J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

Journal ofNeurology, Neurosurgery, andPsychiatry, 1976, 39, 453460

Peripheral motor nerve function in diabetic autonomic neuropathy

D. J. EWING', A. A. BURT, I. R. WILLIAMS, I. W. CAMPBELL, AND B. F. CLARKE From the University Departments ofMedicine and Medical and the Diabetic and Dietetic Department, Royal Infirmary, Edinburgh

SYNOPSIS Motor conduction velocity was measured in the median, ulnar, and common peroneal nerves of 32 diabetics with clinical features of autonomic neuropathy. The responses to the Valsalva manoeuvre and sustained handgrip, and the postural fall in blood pressure were used to assess the integrity of the autonomic . Abnormalities in the three autonomic function tests were significantly correlated with the forearm conduction velocity of the ulnar nerve, the conduction velocity and motor latency of the common peroneal nerve, and the H reflex. These results show that, in diabetics with autonomic neuropathy, abnormalities in the parallel Protected by copyright. changes in the peripheral nerves. Any diabetic with should be examined for evidence of autonomic nervous system involvement.

Peripheral neuropathy is a well-recognized METHODS complication of mellitus which can be PATIENTS Thirty-seven diabetics with features detected even at diagnosis (Ward et al., 1971). suggestive of autonomic neuropathy were selected With increasing duration of diabetes there is for study. All patients had normal levels of serum increasing impairment of nerve conduction vitamin B12, except one who was excluded; four others (Gregerson, 1967). Autonomic neuropathy is were excluded as complete nerve conduction data likewise common, and, although it does not could not be obtained. The ages of the remaining 32 become clinically manifest until the later stages, patients ranged from 26 to 63 years (mean 48 ± 10 tests of autonomic function show abnormalities years), and the duration of diabetes from two to 33 years (mean 19±9 years). All but four were before the onset of symptoms (Sharpey-Schafer or insulin dependent, and none had cardiac failure http://jnnp.bmj.com/ and Taylor, 1960; Ewing et al., 1974; Murray et was receiving hypotensive therapy. None of the al., 1975). patients was known to be an alcoholic, all had normal This study compares autonomic and peripheral protein bound iodine values, and 18 had blood urea nerve function in diabetics with clinical features levels within normal limits (less than 6.6 mmol/l). suggestive of autonomic neuropathy to see Of the remainder 11 had blood urea levels between whether the nerve damage known to occur in the 6.6 and 10.0 mmol/l, while only three had higher peripheral motor nerves was paralleled by similar levels of 10.5, 10.6, and 12.1 mmol/l. All patients had changes in the autonomic nerves. A preliminary one or more of the following features: impotence on October 1, 2021 by guest. report of nerve conduction in diabetic autonomic (25 patients); symptoms of postural hypotension elsewhere (Ewing (seven patients); intermittent diarrhoea (nine neuropathy has been published patients); gastric fullness or delay in emptying (four et al., 1973); this paper gives the findings in more patients); a history of hypoglycaemic unawareness detail and analyses them in relation to the (11 patients); gustatory sweating (two patients); and autonomic function tests. reduced sweating in the legs (one patient). Ten patients had symptoms of a peripheral neuropathy in I Address for correspondence: Dr D. J. Ewing, Department of including numbness and painful cramps the legs Medicine, Royal Infirmary, Edinburgh EH3 9YW. but without signs. Nineteen had signs of peripheral (Accepted 13 January 1976.) neuropathy on clinical examination including absent 453 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

