RESEARCH REPORT

Dorsal sural nerve conduction study in vitamin B12

deficiency with megaloblastic anemia

BurhanTurgut1, NildaTurgut2, Seval Akpinar1, Kemal Balci2, Gu« lsu« m E. Pamuk1,EmreTekgu« ndu« z1, and Muzaffer Demir1

1Department of Medicine, Division of Hematology; and 2Department of Neurology, Trakya Medical Faculty,University of Trakya, Edirne, Turkey

Abstract Peripheral neuropathy is frequently observed in B12 deficiency. In spite ofthis, there is little knowledge about peripheral neuropathy in B12 deficiency because theseverity of clinical involvement of the central nervous system clearly outweighs signs andsymptoms due to peripheral nervous system involvement. We primarily investigatedperipheral neuropathy with dorsal sural conduction study, which is a new method fordetection of early peripheral neuropathy, in B12 deficiency with megaloblastic anemia.Conventional nerve conduction studies and tibial sensory-evoked potential (SEP) recordingwere also performed. Twenty-eight B12-deficient patients (15 male, 13 female, mean age65.8 years) with megaloblastic anemia and 18 age- and sex-matched controls wereincluded in the study. Although dorsal sural sensory nerve action potentials (SNAPs) werenot recorded in 15 (54%) of 28 patients, only 9 (32%) of them were found to have poly-neuropathy by conventional conduction studies. Furthermore, patients with dorsal suralSNAP had mean lower amplitude, mean longer latency, and slower velocity responsewhen compared with controls. Twenty patients (71%) were diagnosed as having myelop-athy by the combination of tibial SEP and neurological findings. Two patients whose dor-sal sural SNAPs were not recorded had normal tibial SEP responses; therefore, thesepatients were considered to have isolated peripheral neuropathy. As a result, we concludethat dorsal sural nerve conduction study is a reliable method for detection of early periph-eral neuropathy in B12 deficiency.

Key words: B12 deficiency, dorsal sural nerve, peripheral neuropathy, SEP

IntroductionVitamin B12 deficiency is a systemic disease that

often affects the nervous system. It may causeperipheral neuropathy and lesions in the posterior andlateral columns of the spinal cord and in the cerebrum(Toh et al., 1997). Demyelination with subsequentaxonal disruption and gliosis can affect all parts of

the central nervous system (Carmel, 2003). Peripheralnerves, however, tend to show axonal degenerationwithout demyelination (McCombe and McLeod,1984; Tomoda et al., 1988). Peripheral neuropathy,which is usually sensory, may be sensorimotor(Hemmer et al., 1998), and posterior column damageis often hard to differentiate clinically (Steiner et al.,1988; Healton et al., 1991).

In peripheral neuropathies, the most distal sen-sory fibers of the lower extremities are generally firstaffected, but these nerves cannot be evaluated bysural and superficial peroneal conduction studies, whichare used routinely for the diagnosis of polyneuropathy

Address correspondence to: Burhan Turgut, MD, Department ofHematology, Trakya Medical Faculty, University of Trakya, 22030Edirne, Turkey. Tel: þ90 532 433 36 76; Fax: þ90 284 235 76 52;E-mail: burhanturgut@trakya.edu.tr

Journal of the Peripheral Nervous System 11:247–252 (2006)

ª 2006 Peripheral Nerve Society 247 Blackwell Publishing

(Killian and Foreman, 2001). Dorsal sural nerve of thefoot is the most distal sensory nerve. It has beenreported that dorsal sural nerve conduction study isa sensitive marker of peripheral neuropathies thatmay show abnormalities before they are found inproximal sural nerve (Lee et al., 1992; Killian andForeman, 2001; Turgut et al., 2004).

In this study, we investigated peripheral neuropa-thy by dorsal sural nerve conduction studies in additionto conventional conduction studies in B12-deficientpatients with megaloblastic anemia. Also, becauseposterior column involvement is the most frequentlyreported and complicated neuropathy in B12 defi-ciency, tibial sensory-evoked potentials (SEPs) werestudied in all patients.

Materials and MethodsSubjects

The study was carried out between December2004 and October 2005. Twenty-eight consecutivepatients and 18 healthy age- and sex-matched con-trols were included in the study. Informed consentwas obtained from each subject. The patients met allthe following criteria: the presence of megaloblasticanemia with characteristic features such as macroova-locytosis and hypersegmentation and evidence ofcobalamin deficiency based on a low serum cobalaminlevel (,200 ng/l).

