Abnormal Carbohydrate Metabolism in

Chronic Renal Failure

THE POTENTIAL ROLEOF ACCELERATEDGLUCOSEPRODUCTION,

INCREASEDGLUCONEOGENESIS,ANDIMPAIRED GLUCOSEDISPOSAL

SHELDONRUBENFELDand ALAN J. GARBER,Division of Endocrinology andMetabolism, Departments of Medicine and Cell Biology, Baylor College ofMedicine, Houston, Texas 77030

A B S T RA C T To delineate the potential role of dis-ordered glucose and glucose-precursor kinetics in theabnormal carbohydrate metabolism of chronic renalfailure, alanine and glucose production and utilizationand gluconeogenesis from alanine were studied in pa-tients with chronic compensated renal insufficiencyand in normal volunteers. With simultaneous primedinjection-continuous infusions of radiolabeled alanineand glucose, rates of metabolite turnover and pre-cursor-product interrelationships were calculated fromthe plateau portion of the appropriate specific activitycurves. All subjects were studied in the postabsorp-tion state. In 13 patients with chronic renal failure(creatinine = 10.7±1.2 mg/100 ml; mean+SEM),glucose turnover was found to be 1,035+99.3 ,umol/min. This rate was increased 56% (P = 0.003) overthat observed in control subjects (664±33.5 gmol/min).Alanine turnover was 474±96.0 gmol/min in azotemicpatients. This rate was 191% greater (P = 0.007) thanthe rate determined in control subjects (163±19.4gmol/min). Gluiconeogenesis from alanine and the per-cent of glucose production contributed by gluconeo-genesis from alanine were increased in patients withchronic renal failure (192% and 169%, respectively)as compared to controls (P < 0.05 for each). Alanineutilization for gluconeogenesis was increased from 40.2±3.86 gmol/min in control subjects to 143+39.0,umol/min in azotemic patients (P < 0.05). The percentof alanine utilization accounted for by gluconeogenesiswas not altered in chronic renal insufficiency. In non-diabetic azotemic subjects, mean fasting glucose and

Dr. Garber is an investigator at the Howard Hughes Medi-cal Institute.

Received for publication 7 October 1.977 and in revisedform 8 Februartj 1978.

immunoreactive insulin levels were increased 24.3%(P = 0.005) and 130% (P = 0.046), respectively.

These results in patients with chronic renal failuredemonstrate (a) increased glucose production andutilization, (b) increased gluconeogenesis from ala-nine, (c) increased alanine prodtuction and uitilization,and (d) a relative impairment to glucose disposal. Weconclude that chronic azotemia is characterized by in-creased rates of glucose and glucose precursor flux andby a relative impairment to glucose disposal. Thesefindings may suggest an underlying hepatic and periph-eral insensitivity to the metabolic action of insulin inpatients with chronic renal insufficiency.

INTRODUCTION

Disordered carbohydrate metabolism is widely recog-nized in patients with chronic renal failure (1).These abnormalities most commonly include previ-ously unrecognized carbohydrate intolerance ingreater than 50% of azotemic individuals (2, 3) andeven overt fasting hyperglycemia (3, 4). However, thedevelopment of spontaneous hypoglycemia has alsobeen reported (5, 6). On the other hand, an ameliora-tion of the diabetic state with decreasing insulinrequirements is a well-known clinical concomitant ofadvancing renal failure in patients with previouslyestablished diabetes mellitus (7).

Recent work in patients with diabetes mellitus hassuggested that an inappropriate and excessive rate ofglucose production may function as an important deter-minant of the degree of hyperglycemia (8). Addi-tionally, increased gluconeogenesis from a variety ofprecursors such as alanine, the principal gluconeo-genic amino acid, has been demonstrated in patientswith diabetes mellitus (9, 10). Elevated rates of alanine

J. Clin. Invest. © The American Society for Clinical Investigation, Inc., 0021-9738/78/0701-20 $1.0020

formation and release may have been described inskeletal muscle of chronically uremic rats (11). In-creased hepatic gluconeogenesis has been observed inthe acutely uremic rat (12) and in rat liver slices in-euibated in serum obtained from uremic patients (13).

In light of these data, investigations of glucose andglucose-precursor kinetics were performed in patientswith chronic renal insufficiency. Turnover rates ofalanine and glucose were determined in chronicallyazotemic patients and in normal control subjects withsimultaneous primed injection-continuous infusions of[U-14C]alanine and [2-3H]Hglucose. Alanine-glucose,precursor-product interrelationships were assessedand compared in the two groups. The roles of ac-celerated glucose fluix, of increased gluconeogenesisfrom alanine, and of impaired glucose disposal in theabnormal carbohydrate metabolism of chronic renalfailure are discussed.

NI ETHODS

10 healthy adults (5 men and .5 women) and 13 nondialyzedpatients with chronic renal f;ailure (11 men and 2 women) werestudied after obtaining informed consent in writing. All con-trol subjects were wvithin 10% of' their desirable body weight(Mletropolitan Life Inisurance Co. tables), and they ranged inage f'rom 22 to 36 yr (25.8±1.29 yr; mean±+SENI). Glucosetolerance was normal as determined by standard 5-h glucosetolerance testing interpreted by the criteria of Fajans andConn (14). Routine screening blood chenmistries, hemograms,and urinalyses were within normal limits as were tests ofadrenal and thyroid f unction.

