Insulin Secretion in Protein-Calorie Malnutrition I. Quantitative Abnormalities and Response to Treatment Dorothy J. Becker, M.B., B.Ch., Bernard L. Pimstone, M.D., MRCP. John D. L. Hansen, M.D., F.R.C.P., D.C.H., and S. Hendricks, Cape Town, South Africa SUMMARY Im woreactive insulin (IRE) responses to oral and/or intravenous (IV) glucose were measured in kwashiorkor and tmarasnnus, Responses were expressed either as peak IRE Teyels attained or a5 insulimglucose ratios, Similar measure iments were made in normal conteols ‘After oral glucose, insulin secretion was abnocmally low in every patient with washiorkor and in most of the maras mie eases, Mfter IV glucose, normal insulin responses were froted more often, presumably associated with the greater Iycenie stiwulve, This was particulaely noticeable in Tn both groups of subjects there was improvement after thece to sis weeks therapy but this was much more striking after oral glicose, However, in mats patients, insulin seere- tio remained subnormel or even deteriorated at this time, When nutriionat status had greatly improved. Tico to ten months Tater howexer, insulin levels were judzed to be “The high protein turnover of the pancreas? is thought to make it especially vulnerable to protein depletion Presumably as a consequence, acinar atrophy" and im- paired exocrine pancreatic function’ occur in proveit- calorie malnutrition (PCM). There is far less clarity concerning the effects of protein depletion on the pan- creatic islee cells. Isler cell damage and poor insulin secretion have been suggested to explain the abnormal ‘lucose tolerance of kwashiorkos,** but studies reporced iin this contexe are often incomplete and confusing. From the MRC Clinical Nutrition Research Unit, Depare ment of Pediatrics and Child. Health, and the Kotope Unit, Department of Medicine, University of Cape Toxn, South Africa Requests for reprings should be addressed w De. B. L. Pim sone, Depactment of Medicine, University of Cape Town, South Africa. saz J within normal fimits in these eases Children known t0 have Iaflered from kwathiocKor ten years presiovsly likewise had normal insulin responses in 90 per cent of cases, Augmented stimulation by slucoge pls glucagon revested some panereateinslin esere i hall the untreated patients Sued. Yet the responses to. augiented testing improved Sill farther in three of ive eases after therapy nein secretion i gromlytnpaiced in hwashiorkor and D great-ayeemie stimulus or by adding glucagon to the flacore Toad. This euggesta © “luguishnes” of fnolin re Tease afer elucose under conventional conditions of teeting in there caren The. disproportionate improvement after therapy inthe insulin teaponse to afl, as opposed to 1V, flucose. nay_provide some evidence that an paired zu fetaytotrophie mechanism i parly tered eles mechaniss, Deataes 20:522551, August, 91 many eater Bowie concluded that insulin deficiency seemed unlikely fas exogenous insulia failed to increase the glucose dis- appearence rave." Daca based on the bioassay of insulin- Tike activigy (ILA)% are difficult to interpret because of the nonspecifcty of this substance which persis after panc 24 Variable levels of fasting im- ‘munoreactive insulin (IRI) in PCM®* may simply be consequence of the known inadequacy of the assessment cof insulin. secretion by basal measurements only. In) scudies of insulin responses to glucose stimulation, IRL was reporied ro have a normal bue unsustained peak," and 10 be decreased" with improvemenc after feed- ing, Deficien: insulin responses to arginine’ and amino acid’? stimulation in PCM are also recorded. However, the small aumbers of patients used in most studies and pooling of individual data have obscured differences in the pattern of insulin secretion in PCM, In some re- com. DIABERES, VOL. 20, NO. &ports insulin responses to glucose were only measured days after the introduction of protein into the diet, at a time when they may have alzeady chan mencal PCM, Heard has ced early hyperinsulin- ism'* and subsequent insulin deficiency." To further complicate the issue, insulin antagonism has been roved ®t We report here fifty-four pacients with kwashiorkor and marasmus who were studied in an attempr to quan. titate individual IRE responses to oral and inceavenous slucose. We have also assessed the response 10 an aug- meored stimulus of glucose plus glucagon’” before and after treatment, in