- Email: [email protected]
Changes in measures of motor axon excitability with age
Clinical Neurophysiology 118 (2007) 1397–1404 www.elsevier.com/locate/clinph
Changes in measures of motor axon excitability with age S.K. Jankelowitz *, P.A. McNulty, David Burke Institute of Clinical Neurosciences, University of Sydney and Royal Prince Alfred Hospital, Sydney, Australia Accepted 21 February 2007 Available online 23 April 2007
Abstract Objective: Threshold tracking is a novel technique that permits examination of the excitability of human axons in vivo. Protocols have been validated for sensory and motor axons, but there are limited data on the changes in the excitability of motor axons with age. This study aimed to determine such changes from the third to the eighth decades. Methods: Sixty healthy subjects aged 22–79, 10 per decade, were studied using the TRONDXM4 protocol of the QTRAC thresholdtracking program to assess motor axon function. The median nerve was stimulated at the wrist and the compound muscle action potential was recorded from the thenar muscles. Results: There was an increase in threshold in elderly subjects, associated with a decrease in slope of the stimulus–response curves. Strength-duration time constant and threshold electrotonus to depolarising and hyperpolarising currents of up to 40% did not change significantly with aging. The current–threshold relationship was similar across all decades for subthreshold depolarising currents, but the slope of the current–threshold relationship was significantly steeper the older the subjects for hyperpolarising currents, particularly those greater than 40% of threshold. There was also a significant decrease in supernormality in the recovery cycle with increasing age. Conclusions: The threshold of axons increases with age and the extent of supernormality decreases. There may also be greater inward rectification in motor axons, perhaps due to greater activity of IH, the hyperpolarisation-activated conductance, though this is only significant with hyperpolarising currents greater than 40% of the threshold current. Significance: Many indices of axonal excitability, such as strength-duration time constant, the relative refractory period, late subnormality, threshold electrotonus and the depolarising side of the current–threshold relationship, do not change significantly with age. For other indices, age-related changes may be due to a combination of non-neural factors that alter current access to the node of Ranvier, changes in the axon and its myelination and, possibly, changes in channel activity and/or changes in extracellular [K+]o. 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Axonal excitability; Age; Threshold tracking; Supernormality; Inward rectification
1. Introduction Changes occur throughout the peripheral nervous system with age. For example, age-related structural changes include neuronal loss in the dorsal and ventral columns (Kawamura et al., 1977; Tomlinson and Irving, 1977), fibre loss in peripheral nerves, especially thick myelinated fibres, and changes in internodal length (Lascelles and Thomas, * Corresponding author. Address: Medical Foundation Building K-25, University of Sydney, Sydney, NSW 2006, Australia. Tel.: +61 2 9036 3091; fax: +61 2 9036 3092. E-mail address: [email protected] (S.K. Jankelowitz).
1966; Jacobs and Love, 1985; Vital et al., 1990). The importance of these changes increases with age, becoming noticeable in the fifth decade and significant in the seventh. A reduction in the number and density of human myelinated fibres with old age has been demonstrated in spinal roots (Kawamura et al., 1977) and in the radial (O’Sullivan and Swallow, 1968), ulnar (Rafalowska et al., 1976), sciatic (Takahashi, 1966), peroneal (Stevens et al., 1973), tibial (Swallow, 1966) and sural nerves (O’Sullivan and Swallow, 1968; Jacobs and Love, 1985). There are also changes in peripheral motor unit function with age: muscle mass/area is reduced and there may be an increase in slow (type I) muscle fibres compared to fast (type II) fibres. Motor force
1388-2457/$32.00 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2007.02.025
1398
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
and strength are reduced by 20–40% in subjects after the age of 70 although changes are observed from the sixth decade. A single motor axon innervates more muscle fibres, producing polyphasia on macro EMG (see Roos et al., 1997 for review). Touch-pressure, vibratory and cooling detection thresholds increase with age, and there is a reduction in the density and size of Meissner corpuscles in the fingers (Bolton et al., 1966) and the plantar surface of the big toe (Schimrigk and Ru¨ttinger, 1980). The threshold for perception of electrical stimuli increases with age (Nadler et al., 2002). Axonal excitability studies using threshold tracking are a recent addition to the neurophysiological armamentarium available to assess peripheral nerve function. They are now being employed to determine changes in excitability of axons in various pathologies to help determine the underlying pathophysiology. There has been, so far, no systematic analysis of age-related changes in the different measures of the excitability of motor axons, apart from a recent study showing no significant correlation between strengthduration time constant and age (Tamura et al., 2006). There has also been a study that documented the changes in excitability of sensory axons with age (Kiernan et al., 2001). Many pathologies affect only the elderly or are more severe the older the subject, and it is therefore relevant whether different excitability indices vary with age. This study was undertaken to determine if excitability indices of motor axons change with age, to determine the relationship with age and, hopefully, to shed light on the mechanisms. 2. Methods 2.1. Subjects Nerve excitability studies were performed on motor axons in 60 subjects of both genders, from 22 to 79 years, 10 per decade, 28 males and 32 females (Table 1), using the protocol and techniques described in full elsewhere (Bostock et al., 1998; Kiernan et al., 2000). The subjects provided informed consent to the procedures that had the approval of the Human Research Ethics Committee of the University of Sydney. For all experiments, skin temperature was monitored close to the stimulation site and was maintained above 32 C using blankets.
The median nerve was stimulated via non-polarisable electrodes (Maersk Medical, Stonehouse, UK) with the cathode over the median nerve at the wrist and the anode 10 cm proximal over muscle on the radial aspect of the arm. Compound muscle action potentials (CMAPs) were recorded from thenar muscles using surface electrodes over abductor pollicis brevis (APB) with the active electrode at the motor point and the reference electrode on the proximal phalanx. The electromyographic (EMG) signal was sampled at a rate of 10 kHz, amplified, filtered (3–500 Hz), and digitized by computer (Pentium PC) with an analogue-to-digital board (DT2812, Data Translation, Marlboro, Massachusetts). Stimulus waveforms were generated by the computer and converted to current by an isolated linear bipolar constant-current source (maximal output ± 50 mA). Motor axons were studied using the multiple excitability protocol, TRONDXM4 (QTRAC version 8.2, Professor H. Bostock, Institute of Neurology, London), as described elsewhere (Kiernan et al., 2000). Test current pulses of 0.2or 1-ms duration were delivered regularly at 0.8-s intervals to produce a target potential that was on the fast rising phase of the stimulus–response curve, 40% of the maximal CMAP. The computer was programmed to maintain the test potential at this target size despite changes in stimulus duration or the delivery of conditioning stimuli, either suprathreshold stimuli or subthreshold polarising currents. The amplitude of the CMAP was measured from baseline to the negative peak. Stimulus–response curves were recorded separately for test stimuli of 0.2- and 1-ms duration. These data were used to optimise threshold tracking (Kiernan et al., 2000) and to calculate the strength-duration time constant (sSD) according to Weiss’ Law, based on the linear relationship between stimulus charge (mA · ms) and stimulus duration. The sSD is a nodal property and reflects the rate of decrease of threshold current as the duration of the stimulus pulse increases. Prolonged subthreshold currents were used to alter the potential difference across the internodal membrane. The changes in threshold associated with electrotonic changes in membrane potential are termed threshold electrotonus and generally reflect the underlying changes in membrane potential. The subthreshold polarising currents were of 100-ms duration and set to be +40% and +20% (depolarising) and 40% and 20% (hyperpolarising) of the control
Table 1 Demographics and nerve conduction indices Decade
Mean age (range)
Sex ratio M:F
CMAP amplitude (mV)
Onset latency (ms)
Mean threshold 50% CMAP (mA)
20–29 30–39 40–49 50–59 60–69 70–79 Change with age
26.3 31.4 42.8 54.7 63.4 74.5
6:4 3:7 5:5 5:5 3:7 6:4
8.53 ± 0.92 9.88 ± 1.05 10.05 ± 0.92 8.46 ± 0.72 9.27 ± 1.08 9.53 ± 1.00 No significant change
2.78 ± 0.11 2.80 ± 0.08 2.64 ± 0.12 2.80 ± 0.12 3.03 ± 0.12 4.03 ± 0.20 Quadratic R2 = 0.21, P = 0.001
8.49 ± 0.71 8.90 ± 0.74 9.73 ± 1.61 10.39 ± 1.53 12.24 ± 1.91 15.83 ± 1.64 Quadratic R2 = 0.24, P < 0.001
Data are mean ± SEM.