The clinical data pertaining to the patients with chronicrenal failure are presented in Table I. Presumnptive etiologies(diagnoses) for renal f:ailure were based upon the following

criteria: chronic glomerulonephritis, biopsy and(or) a historyof nephrotic syndrome in a nondiabetic patient; diabeticnephropathy, insulin-dependenit mellitus of greater than 10 yrduration; obstructive uropathy, chronic lower urinary tractobstruction with bilateral hydronephrosis; nephrosclerosis,severe, inadequately treated hypertension of greater than 10yr duration before the onset of renal f:ailure; and chronic renalf'ailure of unknown etiology, no specific etiology could bef'ound. Mean weight (73.9±3.4 kg) of the azotemic patientswas not significantly different from the value for the controlsubjects (65.6±2.7 kg). Body surface area, calculated as anestimate of lean body mass, was not significantly differentin the patients with chronic renal failure (1.84+0.05 m2) whencompared to the normal volunteers (1.76+0.04 m2). In theazotemic subjects, mean blood urea nitrogen, creatinine,hematocrit, and albumini were 99.8±8.2 mg/100 ml, 10.7+1.2 mg/100 ml, 27+1% and 3.5±0.2 g/100 ml, respectively.None of'the patients had the nephrotic syndrome (24 h proteinexcretion > 3.0 g) or peripheral edema at the time of thestudy. All patients were maintained on a 2-g sodium, 2-g potas-sium, 40-g high biologic value protein diet made isocaloricto include 300 g carbohydrate for at least 3 days bef'orethe study. The patients were all normokalemic and not acidoticat the time of the study. Medications included varying dosagesof dihydrotachysterol, phosphate binders, and sodium citratesolution. Patients taking additional medications required f'orthe management of' associated conditions were included inthis study only if' these medications wvere known not to inter-fere with carbohydrate metabolism. Three patients wereknown to be diabetic bef'ore the onset of renal f:ailure;patient S. B. was taking insulin, patient S. P. was takingchlorpropamiiide. 'Medications were omiiitted the morning of'the study.

Previously described isotope dilution methodology wasmodified slightly for the present study (15). After a 12-hrovernight f'ast, short plastic catheters were inserted in theantecubital veins, bilaterally. 30 min later, base-line bloodsamples were withdrawn and a priminig intravenous injec-tion of [2-3H]glucose (0.1 ,uCi/kg) an(d [U-14C]alanine (0.1

TABLE IClinical Data of Azotemic Patients Studied*

Stirf'acePatient Age Sex W't area Diagniosis Hct BUN Creat Alb

yr kg "I2 % mg/i 00 flml mgll00 nil gll()) nil

R. S. 22 NI 72 1.83 CGN 27 89 13.0 2.5S. B. 22 NI 90 1.95 DN 21 140 19.9 1.7J. P. 35 M .58 1.61 CRFUE 22 149 12.3 4.2WNI. NI. 37 NI 81 2.02 NSC 26 82 11.8 3.4A. G. 42 NI 99 2.28 CGN 27 91 10.2 4.4D. F. 43 N1 76 1.96 CRFUE 25 120 15.7 4.1R. V. 51 NI 72 1.89 OU 29 41 7.0 3.2NI. T. 61 NI 54 1.52 DN 27 98 8.4 3.1F. L. 63 NI 75 1.86 OU 31 78 5.7 3.6K. Hl. 64 NI 74 1.83 NSC 34 130 13.6 3.9A. S. 79 NI 61 1.68 CRFUE 23 110 6.6 3.4S. P. 64 F 69 1.70 CRFUE 29 87 5.6 4.1A. WV. 65 F 80 1.81 CRFUE 25 82 8.7 4.3

Hct, hematocrit; BUN, blood urea nitrogen; Creat, serum creatinine; Alb, serumnalbuimin; CGN, chroniic glomertulonephritis; DN, diabetic nephropathy; CRFUE,chronic renal failure of uniiknown etiology; NSC, nephroselerosis; OU, obstructiveuropathy.

Increased Gltucose Production and Gluconeogenesis in Chronic Azotemia 21

,uCi/kg) was given, followed by a continuous infusion of bothisotopes (0.15 ,uCi/kg per h). Venous blood samples were ob-tained without tourniquet occlusion at 30-min intervals for5 h while the patient was at absolute bedrest. Alanine (16)and glucose (17) were determined in neutralized per-chloric acid extracts of whole blood in triplicate by fluoro-metric enzymatic techniques. Nonglucose reducing sub-stances, present in higher than normal concentrations inazotemic patients (18), are not measured by this glucose-specific assay which relies upon hexokinase (E.C.2.7.1.1.)and glucose-6-phosphate dehydrogenase (E.C.1.1.1.49). Fordetermination of their specific activities, alanine (5) andglucose (19) were removed from neutralized perchloric acidextracts of blood by ion exchange chromatography. Becauseof the redistribution of alanine between erythrocytes andplasma in anephric man (20) and to minimize blood lossfrom the azotemic patients during the study, previouslydescribed recovery techni(lues (15) with plasma were modi-fied for whole blood analysis with small volume samples(5 ml). Glucose recovery was determined for each individualblood sample analyzed and was greater than 95% in all in-stances. Alanine recoveries ranged from 93 to 99%. Theamount of tritiuim label from [2-3H]glucose appearing in thealanine pool was insignificant. Greater than 95% of '14C labelin all blood samples was accounted for by 14C label recoveredin the sum of glucose plus alanine.

