Insulin. Its role in the thermic effect of glucose.
L Christin, … , E Jéquier, K J Acheson
J Clin Invest.
1986;
77(6)
:1747-1755.
https://doi.org/10.1172/JCI112497
.
To investigate the possible role of insulin per se in the thermic response to glucose/insulin
infusions, respiratory exchange measurements were performed on eight healthy young men
for 45 min before and 210 min after somatostatin infusion. Two tests were performed on
separate days and each had two consecutive phases of 90 min each. Test 1. Two different
rates of glucose uptake were imposed, one at euglycemia (phase 1) and the other at
hyperglycemia (phase 2) while insulinemia was maintained constant throughout. Test 2.
Glucose uptake was maintained constant throughout while insulin was infused at two
different rates: 1 mU/kg per min with hyperglycemia (phase 1) and 6.45 mU/kg per min with
"euglycemia" (phase 2). The thermic effect of glucose and insulin, obtained from phase 1 in
both tests, was 5.9 +/- 1.2 and 5.8 +/- 0.5% (NS) of the energy infused, respectively. A step
increase in glucose uptake alone, test 1, phase 2, (0.469 +/- 0.039 to 1.069 +/- 0.094 g/min)
caused an increase in energy expenditure of 0.14 0.03 kcal/min (thermic effect 5.9
+/-1.1%). When insulin was increased by 752 +/- 115 microU/ml, with no change in glucose
uptake, energy expenditure rose by 0.05 +/- 0.02 kcal/min, which correlated with the
increase in plasma catecholamines. It is concluded that a large proportion of the thermic
response to […]
Research Article
Find the latest version:
Insulin
Its Role inthe Thermic Effect of Glucose
L.Christin, C.-A. Nacht,0.Vemet,E.Ravussin,E.Jequier, andK.J. Acheson
Institute ofPhysiology, UniversityofLausanne, 1005LausanneandNestleResearchDepartment, 1800 Vevey,Switzerland
Abstract
Toinvestigate the possibleroleofinsulin persein thethermic
response to glucose/insulin infusions, respiratory exchange measurements wereperformed on eight healthy young menfor
45 min before and 210 min after somatostatin infusion. Two
testswere performed on separate days and each hadtwo
con-secutive phases of90min each. Test 1. Twodifferentrates of glucose uptake wereimposed, oneat euglycemia(phase 1)and the other at hyperglycemia (phase 2) while insulinemia was
maintained constant throughout.
Test 2.Glucoseuptake was maintainedconstantthroughout
while insulin was infused at twodifferent rates: 1 mU/kg per minwith hyperglycemia (phase 1) and6.45mU/kgper min with
"euglycemia" (phase
2).
The thermiceffectof glucose andinsulin, obtainedfrom phase
1 in both tests, was5.9±1.2 and 5.8±0.5% (NS) of the energy infused, respectively. A step increase in glucose uptake alone,
test1, phase2,
(0.469±0.039
to1.069±0.094g/min)causedanincrease in energy expenditure of0.14±0.03kcal/min (thermic effect 5.9±1.1%).When insulinwasincreasedby752±115
AUI
ml, withnochange in glucose uptake,energyexpenditure rose by 0.05±0.02
kcal/min,
which correlated with the increase in plasma catecholamines.Itis concluded that a large proportion of the thermic response
toglucose/insulininfusions isduetoglucosemetabolism alone. Thethermic effect of insulin is small
aind
appearstobemediated by the sympatheticnervoussystem; thusatphysiologicalinsulinconcentrations,thethermic effect of insulin perseisnegligible.
Introduction
The
thermic
effect of glucose inleanindividuals either afteroralingestionorduring intravenous
infusions
isgreater(1, 2)thancanbeattributedto the energycostofstoring glucoseasglycogen (3). Onthe otherhand,obesesubjectsoften(4-8)butnotalways (9-11) present with a reducedthermic effect when compared
with lean controls, which is aggravated by increasing insulin
resistance(5).Althoughadiminished thermic effect offoodhas been proposedtocontributetothepathogenesisofobesity( 12-14), itis uncertain whether thisisacause or aconsequence of theobese state,particularlysinceithas been demonstrated that when the glucose uptakeis the same in lean and obeseindividuals
Addressreprint requests toDr. K.J.Acheson,Instituteof Physiology, University of Lausanne,Rue duBugnon 7, CH-1005, Lausanne, Swit-zerland.
Receivedfor publication 7November 1985 and in revised form 14
February1986.
thethermic effect ofglucose/insulin is also the same (15). Roth-well and Stock( 16)have reported that the increase in metabolic
rateinresponse to amealwasreduced instreptozotocin-induced
diabetic rats butwas reestablished in those receiving insulin. This has led to thesuggestion that insulin action is involved in thethermicresponse tofood ingestion (16, 17).
Since the effects of glucose and insulin metabolismare so
intimately related,it is difficult to conclude if it is a directeffect
of insulinindependentofglucosemetabolism thatisresponsible for the observed increase in thermogenesis. This study was
therefore designed toseparately measurethe thermic effect of
glucose alone and the influence of insulin itselfonenergy
ex-penditure in man. For this purposeeuglycemic and hypergly-cemicinsulin clampingwasperformed inhealthy youngmale
subjects with or without somatostatin infusionin combination
with continuousrespiratory exchangemeasurements. This
ap-proach allowed us todetermine whether insulin per se plays a rolein the thermic response toglucose/insulin infusions.
Methods
Subjects
Eight healthy young male subjects, whose physical characteristics are presented in TableI, were studied. No subject had a family history of diabetes mellitus andnone wastaking any medication during the
ex-perimentalprotocol. The protocolwassubmitted,reviewed, and accepted by the ethical committee of the Faculty ofMedicine, University of Lau-sanne.After receivingadetailed explanationof the experimental protocol, written consent was obtained from each volunteer before acceptance into the study.
Experimentalprotocol
For3 d before each test the subject was instructed to eat a diet containing atleast250 gof carbohydrate, supplemented with sugared fruit juice providing an additional 60 g of hexose sugars per day. Each subject was requested to keep a logofhis diet and physical activity during this period sothat he couldreplicatesimilar dietary and activitypatterns in the days preceding subsequent tests.
Eachsubject spent thenightbefore thetest attheInstitute. Inthe
morningthesubject was awakened at 6:30 a.m. and after voiding was transferredto the roominwhichthetest was tobe performed. Two venouslines were inserted. One was a venous catheter(Venflon, Viggo, AB,Helsingborg, Sweden) placed inanantecubital veinforinfusionof glucose and hormones and the other,a19-gauge Butterfly (Abbott Ireland Ltd,Sligo, Republic of Ireland)wasinsertedretrogradely into a hand vein forblood samplingand was kept patent with physiologicalsaline.
The handwas then placedinaheated box (-60°C) toachieve arterial-izationof thevenousblood (18).
Thesubject performedfourexperimentalprotocols within a period of2 mo. Thefirst and second (Fig. 1) and the third and fourth experiments wereperformed within 2 wk.
Testlawith constant insulin infusion at two successive steady state
plasmaglucose concentrations. This test was designed to study the effect ofanincrease in glucose uptake on energy expenditure, without a con-comitant changeininsulinemia.