(22–29) (30–35) (41–47) (50–59) (50–59) (70–79)
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
threshold current (i.e., the current required to produce the unconditioned target CMAP, 40% of maximum). Threshold was tested at different time points during and after the 100-ms polarising currents. Five stimulus combinations were tested in turn: test stimulus alone to measure the control threshold, test stimulus plus depolarising current of +20% or +40%, and test stimulus plus hyperpolarising current of 20% or 40%. Threshold was then measured with 1-ms test stimuli during a long-lasting (220 ms) subthreshold polarising current, 200 ms after its onset. The strength of the conditioning current was altered in 10% steps from +50% (depolarising) to 100% (hyperpolarising) of the control threshold. This produced a current–threshold relationship (analogous to the current–voltage relationship) that depends on the rectifying properties of the internodal axolemma. For both threshold electrotonus and the current–threshold relationship, each stimulus combination was repeated until three threshold measures were within 10% of the target response. The final part of the protocol measured the recovery of axonal excitability following a supramaximal conditioning stimulus. Recordings were made at 18 conditioning-test intervals from 2 to 200 ms. The stimulus combinations employed were unconditioned test stimulus of 1-ms duration tracking the control threshold, supramaximal 1-ms conditioning stimulus alone, and conditioning plus test stimuli. The response to the conditioning stimulus alone was subtracted from the response to both stimuli so that the conditioning maximal CMAP did not contaminate the measured response when the conditioning-test interval was short. Each stimulus combination was repeated until four valid threshold estimates were obtained. The results for stimulus threshold, strength-duration time constant, threshold electrotonus, the current–threshold relationship and the recovery cycle were compared across the decades. Statistical analyses were performed using SPSS version 13 (SPSS Corporation Chicago, USA). One-way ANOVA and paired t-tests were per-
1399
formed with Bonferroni corrections where appropriate, and corrected probability values are quoted in the text. Curve estimation and regression analysis were performed using SPSS to determine whether age-related changes were best fitted by a linear, exponential or polynomial function. Where more than one function was significant, the one with the higher R2 value has been reported. In addition, whether the data were linear or not was confirmed by performing regression analyses comparing one decade with each of the other decades in turn. Using the b coefficient, assessment was made whether the observed change was progressive (i.e., linear) or not. Significance was taken at P < 0.05. All data are given as means ± SEM. 3. Results No subject had clinical or neurophysiological evidence of carpal tunnel syndrome. The amplitude of the CMAP was fairly stable across the decades ranging from 8.53 ± 0.92 mV for the third decade to 9.53 ± 1.00 mV for the eighth decade (Table 1). Mean latency to the onset of the CMAP increased from 2.78 ± 0.11 ms in the third decade to 4.03 ± 0.02 ms in the eighth decade (quadratic function, R2 = 0.21, P = 0.001). The threshold for a 50% maximal CMAP increased from 8.49 ± 0.71 mA to 15.83 ± 1.64 mA from the third to the eighth decade (Table 1, Fig. 1a; see next section). For all excitability parameters, there was little difference between the results for the third and fourth decades (though this may have been due to the narrow age range of subjects in the fourth decade; see Table 1). Changes usually became evident in the fifth decade. Tables 2 and 3 summarise the age-related data for the measures of excitability discussed below. 3.1. Stimulus–response curves The stimulus–response curves showed an increase in threshold with age (Fig. 1a). Not only did the curve shift
Fig. 1. Stimulus–response curves. There is an increase in stimulus threshold and a decrease in slope of the stimulus–response curve with age. (a) The mean stimulus–response curves plotted against the absolute stimulus intensity. (b) The stimulus–response curves with stimulus intensity normalised so that the stimulus required for a 50% maximal CMAP is equal to 1. (Black circle = 20’s, white circle = 30’s, grey circle = 40’s, black triangle = 50’s, white triangle = 60’s, grey triangle = 70’s.)
1400
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
Table 2 Indices of axonal excitability for each decade Decade 20–29 30–39 40–49 50–59 60–69 70–79 Change with age
SDTCa (ms) 0.42 ± 0.02 0.52 ± 0.03 0.39 ± 0.02 0.43 ± 0.03 0.46 ± 0.03 0.40 ± 0.03 No significant change
Rheobase (mA)
Refractoriness at 2.5 ms (%)
2.50 ± 0.20 2.43 ± 0.21 3.18 ± 0.55 3.34 ± 0.56 3.74 ± 0.74 4.59 ± 0.54 Quadratic R2 = 0.17, P = 0.005
(n=8)
46.02 ± 5.57 38.06 ± 9.89(n=9) 22.70 ± 8.64(n=9) 20.05 ± 4.28(n=9) 16.34 ± 6.23 26.66 ± 9.99 Quadratic R2 = 0.20, P = 0.003
Relative refractory period (ms)
Supernormality (%)
3.37 ± 0.09 3.20 ± 0.12 2.91 ± 0.15 2.99 ± 0.08 2.88 ± 0.12 3.35 ± 0.37 No significant change
27.54 ± 1.23 28.68± 1.54 22.49 ± 2.55 23.53 ± 1.22 24.24 ± 1.22 18.51 ± 2.46 Linear R2 = 0.19, P < 0.001
Data are mean ± SEM. N = 10, except for refractoriness, where n is indicated in the superscript. a SDTC, strength-duration time constant.
to the right, but the slope of the normalised stimulus– response curve became less steep, especially close to the maximal CMAP amplitude, i.e., high-threshold axons required an even greater stimulus with age (Fig. 1b). The change in stimulus threshold became noticeable at, and gradually increased from, the fifth decade (Fig. 1; Table 1). The change in threshold (for the 50% CMAP) was significantly related to age in linear and quadratic functions, but the better correlation was with the quadratic relationship (R2 = 0.24, P < 0.001). Not surprisingly, the increase in current threshold was associated with an increase in rheobase with age (for the 50% CMAP, range: 2.50 ± 0.20 mA to 4.59 ± 0.54 mA, Fig. 2b; Table 2). The greatest change in rheobase was observed in the eighth decade. Again, the data could be fitted by a linear or quadratic relationship, but the better fit was a quadratic function (R2 = 0.17, P = 0.005). There was no consistent change in sSD with age, 0.42 ms in the third decade and 0.39 ms in the eighth decade (Fig. 2a). There was no clear relationship between sSD and rheobase, suggesting that the increase in rheobase (and presumably other threshold measures) was not solely due to a change in excitability. 3.2. Recovery cycle The recovery cycle examines the change in current required to produce the target CMAP following a supramaximal conditioning stimulus and, for each decade, there was a normal pattern. Axons were relatively refractory at short interstimulus intervals (2–3 ms), then supernormal,
maximally so at 6–7 ms, and finally subnormal, maximally at the sampled interstimulus intervals of 32–56 ms. The extent of refractoriness at the 2.5-ms interval was greatest in the third decade, and there was a significant difference in the degree of refractoriness (but not in the duration of the relative refractory period) with age (data for 2.5-ms interstimulus interval are given in Table 2; quadratic relationship, R2 = 0.20, P = 0.003). Supernormality is a reflection of the depolarising afterpotential, and late subnormality reflects the time course of voltage-gated slow potassium channels activated by the conditioning discharge (Taylor et al., 1992; Bostock et al., 1998; Lin et al., 2000). Supernormality was measured as the mean percentage decrease in threshold seen at conditioning-test intervals of 5, 6.3 and 7.9 ms. There was a significant reduction in supernormality with age (Fig. 3b; Table 2). The reduction in supernormality was greatest between the seventh and the eighth decades, but linear regression provided the best correlation (R2 = 0.19; P < 0.001). There was a significant correlation between peak supernormality and the depolarising peak reached in the threshold electrotonus plot at 20 ms (R2 = 0.32; P < 0.001). This latter finding is not unexpected given that the two measures share common mechanisms, but surprisingly the age-related change in the depolarising peak at 20 ms was not significant (P = 0.400, see below, Table 2). There was a trend for late subnormality, measured as the mean threshold increase at conditioning-test intervals of 32, 42 and 56 ms, to decrease with age, but this change was not significant (P = 0.060).