Turnover rates for both glucose and alanine were calcu-lated for each patient from the plateaui portion of eachappropriate specific activity curve. Point-to-point calculationsof metabolite production and utilization by modifications ofthe equations of Steele and colleagues were also performed asdescribed previously (15, 21). However, becauise these datadid not significantly extend the data derived from the turn-over calculations, they have not been incluided. Calctulationsof precursor-produtct interrelationships for alanine and glui-cose were based upon minor modifications of the equationsdescribed by Kreisberg et al. (19). These modifications cor-rect for loss of 14C label appearing in glucose because offumarase randomization and stubsequent decarboxylation byP-enolpyruvate carboxykinase of four-carbon intermediatesin the pathway of gluconeogenesis (22, 23). Serum insulinwas measured by double antibody radioimmunoassay (24).Statistical evaluation of data was performed with Student'st test after ascertaining normal distribution (25). Results areexpressed as the mean+SEMas the index of dispersion. Com-putational assistance was provided by the CLINFO project.

RESULTS

After initiation of the simultaneous isotope infusions,constant specific activity in glucose and alanine wasobserved in each subject no later than 210 min. In most,constant specific activity was reached between 120 and150 min. As shown in Figs. 1A and 2A, constant specificactivity of the 14C label in the circulating alaninepool was attained after 180 min of infusion in controlsubjects and in azotemic patients. In comparison, con-stant specific activity of 3H label in circulating glucosewas evident after 150 and 210 min in normal volunteersand in patients with chronic renal failuire, respectively(Figs. 1C, 2C). The appearance of' 14C label in the cir-culating glucose pool from [U-14C]alanine was deter-mined to provide an index of the rate of gluconeogene-sis, at least from alanine. Constant specific activity of

the 14C label in glucose was observed by at least 270min in all control and azotemic subjects (Figs. iB, 2B).As shown in Table II, mean fasting blood glucose levelsof the nondiabetic patients with chronic renal insuf:ficiency were increased 24.3% (P = 0.005) as com-pared to control subjects, and fasting immunoreactiveinsulin levels were increased 130% (P = 0.046). Noneof the nondiabetic azotemic subjects had fasting hyper-glycemia (glucose > 6.67 mM). Calculated glucoseturnover in control subjects was found to average 664±33.5 gmol/min. Glucose turnover was increased 55%in both the nondiabetic azotemic patients (P < 0.001)and in the total group of azotemic patients studiecd(P = 0.003). These increments in the data obtainedfrom the azotemic patients were also statistically sig-nificant when expressed either on a per unit weight orbody surface basis. Glucose turnover did not correlate

A_2100-

ZE1500

U'OC 900

300 -

0 L

120 B

L0908E

30 -

0

400 c

U(A 3000-

E 200 -

^ 100 -

0-

0 60 120 180 240 300

TIME (min)FIGURE 1 14C carbon specific activity in the circutlatinigalanine (A) and glucose (B) pools and tritium specific ac-tivity in the circulating gltucose pool (C) in 10 control sub-jects after simultaneous injection-continutouis influsions of[U-'4C]alanine and [2-3H]glucose. Couints per minutte wereobtained after counting to at least the 1% significance level.Valuies are expressed as the mean+SEM.

22 S. Rubenfeld and A. J. Garber

- o

0

,Z E<C EL

a.-

O E

QE

w I

8 6czE= =:CD E

A

20 B

90[60

30[01

I I I I I I

60 120 180 240 300

TIME (min)FIGURE 2 14C carbon specific activity in the eireulating ala-nine (A) and gluicose (B) pools and tritium specific activityin the cireculating gluicose pool (C) in 13 azotemiiic subjectsafter primne(l injection-contintlous infusions of [U-14C]alanine and [2-3H]glucose. Valtues are expressed as themean-+SEN1.

with fasting glucose levels or body surface area ineither controls or nondiabetic azotemic subjects, nordid glucose turnover rates correlate with the age of theazotemic subjects studied.

No differences were found in fasting blood alaninelevels between the various groups studied (Table III).In control subjects, alanine turnover averaged 163+ 19.4 ,mol/min. Rates of alanine turnover in theentire azotemic population and in the nondiabeticsubpopulation were increased 191% (P = 0.007) and234% (P = 0.008), respectively. No correlation couldbe found between age, weight, body surface area, orfasting alanine level and the rate of alanine turnoverin either the azotemic or control group.

