Marmot Indian dispora

8/7/2019 Marmot Indian dispora

1/13

J Clin Qidmhl Vol. 42, No. 7, pp .597409, 1989Printed in GreatBritain .All rightsnscr~ul 08954356/8983.00+ 0.00CopyrightQ 1989PcrgamonPress plc

CORONAR Y HEART D ISEASE IN SOUTH ASIANSOVERSEAS: A REVIEW

P. M. MCKEIGUE ,* G. J.MILLER and M. G. MARMOTDepartment of Comm unity Medicine, University College and Middlesex School of Medicine,66-72 Gower Street, London WClE 6EA, England and MRC Epidemiology and Medical CareUnit, Northwick Park Hospital, Watford Road, Harrow, Middlesex HA1 3UJ, England

(Received in revised form 9 December 1988)

Ahstraet-C oronary heart disease rates have been reported in several parts of the world to beunusually high in people originating from the Indian subcontinent. High coronary disease ratesappear to be common to South Asian groups of different geographical origin, religion, andlanguage. This presents a challenge to the understanding of coronary heart disease: the high ratesin South Asians are not explained on the basis of elevated serum cholesterol, smoking orhypertension. Low plasma HD L cholesterol, high plasma triglyceride levels and high prevalenceof non-insulin-depen dent diabetes have been consistently found in Sou th Asians overseas: thisprobab ly reflects an unde rlying state of insulin resistance. F urther studies are needed to determinewhether this metabolic disturbance can account for the high rates of coronary heart disease inSouth Asians, and to identify possibilities for prevention.Coronary disease occurrence, mortality Dietary fats Smoking Blood pressureCholesterol Diabete s mellitus, non-insulin-depen dent Insulin resistance

INTRODUCTIONCoronary heart disease (CHD) rates have beennoted in several parts of the world to be particu-larly high in people originating from the Indiansubcontinent. A basis for preventive strategiesin these communities depends upon adequateunderstanding of the causes of this epidemic.This review examines the evidence of excessCHD rates in South Asians, the extent to whichexisting knowledge about the pathogenesis ofthe disease can account for this excess, newhypotheses and possibilities for prevention. Inthis review the term South Asian is used todenote origin from the Indian subcontinent.Those who migrated before Partition, and theirdescendants, generally describe themselves asIndian. Migration from India, Pakistan, andBangladesh to the U.K. has occurred mainly

*Author for correspondence. (Tel.: 01-387 -7050 ext. 5703 ).P. M. McKeigue was supported by a Wellcome TrainingFellowship in Clinical Epidemiology during the prepara-tion of this article.

since the partition of the subcontinent: in theU.K. the term Asian is generally used forthose originating from this region.

THE INDIAN DIASPORAIndian people have migrated to other lands

for many centuries. The modern history ofIndian emigration dates from the beginning ofthe colonial system of indentured labour in 1834[l ,2]. This led to the creation of Indian commu-nities in Mauritius (from 1834), the West Indies(from 1838), Natal (from 1860), the StraitsSettlements (from 1877), Fiji (from 1879) andEast Africa (from 1896). The main agenciesresponsible for recruiting indentured labourerswere in Calcutta .and Madras. Indians leavingthrough Calcutta came mainly from the UnitedProvinces and Bihar, with smaller numbersfrom Bengal, Orissa, Oudh, Punjab and North-West Provinces. Those emigrating throughMadras were mostly Dravidian people fromMadras State (now Tamilnadu) and Andhra

C.E. 4217-A 59 7


2/13

59 8 P. M. MCKEIGLJEet al.

Pradesh. Hindus exceeded M uslims by about 9to 1, and many from the South had beenconverted to Christianity before their departure.Other Indians trave lled through the BritishEmpire on their own initiative, particularly toEast and Central Africa. Many of these entre-preneurs and professionals came from Gujarat,Bombay Presidency and Goa. The indenturedlabour system was banned by the Viceroy in1917 but a steady m igration of traders, profes-sionals and their families has continued eversince. Indian migran ts continued to arrive inEast Africa until the enactme nt of legislativebans in 1954. With independence and subse-quent political changes, large numbers of Indi-ans in East Africa were forced to move to India,Europe or North Am erica. M igration to Britainfrom the Indian sub continent increased after1960 and reached a peak around 1966-67 [3].Migrant to Britain from Gujarat and Punjab

$came f m rural areas [4]. Most had beenagricultural or man ual w orkers but their literacyrate was higher than that of the general popula-tion. Migrants from Pakistan and Bangladeshwere from poorer circumstanc es than thosefrom the North of India.

CORONARY HEART DISEASE RATESIN INDIANS OVERSEAS

The earliest report of high CHD rates inIndians compared with other ethnic groups w asfrom Sing apore in 1957 [5]. Prevalence ofCHD -defined in this study as coronary arterydisease with myoca rdial involvement-in aseries of all 9568 autopsies undertake n over theyears 1950-54 was seven times higher in Indianmales than Chinese [5]. Age-standardized CHDdeath rates in 1954-57 were four times higher inIndians than Chinese. Though CHD mortalityin Singapore has more than doubled between1957 and 1978 in both groups, a threefold excessin Indians com pared with Chinese remains [6].In Uganda 43% of deaths among South Asianmen in Kamp ala in the years 195658 werecertified as due to CHD while in the Africanpopulation of Uganda the disease was said to bealmost non-existent [7J In South A frica, C HDmortality in Indian women aged 30-69 was 49%above that for wom en of Europea n descentduring 1955-57 [8]. Mortality was similar inIndian and European men at this time but overthe years 1968-1977 the rate for Indian menaged 15-64 increased to 45% above the rate forEuropeans [9]: around twice the estimated rates

for the U.S.A. and U.K. at this time. In Fijiduring 1970-73, myocardial infarction wasnoted to be much commoner among hospitaladmissions of men of Indian descent than thoseof Me lanesian descent [lo]. Nationa l mortalitydata for 1971-80 showed age-specific mortalityfrom ischaemic heart disease to be about threetimes higher in Indians than in Melanesians,though only 72% of deaths were medicallycertified [111.In Trinidad, where about 30% of the popula-tion is of Indian descent, the prevalence ofcoronary heart disease in different ethnic groupswas assessed in a survey of a total urbancomm unity [12]. The odds ratio for major elec-trocardiographic Q waves (Minnesota codesl-l, 1-2) in Indians vs other ethnic groups wa s3.8 at ages 35-54 and 1.3 at ages 55-69 years.In a more recent ana lysis of mortality in thiscommunity between 1977 and 1985, age-adjusted relative risk of death from cardiovascu-lar disease (mostly CHD ), in comparison withadults of African descent, was 2.6 in Indians, 2.1in Europeans, and 0.3 in adults of mixed de-scent, with no significant sex difference in rela-tive risk estimates [13].In Britain high CHD rates in South Asianswere first observed at the time of the 1 971Cens us. In a comm unity heart attack register inthe East London borough of Tower Ham lets, 40coronary events were recorded over the years1970-72 in South Asian-born men, predomin-antly Bang ladeshi: the numb er expected fromthe rate for all males in the borough wa s 31 [14].Analys is of national data for Englan d an dWales during the same period showed standard-ized mortality ratios (SM Rs) for CH D in mi-grants from the Indian subcontinen t of 119 formen and 128 for women aged 20-69 years,taking the rates for all men and all womenrespectively as 100 [15]. In contrast migra ntsfrom other developing countries had low SM Rsfor coronary heart disease . Ana lysis of sur-names on death certificates showed that ethnicSouth Asians born in Africa also had high CHDmortality ra tes: the proportional mortality ra -tios (PMRs) were 120 for men and 163 forwomen, although these ratios were based onsmall numbers of deaths. The social class gradi-ent of CH D mortality seen in the general popu-lation of England and Wales was absent inSouth Asian s. It is not possible to calculateseparate rates for migrants from differentcountries of the subcontinent but surnam es ondeath certificates of imm igrants from the Indian