Table 3 Threshold electrotonus and the current–threshold plot for each decade Decade
Depolarising peak at 20 ms in TE
Depolarising change at 100 ms in TE
Hyperpolarising change at 100 ms in TE
Current–threshold slope
20–29 30–39 40–49 50–59 60–69 70–79 Change with age
69.33 ± 1.48 71.02 ± 0.97 68.01 ± 1.37 68.47 ± 1.74 70.00 ± 0.81 67.43 ± 1.89 No significant change
46.17 ± 1.45 47.69 ± 2.02 45.69 ± 1.80 45.50 ± 1.25 47.92 ± 0.98 47.85 ± 2.33 No significant change
122.56 ± 38.76 127.15 ± 40.21 121.26 ± 38.3 116.75 ± 36.92 128.11 ± 40.51 127.84 ± 40.43 No significant change
0.241 ± 0.01 0.271 ± 0.01 0.318 ± 0.01 0.316 ± 0.01 0.316 ± 0.01 0.331 ± 0.02 Quadratic R2 = 0.39, P < 0.001
Data are % change in threshold (mean ± SEM, n = 10).
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
1401
Fig. 2. Scatter plots of the raw data for (a) the threshold for the 50% maximal CMAP, (b) rheobase, (c) supernormality and (d) the slope of the current– threshold relationship (hyperpolarising direction). The increase with age was best fitted with a quadratic function for all but supernormality, which showed a linear change.
Fig. 3. Recovery cycles. (a) All decades showed a normal recovery cycle with a period of refractoriness at 2–3 ms, supernormality maximal between 5.0 and 7.9 ms and subnormality, maximal at ISIs of 30–56 ms. (b) The degree of supernormality decreases with age (n = 10 for each decade).
1402
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
3.3. Threshold electrotonus and the current–threshold relationship Threshold electrotonus and the current–threshold relationship provide information about internodal properties. In the depolarising direction (plotted upwards, Fig. 4), the threshold electrotonus plot shows an initial fast phase proportional to the applied current. This is followed by further slow depolarisation, maximal 15–20 ms after the onset of the polarising current, as current spreads to and depolarises the internodal membrane. There is then a lessening of depolarisation, as an accommodative response develops due to slow potassium currents generated by channels on the node and internode. In the hyperpolarising direction (plotted downwards, Fig. 4), there is again an instantaneous change proportional to the applied current, followed by further hyperpolarisation as the closure of internodal potassium channels increases membrane resistance. An accommodative response, due to the inwardly rectifying conductance, IH, develops slowly and, after 100 ms, may be sufficient to produce a lessening in the degree of hyperpolarisation (not apparent in Fig. 4 because the current lasted only 100 ms). The threshold electrotonus plot showed a normal pattern for all decades, with no significant change between the decades (Fig. 4; Table 3). The current–threshold relationship is an analogue of the current–voltage relationship and reflects the rectifying properties of the axon. The steepening of the curve to the top right reflects outward rectification, an accommodative response to depolarisation due to the activation of potassium channels. The steepening of the curve to the bottom left represents inward rectification in response to hyperpolarisation. The slope of the current–threshold relationship was similar across the age spectrum in the depolarising direction (Fig. 5). However in the hyperpolarising direc-
Fig. 5. Current–threshold relationships. There was no difference with age in the accommodation to depolarising subthreshold currents (top right panel). In the hyperpolarising direction, there was a steepening of the curve for the current–threshold relationship with increasing age, suggesting greater accommodation, presumably due to an increase in inward rectification. This change was most prominent for currents greater than 40% of the threshold current (P < 0.001). Again the effect of age was greatest between the fourth and fifth decades. (Black circles = 20’s, white circles = 30’s, grey circles = 40’s, black triangles = 50’s, white triangles = 60’s, grey triangles = 70’s.)
tion, the slope of the curve was steeper for older subjects, especially for hyperpolarisation greater than 40% of the threshold current (P < 0.001; Fig. 5; Table 3). This change in slope was evident from the fifth decade, with little difference between the third and fourth decades. Curve fitting showed that the change in slope of the current–threshold relationship in the hyperpolarising direction could be described adequately by linear, exponential and quadratic relationships, but the latter had the highest correlation coefficient (R2 = 0.39; P < 0.001). 3.4. Sex differences
Fig. 4. Threshold electrotonus. There was no significant difference in the threshold electrotonus plots between the decades. (Black circles = 20’s, white circles = 30’s, black triangles = 40’s, white triangles = 50’s, black diamonds = 60’s, white diamonds = 70’s.)
There were 34 females and 26 males in the study. No significant sex-related differences in the measures of excitability were observed, unlike the findings for sensory axons (Kiernan et al., 2001). In sensory axons, the amplitude of the potential was higher and its threshold and rheobase were lower for females, changes presumably related to extra-neural (soft tissue) factors. In the present study, CMAP amplitudes were 8.9 ± 0.48 mV for females and 9.55 ± 0.62 mA for males (P = 0.403). The threshold for the 50% maximum CMAP increased from 7.54 ± 0.42 mA in the third decade to 15.72 ± 3.18 mA in the eighth decade for females and from 9.44 ± 0.94 mA to 15.84 ± 2.01 mA for males (P = 0.444). Rheobase was 3.47 ± 0.35 mA for females and 3.10 ± 0.26 mA for males (P = 0.418). Mean supernormality was 26.33% (range:
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
29.6% to 24.8%) for females and 21.86% (range: 26.3% to 11.8%) for males (P = 0.054). The slope of the current–threshold relationship was similar for females (0.26 ± 0.01 to 0.35 ± 0.04) and males (0.22 ± 0.01 to 0.32 ± 0.02; P = 0.490, Fig. 5). The change in slope with age was significant for each sex separately (females, P = 0.028; males, P = 0.001), with both sexes showing the greatest change in the eighth decade. 4. Discussion Peripheral nerve conduction varies with age, the changes being described as linear (Dorfman and Bosley, 1979) or exponential (Trojaborg, 1976; Taylor, 1984; Bouche et al., 1993). It has been suggested that the decrease in conduction velocity is due to segmental demyelination and a resultant reorganization of voltage-sensitive sodium channels in the axonal membrane (Adinolfi et al., 1991). Other studies suggest that the change in velocity reflects a decrease in axon diameter (Chase et al., 1992). Axonal energy metabolism is reduced with age and confers resistance to ischaemic conduction block (Low et al., 1986). Therefore alterations in nerve excitability indices with age would not be unexpected, particularly in IH, because the responsible channels belong to the adenyl cyclase-sensitive HCN (hyperpolarisationactivated cyclic nucleotide-gated) channel family (Kaupp and Seifert, 2001). This study has systematically studied the excitability of motor axons in subjects of increasing age with a balanced representation between the sexes. Most changes became evident in the fifth decade, an age when morphological changes begin to appear in peripheral nerves and when there are other neurological changes, e.g., reduction in the ability to accommodate visually for near objects. It is conceivable that minor changes begin to appear earlier but were not apparent in the present study because the subjects in the fourth decade were all between 30 and 35 years. It is of interest that the age-related changes in latency, threshold, rheobase, refractoriness and current– threshold slope were best described by a polynomial (quadratic) relationship, while the age-related change in supernormality was better described by a linear relationship. However, the deviations from linearity were small, and there were only minor differences in the R2 for each form of regression. Nevertheless, it is not intuitively unreasonable for age-related changes to increase to the sixth decade and then to regress, presumably with declining physical activity and other functions. There was a shift of stimulus–response curves to the right, such that stronger stimuli were required to activate axons with increasing age, a change that was particularly prominent for high-threshold axons. The change in slope of the normalised stimulus–response curves suggests that the change in threshold was not simply a reflection of greater difficulty in getting the stimulus to the nerve due to soft tissue changes. Other factors could have been altered, such as myelination and/or internodal length (Lascelles and Thomas, 1966; Rafalowska et al., 1976). How-
1403
ever, in association with the threshold change, there was an increase in rheobase with age, without a significant change in sSD. This latter finding suggests that some of the threshold increase is not due to a change in nodal excitability, but may have involved extra-neural factors that can increase tissue impedance. A decrease in supernormality could be due to depolarisation of axons, but there should have then been appropriate changes in other measures. Specifically, refractoriness, the relative refractory period and sSD should be increased, and the threshold electrotonus plots should have produced lesser maximal changes to depolarising and hyperpolarising currents (so-called ‘‘fanning-in’’). We conclude that there is no consistent evidence suggesting a change in membrane potential with aging. It is possible that age-related thinning of myelin and changes in internodal length (Lascelles and Thomas, 1966; Rafalowska et al., 1976) produce a smaller driving current, less charging of the internodal membrane and a smaller ‘‘back-discharge’’ to produce supernormality. There is some support for this suggestion: the decrease in refractoriness and the tendency for less late subnormality with age are consistent with a decrease in the driving current. However, the precise mechanism for the decrease in supernormality remains uncertain. The current–threshold plot suggested an increase in inward rectification (IH) with age, but the normal threshold electrotonus (with 40% conditioning currents) indicates that this becomes consistent across the age spectrum only with strong conditioning stimuli. Animal studies have demonstrated increases in IH with age in the nucleus accumbens (Belleau and Warren, 2000), in embryonic development of quail ganglionic neurons (Schlichter et al., 1991) and in rat hypoglossal motoneurons (Bayliss et al., 1994; Vianna et al., 1994). The activity of IH was 10-fold greater in adult than neonatal hypoglossal motoneurones with no apparent difference in the voltage-dependence or the reversal potential between neonates and adults. Therefore the greater IH current was thought to reflect increased current density (Bayliss et al., 1994). Using the computer model in QTRACS and the MEMFIT program (Bostock, 2006), the changes in the current– threshold relationship could be explained by variations in IH from the third to the eighth decades: an increase from 4.4 mS in the third decade to 7.2 mS in the sixth decade (+64%) and then a decrease to 5.8 mS in the eighth decade (+32% relative to the third decade), a pattern consistent with the quadratic relationship between this physiological measure of inward rectification and age (see Section 3). In the model, these changes in IH produced minor changes in threshold electrotonus, in keeping with the recorded agerelated data. IH was then allowed to co-vary with other factors, one by one. The greatest decrease in error across the entire data set was produced by changes in IH and small variations in [K+]o, from 4.4 mmol/l to 4.7 mmol/l (maximal error reduction 43%). The changes were similar to, though less prominent than, those described in uraemic patients by Krishnan et al. (2006).