Glucose production from alanine was estimated bythe rate of 14C label appearance in glucose from[U-14C]alanine. In control subjects, glucose producedfrom alanine was 21.7+2.24 ,umol/min, amounting to3.21+0.42% of total glucose production (Table IV).In the nondiabetic azotemic subjects, glucose produc-tion from alanine was increased 221% (P = 0.003),whereas in the entire azotemic population the rate ofgluconeogenesis was increased 192% (P = 0.02). Therelative contribution of gluconeogenesis from alanineto total glucose production was also increased to 9.74+3.00% (P = 0.059) in the nondiabetic subjects withchronic renal failure and to 8.64+2.38% (P < 0.05) inthe entire group of azotemic patients. The rate ofglucose production from alanine correlated well withthe rate of alanine tuirnover in both control subjects(r = 0.729, P = 0.017), and the entire azotemic popula-tion (r = 0.626, P = 0.022) as shown in Figs. 3 and 4,respectively. However, a lesser degree of correlationbetween rates of gluconieogenesis from alanine andalaniine turnover wvas found in the nonidiabetic patientswith renal failure (r = 0.597, P = 0.069).

In control subjects, 40.2+3.86 ,mol/min of alaninewere uitilized for glticoneogenesis, accounting for 25.4

TABLE IIGlucose Kitnetics in Azotemic and Conitrol Snbjects

Patienits Inistulin Glticose Gltucose turnover

AtU/In I mAM ,Amol/min Amol/kg/mnin Amol/mn'min

Normal volunteers 6.61±0.72 4.23±0.22 664±33..5 10.3 ±0.67 379±22.6

Azotemic subjectsTotal 6.29± 1.15 1,035±99.3* 14.2± 1.29t 563±52.5*Nondiabetics 1 5.2±3.44t .5.26±0.23* 1,034-76.5§ 14.2+ 1.29t .556±-+46.1*

\'aliies shown are the means-SENM for each determination. P represenits the probability oft rejectionot the nutll hypothesis of etqtuality between the footnoted population mean and the appropriate controlmean.* P < (.(1.

P < 0.0(5.§ P < (.()01.

Increased Glucose Prodtwctiotn atndI (4,luconeogenesis in C/irotoic Azotemlia 23

TABLE IIIAlanine Kinetics in Azotemic and Control Subjects

Patients Alanine Alanine turnover

mM Amol/min AmolIkg/min ttmolIm'Imin

Normal volunteers 0.233+0.021 163+19.4 2.51+0.29 92.7+10.7

Azotemic subjectsTotal 0.245±0.019* 474±96.0t 6.47±1.36§, 257±53.1tNondiabetic 0.259±0.023* 545±114t 7.47± 1.65§ 295±64.05

Values shown are the means ±SEMfor each determination. P represents the probabilityof rejection of the null hypothesis of equality between the footnoted populationmean and the appropriate control mean.*P > 0.1.* P < 0.01.§ P < 0.05.

+ 1.52% of total alanine utilization (Table V). In thenondiabetic azotemic patients, alanine utilization forgluconeogenesis was increased 305% (P = 0.034).However, the fraction of total alanine utilization ac-counted for by gluconeogenesis was unchanged inazotemic subjects as compared to control subjects.

DISCUSSION

Isotope dilution techniques have been widely used forthe in vivo study of intermediary metabolism, par-ticularly of glucose and glucose precursors (26). Simul-taneous infusions of labeled glucose and gluconeo-genic precursors have been employed previously to ob-tain accurate data concerning gluconeogenesis in vivo(27). The use of [2-3H]glucose provides a relativelygood estimation of total glucose production becausedetritiation occurs by the action of phosphohexose-isomerase (28). However, recent studies have indicatedthat overestimations of glucose utilization rates resultfrom the use of [2-3H]glucose as compared to [3-3H] or[6-3H]glucose because of futile cycling between

glucose and glucose-6-phosphate (29, 30). Never-theless, in a preliminary analysis of our data fromanother study,' no significant difference could bediscerned between the mean rates of glucose turnoverdetermined with [2-3H] or [3-3H]glucose in normalvolunteers (663+33.5 vs. 687+49.7 gmol/min, respec-tively) or in patients with chronic renal insufficiency(1034+86.5 vs. 1074±139 ,umol/min, respectively).With [2-3H]glucose, rates of glucose production ob-served in the normal volunteers of this study (663,umol/min or 172 g/day) appear to be in excellentagreement with values obtained by others with isotopedilution methods (8, 31) or hepatic venous catheteriza-tion techniques (32, 33). Similarly, the fraction of totalglucose production contributed by gluconeogenesisfrom alanine in the normal subjects of this study (3.21±0.42%) is in good agreement with the data of otherstudies with isotope techniques (34, 35).

l S. Rubenfeld and A. J. Garber. Abnormal carbohydratemetabolism in chronic renal failure: suppression ofendogenous glucose overproduction by exogenous glucose.Manuscript in preparation.

TABLE IVGlucose Production from Alanine in Azotemic and Control Subjects

Patients Glucose from alanine Glucose from alanine

gmol/min jAmollkglmin pAmolIms/min %

Normal volunteers 21.7+2.24 0.339+0.041 12.4+1.38 3.21+0.42

Azotemic subjectsTotal 63.3+14.9* 0.888+0.231* 35.1±8.71* 8.64+2.38*Nondiabetic 69.6+18.9* 0.983+0.2951 38.5±11.1* 9.74±3.00t

Values shown are the means±SEMfor each determination. P represents the probabilityof rejection of the null hypothesis of equality between the footnoted population meanand the appropriate control mean.* P < 0.05.* P = 0.059.