3/13

Coronary Heart Disease in South Asians 599subcontinent have been used to distinguishGujaratis, Punjabis, Southerners and Muslims:all four groups had high PM Rs for coronaryheart disease in the years 1975-7 7 [16]. A4O-50 % excess. of myocardial infarctions inSouth Asians was found in analyses of NationalHealth Service hospital admission data forLeicester during 1977-78 [17] and North-WestLondon during 1980-8 2 [18]. The clinical pres-entation a nd anatom ical distribution of diseaseat angiography in South Asian and nativeBritish patients with CHD do not differ exceptthat diabetes is more common in South Asianpatients [19,20]. Nationa l mortality. data bycountry of birth for the period around the 1981Cen sus are not yet -available but data forLondon boroughs with large South Asian popu-lations show CH D mortality to be 4060%higher among South A sians in each boroughthan the average for England and Wales [21].This high mortality is shared by populationsoriginating from Gujarat, Punjab, Bangladeshand Pakistan.The high CHD rates of South Asians inEngland and Wales in 197&72 were an unex-pected finding in recent immigran ts: migran tsare usually selected for fitness, although theexodus from East Africa may have been anexception. The high rates seen in certain othermigran t groups-those from Scotland, Ireland,Poland, U.S.A. and South Africa--correspondto the rates recorded in their countries of origin.It is likely therefore that h igh rates of coronaryheart disease exist in the Indian subcontinen t, atleast in those sections of the population fromwhich migrants to Britain w ere drawn.

CORONARY HEART DISEASE INTHE INDIAN SUBCONTINENTThough Indian clinicians report that CHD iscomm on in urban practice [22] there are no

satisfactory po pulation-based data on coronarymortality or incidence in India, Pakista n orBang ladesh. Hospital adm ission rates for Indianrailwaymen between 1958 and 1962 have beenreported but only adm issions to railway hospi-tals from electrocardiographically-documentedCHD were ascertained and the rates were notstandardize d for age [23]. Two surveys of CH Dprevalence have been conducted in northernIndia [24,25], and the results have been com-pared with those of a survey in Tecumseh,U.S.A. [26]. In Chandigarh, capital of Punjaband Haryana, 2030 adults aged 3 0 years and

over were examined, while in rural Haryana atotal village commun ity of this age wa s includedfor study. Th e prevalence (rate per 100 0) ofelectrocardiographic Q waves (Minnesota codesl-l, l-2) in men aged 40-59 years was 38 inChandigarh, 26 in Tecumseh, and 5 in ruralHaryana. In women the rates were 3, 4 and 3respectively. These results suggested thatmarked urban-rural differences in CHD ratesexisted in this part of India, and that prevalencein urban centres was similar to that in theU.S.A. at the time.

CORONARY RISK FACTORSIN SOUTH ASIANSThough the explanation of high CHD mor-tality in South A sians in one comm unity need

not nece ssarily be the same as that in another,the worldw ide distribution of increased risk inboth sexes from an early age suggests that theremay be a common underlying mechanism. Stud-ies which have compa red the distribution ofknown CHD risk factors between South A siansand other ethnic groups with lower CHD rateshave been reported from several countries: byreviewing them it is possible to reject severalhypotheses and to suggest a possible m echanismconsistent with most of the published findings.Smoking

Cigarette smoking is common among SouthAsian men overseas but unusual in women[13, 18,27,28 ]. In Trinidad the proportion ofmen who smoked 20 or more cigarettes a daywas higher in Indians than in Africans butsmoking habit failed to explain the difference inmortality between these two groups [13 ]. In asurvey in Fiji the proportion of men who ha dever smoked was less in Indian than Melanesianmen [29]. In South Africa during 1975-77 smok-ing rates were higher in South Asian thanEuropean men but much lower in South A sianthan European women. National householdsurvey data in Britain show lower smoking ratesin South Asian men than in the general popula-tion [30]: in this respect the heavy smok ing ofBangladeshi men both in the home country andin England is an exception [27,31,32].Haemostasis

Fibrinogen and factor VII coagu lant activity(VIIc) levels have been found to predict CHDand elevation of fibrinogen levels may me diatesome of the adverse effects of smoking [33].


4/13

60 0 P. M. MCKEIGUEet al.Levels of these factors ha ve been measu red intwo South Asian populations in London. Plas-ma fibrinogen and factor V IIc were similar inIndian (mostly Gujarati) and European men;Gujara tis had a slightly higher platelet cou ntand a lower mean platelet volume [34]. Bangla-deshi men and women had similar fibrinogenlevels to Europe ans but factor V IIc was lower inBangladeshi than European men [27].Hypertension

Hypertension is common in urban and ruralpopulations in the Indian subcon tinent [35,36].In a survey in Durb an, South Africa, hy perten-sion was found to be less common in Indiansthan in Africans or Europeans, though thedesign did not control for observer differences[37]. In Guyan a, blood pr essures were higher inAfrican men and women than in Indians: differ-ences which could be explained by the largerbody size of adults of African descent [38]. InTrinidad, adults of Indian descent ha d a similarprevalence of hypertension to those of Eu-ropean and mixed origin, and lower rates thanadults of African descent [39]. Age-adjusted cardiovascu lar mortality (deaths/l000person-years) among Trinidadian men with sys-tolic pressures below 130 mmH g was 11.3 in

Indians compared with 9.6 in Europeans, 5.1 inAfricans and 1.1 in those of mixed desc ent. InFiji, mean systolic pressures were similar inIndian and Melanesian men [29]. Higher bloodpressures were recorded in Punjabi w omen whohad migrated to England than in the homecountry [40] but two stud ies in which bloodpressures in South Asians and Europeans inEngland were compared directly have not foundblood pressures to be higher in South Asians. InNorth-West London m ean blood pressures weresimilar in Indians and Europeans [34] and inEast London mean systolic (but not diastolic)pressures were 10 mmH g lower in Bangladeshimen and women than in Europeans [27].Serum cholesterol concentration

Levels of smok ing, haem ostatic activity andhypertension do not account for the high CH Drates in South Asians overseas. A high averageserum cholesterol, in excess of 5.2 mm ol/l, isconsidered a factor necessary for the occurrencein a population of CHD on a mass scale [41].Values at or above this level have been reportedfor urban groups of relatively high socio-eco-nomic sta tus in various parts of India, but notin the middle and lower socio-economic groups(Table 1) [42-531 . Com parisons of this type are

Table 1. Survevs of serum cholesterol in India and Pakistan

Ref.Mean serumcholesterol (mmol/l)

AgeYear No. Place Population sampled range Males Females1969 42 Delhi High-income groupIndustrial workersRural workers1982 43 Delhi Middle to moderateincome1959 44 Uttar Pradesh Higher socio-economic

(Agra) Middle incomeLower income197 1 45 Uttar Pradesh Voluntary blood donors(Aliuarh) Paid blood donors

6. 230-59 4.54. 8311-50 4.5 4.25. 331-60 4.84. 5

55 - 4. 74. 01966 46 Bihar @am a) Economicallyprivileged Adults 4.91972 47 Bihar (Ranchi) Industrial employees:19761978

1980195619831982

upper social classes 35 - 5. 9lower social classes 4. 548 Rajasthan Academics(Bikaner) Low income 36-51 ;.;:49 Punjab Hospital outpatients:(Jullundur) higher social classes 30 - 5. 7middle/lower classes 4.4 4.450 Andhra Pradesh Middle and low income 31-60 4.6 4.3(Kakinada) groups51 Tamilnadu High socio-economic(Coonoor) Low socio-economic 4049 g52 Tamilnadu Not stated 30-39 5:7* 5.4.53 Pakistan Manual workers 3-8 4. 8(Peshawar)*Estimations on plasma or whole blood.


