Journal ol'the Neurological Sciences, 1976, 28: 217 223 JC~ Elsevier Scientific Publishing Company, Amsterdam - Printed

in The Netherlands

217

THE NEUROPATHIC FACTOR IN THE AETIOLOGY OF DIABETIC FOOT ULCERS

M. J. G. HARRISON and !. B. FARIS

Departme~It o1" Neurological Studies, The Middlesex Hospital, Mortimer Street, London (Great Britain)
(Received 3 October, 1975)

SUMMARY

The nature and severity of changes due to peripheral neuropathy has been assessed clinically and electrophysiologically in 39 diabetic patients, 23 of whom had penetrating ulcers on the sole of the foot. Patients suffering from diabetic foot ulcers were found to have sensory loss to pin prick in the feet more frequently than "controls". The electrophysiological measurements confirmed the severity of the neuropathy with absent sural nerve action potentials in 95 ~ of the ulcer patients and inexcitability of small foot muscles in 40~. Weakness of small foot muscles due to neuropathy is thought to underlie the foot deformity that leads to maldistribution of weight bearing, and subsequent pressure necrosis of analgesic skin.

INTRODUCTION

Ulceration of the foot represents one of the most serious complications of diabetes mellitus, contributes significantly to the morbidity of the disease, and is a frequent cause for admission to hospital of diabetic patients. Septic, neuropathic, ischaemic and combined aetiologies have been proposed (Oakley, Catterall and Martin 1956). There has been increasing awareness of the importance of the neuropathic role (Kelly and Coventry 1958; Catterall 1973). Sensory loss in the feet due to diabetic neuropathy has been considered a critical factor (Ellenberg 1968), together with deformity of the architecture of the foot (Kelly and Coventry 1958). The latter change with a splayed metatarsus, hammer toes, hallux valgus, and prominent metatarsal heads is well-recognised clinically. Recently the maldistribution of weight bearing of such feet, and the importance of pressure points have been demonstrated quantitatively by Stokes, Faris and Hutton (1975). The role of diabetic neuropathy in producing foot deformity and in predis-

218
TABLE I AGE OF PATIENTS A N D D U R A T I O N OF DIABETES Age (years) 25 34 Footlesion (n 23) Controls (n : 16)
0

35 44 3 0

45 54 5 2

55 64 8 5

65 74

7~, and above

Duration (years) 0 10 F o o t lesion (n 23) Controls

(n .... 16)

11 2(J 9 7

21 30 5 2

31 or more 2

7 7

posing to foot ulcers has been further studied in a group of diabetic patients, some of whom were the subject of the parallel study by Stokes et al. (1975). The special feature of the electrophysiological study has been the attempt to measure the severity of the neuropathy affecting the feet.
METHODS

Twenty-three patients had foot ulcers. It was considered impossible to identify a rigorously-matched control group that would take account of all the variables (age, sex, duration and severity of diabetes, types of treatment and quality of control, presence of other complications, etc.). Instead, patients were chosen from the diabetic clinic or when in the ward to fill gaps in the study programme. An attempt was made to choose patients whose age and duration of diabetes was not too dissimilar from that of the ulcer group (Table 1), but other characteristics were not matched. There were 16 men and 7 women with ulcers, and 6 men and t0 women without:_ A clinical examination noted any evidence of peripheral neuropathy. Pain loss was assessed by the appreciation of pin prick. Electromyographic (EMG) studies included attempts to measure median nerve (index to wrist) sensory action potentials (Gilliatt and Sears 1958), sural nerve action potentials (Burke, Skuse and Lethlean 1974), and median and lateral popliteal motor nerve conduction velocity (Thomas, Sears and Gilliatt 1959). In addition the amplitude of the muscle action potential that could be evoked by stimulation of the peripheral nerves in 2 small foot muscles was recorded. Surface electrodes were positioned over the extensor digitorum brevis and over the abductor hallucis brevis during stimulation of the lateral and medial popliteal nerves at the knee and ankle. The electrode placement was varied tO produce the maximum response with supramaximal stimulation. When no response was

219
TABLE 2 CLINICAL EVIDENCE OF NEUROPATHY Grading 0 Foot lesions (n 23) Controls (n : 16) Grade Grade Grade Grade 0: h I1: I11 : 0 2 I 9 7 11 9 4 I11 5 3

n o r m a l clinical examination. absent ankle jerks sensory change. weakness o f toe m o v e m e n t s . weakness o f foot a n d toe movement.

obtained the maximal stimulus used was a square wave pulse of 0.5 msec at 200 V. The absence of a recordable muscle action potential was often due to visible and palpable wasting of the small foot muscles. However, in some cases a high threshold in the peripheral nerve may have contributed to a failure to produce a response.
RESU LI-S