order to relate IRI levels to the de- agree of hyperglycemia and to decermine pancreatic re- serve sed. In experi= CASE MATERIAL AND METHODS 1. Patients studied (A) Firty-roun cinupren aged eight to thin ight months sufering from PCM were admitied to the metabolic unit ofthe Red Cross Children’s Hospital with cither kovashiorkor or marasmic kwashiorkor on the one thand or marasmus with gross growth retardation on the other, Patients with kwashiorkor were underweighe with edema and hypoalbutinemia (082-259 gm/ico ml, mean 156 [© o07]), and had frequent dermatosis Marasmic childeen had weights Go per cent of less of that expected for theit age and had no edema Their serum slbumin levels ranged from 1.61-380 gim./100 ‘ml, mean 286 (0.18). All patients received a proteinfree diet with 10 gm/ ke. or more carbohydrate in the twenty-four hours prior to resting and were treated with antibiotics, potassium Chloride, vitamins and where indicated intravenous ther apy. Patients with obvious iofection, severe anemia. of 708s diarthea were excluded, (2) Incalinogenic stimuli given on admission fx PCM (Gor details see later), (i) Thicsy-nine children were given an intravenous slucose load, (iy Tweny-nine children were. given oral ghicose (iii) Foor of the above cases teceived intavenous and oral glucose loads on consecutive days in rendom order. Cie) In seven of the fifty-four children, the intial test vas followed by an augmented glacoce stimulus, in ‘which IV glucagon was given either with intravenous shucose (five cases) oF thirty minutes after oral glucose (v0 cases), depending on the initial tese done (b) Tests repeated after three 10 six weeks feeding mall but five childeen, che inital tests were repeatel svousr, srt DOROTHY J. RECKER, tm, RIE, AND ASSOCIATES alter three to six weeks of feeding at a stage of apparent clinical and biochemical reZovery. (©) Tests repeated ajter two to ten monthe Intravenous or oral glocose loading was repeated in eleven patients whose insulin responses had remained abnormal when retested after three to six weeks of feeding, Tests were performed two to ten months after the inital suadies, after careful and continued dietary supervision either aca convalescent home or at the pa. tients’ own homes, (B) A contron cRouP consisted of ten chile rea of similar age group (eighteen to thizty-seven months) whose weight was over the third Boston per- centile, All had recovered from unrelated acute infec: fon or were awaiting surgery, and had serum albumin levels signifcandy above those of our patients (3.42. 420 gm/t00 ml), However, the levels in ewo cases were below 3.5 gm/tco ml which is considered the lowest limit of normal for chis population (table 1). An oral glucose Toad was. given in five and an intras venous infusion in Sve as defined below, (C) Ten-vear Foutow-vp stupy. (a) Ten chil dren were selected ar random from a series at pres. nt being followed to assess the long-term effects of PCM, All had been treaced for an acute episode of Ikwashiorkor about ten yeaes previously and had returned to their original environment. Six of this group were below the third Boston percentile for weight. (b) Ten controls in his group were the closest siblings of the above childien. They were known not to have suf. fered acute malnutrition at any time, However, five were below the thied Boston percentile for weight. Aa intra: enous glucose load (05 gmake, as a 25 pee cent sol tion) was administered in these twenty cases after an tight to nine-hour overnight fast following a minimum of twenty-four hours of high catbohydrace intake, Informed consent was obtained from the patents of cach child in chis study U. Details of sett procedures All tests were preceded by an eight to nine-hour over- nigh fast and_basal blood Sampling for serum albumin, immunoreactive insulin (IRL) and blood sugar (a) Oral glucose load Glucose (2 gm./kg. as a 10 per cent solution) was -iven orally. Serial Blood samples were obtained by vent. puncture from an external jugular or antecubital vein at thiew-minute incervals for 150 minutes, as timed from the completion of glucose administration, (b) Intravenous glucose loed Tnrravenous glucose (1 gm/kg. of a 25 per cent dex trose solution) was infosed over two mioutes and bloodTABLE 1 Serum albumin, IRL peaks, IG ratios and K values in jonirol subjects Oral glucose Serum albumin TRE Age “(gms peak (otis) 100ml) GUL) 1G ratio cw 