5/13

Coronary Heart Disease in SouthAsiansTable 2. Surveys of serum cholesterol in South Asians overseas

Mean serumCholesterol(mmovl)Ref. Ag eYear No. Place Population sampled range Males Fem ales

1959 7 Uganda GP attenders 40- 6.6(Kampala)1968 54 Guyana Lower socio-economic 3954 5.0 5.3(Annandale)1977 5 5 Trinidad Middle and lower 35-69 5.9 6.0(Port-of-Spain) so&-economic1980 53 Surinam Lower socio-economic 33-48 5.3(Nickerie)1985 29 Fiji Urban and rural 30-69 4.8*residents1987 71 Singapore Pre-employment 18-56 4.8screening1982 18 England Urban residents 25-65 5.0 4.3*(NW London)1988 34 England Urban residents 45-54 5.4(NW London)1988 27 England Urban residents 35-69 5.5. 5.4+(East London)

601

*Estimations on plasma.difficult to interpret since population samplingprocedures and laboratory techniques are fre-quently not comparable. Mean concentrationstend to be higher in Indians overseas than in thesubcontinent (Table 2) [7,18,27,29,34,53-551but not necessarily higher than in other ethnicgroups with lower CHD rates. In the early1980s mean serum cholesterol for men aged45-69 years was 5.5 mmol/l in the U.S.A. and6.3 mmol/l in Britain [56,57].Apart from a 1959 study of South Asian menattending general practitioners in Kampala,Uganda [7J,no group originating in the Indiansubcontinent has been reported to have amean serum cholesterol concentration similarto or above that of British men. In Trinidad,serum LDL cholesterol concentrations were0.3 mmol/l higher in Indian than in African menbut this did not account for the difference inCHD mortality. In Fiji, plasma total cholesterollevels were 0.3 mmol/l higher in Indian than inMelanesian men but there was no associationbetween plasma total cholesterol and the pres-ence of ECG abnormalities in Indian men. InNorth-West London and in East London,plasma cholesterol was 0.7 mmol/l lower inSouth Asian than in European men and women,in contrast to local mortality data which showedhigher rates in South Asians [18,21,27,34].Neither the levels of serum cholesterol prevail-ing in the home countries, nor the levels inmigrants to Britain, appear to explain the highCHD mortality of South Asians in England andWales.

Dietary fa tIf the Keys equation relating diet to average

serum cholesterol [58]holds, the low cholesterollevels found in South Asians overseas implieseither that saturated fat and cholesterol intakesare low or that polyunsaturated fat intakes arehigh in comparison with the diets of NorthEuropean and U.S. populations. The only over-seas South Asian populations for which detaileddietary survey data exist are those in Singaporeand North-West London. In an assessment ofdietary intakes by 24hour recall in 528 maleemployees of the Singapore Port Authority,mean total fat intake was highest in Chinese,intermediate in Indians and lowest in Malays[59]. Polyunsaturated fat as a proportion oftotal fat intake was similar in all three groups.In two studies, one using household food inven-tories [18] and the other weighed intakes [34]Gujaratis in North-West London were found tohave low dietary saturated fat intakes and highpolyunsaturated fat intakes compared with theBritish average. Average plasma cholesterol wasclose to the value predicted from the Keysformula [181.Total fat intake was similar to thelevels in the British population. Linoleic acidaccounted for most of the high polyunsaturatedfat intake of Gujaratis. As would be expected ina mainly vegetarian population, dietary intakesand levels in plasma lipids of long-chain polyun-saturated fatty acids of the 03 series were low:these fatty acids are mainly of marine origin andit has been suggested that they protect against


6/13

60 2 P. M. MCKEIGLJEet al.atherosclerosis. Levels of 03 fatty acids inMuslim Ban gladeshis in East London, however,are high, presumably reflecting a fish-containingdiet [27].Ghee, pre pared by heating butter to drive offthe water, is traditionally used in North Indiancooking. One analysis has shown that 12% ofthe sterols in ghee obtained from com mercialand home-prepared sources were in the form ofcholesterol oxides: these comp ounds were notfound in ordinary butter [60]. Since there isevidence from animal studies that cholesteroloxides are more atherogenic than pure choles-terol [61], the presence of these compounds inghee was suggested as a possible cause of highCHD rates in South Asians in Britain andTrinidad [60]. Others have failed to confirm thepresence of cholesterol oxidation products inghee [62]. Ghee is not widely u sed by Indians inTrinidad, or Bangladeshis in London.

Plasma HDL cholesterol and triglyceride levelsLow plasma levels of high-density-lipoprotein

(HDL ) cholesterol [63,64] and high levels oftriglyceride [65-68] are predictors of CHD riskin prospective studies an d have been foundfairly consistently in South A sians overseascompared with other groups. In urban Trinidad,age-adjusted mean HDL cholesterol concentra-tions in Indian men (0.9 mmol/l) and w omen(1.2 mmol/l) were 0.1 mmol/l lower and tri-glyceride levels were 40% higher than inAfricans. These differences did not explain fullythe excess of CHD deaths in Indian adults inthis community l-551. In the U.S . also In dianmigra nts have been reported to have lowerHDL ch olesterol and higher triglyceride levelsthan the general population [69,70]. In Fijiplasm a triglyceride levels were 30% higher inIndians than in Melanesians: plasma HD Lcholesterol levels were not measu red [29]. InSingapore levels of HDL cholesterol, apolipo-protein A-I and apoplipoprotein A-II werelower in Indian than in Chinese men attendingfor pre-employm ent medica l screening [71].Howeve r in Malay men levels of these lipo-proteins were similar to the levels in Indianseven though Indians have more than twice theCHD mortality of Malays in Singapore.In North-West London, HDL cholesterollevels were 0.1 mmol/l lower in South Asian(mainly Gujarati) than in European men but theratio of HDL to total cholesterol did not differsignificantly between these groups [34].

Triglyceride levels were lower in South Asianthan in European men, in contrast with findingsin other populations. In East London HDLcholesterol levels were 0.3 mm ol/l lower inBangladeshi men and women than in Europeansand the HDL /total cholesterol ratio was alsolower [27]. Triglyceride levels after a glucoseload were 50% higher in Bangladeshis. Inneither G ujaratis nor Bangladeshis w as there asex difference in HDL cholesterol [18,271.Absen ce of a sex difference in HDL cholesteroland plasma apoprotein A-I concentrations hasbeen reported in other Indian com munities[43,52] and is unlike the usual finding in Euro-pean communities at high CHD risk [63].There is an inverse association between H DLcholesterol and triglyceride levels [63,72] so thatit is difficult to distinguish on the bas is ofepidemiologic data the extent to which lowHDL cholesterol and high triglyceride levelsmay be separately atherogenic. Both factors areassociated with non-insulin-dependent diabetesin European populations [73].