All the patients underwent detailed general andneurological examination. Routine blood chemistry,complete blood cell count, vitamin B12 analysis, andfolate serum concentration analysis were performedin all patients. Blood smears were examined bya hematologist. They also had no history of folatedeficiency and other spinal cord or peripheral nervediseases. Patients with carpal tunnel syndrome,radiculopathy, plexopathy, and mononeuropathy wereexcluded from the study. Diabetes mellitus, uremia,alcohol abuse, drug use, and other possible causesof polyneuropathy were excluded by blood chemistryanalysis and history of the patients.

Electrophysiological investigations

Conventional motor and sensory nerveconduction studies

In all patients and control subjects, left median,ulnar, common peroneal and tibial motor nerves andleft median, ulnar, sural sensory nerves were studiedwith a Medelec Synergy electromyography machine.Median and ulnar nerve compound muscle actionpotentials (CMAPs) were recorded from the abductorpollicis brevis and the abductor digiti minimi muscles,

and stimulation was delivered at the wrist and at theelbow. Tibial and common peroneal nerve CMAPswere recorded from the abductor hallucis and exten-sor digitorum brevis muscles, and stimulation wasdelivered at the ankle and at the poplitea for tibialnerve and at the fibula head for the peroneal nerve.Median and ulnar sensory nerve action potentials(SNAPs) were recorded with ring electrodes from digit2 and digit 5, and stimulation was applied at the wrist.Sural antidromic sensory nerve conduction studieswere performed by recording the SNAP posterior tothe lateral malleolus. Stimulation was carried out 14cm proximally in the midcalf (Oh, 1993).

Filter settings were 20–2,000 Hz bandpass forsensory nerve studies, and 2–10,000 Hz bandpass formotor nerve studies. An average was used for thesensory nerve recording, and the mean number ofaverages required to clearly define the potential was8. The averaging was repeated to confirm the repro-ducibility, if necessary. In patients whose sensorynerve response could not be obtained, the averagingwas repeated at least three to five times. Limb tem-perature was maintained between 31 and 34�C in allsubjects. Polyneuropathy was diagnosed on the basisof abnormal nerve conduction studies when abnor-malities were found in two or more nerves (Feldmanet al., 1994).

Dorsal sural nerve recordingSurface bar recording electrodes (Teca Corp.)

were used to test dorsal sural nerve SNAP. A record-ing electrode was placed over the lateral dorsal sur-face of the foot, with the distal electrode at the originof digits 4 and 5 and the proximal active electrode3 cm from the distal electrode. The stimulation sitewas posterior to the lateral malleolus, with the cathodeplaced 14 cm proximal from the recording electrode.A ground electrode was placed on the dorsum of thefoot equidistant between the recording and the stimu-lating electrodes. Low filters were set at 20 Hz andhigh filters at 2 kHz. Sweep speed was 1 ms per divi-sion, sensitivity was adjusted between 5 and 20 mVper division, and stimulus duration was 0.2 ms ata stimulus rate of 0.5 Hz. At least eight potentialswere averaged to record an appropriate response,and the averaging was repeated if necessary. Distallatency was measured from stimulus onset to thenegative peak of the SNAP. SNAP amplitudes weremeasured from baseline to peak (Killian and Foreman,2001). A representative figure demonstrating sural anddorsal sural nerves action potentials is shown in Fig. 1.

Tibial somatosensory-evoked potential recordingThe active electrode was placed at the Cz0 (2 cm

posterior to Cz). The reference electrode was placed

Turgut et al. Journal of the Peripheral Nervous System 11:247–252 (2006)

248

at Fpz0 (2 cm behind Fp). The ground electrode wasplaced as close to the first pair of recording electrodes.With the patient in the supine position, surface stimu-lating electrodes were placed behind the medial mal-leolus. The exact point of stimulation was identifiedby measuring 1 cm behind and 1 cm below the navic-ular tubercle (medial side of the foot). This point ap-proximates the middle of the belly of the abductorhallucis muscle. Next, measure 8 cm proximal to thispoint, following the course of the tibial nerve 1 cmposterior to the medial malleolus. The electrode atthis point was the anode. The cathode was placed4 cm proximal to the anode. Analysis time was 50–100ms, gain was 10 mV, frequency bandwith was 0–2,000Hz. The N35 and P40 latencies were evaluated (Delisaet al., 1994).

