Chronic Idiopathic Axonal Polyneuropathy and Successful Aging of the Peripheral Nervous System in Elderly People FREE

AFTER extensive investigations, no cause can be identified in 10% to 20% of the patients who have polyneuropathy.^{1 - 5} In most of these patients the polyneuropathy presents insidiously in the sixth decade of life with predominantly sensory or sensorimotor symptoms and a slowly progressive course. Electrophysiologic studies invariably show axonal polyneuropathy. For this clinical entity the term "chronic idiopathic axonal polyneuropathy" (CIAP) has been introduced.⁶ Others have used "cryptogenic (sensory) polyneuropathy."^{3 ,5} Severe disability or ambulatory impairment has not been observed 5 to 10 years after disease onset.^{2 - 3 ,5 - 6}

Neurologic symptoms and signs suggestive of polyneuropathy occur more frequently with successful aging.^{7 - 12} Also, there is a decline of nerve conduction parameters.^{8 ,11 ,13 - 19} Furthermore, age-related degenerative changes have been demonstrated in pathologic studies of peripheral nerves in subjects without peripheral nervous system disease.^{20 - 21}

The age-dependent decline could affect the clinical course of CIAP with differential consequences regarding disability and prognosis for younger and older patients. However, it is unknown whether CIAP with onset at a younger age differs from CIAP with onset at an older age. Moreover, the disease duration could obscure the influence of aging. To our knowledge, the severity of CIAP after a long disease duration (which is more likely to occur in patients with disease onset at a younger age) has not been studied before. The aims of this study were to determine if CIAP is influenced by the superposition of the effects of aging and to evaluate the severity of CIAP according to disease duration. We compared the neurologic symptoms and signs, disability status, and electrophysiologic parameters of 127 patients with CIAP with those of 108 successfully aged control subjects.

A total of 127 patients (87 men and 40 women) with CIAP and 108 successfully aged controls (72 men and 36 women) were included. Fifty-seven patients with CIAP also participated in a quality-of-life study.²² Patients were referred to our outpatient department for neuromuscular diseases after previous evaluation by other neurologists. The diagnosis of CIAP was made according to the following criteria^{2 ,4} : (1) symmetrical distal sensory or sensorimotor symptoms and signs of the limbs, compatible with polyneuropathy; (2) insidious onset and slow or no progression of the disease over at least 6 months; (3) no identifiable cause after extensive clinical and laboratory investigations; (4) no indication of hereditary polyneuropathy; and (5) electrophysiologic findings that agreed with the diagnosis of axonal polyneuropathy.

Controls were recruited from relatives or friends of patients visiting our clinic and had to be free of diseases or conditions known to be associated with polyneuropathy. They had not previously consulted a neurologist for symptoms or signs.

Patients were classified into 2 groups—those with disease onset before the age of 65 years (hereafter referred to as the "early-onset CIAP group" [n = 61 patients]) and those with disease onset at 65 years or older (hereafter referred to as the "late-onset CIAP group" [n = 66 patients]). Disease onset was defined as the age at which the symptoms or signs were first noted by the patient to be consistently present. Controls were classified into a group of persons younger than 65 years at study enrollment (hereafter referred to as the "young controls" [n = 47 subjects]), and a group of persons 65 years or older at study enrollment (hereafter referred to as the "old controls" [n = 61 subjects]). The cutoff was set at the age 65 years, because it coincided with the usual retirement age and health expectancy (ie, the average age at which a person can still be expected to be disability-free or in good self-reported health) in the Netherlands.²³

To evaluate if early-onset CIAP is different from late-onset CIAP, the data at referral were compared. To study the relationship between the disease duration and the severity of CIAP, the latest clinical data of 89 patients who had a short disease duration (<10 years) were compared with the data of 38 patients who had a long disease duration (≥10 years). To determine the normal manifestations of successful aging, young controls were compared with old controls. To evaluate the differences between CIAP and successful aging, patients with late-onset CIAP were compared with old controls. Comparison of patients with early-onset CIAP with young controls was considered inappropriate, because part of the patients with early-onset CIAP had already reached the age of 65 years at the time of study enrollment. The study protocol was approved by the Medical Ethical Committee of the University Medical Center Utrecht, Utrecht, the Netherlands.

In all patients and controls, history taking and clinical examination (muscle strength, sensory function, tendon reflexes, and the Romberg test) were performed in a standardized fashion.^{4 ,22} A detailed inquiry was made about specific motor and sensory symptoms (Table 1), medical history, medication use, and toxic substance exposure. The occurrence of polyneuropathy in family members was thoroughly queried, and affected family members underwent neurologic and electrophysiologic examination, if necessary. If this could not rule out the possibility of a hereditary neuropathy, subjects were excluded.