DIABETES AND IMPAIREDGLUCOSE TOLERANCE

Non-insu lin-depende nt diabetes mellitus is adisease of antiquity in India and ha s long beenrecognized as prevalent in the middle-ag ed,affluent and obese [74,75]. Among overseasIndians the first report of an excess of diabetesin comparison with other ethnic groups w asfrom Saigon in 1913 [76]. Similar repo rts havecome more recently from Malaysia [77],Trinidad [78], South A frica [79], Singapore [80],Fiji [81] Surinam [82] and England [83]. Preva-lence estimates from published surveys in SouthAsian populations [13,27,77,81,84-871 areshown in Table 3 together with data from theU.S. National Health and Nutrition Examina-tion Survey for compa rison [88]. To allow com-parison between recent surveys based on the1980 WH O Expert Com mittees diagnosticcriteria [89] and earlier studies using a 50 gglucose load, prevalence figures based oncomp arable cut-off values for 2-hour bloodglucose have been extracted from the earlierstudies where sufficient data are given. Thelow prevalence rates reported in East Paki-stan an d Orissa up to 1971 [84] are in strikingcontrast to later findings in India a nd overseas[13,27,81,86,87]. Comparison of urban andrural rates in all individuals over 10 years of ageis possible from the published Orissa data: 2%


7/13

Coronary Heart Disease in South Asians 60 3Table 3. Surveys of diabetes prevahmce in South Asians

Ref.Year No. Place Populationsampled2-hour samplecut-off value

@mol/B Sex Ag erange Prevalenceusing a1966 5O g glucose load17 EastPakistan Urban and rural Whole bloodresidents alucose > 9.51971 84 Orissa Urban and rural Whole bloodresidents glucose > 9.51973 85 Calcutta Low-income Whole bloodoutpatients glucose > 10.5or >7.8197 7 13 Trinidad Urban residents Whole bloodglucose j 8.9Using a 75g glucose load and 2-hour plasma glucose > Il.Ommol~l1983 81 Fiji Urban and ruralresidents1985 86 Durban Residents198 8 87 Karnatka Urban residents

198 8 27 London Residents

For comparison198 7 88 U.S.A. WhitesBlacks

M&F 30-M&F 30-M&F 30-

MF 35-69

MF 35 -M&F 30-M&F :z

MF 35-69M&F 45-64

1%2%

5-15%.

19 %22 %25 %22 %22 %9%29 %22 %23 %10 %18 %

*An interval is aiven since break-noin ts in the original table lie either side of the cut-off value thatwould be equivalent to 19 80 WH O criteria.

in urban Cuttack and 0.5% in rural Badachanawere .diabetic by the criteria in Table 3. In theFiji survey the higher prevalence of diabetes inIndian m en than M elanesian men was not ex-plained by obesity (m easured as triceps skinfoldthickness) or physical activity [90 ].

Preliminary analyses of the follow-up data inthe Trinidad study suggested that age - anddiabetes-ad justed relative risks of cardiovascu-lar death were similar in men of Indian andEuropean descent, and that diabetes mellitusmight explain the difference in CHD rates be-tween these groups [13]. A more recent a nalysisof cardiovascular morbidity and mortality failedto confirm this (unpub lished). However theEuropean group in this study was small: similarcomp arisons would be easier to perform inBritain. Diabetes mellitus did not explain morethan a small part of the excess risk of CHD inIndians above that of Africans and men ofmixed descent in Trinidad: the lowest mortalitywas in men whose fasting glucose was in therange 4.2-4.6 mmol/l and in this stratum age-adjusted cardiovascu lar m ortality (deaths/1000person-years) was 6.7 in Indians, 3.4 in Africansand 0 in men of mixed descent.Quan titative consideration of the relationshipbetween glucose intolerance and CHD m ortalitysuggests that markedly elevated CHD risk foran entire population is not easily explained bya high prevalence of glucose intolerance.. The

relationship between plasma glucose and mor-tality is non-linear and there-appears to be athreshold for-elevation of CH D risk, at a 2-hourglucose level approximately equivalent to theWHO criterion for impaired glucose tolerance[91,92]. In the Whitehall Study, age-adjustedCHD mortality in men in the top 5% of the2-hour blood glucose distribution, and in afurther 1% who were known diabetics, wastwice that in the rest of the cohort [91]. Thisimplies that the proportion of all CH D deathswhich are attributable to glucose intolerance-the aetiologic fraction-is about 6% in thispopulation. If this relationship between glucoseintolerance and mortality holds across popula-tions then even an exceptionally high prevalenceof diabetes a nd impaired glucose tolerance inSouth Asians overseas would produce only amodest elevation of total .CHD m ortality. A30% prevalence of glucose intolerance in middleage, instead. of 6% a s in Whitehall, woulddouble the risk for 30% rather than 6% of thepopulation but this wou ld yield only a 23%excess total mortality, other risk factors be ingequal. On this basis it is unlikely that a highprevalence of diabetes an d impaired glucosetolerance could alone explain a threefold rela-tive risk of CHD death in Indians comparedwith Chinese in Singapore and in Indians com-pared with Melanesians in Fiji. In any case therelationship between plasma glucose and CHD


8/13

604 P. M. MCKEIGUEet al.mortality is unlikely to be directly ca usal a nd itis more likely that both have common under-lying determin ants [93].

HYPERINSULINAEMIA ANDINSULIN RESISTANCE

In population surveys low HDL , high tri-glycerides and non-insulin-dep endent diabetesare associated w ith hyperinsulinaemia [94-961.Hype rinsulinaemia increases VLD L triglyceridesynthesis [94,97] and this may lead to lowerHDL cholesterol levels [72]. The consistentfindings of low HDL cholesterol, high trigly-cerides and high prevalence of non-insulin-dependent diabetes in South A sians overseaspoint strongly to an underlying state of insulinresistance [27]. This hypothesis is supported bythe few studies in which insulin levels have beenmea sured. In South A frica, insulin levels aftera glucose load w ere higher in students [98],children [99] and nurse s [loo] of Indian originthan in those of Europea n origin. In EastLondon insulin levels after a glucose load w eretwice as high in South Asians (predominantlyBangladeshi) as in non-Asians [27]. Similarfindings have been reported for hospital out-patients in We st London [loll and in a studyof vegetarians in Washington, D.C. [102].The definitive demonstration of insulin resis-tance depends on steady-state measurements ofplasma insulin levels and glucose disposal butfrom the co-existence of hyperinsulinae mia, ele-vated insulin-glucose ratio and a high diabetesprevalence it is resonable to infer that insulinresistance is present in Bangladeshis in EastLondon, and probably in South Asians overseasgenerally.

Elevated plasma insulin has been found to bean independe nt predictor of CHD in threeprospective studies [10 3-105 1: it may exertatherogenic effects either directly or throughdisturbanc es of HDL and triglyceride levels[106]. Some of the association of hypertensionand non-insulin-dependent diabetes with CHDrisk may also be med iated by hyperinsulin-aemia . In the absence of any explanation basedon more well-established risk factors, insu linresistance is a plausible mech anism for the highrates of CHD in South Asians overseas [27].Similar findings in other populations may berelevant in this respect. E dinburgh men in earlymiddle age have a threefold higher m ortalityfrom ischaemic heart disease than men in Stock-holm: a study of healthy men in these two cities

found that the Scottish men had lower HDLcholesterol, higher triglyceride levels and higherlevels of insulin after a glucose load than theSwedish m en [107]. In Pimas, whose prevalenceof diabetes exceeds even that in South Asiansoverseas, sim ilar lipoprotein findings and hyper-insulinaemia have been reported [108,109]. NoCHD mortality data for Pimas have been re-ported but CH D prevalence in life and at post-mortem appears to be lower in Pimas than in thegeneral population of the U.S. [l lo]. Althoughthis finding appears inconsistent with the insulinresistance hypothesis, it may possibly b e ex-plained by the much lower average plasma totalcholesterol and LDL cholesterol levels in Pima sthan in the general po pulation of the U.S.A .[108].Insulin resistance is strongly associated withobesity, and particularly with the relative pro-portion of body fat deposited intra-abdom i-nally: it has been suggested that elevated levelsof free fatty acids released by omen tal adipo-cytes are responsible for inhibiting glucose up-take in peripheral tissues [l 111.There are nopublished data on the extent to which thepattern of adipose tissue distribution differsbetween South Asians and other ethnic groups.In seeking causes of insulin resistance in SouthAsia ns it is of crucial im portance to determinewhether it can be explained on the basis of bodyfat patterning.