1404
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
In the only other study of age-related changes in these measures, Kiernan et al. (2001) reported that, in cutaneous afferent axons, age-related changes were restricted to threshold measures and stimulus–response slope. In the present study on motor axons, there were changes in supernormality and the current–threshold relationship. The absence of these changes in sensory axons could be related to three factors. First, motor studies are intrinsically less variable because the size of the CMAP means that small but significant changes can be defined more easily. Second, the present study was a systematic study of age effects across the relevant decades. Third, the lesser expression of IH on motor than sensory axons (Bostock et al., 1994; Lin et al., 2002) could result in greater potential for increase in motor axons. However a common conclusion from both studies is that the majority of excitability indices undergo minimal or no change with age, and there is no real evidence for agerelated changes in channel function, apart from IH. Acknowledgements This study was supported by the NHMRC of Australia. S.K. Jankelowitz was the recipient of a Pfizer Neuroscience Award. The authors are grateful to Dr. Val Gebski for statistical advice and Mr. Tim Howells for the modelling studies. References Adinolfi AM, Yamuv J, Morales FR, Chase MH. Segmental demyelination in peripheral nerves of old cats. Neurobiol Aging 1991;12:175–9. Bayliss DA, Viana F, Bellingham MC, Berger AJ. Characteristics and postnatal development of a hyperpolarization-activated inward current in rat hypoglossal motoneurons in vitro. J Neurophysiol 1994;71:119–28. Belleau ML, Warren RA. Postnatal development of electrophysiological properties of nucleus accumbens neurons. J Physiol 2000;84:2204–16. Bolton CF, Winkelmann RK, Dyck PJ. A quantitative study of Meissner’s corpuscles in man. Neurology 1966;16:1–9. Bostock H, Burke D, Hales JP. Differences in behaviour of sensory and motor axons following release of ischaemia. Brain 1994;117:225–34. Bostock H, Cikurel K, Burke D. Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 1998;21:137–58. Bostock H. MEMFIT: a computer program to aid interpretation of multiple excitability measurements on human motor axons. Clin Neurophysiol 2006;117:S85. Bouche P, Cattelin F, Saint-Jean O, Leger JM, Queslati S, Guez D, et al. Clinical and electrophysiological study of the peripheral nervous system in the elderly. J Neurol 1993;240:263–8. Chase MH, Engelhardt JK, Adinolfi AM, Chirwa SS. Age-dependent changes in cat masseter nerve: and electrophysiological and morphological study. Brain Res 1992;586:279–88. Dorfman LJ, Bosley TM. Age-related changes in peripheral and central nerve conduction in man. Neurology 1979;29:38–44. Jacobs JM, Love S. Qualitative and quantitative morphology of human sural nerve at different ages. Brain 1985;108:897–924. Kaupp UB, Seifert R. Molecular diversity of pacemaker ion channels. Ann Rev Physiol 2001;63:235–57. Kawamura Y, O’Brien P, Okazaki H, Dyck PJ. Lumbar motoneurons of man II: the number and diameter distribution of large- and interme-
diate-diameter cytons in ‘‘motoneuron columns’’ of spinal cord of man. J Neuropathol Exp Neurol 1977;36:861–70. Kiernan MC, Burke D, Andersen KV, Bostock H. Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 2000;23:399–409. Kiernan MC, Lin CS-Y, Andersen KV, Murray NMF, Bostock H. Clinical evaluation of excitability measures in sensory nerve. Muscle Nerve 2001;24:883–92. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bostock H, Kiernan MC. Neuropathy, axonal Na+/K+ pump function and activity-dependent excitability changes in end-stage kidney disease. Clin Neurophysiol 2006;117:992–9. Lascelles RG, Thomas PK. Changes due to age in internodal length in the sural nerve in man. J Neurol Neurosurg Psychiatry 1966;29:40–4. Lin CS-Y, Mogyoros I, Burke D. Recovery of excitability of cutaneous afferents in the median and sural nerves following activity. Muscle Nerve 2000;23:763–70. Lin CS-Y, Kuwabara S, Cappelen-Smith C, Burke D. Responses of human sensory and motor axons to the release of ischaemia and to hyperpolarizing currents. J Physiol 2002;541:1025–39. Low PA, Schmelzer JD, Ward KK. The effect of age on energy metabolism and resistance to ischaemic conduction failure in rat peripheral nerve. J Physiol 1986;374:263–71. Nadler M, Harrison LM, Stephens JA. Changes in cutaneomuscular reflex responses in relation to normal ageing in man. Exp Brain Res 2002;146:48–53. O’Sullivan DJ, Swallow M. The fibre size and content of the radial and sural nerves. J Neurol Neurosurg Psychiatry 1968;31:464–70. Rafalowska J, Drac H, Rosinka K. Histological and electrophysiological changes of the lower motor neurone with aging. Pol Med Sci Hist Bull 1976;15:271–80. Roos MR, Rice CL, Vandervoort AA. Age-related changes in motor unit function. Muscle Nerve 1997;20:679–90. Schimrigk K, Ru¨ttinger H. The touch corpuscles of the plantar surface of the big toe, histological and histometrical investigations with respect to age. Eur Neurol 1980;19:49–60. Schlichter R, Bader CR, Bernheim L. Development of anomalous rectification (IH) and of a tetrodotoxin-resistant sodium current in embryonic quail neurones. J Physiol 1991;442:127–45. Stevens JC, Lofgren EP, Dyck PJ. Histometric evaluation of branches of peroneal nerve: technique for combined biopsy of muscle nerve and cutaneous nerve. Brain Res 1973;52:37–59. Swallow M. Fibre size and content of the anterior tibial nerve of the foot. J Neurol Neurosurg Psychiatry 1966;29:205–13. Takahashi K. A clinicopathological study on the peripheral nervous system of the aged. Sciatic nerve and autonomic nervous system. Geriatrics 1966;21:123–33. Tamura N, Kuwabara S, Misawa S, Kanai K, Nakata M, Sawai S, et al. Increased nodal persistent Na+ currents in human neuropathy and motor neuron disease estimated by latent addition. Clin Neurophysiol 2006;117:2451–8. Taylor PK. Non-linear effects of age on nerve conduction in adults. J Neurol Sci 1984;66:223–34. Taylor JL, Burke D, Heywood JL. Physiological evidence for a slow K+ conductance in human cutaneous afferents. J Physiol 1992;453:575–89. Tomlinson BE, Irving D. The numbers of limb motor neurons in the human lumbosacral cord throughout life. J Neurol Sci 1977;34:213–9. Trojaborg W. Motor and sensory conduction in the musculocutaneous nerve. J Neurol Neurosurg Psychiatry 1976;39:890–9. Vianna F, Bayliss DA, Berger AJ. Postnatal changes in rat hypoglossal motoneuron membrane properties. Neuroscience 1994;59:131–48. Vital A, Vital C, Rigal B, Decamps A, Emeriau JP, Galley P. Morphological study of the aging human peripheral nerve. Clin Neuropathol 1990;9:10–5.