10 397 was sampled at 5, 20, 45, Go and 90 min, () Augmented glacose stimulus Ceystalline glucagon, x mg. (Lilly, containing 300 icrounits of insulin pet mg.) was given intravenously, either with the IV glucose or thisry minutes after oral shucose. FIG. 1. IRI" responses to. oral glucose ig trentyrnine cater of Lwashintor and orgumus, Mean level are. sig Conily” ciferent. before “and ater Tete soit weeks of therapy thirty rinses {p< .08003) and sishy Minutes (p< 0.0007) after the glu owe toed in psa UL, Methods Serum was immediately separated and frozen until IRI was measured by a modification of the Morgan and Lazarow radioimmunoassay? with a sensitivity of 2 microunits/ml. The standard and radioactive tracer wes pork insolin (Novo) and the antibody purchased from Wellcome Reagents (Ltd). All specimens of one iadi- vidual were run in the same assay, uc the variation of the assays from rua to run did not exceed 5 per cent. Blood glucoze was estimated by the Somogyi-Nelson technic and serum albumin by the Biures method.” Tnsulin-glacose ration (VG ratios) were devermined by calculating the ratios beeween the areas under the re- spective curves (as measured by planimetry) after se rial arichmetic plotting of the incremental changes of plasma glucose in. mg /100 ml. and IRT in microunes/a The glucose disoppearance rate constant (KY after 0.693, intravenous infusion was calculated as per cont per minuse wheie TY == the half disappearance time fon semilogarithmic plotting of the focal Blood glucose values? Statistics. Levels of significance were calculated by the Mana-Whitney-U tes. RESULTS L IRI and 1G ration in controls (sable 1). ‘The raage of insulin responses to oral or intravenous Oral GT. DIABETES, VOL. 20, NO. 8DOROTHY J, BECKER, MR, HEH, AND assoctarEs 40} 20 Fis. 2. IRE texpontet to IV glucote in thiry nine’ eases of Evashioror and marat: wor Me Sit woeks of therapy five minutes (p < 00007)" and twenty minutos {p © 0207) after the glucore lod slucose is shown in table 1. Except for two cases in the intravenous group, all peak insulin responses exceeded 20 microunits/ml. Mean peak value after oral. glucose was 346 (= 287) mictounits/ml. and after TV gla- cos, 338 (+ 1071) microunits/ml, The two excep- tions in the later group had abnormal K values (less than 2.0)" and the lowest LG ratios, These were n0t however associated with significantly abnormal serum albumin levels, Apare from these two cases all LG ra- Gios after intravenous glucose were greater than 1.0 and after oral glucose they were all greater than 3.0. U, IRE after oral or IV glucose in kwasbiorkor and ‘marasmus (figures 1 and 2). ‘The IRI responses to oral glucose were abnormal in all «wenry-nine patients tested. Peaks were less than 20 microunits/ml. except for thice cases. where responses Were unduly delayed and sustained. After IV glucose, eventy-xwo of thirty-nine peaks were below 20 microunits/ml. The remainder showed higher insulin levels compared to those following oral glucose, associated with the greater increment of blood glucose following the intravenous loading procedure. After three to six weeks of treatment, the over-all re- sults in Both groups showed statistically significanly improved insulin responses co the respective stimuli However, in sinceen of the forty-nine children peaks were below 20 miccounits/ml Avousr, tore IM Gr, ene Bator a I ae 35 96 Time in Minutes IL 1G ratios in kwwashiorkor and marasmus (Bguee 3 and table 2). 1G ratios were significantly lower than controls, whether the glucose load was given orally or intrave- ‘ously. The abnormalities after intravenous loading were proportionately less than after oral loading and were rhormal in half the cases of marasmus. In both groups of subjects, improvement was noted after therapy bur was much more striking after oral glucose. In twenty-two of the forty-nine patients IG ratios remained below 1.0 after IV_ glucose and below 3.0 after an oral load. Over-all improvement was less sig nificant in marasmus as compared (© kwashiorkor. Ta- deed the mean 1-G ratio after IV glucose actually de- clined after cherapy ia marasmus. = IWeInsalin reiponsei-after- augmented jnsalin stim lation (figure 4). In untreated kwashiorkor, glucegon-augmented glu- cose stimulation of IRI failed to produce a greater area under che insulin curves than achieved after glucose alone in three of the seven patients tested. This sug gests substantial impairment of pancreatic insulin re- Serve, In the other four patients some increase in the insulin area occurred. After teatment there was a further increment in the response of three of these four cases following augmented stimulation, white the other showed a marked deterioration. One of the patients sas=o Ke fs esenar Mon — eae —— 1 ora G4 etre | Ate toe aerer 7 D : Creme " a 19 : : . 