Symmetric sensorimotor polyneuropathy is a common complication of diabetes. Sensory and motor evoked amplitudes and conduction velocities are reduced. Both demyelination and axon loss have been reported in pathologic studies. Conduction block (CB),a manifestation of segmental demyelination, has not been previously studied in diabetic neuropathy. We determined the prevalence of conduction block in patients with diabetes by analyzing electrodiagnostic data from 24 diabetics. Conduction block was defined as a >20% drop in peak-to-peak amplitude, and a 20% compared with the distal site, and the -p duration increased at the proximal site by > 15% compared with the distal site. If the change in -p duration was >15%, and the p-p amplitude decreased by >20%, this was designated abnormal temporal dispersion (TD).6 Because of the difficulty in the determination of CB in nerves with reduced CMAP amplia1 nerves in which the dist u d e ~ , ' w~e~excluded '~
Conduction Block in Diabetic Neuropathy
Table 1. Nerve segments with CB in diabetes. ~~~
~~
Nerve segment
~~
~~
~
Number with CBltotal segments
Median Forearm Upper arm Ulnar Forearm Peroneal Leg Total
1I30 016
3/20 2/20 6/76
MUSCLE & NERVE
September 1991
859
proximal sites was 27% (range 16% to 47%). The CV of the segments with T D were: in the median forearm segment, mean 55 m/s (range 47 to 63 m/s); and in the peroneal nerve, 38 m/s (range 23 to 46 m/s) (Table 2). Considering all nerve segments together, the mean CV in our 24 patients was as follows: median nerve in the forearm, 49 m/s (range 38 to 63 m/s); median nerve in the arm, 54 m/s (range 40 to 63 m/s); ulnar nerve in the forearm, 54 m/s (range 32 to 74 m/s); peroneal nerve below the fibular head, 39 m/s (range 23 to 48 m/s). The ranges of evoked CMAP amplitudes (-p) in the nerve segments under study were: 1400 to 15,000 pV in the median nerve, 600 to 10,000 p-V in the ulnar nerve, and 140 to 7000 IJ.Vin the peroneal nerve. i n order to compare our data for -p duration and p- p amplitude change between proximal and distal sites to those of Brown and Feasby,' we calculated these values (a) for all nerve segments (n = 76); (b) for the segments with CB (n = 6); and (c) for the segments without CB, including both normal and those with T D (n = 70). The mean decrease in p- p amplitude between proximal and distal stimulation sites in all 76 segments was 12% (range 0% to 40%), and the mean increase in -p duration between the 2 sites was 10% (range 0% to 47%). The mean decrease in p-p amplitude in the CB segments was 28% (range 21% to 40%), and the mean decrease in -p duration was 8% (range 0% to 14%).The mean decrease in p-p amplitude in segments without CB was 1 1 % (range 0% to 35%), and the mean increase in -p duration was 10% (range 0% to 47%). DISCUSSION
Conduction block, a physiologic finding, has been correlated with segmental demyelination in pathologic ~tudies.".'~CB is encountered clinically in several conditions,5 including acute compressive
Table 2. Nerve segments with temporal dispersion in diabetes. Nerve segment Median Forearm Upper arm Ulnar Forearm Peroneal Leg Total
860
Number with TDItotal segments 2/30 016 0120 7/20 9/76
Conduction Block in Diabetic Neuropathy
neuropathy; nerve entrapments,? which occur frequently in diabetics; acquired demyelinating polyneuropathies, such as Guillain- BarrC syndrome6 or chronic inflammatory demyelinating polyneuropathy (CIDP)1917;and selected motor neuropathies." Hence, the finding of CB on an electrophysiologic study has important implications. In addition to limiting differential diagnosis, the finding of CB along nerve trunks suggests that segmental demyelination is an important part of the underlying pathology which should provide clues to pathogenesis. Several hypotheses for the pathogenesis of symmetric diabetic polyneuropathy have received support from experimental studies (for review see ref. 2). In human diabetic neuropathy, pathologic studies have shown both segmental demyelination and axon and debate continues as to the relative contribution of each.' If segmental demyelination were common in diabetes, then CB should be present in patients with diabetic polyneuropathy. We addressed this question in a series of diabetic patients referred for electrodiagnostic studies. T o determine if CB was found along a nerve segment, we applied the criteria for CB developed by Brown and Feasby.' They studied the median, ulnar, and peroneal nerves of 30 control subjects and found that the decline in p-p amplitude, between proximal and distal stimulation sites, never exceeded 20%. Likewise, the increase in duration between the 2 sites never exceeded 15%. Comparing these control values with those from a series of patients with GBS, they determined that CB was frequent in GBS, occurring in 68% of patients. In other studies of CB, the decline in p-p amplitudes used to determine abnormalities was set at 50%'? and 40%*.';however, it is