r~;~"~ ~'"~" ;:..,;:,', ,. . .,' . ..'

"

1

~.'I'

I.: I[;U-.:JlK".~ - ...~I!!I""""'",.-.';.,r;-.'~...1II:.'j

a'j . (\)II Malnutrition and Diabetes in the Tropics

.4'1.:II

.~,,""._~

"';r""...

'..

('J ~\!"/'.-

. .

REPORT OF THE end (BMI-:;18.5 kwm2) of the spectrumINTERNATIONAL WORKSHOP of NIDDM, to replace the term leanON TYPESOFDIABETES NIDDM., ,

PECULIAR' YO ,THE-.TRO,ICS . Fromourdiscussions,the follow~The International Workshop on type.<;of . ing three consensus statements were itd~Dia15etesPeculiar to the Tropics was held van~ed:in Cuttack, India, 17-19 October 1995.The workshop was attended by 60 dele-gates, most from four developing coun-tries (India, Bangladesh, China, and Ethi-opia), with observers from Europe andthe u.s. It is now well recognized thatdbbetf;'5 in tropical regions and in somedeveloping countries presents clinicallydifferently from IDDM and NIDDM in theWestern wor:d and developed countries.The World Health Organization (WHO)acknowledged this in 1985 by recogniz-ing a ~ird class of diabetes; designatedmalnutrition-re-lated diabetes mellitus(MRDM) and subdivided into fibrocalcu-

. . lous pancreatic diabetes (FCPD) and pro-tein-deficient pancreatic diabetes (PDPD)(subsequently often called protein-defir;:ient dfabetes mellitus (PDDM)).Since then, information has been col-

. lected about diabetes in tropical regions.Such knowledge was presented and dis-cussed at the workshop to serve as a stim-ulus to the funher research that is neces-

sary into the etiologic, epidemiologi'cal,and clinical aspects of this class of diabe-tes.

Am..ther category of diabetes wasalso discussed: NIDDM in subjects of lowbody weight (also referred to as "leanNIDDM"). All cases of diabetes can be di-vided into IDDM and NIDDM accordingto the patient's current clinical and meta-bolic state. It was the co~nsus of theworkshop that .although it has served avaluable purpose, the, term MRDM nowbe sUpplanted by the f6llo\ving: 1) malnu-trition-modulated diabetes mellitus

(MMDM), to replace PDPD; 2) FCPD, tobe considered as a specific form of diabe-tes, clinically either NIDDMor, less fre-quently, IDDM; 3) the low-body weight

PDPD. PDDMThere is a clinical syndrome of diabetesthat occurs in young malnourished indi-viduals in. developing countries. Thesepatients' . clinical features differ from theusual clinical presentations of IDDM orNIDDM as described in developed coun-tries and from that of FCPD, These pa-tients are insulin-requiring but not keto-sis-prone. The designation recommendedfor this group is MMDM. This term re-places all other nomenclatures, such asMRDM; PDPD, and PDDM. The clinicalcharacteristics of MMDM include: .

, 1. having early. onset of diabetes,usually before 30 years of age; .

. 2.requiringinsulin(toobtainad-equateglycemiccontrol); - '-

3. not being prone to ketosis;4. ha\ing an absence oCimaging

evidence of pancreatic' calculi or ductaldilation; -

5. having low BMI(usually <17),with other clinical features of malnutri-tion and often growth retardation.

FCPD1. FCPD is a form of diabetes with

a high prevalence in tropical and develop-ing countries. .-

2. FCPD is due to chronic calcu-

lous pancreatopathy not to chronic alco-holism or other recognized ascribablecauses such as hyperparathyroidism.

- ;3.It is usually seen in young andmalnourished individuals b\!1..aISooccursin others. -

4. Diabetes and pancreatic calculiand/or ductal dilatation are essential fea- ,tures. Recurrent abdominal pain andste-atorrhoea are other important features,........................................................

Recei\'ed for publicalion 13 Febntal)' 1996 and accepled 21 February 1996.Address correspondence and repJim requestS 10 R. Harsha Rao. MO', N-810 Momcliorc Universily Hos-

pilal, 200 LOlhrop51..Pillsburgh. PA 152\3. ,

FCPO. librocalculous pancrealic diabetes; M~IDM. malnutrition.modulaled diabetes mellilus; ~.IROM.malnutJition-reialed diabeles mellims; POPO. pr<'lein-deliciem pancrearic diabetes; PODM. protein-deficient diabeles mellilus; WHO. World Health Organization.