24 S. Rubenfeld and A. J. Garber

y= 0.084x+7.98( r=0. 729)

35

C

;C

0..E E

%A

E

0

CD

200 [

301-

25

201

a.a

E E0 _L- o

qZE

84A0 0

15

`80 120 160 200 240 280 320

Alanine Turnoverlmol /min

FIGURE 3 Relationships between gluconeogenesis from ala-nine and alanine turnover rate in 10 control subjects. Eachpoint is the result of a 300-min primed injection-continuousinfusion of [U-'4C]alanine and [2-3H]glucose in separate pa-tients. The solid diagonal line inidicates the least s(quare linesof regresssion.

Although observations concerning glucose andglucose-precursor kinetics in chronic renal failureare previously unreported (1, 36), WZestervelt observedthat insulin-stimulated rates of glucose uptake in fore-arm studies from azotemic patients were decreased75% compared to rates observed with control subjects(37). Decrements in blood gltucose concentration afterintravenous insulin injection (2, 4, 38) or tolbtutamideinjection (18, 38) were diminished and slower in de-velopment in azotemic as compared to control sub-jects. In this study, as in other studies, increasedfasting immunoreactive instulin levels and normalfasting glucose levels were found in chronic azotemia(2, 3, 18, 38, 39). However, the extent to which these

y 0.098x +17.1( r=0. 626 )

160-

120 r

80

40

* d' 0

600

Alanine Turnoverpmol /min

800 1000 1200U0 200 400

FIGURE 4 Relationship between gltuconeogenesis fromalanine and alanine turnover rate in 13 azotemic subjects.The open circles represent nondiabetic azotemic patientsand the filled circles represeent diabetic azotemic patients.

increased immunoreactive insulin levels may reflectthe accumulation of proinsulin was not ascertained.The results of previouis stuidies have been interpretedto suggest the presence of insulin resistance in chronicrenal failure. The results of this study demonstrateincreased basal glucose flux in chronic renal insuffi-ciency. This increased glucose prodtuction observed inthe postabsorptive state may result in part from anincreased rate of gluconeogenesis from alanine be-cause the latter was inereased 192% in the azotemicsubjects as compared to conitrol subjects. The increasedgluconeogenesis in turn appears to depend in part uponan increased rate of alanine productioni, without whichgluconeogenesis could not be sustained for any sig-nificant period. In addition, an independent activationof hepatic alanine metabolism towards gluconeogene-

TABLE VAlanine Utilization for Glucorneogenesis in Azotemic antd Control Subjects

Patients Alanine to glutcose Alanine to glucose

,rniolinin ,umol/kg/min jAmolIm2/min Y

Normal volunteers 40.2±3.86 0.624±0.069 22.9+2.33 25.4+1.52

Azotemic stubjectsTotal 143 +39.0* 2.01 +0.600* 79.4 ±22.8* 30.1 +4.061Nondiabeties 163+±48.9* 2.31 ±0.760§ 90.6±28.8* 29.3+5.204

Values shown are the means±SENM for each determination. P represents the probabilityof rejection of the null hypothesis of equality between the footnoted population meanand the appropriate control mean.* P < 0.05.

P > 0.1.§ P = 0.055.

Increased Glucose Production and Gluconeogenesis in Chronic Azotemia

40 r 240 r

0

0

00

0

0

inI

25

sis mi-ust also l)e present because the percentage of'total glucose prodtictioni conitrib)uted l)y gliuconeogenie-sis w-as inicreased 169% ini azotemiic subljects, and I)e-cause alaninie lev-els were unchaiiged desp)ite a thre-e-folId inicrease in atlaninie delivery. In niormials, in-creased alaniniie delivery as providled by ani alaniniieinifutsion resuilts ini incereased alaninie levels (40). Thesefinidings of' increased glutconeogeniesis lin atzoteinicimiani correlate wvell with finin(iigs of increased in vitrohepatic gluiconeogeniesis in the urem-ic rat (12, 13).

The total inceremienit ini glucose productioni observedini patienits with chronici renial fatiluire cannot lbe ac-

Counted fo(r en-tirely, by the increased rate of' glutco-nieogeniesis f'romi alaniniie. Althoutgh niot specificallydetermiiined in this stud(y, glucose produiction f'romprecursors suchl as glycerol, lactate, aindc other aminiioacidis may also be incereasedl. Even so, it seem-is un-

likely that increased glmiconieogeniesis alon-e cani pro-dutce the enitire elev-ationi in glucose prodllctimn l)e-cautse th-e uiptake of' all gluicose precursors accountsfu(r onily abouit 25% of' hiepatic gluicose produtctioni inip)ostabsorptive niormial miani (41), and only :32% in dlia-betic man (42). Becauise total gluicose productioni in

azotemice subjects w-as inicreasedI by miore thian 50%/,,somie elemienit of' incereatsed glucogenolysis as wvell as

incereasedI gluconeogenesis seemis likely'.The datai of the presenit study SUggeSt that ani inisensi-