unclear how these values were determined. More recently, Rhee et al.," using a computer simulation technique, have shown that temporal dispersion and interphase cancellation alone can produce decrements of CMAP amplitude of >50%, and of CMAP area of up to 50%. However, it is unknown whether these findings apply to humans. Criteria for determination of CB in chronic peripheral nerve disease, such as diabetic polyneuropathy, have not been developed. Ideally, the control values for determination of CB should be determined by studying nerves of patients with pure axonal polyneuropathy, as these would be the most precise controls. Another control population might be patients with motor neuron disease. In the absence of such control values for CB in
MUSCLE & NERVE
September 1991
chronic neuropathies, we used the Brown and Feasby criteria. We found CB in only 6 of 76 nerve segments. Among these nerve segments are potential, although infrequent, sites of nerve entrapment12 that could lead to CB. These include the pronator teres syndrome (median forearm segment) and flexor carpi ulnaris entrapment of the ulnar nerve (ulnar forearm segment). It is possible that with either “inching” techniques or more extensive conduction studies and EMG, these could have been localized more precisely. Assuming that CB in these segment was not related to nerve entrapment, our findings still suggest that CB in diabetic neuropathy is rare. The relative rarity of CB in our diabetic patients suggests that segmental demyelination is not a prominent part of the underlying pathology. Additionally, it is likely that nerve compression and immunologically mediated mechanisms, which commonly lead to peripheral nerve demyelination, do not play a role in pathogenesis of diabetic polyneuropathy. Thus, while our studies do not provide support for any of the current theories of pathogenesis of diabetic polyneuropathy, they provide indirect evidence against these two mechanisms. The magnitude of CB in the diabetic patients was modest compared with that of patients with GBS studied by Brown and Feasby.‘ Most of the latter patients had reductions in CMAP amplitude exceeding SO%, while the mean CMAP decrease in our patients was 28%, with the highest being 40%. This suggests that in DM, unlike in GBS, segmental demyelination is not widely distributed along the length of nerves. The GBS patients also had frequent CB at entrapment sites (30% of ul-
nar nerve across-elbow segments), while among our 8 patients, who had a total of 9 across-elbow ulnar nerve entrapments, only 1 showed CB. There are 2 potential sources of error in our study. First, apart from CB, a drop in p-p amplitude between distal and proximal stimulation sites without a change in -p duration could be due to failure to deliver maximal stimulation at the roximal site, interphase shift and cancellation,“ and axonal interruption between the stimulating electrodes.” We tried to exclude each of these as possible causes for apparent CB by attending to technical factors during the studies, and by excluding nerve segments with severely reduced CMAP amplitudes from our study. If we counted a nerve segment as having CB when, in reality, there was another explanation, then we would be overestimating the frequency of CB rather than underestimating it; thus, CB may be even less frequent than 7%. Second, if the degree of polyneuropathy in our patient population was mild, CB and segmental demyelination might not have been present. We think this is an unlikely explanation for our findings. We studied patients with a range of degrees of diabetic neuropathy, and the patients with CB were not found exclusively in those with clinically or electrophysiologically severe neuropathy. Nevertheless, another study specifically evaluating patients with severe diabetic polyneuropathy would be valuable. In conclusion, CB in diabetic neuropathy is uncommon. When present, the finding of CB in diabetics, especially in more than one nerve segment, should suggest alternative or additional causes for neuropathy, such as chronic inflammatory demyelinating polyne~ropathy.~
REFERENCES 1 . Albers JW: Inflammatory demyelinating polyradiculoneuropathy, in Brown WF, Bolton CF (eds): Clinical Electromyografhy. Boston, Butterworths, 1987, pp 209-244. 2. Asbury AK: Understanding diabetic neuropathy. N Engl J Med 1988;319:577-578. 3. Behse F, Buchthal F, Carlsen F: Nerve biopsy and conduction studies in diabetic neuropathy. J Neurol Neurosurg Psychiatty 1977;40: 1072- 1082. 4. Brown MJ, Asbury AK: Diabetic neuropathy. Ann Neurol 1984;15:2- 12. 5. Brown WF: The Physiological and Technical Basis of Electromyography. Boston, Butterworths, 1984. 6. Brown WF, Feasby TE: Conduction block and denervation in Guillain-Barre polyneuropathy. Brain 1984; 107:219239. 7. Brown WR, Yates SK: The quantitative assessment of conduction block in human entrapment neuropathies. Eleclroenceph Clin Neurophysiol 1983 ;56:(S52).