10H,

1iIIIDF~

- \S :~0-Ja;.-,

. f~ I.. ~.-j.,~

}

but absence of these latter does not pre-clude the diagnosis.

5. The clinical pr.ofile of diabetesshows a spectrum of hypergl}lcemia vary-

. ing from severe to mild: Ketosis is uncom-mon.

6. Pancreatic calculi are usuallylarge multiple and intraductal. Markedductal dilatation and fibrosis are usual;inflammatory chang~s are uncommon. -

7. Abnormal exocrine pancreaticfunction is invariably presentbut is often

.demonstrable only following investiga-tion.

8. FCPD is associated with an in-

creased risk of pancreatic carcinoma.9. Management of FCPD includes

treatment of diabetes, oral enzyme re-placement, and relief of pain. Surgerymay be required for severe intractablepain and for other indications. .

10. The etiology of FCPD is uncer-tain:. The roles of nutritional, environ-

mental, and genetic factors need' funherim'estigation.

NIDDM in lean subjects, .This group supponsthe WHO classifica-tion of NIDDM into obeseand nonobesesubclasses.In somedeveloping countries,nonobese patients constitute the morecommon category, and a proportion ofthem have BMI of <18.5. There are manyfactors that are not well understood inthese subjects with NlDDMand low bodyweight; funher research is rt:quired in thisgroup.

JOSEPHJ. HOET

B.B. TRIPATHY

CHAIRPERSONS

COMMENTARY, TIME TORETHINK MALNUTRITIONAND DIABETES IN THETROPICSThe hist 3 decades have seen a fascin:uinge\'olution in concepts regarding-the roleof nutrition in the etiopathogenesis of di-abetes in tropical countries. The pendu-Illm has swung from one extreme ("Theevidence that malnutrition prot~ct~ :Jdullpopulations from diabetes seems unaS-sailable." (1)) to the other ("Malnutritionis probably a major determinant of diabe-

DIA8HES C\RE, VOLUME19, NUMIIER.9, SEPTEMBER1996

r!"'''7.,~,.\:''%;' -'~~~e.,..~'~.--~~~~"". "': ," """'?~P''';:~F",,,,,~,r~1':3

,f.-'!!j

r

\1:

It:

l

ih

b'f';"'>, .-

, i

I.11I

F

r

i'J

. III ;;1;I

. rI . I'.1Ii

i

tes." !2)). The turnaround has been so. complete that the WHO has recognized

MROM as a form of diabetes that is dis-tinct from classical lDOM and NlDO?v!(3). The umbrella term MROM highlightsthe association (frequent but not invari-able) -ofmalnutrition '~ith certain clinicaldiabetic syndromes in the tropics and hassome utility because these syndromeshave features that set them apart fromboth lDDM illad'NlDDM. Unfortuniuely.,the creation of a third major form of dia-betes that is neither insulin-dependentnor non-insulin-dependent presents ma-jor problems. In particular, the implica-tion of etiologic distinctness conveyed bythe term MRDM obfuscates the relation-

ship of the tropical syndromes to the in-ternational nomenclature (i.e.,lDDM ver-sus N!DDM). The internationally acceptedclassification is based on clinical pauernsnot on etiology, and there appears to belittle justification for such a radical depar-ture from it.

In the decade since the publica-tion' of the WHO .report, fresh evidencehas become available regarding the inter-action of malnutritiol1 and diabetes in thetropiCs. It has recently been argued thatthe tnith probably lies Somewhere be-tween the tWOextremes of either indicting;l1alnutrition as the proximate cause forthe peculiar diabetic syndromes de-scribed in the tropics or rejecting a role forit entirely (4). This centrist position hasbeen affirmed in the above statement,leading to a recommendation that thenomenclature for the tropical diabeticsyndromes be changed in a manner thatreflects current thinking. The recom-mended chc.nges may be summarized asfollows:

L The term MRDM is inappropri-ate, and the continued use of this termi-nology is inadvisable for the followingreasons: it carries the implication of a uni-form etiopathogenetic mechanism forseveral diverse clinical syndromes; epide-miological studies have failed to show aninvariable association between malnutri-tion and these fom1s of diabetes; and itsdiagnosis relies exclusively on a low 8MIbeing present at diagnosis of diabetes. Theproblem with this heavy reliance on a lowBMI has to do with the fact that subjectswith these syndromes of diabetes belong10 the lower socioeconomic groups and

. are hence thin even before the onset ofdiabetes. Moreover, a history of pro-longed neglect of symptoms can almost

IDIAnETESCARE, VOlllME 19, NIJMIIER9, SrruMIII'R 1996

' '-"--:'~-'-" ......