After 45 min ofbaseline measurements, somatostatin (Stilamin,
Ser-ThermicEffect ofInsulin 1747
J.Clin. Invest.
© The American Society for Clinical Investigation, Inc. 0021-9738/86/06/1747/09 $ 1.00
Table LPhysicalCharacteristicsoftheSubjects
Name Age Weight Height Bodyfat FFM
yr kg cm %kg
C.M. 20 68.24 176.5 12.9 59.6
M.L. 20 78.45 185.5 16.2 65.74
N.F-Y. 21 65.05 175.5 0.6 58.2
C.C. 20 85.60 181.0 14.7 73.0
LU. 19 62.64 181.5 11.6 55.37
G.T. 22 67.75 186.0 12.0 59.62
B.P-Y. 22 81.72 194.5 15.5 69.05
E.K. 22 67.98 180.0 14.4 58.2
Mean±SD 20.8±1.2 72.2±8.5 182.7±6.1 13.5±2.0 62.3±6.2
onoS.A.,CH-1 170Aubonne,Switzerland)wasinfusedatacontinuous rate (7
iAg/min)
forthe.
next 2 hof theexperiment. Thisinfuision ratewaschosentoeffectivelyinhibitendogenousinsulinsecretion(19, 20).
After 30 min of somatostatin infusion aloneaprimingdose of
bio-synthetichumaninsulin(Huminsulin Normal,EliLilly, Indianapolis,
Euglycemic Hyperglycemic 1
clamp clamp
Insulin 1mrnWkg.min
G lucose 20%
M-0.47±0.4' M.1.07k009~
I 7pg/min
9
9
gg/min---I
Roespieatory Exchange Measurements-60 Blood f
samples
Test2a
0 60 120 180minutes
Hyperglycemic 'Euglycemic'
clamp clamp
Insulin
G lucose 20%
M-0.88±t.10 F-- M-0-93±0O40
Somatosta,tin ,mn..1
~~~~~p/i 1~ 9p mn~
-60 Blood t
saimples
0 60 120 180 minutes
Figure1.Experimentaldesign.Thestudywasdivided intotwo tests:
(Test la)Plasmainsulin concentrationsweremaintainedat '--80UU/
mlthroughoutthetest.Gluco'seuptake (M)wasraised from 0.47±0.04to 1.07±0.09g/min (Mean±SEM)at120minin orderto
raiseglycemiafrom 91 to171 mg/dlandmaintain itatthatlevel for another90mmn.(Test2a)Mwasmaintainedconstant at ~--0.9g/min throughoutthetest. At 120minplasmainsulin concentrationswere
raised from 95to848usU/ml.Since Mwasconstant,glycemia de-creasedfrom 170to 1mg/dl.
IN)wasgiveninadecreasingmanner over aperiodof 10 minas
pre-viouslydescribed (18)andwastheninfusedatacontinuousrateof 1
mg/kgpermin foir the remainder of thetest.Euglycemiawasmaintained inarterializedvenousbloodsamplesobtainedat5-minintervals,during
the somatostatin and insulininfusions, byvaryingthe infusionrateofa
20%,glucosesolution(18).After90min ofhyperinsufinemic euglycemia,
aprimingdose ofthe20%glucosesolutionwasgiveninalogarithmically
decreasingmanner(18)andwasthenvaried in ordertoraise and maintain the blood glucoselevel at 170-180 mg/dlfor another 90 min. When hyperglycemiawasimposed,acondition that stimulatesendogenous in-sulinsecretion,therateof somatostatin infusionwasincreasedto9 utg/
mintoensurethat endogenous insulindid not escape somatostatin's inhibitoryeffects.
Test Ib: SameprotocolasIlabut without somatostatininfusion.This
test wasusedas acontrol fortest Ia, inordertocheck whether
somna-tostatin,whichwasgivenintest1a,hadaneffectonenergyexpenditure.
Test 2a with constantglucose infusion at two successive levels of insulinemia. After 45 min of baseline measurements, somatostatinwas
infusedat7
gg/min
for2 h. Whensomatostatin had been infused for 30 minaprimecontinuous infusion of Huminsulinwasbegunatarateof 1 mU/kg perminasdescribed above (test la).Atthesametimea
primingdose ofa20%glucosesolutionwasinfused and then varied in ordertoraise and maintainglycemiaat170-180mg/dlfor 90 min. After 90 min ofhyperinsulinemichyperglycemia, the somatostatin infusion
ratewasincreasedto9
gg/min
and insulinemiawasincreasedbyaprimecontinuousinfusion(6.45 mU/kgpermin)forafurther 90min, while
maintainingtheglucoseinfusionconstant.
Test2b: Sameprotocol'as2abut without somatostatininfuion.(This
test wasusedas acontrol fortest2a). Threesubjectswhoparticiae in the above
experiments
also volunteered fortwoadditionaltestswithoutsomatostatin,inwhich either the 1-mU/kgperminhyperinsulinemic, euglycemic clampor thehyperinsulinemic, hyperglycemic clampwas
continued for the wholedurationof thetestin ordertoobserve 'whether the thermiceffect ofglucose/insulin infusionswerechanged withtime.
Duringeache'xperiment,continuousrespiratoryexchange measu're-mentswere performedfor the duration of thetest(baseline and test) usingaventilatedhood,open-circuit,indirectcalorimeter,thedetails of which have beendescribed elsewhere(21).
Heartratewasrecordedthroughoutthetestbymeansofaportable, lightweight, electronic device(heart rate memory, Difa BeneluxBY,
Breda,TheNetherlands).
Bloodsamplesweretaken every 5min afterthebaseline forglucose analysis. Samples werealsotakenatregularintervals (Fig. 1)for free fattyacid,bloodureanitrogen,and hormoneanalyses.
Urinewascollectedatthebeginningand end of thetestandanalyzed
forglucose, nitrogen, and catecholamines.
Analyses
BloodglucosewasanalyzedinduplicateonaBeckmanIIglucose analyzer (BeckmanInstrumnents,Inc.,Fullerton, CA).Plasmasampleswerealso
analyzedfor insulin(22), C-peptide (kit,Byk-Mallinckrodt, Diezenback,
FederalRepublico'fGermany)andglucagon (23) byradioimmufioassay,
freefattyacids(24)onthe Doleextract(25),andcatecholaminesbyhigh performance liquidchromatography(26, 27).
Urinarynitrogenwasanalyzed,afterdigestion,on aTechnicon
au-toanalyzer (Technicon Instruments, Corp., Tarrytown, NY) (28),and
urinary catecholamines were measured by fluorometry (29). Urinary glucosewasa'nalyzedwithglucoseteststicks(Gluketur-test, Boehringer, Mannheim, Federal Republic ofGermany). Glucosureawasonly ob-tainedinonetestandwasfurtheranalyzedinduplicateontheBeckman IIglucose analyzer.
Data
analysis
Fordatapresentation,themeanvaluesobtainedduringthe last 30 min ofeach relevantphaseineachprotocolwereconsidered.IntestsIlaand
lb,phase 1 representsthehyperinsulinemic (-.80
jtU/ml)
euglycemicclam- an phAs21- th hyernulnei hyegye ic lamp (-' 171
mg/dl). In tests 2a and 2b, phase 1 represents thehyperinsulinemic, Teat la
I
hyperglycemic (- 170 mg/dl) clamp, and phase 2 the "euglycemic" ( 101 mg/dl)clampwithastep increaseininsulinemia (insulinemia
850,uU/ml).
