8/13/2019 Ncs in 6 Weeks Bsl Control

1/4

Changes in Nerve Conduction VelocityAfter Six Weeks of GlucoregulationWith Portable Insulin Infusion PumpsANG EL PIETRI, ALBERT L EHLE, AND PHILIP RASKIN

SUMMARYNear normal glucoregulation was maintained in 10 pa-tients with insulin-dependent type I) diabe tes m ellitusfor 6 wk with preprogrammed continuous subcutane-ous insulin infusion using a portable battery-poweredinfusion pump CSII). This form of therapy resulted ina statistically significant increase in motor nerve con-duction velocity in the median and peroneal nervescompared with baseline values. There was no signifi-cant change in the motor nerve conduction velocity inthe ulnar nerve or in the sensory nerve conductionstudies. No changes occurred in five additional pa-tients studied in similar fashion while on a conven-tional insulin regimen. These results suggest that theprevention of sustained hyperglycemia with CSII couldtheoretically result in the prevention of diabetic neu-ropathy. However, only long-term studies of CSII willprovide the information necessary to determine theclinical relevance of the findings. DIABETES29:668-671,August 1980.

T he role of hyperglycem ia in the pathoge nesis of di-abetic microvascular disease has not yet beenproved.1 Since most diabetic patients receivingconventional, intermittent subcutaneous insulin in-jections are hyperglycemic for a major portion of eachday,

2-3it is difficult to define the role of hyperg lycem ia in thedevelopment of diabetic complications.The recent developme nt of portable op en- loop insulininfusion pumps has made possible the maintenance of nor-mal or near-normal 24-h plasma glucose profiles,37 allow-ing correction of certain of the metabolic and endocrine ab-normalities found in diabetics receiving insulin byconventional methods.710 There is as yet no evidence that

From the University of Texas Health Science Center at Dallas, SouthwesternMedica l School, and the Dallas Veterans Administration Med ical Center, Dal-las, Texas.Address reprint requests to Philip Raskin, M.D., University of Texas HealthScience Center at Dallas, 5323 Harry Hines Blvd., Dallas, Texas 75235.Received for publication 23 May 1980.

this improved level of glycem ic con trol has any effect on theprogression of microvascular diabetic complications.In this communication we report improvement in motornerve conduction velocity in type I diabetics after 6 wk ofcontinuous subcutaneous insulin infusion (CSII) deliveredwith a portable insulin pump. No changes occurred in dia-betic patients treated with conventional insulin deliverymethods over a similar period of time.METHODSTen patients with typical type I (insulin-depe nden t) diabetesmellitus of 5-yr duration or longer (mean age, 15.8 2.5 yr)were stud ied. Four of the patients had clinica lly evident d ia-betic neuropathy, two had symptoms typical of the carpaltunnel syndrome, and the other two had mild polyneurop-athy with painful paresthesia of the feet.

All patients were admitted to the General Clinical Re-search Unit of Parkland Memorial Hospital and placed on ametabolic diet consisting of 45% carbohydrate, 35% fat,and 20% p rotein. The meal plan and CSII protocol has beenpreviously described.7With each patient receiving his or her usual daily dose ofsubcutaneous insulin, the initial baseline studies were per-formed. These studies included a 24-h glucose profile,measurement of glycosylated hemoglobin, and the nerveconduction tests. No attempt was made to achieve optimalcontrol on conventional treatment. The Auto-Syringe pumpmode l AS-2C (Auto-Syringe, Inc., Hooksett, New Ham pshire)was used in all patients but one; that patient was treatedwith the Mill-Hill infuser (Muirhead Ltd., Beckenham, Kent,England).

All patients were ho spitalized for at least the first 2 wk ofCSII therapy, after which they were sent home and followedas outpatients. While at home they were required to adhereto their diabetic d iet and to monitor their capillary blood glu -cose levels several times each day. Capillary b lood was ob-tained by finger stick and the glucose level measured on theAmes Eyetone or Dextrometef.* All managed to continue* A generous gift of the Ames Division, Miles Laboratory, Elkhart, Indiana.