PSYCHOSOCIAL FACTORS

Historically CHD has tended to appea r firstin the more affluent socio-economic groups, a ndthen to spread into the rest of the population,eventually be coming c omm onest in those oflowest socio-economic status [112, 1131. Theabsence of a social class gradient in the CHDmortality of South A sians in Britain ma kes itunlikely that excess risk in this group is relatedto present economic status, or to work p atternscharacterized by low control and high demandwhich have been associated with CHD risk[I 141. It has been suggested that poverty anddeprivation in early life lead to increased CHDrisk in middle age: this is supported by therelationship of short stature to CHD [I 151 andby the relationship between the infant mortalityand subsequ ent CHD m ortality of differentbirth cohorts [116]. No similar data for SouthAsia ns are available: it would be of interest toexamine the relationship between height andcardiovascu lar mortality in existing datasets an d


9/13

Coronary Heart JXsease.in South Asians 60 5life insurance records. .Early deprivation is un-likely to be a sufficient explanation for highCHD rates in South Asians in Britain, sinceAfro-Caribbeans and others who have migratedfrom less developed regions hav e low CH Dmortality rates [151.Emphasis has been placed on the role ofpsycho-social stress associated with migrationand racial tension in the high CHD risk amongSouth Asia ns in Britain [117]. The limited evi-dence available does not support the view thatSouth A sians are experiencing higher levels ofstress than the native British population. In apopulation survey Indian imm igrants reportedfewer psychological symptom s than nativeBritish urban residents [I 181; this is consistentwith the effects of selection a t migration. Theinverse relationship between symptom scoresand social class in British women was reversedin Indian wom en. Psychological symptomscores in Pakistani immigrants were similar tothose in the native British population [119]. It ispossible that standar d questionnaires fail todetect psychological morbidity in a differentcultural setting from that for which they wereoriginally developed. Qu alitative s tudies areneeded to resolve this uncertainty. Lack ofsocial support has been found to predict C HDin prospective studies [ 120, 1211 but survey datado not suggest tha t social isolation is particu-larly common among South Asians in Britain[122]. Adve rse psycho-social effects of migrationand reception by the host country may con-tribute to risk in South Asia ns in Britain butthese factors cannot account for high CHDrates in long-established Indian com munitiessuch as those in the Caribbean and Pacificregions.

GENETIC INHERITANCEAny environme ntal factor invoked to explainthe high rates of the disease in South Asianpopulations around the world must be commonto all the main ethnic groups of the Indiansubcontinent, to both sexes and persist severalgenerations after migration. In other migran t

groups, such as Japanese in the U.S.A., CH Drates in those settled ov erseas longest an d intheir descendants converge to those in the hostpopulation [123]. This ha s led to the view thatbetween -population differences in CH D ratesare environmentally rather than geneticallydetermined. In contrast C HD mortality anddiabetes prevalence in comm unities of Indian

descent settled overseas for many generationscontinue to diverge from the rates in otherethnic groups in the same country. Indian popu-lations overseas originate from several differentparts of the subcontinen t and are likely to begenetically dissimilar: it is not obvious how theconsistent findings of high rates of diabetes andCHD can be reconciled with either environmen-tal or genetic explanations. A report of a rela-tion between histocompatibility (HLA) profileand CHD risk [124] has not been confirmed[125, 1261. In India, the gene frequencies of classI and class II antigens are similar to those inEuropea ns [1271 and the frequency profile inyoung Indian men with a history of myocardialinfarction has not been found to be distinctive.Sibling and twin studies have pointed to theimportance of genetic factors in the aetiology ofnon-insulin-dependent diabetes [128,129] butgenetic markers for this condition have not beenconsistently identified [1 30]. A selective advan-tage for the diabetic genotype under conditionsof food scarcity has been suggested to explainthe high prevalence of the condition in popula-tions exposed to such conditions in earliergenerations who have undergone a recent trans-ition to relative affluence [ 13 11. Further progressin this field will depend upon identification ofnew genetic markers for diabetes and coronaryheart disease.

CONCLUSIONCurrent recommendations for CHD preven-

tion emph asize reduction of dietary fat andserum cholesterol, avoidance of smok ing andcontrol of hypertension. Wh ile these factorsmay predict CHD risk in Indians as in otherethnic groups they do not explain the excessrates in South Asians overseas. The evidencereviewed here suggests that insulin resistancemay be common in South Asian populationsoverseas, leading to the high prevalence ofnon-insulin-depe ndent diabetes and other meta-bolic disturbanc es that are possibly responsiblefor the high rates of CH D. The apparen t rarityof CH D in rural In dia imp lies that the diseaseis as preventable in Indian as it is in Europea npopulations, but identification of appropriatepreventive strategies w ill depend on furtherresearch. The available evidence does notprovide a basis for recomm ending dietarychange to South A sians in Britain: data onaverage dietary intake are available only forGujara tis and even for this group it is not clear


10/13

606 P. M. MCKEIGUEer al.what dietary changes would be appropriate toachieve further re duction in serum cholesterolor how effective this would be in reducing CHDrisk. The importance of smok ing as an individ-ual risk factor is not in doubt but measu res toreduce the habit w ill have only mode st impacton total mortality in South A sian men and verylittle effect in South Asian women who rarelysmoke. Even if antihypertensive treatment wereeffective in reducing CHD risk the widespreaduse of agen ts containing thiazides could increa serisk by further impairm ent of glucose tolerancein a group already at high risk for diabetes. Ifinsulin resistance is widespread among SouthAsians overseas, as this review suggests, then forthose at risk ameliorative measures such asincreased physical activity and control of ob-esity may be the most effective means of pre-venting CH D. Population studies are under waywhich m ay elucidate the possible role of insulinresistance and related metabolic d isturbances inSouth Asians overseas and identify possibilitiesfor modifying CHD risk. Further studies ofCHD rates in urban and rural populations inIndia are also planned by the Indian Council ofMedical Research.

1.

2.3.4.5.

6.

7.

8.

9.

10 .

11 .

12 .

REFERENCESKondapi C. Indians Overseas 18 38-19 49. New Delhi:Indian C ouncil of World Affairs; 1951 .Tinker H. A New System of Slavery. London: OxfordUniversity Press; 1974.Lomas GB. Census 1971: the Coloured Population ofGreat Britain. London: R unnymede Trust; 1974.Rose EJB. Colour and Citizenship. London: OxfordUniversity Press; 1969.Danaraj TJ, Acker MS, Danaraj W, Ong WH, YamTB. Ethnic group differences in coronary heart dis-ease in Singapore: an analysis of necropsy records.Am Heart J 1959; 58: 516-526.Chen AJ. Recent trends in the mortality and mor-bidity of cardiovascular diseases. Ann Acad MedSingapore 1980; 9: 411-415.Shaper AG, Jones KW . Serum cholesterol, diet andcoronary heart disease in Africans and Asians inUganda. Lancet 1959; 2: 534-537.Adelstein AM. Some aspects of cardiovascular mor-tality in South Africa. Br J Prev Sot M ed 1963; 17:29-40.Wyndham CH. Trends with time of cardiovascularmortality in the populations of the RSA for theperiod 1968197 7. S Afr Med J 1982; 61: 987-993.Sorokin M. Hospital morbidity in the Fiji islandswith special reference to the saccharine disease. S AfrMed J 1975; 49: 1481-1485.Tuomilehto J, Ram P, Eseroma R, Taylor R , ZimmetP. Card iovasc ular diseases in Fiji: analysis of mor-tality, morbidity and risk factors. Bull WH O 1984 ;62: 133-143.Miller GJ, Beckles GLA, Alexis SD, Byam NT, PriceSGL. Serum lipoproteins and susceptibility of men ofIndian descent to coronary heart disease. The StJames survey, Trinidad. Lancet 1982 ; 2: 200-20 3.

13 .

14 .

15 .

16 .

17 .

18 .

19 .

20 .

21 .

22 .

23 .

24 .

25 .

26 .

27 .

28 .29 .

30 .

31 .32 .

33 .