Changes in measures of motor axon excitability with age S.K. Jankelowitz *, P.A. McNulty, David Burke Institute of Clinical Neurosciences, University of Sydney and Royal Prince Alfred Hospital, Sydney, Australia Accepted 21 February 2007 Available online 23 April 2007
Abstract Objective: Threshold tracking is a novel technique that permits examination of the excitability of human axons in vivo. Protocols have been validated for sensory and motor axons, but there are limited data on the changes in the excitability of motor axons with age. This study aimed to determine such changes from the third to the eighth decades. Methods: Sixty healthy subjects aged 22–79, 10 per decade, were studied using the TRONDXM4 protocol of the QTRAC thresholdtracking program to assess motor axon function. The median nerve was stimulated at the wrist and the compound muscle action potential was recorded from the thenar muscles. Results: There was an increase in threshold in elderly subjects, associated with a decrease in slope of the stimulus–response curves. Strength-duration time constant and threshold electrotonus to depolarising and hyperpolarising currents of up to 40% did not change significantly with aging. The current–threshold relationship was similar across all decades for subthreshold depolarising currents, but the slope of the current–threshold relationship was significantly steeper the older the subjects for hyperpolarising currents, particularly those greater than 40% of threshold. There was also a significant decrease in supernormality in the recovery cycle with increasing age. Conclusions: The threshold of axons increases with age and the extent of supernormality decreases. There may also be greater inward rectification in motor axons, perhaps due to greater activity of IH, the hyperpolarisation-activated conductance, though this is only significant with hyperpolarising currents greater than 40% of the threshold current. Significance: Many indices of axonal excitability, such as strength-duration time constant, the relative refractory period, late subnormality, threshold electrotonus and the depolarising side of the current–threshold relationship, do not change significantly with age. For other indices, age-related changes may be due to a combination of non-neural factors that alter current access to the node of Ranvier, changes in the axon and its myelination and, possibly, changes in channel activity and/or changes in extracellular [K+]o. 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Axonal excitability; Age; Threshold tracking; Supernormality; Inward rectification
1. Introduction Changes occur throughout the peripheral nervous system with age. For example, age-related structural changes include neuronal loss in the dorsal and ventral columns (Kawamura et al., 1977; Tomlinson and Irving, 1977), fibre loss in peripheral nerves, especially thick myelinated fibres, and changes in internodal length (Lascelles and Thomas, * Corresponding author. Address: Medical Foundation Building K-25, University of Sydney, Sydney, NSW 2006, Australia. Tel.: +61 2 9036 3091; fax: +61 2 9036 3092. E-mail address: [email protected] (S.K. Jankelowitz).
1966; Jacobs and Love, 1985; Vital et al., 1990). The importance of these changes increases with age, becoming noticeable in the fifth decade and significant in the seventh. A reduction in the number and density of human myelinated fibres with old age has been demonstrated in spinal roots (Kawamura et al., 1977) and in the radial (O’Sullivan and Swallow, 1968), ulnar (Rafalowska et al., 1976), sciatic (Takahashi, 1966), peroneal (Stevens et al., 1973), tibial (Swallow, 1966) and sural nerves (O’Sullivan and Swallow, 1968; Jacobs and Love, 1985). There are also changes in peripheral motor unit function with age: muscle mass/area is reduced and there may be an increase in slow (type I) muscle fibres compared to fast (type II) fibres. Motor force
1388-2457/$32.00 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2007.02.025
1398
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
and strength are reduced by 20–40% in subjects after the age of 70 although changes are observed from the sixth decade. A single motor axon innervates more muscle fibres, producing polyphasia on macro EMG (see Roos et al., 1997 for review). Touch-pressure, vibratory and cooling detection thresholds increase with age, and there is a reduction in the density and size of Meissner corpuscles in the fingers (Bolton et al., 1966) and the plantar surface of the big toe (Schimrigk and Ru¨ttinger, 1980). The threshold for perception of electrical stimuli increases with age (Nadler et al., 2002). Axonal excitability studies using threshold tracking are a recent addition to the neurophysiological armamentarium available to assess peripheral nerve function. They are now being employed to determine changes in excitability of axons in various pathologies to help determine the underlying pathophysiology. There has been, so far, no systematic analysis of age-related changes in the different measures of the excitability of motor axons, apart from a recent study showing no significant correlation between strengthduration time constant and age (Tamura et al., 2006). There has also been a study that documented the changes in excitability of sensory axons with age (Kiernan et al., 2001). Many pathologies affect only the elderly or are more severe the older the subject, and it is therefore relevant whether different excitability indices vary with age. This study was undertaken to determine if excitability indices of motor axons change with age, to determine the relationship with age and, hopefully, to shed light on the mechanisms. 2. Methods 2.1. Subjects Nerve excitability studies were performed on motor axons in 60 subjects of both genders, from 22 to 79 years, 10 per decade, 28 males and 32 females (Table 1), using the protocol and techniques described in full elsewhere (Bostock et al., 1998; Kiernan et al., 2000). The subjects provided informed consent to the procedures that had the approval of the Human Research Ethics Committee of the University of Sydney. For all experiments, skin temperature was monitored close to the stimulation site and was maintained above 32 C using blankets.
The median nerve was stimulated via non-polarisable electrodes (Maersk Medical, Stonehouse, UK) with the cathode over the median nerve at the wrist and the anode 10 cm proximal over muscle on the radial aspect of the arm. Compound muscle action potentials (CMAPs) were recorded from thenar muscles using surface electrodes over abductor pollicis brevis (APB) with the active electrode at the motor point and the reference electrode on the proximal phalanx. The electromyographic (EMG) signal was sampled at a rate of 10 kHz, amplified, filtered (3–500 Hz), and digitized by computer (Pentium PC) with an analogue-to-digital board (DT2812, Data Translation, Marlboro, Massachusetts). Stimulus waveforms were generated by the computer and converted to current by an isolated linear bipolar constant-current source (maximal output ± 50 mA). Motor axons were studied using the multiple excitability protocol, TRONDXM4 (QTRAC version 8.2, Professor H. Bostock, Institute of Neurology, London), as described elsewhere (Kiernan et al., 2000). Test current pulses of 0.2or 1-ms duration were delivered regularly at 0.8-s intervals to produce a target potential that was on the fast rising phase of the stimulus–response curve, 40% of the maximal CMAP. The computer was programmed to maintain the test potential at this target size despite changes in stimulus duration or the delivery of conditioning stimuli, either suprathreshold stimuli or subthreshold polarising currents. The amplitude of the CMAP was measured from baseline to the negative peak. Stimulus–response curves were recorded separately for test stimuli of 0.2- and 1-ms duration. These data were used to optimise threshold tracking (Kiernan et al., 2000) and to calculate the strength-duration time constant (sSD) according to Weiss’ Law, based on the linear relationship between stimulus charge (mA · ms) and stimulus duration. The sSD is a nodal property and reflects the rate of decrease of threshold current as the duration of the stimulus pulse increases. Prolonged subthreshold currents were used to alter the potential difference across the internodal membrane. The changes in threshold associated with electrotonic changes in membrane potential are termed threshold electrotonus and generally reflect the underlying changes in membrane potential. The subthreshold polarising currents were of 100-ms duration and set to be +40% and +20% (depolarising) and 40% and 20% (hyperpolarising) of the control
Table 1 Demographics and nerve conduction indices Decade
Mean age (range)
Sex ratio M:F
CMAP amplitude (mV)
Onset latency (ms)
Mean threshold 50% CMAP (mA)
20–29 30–39 40–49 50–59 60–69 70–79 Change with age
26.3 31.4 42.8 54.7 63.4 74.5
6:4 3:7 5:5 5:5 3:7 6:4
8.53 ± 0.92 9.88 ± 1.05 10.05 ± 0.92 8.46 ± 0.72 9.27 ± 1.08 9.53 ± 1.00 No significant change
2.78 ± 0.11 2.80 ± 0.08 2.64 ± 0.12 2.80 ± 0.12 3.03 ± 0.12 4.03 ± 0.20 Quadratic R2 = 0.21, P = 0.001
8.49 ± 0.71 8.90 ± 0.74 9.73 ± 1.61 10.39 ± 1.53 12.24 ± 1.91 15.83 ± 1.64 Quadratic R2 = 0.24, P < 0.001
Data are mean ± SEM.