.! ---:"

(2)invariably be elicited in paticms from arural background. Thus, a significantcomponent of the low BM!at diagnosismay be attributable to diabetes-relatedmalnutrition..and its clinical use as an in-variable marker of malnutrition-relateddiabetes may not be entirely justified.

7:.FCPD, one of the two clinical.syndromes incorporated under the um-brella term MRDMin the WHO classifi-cation, is ofte'nassociated with mah~lUtri-tion as a co~quence of severe exocrineand endocrine pancreatic deficiency,which usuall)'has remained untreated foran extended. period before diagnosis.Since several studies show a frequent as-sociation between this syndrome and alow .socioeconomic status, malnutritionmay playa p.::rmissiverole in its patho-genesis. However, compelling evidenceforsuch a postulation is lacking,and thereis a general agreement that this is a (sec-ondary) formofpancreatic diabetes of un-certain etiology seen in tropical countries. .

lts relationship to nonalcoholic variantsof pancreatic diabetes seen in temperatezones has not been determiried.

3. There is no evidence implicat-ing eitlJerprotein deficiencyor a pancre-atopathy in the causationo( so-calledPDPD,the second of the two clinical syn-dromesthatcompriseMRDM.Thus,thisterm is erroneous and should be dis-carded. EXperimentaland clinical datasuppon the contention that malnutritioncan both playa permissive role in thepathogenesis of diabetes in a number ofsituations and modify the clinical picture

. of coexistent diabetes to give rise to"unique" fe:Hures,such as insulin resis-tance and ketosisresistance.Thus, theen-tity that was formerlycalledPDPDismoreappropriately termed malnutrition-modulated diabetes mellitus (MMDM).

4. NIDDMin malnourished indi-viduals may represent one end of a spec-trum of body weights associated withNIDDM; tlms, lean NIDDM in tropicalcounuies maynot be an etiolegically(Iis-linct enlity. However, thi,s'"syndromeoc-curs even in relath.ely amuent paiiems,and responsi\'enessto oral hypoglycemicagents is often preserved for prolongedperiods aCta diagnosis. These featuressuggest that it ma)' not be correct to dis-miss it simplyas a nutritionally modifiedvariant of cbssic heredofamilial NIDDM.Fu'r.therrc~.'~lrchis nec~'$Saryto elucidateits etiopallll1gmesisand natural histOty.

Thl' reCl1llllllcndcdchanges in no-

.'

Report and tonUtumtary

menclalure reItecta major shift inempha-Sts,w\\icn :iscnotoeodin Wee cooce,pc.s.Pilefirst and most fundamemal oh.hcll!' COt...cepts is thilt malnutrition may not be theprimary cause of diabetes in the tropics.The reasons for such a reevaluation .maybe found in the existing literature, ex-haustively reviewed elsewhere (4-1a).The preponderance of evidence indicatesthat in the majority 6f individuals withcoincident malnutrition and diabe.tes,malnutrition cannot be indicted as the'sole causative factor for diabetes. In"par-.ticular, there is no evidence,experimentalor human, to support the contention that"continued protein deprivation wpuld.ul-timately result in irreparable damage tobeta cells" (2) of a severity sufijcient tocau~ diabetes in and of itself..Thus, it isclear that, notwithstanding the WHOclassification ofMRDM (and PDDM) as anetiologically distinct syndrome of diabe-tes, malnutrition is not the primary oreven the most proximatecause of diabeticsyndromes peculiar to the tropics. In thecase of the syndrome of FCPD, the asso-

. dation is even more tenuoU$because italso occurs in well-nourished subjectsand malnutrition can be excluded as theprimary etiologic factor in such patients(8-10). Thus, at the present time, FCPDis .best consideredas a secoridaryfOrmofdiabetes. However, the conclusion thatmalnutrition does not playa primary rolein the etiologyof diabetesin the tropicsdoes not mean that malnutrition plays norole whatsoever in the causation of diabe-tes in the tropics. Thus, a deficiency ofantioxidants in the diet could set up un-mitigated oxidative damage in the pan-creas from xenobiotics abundanti)'present in the tropical environment, .

thereby contributing to the pathogenesisof FCPD (II). .