Statistical analysis

Statistical analysis was performed using SPSSsoftware (SPSS, Inc.). Mann–Whitney U test was usedfor comparing groups. Relationships between cate-gorical variables were compared using the chi-squaretest. Coefficients of correlation (r ) were calculatedusing the Spearman’s rank test. The level of signifi-cance in all statistical analysis was set at p , 0.05.

ResultsClinical evaluations

Clinical and laboratory data of the patients areshown in Table 1. All the patients had a B12 serumlevel less than 200 pg/ml and obvious megaloblasticanemia. We did not perform a schilling test or intrinsicfactor antibody test for diagnosis of pernicious anemia,but eight patients underwent endoscopy and allhad type A atrophic gastritis. Moreover, gastrectomy,malabsorption syndromes, vegetarianism, and otherrare causes of B12 deficiency were absent in all thepatients; therefore, the diagnosis of pernicious anemiain our patients had a high probability.

Electrophysiological findings

The results of conventional nerve conductionstudies of the patients and control subjects areshown in Table 2. Conventional nerve conductionwas impaired in 9 (32%) of 28 patients; 4 patients hadaxonal sensory polyneuropathy, 4 had axonal sensori-motor polyneuropathy, and 1 had mixed (axonal anddemyelinating) sensorimotor polyneuropathy.

Dorsal sural nerve conduction studiesThe patients and control subjects underwent dor-

sal sural nerve conduction studies. Dorsal sural nervewas recordable bilaterally in all the control subjects.Dorsal sural nerve SNAPs were bilaterally absent in15 (54%) of 28 patients, and 9 of them had impairedconventional conduction abnormalities simultaneously.In the other 13 (46%) patients, the mean distal latencyof the dorsal sural nerve response was longer (p ¼0.003), mean amplitude was lower (p ¼ 0.008), andmean velocity response was slower (p ¼ 0.001) thanthose of 18 healthy controls (Table 3). However, the

Table 1. Clinical and laboratory characteristics of the patients.

Number of the patients (N) 28Age (years), median (range) 66 (39^91)Gender (N), F/M 15/13Neurological findings, N (%)Abnormal vibration or proprioception 15 (54)Abnormal pinprick sensation 6 (21)Gait disturbance 9 (32)Loss of deep tendon reflexes 3 (10)Pyramidal tract finding 1 (3,6)Mental impairment 1 (3,6)Hemoglobulin (g/dl), mean� SD 7.57� 2.44WBC (ml), mean� SD 4,515� 1,825Platelet (ml), mean� SD 123,892� 79,017MCV (fl), mean� SD 110.1�23.9Serum vitamin B12 (pg/ml), mean� SD 109.6� 21.5Serum folate (ng/ml), mean� SD 7.98� 4.11LDH (U/l), mean� SD 2,121�2,295

LDH, lactate dehydrogenase; MCV, mean cell volume; WBC, whitebloodcell.

Figure 1. The top traces show the sensory-evoked poten-tial (SEP) from sural nerve and the bottom traces show theSEP from dorsal sural nerve of a control subject.

Turgut et al. Journal of the Peripheral Nervous System 11:247–252 (2006)

249

action potentials of sural nerve did not differ betweenthe same patients and the control subjects (Table 3).

Tibial SEPsTibial SEPs were abnormal in 20 (71%) of 28

patients with delayed (14 patients) or absent (6 patients)responses. Nine patients were found to have neuro-pathy by conventional conduction studies, and theyhad impaired tibial SEP simultaneously. Three patientswho had no SEP response had severe tibial nerve con-duction impairment, so impaired SEPs may be associ-ated with severe motor neuropathy, but they had theclinical finding of myelopathy. Two patients whosedorsal sural SNAPs were absent had normal SEPresponses and seven patients with abnormal SEPresponses had dorsal sural SNAPs. Tibial SEPs resultsare shown in Table 4.