(22–29) (30–35) (41–47) (50–59) (50–59) (70–79)
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
threshold current (i.e., the current required to produce the unconditioned target CMAP, 40% of maximum). Threshold was tested at different time points during and after the 100-ms polarising currents. Five stimulus combinations were tested in turn: test stimulus alone to measure the control threshold, test stimulus plus depolarising current of +20% or +40%, and test stimulus plus hyperpolarising current of 20% or 40%. Threshold was then measured with 1-ms test stimuli during a long-lasting (220 ms) subthreshold polarising current, 200 ms after its onset. The strength of the conditioning current was altered in 10% steps from +50% (depolarising) to 100% (hyperpolarising) of the control threshold. This produced a current–threshold relationship (analogous to the current–voltage relationship) that depends on the rectifying properties of the internodal axolemma. For both threshold electrotonus and the current–threshold relationship, each stimulus combination was repeated until three threshold measures were within 10% of the target response. The final part of the protocol measured the recovery of axonal excitability following a supramaximal conditioning stimulus. Recordings were made at 18 conditioning-test intervals from 2 to 200 ms. The stimulus combinations employed were unconditioned test stimulus of 1-ms duration tracking the control threshold, supramaximal 1-ms conditioning stimulus alone, and conditioning plus test stimuli. The response to the conditioning stimulus alone was subtracted from the response to both stimuli so that the conditioning maximal CMAP did not contaminate the measured response when the conditioning-test interval was short. Each stimulus combination was repeated until four valid threshold estimates were obtained. The results for stimulus threshold, strength-duration time constant, threshold electrotonus, the current–threshold relationship and the recovery cycle were compared across the decades. Statistical analyses were performed using SPSS version 13 (SPSS Corporation Chicago, USA). One-way ANOVA and paired t-tests were per-
1399
formed with Bonferroni corrections where appropriate, and corrected probability values are quoted in the text. Curve estimation and regression analysis were performed using SPSS to determine whether age-related changes were best fitted by a linear, exponential or polynomial function. Where more than one function was significant, the one with the higher R2 value has been reported. In addition, whether the data were linear or not was confirmed by performing regression analyses comparing one decade with each of the other decades in turn. Using the b coefficient, assessment was made whether the observed change was progressive (i.e., linear) or not. Significance was taken at P < 0.05. All data are given as means ± SEM. 3. Results No subject had clinical or neurophysiological evidence of carpal tunnel syndrome. The amplitude of the CMAP was fairly stable across the decades ranging from 8.53 ± 0.92 mV for the third decade to 9.53 ± 1.00 mV for the eighth decade (Table 1). Mean latency to the onset of the CMAP increased from 2.78 ± 0.11 ms in the third decade to 4.03 ± 0.02 ms in the eighth decade (quadratic function, R2 = 0.21, P = 0.001). The threshold for a 50% maximal CMAP increased from 8.49 ± 0.71 mA to 15.83 ± 1.64 mA from the third to the eighth decade (Table 1, Fig. 1a; see next section). For all excitability parameters, there was little difference between the results for the third and fourth decades (though this may have been due to the narrow age range of subjects in the fourth decade; see Table 1). Changes usually became evident in the fifth decade. Tables 2 and 3 summarise the age-related data for the measures of excitability discussed below. 3.1. Stimulus–response curves The stimulus–response curves showed an increase in threshold with age (Fig. 1a). Not only did the curve shift
Fig. 1. Stimulus–response curves. There is an increase in stimulus threshold and a decrease in slope of the stimulus–response curve with age. (a) The mean stimulus–response curves plotted against the absolute stimulus intensity. (b) The stimulus–response curves with stimulus intensity normalised so that the stimulus required for a 50% maximal CMAP is equal to 1. (Black circle = 20’s, white circle = 30’s, grey circle = 40’s, black triangle = 50’s, white triangle = 60’s, grey triangle = 70’s.)
1400
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
Table 2 Indices of axonal excitability for each decade Decade 20–29 30–39 40–49 50–59 60–69 70–79 Change with age
SDTCa (ms) 0.42 ± 0.02 0.52 ± 0.03 0.39 ± 0.02 0.43 ± 0.03 0.46 ± 0.03 0.40 ± 0.03 No significant change
Rheobase (mA)
Refractoriness at 2.5 ms (%)
2.50 ± 0.20 2.43 ± 0.21 3.18 ± 0.55 3.34 ± 0.56 3.74 ± 0.74 4.59 ± 0.54 Quadratic R2 = 0.17, P = 0.005
(n=8)
46.02 ± 5.57 38.06 ± 9.89(n=9) 22.70 ± 8.64(n=9) 20.05 ± 4.28(n=9) 16.34 ± 6.23 26.66 ± 9.99 Quadratic R2 = 0.20, P = 0.003
Relative refractory period (ms)
Supernormality (%)
3.37 ± 0.09 3.20 ± 0.12 2.91 ± 0.15 2.99 ± 0.08 2.88 ± 0.12 3.35 ± 0.37 No significant change
27.54 ± 1.23 28.68± 1.54 22.49 ± 2.55 23.53 ± 1.22 24.24 ± 1.22 18.51 ± 2.46 Linear R2 = 0.19, P < 0.001
Data are mean ± SEM. N = 10, except for refractoriness, where n is indicated in the superscript. a SDTC, strength-duration time constant.
to the right, but the slope of the normalised stimulus– response curve became less steep, especially close to the maximal CMAP amplitude, i.e., high-threshold axons required an even greater stimulus with age (Fig. 1b). The change in stimulus threshold became noticeable at, and gradually increased from, the fifth decade (Fig. 1; Table 1). The change in threshold (for the 50% CMAP) was significantly related to age in linear and quadratic functions, but the better correlation was with the quadratic relationship (R2 = 0.24, P < 0.001). Not surprisingly, the increase in current threshold was associated with an increase in rheobase with age (for the 50% CMAP, range: 2.50 ± 0.20 mA to 4.59 ± 0.54 mA, Fig. 2b; Table 2). The greatest change in rheobase was observed in the eighth decade. Again, the data could be fitted by a linear or quadratic relationship, but the better fit was a quadratic function (R2 = 0.17, P = 0.005). There was no consistent change in sSD with age, 0.42 ms in the third decade and 0.39 ms in the eighth decade (Fig. 2a). There was no clear relationship between sSD and rheobase, suggesting that the increase in rheobase (and presumably other threshold measures) was not solely due to a change in excitability. 3.2. Recovery cycle The recovery cycle examines the change in current required to produce the target CMAP following a supramaximal conditioning stimulus and, for each decade, there was a normal pattern. Axons were relatively refractory at short interstimulus intervals (2–3 ms), then supernormal,
maximally so at 6–7 ms, and finally subnormal, maximally at the sampled interstimulus intervals of 32–56 ms. The extent of refractoriness at the 2.5-ms interval was greatest in the third decade, and there was a significant difference in the degree of refractoriness (but not in the duration of the relative refractory period) with age (data for 2.5-ms interstimulus interval are given in Table 2; quadratic relationship, R2 = 0.20, P = 0.003). Supernormality is a reflection of the depolarising afterpotential, and late subnormality reflects the time course of voltage-gated slow potassium channels activated by the conditioning discharge (Taylor et al., 1992; Bostock et al., 1998; Lin et al., 2000). Supernormality was measured as the mean percentage decrease in threshold seen at conditioning-test intervals of 5, 6.3 and 7.9 ms. There was a significant reduction in supernormality with age (Fig. 3b; Table 2). The reduction in supernormality was greatest between the seventh and the eighth decades, but linear regression provided the best correlation (R2 = 0.19; P < 0.001). There was a significant correlation between peak supernormality and the depolarising peak reached in the threshold electrotonus plot at 20 ms (R2 = 0.32; P < 0.001). This latter finding is not unexpected given that the two measures share common mechanisms, but surprisingly the age-related change in the depolarising peak at 20 ms was not significant (P = 0.400, see below, Table 2). There was a trend for late subnormality, measured as the mean threshold increase at conditioning-test intervals of 32, 42 and 56 ms, to decrease with age, but this change was not significant (P = 0.060).