A second concept important tounderstanding the interaction betweenmalnutrition and diabetes in the tropics isthat malnutrition could h.we an indirectrole in the etiopathogenesi.s of diabetes inthe tropics. Experiment.tl data indicatethat malnutrition might well play whathas been termed a secondary role,whereby it unmasks or amplifies the ef-fectsof a primarydiabetl1genicinfluencethat in itself mightlllit be operating to itsfull effect (4). Recent epidemiologicalslIIdt.cs frcir.'J1the 'U.\( Shl~,\' a.strOl~ as~~'-

. ci;1rionbct\\'l'i:.n cady (fl'uLmd rv:(':n;lI~mmalnutrition ~'nd Ih<:risk 0fdiabl'l'Cs and

s)'ndrome X in latcr lire (12-I-f). Thc ar-

1015

,.

,.'

;'

I I

i rI

l-II

"\,I

t!

\

~,

~

,

...""

,I

,"

,

'

I,

it;!,

1

,

1

~;I

~ftl. J.ir~.'

>",,-'

,!

i,>.:t,~';r~.;,~1~..'i:'~;,'::::.? ;:.,

aRei]o,.tClndcornrncntcuy

gumcnts that the authors of those slUdiesmake in support of this proposed link(15) are very similar to those advanced byone of us previously in relation to diabetesin the tropics (5,6), i.e., that the increasedrisk of diabetes .is a consequence of fetal

."malnutritio_n (possibly due to malern.alnutritional deprivatiQn). According tothese arguments, on the one hand, im-paired development of /3-cells during a

. critical period of intrauterine and early.neonatal life results in an irrev~rsible dim-inutionin l3-cell number and mass. Onthe other, resistance develops to the ac-tions of groWth-promoting factors, suchas insulin, growth hormone, and IGF-I, intissues and. persists into postnatal life06,17). These two processes may interactwith each othz:- or with other diabeto-

genic influences (genetic or environmen-tal) to increase susceptibility to the futuredevelopment of diabetes. Hales andBarker (15) have called this concept "thethrifty phenotype hypothesis, fta term thatcarries the teleological implication of along-term price that has to be paid to en-sure the shon-term survival and growthof the developingorganispl.Themost ob-vious featureof the thrifty p~enotype is alow binh weight,but low ponderalindex',shon length, and small head and' waistcircumferencesat birth have been associ-ated with adult diabetes or other cardio-vascular risk factors in adult life.Similarreports of the association of low birthweightwith adult diabetesand insulin re-sistance are available from the U.S. andSweden (18-20). The thrifty phenotypeconcept is relevantto the causationof di-abetes in the developingcountries of thetropics.Malnutritionin pregnantmothersand fetal malnutrition are common: athird ofallbabiesborn in Indiahavea lowbinh weight, predominantly because ofintrauterine growth retardation (21).Studies fromboth India and the U.K.linkfetal malnut,rition with an insulin-resistant state in adult life, characteriZedby diabctes-and coronary'artery..disease(14,22). Early life malnutrition coupled,with overnutrition later in life seems toconstitute a particularly deleterious com- '

bination of factors,suggestingan ongoing'interaction between the two ends of thespectrum of dysnu'trition in the patho-genesis of adult diabetes (23); Thus, theconcept Qfa thrifty phenotype that is ren- ,

dered detrimentalby subsequent overnu-trition provides insight into some of thewell-known peculiarities of diabetes in

1016

.!

.,_..Co. '..: ,,:',; ,'-0,..;'. "",,", " ...' "..<C'..:, ~,

dmloping coun"'" in th,Q2UChas India. As might be predicted from thethrifty phenotype concept, subjects inthese countries develop diabetes at a anearlier age and with a relatively lowerBMI, characterized by marked centra"obesity and more severeinsulin resistancethan in individuals in developed coun-tries and display several other cardiovas-cular.risk factors, despite being thin (24~.26).Thus, early malnutrition followedbypredominantly central weight gain couldlead to the insulin-resistance syndrome,supporting the secondary hypothesis forthe role of malnutrition in diabetes (5).'This framework alsoprovides an explana-tion for the high prevalence of diabetes inmigrant Indians (first generation), whohave a higher BMIthan native Indians andgreater central obesity than indigenouspopulations (27). Some have argued thatsuch observations indicate that the thriftyphenotype actually represents a selectivesurvival manifestation of the "thrifty ge-notypeft (19). However, this interestingvarjation on the overall theme must re-

, main in the realm of speculation until de-finitiv~studieson genetic markers arecar-ried out.