Conduction Block in Diabetic Neuropathy
8. Cornblath DR, Asbury AK, Albers JW, Feasby TE, Hahn AF, McLeod JG, Mendell JR, Parry GJ, Pollard JD, Thomas PK: Criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology (in press). 9. Cornblath DR, Drachman DB, Griffin JW: Demyelinating motor neuropathy in patients with diabetic polyneuropathy. Ann Neurol 1987;22:126. 10. Cornblath DR, Sumner AJ: Conduction block in neuropathies with necrotizing vasculitis. Muscle Nerve 1991,14: 185. 11. Daube JR: Electrophysiological testing in diabetic neuropathy, in Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D (eds): Diabetic Neuropathy. Philadelphia, Saunders, 1987, pp 162- 176. 12. Dawson DM, Hallett M, Millender LH: Entrapment Neuropathies, Boston, Little, Brown, 1983. 13. Dyck PJ: Pathology, in Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D (eds): Diabetic Neuropathy. Philadelphia, Saunders, 1987, pp 223-236.
MUSCLE & NERVE
September 1991
861
14. Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D: Diabetic Neuropathy, Philadelphia, Saunders, 1987. 15. Feasby TE, Brown WF, Gilbert JJ, Hayn AF: The pathological basis of conduction block in human neuropathies. J Neurol Neurosurg Psychiatry 1985;43:239-244. 16. Halar EM, Graf RJ, Halter JB, Brozovich FV, Soine TL: Diabetic neuropathy: A clinical, laboratory, and electrodiagnostic study. Arch Phys Med Rehab 1982;63:298-303. 17. Lewis RA, Sumner AJ: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 1982;32:592-596. 18. Pestronk A, Cornblath DR, Ilyas AA, Baba H, Quarles RH,
862
Conduction Block in Diabetic Neuropathy
Griffin JW, Alderson K, Adams RN: A treatable multifocal motor neuropathy with antibodies to GM 1 ganglioside. Ann Neurol 1988;24:73-78. 19. Rhee EK, England JD, Sumner AJ: A computer simulation of conduction block: Effects produced by actual block versus interphase cancellation. Ann Neurol 1990;28:146- 156. 20. Thomas PK, Lascelles RG: The pathology of diabetic neuropathy. Q J Med 1966;35:489-509, 21. Trojaborg W, Lange DJ, Latov N, Younger DS, Lovelace RE, Rowland LP: Conduction block and other abnormalities of nerve conduction in motor neuron disease. Neurology 1990;4O(suppl 1): 182.
MUSCLE & NERVE
September 1991
Conduction Block in Diabetic Neuropathy
Table 1. Nerve segments with CB in diabetes. ~~~
~~
Nerve segment
~~
~~
~
Number with CBltotal segments
Median Forearm Upper arm Ulnar Forearm Peroneal Leg Total
1I30 016
3/20 2/20 6/76
MUSCLE & NERVE
September 1991
859
proximal sites was 27% (range 16% to 47%). The CV of the segments with T D were: in the median forearm segment, mean 55 m/s (range 47 to 63 m/s); and in the peroneal nerve, 38 m/s (range 23 to 46 m/s) (Table 2). Considering all nerve segments together, the mean CV in our 24 patients was as follows: median nerve in the forearm, 49 m/s (range 38 to 63 m/s); median nerve in the arm, 54 m/s (range 40 to 63 m/s); ulnar nerve in the forearm, 54 m/s (range 32 to 74 m/s); peroneal nerve below the fibular head, 39 m/s (range 23 to 48 m/s). The ranges of evoked CMAP amplitudes (-p) in the nerve segments under study were: 1400 to 15,000 pV in the median nerve, 600 to 10,000 p-V in the ulnar nerve, and 140 to 7000 IJ.Vin the peroneal nerve. i n order to compare our data for -p duration and p- p amplitude change between proximal and distal sites to those of Brown and Feasby,' we calculated these values (a) for all nerve segments (n = 76); (b) for the segments with CB (n = 6); and (c) for the segments without CB, including both normal and those with T D (n = 70). The mean decrease in p- p amplitude between proximal and distal stimulation sites in all 76 segments was 12% (range 0% to 40%), and the mean increase in -p duration between the 2 sites was 10% (range 0% to 47%). The mean decrease in p-p amplitude in the CB segments was 28% (range 21% to 40%), and the mean decrease in -p duration was 8% (range 0% to 14%).The mean decrease in p-p amplitude in segments without CB was 1 1 % (range 0% to 35%), and the mean increase in -p duration was 10% (range 0% to 47%). DISCUSSION
Conduction block, a physiologic finding, has been correlated with segmental demyelination in pathologic ~tudies.".'~CB is encountered clinically in several conditions,5 including acute compressive
Table 2. Nerve segments with temporal dispersion in diabetes. Nerve segment Median Forearm Upper arm Ulnar Forearm Peroneal Leg Total
860
Number with TDItotal segments 2/30 016 0120 7/20 9/76