Anindexofproteinoxidationwascalculatedfromthemeanurinary
nitrogenexcretionduringthe test, after correctionfor changesinnitrogen
in theureapool(30).
"Glucosestorage"ornonoxidative glucosedisposalwascalculated
bysubtractingtherateofglucose oxidationfrom therateofthe corrected
glucose uptake(M),i.e.,bytakingintoaccountthechangesin theglucose poolassumingadistribution volume of 0.21 liter/kgperbodywtand
subracing
anyurinaryglucoselosses.Urinary glucosewasonlyobservedoncein Isubject,duringtest2b and amountedtoanexcretionrateof 0.01 mg/kgper min.Hepatic glucoseproductionwasnotmeasured in the presentstudyandwasassumedtobetotallysuppressedduringthe glucose/insulininfusions (31, 32).
Thethermiceffectofglucose/insulin
(TEe)
wascalculatedbydividing theincrease in energyexpenditure (EE)(EEp.
I-EE.)bytheenergy content of the increase in glucose uptake (GU)
(GUp1,,
I-
GQU,.tmd).
Testslaand 2a
TEG,=
EEp
-EE,tw(GUpI
-GUw)X3.74X 100. Thethermic effect ofglucosealone(TEG)wascalculatedasfollows:Test Ja
TEG
=EEp
2- EE1/(GUphm
2-GUp,,
1)
X 3.74X100,
where EE is thesteadystateenergyexpenditure kcal/min,GUis the glucoseuptakeg/min,and3.74 is the energy value Igglucosekcal/g.
Statistical analyses were performed using Student's paired t test. Stepwise regression analysiswasused to determine which parameter change (glucose uptake, plasma concentration of insulin, norepinephrine, andepinephrine) correlated with each change in energy expenditure. Results are expressed as mean±SEM unless stated otherwise.
Results
Theaim of the presentprotocols wereintest la, to maintain insulinemiaat a constantrateduringtwodifferent steadystate
conditions ofglucose uptake and intest2atomaintainaconstant
glucose uptakeduringtwodifferent steadystate conditions of insulinemia. The results of the blood parameters and heartrates
measuredduring the baseline and the different phasesofeach experimentaresummarized in TableII.
Testla. Itcanbeseenthat with somatostatin infusion there
was a slight butsignificant decrease in both glycemia and in-sulinemia between the baseline and somatostatin phases after which blood glucosewasmaintainedat91 ±1 mg/dl when plasma insulinwas80±4
AU/ml
(phase 1). As expected, C-peptidecon-centrations fell dramatically with somatostatin and decreased furtherto 0.18±0.01 ug/ml in phase 2atwhich time plasma insulin concentrations were 71±3 uU/ml.
Test lb. Glycemic values were similartothose in test la. Withoutsomatostatininfusion, C-peptideconcentrationswere moreelevated than in test Ia, especially during the hyperglycemic
TableI. ChangesinthePrincipal Blood Parameters andHeart Rate intheTestswith and withoutSomatostatin(n= 8) (Mean±SEM)
Baseline Somatostatin Phase I Phase 2 Baseline Phase 1 Phase 2 Parameter (-304 min) (0-30min) (90-120min) (180-210min) (-30-0min) (60-90min) (150-180 min)
Test ia(withsomatostatin) Testlb(without somatostatin)
Euglycemia followed by hyperglycemia
Glucose(mg/dl) 99±1* 95±2 NS
91±1t
171±2 101±2* 91±1* 170±3Insulin
(,uU/ml)
11.2±0.9§ 6.7±1.0t 80±4t 71±312.0±0.9t
81±7t 108±9C-peptide(ng/ml) 1.54±0.07t 0.59±0.02t 0.23±0.01t 0.18±0.01
1.74±0.2t
0.72±0.13t
3.70±0.30Glucagon(pg/ml)
100±12t
46±6 NS 44±8 NS 43±5 81±9 NS69±12t
53±10Freefattyacids 249±25t 344±47t 128±9t 112±11 268±16t
135±8t
118±6(Amollliter)
Norepinephrine 209±16NS 204±18 NS 229±19 NS 234±19 205±10§ 232±18 NS 242±13
(pg/ml)
Epinephrine(pg/ml) 74±5 NS 72±3NS 79±5 NS 82±6 71±3 NS 76±4 NS 77±3
Heartrate(beats/min) 67±3 NS 65±2NS 69±3* 73±3 68±2t 74±2* 76±2
Test 2a(withsomatostatin) Test2b(withoutsomatostatin)
Hyperglycemia followed byhyperinsulinemia
and euglycemia
Glucose(mg/dl) 101±2* 97±2* 170±2* 101±2 100±2t 171±1t 122±3
Insulin(AU/mlm 12.8±1.0t 6.0±1t 95±7t 848±121 10.9±1.6t 128±8* 843±43
C-peptide
(ng/mO
1.62±0.10t
0.55±0.04t 0.25±0.03§ 0.02±0.01 1.56±0.13t 4.40±0.21t 2.82±0.24Glucagon(pg/ml
82±11t
41±6§ 31±4NS 36±7 86±9t 46±7 NS 44±5Freefatty acids
(;tmol/
280±18* 439±46t 119±14 NS 111±12 259±30t129±9*
114±7liter)
Norepinephrine 208±15 NS 204±10 NS 207±15* 244±12 213±21 NS 225±17t 274±25
(pg/ml)
Epinephrine(pg/ml) 76±4 NS 75±4 NS 75±5* 84±4 76±5 NS 81±6t 91±6
Heart rate(beats/min) 63±2NS
62±1t
67±2t
73±265±2t
74±2t 80±3*<0.01. t<0.005. § <0.05.
phase(3.7±0.3 vs. 0.18±0.01 ng/ml, P<0.001). Plasmainsulin concentrations were similar during theeuglycemic phase (phase 1) in tests la and Ib, but was much higher during the hypergly-cemicphase (phase 2) in test lb than in test la (108±9 vs. 71±9 ,uU/ml,respectively, P <0.001). Thesomatostatin infusion in test lathuseffectively inhibited endogenous insulin secretion
during hyperglycemia.
Test2a. During phase 1 (hyperinsulinemic hyperglycemic clamp)the plasmaglucoseconcentrationreached asimilarvalue tothatinphase 2 oftest la (170±2 vs. 171±2 mg/dl,
respec-tively).Attheendofphase2, thehighinsulin infusionrate(6.45 mU/kg per min) induced a very high insulinemia
(848±121,uU/
ml)withaconcomitant decrease in plasma glucose concentration towardseuglycemia (101±2 mg/dl). Plasma C-peptideconcen-trationwasmarkedly reduced by thesomatostatin infusionfrom 1.62±0.1 to 0.25±0.03 and 0.02±0.01 ng/ml during baseline, phase 1 and phase 2, respectively.
Test2b. Similar glycemicconcentrationswere obtainedin test2b asintest2a, whileinsulinemiawashigher duringphase 1 oftest2b (128±8 vs. 95±7
,uU/ml,
P<0.001 in tests 2b and 2a,respectively). Plasma C-peptideconcentrationsalsoincreasedduring hyperglycemiabut weresignificantly inhibitedatthehigh insulinconcentrations (843±43 ,AU/ml) obtained in phase 2.