Manual muscle strength testing was scored with the Medical Research Council (MRC) scale in the subject's forearms, hands, lower legs and feet (ie, wrist extensors, finger extensors, finger flexors, interossei muscles, anterior tibial muscle, gastrocnemius muscle, peroneal muscles, and the long extensor muscle of the big toe). Summation of the MRC scores could total a maximum of 80. Sensory function was examined and scored bilaterally. Pinprick and light touch sense were scored as follows: normal, 4; distal to wrist/ankle abnormality, 3; distal half forearm/leg abnormality, 2; distal to elbow/knee abnormality, 1; and above the elbow/knee abnormality, 0. Vibration sense (128-Hz tuning fork perception) was scored as follows: normal perception on middle finger (at least 15 seconds)/or big toe (at least 10 seconds), 4; decreased or absent on middle finger/big toe, but with perception on styloid/malleolus, 3; absent on styloid/malleolus, but with perception on elbow/knee, 2; absent on elbow/knee, but with perception on clavicle/crista iliaca, 1; and absent on clavicle/crista iliaca, 0. Joint position sense (perception of movement at distal interphalangeal joint of middle finger/big toe)was scored as follows: immediate or normal perception, 2; delayed or diminished perception, 1; and absent perception, 0. Summation of all sensory scores could total a maximum score of 56. Sensory scores were dichotomized at the following values: pinprick/light touch/vibration sense below 7 (abnormal) vs 7 or 8 (normal), joint position sense below 3 (abnormal) vs 3 or 4 (normal). Disability status was scored using the modified Rankin scale.²⁴

Extensive laboratory tests were performed in all patients to exclude any cause of the polyneuropathy.^{1 - 2 ,4} Additional studies (eg, radiography of the chest or spinal axis or cerebrospinal fluid analysis) were done when indicated and the findings or results had to show no abnormality or be within a normal range. Sural nerve biopsy was performed if there was any suggestion, based on clinical or laboratory findings, of an underlying disorder such as vasculitis or amyloidosis. Laboratory investigations were not performed in controls.

Nerve conduction studies with surface electrodes and concentric needle electromyography according to standard techniques²⁵ were consistent with axonal polyneuropathy in all patients. In 96 patients and 49 controls additional standardized conduction studies were performed after warming the participants' limbs for 30 minutes in water of 37°C.^{26 - 27} Motor conduction was investigated in the median nerve up to the axilla, and in the tibial nerve up to the knee; sensory conduction on distal stimulation was investigated in the median and sural nerve.²⁷ Concentric needle electromyography was performed in the anterior tibial muscle. Criteria for demyelination were as described previously.²⁷

Frequencies were compared with χ² test statistics. For differences between not normally distributed continuous variables (ie, age and disease duration) and ordinal or scaled variables (ie, motor and sensory sum scores, modified Rankin scores), the Kolmogorov-Smirnov test or Mann-Whitney test were used, respectively. The means of normally distributed variables (electrophysiologic parameters) were compared using the t test. Statistical significance was set at P<.05.

Age distributions at study enrollment of patients and controls are shown in Figure 1. The men-women ratio was 2:1 in patient and control groups.

Profiles of symptoms and signs are shown in Figure 2A. In all patients the polyneuropathy had started distally in the legs; at the time of referral, 60 patients had symptoms distally in the legs and hands.

All patients had distal sensory signs in the legs, and 55 had sensory signs in the hands. Weakness of distal leg muscles was noted in 80 patients, and 17 also had weakness of the hand muscles. A total of 62 patients with signs distally in the legs also had signs in the hands.

Thirty-five patients used walking aids such as a cane, adjusted shoes, ankle-foot orthoses, or a wheeled walker. One patient used a wheelchair because of a hemiparesis from cerebral infarction. The scores on the modified Rankin scale were as follows: 1, 70 patients; 2, 52 patients; and 3, 5 patients. No patients had modified Rankin scores of 4 or 5.

Patients with early-onset CIAP significantly more often had tingling or prickling, pain, numbness or a cotton-wool sensation, muscle cramps, stiffness, and weakness (Table 1), but there were no statistically significant differences in the presence of signs (Table 2).

In patients with early-onset CIAP who had a short disease duration, modified Rankin scores were significantly higher than in patients with late-onset CIAP who had a short disease duration (Table 3). In patients with late-onset CIAP who had a short disease duration, sensory and motor total scores of the legs were significantly higher, and modified Rankin scores were significantly lower than in patients with late-onset CIAP who had a long disease duration (Table 3).