tivity or resistancee to the metabl)(ic actioni of' inisulinimay be ani important mechtanism in the abnormalglutcose andi( glucose-precursor ki imetics ol)serve(l inichironici renial insufficiencly. The fin dlings of' ele\atedfimmunoreactivc inisulini levels togethier withi inicreatsedglycogenolysis and gIuconieogenesi s suiggest an in-sensitivity to inisulin mi-odutlationi or hepatic glucoseproduictioni. The demonstration of' tlimini ished( inistil inbinding to hiepatic me-nblranes isolatted f'romi experi-mientally uiremici rats suiggests a potential mi-echanismifor the hiepatic inisulini insensitiv~itv because dliimin-ishied receptor binding may produce (liminiiished lhor-inione actioni (43), Additionally, imipairedI inisulini ac-tioni oni skeletal muitscle maxit accounit f'Or the inc-reatsedalaniniie turnover observed ini patienits with chroniicrenal insuifficiencly. Becauise the action o)f' inisulini isimiportanit to the regulation of' both proteini synthe-sins and proteini (legradatini-m ini skeletal muiLscle (44),insensitivity or resistanice to insulin miighit be asso-ciated with either acceleratedI muiscle p)roteolys5is ordiminished amiiino) acid incorporation inito muiiscle pro-tein or both. As a conisequtence of' either process, ac-celeratedI alaninie f'orm-ationi and( release would beaniticipated (45). These finin(iigs of' incereased alaninieproductimn in postabsorptiv e azoteinic miani extenid ourprior observationis of' ncreased skeletal muiscle alaninef'ormation and release observed ini miuscle preparattionisf'romi experinentally uiremici rats (1 1). Ani imtpairmientof insulin-mnediated glucose uitilization maiiy also be in-f'erred f'romt the dlata of' this stud(y. M/eani f'asting im-

mutnoreactive insuilini levels wvere incereased 130% anidmiean fhsting gluicose levels were increased 20% inazotemici suibjects. These finin(ilgs, coupled with the ohb-serv'ation of a 56% inicrease in gluicose fluix, suggestani imipairmienit of' gluicose disposal. Otherw\ise, un-

chaniged gluicose levels would be anticipated dlesp)itethle Substantial inicIimmentcts in gluicose fluix.

Althouigh ani insensitivity or resistancee to insutlinactioni cani accounit fo(r miost of' the dleranigemienits ingluicose and( glucose-precursor kinietics observ'ed inazotemnic' subjects. abnIormnalitie s in the action anidcirculating levels of o)ther hiormionies have also b)eeni re-

ported in chironiic rencal insufficiency. Gluicacgomi en--hiances both hepatic alainine extractioni and( gluiconco-geniesis (34). Because increasedI circutlatinig glucagonlevels, incluidinig the 3,500 miol \\t f'raction, havebeeni reported in chironiic renial f'ailutre (39), pathologichypergliucagoniemiia may account in part fu(r the in-creasedI gluiconeogenesis obIserv,ed in azoteiiic uman.Parathyroid hiorm-ione, like glucagon, appears to stinimu-late hiepatic adeniylate cyclase activity (46) and( hepat-tic gluiconieogeniesis f'romi thiree carbon prectirsors (47).Whereas the eff'ect of' parathyroid hiormionc o)n adlenyl-ate cyclase activity is less thian- the stiimutlation pro-dutced by epiniephri ne o)r glutcagoni, parathyroid h-or-mionie levels are suibstanltially elevated in chironiic renialf'ailuire (:36) andi( may' contributetc to the inicreased1hiepatic gluiconieogemmesis. Ihoweve-r, a single study hiassuiggestedI that parathyroi dectomy h-as nio ef'fect tiponthie abnormal gluicose mectabolism of' chiron-ic- renialfhailutre (48).

In sumnmarv, the finidinigs of the p)rcsent studyt demin)ml-strate incrcasedl gluicose production, uitilizationi, andgluconeogenesis in patienits with chlronici renial t'aluire.InicreasedI glucose produiction f'romi alaniIiie appears toderive f'romi a primary abnormality of'hepatic gILucosemietabolismi. Rates of' alaniniie produtction atlso wveresubstantially incereatsed su ggestliug ani indltepenidenit ab-niormiality )1' alanine metabolism in L)eripberal tissuiessutch as skeletal inn iscle. Relative elevationis of' f'astiinggluicose and ilInMuLnoreactive inisulini levels were

ol)served in azotemi-ic pattienits. Taken as at whole, thefinin(iigs ofI' incereasedi alaninie producttioni, inicreatsediglucose production, incereatse(l gluconeo)genesi s, andimipaired gluicose dlisposal, all o)f which are processesreguilatedi by insuinii, suiggest ani undierlyinig loepatticas well as peripheral iinsensitivitv to insutlini actioni inchironici renial insufficiencx'.

ACKNO\VLEDGMIENTS

We are jildiebtedi to Nuirit Sh-inionii, Catrul Miaillet, SherryBarlow, aidl Lande Siindu fb(r exc-ellent technical assistanee.The advice aidi commentary of' Dr. Woadhi Suiki is grateftillyackniowlediged. WNe are also indeb)tedl to Ducturs CharlesCrumlb, Edward Wseinman, andl Carmielo Dichusu fo(r ref'erralof their patients fior this study.

The researchi fo(r this report was supported in part bx

26 5. Rtuben]j'eld anoi Aj. G.(arber

contract AM-4-2215 from the Artificial Kidney-ChronicUremia Program; by the CLINFO Project, contract RR-5-2218; grant RR-0350 from the Division of Research Resources,National Institutes of Health; and by a grant from theKelsey-Leary Foundation.