The third concept underlying thechanged emphasis reflected in the con-sensus statement is that malnutritionmodifies the clinical picture of diabeteswhen the two conditions occur coirici-dentally. This has been referred to as the"tertiary" role (5). Thus, in both humansand experimentalanimals, malnutrition isassociatedwith insulinopenia, insulin re-sistance, and an absence of ketosis, three"uniqueftfeaturesof the clinicalsyndromethat was formerly called PDPD (4-6,28).Furthermore, experimental data indicatethat the insulin resistance and insulin de-ficiencyof malnutrition can be expressedin the context of coexisting diabetes (29-32). These observations indicate that it isprobable that the clinical picture of hu-

''man diabetes ca.nbe modified by coexist-ing malnutrition., Thus, the interactionbetwee'n malriutrition and diabetes mayhavenothing to do with etiology and ev-erything to do with the clinical presenta-tion, A shift in emphasis from etiologytoclinical patterns would also make it easierto determine the proper placement of theclinical syndromes ofdiabetes in the trop-ics within the internationally al;ceptedclinical classification.Thus, instead of theetiologicallyderived and hence inappro-priate tenninology espoused by th~WHO

:], '''J'

-(POI'D), it has been suggested Ihal Iterm,MMOM (malnutrition-modulaldiabetes mellitus) bc used. The new teltacitly acknowledges that the clinical s)'drome may mask the underlying formdiabetes. It also implies that it is an inhtenlly heterogeneous entity, made updiverse syndromes that may be clinicalsimilar but etiologi!:ally different. In othtwords, MMDM may occur ov~r a templalof 'either IDDM or' N)ODM. This simp]concept takes into account the contradi<tory obse'rvations reported by differenauthors in studies of patients with thsame clinical syndrome. For instance, nutritional rehabilitation of these subject~can uncover an underlying pattern of ei.ther NIDDM or IDDM (33,34). Similarly.some .cports indicate thal these subjectsare islet cell antibody positive (35) andha\'e partial HlA identity with classicalIDDM (36), whereas others show quitethe opposite (37,38). Thus, MMDM maybe thought of as a mixed bag: it includesnutritionally modified forms of both clas-sical types of diabetes" the .basic featuresof the underlying syndrome beingmasked by the clinical and metabolic fea-turC$'of malnutrition. The ne* conceptsdiscussed above thus provide an inter-nally consistent framework within whichare incorporated all of the sometimes con-tradictory observations regarding the in-teraction between malnutrition and dia-betes. The framework is based on the two

hy.potheses for which there is some evi-dence, the secondary and the tertiary hy-potheses described earlier (5), and ac-cords with the available data, both clinical jand experimental, regarding the interac-tion. Therefore, at the present time, itseems reasonable to conclude that malnu-

trition plays an imponant role in precip-itating or modifying diabetes in develop-ing countries, but that it is neither the solenor even the proximate cause of the syn-dromes that are peculiarly seen in mal-nourished PQPuladons. .However, it is,im-portant to r~cognize that the last word onthe subject has not been wriuen'. At a timewhen our understanding of this interac-tion continues to evolve, it is vital to keepan open mind on this fascinating ~u~iect.Any connection between malnutritionand diabetes may appear esoteric and re-mote to diabetologists uscd to seeingobese patients. in the de\'cloped world.bUt the field has nonetheless provided awholly unexpected insight into the patho-genesis of diabetes itself, based on lhe ep-

: I

- ,- , ~. ,-' ".~"':~ ~'"" '~""' 19,:':'9';::~~'::~l".; -- '., ' .. _:~~~

,,;r-~.7""'~;::-'~', :~:;~,i:~)'-:.'" "';,:'T'

--~~:;,;<~~~~/'" ,',.."'. ,... =\""",-.,.'

, I,ijt"

,

I

fii~1 .

~ .

.

idemiological evidence that is at the heart01the thrihy phenotype hypoll1csis.

R. HARSHA RAO, MD, MRCP

CIIITTAR,\NJAN 5, YAJNIK, MD

ii1'J

l-t

.-From lhe-UniversilY of Pitlsburgh Sch~l of Medi-

cine. Piltsburgh. Pennsylvania; and lhe DiabelOlogy'Research Center, K.E.M. Hospilal and Research

- u'1ter, Pune, India.