Plasma
concentration of norepinephrine and epinephrine.Plasma norepinephrine concentrations in tests 2 were signifi-cantly increased during thehyperinsulinemic phase (phase 2) when comparedwith values in phase 1 (244±12 vs. 207±15 pg/ ml, P<0.01; 274±25 vs. 225±17 pg/ml, P<0.005,in tests 2a and2b, respectively). Similarly,plasmaepinephrine
concentra-tionswere slightly increasedin tests 2aand 2bduringthe
hy-perinsulinemic phase.
Plasma concentration of glucagon. The plasma glucagon
concentrationwaslowered bysomatostatininfusionsin tests la
and 2a. In test 2b, plasma glucagon concentrations were
de-creasedfrom 86±9 (baseline)to46±7pg/ml (phase 1; P<0.005)
following
the stepincreaseininsulinemia.Glucose uptake, glucose oxidation, andglucose storage. The
changes inglycemia,glucose uptake (M), glucose oxidation, and
glucosestorage areillustratedin Fig. 2 for tests laand 2a. During thebaseline
period
glucoseoxidation was0.136±0.020 g/min(test la) and0.112±0.021 g/min (test 2a). Lipid oxidation at
this timewas0.061±0.007 g/min in both tests. With
somato-statin,
glucose oxidation fellto0.110±0.017 g/min, P<0.02,(test la)and0.098±0.015g/minNS (test2a)eventhoughglucose
uptake was 0.093±0.018 g/min and 0.094±0.017 g/min,
re-spectively.
Duringthe last 30 minof thehyperinsulinemic (80±4
,U/
ml),euglycemic clamp (test la)glucoseuptakewas0.469±0.039E 12 .0
n
2 0.8
0 co
0
05
0.4,
O-L
-60 0 60 120 180
Time(min)
E
.L
0
D
V
0
-60 0
g/min, of which 0.255±0.021 g/min could be accounted for by oxidation and the remainder, 0.214±0.036 g/min, by glucose storage. At this time lipid oxidation had decreased to almost half the postabsorptive value 0.034±0.006g/min.
Astepchange in the rate ofglucose uptake from 0.469±0.039 g/min to 1.069±0.086 g/min resulted in an increase in both glucose oxidationto0.385±0.029 g/min (P < 0.005) and storage to 0.684±0.077 g/min (P < 0.001), while lipid oxidation de-creased to negligible values.
During the last 30 min of test 2a phase 1, where hypergly-cemia wasmaintainedat170±2mg/dl ataninsulinemia of 95±7
AU/ml
glucose uptake was 0.884±0.101 g/min, one-third of which was due to oxidation (0.284±0.029 g/min) and 0.600±0.086 g/min to glucose storage. The step change in plasmainsulin concentration from 95±7 to 848±121
AU/ml
(test 2a, phase 2) caused a slight but significant increase in glucose uptake from 0.884±0.101 to 0.926±0.102 g/min, P < 0.001, which was due toafallin glucoseconcentrationin the glucose space. Thedisposal of thissmallincreasein glucose uptake could be entirely accounted for byanincrease inglucose oxidation since glucose storage remainedunchanged (0.570±0.080 g/min, NS).
During both clamp phases lipid oxidation decreased signif-icantly (0.020±0.004 g/min, P<0.001 and 0.006±0.006 g/min, P < 0.001, phases 1 and 2, respectively) when compared with the baseline (0.061±0.007 g/min) and somatostatin
(0.067±0.004 g/min)phases.
Table III presents the resultsforglucoseuptake,oxidation, and storageobtainedin the twotestswithout somatostatin. In theseconditions, although glucose uptakewasslightly greater, thecontribution of oxidation and storagetoglucose uptake was the same(within 4%) in each of the different phasesas in the
testswith somatostatin. Wheninsulinemiawasincreased from 128±8 to 843±43 ,uU/mlglucose uptakeincreased slightly and wasaccompaniedby asignificantincrease in
glucose
oxidation.The changes in energyexpenditure, glucose uptake and in-sulinemia during the twosomatostatin tests are illustrated in
Fig. 3. Glucose uptake has been divided into itsconstituents, oxidation,and storage. Itcanbeseen(test la) thatanincrease in both glucose uptake andinsulinemia of0.377±0.044g/min
and 73±4
ALU/ml
(phase 1 minus somatostatin), respectively,causedanincreasein energyexpenditure of0.08±0.02 kcal/min or athermiceffect of5.9±1.2%. When therateofglucose uptake
wasincreased by a further 0.599±0.07g/min, energy expenditure also increased by 0.14±0.03 kcal/min, which resulted in a thermic effectofglucosealone of 5.9± 1.1%.
Intest2a, theincreasein bothglucoseuptakeandinsulinemia
of0.791±0.110 g/minand 90±7 ,uU/ml, respectively, caused
anincrease in energyexpenditure of0.17±0.03kcal/min.Inthis
TEST 2a
plasma glucose (mg/d)|
1180
1140
100
60 Figure2.Changeinplasma glucose
.*,glucose
uptake
and its compo-C 9]20nents,
oxidationu,andstorageoduring20 thetwo testswithsomatostatin (n=8:
60 120 180 Mean+SEM).
Table III. Glucose Metabolism and Metabolic Variables during the Two Testswithout Somatostatin (n 8,Mean±SEM)
Baseline Phase I Phase 2
-300 min 6090 min 15S-180 min
Test Ib: Euglycemia followed by hyperglycemia
Glucose uptake(g/min) 0.524±0.040* 1.168±0.094
Glucoseoxidation(glmin) 0.116±0.017t 0.281±0.088t 0.378±0.024
Glucose storage(g/min) 0.243±0.026t 0.790±0.088
AEnergy expenditure(kcal/min) 0.14±0.03* 0.27±0.03
Thermic effect (%) 7.1±1.4 NS 6.0±0.5
Theoretically derived thermiceffect"l(%) 2.5±0.2 3.6±0.1
Lipid oxidation
(glmin)
0.066±0.006§ 0.032±0.008§ 0.006±0.004Test 2b: Hyperglycemia followed by hyperinsulinemia
Glucose uptake(glmin) 1.076±0.088 NS 1.097±0.090
Glucose oxidation(glmin) 0.137±0.013t 0.360±0.023t 0.391±0.020
Glucose storage(glmin) 0.716±0.075 NS 0.705±0.075
A Energy expenditure(kcal/min) 0.25±0.02* 0.31±0.03
Thermic effect (%) 6.5±0.2§ 7.6±0.3
Theoreticallyderived thermic effect (%) - 3.5±0.1 3.4±0.1
Lipid oxidation (g/min) 0.058±0.004t 0.009±0.006 NS 0.003±0.005
* <0.005. t<0.01. § <0.05. 11The theoretically derived thermic effect=glucose storageX5.3 + glucose uptake, where 5.3% is the thermic
effect of glucose storage as glycogen (3).
condition glucose storage was 0.605±0.099 g/min, and the
thermic effectwas5.8±0.5%,whichwasthe same as that observed in phase 2 of test la. When glucose uptake and storage was
maintainedconstant andinsulinemia was increased by 752±115 ,uU/ml, energy expenditure increased by 0.05±0.02 kcal/min (P
<0.025).