Clinical evidence of neuropathy

There was evidence of peripheral neuropathy in all ulcer patients. Two of the control patients had no signs of neuropathy. The patients were placed in one of 3 categories according to the severity of the neuropathy as defined in Table 2. There is a tendency for the ulcer patients to show a more extensive neuropathy with more of the ulcer patients having detectable weakness (14/23, 61 ~ cf. 7/16, 44~/o) though the difference is not statistically significant. The sensory examination appeared to show a real difference, however. All but one of the ulcer patients had a demonstrable disturbance of appreciation of pin prick (Table 3), and this was more extensive than in the control group. In each case the sensory disturbance was in a distal stocking distribution. The single exception was a patient who had evidence
TABLE 3 LOSS O F P A I N S E N S I B I L I T Y N o loss F o o t lesions (n 23) Controls (n : 16) I I0 Toes 5 2 Foot 5 1 Below shin 2 1 Below knee 9 1 Below thigh

Pain loss in 22/23 foot lesion patients, and 6/16 controls; Z'~

13.0, P < 0.001.

220 TABLE 4 EMG FINDINGS No response from No response from extensor digitorum abductor hallucis brevis brevis No sural nerve action potential Median nerve sensor3action potentia! small

Foot lesion patients Controls

Z"P

18:41 (44 %) 5"29 (I 7 %} 4.8 0.05

19'44 (43 '!,) 1'30 (3 ".) 1~.5 I).l){)I

36.38 (95 "i0 t I,,25 (44 "<,l I8,9 0.00 [

(,23 t70 '!D (, 15{41) % I ~ N.N

of unilateral medial and lateral popliteal nerve lesions with wasting and ~,'eakness of the ulcerated foot. There was unilateral disturbance of touch, vibratiol~ and oint position sense but not a p p a r e n t l y to pin prick.

EMGfindings (Table 4) Median nerve sensory action potential. This was sought in all 23 ulcer patients.
and was small 1 - 6 #V) or absent in 16 (70i,). It was sought in 15 of the control patients and was small or absent in 6 (400;). This trend is not statistically significant. Sural nerve action potential. This was sought in 38 limbs in the ulcer group and was absent in 36 (95 '}{i). It was recordable, however, in 14 of 25 legs tested in the control group. This difference is highly significant (Z" 18.9. P . : 0.001 ).

Musele action potential in the extensor digitorum brevis. The response from the
extensor digitorum brevis was measured in 41 legs from the ulcer group, tn 18 instances there was no response (44,:YI;). in the control group no response could be obt a i n e d in 5/29 ( 17 7{j. This difference is significant at the P -:: 0.05 level. Muscle action potential in the abductor hallucis brevis. There was no response from this muscle in 19/44 instances in the ulcer group (43%) but a response was recorded from all but 1 of 30 limbs tested in the control patients. This difference is highly significant (Z" 13.5, P - : 0.0011.

TABLE 5 MOTOR CONDUCTION VELOCIT'~ Lateral popliteal nerve n mean (m/sec) 35.5 41 50 range (m/secl 25-45 28 56 35.5-63.5 Medial popliteal nerv~ n mean (m/sect 37 40 57 range (m/see) 25 47 28 59 52 67

Ulcer patients Control patients Normal subjects~' Thomas et al. 1959.

19 27 30

23 26 25

221

Motor conduction velocity. In many cases a conduction velocity could not be calculated since no response was obtained from the appropriate foot muscle. The data are therefore difficult to assess and comparisons are of doubtful significance. The figures obtained are shown in Table 5. There was a slight tendency for lower velocities to be found in the ulcer patients, particularly in the lateral popliteal nerve.
DISCUSSION

Only limited comparisons can be made with the "control" group since the matching was not exact. For example, there was a tendency for the ulcer patients to have more long-standing diabetes, and this factor may have contributed to the greater severity of the neuropathy that they showed (Pirart 1965). However, 2 features of the neuropathy in the ulcer patients would appear to be of particular importance, one sensory, and one motor. Firstly the clinical examination showed that all but one of the patients with diabetic foot ulcers had distal sensory loss including loss of the appreciation of superficial pain, and that this loss was often extensive. The electrophysiological measurements paralleled the clinical findings in that it was rarely possible to record a sural nerve action potential in the patients with foot ulcers. The median sensory action potentials also tended to show more severe changes than in the control group, but the difference was not marked, and was not statistically significant. The clinical and electrophysiological evidence thus supports the clinical view that sensory neuropathy in the lower limbs is important in the development of the characteristic ulcer. On the motor side, weakness of leg muscles was detected in one half of the patients with ulcers, with particular weakness of toe or foot movements. The prevalence of the changes was greater in the ulcer patients than in the control group but the difference was not statistically significant. This makes the evidence of abnormality of small foot muscles in the ulcer patients all the more interesting. It is difficult to measure clinically the power of intrinsic foot muscles, though the clawed deformity can be attributed to weakness of small muscles (Ellenberg 1968). The E M G findings in the present study, however, demonstrate that the small foot muscles, particularly abductor hallucis brevis in the sole are often inexcitable in the ulcerated foot, due to the severity of peripheral neuropathy. The importance of neuropathic weakness of intrinsic foot muscles is thought to be in the effect it has on the architecture of the foot, and thereby on the distribution of weight bearing. Stokes et al. (1975) have recently shown that diabetic patients show a lateral shift of the highest maximum load on the forefoot during walking and a decrease in the load carried by the toes. Six patients with ulcers showed maximum loading at the site of their ulcer. The evidence from the present study together with that by Stokes et al. (1975) would suggest that diabetic neuropathy in the lower limbs may lead to intrinsic foot muscle we~.kness, and thereby to altered weight bearing on the heads of the metatar-