100) I 2) * e = 101 A g Z “4 . *} p__| x +o] 20. 4 cP ° 23 FG. 4 wud o rent ender the eons. cae fe uci and lect pl ghcagonstmcain best ° SNS" Deiere and afer three fo se wosts of therapy FIG. 2. 1G ration in beaniror end poramus ater ral and 1G ratio in knots ond panna afer sul shiotkor and marasmus, In no instance was there an ade oi {quae insulin response co oral glucose in childsen srieh swith aa initially poor response was retested after therapy untreated Kwashiorkor. A striking feature was the lr and was found 20 have remained unchanged percentage of impaized insulin responses remaining #- V. Quantitative imsalin responses before and after normal in both groups, especilly in marasmus. In tree to six weeks of therapy (rable 3). fddition, in a few patents the initially normal values ‘ised on LG ratios, four groups of quantitative re- + actually dropped although often remaining within nor sponses were defined before and after therapy in kwa- mal limits, Again this was more noticeable in marasmus ‘TABLE 2 1.6 ration in contiols and in kwashiorkor and marasmus before and after tee to sb weeks of therapy 7 — (rat glucose load 7 Tiravenous glucose load store Alter Before ‘ater Gontole 45(= 050 1 (2 058) Rivashorkor 203029 166 (2 140 £03009 taco Navsemue BESS HOC 130) #1315083) osteo | Values in untreated kwashiorkor and marasmus signifieantly different from controls. Kwashiorkor af cose (p © O01) Marasmus after oral glucase (p< 001) TvSlues in teested kwashiorkor signfeantly diferent from those in untreated patients (p < 0.0000). [Values in trated marasmus not significantly diferent fram those in untceated patients (p > 0.03) FValues in marasmus signifeantly higher than those In hwashioskor (p < 0.02), 546 DIABETES, VOL. 20, NO.DOROTHY J. BECKER, MCA, BCH, AND ASSOCIATES TABLE 3 © patterns of insulin response to glucose (based on 1-G ratios) before and After three to six weeks therapy in PCM. Insulin seeretion Number of cases 6 ration) Kwashiorkor Marasmus Before Aer Oral Vv ont v terapy therapy slocore elscose locate slucose Normal Normal or 302) 207) 1a) Impaired Normal nes na 403) 00) Impaired Improved but 4445) sulbnormat 405 200 502) Normal Deteriorated 20) 2@ 16) 4445) “Total cases 7 6 2s 2 2 “The numbers in parentheses are per cent of Toul eee © Vi. Quenttative insulin responses after two to ten DISCUSSION sombs of therapy in patients showing persistence of abnormal insulin secretion at three to six weeks (table 4). In the eleven cases that were followed eight had im. proved insulin responses resulting in I-G ratios similar to those of the controls. The other three had peak insulins over 20 microunits/ml. after IV glucose but im. -paired glucose tolerance resulted in LG ratios lower than 1.0. VIL. Insulin secretion ten years after kwashiorkor (able 5). Nine of ten recovered cases and nine of ten controls showed peak insulin levels of over 20 microunits/ml ‘The mean I-G ratio in recovered cases (3.9 (+ 084]) ‘was noc significantly differene from that of their cone ttols (3.0 + 044). One case in each group had an 1-G ratio lower than 10. K values were low in @ num. ber of-cases without corresponding abnormalities of in- sulin secretion, Exocrine pancreatic function is extremely vulnerable to protein depletion as evidenced by marked acinar atrophy and the low basal enzyme levels and poor responses to stimulation described in children with PCM. The position regarding endocrine function is confused, as decreased granulation and atzophied,:92*-* hypertrophied*** and normal islets of Langerhans??? have been variously noted. The limitations of previous seudies of ILA and IRI have already been mentioned, Yet the bulk of evidence suggests some deficiency of insulin secretion in PCM. ‘The interpretation of the quantitative assessment of insulin secretion in environments where malnutrition is common is difficult because of the dearth of suitably ‘matched normal controls of the same racial group. Io our study, some age-matched children who appeared co have