Six patients had stocking/glove type sensory loss,and all of them had sensory or sensorimotor polyneuro-pathy. Three patients with sensorimotor polyneuro-pathy had absent deep tendon reflexes. However, theother three patients with polyneuropathy had no sen-sory loss and abnormality of deep tendon reflexes.Of 20 patients with abnormal SEP responses, 15had abnormal vibration or proprioception, and nine

patients had unsteady walking with positive Romberg’stest, and SEP was abnormal (6 absent, 3 delayed) inall. Spasticity was evident in the lower limbs of apatient with extensor plantar response. No motor deficitwas found in any of these patients.

Correlation analysis

There was no difference between female andmale patients for all the electrophysiological studies.There was no correlation between all the electrophys-iological findings and serum vitamin B12 and folatelevels, hemoglobin level, mean cell volume, leuko-cyte and platelet counts, and lactate dehydrogenaselevel.

DiscussionThere is a wide range of neurological disturbances

in patients with B12 deficiency including myelopathy,peripheral neuropathy, altered mental status, andoptic neuropathy (Healton et al., 1991; Pandey et al.,2004). It is difficult to distinguish findings due to theinvolvement of the central nervous system fromthose due to peripheral nervous system dysfunction(Roos, 1978; Healton et al., 1991; Shevell and Rosen-blatt, 1992), and there are no commonly acceptedinclusion criteria in studies that investigated neurologi-cal disturbances because there are no exact criteriafor the diagnosis of B12 deficiency. For this reason,the frequency of peripheral neuropathy in vitamin B12

deficiency sharply differs among studies. This studyincluded B12-deficient patients with overt megaloblas-tic anemia; the cause of B12 deficiency was perni-cious anemia with high probability. We primarilyinvestigated peripheral neuropathy in these patientsby dorsal sural nerve conduction studies, in addition toconventional electrophysiological studies. Dorsal suralSNAPs were not recorded in 54% of the patients.Only 32% of patients were diagnosed as having

Table 3. The results of sural and dorsal sural nerve actionpotentials of patients whose dorsal nerve was recordable andthe comparison of these results with controls.*

NervesPatients(n¼13)

Controls(n¼18) p

Dorsal suralAmplitude (mV) 4.78� 1.62 6.99� 2.24 0.008Velocity (m/s) 34.5� 6.22 46.8� 11.2 0.001Latency (ms) 3.89� 0.70 3.09� 0.59 0.003

SuralAmplitude (mV) 14.1�5.68 12.4� 4.39 0.70Velocity (m/s) 50.3� 8.33 55.9� 12.7 0.17Latency (ms) 2.86� 0.77 2.68� 0.63 0.37

*The values are expressed as mean � SD.

Table 2. The results of action potentials in patients with B12 deficiency who had normal conventional electrophysiological studiesand control subjects.*

B12-deficient patients with normal conventionalconduction studies (n¼19) Control subjects (n¼18)

Nerves Latency (ms) Amplitude (mV) Velocity (m/s) Latency (ms) Amplitude (mV) Velocity (m/s)

Median (s) 3.31�0.40 20.2� 9.50 54.1�6.08 2.97� 0.24 23.1�7.71 53.5� 4.27Ulnar (s) 2.82� 0.31 18.8� 8.89 53.5� 4.27 2.46� 0.22 22.8� 7.92 63.9� 5.85Tibial 4.22� 0.83 9.31�3.97 44.9� 3.01 4.14� 0.63 7.63� 2.45 48.8� 3.98Median (m) 3.52� 0.43 7.86� 2.64 53.7� 3.61 3.32� 0.37 7.53� 2.14 59.1�4.54Ulnar (m) 2.60� 0.33 9.78� 2.04 66.2� 5.37 2.27� 0.22 9.38� 2.28 70.6� 7.16C. Peroneal 4.37� 0.67 4.02� 1.67 48.4� 4.96 3.68� 0.76 4.28� 1.92 54.1�5.71Sural 2.89� 0.81 13.3� 5.71 49.2� 7.66 2.68� 0.63 12.4� 4.39 55.9� 12.7

*The values are expressed as mean � SD.