668 DIABETES, VOL. 29, AUGUST 1980

8/13/2019 Ncs in 6 Weeks Bsl Control

2/4

ANGEL PIETRI, ALBERT L. EHLE, AND PHILIP RASKINwith their usual da ily activities, returning to work or to school.Most reported a marked increase in their feeling of wel l-being after institution of CSII. They were readmitted to theClinical Research Unit periodically for reevaluation, duringwhich time another profile was done. All studies were re-peated after 2 and 6 wk of treatment.

Blood samples for the 24-h glucose profile were obtainedat hourly intervals from 0700 to 2300 and then at 2-h inter-vals from 2300 to 0700 through an indw elling 19-gauge but-terfly needle placed in a large forearm vein. The glucosewas measured immediately on a Beckman glucose ana-lyzer. Total glycosy lated he mog lobin was determined by ionexchange chromatography using a prepacked microcolumn(Quik-Column, Helena Laboratories, Beaumont, Texas).An additional 5 type I diabetic patients (13 2 yr dura-tion) who served as controls were also studied. Each under-went the nerve conduction studies on three occasions(base line, 2 wk, and 6 wk) while on their usual daily dose ofintermittent subcutaneous insulin. Before the nerve conduc-tion studies at the baseline and 6-wk time points, a bloodsample was drawn for measurement of total glycosylatedhemoglobin.

The motor nerve conduction velocities were measured inthe right median and ulnar nerve forearm segments and inthe peroneal nerve in the right lower extremity. In the upperextremity the distal stimulating electrodes were located 1.5cm proximal to the palmar crease with an electrode spacingof 2 cm. For the median nerve a proximal electrode wasplac ed in the antecu bital fossa, and for the ulnar nerve prox-imal stimulation was performed app roximately 2 cm distal tothe elbow. At the time of initial rec ording, the distance of theproximal electrode from the distal stimulating site was re-corded and this distance was used subsequently to deter-mine the proximal site of stimulation. Latencies were deter-mined to the onset of the surface-recorded muscle actionpotential.The peroneal cond uction velocity was performed in a sim-ilar manner. Proximal stimulation was just distal to thecourse of the nerve around the head of the fibula. Distalstimulation was performed at the ankle and the distance re-corded for subsequent reproducibility. Normal motor nerveconduction velocity is greater than 40 m/s for the peronealnerve and greater than 50 m/s for the ulnar and mediannerves.TABLE 1Mean 24-h plasma glucose and g lycosylated hemoglobin levelsin type I diabetics after either 6 wk of continuous subcutaneousinsulin infusion or conventional therapy

CSIIBaseline2 wk6 w kConventionaltherapyBaseline6 w k* Mean SEM.f P < 0.01 by paired

Plasmaglucose(mg/dl)

214 17*107 9t103 8f

analysis.

Glycosylatedhemoglobin(%)

10.4 0.9f7.3 1t6.0 0.4|

10.1 0.99.5 0.9

The sensory conduction study was performed using theright median nerve. Stimulating ring electrodes were placedat the distal interphalangeal crease of the index finger. Therecording electrode was placed over the median nerve atthe wrist with the 2-cm spacing and the distal electrode lo-cated 1.5 cm from the palmar crease. An averager was usedto average 10 responses, and the latency to the negativepeak of this sensory-action potential and the peak-to-peakamplitude of the deflection was recorded for all subjects.Skin temperature was measured on the back of the righthand . All measurem ents were within 1C for the stud ieson each subject.The Student's test for paired groups was used for com-parison within groups. The test for nonpaired groups andchi-square a nalysis was used for comp arison between treat-ment groups.