an overtly normal nutritional stars on examina tion, probably hed marginal deficiency states. Our find ing of serum albumin levels less than 3,5 gm. per cent TABLE 4 IRI peaks and I-G ratios in PCM on admission and after three to six weeks and ‘0 to fen months feeding TRE peak GU./mh) 1G ratio Oral glucose IV glucose Oral slucose IV glucose Before After Afler Before After’ After Before After After Before After After Beate 36 210 wee 362.10 treats 36 S10 ease SIGE ment wis. mths. ment «wks, mths. ment wit. mubs. «ment wie, mali RM 3 Ae 06 24124 Be Boos ae ot ia '77 IK ton 8 o 20 70 SG Bw on 07 on og cs 23 a6 03 og 4 BB 4 SS 0a) 05 20) BM Be ezie 2 ot 03 Oa Py re) 04 028 Gs woo og 3306 32. cs Sf 8 2 wa 0s 08 33) 02 os an BS. esa nae ee aueust, 297TABLE 5 Serum slbumin and peak IRI, 1G ratios and K values after IV glucose in children ten years ‘after kwashiorkor, and it contol siblings Patients 10 years after kwashioskor ‘Serum we. albumin IRE peak percentile (em /100 ml) GULF) 16 ratio K 1 310th 3.83 ea 63 Ls 2 Bra 391 96 13 27 3 0.05). 1) had insulin peaks greater than 20 4U,/ml, The «wo exceptions are considered likely to be abnormal as they are associated with K values below the lower limit of normality for this age group (2.0)2+* We have there- fore arbitracily considered an insulin peak of 20 pUl./ml, as the probable lowest limie of normal in this study. ‘The results in our controls were vastly different feom those in untreated PCM and similar co recovered cases tested two to tea months later. A group of subjects who had suffered from kwashiorkor cen years before and whose insulin responses to glucose were retested in com pany with those of previously unaffected siblings, showed higher insulin levels, However, these children were considerably older chan our subjeces with PCM and cannot strictly be considered controls. The use of pa tients as their own controls during short-term follow- up also proved difficult ro interpret as there appeared t0 be evidence that a number of such children had not attained acceptable levels of insulin secretion after three to six weeks therapy. (Discussed below) Figures 1 and 2 show quite clearly the low absotate insulin values throughout glucose colerance tests and the tendency soward improvement after therapy. However, DIANETES, OL. 20, NO. 8such an assessment rakes no cognizance of th alycemic stimulus. This is particulatly relevai because of the possibitity of malabsorption of glucose in some cases"* and sustained hyperglycemia following slucose. As changes of insulin secretion are only means ingful if the glycemic stimulus is taken into account, in- sulin-glucose (I-G) ratios were calculated. Such ratios evaluate more closely both the sata! insulin seccetion uring the course of a glucose tolerance test (not merely the peak height attained) and its relationship to the oral hyperglycemic stimulus. There ate no reports on 1G ratios in children. After IV glucose loading, the con- tol children in this study had LG ratios of greater chan 20 with the two exceptions mentioned above, After ‘oral glucose loading all five control children had IG ratios of greater than 3,0, 1G ratios were depressed in kwashiorkor and maras. ‘mus after oral glucose. The mean I-G ratio was some: what but nor significantly higher in marasmus (1.8) than kwashiorkor (0.9), in spice of similarly decreased IRI levels in both conditions. However the low IRI peaks in some cases of marasmus were probably a more adequate response to the smaller glycemic stimulus of their more normal glucose tolerance.” After intravenous ‘glucose, 1-G ratios were low in kwashiorkor, but five of rine marasmie children showed normal I-G ratios, aso- ciated in this instance with 2 greater frequency of nor- smal IRT peaks. degree of in PCM In kewashiorkor, a substantial improvement in IG ratio after oral glucose loading occurred when the nu tirional status improved after three to six weeks of feeding. After IV glucose, by contrast, though LG ra tios were somewhat higher initially, improvement afer therapy was less pronounced. Ia marasmus, the mean LG ritio following intravenous glucose was not sigai- ficancly different after this period of therapy. The poor 1G ratios after oral glucose at a time when greater re- sponses frequently followed intravenous infusion, sug. Bests a raised threshold or sluggishness of the insulin releasing mechanism, partly overcome by elevating che slycemie stimulus with IV-glacose. Some pancreatic re serve is therefore present: The disproportionate improve- ment of I-G ratios after oral glucose on treatment may ‘mean that the sluggishness is at least in part, due co a Seficiency and later recovery of the gut betacytotsophic factors.