Conduction Block in Diabetic Neuropathy
neuropathy; nerve entrapments,? which occur frequently in diabetics; acquired demyelinating polyneuropathies, such as Guillain- BarrC syndrome6 or chronic inflammatory demyelinating polyneuropathy (CIDP)1917;and selected motor neuropathies." Hence, the finding of CB on an electrophysiologic study has important implications. In addition to limiting differential diagnosis, the finding of CB along nerve trunks suggests that segmental demyelination is an important part of the underlying pathology which should provide clues to pathogenesis. Several hypotheses for the pathogenesis of symmetric diabetic polyneuropathy have received support from experimental studies (for review see ref. 2). In human diabetic neuropathy, pathologic studies have shown both segmental demyelination and axon and debate continues as to the relative contribution of each.' If segmental demyelination were common in diabetes, then CB should be present in patients with diabetic polyneuropathy. We addressed this question in a series of diabetic patients referred for electrodiagnostic studies. T o determine if CB was found along a nerve segment, we applied the criteria for CB developed by Brown and Feasby.' They studied the median, ulnar, and peroneal nerves of 30 control subjects and found that the decline in p-p amplitude, between proximal and distal stimulation sites, never exceeded 20%. Likewise, the increase in duration between the 2 sites never exceeded 15%. Comparing these control values with those from a series of patients with GBS, they determined that CB was frequent in GBS, occurring in 68% of patients. In other studies of CB, the decline in p-p amplitudes used to determine abnormalities was set at 50%'? and 40%*.';however, it is unclear how these values were determined. More recently, Rhee et al.," using a computer simulation technique, have shown that temporal dispersion and interphase cancellation alone can produce decrements of CMAP amplitude of >50%, and of CMAP area of up to 50%. However, it is unknown whether these findings apply to humans. Criteria for determination of CB in chronic peripheral nerve disease, such as diabetic polyneuropathy, have not been developed. Ideally, the control values for determination of CB should be determined by studying nerves of patients with pure axonal polyneuropathy, as these would be the most precise controls. Another control population might be patients with motor neuron disease. In the absence of such control values for CB in
MUSCLE & NERVE
September 1991
chronic neuropathies, we used the Brown and Feasby criteria. We found CB in only 6 of 76 nerve segments. Among these nerve segments are potential, although infrequent, sites of nerve entrapment12 that could lead to CB. These include the pronator teres syndrome (median forearm segment) and flexor carpi ulnaris entrapment of the ulnar nerve (ulnar forearm segment). It is possible that with either “inching” techniques or more extensive conduction studies and EMG, these could have been localized more precisely. Assuming that CB in these segment was not related to nerve entrapment, our findings still suggest that CB in diabetic neuropathy is rare. The relative rarity of CB in our diabetic patients suggests that segmental demyelination is not a prominent part of the underlying pathology. Additionally, it is likely that nerve compression and immunologically mediated mechanisms, which commonly lead to peripheral nerve demyelination, do not play a role in pathogenesis of diabetic polyneuropathy. Thus, while our studies do not provide support for any of the current theories of pathogenesis of diabetic polyneuropathy, they provide indirect evidence against these two mechanisms. The magnitude of CB in the diabetic patients was modest compared with that of patients with GBS studied by Brown and Feasby.‘ Most of the latter patients had reductions in CMAP amplitude exceeding SO%, while the mean CMAP decrease in our patients was 28%, with the highest being 40%. This suggests that in DM, unlike in GBS, segmental demyelination is not widely distributed along the length of nerves. The GBS patients also had frequent CB at entrapment sites (30% of ul-