RESULTSTable shows the mean 24-h glucose level and total glyco-sylated hemoglobin values before and after 2 and 6 wk ofCSII treatment. The 24-h glucose profile was normal after 2wk of CSII treatment, and after 6 wk of treatment the glycosy-lated hemoglobin level had fallen into the normal range,demonstrating the effectiveness of this form of treatment inachieving long-term normal or near-normal levels of glyce-mia. Also shown in this table are the glycosylate d hemo glo-bin values for those patients studied over a 6 wk period onconventional insulin treatment. Total glycosylated hemoglo-bin levels were elevated initially and remained high for theperiod of observation.

Table 2 shows the results of the nerve conduction tests.For those diabe tics treated with CSII,the median nerve con-duction velocity averaged 50.7 1.2 m/s before the initia-tion of CSII and was un cha nge d at 51.4 0.8 m/s after 2 wkof therapy. However, after 6 wk ofCSII,there was a statisti-cally significant increase (P < 0.05) to 53.5 0.8 m/s. Asimilar significant increase in the motor nerve conductionvelocity was seen in the peroneal nerve, which increasedfrom 39.5 1.6 to 41.7 2 m/s at 2 wk (P < 0.05) and to44.4 1.8 m/s after 6 wk (P < 0.01) ofCSII.Of the 30 indi-vidual motor nerve conduction tests performed after CSIItherapy, 23 improved while conduction slowed in seveninstances (Table 3). Motor nerve conduction velocity in-creased in all but one case in which the initial velocity wasabnormal. There were no significant changes in the motornerve conduc tion velocity in the ulnar nerve or in the m ediannerve sensory conduction studies. In those diabeticstreated by conventional means, there were no statisticallysignificant differences in the baseline values from thosepatients treated withCSII,and no significant changes in thenerve conduction studies occurred on repeated testing. Incontrast to those patients treated w ithCSII,motor nerve con-duction velocity tended to decrease with time, although thisdecrease was not statistically significant. Of the 15 individ-ual motor conduction velocities performed, nerve conduc-tion velocity diminished or was unchanged in 10 cases andimproved in only 5 (Table 3). This is significantly differentfrom the CSII-treated group (P < 0.01 by chi-square anal-ysis).There was no correlation between the duration of dia-betes and the nerve conduction studies.

All 4 patients with clinica l neuropathy treated w ith CSII re-ported a marked improvement in their symptoms within 6

DIABETES, VOL. 29, AUGUST 1980 669

8/13/2019 Ncs in 6 Weeks Bsl Control

3/4

NERVE CONDUCTION VELOCITY AFTER INSULIN PUMP GLUCOREGULATIONTABLE 2Nerve conduction tests in type I diabetics after either 6 wk of continuous subcutaneous insulin infusion or conventional therapy

CSIIBaseline2 wk6 w kConventionaltherapyBaseline2 wk6 w k* Mean SEM.t P < 0.05 by paired P < 0.01 by paired

Motor nerveUlnar

52.9 1.6*53.9 1.354.2 1.3

55.8 2.251.0 1.453.4 2.3

analysis.analysis.

conduction velocityMedian

50.7 1.251.4 0.853.5 0.8f

53.0 0.951.7 1.152.0 1.5

(m/s)Peroneal

39.5 1.641.7 2 .0 f44.1 1.8$

41.6 2.736.1 5.839.6 3.2

MedianLatency (ms)

3.5 0.33.5 0.23.4 0.2

3.4 0.13.5 0.23.4 0.2

nerve sensoryAmplitude (ptV)