*-* This witl be fully evaluated in a future com- Some I-G ratios after oral ghucose rose quite remark« ably 10 10 oF mote, suggesting the possibility of a transicat stage of insulin hypersecretion on recovery in a few cases DOROTHY J. BECKER, ML, RCH, AND ASSOCIATES To cest pancreatic insulin reserve more efficiently, a0 enhanced stimulus was given; a aumber of patients were assessed both after glucose and glucose-elucagon admin- istration, Further insulin reserve was shown by this tech. Dic, bur ie was considerably impaired as compared 10 the responses following. identical stimulation in mose of the treated cases Table 2 illustrates four different groups of quantita: ‘ive insulin responses before and afer short-term treat ment ia both kwashioskor and marasmus. Te can be seen that no patient with kwashiorkor showed adequate insulin secretion after oral glucose. In marasmus the ap- prcently normal responses in a few subjects were a re. sult of rapid clearing of their glucose load and conse- quent normal 1-G ratios despite peek insulin levels of Jess than 20 U/ml. Most striking are the num bers of patients, particularly those sufering from maras- mas, whose insulin secretion was initially normal and deteriorated, or failed to improve after apparently effcc- tive therapy. As toca calorie depletion plays a more striking role inthe marasmic cases, ie is conceivable that the insulin required to dispose of their meager catbo- hydrate end prosein intake does not stress even an im- paited pancreatic insulin reserve, However, the e- tmendous increase in protein synthesis and energy £0- uirements during recovery must make substantial de- ‘mands on insulin production which might not be able (0 keep pace for a period. More prolonged follow-up (evo t0 ten months) in patients with early persistent abnormalities showed a retuen 10 normal of LG ratios in all bue three cases who, however, had normal peak values. This ulimate improvement is borne out by the normal inulin esponse in all but one patient tested ten years after kwwashiorkor. ‘A number of tentative conclusions can be drawn from this study. Pancreatic endocrine function is seriously dis- turbed in proteincalorie malnutrition. Persistently ab. normal insula levels are frequently found on short-term follow-up particularly in marasmus. Ths has recently been reported by James" who concluded chat impaired insulin secretion in PCM might be a permanent phenomenon However, the vast majority of patients with both maras ‘mos and kwashiorkor showed substantial. improvement months later, while a group of children knowin t0 have soffered acute PCM ten years before, likewise had nor- ral insvlin levels. Clearly impairment of insulin sere. sion is not necessarily permanent after acute PCM and js unlikely to be a major cause of adule diaberes in underprivileged populations of tropical countries*™* Insulin deficiency is probably not incriminated in the genesis of the persistently abnormal glacose colerance 349id years after acute PCM by Cook'* and in our ten: year follow-up sudy in which similae K values after 1V slucose are noted (table 5). However, our dain does ‘hot exchide the possible permanent effect on pancreatic endocrine fanetion of more persistent and severe mal nutrition dating childhood. Impairment of iasulin secretion may not be dve failuze of insulin production alone, as a degree of pan- ceeatic insulin reseeve is present although cleatly im- paired, This study does nor allow fem conclusions to be drawn regarding the possible mechanisms of the abnor smal secretion, However, diminished sensitivity of insulin release is suggested by the rather better insulin peaks and, LG ratios after the larger glycemic stimulus of intea- enous as opposed to oral glucose in our untreated case. ‘The disproportionate improvement after oral glucose when the nutritional staus has improved, suggests that fan impaired gut beracytosrophic mechanism" is a contributing factor. Many other possible factors respon: sible for abnormal insulin secretion remain ro be evali- ated, for example the role of a low amino acid intake (amino acids have been