Stepwise regression analysiswasperformedonthechanges
in glucose uptake, plasmainsulin, and catecholamine concen-trations,in the somatostatintests todeterminewhichparameters most influenced the observed increase inenergy expenditure
(Table IV). It can beseenthatintestla,astepchangeinglucose uptake correlated
significantly
with the increase in energy ex-penditure (r=0.823, P<0.01)and accounted for 68%(r2)
of this response (P<0.025).The correlationimproved slightlyandremained significantwhentheincreaseinplasma epinephrine concentrationswereaddedtotheequation(r=0.837,P<0.05). In test 2a whereglucose uptakewasmaintainedessentially constantand plasmainsulin concentrationswereincreased by
752±115 uU/ml, the change ininsulinemia didnotcorrelate
withthechange inenergyexpenditure. Althoughthe increases in both plasmanorepinephrineandepinephrine concentrations
TEST la
E
i-aco
a5
a
w
aGIucoseUptake
Q61.2 (g/min)
I
OA4 1JD
L020.4
iI
0
AIRII*ooO
(pu/mi)1
800
1600
1400
1200
10
correlated with energy expenditure (r = 0.745 and0.689, re-spectively, Fig. 4) only norepinephrine correlated significantly when analyzed bystepwiseregression analysis (P<0.05).
Table V presentsboththemetabolicand blood parameters ofthe tests where eithereuglycemiaorhyperglycemiawas main-tained continuously for3 h.These resultsdemonstrate thatonce
steady state conditions were attained, no further changes
oc-curredinanyofthemeasuredparameters, except in the
eugly-cemicclamp,where bothplasmainsulin and freefattyacid con-centrations fellslightlybutsignificantly.
Heartrate.Fig.5 presents thechangesinheartrateduring
theexperimentswith and without somatostatin. In the
postab-sorptivestate, theresting supineheartratewas66±1beats/min
in all tests. Whensomatostatinwasinfused,therewas a
transitory
fallthatwas
significant
(68±3 to65±2beats/min,
P<0.02in test 1;63±2to62±1beats/min,
P<0.05 intest2)and lasted- 10-15min.Duringthedifferentclamp phases,the heartrate
increased significantly intratest. Although the heartratetended tobe lowerwithsomatostatinintest 1,thedifferenceswere not
significant.
However,intest2,the differencesweremoreevident andtheyweresignificantly
different.TEST 2a
Figure 3. Changes in energy
expen-diturem, glucose oxidation o, glu-cosestorage o, and plasma insulin concentrations a, over the somato-statin baseline values during phaseI ontheleft and phase2onthe right of each of the tests in which somato-statinwasinfused.Notethat glucose uptake is the sum of glucose oxida-tion +glucose storage (n=8;
mean±SEM).
TableIV.Simpleand Partial CorrelationCoefficients by Stepwise Regression Analysis Between
theChangesin EnergyExpenditure (Dependent Variable) and the ChangesinGlucose Uptake,
Plasma Insulin, Norepinephrine and Epinephrine Concentrations (Independent Variables) (n = 8)
Dependentvariable A energyexpenditure kcal/min
Independentvariables Simplecorrelation Partial correlation P
Test la:Steady state insulin concentrations with a step increase in
glucose uptake (phase 2-phasel) \
A Glucose uptake(glmin) 0.823 <0.025
A Plasma epinephrine(pg/ml) 0.450 0.837 <0.05
APlasma norepinephrine
(pg/lm)
-0.583 0.856 NSTest 2a:Steady state glucose uptakewith astepincrease inplasma
insulin concentration (phase2 -phasel)\1
APlasma norepinephrine(pg/mI) 0.745 <0.05
APlasma epinephrine(pg/mI) 0.689 ) 0.754 NS
A Plasma insulin(uU/ml) 0.333 J 0.722 NS
Urinarycatecholamines. During test1aurinary norepineph-rine and epinephrine excretions were 1.67±0.45
gg/h
and 0.35±0.18,ug/h,
respectively, and were notsignificantly different from the excretion rates observed in test 2a (1.49±0.41ug/h
norepinephrineand 0.40±0.28ag/hepinephrine). In the control testswithoutsomatostatin the rate of norepinephrine excretion was1.60±0.45 and 1.92±0.37
,g/h
(testlb and 2b, respectively) andthatof epinephrine was 0.39±0.23 and 0.41±0.15,ag/h
(testlb and2b, respectively).
Discussion
This
study
investigated
theinfluence ofglucose
metabolismper seandinsulinper se uponthe thermiceffectof glucose/insulininfusions. Theresults demonstrate that a step change in glucose
0-12
0.10
c008
E
"I
006
a004
002
0 l
- * 0
0 * 0
0 *
0
I
,,o'
-10 0 10 20 30 40 50 60 70
Aplasma Norepinephrine pg/ml
-5 0 5 10
A plasma Epinephrine pg/mi
15 20
Figure 4. Correlations between the increase inenergyexpenditure (mean kilocaloriesperminute)and thechangesinplasma catechol-amine concentrations withastepincrease in insulinemia of 752±115 ,uU/ml(test2a,phase2minus phase 1).Simplelinearcorrelation
analysisshowed that bothnorepinephrine (* *, y=0.00 lx + 0.008;r=0.745,P<0.05)andepinephrine
(-
-- -o,y=0.004x +0.015;r=0.689,P<0.05)weresignificantlycorrelated withmeankilocaloriesperminute. Withstepwiselinearcorrelationanalysis only the increase inplasmanorepinephrineconcentrationswas
significantly
correlated withmeankilocaloriesperminute and couldaccountfor 57%(r2)of this increase (n=8).
uptakeishighlycorrelated with acorrespondingchangein energy
expenditure. Althoughastepchangeininsulinwasaccompanied byaslightincrease inenergyexpenditure,the latter was
signif-icantlycorrelated with aconcomitant rise inplasma catechol-amineconcentration rather than with the change in insulinemia. Theinfluence of insulinemia upon plasma norepinephrine concentrations supports the observations of Rowe et al. (33) that
hyperinsulinemia, rather than hyperglycemia, stimulates
sym-patheticnervousactivity.The fact that stepwise linear regression analysis showsasignificant correlation between the increasein
plasmanorepinephrine and that of energy expenditure suggests that, when present inlarge concentrations, insulin can cause a
small increase in energyexpenditure that is mediated via the sympatheticnervoussystem.Such a role for insulin has already beensuggested(34), possibly acting via receptors that have been found in theventromedialhypothalamus(35). Insulin
concen-trations of 840 ,U/ml are rarely or never observed, even in
pathological conditions. It is thus likely that physiological changes in insulin concentrations can account for onlyavery smallproportionof the observed change in energyexpenditure, i.e.,0.05kcal/minper 752
,U
per mlinsulinor0.007kcal/minfor every 100 gU/ml increase in insulin concentration ifone assumes alinearrelationship between the riseinplasmainsulin concentrationand theincreasein energyexpendituremediated via thesympathetic nervous system.