Table 3 Threshold electrotonus and the current–threshold plot for each decade Decade
Depolarising peak at 20 ms in TE
Depolarising change at 100 ms in TE
Hyperpolarising change at 100 ms in TE
Current–threshold slope
20–29 30–39 40–49 50–59 60–69 70–79 Change with age
69.33 ± 1.48 71.02 ± 0.97 68.01 ± 1.37 68.47 ± 1.74 70.00 ± 0.81 67.43 ± 1.89 No significant change
46.17 ± 1.45 47.69 ± 2.02 45.69 ± 1.80 45.50 ± 1.25 47.92 ± 0.98 47.85 ± 2.33 No significant change
122.56 ± 38.76 127.15 ± 40.21 121.26 ± 38.3 116.75 ± 36.92 128.11 ± 40.51 127.84 ± 40.43 No significant change
0.241 ± 0.01 0.271 ± 0.01 0.318 ± 0.01 0.316 ± 0.01 0.316 ± 0.01 0.331 ± 0.02 Quadratic R2 = 0.39, P < 0.001
Data are % change in threshold (mean ± SEM, n = 10).
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
1401
Fig. 2. Scatter plots of the raw data for (a) the threshold for the 50% maximal CMAP, (b) rheobase, (c) supernormality and (d) the slope of the current– threshold relationship (hyperpolarising direction). The increase with age was best fitted with a quadratic function for all but supernormality, which showed a linear change.
Fig. 3. Recovery cycles. (a) All decades showed a normal recovery cycle with a period of refractoriness at 2–3 ms, supernormality maximal between 5.0 and 7.9 ms and subnormality, maximal at ISIs of 30–56 ms. (b) The degree of supernormality decreases with age (n = 10 for each decade).
1402
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
3.3. Threshold electrotonus and the current–threshold relationship Threshold electrotonus and the current–threshold relationship provide information about internodal properties. In the depolarising direction (plotted upwards, Fig. 4), the threshold electrotonus plot shows an initial fast phase proportional to the applied current. This is followed by further slow depolarisation, maximal 15–20 ms after the onset of the polarising current, as current spreads to and depolarises the internodal membrane. There is then a lessening of depolarisation, as an accommodative response develops due to slow potassium currents generated by channels on the node and internode. In the hyperpolarising direction (plotted downwards, Fig. 4), there is again an instantaneous change proportional to the applied current, followed by further hyperpolarisation as the closure of internodal potassium channels increases membrane resistance. An accommodative response, due to the inwardly rectifying conductance, IH, develops slowly and, after 100 ms, may be sufficient to produce a lessening in the degree of hyperpolarisation (not apparent in Fig. 4 because the current lasted only 100 ms). The threshold electrotonus plot showed a normal pattern for all decades, with no significant change between the decades (Fig. 4; Table 3). The current–threshold relationship is an analogue of the current–voltage relationship and reflects the rectifying properties of the axon. The steepening of the curve to the top right reflects outward rectification, an accommodative response to depolarisation due to the activation of potassium channels. The steepening of the curve to the bottom left represents inward rectification in response to hyperpolarisation. The slope of the current–threshold relationship was similar across the age spectrum in the depolarising direction (Fig. 5). However in the hyperpolarising direc-
Fig. 5. Current–threshold relationships. There was no difference with age in the accommodation to depolarising subthreshold currents (top right panel). In the hyperpolarising direction, there was a steepening of the curve for the current–threshold relationship with increasing age, suggesting greater accommodation, presumably due to an increase in inward rectification. This change was most prominent for currents greater than 40% of the threshold current (P < 0.001). Again the effect of age was greatest between the fourth and fifth decades. (Black circles = 20’s, white circles = 30’s, grey circles = 40’s, black triangles = 50’s, white triangles = 60’s, grey triangles = 70’s.)
tion, the slope of the curve was steeper for older subjects, especially for hyperpolarisation greater than 40% of the threshold current (P < 0.001; Fig. 5; Table 3). This change in slope was evident from the fifth decade, with little difference between the third and fourth decades. Curve fitting showed that the change in slope of the current–threshold relationship in the hyperpolarising direction could be described adequately by linear, exponential and quadratic relationships, but the latter had the highest correlation coefficient (R2 = 0.39; P < 0.001). 3.4. Sex differences
Fig. 4. Threshold electrotonus. There was no significant difference in the threshold electrotonus plots between the decades. (Black circles = 20’s, white circles = 30’s, black triangles = 40’s, white triangles = 50’s, black diamonds = 60’s, white diamonds = 70’s.)
There were 34 females and 26 males in the study. No significant sex-related differences in the measures of excitability were observed, unlike the findings for sensory axons (Kiernan et al., 2001). In sensory axons, the amplitude of the potential was higher and its threshold and rheobase were lower for females, changes presumably related to extra-neural (soft tissue) factors. In the present study, CMAP amplitudes were 8.9 ± 0.48 mV for females and 9.55 ± 0.62 mA for males (P = 0.403). The threshold for the 50% maximum CMAP increased from 7.54 ± 0.42 mA in the third decade to 15.72 ± 3.18 mA in the eighth decade for females and from 9.44 ± 0.94 mA to 15.84 ± 2.01 mA for males (P = 0.444). Rheobase was 3.47 ± 0.35 mA for females and 3.10 ± 0.26 mA for males (P = 0.418). Mean supernormality was 26.33% (range:
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
29.6% to 24.8%) for females and 21.86% (range: 26.3% to 11.8%) for males (P = 0.054). The slope of the current–threshold relationship was similar for females (0.26 ± 0.01 to 0.35 ± 0.04) and males (0.22 ± 0.01 to 0.32 ± 0.02; P = 0.490, Fig. 5). The change in slope with age was significant for each sex separately (females, P = 0.028; males, P = 0.001), with both sexes showing the greatest change in the eighth decade. 4. Discussion Peripheral nerve conduction varies with age, the changes being described as linear (Dorfman and Bosley, 1979) or exponential (Trojaborg, 1976; Taylor, 1984; Bouche et al., 1993). It has been suggested that the decrease in conduction velocity is due to segmental demyelination and a resultant reorganization of voltage-sensitive sodium channels in the axonal membrane (Adinolfi et al., 1991). Other studies suggest that the change in velocity reflects a decrease in axon diameter (Chase et al., 1992). Axonal energy metabolism is reduced with age and confers resistance to ischaemic conduction block (Low et al., 1986). Therefore alterations in nerve excitability indices with age would not be unexpected, particularly in IH, because the responsible channels belong to the adenyl cyclase-sensitive HCN (hyperpolarisationactivated cyclic nucleotide-gated) channel family (Kaupp and Seifert, 2001). This study has systematically studied the excitability of motor axons in subjects of increasing age with a balanced representation between the sexes. Most changes became evident in the fifth decade, an age when morphological changes begin to appear in peripheral nerves and when there are other neurological changes, e.g., reduction in the ability to accommodate visually for near objects. It is conceivable that minor changes begin to appear earlier but were not apparent in the present study because the subjects in the fourth decade were all between 30 and 35 years. It is of interest that the age-related changes in latency, threshold, rheobase, refractoriness and current– threshold slope were best described by a polynomial (quadratic) relationship, while the age-related change in supernormality was better described by a linear relationship. However, the deviations from linearity were small, and there were only minor differences in the R2 for each form of regression. Nevertheless, it is not intuitively unreasonable for age-related changes to increase to the sixth decade and then to regress, presumably with declining physical activity and other functions. There was a shift of stimulus–response curves to the right, such that stronger stimuli were required to activate axons with increasing age, a change that was particularly prominent for high-threshold axons. The change in slope of the normalised stimulus–response curves suggests that the change in threshold was not simply a reflection of greater difficulty in getting the stimulus to the nerve due to soft tissue changes. Other factors could have been altered, such as myelination and/or internodal length (Lascelles and Thomas, 1966; Rafalowska et al., 1976). How-
1403
ever, in association with the threshold change, there was an increase in rheobase with age, without a significant change in sSD. This latter finding suggests that some of the threshold increase is not due to a change in nodal excitability, but may have involved extra-neural factors that can increase tissue impedance. A decrease in supernormality could be due to depolarisation of axons, but there should have then been appropriate changes in other measures. Specifically, refractoriness, the relative refractory period and sSD should be increased, and the threshold electrotonus plots should have produced lesser maximal changes to depolarising and hyperpolarising currents (so-called ‘‘fanning-in’’). We conclude that there is no consistent evidence suggesting a change in membrane potential with aging. It is possible that age-related thinning of myelin and changes in internodal length (Lascelles and Thomas, 1966; Rafalowska et al., 1976) produce a smaller driving current, less charging of the internodal membrane and a smaller ‘‘back-discharge’’ to produce supernormality. There is some support for this suggestion: the decrease in refractoriness and the tendency for less late subnormality with age are consistent with a decrease in the driving current. However, the precise mechanism for the decrease in supernormality remains uncertain. The current–threshold plot suggested an increase in inward rectification (IH) with age, but the normal threshold electrotonus (with 40% conditioning currents) indicates that this becomes consistent across the age spectrum only with strong conditioning stimuli. Animal studies have demonstrated increases in IH with age in the nucleus accumbens (Belleau and Warren, 2000), in embryonic development of quail ganglionic neurons (Schlichter et al., 1991) and in rat hypoglossal motoneurons (Bayliss et al., 1994; Vianna et al., 1994). The activity of IH was 10-fold greater in adult than neonatal hypoglossal motoneurones with no apparent difference in the voltage-dependence or the reversal potential between neonates and adults. Therefore the greater IH current was thought to reflect increased current density (Bayliss et al., 1994). Using the computer model in QTRACS and the MEMFIT program (Bostock, 2006), the changes in the current– threshold relationship could be explained by variations in IH from the third to the eighth decades: an increase from 4.4 mS in the third decade to 7.2 mS in the sixth decade (+64%) and then a decrease to 5.8 mS in the eighth decade (+32% relative to the third decade), a pattern consistent with the quadratic relationship between this physiological measure of inward rectification and age (see Section 3). In the model, these changes in IH produced minor changes in threshold electrotonus, in keeping with the recorded agerelated data. IH was then allowed to co-vary with other factors, one by one. The greatest decrease in error across the entire data set was produced by changes in IH and small variations in [K+]o, from 4.4 mmol/l to 4.7 mmol/l (maximal error reduction 43%). The changes were similar to, though less prominent than, those described in uraemic patients by Krishnan et al. (2006).