454 D. J. Ewing, A. A. Burt, L R. Williams, L W. Campbell, and B. F. Clarke

knee and ankle jerks and diminished sensation in the However, they had a significantly longer duration lower limbs. of diabetes, their conduction velocity was slower Autonomic function was assessed by the Valsalva and terminal latency longer. manoeuvre, the response to sustained handgrip, and the postural fall in blood pressure. The Valsalva manoeuvre was carried out using a standardized RELATION BETWEEN THE CLINICAL FEATURES OF technique as previously described (Ewing et al., 1973). AUTONOMIC AND PERIPHERAL NEUROPATHY The heart rate response was measured by a simul- (Table 3) Five patients with impotence alone taneous electrocardiograph (ECG). The results were had signs of a peripheral neuropathy in contrast expressed as theValsalva ratio(the ratio ofthelongest with 14 of those with other features of autonomic RR interval in the ECG after the manoeuvre to the neuropathy (x2=5.15, P < 0.05). shortest RR interval during the manoeuvre), a ratio of 1.10 or less being defined as an abnormal response, CLINICAL FEATURES OF AUTONOMIC NEUROPATHY 1.11 to 1.20 as borderline, and 1.21 or more as a AND NERVE CONDUCTION (Table 4) When the normal response. A standardized sustained handgrip MCV results were analysed in terms of the test was also performed as described previously (Ewing et al., 1974). A rise in diastolic blood pressure clinical features of autonomic neuropathy, there of less than 10 mmHg was defined as abnormal, were significant differences between the groups 11-15 mmHg as borderline, and 16 mmHg or more as only in the terminal motor latency ofthe common normal. Postural hypotension was defined as a fall in peroneal nerve (P < 0.02). systolic blood pressure of 30 mmHg or more im-

mediately on standing up from the supine position. CLINICAL FEATURES OF PERIPHERAL NEUROPATHY Protected by copyright. Motor conduction velocity (MCV) was measured AND AUTONOMIC FUNCTION TESTS (Table 5) in the median, ulnar, and common peroneal nerves. Table 5 shows the results of the autonomic The latency of the H reflex in the triceps surae was function tests when the subjects were grouped measured after stimulation of the medial popliteal according to the clinical features of peripheral nerve at the popliteal fossa. Surface electrodes were used for stimulating with 1 cm silver discs placed neuropathy. 5 cm apart. Standard DISA surface recording elec- trodes and a DISA two channel electromyograph NERVE CONDUCTION AND AUTONOMIC FUNCTION were used in all studies, which were performed in the TESTS MCV and Valsalva manoeuvre (Table same environment with skin temperatures maintained 6) The Valsalva ratio was significantly cor- above 30°C. related in the group as a whole only with the The physical examination of the patients, the auto- forearm conduction velocity in the ulnar nerve nomic function tests, and the nerve conduction (r=0.375, P<0.05); but, when the group was studies were carried out by three independent divided into those with normal and abnormal observers. Standard statistical methods were used to Valsalva ratios, those with abnormal ratios had calculate the significance of the results. significantly more abnormal MCV measurements http://jnnp.bmj.com/ in the forearm conduction velocity in the ulnar RESULTS nerve (P<0.01), the motor latency (P<0.005), and conduction velocity (P < 0.02) ofthe common CLINICAL FEATURES OF AUTONOMIC NEUROPATHY peroneal nerve and the H reflex latency (P < 0.05). AND AUTONOMIC FUNCTION TESTS (Table 1) When the patients were grouped into those with MCV and response to sustained handgrip (Table impotence alone and those with other features of

6) Within the whole group the rise in diastolic on October 1, 2021 by guest. autonomic neuropathy, there were significant blood pressure correlated significantly with the differences between the results of the autonomic forearm conduction velocity in the ulnar nerve function tests in the two groups. (r=0.356, P <0.05), the motor latency of the common peroneal nerve (r=0.552, P < 0.01) and CLINICAL FEATURES OF PERIPHERAL NEUROPATHY the conduction velocity of the common peroneal AND NERVE CONDUCTION (Table 2) Patients nerve (r=0.390, P <0.05). When normal and with symptoms and signs of a peripheral neuro- abnormal handgrip responses were separated, pathy did not, as a group, differ significantly in there was a significant difference only in the motor mean age from those having no such features. latency ofthe common peroneal nerve (P < 0.005). J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