REFERENCES

1. DeFronzo, R. A., R. Andres, P. Edgar, and W. G. Walker.1973. Carbohydrate metabolism in uremia: a review.Medicine (Baltimore). 52: 469-481.

2. Horton, E. S., C. Johnson, and H. E. Lebevitz. 1968.Carbohydrate metabolism in urea. Ann. Intern. Med. 68:63-74.

3. Lowrie, E. G., J. S. Soeldner, C. L. Hampers, and J. P.Merrill. 1970. Glucose metabolism and insulin secretionin uremic, pre-diabetic, and normal subjects. J. Lab.Clin. Med. 76: 603-615.

4. Perkoff, G. T., C. L. Thomas, J. D. Newton, J. C. Sellman,and F. H. Tyler. 1958. Mechanisms of impaired glucosetolerance in uremia and experimental hyperazotemia.Diabetes. 7: 375-383.

5. Garber, A. J., D. M. Bier, P. E. Cryer, and A. S. Pagliara.1974. Hypoglycemia in compensated chronic renal in-sufficiency: substrate limitation of gluconieogenesis.Diabetes. 23: 982-986.

6. Frizzell, M., P. R. Larsen, and J. B. Field. 1973. Spontane-ous hypoglycemia associated with chronic renal failure.Diabetes. 22: 493-498.

7. Runyan, J. W., D. Hurwitz, and S. L. Robbins. 1955.Effect of' Kimmelstiel-Wilson syndrome on insulin re-quirements in diabetes. N. Engl. J. Med. 252: 385-391.

8. Bowen, H. F., and J. A. Moorhouse. 1973. Glucoseturnover and disposal in maturity-onset diabetes. J.Clin. Invest. 52: 3033-3045.

9. Felig, P., T. Pozefsky, E. Marliss, and G. F. Cahill, Jr.1970. Alanine: key role in gluconeogenesis. Science(Wash. D. C.). 167: 1003-1004.

10. Wahren, J., P. Felig, E. Cerasi, and R. Luft. 1972.Splanchnic and peripheral glucose and amino acidmetabolism in diabetes mellitus. J. Clin. Invest. 52:1870-1878.

11. Garber, A. J. 1977. Skeletal miuscle alanine and glutamineformation and release in experimental uremia. Clin. Res.52: 32A. (Abstr.)

12. Frohlick, J., J. Scholmerich, G. Hoppe-Seyler, K. P.Maier, H. Talke, P. Schollmeyer, and W. Gerok. 1974. Theeffect of acute uremia on gluconeogenesis in isolated per-fused rat livers. Euro. J. Clin. Invest. 4: 453-458.

13. Dzurik, R., T. R. Niederland, and P. Cernacek. 1969.Carbohydrate metabolism by rat liver slices incubated inserum obtained from uremic patients. Clin Sci. (Oxf.).37: 409-417.

14. Fajans, S. S., and J. XW. Conn. 1959. The early recogni-tion of diabetes mellitus. Ann. N. Y. Acad. Sci. 82:208-218.

15. Garber, A. J., P. E. Cryer, J. V. Santiago, M. XV. Haymond,A. S. Pagliara, and D. M. Kipnis. 1976. The role ofadrenergic mechanisms in the substrate and hormonalresponse to insulin-induced hypoglycemia in man. J.Clin. Invest. 58: 7-15.

16. Karl, I. E., A. S. Pagliara, and D. M. Kipnis. 1972. Amicrofluorometric enzymatic assay for the determina-tion of alanine and pyruvate in plasma and tissues.

J. Lab. Clin. Med. 80: 434-441.17. Lowry, 0. H., and J. V. Passoneau. 1972. A Flexible

System of Enzymatic Analysis. Academic Press, Inc.,New York. 174-177.

18. Cerletty, J. M., and N. H. Enbring. 1967. Azotemia andglucose intolerance. Ant. Intern. Med. 66: 1097-1108.

19. Kriesberg, R. A., A. M. Siegal, and W. C. Owen. 1971.Glucose-lactate interrelationships: Effect of' ethanol.

J. CliGu. Invest. 50: 175-185.20. Ganda, 0. P., T. T. Aoki, J. S. Soeldner, R. S. Morrison,

and G. F. Cahill, Jr. 1976. Hormone-fuel concentrationsin anephric subjects. Effects of hemodialysis (withspecial reference to amino acids). J. Clitn. Invest. 57:1403-14011.

21. Steele, R., H. Rostami, and N. Altszuler. 1974. A two-compartmenit caleulator for the dog glucose pool in thenonsteady state. Fed. Proc. 33: 1869-1876.

22. WN'alter, P., V. Paetkau, and H. A. Lardy. 1966. Paths ofcarbon in gluconeogenesis and lipogenesis. III. The roleand regulation of' mitochondrial processes involved insupplying precursors of phosphoenolpyruvate. J. Biol.Clieni. 241: 2523-2532.

23. Hanson, R.VW., and A. J. Garber. 1972. Phosphoeinol-pyruvate carboxykinase. I. Its role in gluconeogenesis.Am.]. GnCli. Nutr. 24: 1010-1021.

24. Hales, C. N., and P. J. Randle. 1963. Immunoassay of'insulin wvith insulin-antibody precipitate. Biocheem. J.88: 137-146.