shown to participate in normal insulin release)" and the extent of insulin antagonism in selected cases. These will be the subject of commu- rications currently being prepared. ACKNOWLEDGMENT. ‘This study was supported by the US. Public Health Service Grant AMO-3995 and grants from the South African Medical Research Council, Atomic Energy Board and Wellcome Teust, The pork insulin used in the redio- immunoassay was kindly donated by Navo Laboratories, Denmark, Thanks are due to Mrs. B. MacHutchon and Miss A, Drysdale for technical assistance and to Miss A. ‘Moodie for the transport of follow-up cases. REFERENCES "Wheeler JE, Lukens, F, D. W., and Gyorgy, Pes Saties fon the localization of tsgsed methioaine within the pancreas, Proc. Soe Exp. Biol. Med. 70:187, 1949. "Trowell, H. C, Davies, J. N. P, and Dean, RF A Kwathiorkor, London, Edward Arnold Lid, 1954 "Davies J. No P.: The exential pathology of kwashiorkor Lancet #3317, 1948. ‘Barber, G. 0, and Hansen, J. D. Lt The exocrine pan seas and prorein-ilorie malnutriion, Podiaties 42:77, 1968. Bowe, J. Py Dey V. Nu and Mukerjee, Ps A preliminary study of the Biochemical changes ia starvation cases, Ind. J Med, Res. 34:145, 1946 Heard, C. RC. fects of severe proteiacalorie deficiency fon the endocrine control of carbohydeate metabolism, Dis betes 15:78, 1966. "Cook, G. C.: Glucose tolerance after kwashiorkor. Nocure 21511295, 1967 Baig, H. A. and Blowien, J. Cs Carbohydeme mecabotism in kwathiorkor. Lancet 4:662, 1965. Bose, M. Du: Ineavenous glucose tolerance in kwashiorkor mus S. A. Med. J. 381328, 1964 petimensal_ pe deficiency in Mammalian, Proin Meabotism. Vol. Il, H. N. Munro and J. B. Allison, New York, Academic Press, 1964, p. 443 Heard, C. RoC, and Turner, M. Rs Glucote tolerance tnd related factors in dogs fed diets of suboptimal prowin talue. Diabeces 16:96, 1967. © Leona, J. R: Inulinlike activity of Blood, What is 12 Fed, Proc. 181272, 1959 1 Steinke, J, Sitek, Ay Laiens, V, Lukens, FD. W., Renold, A. E, Ginsberg, M. and Hanus, Y.: Messutement of small quantves of ivvlinlike activity with rat adipose tissue UL Persistence of serum insslinlike activity after panctea- rectomy. J. Clin. Invest. 41:1699, 1962. W Hadden, D. R.z Glucose, fe faty acid and insula intr telaions in kwashiorkor and marasmus, Lancet 1589, 1967. "James, W. P. T., and Coore, H. G.: Persistent imps: rent of insulin secretion and glucose wolerance after msl- rvtrtion. Amer. J. Clin, Nutr. 23:386, 1970. G Milner, R- D. G.: Hormonal and mecaboic inertelation ships in malnutrition Pediat. Res. 4:213, 1970 "Geahum, GG, Condano, Ay Blizard, RM, and Cheek, D. Bi: Infantile malnutrition: Changes in body com: position dusing rehabilitation. Pediat. Res. 32579, 1960. W Heard, C. RC, and Hensy, F. A. J: The iasulinogenic response to inuavenous alucose ia dogs fed dies of dif: protein value. J. Endoctinol. 45:375, 1960. Ww Ryan, W. G., Nibbe, AU, and Schwa, T. B.: Bers cyotrophic effects of glucose, glucagon and tolbutamide ip man Lancet #:1255, 1967. sulin: Two antibody systems. Diabetes 12:115, 1963. 51 Somozyi, M: Notes on sugar determination. J. Biol (Chem. 195:19, 1952 = Wallson, W. Q, Coho, G., Calvaty, Ey and Ichaba, F Seudies ia serum proseins. Amer. J. Clio. Path. 18:723, 1948 Wkkos, D., and Luft, R2 On the inwavenous glucose tl cexance test Aca Endoctinol. 25:312, 1957 Loeb, HL: Vatiaions in glucose tolerance dusing infancy and childhood. J. Pediat. 68:237, 1966. 2Plac, B.S, and Scewart, R. J. C: Experimental protein alotie defcieng): Histopachologieal changes in the endocrine lands of pigs. J. Endoctinol. 382121, 1967, Stenar, Ro J. Cy and Heard, CR. C: Pancreatic ister cells and Blood sugar regulation in pigs maintained on Low protein diets. Proc. Nuts. Soc. 18:x, 1959. "Heard, CRC, Plat, B.S, and Stewart, RJ. C.: The effects on pigs of a low proscin diet with and without addi tional carbohydrate, Proc. Nutz. Soe. 17:ti, 1958 Camain, Re, and Pietchoa, M Lesions of the pancreas in kwashiorkos. Rep. 2d latee-African Conference on Nut tion. Gambia, 1952, p- 146 Campbell, J. A HL: The morbid anatomy of iafeatile malnutrition in Cape Town, Acch Dis. Child, 32:310, 1956 3° Veghelyi, P. V., Kemeny, T. T., Possony, J, and 5és, J. Diewary lesions of the pancreas. Amer. J. Dis, Child. 79:638, 1950. DIABETES, VOL. 20, NO. 8 "Moresa, CR. and Lazsrow, Av: Immunoassay of in:DOROTHY J. BECKER, MB, BCH, "Geant, D. Bs Seruminsolin levels in childen during lucose tolerance tess. Acta Palit, Scand. 