Peripheral motor nerve function in diabetic autonomic neuropathy 455

ei e~i

-H -H-i -H

'tH -H -H -H

~~~~~f -H H1- -H

N0~~~~~~~0W

06 Ni e~) 0 -H - H110 -H

Li 00 CR tn~~HiQ N;

~ ~ ~ ~ ~ r t

0 1:4 616 e'it t, Protected by copyright.

- ~0 L ~IZ ~ 00

H00~~~~~~~~~~~~~~~~~~ ~~~~0 0

ei~~~~~~~~~e

CA O~~~~~~~~0

V~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- CACo http://jnnp.bmj.com/ U~~ ~ ~ ~ ~ ~ ~

.0

z~~~~~~~~~

mC~~~Z on October 1, 2021 by guest.

0) J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

456 D. J. Ewing, A. A. Burt, L R. Williams, L. W. Campbell, and B. F. Clarke

N en -H -H on

en -n t-n

6 CZ,0 0~ oIR-H r-_+00 L. 'I± ±6Z 0~ 0: 0s 0~ N t- CU. 0: 00 01, 00c "t £c a

L

L. 0x rl 0~ 01 n U \_-H en_ z tt el; U U 0u 0 L; z 0 0 0x O.-H LL +00

+ Protected by copyright. o4 6 ~oo Z 4 o 0

0 00 z _- F- zw eO " Ci t0t .!L. 12-H ( z4 -f F- w -z

110 C4 u A. .2 E 0 (L) U

E 0 - 0. o :00 1 % - 0 .0 vo0. z0 0 E &a r. -H -H0

0 on October 1, 2021 by guest.

0

.0 ._ 0O0 _04) 0 4) 0

U) eo 0E S J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

Peripheral motor nerve fuinction in diabetic autonomic neuropathy 457

r-~C 00r

4h+0/) -4A-+ -H -H0En 0 0 ~~~~~~~~~~~~~ oooo7~~~~~~~~~~~~0W 9- v r_-.7 ~

00 r'- - 0O 06 0- 0~~~~~~~4C 0.~~~~~~~~~~~ C4N~ ~ ~ ~ -

LL.~~~~~~~z C E 0 O~~~~~~~~~~~~~

LU ro- (A 0 0C3000 -H±1 Q~~~ r--~00O r 'I 42 W)~~~~~~~~~~~~~~~r~rorW)) 0 0-- ~~~~~~~~~~~~LU0

U W)~~~~~~ 0. ~~~~~~~~~~~~~~~LU0 Ho-r + H -Ho0 -Hooo Ztc ~~~~~LU* ~r-~ Protected by copyright. H00~~~~~~~~~.00 ~ ~ ~ ~ ~ ~ ~..0V,W ~ ~ 7 0 Z. 7. -H 7~ 00~ I ~0 ~0'r- \00 or- O ~~~HH U H r-0~~~~~~~~~~~~~~~~~~ r-1r- ~ r-

o~~~~~ ~ ~ ~ ~ ~ ~ ~ LUH 0~~~~~~~~~~~~~~~-

Z~ < - H s- 0O0)~~~~~0 7LU - Z 0 LU C.~~~~~~~~~~~~~~~~~~~~77~~~~~~~~~~~~~~~LU

0.0 ~~~~~~~00(10 iO0 CZ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C

7U HU

D 1~~~~~~~~0o-r-00 00 "C - 00~~~~~~~~~~~~ -~~~~~~~~~~~~ 0. ~ ~ ~ ~ ~ ~ ~ n c http://jnnp.bmj.com/ Ho > 0 +0 ++ 0 +02 . 0 U.,~~~~~~~~0~~*~o-7--~r.o

E 0 0

7-~~~~~~~~~~~~o0 0. 0~~~~~~~~~~~~~~~C a0 on October 1, 2021 by guest. (.~~~~~~)~~~~~....o0 -~~~~~~~~~~~~~~~~ ~ Eo C 0).-C. .-~~~~~~~~~~~~~~~~~~~~~~~ ~ ~ ~ o 0 ;o. 0~~~~~~~~~~~~~~~~~~~~~~~~~~0(Z.

0~~~~~~~~~~~~~~~~~ 0 0~~~~~~~~~~~~~~~~~~~~~~~> O J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

458 D. J. Ewing, A. A. Burt, I. R. Williams, I. W. Campbell, and B. F. Clarke

MCV and posturalfall in blood pressure (Table more specifically at the peripheral and the auto- 6) The fall in systolic blood pressure also cor- nomic nervous systems independently. Martin related significantly with the forearm conduction (1953b) showed abnormalities of sudomotor and velocity in the ulnar nerve (r=0.466, P<0.01), vasomotor function in 20 patients with diabetic the motor latency of the common peroneal nerve peripheral neuropathy, and another study at- (r=0.681, P<0.001), the conduction velocity of tempted to assess the amount ofautonomic nerve the common peroneal nerve (r=0.573, P < 0.001), damage in patients with peripheral neuropathy and also the H reflex latency (r=0.374, P < 0.05). by using the Valsalva manoeuvre (Bishnu and Again, however, when those with a normal and an Berenyi, 1971). The present study, however, has abnormal fall in systolic blood pressure were examined the relationship in more detail from the separated, there was a significant difference only opposite viewpoint by measuring peripheral in the motor latency of the common peroneal nerve function in patients with autonomic nerve (P < 0.05). neuropathy. As has been demonstrated previously (Ewing Renal function, MCV, and autonomic function et al., 1973), the study shows that diabetics with tests The level of the blood urea correlated impotence alone, which is often stated to be the significantly, in the group as a whole, with the only manifestation of diabetic autonomic neuro- forearm