Beckles GLA, Miller G+I,Kirkwood BR, Alexis SD,Carson DC, Byam NTA. High total and cardiovascu-lar disease mortality in adults of Indian descent inTrinidad, unexplained by major coronary risk fac-tors. Lancet 1986; 1: 1298-1301.Tunstall Pedoe H, Clayton D, Morris JN, Brigden W,McDonald L. Coronary heart attacks in EastLondon. Lancet 1975; 2: 833-838.Marmot MG, Adelstein AM, Bulusu L. ImmigrantMortality in England and WaIes 1970-78. OPCSStudies of Medical and Population Subjects No 47.London: HMSO; 1984.Balarajan R, Adelstein AM, Bulusu L, Shukla V.Patterns of mortality am ong m igrants to England andWales from the Indian subcontinent. Br Med J 1984;289: 1185-l 187.Donaldson LJ, Taylor JB. Patterns of Asian andnon-Asian morbidity in hospitals. Br Med J 1983;286: 949-95 1.McKeigue PM, Marmot MG, Adelstein AM et al.Diet and risk factors for coronary heart disease inAsians in north-west Lond on. Lancet 1985; 2:10861090.Lawrence RE, Littler WA. Acute myocardial infarc-tion in Asians and whites in Birmingham. Br Med J1985; 290: 1472.Lowry PJ, Glover DR, Mace PJ, Littler WA. Coron-ary artery disease in Asians in Birminghm. Br HeartJ 1984; 52: 610-613.McKeigue PM, Marmot MG. Mortality from coron-ary heart disease in Asian comm unities in London. BrMed J 1988; 297: 903.Agarwal BH. A reappraisal of pattern of cardiovascu-lar disorders in India. J Indian Med Asaoc 19 75; 65:125-129.Malhotra SL. Epidemiology of ischaemic heart dis-ease in India w ith special reference to causation. BrHeart J 1967; 29: 895-905.Sarvotham SG, Berry JN. Prevalence of coronaryheart disease in an urban p opulation in northernIndia. Circulation 1968; 37: 939-953.Dewan B D, Malhotra KC, Gupta SP. Epidemiologi-cal study of coronary heart disease in a rural commu-nity in Haryana. Indian Heart J 1974; 26: 68-78.Ostrander LD, Bradt RL, Kjelsberg M O, EpsteinFH. Electrocardiographic findings among the popu-lation of a total natural co mmunity: Tecumseh,Michigan. Circulation 1965 ; 31: 888-89 3.McKeigue PM, Marmot MG, Syndercombe CourtYD, Cottier DE, Rahman S, Riemersma RA. Dia-betes, hyperinsulinaemia and coronary risk factors inBangladeshis in East London. Br Heart J 19 88; 60:39&396.Yach D, Townshend GS. Smoking and health inSouth Africa. S Afr Med J 1988 ; 73: 391-39 9.Sicree RA, Tuomilehto J, Zimmet P et al. Electro-cardiographic abnormalities amongst Melanesianand Indian men of Fiji: prevalence and associatedfactors. Int J Cardlol 1988 ; 19: 27-38.Balarajan R, Yuen P. British smoking and drinkinghabits: variation by country of birth. Comm m~ Med1986; 8: 237-239.Cohen N. Smoking, health and survival: prospects inBangladesh. Lancet 1981; 1: 1090-1093.Silman A, Loysen E, de Graef W, Sramek M. Highdietary fat intake and cigarette smoking as riskfactors for ischaemic heart disease in Bangladeshimale immigrants in East London. J EpidemiolCommun Health 1985; 39: 301-303.Meade TW , North W RS, Chakrabarti R et al.Haemostatic function and cardiovascular death: earlyresults of a prospective study. Laaeet 1977; 1:1050-1054.


11/13

60 875 .

76 .

77 .

78 .

79 .

80 .

81 .

82 .

83 .

84 .

85 .

86 .

87 .

88 .

89 .90 .

91 .

92 .

P. M. MCKBIGUEer al.McCay D, Banerjee SC, Ghosal LM, DuttaMM, Roy C. Observations on the sugar of th eblood and the sugar in the urine in varying conditionsof health in the Bengali. Bu llan J M ad Res 1916 ; 4:l-27.Monte1 MLR. Une question sur le diabete. In:Cemptas Per&m &I Travanx da TmBi&ns Co&&sBkenlaI tenu & Balgon. Saigon: Fa r Eastern Associa-tion of Tropical Medicine; 1913 : 522-52 3.West KM, Kalbfleisch JM. Glucose tolerance, nutri-tion and diabetes in Uruguay, Venezuela, Malay a andEast Pakistan. Diabetes 1966 , 15: 9-18.Poon-King T, Henry MV, Rampersan F. Prevalenceand natural history of diabetes in Trinidad. Lancet1968; 1: 155-160.Marine N, Vinik AI, Edelstein I, Jackson WPU.Diabetes, hyperglycaemia and glycosuria among In-dians. Malavs. and Africans (Bantu) in Caoe Town .South Africa. Dmhetes 1969 ; 18: 84& 857. -Cheah JS, Lui KF, Yeo PPB, Tam BY, Tan YT, NgYK. Diabetes mellitus in Singapore: results of acountry-wide population survey. In: Ahuja MM S,Ed. Epidemiology of Diabetes in Developing Ceuntrles.New Delhi: In&print; 1979 : 93-102 . _ _Zimmet P. Tavlor R. Ram P et al. Prevalence ofdiabetes and impaired glucose tolerance in the bira-cial (Melanesian and Indian) population of Fiji: arural-urban comparison. Am J Epidemlol 1983; 118:673688.Schaad JDG , Terpstra J, Gemrawsingh I, Nieuwen-huijzen Kruseman AC, Bouwhuis-Hoogerwerf ML.Diabetes prevalence in the three main ethnic groupsin Surinam (South America): a population survey.Netb J M ed 1985; 28: 17-22.Mather HM, Keen H. The Southall Diabetes Survey:prevalence of known diabetes in Asians and Euro--peatis. Br Med J 1985; 291: 1081-1084 .Trioathv BB. Panda NC. Tei SC. Sahoo GN. Kar BB .S&vey for detection of glycosuria, hyp erglycaemiaand diabetes mellitus in urban and rural areas ofCuttack district. J Asset Physklans India 1971 ; 19:681-692.Murkerjee AB, Sen S, Dey P. Epidemiological surveyof diabetes m ellitus in a-mixed popu lation of Cai-cutta. J Bullan Med Assee 1973 : 61: 17-22.Gmar MAK, Seedat MA, Dyer RB, Rajput MC,Motala AA, Joubert SM. The prevalence of diabetesmellitus in a large group of South African Indians.S Afr Med J 1985; 67: 924-926.Ramacha ndran A, Jali MV, Mohan V, Snehalatha C,Visnawathan M. High prevalence of diabetes in anurban population in south India. B r Med J 198 8; 297:587-590.Harris MI, Hadden WC, Knowler WC, Bennett PH.Prevalence of diabetes and impaired glucose toleranceand plasma glucose levels in U.S. population aged20-74 yr. M&es 1987; 36: 523-534. - -WH O Exwrt Committee on Diabetes Mellitus. Set-ond Report. WHO Tech Rep Ser 1980; 646: l-80.Taylor R, Ram P, Zhnmet P, Raper LR, Ringrose H.Physical activity and prevalence of diabetes inMelanesian and Indian men in Fiji. Mabstologlr1984; 27: 578-582.Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H.Coronary heart disease risk and impaired glucosetolerance: the Whitehall Study. Lancet 1980; 1:1373-1376.Pyorala K , Savolainen E, Kauko la S, Haapak oski J.Plasma insulin as coronary heart disease risk factor:relationship to other risk factors and predictive valu eduring 9 l/2-year follow-up of the Helsinki PolicemenStudy population. Acta Med ScaruJ 1985 ; 701(Suppl.): 38-52.

93 .

94 .

95 .

96 .

97 .

98 .

99 .

100.

101.

102.

103.

104.

105.

106.107.

108.

109.

110.