222

sals. When this is combined with analgesia due to sensory neuropathy utcc~ation ma} occur. There has been much debate about a possible vascular factor in the aetiology of these ulcers. Catterall (1973) found that there was rarely any evidence of atheromatous disease in major leg arteries in patients with diabetic foot ulcers. Disease of smaller vessels and capillaries may play a role however. Ferrier (1967) showed thai metatarsal and plantar arch arteries in the diabetic were often abnormal and ~howed obstructive changes even when clinical examination had revealed normal loot pulses. Faris (1975) has found that the blood pressure gradient between anktc and toe is steeper than normal in the ulcer patients investigated in the present study, in addilion there is some evidence of thickening of capillary basement membranes in the skin ~f the toes of diabetics (Banson and Lacy 1964). The combination of obstructive changes in foot arteries and capillaries could well add an ischaemic factor to the vulnerability of the sole of the foot. In addition to these possible obstructive changes in the vessels Moorhousc. Carter and Doupe (1966) and Ozeran, Wagner, Reimer and Hill (1970) have i~resented evidence that autonomic neuropathy in diabetic patients may cause vasoconstriction. The study by Moorhouse et al. (1966) showed that many patients with evidence of diabetic neuropathy had lost reflex vasoconstrictor and vasodilator rcsponscs in digital vessels in response to body heating and cooling. Furthermore digital vessel> showed cold hypersensitivity. The consequence of these 2 abnormatitie,, was that autonomous vasospasm might occur in digital vessels if the periphery of the limb was cold. and that central warming would fail to overcome this vasoconstrictitm. Ozeran et al. (1970), using galvanic skin resistance measurements, also found cvider~ce of vasospasticity in the feet o1"some diabetics. Further studies are needed to delineate the possible role of these vascular factors and assess their contribution to the vulnerability of the skin of the sole of the foot to ischaemic necrosis under pressure. ACKNOWLEDGEMENTS We are grateful to Dr, J. D. Nabarro and Professor Le Quesne for permission to study patients under their care and to Drs. C. J. Earl and P. M. Le Quesne tk)r helpful discussions,

REFERENCES Banson, B. B. and P. E. Lacy (1964) Diabetic microangiopathy in human toes, Ame: J. Path., 45: 41-58. Burke, D,, N. F. Skuse and A. K. Lethlean (1974) Sensory conduction of the sura~ nerve in poly~ neuropathy, J. Neurol. Neurosur~,. Psyehiat., 37: 647-652. Catterall, R. C. F. (1973) The diabetic foot, Proc. roy. Soc. Med., 66:201 21 I. Ellenberg, M., (1968) Diabetic neuropathic ulcer, J. M t Sinai Hosp., 35:585 594. Faris, I. B. (1975) Unpublished observations. Ferrier, T. M. (1967) Comparative study of arterial disease in amputated lower [imbs~from diabetics and non-diabetics (with special reference to feet arteries), Med. J. Aust., 1 : 5-1 I.

223
Gilliatt, R. W. and T. A. Sears (1958) Sensory nerve action potentials in patients with peripheral nerve lesions, J. Neurol. Neurosurg. Psychiat., 21 : 109-118. Kelly P. J. and M. B. Coventry (1958) Neurotrophic ulcers of the feet, J. Amer. reed. Ass., 168: 388 393. Moorhouse, J. A., S. A. Carter and J. Doupe (1966) Vascular responses in diabetic peripheral neuropathy, Brit. reed. J., 1 : 883-888. Oakley W., R. C. F. Catterall and M. M. Martin (1956) Aetiology and management of lesions of the feet in diabetics, Br#. reed. J., 2: 953-957. Ozeran R. S., G. R. Wagner, T. R. Reimer and R. A. Hill (1970) Neuropathy of the sympathetic nervous system associated with diabetes mel[itus, Surgery, 68: 953-958. Pirart, J. (1965) Diabetic neuropathy - - A metabolic or a vascular disease? Diabetes, 14: 1-9. Stokes, I., I. B. Faris and W. C. Hutton (1975) The neuropathic ulcer and loads on the foot in diabetic patients, Acta orthop, scand., 46:839 847. Thomas P. K., T. A. Sears and R. W. Gilliatt (1959) The range of conduction velocity in normal fibers to the small muscles of the hand and foot, J. Neurol. Neurosurg. Psychiat., 22: 175-181.