conduction velocity in the ulnar nerve pathy in some patients (British Medical Journal, (r=0.447, P < 0.05), and the motor latency of the 1974), had more normal autonomic vascular common peroneal nerve (r=0.452, P <0.01). It reflexes than those with other features of was also significantly correlated with the sustained autonomic neuropathy. Similarly, the clinical Protected by copyright. handgrip response (r=0.356, P<0.05) but not symptoms and signs of peripheral neuropathy with the Valsalva ratio (r=0.286 NS) nor with were closely reflected by the changes in MCV. fall in systolic blood pressure (r=0.214 NS). As might be expected, therefore, few patients with impotence alone also had signs of a peripheral neuropathy. DISCUSSION In addition, this study now provides more This study shows that, in diabetics with auto- detailed information of the relation between nomic neuropathy, abnormalities in the auto- MCV and autonomic vascular reflex responses. nomic nervous system parallel changes in Whether one examines the regression relationship peripheral nerves. Previous workers have between the individual values, or the differences demonstrated that peripheral nerve function, as between groups with and without abnormal determined by motor nerve conduction velocity, is autonomic function tests, it is clear that changes impaired in diabetics and that, although this is in the common peroneal nerve and in the forearm of of the ulnar nerve are reflected by the presence peripheral neuro- conduction velocity http://jnnp.bmj.com/ pathy, some diabetics have early changes ofMCV closely related. The three autonomic responses without clinical evidence ofa neuropathy (Mulder measured-the Valsalva manoeuvre, sustained et al., 1961; Lawrence and Locke, 1961; Skillman handgrip, and postural fall in blood pressure- et al., 1961). Others have documented abnormal are all cardiovascular reflexes. Although auto- autonomic function tests in diabetics both with nomic neuropathy affects all systems ofthe body, and without clinical features ofautonomic neuro- it is the vascular responses that are both simplest pathy (Sharpey-Schafer et al., 1960; Ewing et al., to measure non-invasively and easiest to quantify, 1974; Murray et al., 1975). It is surprising perhaps and they bear a close relation to symptoms on October 1, 2021 by guest. that there has been little attempt to relate the two, (Ewing et al., 1973). Their reflex pathways are either in terms of clinical features, or in terms of not clear in their entirety, but probably involve objective testing. Early authors considered auto- efferent fibres to the blood vessels in the lower nomic neuropathy as one part of the spectrum of limbs in all three responses and this would (Rundles, 1945; Martin, therefore suggest that damage to the nerves 1953a); but, with the development of objective conveying the reflexes occurs in parallel with the means of testing both nerve conduction and motor nerve damage measured by the nerve autonomic function, later workers have looked conduction studies. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

Peripheral motor nervefunction in diabetic autonomic neuropathy 459

Peripheral neuropathy in diabetics is common 1974), the greater splanchnic nerves (Low et al., in the lower limbs, but rare in the upper limbs 1975), and the lumbar sympathetic ganglia (Kott (Noel et al., 1971; Noel, 1973), and, although the et al., 1974). neuropathy is predominantly sensory, changes in Diabetic peripheral neuropathy and diabetic the motor fibres correlate well with the clinical autonomic neuropathy are often considered as picture (Lamontagne and Buchthal, 1970) and distinct clinical entities, with their different with changes in the sensory nerves (Gilliatt and symptomatology and features. They share, Willison, 1962). On the basis of these observa- however, a common pathogenesis, and the