~

References

,

J'1

1.,tt

"I

j

I'Jrtf:;'/

!,}

!'jj,j

:j!

1. World Health Organization: DiabetesMel-litus:Reporlof a WHO Expert Committee.Geneva, World Health Org.; 1965, p.16-17 (Tech. Rep. Ser., no. 310)

2. WorldHealth Organization:WHO ExpertCommitteeon DiabetesMellitus:SecondRe-port Geneva, World Health Org., 1980, p.

. 23-24 (Tech. Rep. Ser., no. 646)3. World Health Organization; Diabetes Mel-

:;tus:Reportof a WHO Study Group. Ge-neva, World Health Org., 1985 (Tech.Rep. Ser., no. 727)

.r Rao RH; Is tropical pancreatic diabetesmalnutrition-related? Diabetes Care 16;941-945, 1993 .

5. Rao RH; Diabetes in the undernourished:

coincidence or consequence? EndocrineRev9;67-87,198S .

6. Rao RH; The role of undernutrition in the

pathogenesisof diabetes mellitus. Diabe-. Its Cart 7:595-601, 1984

7.Yajnik CS: Malnutrition related diabetesmellitus(Editorial).:Int J DiabetesDcvCountries13;1-2,1993

-7a.Yajnik CS, Shelgikar KM: Fibroca\culouspancreatic diabetes in Pune, India. Diabe-les Cart 16:916-921, 1993

8. Mohan V, Chari ST, Ramachandran A,Malathi 5, Madanagopalan N, ViswanathanM; Fibrocalculouspancreaticdiabetesandobesity.DiabetesResClinpract 8:161-166, 1990

9. Yajnik CS: Diabetes secondary to tropicalcalcificpancreatitis.Clin EridocrinolMctab6:777-796,1992

10. BraganzaJM: Oxidant stress: common de-nominator in the pathogenesis of temper-ature zone and tropical chronic pancreati;tis? .In Proceedingslif WHO' Workshop 0'/1Malnutrition-Related DiabA!(es. AlbeniKGMM, Keen H, Parry E, 'Eds., London,Oxford Univ. Press. In press

12; Hab CN, Barker DJP, Clark PMS, Cox LJ,Fall C, Osmond C, Wimer PD; Fetal and

infani growth and impaired gluwsc !Oler-ance al age 64. Br :>.It't/)303: 1019-\022.1991

13. Phipps K, Barka DJP, Hales CN, FallCHD, Osmond C, Clark P~IS; Fl'utgfl1\\.ihand impaired glucosc tolerance in

DIAIIUb CAR[, VOlllME 19, NLJMIII:~9, SU'lHIIII'~ 1996

,

~men and women. Diabetologia35:225-228,1993

If. Barker DJP, Hales CN, Fall CHD, Os-mond C, Phipps K, Clark PMS:Type 2(non-insulin-dependem) diabetes .melli-tus, hypertension, and hyperlipidemia(syndrome X): relation to reduced fetalgroWth.Diabctologia36:62-67, 1993

15. HalesCN, Barker DJP;T}'pe2 (non-insu-lin-dependent) diabetes mellitus: the

. thrifty phen.Olypehypothesis. Diabetolo-

. gia 3'5:595-601, 1992 . .16. BarkerDJP:Fetal originsof coronaryheart.

disease.Br MedJ311:171-174,199517. FallCHD,PanditAN, LawCM,YajnikCS,

Clark PM,Brier B,Osmond C, ShiellAW,Gluckman PD, Barker DJP: Size at birthand plasma insulin-like growth factor-lconcentrations.Arch Dis Chila 73:287-293, 1995

18. Valdez R, Athens MA, Thompson GH,Bradshaw BS, Stem MP: Binhweight andadult health outcomes in a biethnic USpopulation. Diabetologia 37:624-631,1994

19. McCance DR, Pettit DJ, Hanson RL,Ja-cobsson LTM,Knowles WC, Bennet PH:Binh weight and non insulin dependentdiabetes: "thrifty genotype", "thrifty phe-notype", or "surviving small baby gene-

: type"?BrMedJ308:9H-9.45, 1994 .20. McKeiguePM, Leon DA, BerglundI, Me-

hren R, Lithell HO: Relationshipof binhweight and ponderal index to non-insulindependent diabetes and insulin responseto glucose challenge in men.aged 50-60years. DiabeticMtd 11(Suppl. 17):5i-S7,1994