57:297, 1968. © Welbor, T. A, Rubeastein, AH, Haslam, Ry and Fraser, Rs Normal insulin response to glucose. Lancet /:28, 1966, “Ehrlich, R. Mand Bambess, G.: Immunolasic aay of insulin in plasma of children, Diabetes 13:172, 1964. " Milunshy, A., Lowy, C, Rubenstein, A. H., and. Wright, A. Di: Carbohydrate tolerance, growth hormone and insulin levels in mongolism. Devel. Med. Child. Neurol, 10:25, 1968. © Buchanan, K, D., and MeKiddie, M. T. sponse to glucose, Comparison between otal and intravenous olerance fess, J. Endoctinol. 39:13, 1967, James, W. P. T Intestinal absorption in pr malnutrition. Lancet 333, 1968. © Malone, Ni, Holdsworth, C. ., and Turner, D. SA Insulin incalorie new interpretation of oral glucose tolerance. Lancet #20, 1964 O*Elrck, HL, Simmer, Ly Hla, C. Jy and Ari, ¥. Plasma insulin’ response to oral and intcavenous glucose al ministration. J. Clin, Endocrinol. 24:1036, 1964 Shaper, A. G.z Chronic pancreatic disease and protein mmalautttion, Laneet £1223, 1960. ” Incerface of protein muttition and medicine in the topics. Lancet i733, 1965, * Cook, G. C: Glucose and starch tolerances after recovery from kwashiorkor, Metabolism 17:1073, 1968. "Floyd, J. Gy Jeu Faans, SS, and Conn, J. W.: Stimula- tion of insulin secretion by amino acids. J. Clin. lavest. 43 1387, 196 © Fajans, &. S, Knopf, RE, Floyd, J. Cx Jty and Conn, J. Was The effec of amino acids and proteint on insulin ‘ecretion in man. Rec. Prog. Hosea. Res. 23°617, 1967 Ramalingsswam, V, Influence of Age and Strain on the Response of Rats to Dietary Fructose and Sucrose (Continued from page 518) the activity of hepatic glocose-6-phosphatase was assayed Ras from five diferent stains were used: t80 strains of Hooded rats 4 Sprague-Dawley stain, Lister rats, and Wistar sats, The resus demonsuated quite a marked srtin difference jn response 10 dietary sucrose Sucrose ingestion significantly depressed far synthesis from ghicose in three groups: the Sprague-Dawley, Wistar, and Lister stains, but not in the 280 Hooded sersins. Three stains (Lister, Wistar, and one Hooded stcain) showed a depressed sate of hepatic glucose oxidat and an increased activity of aluccte-Gphosphatate. Neither of these parameters was influenced by sucrose ingestion in the remaining two scvns ‘The authors interpret these observations to suppote the sus: seston that glucose oxidation by liver slices of rats fed sucrose fs raluced because of an increate in glucoee G:phosphatae ac- tivitg. One of the more significant aspects of these result is ‘hac they illustrate the fact thatthe metabolic esponses to 8 ‘rose ingestion are diferent in animals having diferent genetic backgrounds. Theve studies are consistent with earlier reports showing strain. difeences in response 1a. dietary. sucrose OM. W. Marshall and HE. Hiklebrand, J. Nowe 79-237, 1963; A. M, Durand, M. Fisher, and M. Adams, Arch, Path 85:318, 1968) ‘Thee two reports taken together are of considerable im. Avousr, 1978 portance in illustrating some aress in which crution should be observed ia animal and human experimentation related to the Ietabolic effeas of dietary carbohydrates, particalacly si rose and fructose. Ie is well recognized that individual ditfer fences exist in the human population with regard co carbo hydrate metabolism, The maay abnormalities of carbohydrate ‘metabolism which have been reported llestrate this point (Nutt. Rev, 28:153 and 204, 1970). However in dealing ith laboratory ray we often assuine that all aze similar ‘The report of Bender et al. reminds ve cha thie is not rccesasly the case, and that this needs to be considered in cererapolating results from one stetin of rat to another, and fstapolation from one species 19 another should be atempied only with extceme caution. The repose of Hill demonstrate that the age of animals is also an important facor to consider in studying metabolic responses 10 dietary carbohydrate, This observation may well explain some of the apparent incom sistencies found in the liersture, The significance of age with regard to the logenic effecs of various dietary carbohydrates jn fuman beings remains to be studied. In view of the cute rene discussions 25 t0 the possible role of dietry sucrote in the development of cardiovascular divas this is @ point that needs to be resolved From Nutrition Reviews, Vol. 29, No. 5 May 1971, pp. 124-26