tions, it would be expected that the leg MCV present study shows that abnormalities occurring gives a better correlation than the arm MCV with in the autonomic nervous system are paralleled the autonomic function tests. by changes in the peripheral nervous system. In Although the level of blood urea correlated practice, therefore, as diabetic peripheral neuro- both with the MCV findings and the autonomic pathy has been more closely studied and function tests, and could be presumed to be the understood, it means that any patient with cause of both abnormalities, this does not stand peripheral neuropathy due to his diabetes should up to critical analysis. In chronic renal failure be examined carefully for evidence of autonomic the blood urea, although a crude index of renal nerve involvement. function, is usually much higher than the levels changes in MCV are detected. in this study before REFERENCES

Of those patients with normal blood urea values, Protected by copyright. seven had clinical signs ofperipheral neuropathy; Appenzeller, O., and Richardson, E. P. (1966). The sympathetic chain in patients with diabetic and alcoholic and although changes of autonomic function 1205-1209. have been described in chronic renal failure . Neurology (Minneap.), 16, Bishnu, S. K., and Berenyi, M. R. (1971). Circulatory (Ewing and Winney, 1975), eight patients with reflex response in diabetic patients with or without normal blood urea levels had abnormal auto- peripheral neuropathy. Journal of the American Geri- nomic function tests. Impaired renal function, atrics Society, 19, 159-166. although it cannot be incriminated causally, may British Medical Journal (1974). Editorial: diabetic nevertheless contribute to the neuropathy. autonomic neuropathy, 3, 2-3. Fibres mediating the autonomic responses are Chopra, J. S., Hurwitz, L. J., and Montgomery, D. A. D. both myelinated and unmyelinated, while the (1969). The pathogenesis of sural nerve changes in MCV is measured in large myelinated fibres diabetes mellitus. Brain, 92, 391-418. ending at neuromuscular junctions. The most Dyck, P. J. (1971). The concept of secondary segmental demyelination. Proceedings of the 2nd International constant underlying lesion in the peripheral Congress of Muscle Diseases. Excerpta Medica Inter- neuropathy of diabetes has been shown to be national Congress Series, no. 237, 74 segmental demyelination (Thomas and Lascelles, Ewing, D. J., Campbell, I. W., Burt, A. A., and Clarke, http://jnnp.bmj.com/ 1965, 1966; Locke, 1967). This is usually thought B. F. (1973). Vascular reflexes in diabetic autonomic to be due to disturbance of the Schwann cells, neuropathy. Lancet, 2, 1354-1356. but it has been suggested that segmental Ewing, D. J., Irving, J. B., Kerr, F., Wildsmith, J. A. W., demyelination may result from axonal dis- and Clarke, B. F. (1974). Cardiovascular responses to sustained handgrip in normal subjects and in patients turbance (Dyck, 1971; Thomas, 1971). Axonal with diabetes mellitus: a test of autonomic function. loss has been demonstrated in more severe cases Clinical Science and Molecular Medicine, 46, 295-306.

(Chopra et al., 1969). Histological studies of Ewing, D. J., and Winney, R. (1975). Autonomic function on October 1, 2021 by guest. autonomic nerves in diabetics are limited, but in patients with chronic renal failure on intermittent widespread changes have been noted in the haemodialysis. Nephron, 15,424-429. sympathetic chain and ganglia in patients dying Faerman, I.. Glocer, L., and Celener, D. (1973). Auto- from various complications (Appenzeller and nomic nervous system and diabetes. Histological and Richardson, 1966; Olsson and Sourander, 1968). histochemical study of the autonomic nerve fibres of the in diabetic patients. Diabetes, 22, Damage has also been noted to the autonomic 225-237. fibres supplying the bladder wall (Faerman et al., Faerman, I., Glocer, L., Fox, D., Jadzinsky, M. D., and 1973); the corpora cavernosa (Faerman et al., Rapaport, M. (1974). Impotence and diabetes. Histo- 1974), the innervation of the oesophagus (Smith, logical studies of the autonomic nervous fibres of the J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.39.5.453 on 1 May 1976. Downloaded from

