysfunction. We did not determine in the present study, however, whether EFR were affected by pain because patients suffering from severe pain provoked by wrist manipulation had been excluded.
The fetus in the uterus lies curled up while in the endstage of life the same posture is assumed as the functioning brain deteriorates; neck, trunk, elbows, wrists, finger joints, hips, knees, and ankle joints, as well as toes, are flexed (see Fig. 7.2 in Adams et al.13) in both stages, primitive reflexes are known to appear.4,7 These observations suggest that the curled-up or fetal posture has a strong association with primitive reflexes and brain dysfunction, especially that occurring in the frontal lobe. We suspect that EFR might be provoked by disinhibition of the motor system controlled by lower brain activity, and furthermore that EFR can be assumed to be another primitive reflex. Grasp reflex, one of the typical primitive reflexes, shows flexion of all fingers provoked by stimulation of the palm, and tonic foot reflex also manifests flexion of the toes by stimulation of the sole, while EFR presents a flexion of the elbow by manipulation of the forearm. All of these phenomena are seen in the fetal posture. We postulate that pronation–supination procedure in the forearm might act the same as stimulation of the palm in the grasp reflex. We therefore propose that EFR could be assumed to be a variation, occurring in the more proximal or axial part of the body, of the grasp reflex.
Here, however, we must consider another important phenomenon attributed to the brain dysfunction, gegenhalten, which is a resistance to a passive stretch of a muscle. Because the function of biceps brachii is not only flexion of the elbow but also supination of the forearm, passive pronation of the forearm means extension of the muscle. In addition, gegenhalten increases with the velocity of the movement.14 Accordingly, resistance of the muscle on swift forearm pronation may provoke flexion of the muscle leading to elbow flexion. Interestingly, gegenhalten is often accompanied by a grasp reflex.14 Although we were not able to evaluate the effect of pronation and supination separately in the present study, there seems to be no association of EFR with supination of the forearm. Because there is a similarity in the causative lesion between gegenhalten and primitive reflexes, and because we were not able to exclude the possibility of some contribution of gegenhalten toward EFR, further neurophysiological study will be desirable to clarify the nature of EFR including its relationship with gegenhalten. Furthermore, instinctive grasp reaction can be easily ruled out,15 because only touch or rubbing to any part of the upper extremity did not provoke EFR.
Although primitive reflexes are suspected to appear in the limb contralateral to the brain lesion,4,8 the present results failed to confirm significant asymmetry in EFR between intact and affected limbs, even in the case that the brain lesion was asymmetric or localized; only in the intact limbs was EFR affected by age at the time of study. To explain the lack of asymmetry in EFR, we postulate that elbow flexion failed to manifest in the paralysed limb (although we did not estimate the degree of paresis in each patient in the present study) and in some cases the connecting fibers crossing over to the opposite hemisphere might be damaged simultaneously. In the present study, a simple manual procedure included in the minimal neurological check-up of the upper extremities did provide information about the state of brain function. We recommend that clinicians be aware of this phenomenon and employ these simple procedures to obtain preliminary information on brain function.
ACKNOWLEDGMENTS
We would like to thank the late Dr Naoki Ito (d. 25 November 1998) for his suggestion regarding the significance of EFR. We are indebted to four members of EFR Research Council in Toya Kyokai Hospital who took charge of the evaluation of patients: Mr Satoru Yorozuya, a speech therapist, Mr Takahisa Muraoka, a registered occupational therapist, Mr Akira Yamada, a registered physical therapist, and the registered nurses of the rehabilitation ward, Ms Yuriko Okuno and Ms Keiko Saito.