Symptoms and signs were also found in controls, and more often in old controls (Figure 2B). In 54 controls symptoms were located distally in the legs, and 26 (6 young and 20 old controls; P≤.01) had symptoms in the hands as well.

In 70 controls (25 young and 45 old controls; P≤.02) sensory signs were found distally in the legs, and 17 (5 young and 12 old controls; P≤.01) also had sensory signs in the hands. Muscle strength of the arms and legs was normal in all but 2 controls (ages 73 and 74 years) who had an MRC grade 4 weakness of the long extensor muscle of the big toe or anterior tibial muscle.

The walking aids used by 2 controls (ages 79 and 80 years) were a cane and a wheeled walker. The modified Rankin scores for the control subjects were as follows: 0, 51 controls; 1, 54 controls; and 2, 3 controls. No controls had a modified Rankin score of 3 or higher.

Old controls significantly more often mentioned tingling or prickling and weakness (Table 1). Furthermore, old controls had significantly lower sensory total scores of the legs (Table 2), and they significantly more often had abnormal light touch and vibration sense, and absent ankle tendon reflexes (Figure 3; sensory functions were considered abnormal if the sensory scores in the arms or legs were below 7 for pinprick, light touch, or vibration sense, and below 3 for position sense).

In general, patients with late-onset CIAP significantly more often had symptoms and signs as well as significantly lower sensory and motor total scores (Table 1 and Table 2). However, old controls significantly more often had pure motor symptoms (0 patients with late-onset CIAP and 23 old controls; P<.001) or pure sensory signs (23 patients with late-onset CIAP and 46 old controls; P<.001).

In the legs of old controls pinprick (75%), light touch (82%), and vibration sense (83%) abnormalities were restricted to the feet, whereas in patients with late-onset CIAP these rates were significantly lower, that is, 38%, 47%, and 63%, respectively (P<.001 for all rates). Patients with late-onset CIAP significantly more often used walking aids (P≤.01) and had significantly higher modified Rankin scores (P<.001).

None of the electrophysiologic parameters in patients or controls fulfilled our criteria for demyelination. Table 4 summarizes the findings of the standardized electrophysiologic studies in 96 patients with CIAP and 49 controls.

In patients with late-onset CIAP the mean distal motor latency of the tibial nerve was significantly longer. Absence of the sural nerve sensory nerve action potential (SNAP) or the presence of spontaneous muscle fiber activity (ie, fibrillation potentials, positive sharp waves, or complex repetitive discharges) in the anterior tibial muscle were found in most patients (72%) (33 patients with early-onset CIAP and 36 patients with late-onset CIAP).

In old controls the mean distal compound muscle action potential amplitude of the tibial nerve and the mean distal SNAP amplitudes of the median nerve and sural nerve were significantly lower. The SNAP of the sural nerve was absent in 1 old control (age 84 years). In another old control (age 67 years) concentric needle electromyography of the anterior tibial muscle showed positive sharp waves at 1 insertion site.

In patients with late-onset CIAP the mean distal motor latencies were significantly longer, and the mean distal compound muscle action potentials, SNAP amplitudes, the mean motor nerve conduction velocities and sensory nerve conduction velocities were significantly lower. Absence of the sural nerve SNAP or the presence of spontaneous muscle fiber activity in the anterior tibial muscle were found significantly more often in patients with late-onset CIAP.

There were no important differences in clinical features, electrophysiologic parameters, or disease course between patients with early-onset and late-onset CIAP. Patients with early-onset CIAP who had a short disease duration had worse modified Rankin scores than patients with late-onset CIAP who had a long disease duration. Basic activities of daily living appear fairly resistant to aging, whereas complex or instrumental activities of daily living (eg, preparing meals, housekeeping, occupational skills) are less so.²⁸ The modified Rankin score is an overall disability scale that integrates these levels of activities of daily living. The polyneuropathy probably has a more negative influence on the daily life of the nonretired patients with early-onset CIAP, although they are not more severely affected than the retired patients with late-onset CIAP. The low total scores of the legs and the worse modified Rankin scores in patients with late-onset CIAP who had a long disease duration are fallible because of the small sample size. The advanced age (at least 75 years old) of patients with late-onset CIAP who had a long disease duration is less likely to provide an explanation, because the scores are equally low in patients with early-onset CIAP who had a similar disease duration. Moreover, as we have shown, age does not importantly affect CIAP, nor are the neurologic impairments or disability in successfully aged old controls of such severity that they can be expected to have a significant influence. In the successfully aged person 65 years and older, there is a decline of sensory function distally in the legs, ankle reflexes are more often absent, and electrophysiologic studies most notably show lower distal compound muscle action potentials and SNAP amplitudes.