460 D. J. Ewing, A. A. Burt, 1. R. Williams, L W. Campbell, and B. F. Clarke

corpora cavernosa in impotent diabetic males. Diabetes, Noel, P. (1973). Sensory nerve conduction in the upper 23, 971-976. limb at various stages of diabetic neuropathy. Journal of Gilliatt, R. W., and Willison, R. G. (1962). Peripheral Neurology, Neurosurgery, and Psychiatry, 36, 786-796. nerve conduction in diabetic neuropathy. Journal of J. Neurology, Neurosurgery and Psychiatry, Noel, P., Lauvaux, P., and Pirart, J. (1971). Upper limb 25, 11-18. diabetic neuropathy: a clinical and electrophysiological Gregerson, G. (1967). Diabetic neuropathy: influence of study. Hormone and Metabolic Research, 3, 386-392. age, sex, metabolic control and duration of diabetes on motor conduction velocity. Neurology (Minneap.), 17, Olsson, Y., and Sourander, P. (1968). Changes in the 972-980. sympathetic nervous system of diabetes mellitus. A preliminary report. Journal of Neurovisceral Relations, Kott, I., Urca, I., and Sandbank, U. (1974). Lumbar 31, 86-95. sympathetic ganglia in atherosclerotic patients, diabetic and non-diabetic. Archives ofSurgery, 109, 787-792. Rundles, R. W. (1945). Diabetic neuropathy. General Lamontagne, A., and Buchthal, F. (1970). Electro- review with report of 125 cases. Medicine (Balt.), 24, physiological studies in diabetes mellitus. Journal of 111-160. Neurology, Neurosurgery and Psychiatry, 33, 442-452. Sharpey-Schafer, E. P., and Taylor, P. J. (1960). Absent Lawrence, D. G., and Locke, S. (1961). Motor nerve circulatory reflexes in diabetic . Lancet, 1, conduction velocity in diabetes. Archives of Neurology 559-562. (Chic.), 5, 483-489. Skillman, T. G., Johnson, E. W., Hamwi, G. J., and Locke, S. (1967). Axons, Schwann cells and diabetic Driskill, H. J. (1961). Motor nerve conduction velocity neuropathy. Bulletin of the New York Academy of in diabetes mellitus. Diabetes, 10, 46-51. Medicine, 43, 784-791. Smith, B. (1974). Neuropathology of the oesophagus in Low, P. A., Walsh, J. C., Huang, C. Y., and McCleod, diabetes mellitus. Journal of Neurology, Neurosurgery, Protected by copyright. J. C. (1975). The sympathetic nervous system in diabetic and Psychiatry, 37, 1151-1154. neuropathy-a clinical and pathological study. Brain, 98, 341-356. Thomas, P. K. (1971). The morphological basis for Martin, M. M. alterations in nerve conduction in peripheral, neuro- (1953a). Diabetic neuropathy. A clinical pathy. Proceedings of the Royal Society ofMedicine, 64, study of 150 cases. Brain, 76, 594-624. 295-298. Martin, M. M. (1953b). Involvement of autonomic nerve fibres in diabetic neuropathy. Lancet, 1, 560-565. Thomas, P. K., and Lascelles, R. G. (1965). Schwann cell abnormalities in diabetic neuropathy. Lancet, 1, Mulder, D. W., Lambert, E. H., Bastron, J. A., and 1355-1357. Sprague, R. G. (1961). The neuropathies associated with diabetes mellitus. A clinical and electromyographic Thomas, P. K., and Lascelles, R. G. (1966). The pathology study of 103 unselected diabetic patients. Neurology of diabetic neuropathy. Quarterly Journal ofMedicine, (Minneap.), 11, 275-284. 35, 489-509. Murray, A., Ewing, D. J., Campbell, I. W., Neilson, Ward, J. D., Barnes, G. C., Fisher, D. J., Jessop, J. D., and J. M. M., and Clarke, B. F. (1975). RR interval varia- Baker, R. W. R. (1971). Improvement in nerve con- tions in young male diabetics. British Heart Journal, duction following treatment in newly diagnosed 37, 882-885. diabetes. Lancet, 1, 428431. http://jnnp.bmj.com/ on October 1, 2021 by guest.