Jarrett RJ. Type 2 (non-insulindependent) diabetesmellitus and coronary heart di -hicken,neither? Di&teb& 1984 .26: 99-102 . egg orOkfsky JM, Farquhar JW, Reaven GM, Reappraisalof the role of insulin in hypertriglyce ridemia. Am JMed 1974; 57: 551-560.Stalder M, Pometta D, Suenram A. Relationshipbetween high plasma insulin levels and high densitylipoprotein cholesterol levels in healthy m en. Ma-het&gla 1981; 21: 54&548.Orchard TJ, Becker DJ. Bates M, Kuller LH, DrashAL. Plasma insulin and lipoprotein concentrations:an atherogenic association? Ant J Epidemlel 1983 ;118: 326-337.Reaven GM, Lemer MP, Stem MP, Farcmhar JW.Role of insulin in endogenous hypertriglyc&laemia.J Clin Invest 1967 : 46: 175 61767.Keller P, Schatx K, Jackson WPU. Immunoreactiveinsulin in various South A frican population groups.S Afr Meal J 1972; 46: 15 2-157.Walker ARP, Bernstein RE, Du Plessis I. Hyper-insulinaemia from glucose dose in South AfricanIndian children. S Afr Med J 1972 ; 46: 1916.Notelovitx M . An interracial com parison of insulinsecretion in normal non-pregnant women. S Afr BledJ 1976; 50: 12141217.Mohan V, Sharp PS, Cloke HR, Burrin JM, SchumerB, Kohner EM. Serum immunoreactive insulinresponses to a ghtcose load in Asian Indian andEuropean Type 2 (non-insulin-dependent) diabeticpatients and control subjects. Diahetobgla 1986 ; 29:235-237.Schollield DJ, Beh all KM, Bhathena SJ, Kelsay J,Reiser S, Revett KR . A study on Asian Indian andAmerican vegetarians: indications of a racial pm-disposition to glucose intolerance. Am J CBn Nutr1987; 46: 955-961.Pyorala K. Relationship of glucose tolerance an dplasma insulin to the incidence of coronary heartdisease. Results from two popu lation studies inFinland. Mabetes Care 19 79; 2: 131-14 1.Ducimetiere P, Eschwege E, Papoz L, Richard JL,Claude JR , Rosselin G . Relationship of plasma insu-lin levels to the incidence of myoc ardiil infarctionand coronary heart disease mortality in a middle-agedpopulation. Dlahetelegla 1980 ; 19: 205-21 0.Welbom TA, Weame K. Coronary heart disease-incidence and mortality in Busselton with reference toglucose and insulin levels. Mabetw Care 1979 ; 2:154-160.Stout RW. Insulin and atheroma-an update. Lancet1987; 1: 1077-1078.Logan R L, Riemersma RA, Thomson M et al. Riskfactors for ischaem ic heart disease in norm al m enaged 40. Edinburgh-Stockholm Study. Lancet 1978;1: 949-954.Howard BV, Davis MP, Pettitt DJ, Knowler WC,Bennett PH. Plasma and lipoprotein cholesterol andtriglyceride conce ntrations in the Pima Indians: distri-butions differing from those of Caucasians. ClreuIa-tlon 1983; 68: 714-724.Aronoff SL, Bennett PH, Gorden P, RushforthN, Miller M. Unexplained hyperinsulinaemia innormal and prediabetic Pima Indians comparedwith normal Caucasians. An example of racial differ-ences in insulin secretion. Diabetes 1977; 26:827840.Ingelfinger JA, Bennett PH, Liebow IM, Miller M.Coronary heart disease in the Pima Indians. Electro-cardiographic findings and postmortem evidence ofmyocardial infarction in a population with a highprevalence of diabetes mellitus. Dlaketa 1976 ; 25:561-565.


12/13

Coronary Heart Disease in South Asians 60 734 .

35 .

36 .

37 .

38 .

39 .

40 .

41 .42 .

43 .

44 .

45 .46 .

47 .

48 .

49 .

50 .

51 .52 .

53 .

54 .

55 .

Miller GJ, Kotecha S, Wilkinson WH et al. Dietaryand other characteristics relevant for corona ry h eartdisease in men of Indian, West Indian a nd Europeandescent in London. AtheroscIemeIs 1988 ; 70: 63-72.Gupta SP, Siwach SB, Gupta MS. Hypertension andblood pressure trends in the general population ofHaryana (based on total community surveys). J AsseePhysicians India 1979 ; 27: 119-12 6.Islam N, Khan M, Latif ZA. Hypertension in therural population of Bangladesh-a preliminary sur-vey. Bangladesh Med Res Coonc Boll 1983; 9: 1I-14.Scedat YK, Seedat MA. An inter-racial study of theprevalence of hypertension in an urban S outh Africanpopulation. Trrns R Sot Trop Med Hyg 19 82; 76:62-7 1.Florey C Du V, Ashcroft MT, M iller GJ. Bloodpressure levels of Guyanese adults of African andIndian origin. Ant J Epidemiol 1 971; 94: 419-42 4.Miller GJ, Beckles GLA, Byam NTA, Kirkwood BR,Carson DC. Blood pressure, ethnic group and all-cause mortality in urban Trinidad. W Ind Med J1985 ; 34 (Suppl.): 4 2.Keil JE, Britt RP, Weinrich MC, Hollis Y, Keil BW.Hypertension in Punjabi females: a comparison be-tween migrants to London and natives in India. HumBiol 1980; 52: 423-433.WH O Expert Committee. Prevention of coronaryheart disease. WH O Tech Rep Ser 198 2; 678: l-53.Padmava ti S, Gupta S, Pantalu GV. Dietary fats,serum cholesterol levels and incidence of athero-sclerosis in Delhi. Circulation 1959; 19: 849.Gandhi BM. Lipoprotein composition of normalhealthy subjects in northern India. Indlao J Med Res1982; 75: 393401.Mathur SK, Wahi PN, Malhotra KK, Sharma RD,Srivastava SK. Dietary fats, serum cholesterol andserum lipids in different socioeconom ic groups inUttar Pradesh. J Indian Med Assoc 1959 ; 33: 303-30 9.Tyagi SP. Serum cholesterol and ABO blood group.J Assoc Physicians India 1971 ; 19: 257.Srivastava BK, Sinha AS. Observations on serumcholesterol and lipid phospho rus levels in relation toABO blood groups. J Iodian Med Assoc 1966; 47:261-262.Ban B. Serum cholesterol levels in male industrialpopulation of Heavy Engineering Corporation. JA& c Pkysiciaos India 1972; 20: 495-499.Raman BN. Vii SC. Maitrva BB. Blood cholesteroland serum magnesium in health and ischaemic heartdisease. J. Indian Med Assoc 1 976; 66: 124-126 .Werner G T, Sareen DK. Serum cholesterol levels inthe population of Punjab in north-west India. Am JChin Nutr 1978; 31: 1479148 3.Rao PN, Sastry NS . Serum cholesterol levels of malesand females in different age groups in South India.Am J Clln Nutr 1980; 33: 181-182.Gopalan C, Ramathan KS. Fats and disease. Lancet1956; 2: 1212-1213.Shanmu gasundaram KR, Suresh S, Misra KP,Jayakrishan TK. Plasma lipoprotein cholesterol inIndia in healthy persons and those with coronaryheart disease. Atherosclerosis 1983 ; 46: 129-135 .Knuiman JT, West CE, Burema J. Serum total andhigh density lipoprotein cholesterol con centrationsand body mass index in adult men from 13 countries.Am J Epidemiol 1982; 116: 631-642.Ashcroft M T, Beadnell HMSG , Bell R, Miller GJ.Charac teristics relevant to cardio vascula r diseaseamong adults of African and Indian origin inGuyana. BuIl WHO 1970; 42: 205-223.Miller GJ, Beckles GLA, Byam NT A er al. Serumlipoprotein concen trations in relation to ethnic com-position and urbanization in men and women of

56 .

57 .

58 .

59 .

60 .

61 .