Turgut et al. Journal of the Peripheral Nervous System 11:247–252 (2006)

250

polyneuropathy by conventional conduction studies.Furthermore, when compared with normal in-dividuals, patients with recordable dorsal sural SNAPshad lower mean amplitude, slower mean velocity,and longer mean latency. It was reported that therewas high variability in dorsal sural SNAPs in normal in-dividuals, and standard cutoff values for dorsal suralSNAPs for adults were not designated in the literature(Killian and Foreman, 2001). For this reason, we didnot use the quantitative values of the action poten-tials of dorsal sural nerve to diagnose peripheralneuropathy.

Dorsal sural nerve conduction study is describedas a new method providing an opportunity for theevaluation of most distal nerves of lower extremitiesthat are usually first affected in peripheral neuropathy(Lee et al., 1992; Killian and Foreman, 2001). In dia-betic children, this method is reliable to show sub-clinical neuropathy (Turgut et al., 2004). Our resultsdemonstrated that dorsal sural SNAP can aid to diag-nose the peripheral neuropathy component of neuro-logical disturbances of B12 deficiency.

Myelopathic presentation has been reported tobe more common than peripheral neuropathy in B12

deficiency (Misra et al., 2003). Furthermore, it is con-troversial whether polyneuropathy may be present inthe absence of myelopathy (Saperstein et al., 2003).In concordance with the literature, most of our pa-tients presented with myelopathy. With clinical evalu-ations and SEP studies, we found that 71% of ourpatients had posterior tract damage. Myelopathypredominantly affects the posterior columns followedby the anterolateral and anterior tracts (Hemmeret al., 1998). We did not measure motor-evoked po-tentials of our patients, but the presence of signs ofcentral motor pathway damage in only one patient onneurological examination made us think that damageto the pyramidal tract was not frequent in B12 defi-ciency as previously reported (Hemmer et al., 1998).Two patients with peripheral neuropathy diagnosedwith dorsal sural nerve conduction study did not havemyelopathy; this observation demonstrated that iso-lated neuropathy may be present in B12-deficientpatients. Like our study, Misra et al. (2003) reportedthat 1 of 17 vegetarian patients with B12 deficiency hadisolated neuropathy, and Puri et al. (2005) also reported

that 4 patients had isolated neuropathy on the basisof clinicoelectrophysiological profiles of 40 patients withB12 deficiency.

There are a few studies proposing a correlationbetween clinicolaboratory findings and neurologicalfindings in B12 deficiency. It has been reported thatthe severity of neurological abnormalities increaseswith increasing hematocrit levels (Healton et al.,1991; Savage and Lindenbaum, 1995). In anotherstudy, serum vitamin B12 level was found to correlatewell with sural SNAP and prolonged tibial SEP res-ponses (Puri et al., 2005). In spite of this, we did notfind any correlation between electrophysiological find-ings and clinicolaboratory findings. The studies thatreported an inverse correlation between hematocritand the severity of neurologic findings includedpatients without anemia or mild anemia. Even in thestudy of Healton et al. (1991), this correlation wasreported only in patients with normal hematocrit level.Discordance between these studies and our studycan be explained by the fact that we included onlypatients with overt anemia. Exact reasons for theinverse correlation between hematocrit and theneurologic involvement are not known. Recently,Carmel et al. (2003) showed significant differences inhomocysteine metabolism among patients with B12

deficiency. They found that patients with B12 defi-ciency who had neurologic defects had higher folate,S-adenosylmethionine, cysteine, and cysteinylglycinelevels than neurologically unaffected patients and thatlow S-adenosylmethionine and glutathione levelswere independent predictors of anemia.

As a result, with dorsal sural SNAP, most of theB12-deficient patients with megaloblastic anemia canbe detected to have peripheral neuropathy. On theother hand, in concordance with previous studies,dorsal tract involvement is more common thanneuropathy in B12 deficiency.

ReferencesCarmel R (2003). Megaloblastic Anemias: Disorders of

Impaired DNA synthesis. In: Wintrobe’s Clinical Hematol-ogy, 11th Edn. Greer JP, Foerster J, Lukens JN, Rodgers

Table 4. The results of tibial SEP studies.*

P40 N35 N50No

responseDelayedresponse

SEP right (n¼ 28) 46.2� 7.6 40.1�7.27 54.4� 6.32 6 patients 14 patientsSEP left (n¼ 28) 47.1�8.09 40.0� 7.25 58.7�11.4 6 patients 14 patients

SEP, sensory-evoked potential.*The values are expressed as mean � SD.