Flatt has calculated the energy cost of storing glucoseas gly-cogen basedontheratio of the numberofmolesof ATP used
to those produced during glucoseoxidation andglucose
con-version to glycogen, i.e., 5.3% of the energy content of glucose
or0.2 kcal/g (3). Bycomparingthethermic effects of
glucose/
insulin,glucosealone, and insulin aloneobtainedexperimentally
with those calculatedasabove, it ispossibletoobtainanindex of thecontributionof theobligatoryand facultativethermogenic
components(36) duringthedifferentphasesof the
study.
Theenergy cost of glucose storage (obligatory
thermogenesis)
ac-counted for -50-60% of the measured thermic effect(Tables III and V) when no somatostatin was infused and 60% with
somatostatin infusionduring both hyperglycemic phases (test la,phase 2 andtest2a, phase 1).With astep increase in
insu-linemia, thecontribution of the facultative
thermogenesis
in-creasedfrom 42to51%with somatostatinand from 46to55%
withoutsomatostatin,demonstratingthatinsulinand/or
Table V.Blood Parameters andMetabolic VariablesduringEitherConstantEuglycemiaorHyperglycemia (n= 3, Mean±SEM)
Euglycemia Hyperglycemia
Baseline Phase I Phase 2 Baseline Phase 1 Phase 2 (-30-0 min) (60-90min) (150-180min) (-30-0min) (60-90min) (150-180 min)
Glucose uptake(g/min) - 0.505±0.026 NS 0.536±0.035 - 1.195±0.177 NS 1.178±0.152
Glucose oxidation
(glmin) 0.146±0.051* 0.286±0.037 NS 0.289±0.026 0.137±0.033* 0.407±0.059 NS 0.403±0.042
Glucosestorage
(glmin)
0.220±0.053 NS 0.247±0.061 - 0.788±0.135 NS 0.775±0.121AEnergy expenditure
(kcal/min) 0.17±0.05NS 0.15±0.06 0.29±0.05 NS 0.28±0.07
Thermic effect (%) 8.9±2.2 NS 6.9±2.4 6.7±0.9 NS 6.3±1.3
Theoretically derived
thermiceffectt(%) 2.3±0.5 2.7±0.2 - 3.4±0.2 3.4±0.1
Lipidoxidation(g/min) 0.048±0.014 NS 0.028±0.009 NS 0.024±0.016 0.061±0.014 NS -0.002±0.014 NS 0.007±0.001
Glucose(mg/di) 95±1 NS 89±2 NS 93±3 98±0.4§ 170±1 NS 171±2
Freefatty acids
(Amol/liter)
273±76NS 142±14§ 130±13 267±37 NS 135±16 NS 133±19Insulin(,U/mi) 12.6±1.8§ 84±6§ 78±7 12±2§ 128±22 NS 137±18
C-peptide (ng/ml) 1.58±0.17§ 0.72±0.16 NS 0.58±0.12 1.81±0.21§ 5.30±0.68 NS 5.88±0.74
Glucagon(pg/mI) 89±47 NS 63±28 NS 61±25 72±21§ 35±15 NS 33±13
Norepinephrine(pg/mi) 205±7§ 252±16 NS 239±7 233±30 NS 253±23 NS 242±30
Epinephrine (pg/ml) 64±10§ 75±11 NS 72±10 76±8 NS 84±7NS 77±7
Heart rate(beats/min) 60±4§ 69±4NS 69±5 61±3§ 70±5 NS 72±5
*<0.05. t Thetheoreticallyderivedthermic effect="glucose storage"X5.3 . glucoseuptake;where5.3 is thethermiceffectof
glucose
storageasglycogen (3); § <0.005.
mediated factorscausedasmall increase in energyexpenditure
thatwasnotrelatedtoglucoseuptake.
Itisinterestingtonotethat the increase in energyexpenditure overthesomatostatin baseline duringthehyperglycemic, high
plasmainsulinclamp (test 2a, phase 2) was 0.23±0.03 kcal/min
at aglucoseuptake of 0.926±0.102 g/min of which 0.05±0.02 kcal/minor 22% can be accounted forbyinsulinand/or stim-ulation of the sympathetic nervous system. If one assumes,as
observed elsewhere(37), thatacertainsympatheticcomponent
totheincrease inenergyexpenditurewasalreadypresentbefore
the step increase in plasma insulin concentration, then these resultsconfirm ourpreviousobservations (36, 37) that stimu-lation ofthesympatheticnervous system canaccountfor - 30%
of theincreasein energyexpenditure during glucose/insulin
in-fusions.
Although it would appear thatinsulinviastimulation ofthe
sympatheticnervoussystemcanaccountfor20-25%ofthe
fac-TESTS la and lb
ultativethermogenesis (36)it is possiblethat otherfactorsmay beinvolvedbutwhichcannotbedetermined from ourdata.
In this study somatostatin wasused as a tool to suppress
endogenous insulin secretion. Itis unlikely that it influenced ourresultssubstantially sincetests lband2bcarriedoutwithout somatostatingavesimilar resultsof glucose metabolism.
Certainaspectsofsomatostatin actionswereobserved,
how-ever. Somatostatin infusion causedadecrease in bloodglucose concentration, which necessitatedtheinfusion of glucoseat a
rateof1.3 mg/kg per minin bothtests, tomaintain euglycemia.
Itiswelldocumented thatthis decreaseinglycemia is principally
duetoinhibition ofglucagon-stimulated glycogenolysisby so-matostatin, whichresults ina50-80%reduction inhepatic
glu-coseproduction (38, 39). Although hepatic glucose production
wasnotmeasured in the present study, one can assume a basal
hepatic glucose production of -2.2 mg/kg per min (40, 41);
somatostatin thus caused a 60% decrease in hepatic glucose
TESTS 2a and 2b
co0
$
-60 0 60 120 180
Time(min)
-60 0 60 120 180
FigureS.Change inrestingheartrate
with(- *) and without somatostatin (o o)infusionduringthe twoclamp protocols (n=8; mean±SEM).
Thermic EffectofInsulin 1753 5)
(a
cc
0
0)
production in our study consistent with the results of others (38, 39).
During the last 30 min of the hyperinsulinemic euglycemic clamp with somatostatin, glucose uptake was somewhat lower (6.5±0.4 mg/kg permin) than during thesamephase without somatostatin (7.2±0.5 mg/kg per min), but the differencewas
notsignificant. While otherstudieshavereported decreased pe-ripheral glucose utilization after long periods ofsomatostatin infusion (38, 42), they have beenquestionedby Bergman et al. (43) who observed an increased glucose clearancewith
soma-tostatin infusion andinsulin in dogs. They proposed that
so-matostatinincreases insulin receptor sensitivity to insulin but
theirrateof somatostatininfusionwasapproximately eight times
the rate used in the present study. Such large doses could cause nauseaandabdominal pain in humansubjects(44).
Theinitiation of somatostatin infusionnotonlyinfluenced
hepatic glucose production but other metabolicparameters as
well. Transient falls were also observed in metabolic rate and glucose oxidation, conversely fat oxidation rose concomitant withanincrease in plasmafreefatty acids. The latter was
prob-ably duetoincreasedlipolysisby the removaloftheinhibitory effect of insulin. Whether the fall in glucose oxidation was a resultof reduced hepatic glucose production (38, 39), inhibition
ofglucoseoxidationby the glucosefatty-acid cycle(45)orboth
cannotbedetermined fromourdata.