62 .63 .

64 .

65 .

66 .

67 .

68 .

69 .

70 .

71 .

72 .

73 .

74 .

Trinidad, West Indies. Iat J Epidemiol 1984; 13:413-421.US Department of Health and Hum an Services. TheLipid Research Clioics Populatioo Stodies Data Book,Vol. 1. The Prevalence Study, NIH Pu blication No.80-1527. Washington: NIH; 1980.Thelle DS, Shaper AG, Whitehead TP, Bullock DG,Ashby D, Pateil. Blood lipids in middle-aged Britishmen. Br Heart J 1983: 49: 205-213.Keys A, Anderson JT,.Grande F. Serum cholesterolresponses to changes in the diet. III. Differencesamong individuals. Metahollsm 1965 : 14: 766-77 5.Tan PY, Ng TB, Saha N. Dietary intakes of threeethnic groups of Singapore. Proe Nu tr SIX 1984; 43:135A.Jacobson MS. Cholesterol oxides in Indian ghee:possible cause of unexplained high risk of athero-sclerosis in Indian immigrant populations. Laocet1987; 2: 656658.Imai H, Werthessen NT, Taylor CB, Lee KT. Angio-toxicity and arteriosclerosis- due to contam inants ofUSP-nrade cholesterol. Arch Pathol Lab Med 1976:loo: 365-572.Surendra Nath B, Ram a Murthy NK. Cholesterol inIndian ghee. Lmcet 1988; 2: 39.Miller GJ, Miller NE. Plasma high-density lipo-protein concentration and development of ischaemicheart disease. Laocet 1975; 1: 1619.Gordon T, Castelli WP, Hjortland MC, Kannell WB .High density lipoprotein as a protective factor againstcoronary heart disease. The Framingham Study. AmJ M ed 1977; 62: 707-714.Pelkonen R, Nikkila EA, Koskinen S, Penttinen K,Sarna S. Association of serum lipids and obesity withcardiovascular mortality. Br Med J 19 77; 2:1185-1187.Carlson LA, Bottiger LE, Ahfeldt PE. Risk factorsfor myocardial infarction in the Stockholm prospec-tive study: a 14year follow-up focusing on the role ofplasma triglycerides and cholesterol. Acta Med Scaod1979; 206: 351-360.Petersson B, Trell E, Hood B. Premature death andassociated risk factors in urban middle-aged men. AmJ Med 1984; 77: 418426.Cambien F, Jacqueson A, Richard JL, Wamet JM ,Ducimetiere P, Claude JR. Is the level of serumtriglyceride a significant predictor of corona ry deathnormocholesterolaemic subjects? The ParisEospective Study. Am J Epidemiol 1986; 124:624-632.Reddy M N, Krishnaiah KV, Sehadri B. Total choles-terol fTC), high density lipoprotein cholesterol(HDL ) and its subfractions (HDL-2) an d (HDL-3),and triglyceride (TG) levels in people of Asian Indianorigin. Clin Ckem 1984; 30: 991.Thomas I, Gupta S, Sempos C, Cooper R. Serumlipids of Indian physicians living in the U.S. com-pared to U.S.-born physicians. Atherosclerosis 1986;61: 99-106.Saha N. Serum high-density lipoprotein cholesterol,apolipoprotein A-I, A-II and B levels in Singaporeethnic groups. AtberoseIerosds 1987 ; 68: 117-12 1.Schaeffer EJ, Levy RI, Anderson DW, Danner RN,Brewer HB, Blackwelder WC. Plasma triglycerides inregulation of HDL cholesterol levels. Lancet1978; 2:391-392.Laakso M, Pyorala K, Voutilainen E, Mamiemi J.Plasma insulin and sertm~ lipids and lipoproteinsin middle-aged no n-insulin dependent diabetic andnon-diabetic subjects. Am J Epidemtol 1987; 125:611621.Mukerjee AB. Pathogenesis of diabetes m ellitus.J Indian Med Assoc 1973 ; 61: l-6.


13/13

Coronary Heart Dim tse in South AsiansSmith U. Obesity and insulin resistance-po-tential importance of adipose tissue metabo lism.In: Vague J, Bjomtorp P, Guy-Grand B, Rebu&-Strive M, Vague P, Eds. Metabolic Complka-tiom of Humm Obesity. Amsterdam: Elsevier; 1985:173-178.Marmot MG. Socio-economic and cultural factorsin ischaemic heart disease. Adv Cardiol 1982; 29:68-76.Marmot MG, Adelstein AM, Robinson N , Rose GA.Chan ging social class distribution of heart disease. BrMed J 1978; 2: 1109-l 112.Alfredsson L, Karasek R, Theo&l T. Myocardialinfarction risk and psychosocial work environment:an analysis of the male Swedish working force. SetScl Med 1982; 16: 46W7.Marmot MG, R ose G, Shipley M, Hamilton PJS.Employment grade and coronary heart disease inBritish civil servants. J Epidemiol Commm Health1978; 32: 244-249.Forsdah l A. Are poor living conditions in childhoodand adolescence an important risk factor for arterio-sclerotic heart disease? Br J Rev Sot Med 1977; 31:91-95.

121. Orth-Gamer K, Johnson JV . Social network interac-tion and m ortality. A six year follow-up study of arandom sample of the Swedish population. J ChrenDfs 19 87; 40: 949-957.122. Stopes-Roe M, Cochrane R. Mental health and inte-gration: a comparison of Indian, Pa kistani a nd Irishimmigrants to England. Ethnic and Racial Stodka1980; 3: 316341.123. Syme SL, Marmot MG, Kagan A, Kato H, RhoadsGG. Epidemiologic studies of coronary heart diseaseand stroke in Japanese men living in Japan, Hawaiiand California: introduction. Am J Epidemiol 1975;102: 477-480.124. Mathews D . Ischaemic heart disease: possible geneticmarkers. Lmcet 1975; 2: 681-683.125. Scott BB, McGu Bln P, Rajah SM , Stoker JB,Losowsky MS. Histocompatibility antigens and myo-cardial infarction. Tlasue An&tens 1976 : 7: 187-188.

Coronary Prevention Group. Coronary Heart Diseaseand Adam in Britain. London: Confederation ofIndian Organisations; 1986 : 37-45.Gxhrane R, Stopes-Roe M. Psychological symptomlevels in Indian immigrants to England-a compari-son with native English. Psychei Med 1981 ; 11:319-327.Cochrane R, Stopes-Roe M. Social class and psychi-atric disorder in-na tives an d immig rants to Britain.M J Set Psvcbiat 1981: 27: 173-182.Berkman LF: Syme SL.Social networks, host resis-tance and mortality: a nine-year follow-up ofAlameda County residents. Ant J Epidemiil 1979;109: 186-204.

126. Logan R L, Oliver MF, M cTa%sh J, D&g C, WhiteAG. Histocompatibility antigens and myocardial in-farction. Tlssne Antigem 1977 ; 10: 361-36 3.127. Mehra NK, Taneja V, Kailash S, Raii N, VaidyaMC. Distribution of HLA antigens in a sample of theNorth Indian Hindu population. &sue Antigem1986; 27: 64-74.128. Baird JD. The role of obesity in the development ofclinical diabetes. Puhl R CoU Physicians Edlnb 1973 ;42: 83-99.129. Harvald B, Hauge M. Selection in diabetes in modemsociety. Acta M&l Scmd 1963; 173: 459-465.130. Mandrun-Paulsen T. Owerbach D. Nerun J.Johansen K, IngerslevJ, Tybjaerg Hansen A. In.&lin:gene flanking sequences, diabetes mellitus andatherosclerosis: a review. DIahetoIogla 1985 ; 28:556564.131. Neel JV. Diabetes m ellitus: a thrifty genotyperendered detrimental by progress. Am J HumGenet 1962; 14: 353-362.

60 9111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

Documents

Marmot Indian dispora