nar nerve across-elbow segments), while among our 8 patients, who had a total of 9 across-elbow ulnar nerve entrapments, only 1 showed CB. There are 2 potential sources of error in our study. First, apart from CB, a drop in p-p amplitude between distal and proximal stimulation sites without a change in -p duration could be due to failure to deliver maximal stimulation at the roximal site, interphase shift and cancellation,“ and axonal interruption between the stimulating electrodes.” We tried to exclude each of these as possible causes for apparent CB by attending to technical factors during the studies, and by excluding nerve segments with severely reduced CMAP amplitudes from our study. If we counted a nerve segment as having CB when, in reality, there was another explanation, then we would be overestimating the frequency of CB rather than underestimating it; thus, CB may be even less frequent than 7%. Second, if the degree of polyneuropathy in our patient population was mild, CB and segmental demyelination might not have been present. We think this is an unlikely explanation for our findings. We studied patients with a range of degrees of diabetic neuropathy, and the patients with CB were not found exclusively in those with clinically or electrophysiologically severe neuropathy. Nevertheless, another study specifically evaluating patients with severe diabetic polyneuropathy would be valuable. In conclusion, CB in diabetic neuropathy is uncommon. When present, the finding of CB in diabetics, especially in more than one nerve segment, should suggest alternative or additional causes for neuropathy, such as chronic inflammatory demyelinating polyne~ropathy.~
REFERENCES 1 . Albers JW: Inflammatory demyelinating polyradiculoneuropathy, in Brown WF, Bolton CF (eds): Clinical Electromyografhy. Boston, Butterworths, 1987, pp 209-244. 2. Asbury AK: Understanding diabetic neuropathy. N Engl J Med 1988;319:577-578. 3. Behse F, Buchthal F, Carlsen F: Nerve biopsy and conduction studies in diabetic neuropathy. J Neurol Neurosurg Psychiatty 1977;40: 1072- 1082. 4. Brown MJ, Asbury AK: Diabetic neuropathy. Ann Neurol 1984;15:2- 12. 5. Brown WF: The Physiological and Technical Basis of Electromyography. Boston, Butterworths, 1984. 6. Brown WF, Feasby TE: Conduction block and denervation in Guillain-Barre polyneuropathy. Brain 1984; 107:219239. 7. Brown WR, Yates SK: The quantitative assessment of conduction block in human entrapment neuropathies. Eleclroenceph Clin Neurophysiol 1983 ;56:(S52).
Conduction Block in Diabetic Neuropathy
8. Cornblath DR, Asbury AK, Albers JW, Feasby TE, Hahn AF, McLeod JG, Mendell JR, Parry GJ, Pollard JD, Thomas PK: Criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology (in press). 9. Cornblath DR, Drachman DB, Griffin JW: Demyelinating motor neuropathy in patients with diabetic polyneuropathy. Ann Neurol 1987;22:126. 10. Cornblath DR, Sumner AJ: Conduction block in neuropathies with necrotizing vasculitis. Muscle Nerve 1991,14: 185. 11. Daube JR: Electrophysiological testing in diabetic neuropathy, in Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D (eds): Diabetic Neuropathy. Philadelphia, Saunders, 1987, pp 162- 176. 12. Dawson DM, Hallett M, Millender LH: Entrapment Neuropathies, Boston, Little, Brown, 1983. 13. Dyck PJ: Pathology, in Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D (eds): Diabetic Neuropathy. Philadelphia, Saunders, 1987, pp 223-236.
MUSCLE & NERVE
September 1991
861
14. Dyck PJ, Thomas PK, Asbury AK, Winegrad AJ, Porte D: Diabetic Neuropathy, Philadelphia, Saunders, 1987. 15. Feasby TE, Brown WF, Gilbert JJ, Hayn AF: The pathological basis of conduction block in human neuropathies. J Neurol Neurosurg Psychiatry 1985;43:239-244. 16. Halar EM, Graf RJ, Halter JB, Brozovich FV, Soine TL: Diabetic neuropathy: A clinical, laboratory, and electrodiagnostic study. Arch Phys Med Rehab 1982;63:298-303. 17. Lewis RA, Sumner AJ: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 1982;32:592-596. 18. Pestronk A, Cornblath DR, Ilyas AA, Baba H, Quarles RH,
862
Conduction Block in Diabetic Neuropathy
Griffin JW, Alderson K, Adams RN: A treatable multifocal motor neuropathy with antibodies to GM 1 ganglioside. Ann Neurol 1988;24:73-78. 19. Rhee EK, England JD, Sumner AJ: A computer simulation of conduction block: Effects produced by actual block versus interphase cancellation. Ann Neurol 1990;28:146- 156. 20. Thomas PK, Lascelles RG: The pathology of diabetic neuropathy. Q J Med 1966;35:489-509, 21. Trojaborg W, Lange DJ, Latov N, Younger DS, Lovelace RE, Rowland LP: Conduction block and other abnormalities of nerve conduction in motor neuron disease. Neurology 1990;4O(suppl 1): 182.
MUSCLE & NERVE
September 1991