Theinfluenceof increasing insulin concentrations on plasma
catecholamine concentrationshas beenmentionedabove.Like
Rowe etal.(33), wefound that these hormonal changes corre-latedwellwithourheartrate measurements. An 11% increase
incirculating plasmacatecholamine concentrationwas
accom-panied byan 11%increasein heartrate.Itis of interestto note
that when thesomatostatin infusionwasinitiated, therewas a significantbuttransient fall inheartrate(seeFig. 5)thatlasted
10-15 min. This effect of somatostatin on the heart rate has beenobservedbefore (46, 47),butnosuggestionsweremadeto explainthisphenomenon. Since somatostatin inhibitspancreatic insulinandglucagonsecretion, thisdecrease in heartratecould be duetoinhibition ofthe positive inotropic action of insulin (48) and/orthepositive inotropic and chronotropic action of glucagon (49). Butthiswouldnotexplainthetransientnature ofthe response. Itisalsopossiblethatsomatostatinmight influ-ence the sympathetic nervous system. Somatostatin or soma-tostatin-likepeptideshave beenfoundin thebrain(50),andit has been suggested that somatostatin and endorphin neurons interacttomodulatesympathetic outflowtothe adrenal medulla
(51). Thesameauthorshave not,however, observed similar ef-fects of somatostatin onperipheral
sympathetic
nerveendings,
norhasit beenfound to decrease plasmaepinephrine or
nor-epinephrineconcentrations (52); this is in agreement with the resultsof this study.Assuming that the slight decrease in heart
rateobservedwith somatostatininfusion represents a change in
sympatheticnervousactivity, this is ofatransitorynaturesince the
resting
heartratewasreestablished before the start of theinsulin infusion.
Inconclusionourresultsdemonstrate thatalargeproportion
of the thermic responseto glucose/insulin infusions is dueto
glucose metabolism alone. The contribution of insulin persein the stimulation of energyexpenditure is small and appearsto bemediated via thesympatheticnervoussystem, thusat phys-iologicalinsulinconcentrations,thethermogeniceffect of insulin
per seis negligible.
Acknowledgments
The authors wouldlike to thank Mss. D.Kock,E.Rossi,E.Temler,D.
Penseyresfortheir technical assistance andMs. J. Braissant fortyping
themanuscript.We would alsolike to thank Serono S.A CH- 1170
Au-bonne, Switzerland,for thegenerous giftof somatostatin(Stilamin)and
theNESTLECo.,Vevey,Switzerland,for its financialsupport.
References
1.Acheson, K.J.,Y. Schutz,T. Bessard, E.Ravussin,E.Jequier,
andJ. P. Flatt. 1984. Nutritional influences onlipogenesisand
ther-mogenesisafteracarbohydratemeal. Am. J.Physiol.246:E62-E70.
2.Thiebaud, D.,Y.Schutz,K. J.Acheson,E.Jacot,R. A.DeFronzo, J. P.Felber,and E.Jequier.1983.Energycost ofglucose storagein man
during euglycemicinsulinclamp.Am.J.Physio.244:E216-E221.
3.Flatt,J.P.1978. Thebiochemistryofenergyexpenditure.In Recent
advancesinObesityResearch. Vol. II.G. A.Bray,editor.Newman
Pub-lishing,London,211-228.
4. Pittet, Ph., Ph.Chapuis, K. J.Acheson,F. deTechterman,and
E.Jequier. 1976.Thermic effect ofglucoseinobesesubjectsstudiedby
direct and indirectcalorimetry.Br. J.Nutr. 35:281-292.
5.Golay,A.,Y.Schutz,H. U.Meyer,D.Thiebaud,B.Curchod,E.
Maeder, J.P. Felber,and E.Jequier. 1982. Glucoseinduced
thermo-genesisin nondiabeticand diabetic obesesubjects. Diabetes. 11:
1023-1028.
6.Kaplan,M. L.,and G.A. Leveille. 1976.Calorigenic responsein
obese and non-obesewomen.Am. J. Clin.Nutr. 23:1108-1113.
7.Schwartz,R.S.,J.B. Halter,and E. L. Bierman. 1983. Reduced
thermogeniceffect offeedinginobesity:role ofnorepinephrine.Metab.
Clin.Exp. 32:114-117.
8. Shetty, P. S.,R. T. Jung, W. P. T.James, M. A. Barrand,and
B. A.Callingham. 1981. Postprandial thermogenesis in obesity. Clin.
Sci. (Lond.).60:519-525.
9. Felig,P.,J. Cunningham,M. Levitt,R.Hendler,and E. Nadel.
1983. Energyexpenditure inobesityinfasting andpostprandialstate.
Am. J.Physiol.244:E45-E51.
10.Sharief,N.N.,and I. MacDonald. 1982.Differences indietary
inducedthermogenesiswith variouscarbohydratesonnormal and
over-weightmen. Am. J. Clin. Nutr. 35:267-272.
11.Welle,S.L.,andR.G.Campbell. 1983. Normal thermic effect
ofglucosein obese women.Am. J. Clin. Nutr. 37:87-92.
12.Jequier,E. 1983. Doesathermogenicdefectplaya role in the
pathogenesisof humanobesity?Clin.Physio.3:1-7.
13.Schutz,Y.,A.Golay,J.P.Felber,and E.Jequier. 1984. Decreased
glucose-inducedthermogenesisafterweightloss in obesesubjects:a
pre-disposingfactor forrelapseofobesity.Am. J. Clin. Nutr. 39:380-387.
14. Bogardus,C.,S. Lillioja,D.Mott,J.Zawadzki,A.Young,and
W. Abbot. 1985. Evidence for reduced thermiceffect of insulin and
glucoseinfusionsinPima Indians. J. Clin. Invest. 75:1264-1269.
15. Ravussin, E.,K. J. Acheson,0. Vernet,E. Danforth, andE.
Jequier.1985. Evidence thatinsulin resistance isresponsiblefor the
de-creased thermic effect ofglucoseinhumanobesity.J. Clin. Invest. 76:
1268-1273.
16.Rothwell,N.J.,andM. J. Stock. 1981.Arole for insulininthe
diet-inducedthermogenesisof cafeteria fed rats. Metab. Clin.Exp. 30:
673-678.
17. Felig,P. 1984. Insulin is the mediator offeeding-related
ther-mogenesis:insulinresistanceand/ordeficiencyresultsinathermogenic
defect which contributesto thepathogenesisofobesity.Clin.Physio.4: 267-273.
18. DeFronzo, R. A.,J. D. Tobin,and R. Andres. 1979. Glucose
clamp technique;amethod forquantifyinginsulin secretionand resis-tance. Am. J.Physio. 237:E214-E223.
19.Gottesman, I.,L.Mandarino,and J.Gerich. 1983. Estimation
and kineticanalysisof insulinindependentglucose uptakeinhuman
20.Sherwin, R. S., W. Tamborlane, R.Hendler, L. Sacca, R. A. DeFronzo, and P.Felig. 1977. Influence of glucagon replacementonthe hyperglycemic andhyperketonemic responsetoprolongedsomatostatin infusion in normal man. J. Clin. Endocrinol. Metab. 45:1104-1107.
21.Ravussin, E.,P.Pahud,A.Doerner, M. J.Arnaud, and E.Jequier.