7.6 1.76.7 1.17.5 1.6

9.0 2.07.2 2.17.6 2.0

wk. However, there were no striking differences betweenchanges in the motor nerve conduction velocities in thissubset of patients as compared with the groups without,symptoms.DISCUSSIONThis study clearly shows that the improved glucoregulationobtained with CSII has an effect on motor nerve conductionvelocity. Both the median and peroneal nerves showed sta-tistically significant increases in conduction velocities afteronly 6 wk of treatment with CSII as compare d w ith values ob-tained in these patients before institution ofCSII.There wasalso a small improvement in the peroneal nerve at 2 wk. Weshowed no effect on conduction velocities in the ulnar nerveor in sensory function. In contrast, in patients studied in asimilar fashion while on a conventional insulin regimen, re-sulting in persistent elevation of total glycosylated hemo glo-bin levels, motor nerve conduction velocities tended to de-crease. However, these changes were not statisticallysignificant.These data fit well with those of Greene et al.11 wh oshowed that in rats with streptozotocin-induced diabetes,significant decreases in motor nerve conduction velocitiesoccur within several weeks of the induction of diabetes.These abnormalities in motor nerve conduction velocitycould be corrected with insulin treatment. However, exqui-site diabetic control was required before a complete returnto normal conduction velocity occurred.These data, although ex citing, must be viewe d ca utiously.While the methods used to study nerve conduction velocityare well established, they provide only a limited evaluationTABLE 3Comparison of changes in motor nerve conduction velocity intype I diabetics after either 6 wk of continuous subcutaneousinsulin infusion or conventional therapy

Decreased orImproved no changeCSIIConventionaltherapy

23*5

7

* P < 0.01 , CSII versus conventional therapy by chi-squareanalysis.

of the overall function of periphera l nerves. Motor nerve con-duction studies provide only information about the fastestconducting fiber population at the time of study. Sensoryconduction studies, while reflecting a different fiber popula-t ion, also examine only the function of large myelinatedaxons. While the amplitude of the potential can provide someindication of the number of fibers present, the low signal-to-noise ratio of this potential m akes detection of sm all chan gesdifficult. Other factors, such as the degree of dispersion ofconduction velocities, also affect amplitude.Although most of our patients had no clinical suggestionof diabetic neuropathy, all 4 who did noted a marked im-provement in their symptoms. One must ask what thesef ind-ings mean in relation to the long-term development of dia-betic neuropathy. The small changes we observed couldhave nothing to do with long-term co mp lications, but merelyreflect a reversible effect due to reduction in the plasma glu -cose levels. Graf et al. showed a correlation between thelevel of hyperglycemia and the slowing of motor nerve con-duction velocity in maturity-onset diabetics.12 The fact thatthe 24-h glucose profile was normal after 2 wk of CSII, yetthe conduction velocity in the median nerve was unchangedfrom baseline values and that the peroneal nerve showedonly a small change, suggests that the improvement inmotor nerve conduction velocity that we observed was notdue merely to the acute reduction in plasma glucose levelsbut was the result of a more long-term effect of improvedglucoregulation.It must be remembered that sustained hyperglycemia hasbeen shown to result in postsynthetic glyco sylation of manyproteins, including hemoglobin,13 albumin,14-15and the pro-teins of the lens.16Perhaps the proteins of nervous tissue arealso glycosylated, which could be one of the the pathogen-etic mechanisms for the development of the neurologiccomplications of diabetes. Prevention of sustained hyper-glycemia with CSII could, thus, theoretically result in theprevention of diabetic neuropathy. Only long-term con-trolled studies of CSII will provide the necessary informationto determine the clinical relevance of these findings.ACKNOWLEDGMENTSThe authors wish to thank Roger H. Unger, M.D., for his ex-pert advice and consultation; Drs. Jaime Davidson, RobertDail,and Aubrey Godfrey for referral of their patients; Mari-