Similar to others,^{7 ,11 ,29} we evaluated each sensory modality by simple techniques. Measurement of the perception time of the vibratory stimulus from a 128-Hz tuning fork was validated earlier by De Michele et al.³⁰ Our normal perception time values of 15 seconds at the middle finger and 10 seconds at the big toe are of similar magnitude as their mean perception times at the ankle and wrist. Hence, our normal values likely classify a larger proportion of elderly persons as having abnormal vibration sense at the feet. Others have advocated the use of the Rydel-Seiffer graduated tuning fork and contended that absent vibration sense at the feet in healthy elderly subject should be considered abnormal.³¹ However, individuals with absent vibration perception in at least one of the sites tested were excluded prior to the analysis. Because vibration perception at the big toe was absent in 28% of our old controls, we are less inclined to denote this as abnormal in elderly persons. Whatever sensory testing techniques used, many studies most notably show a decline of light touch and vibration sense with advancing age.^{7 - 11 ,30 - 32}

Paradoxically, old controls mentioned weakness more often than could be confirmed by neurologic examination. Weakness on standard neurologic examination is unusual in successful aging, but a decline of muscle strength can been demonstrated by more sensitive quantitative techniques (ie, isometric or isokinetic strength testing).³³

The biceps, triceps, and knee jerks are usually retained after the age of 65 years.^{9 - 11 ,34} Some investigators have asserted that the ankle reflex should also usually be present after the age of 65 years,^{35 - 36} but others have found it to be absent as frequently as in our study.^{7 - 11 ,34 ,37} The effort made to obtain the ankle jerk (eg, reinforcement methods) and the selection and age distribution of controls may contribute to observed response rates. Because the ankle jerk may often be absent in persons older than 65 years, it is not a good indicator for the presence of polyneuropathy in elderly people.

The lower mean distal SNAP amplitudes and compound muscle action potentials in old controls compared with young controls are compatible with normal aging.^{8 ,11 ,13 ,15 - 19} We did not find a change with age in distal motor latencies or nerve conduction velocities unlike others,^{11 ,14 ,18 - 19} which may be because of the inclusion of younger subjects in those studies, and our rigorous temperature control.

Because absence of the sural nerve SNAP or presence of spontaneous muscle fiber activity were exceptional in controls but common in patients with CIAP, either one of these phenomena probably best supports the diagnosis of axonal polyneuropathy in a clinically suspect individual. Instead of using standard electrode placement, we palpated the sural nerve at the ankle and placed the recording electrode over the nerve, which may have contributed to our findings on the preservation of the sural nerve responses. Also, our findings are reinforced by other studies that have shown that sural nerve SNAP is absent in only small proportions (0%-9%) of the control subjects, even after the age of 60 years.^{14 - 16 ,19 ,38 - 44} Studies on the normal electromyography of the intrinsic foot muscles of controls have shown conflicting results regarding the presence of spontaneous muscle fiber activity.^{45 - 46} Therefore, we performed electromyography in the anterior tibial muscle, although we are unaware of normal electromyography studies of this muscle.

The superposition of aging in the peripheral nervous system does not importantly affect CIAP. Even after more than 10 years of having the disease, severe disability (modified Rankin score >3) does not occur. Axonal polyneuropathy can probably be confirmed best on the basis of the absence of the sural nerve SNAP or the presence of spontaneous muscle fiber activity in the anterior tibial muscle. In successfully aged subjects older than 65 years, mild pure sensory or pure motor symptoms, slight and pure sensory signs confined to the feet (including absent vibration sense at the big toe), or absent ankle reflexes can be considered as normal aging manifestations. Our findings are valuable when informing patients with CIAP about the prognosis. In view of the aging population, our findings stress the importance of clinical criteria by which symptoms or signs may be attributed to normal aging or conversely to polyneuropathy.

Accepted for publication September 25, 2001.

Author contributions: Study concept and design (Drs Vrancken and Notermans); acquisition of data (Drs Vrancken, Franssen, Teunissen, and Notermans); analysis and interpretation of data (Drs Vrancken, Franssen, and Wokke); drafting of the manuscript (Dr Vrancken); critical revision of the manuscript for important intellectual content (Drs Franssen, Wokke, Teunissen, and Notermans); statistical expertise (Drs Vrancken and Franssen); study supervision (Drs Franssen, Wokke, and Notermans).

Corresponding author: Alexander F. J. E. Vrancken, MD, Department of Neurology, C03.236, University Medical Center Utrecht, PO Box 85500, 3508 GA Utrecht, the Netherlands (e-mail: a.f.j.e.vrancken@neuro.azu.nl).