1979.Substrateutilization duringprolonged exercise preceded by inges-tionof'3C-glucose in glycogen-depleted and control subjects.PJfligers Arch. Eur. J.Physiol.382:197-202.
22.Herbert, V., K. S. Lau, C. W.Gottlieb, and S. J. Bleicher. 1965. Coated Charcoalimmunoassay of insulin. J. Clin. Endocrinol. Metab. 25:1375-1384.
23.Aguilar-Parada, E.,A.M.Eisentraut, and R. H.Unger. 1969. Pancreatic glucagon secretion in normal and diabeticsubjects.Am.J. Med. Sci. 257:415-419.
24.Heindel, J. J., S.W.Cushman, andB.Jeanrenaud. 1974. Cell associatedfattyacid levels and energyrequiringprocesses inmouse adi-pocytes.Am.J.Physiol. 226:16-24.
25. Dole, V.P., and H. Meinertz. 1960. Microdetermination of
long-chainfatty acids in plasma and tissues.J.Biol. Chem. 235:2595-2599. 26. Hallman, J., L.0. Farnebo, B. Hamberger, and G. Jonsson. 1978.Asensitive method for the determination ofplasma catecholamines
using liquid chromatographywith electrochemical detection. LifeSci. 23:1049-1052.
27.Hjemdahl, P., M. Daleskog, and T. Kahan. 1979. Determination of plasma catecholamines byhigh performance liquid chromatography
with electrochemical detection: comparison with a radio-enzymatic
method.LifeSci.25:131-138.
28. Wall,L.L.,Sr., C.W.Gehrke, T. E. Neuner, R. D. Cathey, and P. R.Rexroad. 1975. Totalproteinnitrogen: evaluation and comparison of four different methods.J. Assoc.Off.Anal.Chem. 58:811-817.
29.Crout, J. R. 1961. In Standard Methods of Clinical Chemistry, Vol. 3. E.Segligson, editor. Academic Press, New York. 62-80.
30. Acheson, K. J., J. P. Flatt, andE.Jequier.1982.Glycogen synthesis versuslipogenesis after a 500 gram carbohydrate meal in man. Metab.
Clin.Exp. 31:1234-1240.
31. Kolterman, 0.G., J. Insel, M. Soekow, and J. Olefsky. 1980. Mechanismsof insulin resistance in human obesity. Evidence for receptor and post-receptordefects. J. Clin. Invest. 65:1272-1284.
32. Rizza, R.A., L. S. Mandarino, and J. E. Gerich. 1981. Dose-responsecharacteristics for effects ofinsulinon productionandutilization of glucose in man.Am. J.Physiol. 240:E630-E639.
33. Rowe, J. W., J. B. Young,K.L.Minaker,A.L.Stevens, J.Pallotta, andL.Landsberg. 1981. Effect of insulin andglucose infusionson sym-patheticnervoussystemactivityin normalman.Diabetes. 30:219-225. 34. Young,J.B., and L.Landsberg. 1979.Catecholamines and the sympathoadrenal system: theregulation of metabolism. In Contemporary Endocrinology, Vol. I.S.M.Ingbar, editor.Plenum Press, New York. 24-303.
35. Anand, B. K., G. S. Chhina, K.N.Sharma, S. Dua,and B. Singh. 1964. Activityof single neurons in the hypothalamic feeding centers: effect of glucose.Am.J.Physiol. 207:1146-1154.
36. Acheson, K.J., E.Ravussin,J.Wahren, and E.Jequier. 1984.
Thermic effect of glucose in man; obligatory and facultative thermo-genesis. J. Clin. Invest. 74:1572-1580.
37. Acheson, K. J., E.Jequier, and J. Wahren. 1983. Influence of,B-adrenergic receptor blockade on glucose-induced thermogenesis inman. J.Clin. Invest. 72:981-986.
38. Sherwin,R.S.,R.Hendler, R.A.DeFronzo, J. Wahren, and P. Felig. 1977. Glucose homeostasis during prolonged suppression of glu-cagonandinsulin secretion by somatostatin. Proc. Natl. Acad. Sci. USA. 74:348-352.
39. Wahren, J., S. Efendic, R. Luft, L. Hagenfeldt, 0. Bjorkman, and P.Felig. 1977. Influence of somatostatin on splanchnic glucose me-tabolism inpostabsorptive and 60-hour fasted humans. J. Clin. Invest. 59:299-307.
40. DeFronzo, R. A., E. Jacot, E. J. Jequier, E. Maeder, J. Wahren, and J. P. Felber. 1981. The effect of insulin on thedisposalofintravenous glucose: results from indirect calorimetry and hepatic and femoralvenous catheterisation. Diabetes. 39:1000-1007.
41. DeFronzo, R. A., E.Ferrannini, E. Hendler, P. Felig, and J. Wahren. 1983.Regulation of splanchnic andperipheral glucose uptake byinsulin and hyperglycemia inman.Diabetes. 32:35-45.
42. Sherwin, R. S., J. Wahren, and P. Felig. 1976. Evanescent effects ofhyper- and hypoglucagonemiaonblood glucose homeostasis. Metab. Clin.Exp. 25:1381-1383.
43. Bergman, R. N., M. Ader, D. T. Finegood, and G. Pacini. 1984. Extrahepaticeffect of somatostatin infusion to increase glucose clearance. Am.J.Physiol. 247(Endocrinol. Metab. 10):E370-E379.
44.Efendic, S., T. H6kfelt, and R.Luft. 1978. Somatostatin. In Ad-vances inMetabolic Disorders, Vol. 8. A.J.Szabo,editor. Academic Press, New York. 367-423.
45. Randle, P. J., P. B. Garland, C. N. Hales, and E. A. Newsholme. 1963. The glucose fatty-acid cycle: Its role in insulin sensitivity and the metabolic disturbances ofdiabetes mellitus. Lancet.i:785-789.
46.Alberti, K. G. M. M., N. J. Christensen, S. E. Christensen, Aa. Prange-Hansen, J. Iversen, K. Lundbaek,KSeyer-Hansen, and H. Or-skov.1973. Inhibition of insulin secretion by somatostatin. Lancet.
1299-1301.
47. Prange-Hansen, Aa., H. Orskov, K. Seyer-Hansen, and K. Lund-baek. 1973. Some actions of growth hormone releaseinhibitingfactor. Br.Med. J. 3:523-524.
48.Lucchesi, B. R., M. Medina, and F. J. Kniffen. 1972. The positive
inotropicactionofinsulin in the canine heart. Eur. J. Pharmacol. 18: 107-115.
49.Lucchesi, B. R. 1968. Cardiac actions of glucagon. Circ. Res. 22: 777-787.
50. Brown,M. 1981.Neuropeptides: Centralnervoussystem effects onnutrientmetabolism.Diabetologia.20:299-304.
51. Van Loon,G. R., and N. M. Appel. 1981.,3-endorphin-induced
hyperglycemia is mediated by increased central sympathetic outflow to the adrenalmedulla. Brain Res. 204:236-241.
52.Christensen, N.J., S.E.Christensen, Aa.Prange-Hansen, and K.Lundbaek. 1975. Theeffects of somatostatin on plasma noradrenaline andplasma adrenaline concentrations during exercise and hypoglycemia. Metab. Clin. Exp. 24:1267-1272.