25. Bahn, A. K. 1972. Basic NMedical Statistics. Grune andStratton, Inc., New York. 136-143.

26. Searle, G. L. 1976. The use of isotope turnover tech-niques in the study of carbohydrate metabolism inman. Clin. Enidocrinol. Metab. 5: 783-804.

27. Chiasson, J. L., J. E. Liljenquist, WN7. WV. Lacy, A. S.Jennings, and A. D. Cherrington. 1977. Gluconeogeniesis:Methodological approaches in vivo. Fed. Proc. 36: 229-235.

28. Katz, J., and A. Dunn. 1967. Glucose-2-t as a tracer forglucose metabolism. Biocheemistry. 6: 1-5.

29. Katz, J., H. Tostami, and A. Dunn. 1974. Evaluation of,glucose turnover, body mass and recycling with re-versible and irreversible tracers. Biochein. J. 142: 161-170.

30. Katz, J., A. Duinn, M. Chenowveth, and S. Golden. 1974.Determination of' synthesis, recycling and body mass ofglucose in rats and rabbits in vivo with 3H- and "C-labelled glucose. Biochem. J. 142: 171-183.

31. Paul, P., and XV. M. Bortz. 1969. Turnover and oxidationof plasma glucose in lean and obese hu-mans. Metab.Clin. Exp. 18: 570-584.

32. Wahren, J., P. Felig, G. Ahlborg, and L. Jorfeldt. 1971.Glucose metabolism duiring leg exercise in man. J. Clin.Invest. 50: 2715-2725.

33. Wahren, J., P. Felig, and L. Hagenfeldt. 1976. Effect ofprotein ingestion on splanchnic and leg metabolism innormal man and in patients with diabetes mellituis. J.Clitn. Invest. 57: 987-999.

34. Chiasson, J. L., J. E. Lilienquist, B. C. Sinclair-Smith,and XV. W. Lacy. 1975. Gluconeogenesis from alanine innormal postabsorptive man: intrahepatic simuilatorv ef-fect of' glucagon. Diabetes. 24: 574-584.

35. Chochinov, R. H., and J. A. Moorhouse. 1972. Alanineand gluconeogenesis in health and diabetes. Diabetes.21:(Suppl. 1): 341-342.

36. Feldman, H. A., and I. Siniger. 1974. Endocriniology andmetabolism in uremia and dialysis. A clinlical review.Medicine (Baltimnore). 54: 345-376.

37. Westervelt, F. B. 1979. Uremia and insuilini response.Arch. Interni. Med. 126: 865-869.

38. Spitz, MI. M., A. H. Rubenstein, 1. Bersohn, C. Abrahams,and C. Lowry. 1970. Carbohydrate metabolism in renaldisease. Q. J. Med. 39: 201-226.

39. Sherwin, R. S., C. Gastl, F. 0. Finkelstein, M. Fisher,

Increased Glucose Production and Gluconteogentesis in Chroniic Azotemnia 27

H. Black, R. Hendler, and P. Felig. 1976. Influence ofuremia and hemodialysis on the turnover and metaboliceffects of glucagon. J. Clin. Invest. 57: 722-731.

40. Pagliara, A. S., I. E. Karl, D. C. DeVivo, R. D. Feigin, andD. M. Kipnis. 1972. Hypoalaninemia: a concomitant ofketotic hypoglycemia. J. Clin. Invest. 51: 1440-1449.

41. Felig, P. 1973. The glucose-alanine cycle. Metab. Clin.Exp. 22: 407-413.

42. Wahren, J., P. Felig, E. Cerasi, and R. Luft. 1972.Splanchnic and peripheral glucose and amino acidmetabolism in diabetes mellitus. J. Clin. Invest. 51:1870-1878.

43. Soman, V., and P. Felig. 1977. Glucagon and insulinbinding to liver membranes in a partially nephrectomizeduremic rat model. J. Clin. Invest. 60: 224-232.

44. Fulks, R. M., K. B. Li, and A. L. Goldberg. 1975. Effects

of insulin, glucose, and amino acids on protein turnoverin rat diaphragm. J. Biol. Chem. 250: 290-298.

45. Jefferson, L. S., D. E. Rannels, B. L. Munger, andH. E. Morgan. 1974. Insulin in the regulation of proteinturnover in heart and skeletal muscle. Fed. Proc. 33:1098-1104.

46. Canterbury, J. M., G. Levy, E. Ruiz, and E. Reiss. 1974.Parathyroid hormone activation of adenylate cyclase inliver. Proc. Soc. Exp. Biol. Med. 147: 366-370.

47. Moxley, M. A., N. H. Bell, S. R. Wagle, D. 0. Allen,and J. Ashmore. 1974. Parathyroid hormone stimula-tion of glucose and urea production in isolated livercells. Am. J. Physiol. 227: 1058-1061.

48. Amend, W. J. C., Jr., S. M. Steinberg, E. G. Lowrie, J. M.Lazarus, J. S. Soeldner, C. L. Hampers, and J. P. Merrill.1975. The influence of serum calcium and parathyroidhormone upon glucose metabolism in uremia. J. Lab.Clin. Med. 86: 435-444.

28 S. Rubenfeld and A. J. Garber