Turgut et al. Journal of the Peripheral Nervous System 11:247–252 (2006)

251

GM, Paraskevas F, Glader B (Eds). Lippincott Williams &Wilkins, Philadelphia, Vol. 1, pp 1367–1395.

Carmel R, Melnyk S, James SJ (2003). Cobalamin deficiencywith and without neurologic abnormalities: differences in

homocysteine and methionine metabolism. Blood 101:3302–3308.

Delisa JA, Lee HJ, Baran EM, Lai K, Spielholz N (1994). Man-ual of Nerve Conduction Velocity and Clinical Neurophysiol-

ogy, 3rd Edn. Raven Press, New York, pp 240–250.

Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N,Greene DA (1994). A practical two-step quantitative clinical

and electrophysiological assessment for the diagnosis andstaging of diabetic neuropathy. Diabetes Care 17:1281–1289.

Healton EB, Savage DG, Brust JC, Garrett TJ, Lindenbaum J(1991). Neurologic aspects of cobalamin deficiency. Medi-

cine 70:229–245.Hemmer B, Glocker FX, Schumacher M, Deuschl G, Lucking CH

(1998). Subacute combined degeneration: clinical, electro-physiological, and magnetic resonance imaging findings.

J Neurol Neurosurg Psychiatr 65:822–827.Killian JM, Foreman PJ (2001). Clinical utility of dorsal sural

nerve conduction studies. Muscle Nerve 24:817–820.Lee HJ, Bach HJ, DeLisa JA (1992). Lateral dorsal cutaneous

branch of the sural nerve. Standardization in nerve conduc-tion study. Am J Phys Med Rehabil 71:318–320.

McCombe PA, McLeod JG (1984). The peripheral neuropathyof vitamin B12 deficiency. J Neurol Sci 66:117–126.

Misra UK, Kalita J, Das A (2003). Vitamin B12 deficiency neu-rological syndromes: a clinical, MRI and electrodiagnostic

study. Electromyogr Clin Neurophysiol 43:57–64.Oh SJ (1993). Clinical Electromyography: Nerve Conduction

Studies, 2nd Edn. Williams & Wilkins, Baltimore, pp 3–140.

Pandey S, Kalita J, Misra UK (2004). A sequential study of visualevoked potential in patients with vitamin B12 deficiency neu-

rological syndrome. Clin Neurophysiol 115:914–918.Puri V, Chaudhry N, Goel S, Gulati P, Nehru R, Chowdhury D

(2005). Vitamin B12 deficiency: a clinical and electrophysio-logical profile. Electromyogr Clin Neurophysiol 45:273–284.

Roos D (1978). Neurological complications in patients withimpaired vitamin B12 absorption following partial gastrec-

tomy. Acta Neurol Scand, Suppl 69:1–77.

Saperstein DS, Wolfe GI, Gronseth GS, Nations SP, HerbelinLL, Bryan WW, Barohn RJ (2003). Challenges in the identi-

fication of cobalamin-deficiency polyneuropathy. Arch Neu-rol 60:1296–1301.

Savage DG, Lindenbaum J (1995). Neurological complicationsof acquired cobalamin deficiency: clinical aspects. Baillieres

Clin Haematol 8:657–678.Shevell MI, Rosenblatt DS (1992). The neurology of cobalamin.

Can J Neurol Sci 19:472–486.Steiner I, Kidron D, Soffer D, Wirguin I, Abramsky O

(1988). Sensory peripheral neuropathy of vitamin B12 defi-ciency: a primary demyelinating disease? J Neurol 235:

163–164.Toh BH, van Driel IR, Gleeson PA (1997). Pernicious anemia.

N Engl J Med 337:1441–1448.Tomoda H, Shibasaki H, Hirata I, Oda K (1988). Central vs

peripheral nerve conduction. Before and after treatmentof subacute combined degeneration. Arch Neurol 45:

526–529.Turgut N, Karasalihoglu S, Kucukugurluoglu Y, Balci K, Ekuklu

G, Tutunculer F (2004). Clinical utility of dorsal sural nerveconduction studies in healthy and diabetic children. Clin

Neurophysiol 115:1452–1456.

Turgut et al. Journal of the Peripheral Nervous System 11:247–252 (2006)

252