1404
S.K. Jankelowitz et al. / Clinical Neurophysiology 118 (2007) 1397–1404
In the only other study of age-related changes in these measures, Kiernan et al. (2001) reported that, in cutaneous afferent axons, age-related changes were restricted to threshold measures and stimulus–response slope. In the present study on motor axons, there were changes in supernormality and the current–threshold relationship. The absence of these changes in sensory axons could be related to three factors. First, motor studies are intrinsically less variable because the size of the CMAP means that small but significant changes can be defined more easily. Second, the present study was a systematic study of age effects across the relevant decades. Third, the lesser expression of IH on motor than sensory axons (Bostock et al., 1994; Lin et al., 2002) could result in greater potential for increase in motor axons. However a common conclusion from both studies is that the majority of excitability indices undergo minimal or no change with age, and there is no real evidence for agerelated changes in channel function, apart from IH. Acknowledgements This study was supported by the NHMRC of Australia. S.K. Jankelowitz was the recipient of a Pfizer Neuroscience Award. The authors are grateful to Dr. Val Gebski for statistical advice and Mr. Tim Howells for the modelling studies. References Adinolfi AM, Yamuv J, Morales FR, Chase MH. Segmental demyelination in peripheral nerves of old cats. Neurobiol Aging 1991;12:175–9. Bayliss DA, Viana F, Bellingham MC, Berger AJ. Characteristics and postnatal development of a hyperpolarization-activated inward current in rat hypoglossal motoneurons in vitro. J Neurophysiol 1994;71:119–28. Belleau ML, Warren RA. Postnatal development of electrophysiological properties of nucleus accumbens neurons. J Physiol 2000;84:2204–16. Bolton CF, Winkelmann RK, Dyck PJ. A quantitative study of Meissner’s corpuscles in man. Neurology 1966;16:1–9. Bostock H, Burke D, Hales JP. Differences in behaviour of sensory and motor axons following release of ischaemia. Brain 1994;117:225–34. Bostock H, Cikurel K, Burke D. Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 1998;21:137–58. Bostock H. MEMFIT: a computer program to aid interpretation of multiple excitability measurements on human motor axons. Clin Neurophysiol 2006;117:S85. Bouche P, Cattelin F, Saint-Jean O, Leger JM, Queslati S, Guez D, et al. Clinical and electrophysiological study of the peripheral nervous system in the elderly. J Neurol 1993;240:263–8. Chase MH, Engelhardt JK, Adinolfi AM, Chirwa SS. Age-dependent changes in cat masseter nerve: and electrophysiological and morphological study. Brain Res 1992;586:279–88. Dorfman LJ, Bosley TM. Age-related changes in peripheral and central nerve conduction in man. Neurology 1979;29:38–44. Jacobs JM, Love S. Qualitative and quantitative morphology of human sural nerve at different ages. Brain 1985;108:897–924. Kaupp UB, Seifert R. Molecular diversity of pacemaker ion channels. Ann Rev Physiol 2001;63:235–57. Kawamura Y, O’Brien P, Okazaki H, Dyck PJ. Lumbar motoneurons of man II: the number and diameter distribution of large- and interme-
diate-diameter cytons in ‘‘motoneuron columns’’ of spinal cord of man. J Neuropathol Exp Neurol 1977;36:861–70. Kiernan MC, Burke D, Andersen KV, Bostock H. Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 2000;23:399–409. Kiernan MC, Lin CS-Y, Andersen KV, Murray NMF, Bostock H. Clinical evaluation of excitability measures in sensory nerve. Muscle Nerve 2001;24:883–92. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bostock H, Kiernan MC. Neuropathy, axonal Na+/K+ pump function and activity-dependent excitability changes in end-stage kidney disease. Clin Neurophysiol 2006;117:992–9. Lascelles RG, Thomas PK. Changes due to age in internodal length in the sural nerve in man. J Neurol Neurosurg Psychiatry 1966;29:40–4. Lin CS-Y, Mogyoros I, Burke D. Recovery of excitability of cutaneous afferents in the median and sural nerves following activity. Muscle Nerve 2000;23:763–70. Lin CS-Y, Kuwabara S, Cappelen-Smith C, Burke D. Responses of human sensory and motor axons to the release of ischaemia and to hyperpolarizing currents. J Physiol 2002;541:1025–39. Low PA, Schmelzer JD, Ward KK. The effect of age on energy metabolism and resistance to ischaemic conduction failure in rat peripheral nerve. J Physiol 1986;374:263–71. Nadler M, Harrison LM, Stephens JA. Changes in cutaneomuscular reflex responses in relation to normal ageing in man. Exp Brain Res 2002;146:48–53. O’Sullivan DJ, Swallow M. The fibre size and content of the radial and sural nerves. J Neurol Neurosurg Psychiatry 1968;31:464–70. Rafalowska J, Drac H, Rosinka K. Histological and electrophysiological changes of the lower motor neurone with aging. Pol Med Sci Hist Bull 1976;15:271–80. Roos MR, Rice CL, Vandervoort AA. Age-related changes in motor unit function. Muscle Nerve 1997;20:679–90. Schimrigk K, Ru¨ttinger H. The touch corpuscles of the plantar surface of the big toe, histological and histometrical investigations with respect to age. Eur Neurol 1980;19:49–60. Schlichter R, Bader CR, Bernheim L. Development of anomalous rectification (IH) and of a tetrodotoxin-resistant sodium current in embryonic quail neurones. J Physiol 1991;442:127–45. Stevens JC, Lofgren EP, Dyck PJ. Histometric evaluation of branches of peroneal nerve: technique for combined biopsy of muscle nerve and cutaneous nerve. Brain Res 1973;52:37–59. Swallow M. Fibre size and content of the anterior tibial nerve of the foot. J Neurol Neurosurg Psychiatry 1966;29:205–13. Takahashi K. A clinicopathological study on the peripheral nervous system of the aged. Sciatic nerve and autonomic nervous system. Geriatrics 1966;21:123–33. Tamura N, Kuwabara S, Misawa S, Kanai K, Nakata M, Sawai S, et al. Increased nodal persistent Na+ currents in human neuropathy and motor neuron disease estimated by latent addition. Clin Neurophysiol 2006;117:2451–8. Taylor PK. Non-linear effects of age on nerve conduction in adults. J Neurol Sci 1984;66:223–34. Taylor JL, Burke D, Heywood JL. Physiological evidence for a slow K+ conductance in human cutaneous afferents. J Physiol 1992;453:575–89. Tomlinson BE, Irving D. The numbers of limb motor neurons in the human lumbosacral cord throughout life. J Neurol Sci 1977;34:213–9. Trojaborg W. Motor and sensory conduction in the musculocutaneous nerve. J Neurol Neurosurg Psychiatry 1976;39:890–9. Vianna F, Bayliss DA, Berger AJ. Postnatal changes in rat hypoglossal motoneuron membrane properties. Neuroscience 1994;59:131–48. Vital A, Vital C, Rigal B, Decamps A, Emeriau JP, Galley P. Morphological study of the aging human peripheral nerve. Clin Neuropathol 1990;9:10–5.