21. Gopalan C: Low birth weights: signifi-cance and implications.In Nutrition inChildren.SachdevaHPS,ChoudhuryP,Eds. New Delhi, Cambridge Press, 1994,p. 1-33

22. Yajnik CS, Fall CHD, Vaidya U, PanditAN, Bavdekar A, Bhat DS, Osmond C,Hales CN, Barker DJP: Fetal groWthandglucose and insulin metabolism in four-year-old Indian children. Diabetic Med 12:330-336,1995

23. Yajnik CS: Diabetes and insulin resistancesyndrome in Indians. Bull Nutr found In-cliq16:1-5, 1995 .

24. Shelgikar.KM,Hock:\dayTOR,YajnikCS:; .Cemral. rather' than genera/lsed' obesity is

related to hyperglycaemia in Asian Indiansubjects. Diabetic Mcd 8:712-717,1991

25.. Yajnik CS, Naik 55, Bhat DS, Joshi VM,Shclgikar KM, Alberti KGMM, HockadayTDR: The relationship between obesity,plasma immunoreactive insulin concen-tration and blood pressure in newly diag-nosed Indian type 2 diabetic patients. Di-abetic .\1,..110;146-151,1993 .

26. Mohan V, Alberti KGMM: Diabetes in

. ,-

-Report and commentary

tropics.In International Textbooll oj Diabe-tes. Albeni KGMM, DeFronzo RA, KeenH, Zimmet P, Eds. Chichester, U.K.,JohnWiley, 1990, p. 177-196

27. McKeigue PM, Shah B, Marmot MG: Re-lation of central obesity and insulin resis-tance with high diabetes prevalence andcardiovascular risk in South Asians. lan-

cet 337:382-386, 199128. Heard CRC: The effects of protein energy

malnutrition on blood alucose homeosta-sis. World Rev Nutr Diet 30:'107-147,1979 .

29. RaoRH:Chronic undernutrition accentu-ates insulin deficiency in .rats with mildstreptozocin-induced diabetes. Diabeles40:1404-1409,1991

30. RaoRH, MenonRK: Chronic malnutri-tion impairs insulin sensitivity throughboth receptor and po~i.,,:ceptor defects inrats with mild streptozocin diabetes. Me-tabolism42:772-779,1993

31. Rao RH: Adaptationsin glucose ho-meostasis during chronic nUtritional de-privation in rats: hepatic resistance toboth insulinand glucagon.Metabolism44:817-824,1995

32. Rao RH: Insulin.resis~nce during eugly-cemic clampstudies in chronically under-nourished rats with mild streptozocin di-

. abetes,Metabolism44:1422-1427,1995~3. Tulloch JA, Msc1ntosh D: 1" type diabe-

tes. Lancel ii:119-121 , 196134. Abdulkadir J, Mengesha B, Gebriel ZW,

KeenH, Worku Y, Gebre P, Bekele A,Urga. K,-Taddesse AS: The clinical andhormonal (C-peptide and glucagon) pro-file and liability to ketoacidosis duringnutritional rehabilitation in Ethiopian pa-tients with malnutrition-related diabetes

mellitus. Diabetologia33:222-227, 199035. Hazra DK,Singh R, Singh B, Gupta MK,

Agarwal P, Mittal 5: Autoantibodies in atropical ketO$is resistant but insulin de-pendent diabetes mellitus. In DiabetesMelli.tusin DevelopingCounlries. Bajaj JS,Ed. New Delhi, Interprint, 1984, p. 165-161

36. Abdulkadir J, Worku y, Schreuder GMT,D'Amaro J, DeVries RRP, OttenhoffTHM:HLA-DR al'!tigens in malnUtrition-relatcddiabetes mellitus fn Ethiopians: a cl!1e toits etjology?TissueAntigens34:284-289,-1989 . . .

37. Ahuja MMS: Heterogeneity in tropicalpancreatic diabetes (Letter). Diabctologia28:708, 1985 .

38. SanjeeviCB,Scshiah V, Moller E, Olerup0: Different genetic backgrounds for mal-nutrition-related diabetcs and type I (in-sulin-dependcnt) dia~tes mellitus inSouth In:di:ms. Diaht'l.'1It'b'ia~5:2.83-286,1992

1017J

.. -.