670 DIABETES, VOL 29, AUGUST 1980

8/13/2019 Ncs in 6 Weeks Bsl Control

4/4

ANGEL PIETRI, ALBERT L. EHLE, AND PHILIP RASKINlyn Alford, R.N., for her invaluable help; David Brown, Mar-garet Cason, Alison Hanson, and Arthur Ojirika for theirtechnical assistance; Bette Newton for secretarial assis-tance; and the staff and nurses of the General Clinical Re-search Center.This study was supported in part by NIH Grants AM18179,AM2700, and 1-MO1-RR0063; VA Institutional ResearchSupport Grant 549-8000-1; NIH Contract NO1-AM62219;American Diabetes Association, North Texas Affil iate; EliLilly Co., Indianapolis, Indiana; and by William Wheeler,Portland, Oregon.This work was performed during the tenure of Dr. Raskinas a Clinical Investigator, Dallas Veterans AdministrationMedical Center, Dallas, Texas.REFERENCES1 Raskin, P.: Diabetic regu lation and its relationship to m icroangi-opathy. Metabolism 27:235-52, 1978.2 Raskin, P., and Unger, R. H.: Effect of insulin therapy on the profilesof plasma imm unoreactive glucagon in juvenile-type and adult-type diabetes.Diabetes 27:411-19, 1978.3 Service, F. J., Molnar, G. D., Rosevear, J. W., et al.: Mean amplitudeof glycemic excursions, a measure of diabetic instability. Diabetes 79:644-55, 1970.4 Pickup; J. C , Keen, H., Parson, J. A., and Alb erti, K. G. M. M.: Contin-uous subcutaneous insulin infusion: an approach to achieving normoglyce-mia. Br. Med. J. 7:204-07, 1978.5T amborlane, W. V., Sherwin, R. S., Genel, M., and F elig, P: Reductionto normal of plasma glucose in juvenile diabetes by subcutaneous adminis-

tration of insulin with a portable infusion pump. N. Engl. J. Med. 300:573-78,1979. 6 Irsigler, K., and Kritz, H.: Long-term continuous intravenous insulintherapy with a portable insulin dosage-regulating apparatus. Diabetes28:196-203, 1979.7 Raskin, P., Pietri, A., and Uhger, R. H.: Changes in glucagon levelsafter four to five weeks of glucoregulation by portable insulin infusion pumps.Diabetes 28:1033-35, 1979.8P ickup, J. C , Keen, H., Parson, J. A., Alberti, K. G. M. M., and Rowe,A. S.: Continuous subcutaneous insulin infusion: improved blood-glucoseand intermediary-metabolite control in diabetics. Lancet 7:1255-57, 1979.9Tamborlane, M. W., Sherwin, R. S., Genel, M., and Felig, P.: Restora-tion of normal lipid and amino acid metabolism in diabetic patients treatedwith a portable insulin-infusion pump. Lancet 7:1258-61, 1979.10 Tamborlane, M. W., Sherwin, R. S., Koivisto,.V., Hendler, R., Genel,M., and Felig, P.: Normalization of the growth hormone and catecholamineresponse to exercise in juvenile-onset diab etic subjects treated w ith a porta-ble insulin infusion pump. Diabetes 28:785-88, 1979.11Greene, D. A., De Jesus, P. V., and Winegrad, A. L: Effects of insulinand dietary myoinositol on impaired peripheral nerve conduction velocity inacute streptozotocin diabetes. J.Clin. Invest. 55:1326-36, 1975.12Graf R. J., Halter, J. B., Halar, E., and Porte, D., Jr.:.Nerve conduc-tion abnormalities in untreated maturity-onset diabetes: Relation to levels offasting plasma glucose and glycosylated hemoglobin. Ann. Intern. Med.90:298-303, 1978.13Bunn, F. H., Gabbay, K. H., and Gallop, P. M.: The glycosylation ofhemoglobin: relevance to diabetes mellitus. Science 200:21-27, 1978.14Day, J. F., Thorpe, S. R., and Baynes, J. W.: Nonezym atically glu co-sylated albumin. J. Biol.Chem. 254:595-97, 1979.15 Yue, D. K., Morris, K., McLennan, S., and Turtle, J. R.: Glycosylationof plasma protein and its relation to glycosylated hemoglobin in diabetes. Di-abetes 29:296-300, 1980.16 Stevens, V. J., Rouzer, C. A., Monnier, V. M., and Cerami, A.: Dia-betic cataract formation: potential role of glycosylation of lens crystallins. J.Biol.C hem. 254:595 -97, 1979.

DIABETES, VOL. 29, AUGUST 1980 67