INTRODUCTION

Elevated circulating total homocysteine (tHcy)concentration is a risk factor for cardiovasculardisease1,2 and elevated tHcy or low folate and vitaminB12 concentrations are a risk factor for birth defects,poor pregnancy outcomes and neurocognitiveperformance.3-5 Indians in India6,7 as well as thosemigrated abroad8, 9 have high circulating tHcy concentration compared to other ethnic groups. InCaucasian populations not eating folic acid fortifiedfood, hyperhomocysteinemia is usually explained by lowblood folate concentrations.10 In contrast, hyper-homocysteinemia in Indians living in India is moreattributable to low concentrations of vitamin B12.4

*Diabetes Unit, King Edward Memorial Hospital and Research Centre, Pune, India. **Institute of Basic Medical Sciences, Department of Nutrition, University of Oslo and Institute of Medicine, Section of Pharmacology, University of Bergen, Norway. ***Diabetes and Cardiovascular Disease Academic Unit, University College, London, UK.
Received : 7.2.2006; Revised : 12.7.2006;
Re-revised : 16.8.2006; Accepted : 25.8.2006

Eventhough low circulating vitamin B12concentration have been recognised in Indians for a longtime,11 there is little appreciation of this amongst Indianmedical professionals and policy makers. This may bedue to a number of reasons: 1) vitamin B12 and tHcy arenot routinely measured in clinical practice, 2) despitelow circulating vitamin B12 concentrations, specifichematological and neurological manifestationsconsistent with vitamin B12 deficiency are rare, and 3)the majority of previous reports are clinic based andtherefore may not represent community prevalence.

CRISIS (Coronary Risk of Insulin Sensitivity in IndianSubjects) is a community-based study of insulinresistance and cardiovascular risk in rural and urbanmiddle-aged Indian men. We report our findings onvitamin B12 and folate concentration andhyperhomocysteinemia in the CRISIS study.

Details of design and methods in the CRISIS studyhave been published.12 We used multistage randomsampling to select 30-50 y old men from 2 villages, 2slums and 2 middle class areas from in and around Pune.Those with known diabetes mellitus, hypertension,coronary heart disease were excluded. We studied 441apparently healthy men (149 rural, 142 slum residentsand 150 urban middle-class). Participation rates were86% for rural, 79% for slums and 71% for urban middle-class men.

The study protocol was approved by the EthicalCommittee of the King Edward Memorial Hospital andResearch Centre and by the local community leaders.Every participant gave a signed consent.

Subjects ate their normal diet and performed theirusual activities until the day before the study. Those withacute intercurrent illness (n=15) were rescheduled 4weeks after recovery. Subjects arrived at the DiabetesResearch Unit the night before the study and were thenfed a standard local meal, medically examined and anx-ray chest was done. Trained research staff enquiredabout history of migration, lifestyle factors, nutritionalintake and disease symptoms. Only water was allowedovernight.

In the morning, fasting blood samples were collectedin EDTA tubes. A 75 g anhydrous glucose tolerance test(WHO 1997) was done. One portion of the whole bloodwas used for determination of hematological variables,which were measured within 1 hour of blood collection.The remaining blood was immediately centrifuged at4oC and 2500 g for 15 minutes to obtain EDTA plasma.Aliquots of plasma were stored at –80oC till further analysis.

Standardised anthropometric measurements weremade. Blood pressure was measured by an automatedmachine (UA 767PC, A and D Instruments Ltd,Abingdon, Oxford, UK). Two blood pressure readingswere recorded 5 minutes apart after a 15 minutes rest inthe supine position: the second reading was used in theanalysis.

Self-declared religion was noted. Smoking and alcoholintake were recorded as never, past and current.Education was recorded as completed years of formaleducation, and monthly income was categorized.Medical history included frequency of upper gastrointestinal (acidity, regurgitation, epigastric pain)and lower gastrointestinal symptoms (diarrhoea,worms, blood and mucus in stools). Use of drugs,including vitamin supplements was noted.

A diet-recall assessed the intake of energy,carbohydrates, proteins and fats on an average day usingnutritive values from a local13 and a national database.14A food frequency questionnaire was used to assess thefrequency of consumption of foods which were rich inmicronutrients and antioxidants. Based on the focus ofthe present paper, foods rich in vitamin B12, folate andthose shown in previous studies to be related tohomocysteine were selected and grouped into 5 mutuallyexclusive groups: milk and milk-containing beverages,coffee, green leafy vegetables, other vegetables and non-vegetarian foods. For each food item, average frequencyof consumption over the past year and the normalportion size were ascertained. For some items (vegetables)only seasonal frequency was recorded.

Cardiovascular disease was assessed by the Rose-WHO angina questionnaire15 and a resting 12 leadelectrocardiogram (ECG). Intima media thickness of thecommon carotid artery (CCA) was measured using ATLUM9 Color Doppler Machine (Philips Bothell,Washington, USA). We validated our intima mediathickness measurements against those made by avascular research laboratory (Prof Michiel Bots, JuliusCentre for Health Sciences and Primary Care, UniversityMedical Centre, Utrecht, Netherlands).

Haematological measurements were made on aBeckman Coulter Analyser (AC.T diff TM, Miami, Florida).Plasma levels of glucose, total and HDL cholesterol,triglycerides, creatinine were measured on a Hitachi 911automated analyser (Hitachi Ltd, Tokyo, Japan) usingstandard methods. Plasma insulin was measured usingin-house DELFIA method. Insulin resistance wascalculated from fasting plasma glucose and insulinconcentrations using the homeostasis model assessment(HOMA-R).16

Plasminogen Activator Inhibitor-1 (PAI-1) (HypenBiomed, Andresy, France), C-reactive protein (CRP)(United Biotech Inc. CA, USA) and antibody titres toHelicobacter pylori and intrinsic factor were measuredusing commercial kits. Plasma vitamin B12 and red cellfolate were measured using a radioimmunoassay kit(Diagnostic Products Corporation, USA) and plasmatotal homocysteine (tHcy) was determined using the IMxSystem (Abbott Laboratories, IL, USA).

Data are presented as median (interquartile range)and percentages for the three residential groups. In thestatistical analyses, variables with skewed distributionwere log-transformed to satisfy the assumption ofnormality. Comparison between the three groups was tested using analysis of variance with Bonferroni post-hoc test, or chi square test. Contribution of covariates todifferences in 3 populations was assessed by analysisof covariance. Pooled data from the three residentialgroups was used to test the associations between bloodconcentrations of vitamin B12, tHcy and folate and othervariables. This analysis was either by partial correlationsor analysis of variance, adjusting for age and place ofresidence. In this study, we defined low vitamin B12concentration as plasma vitamin B12 <150 pmol/L, lowfolate concentration as red cell folate <283 nmol/L andhyperhomocysteinemia as plasma tHcy >15 µmol/L. Wecalculated the population-attributable risk forhyperhomocysteinemia of low concentrations of vitaminB12 and folate compared to those who had vitamin B12and folate concentration above the thresholds. Significantdeterminants of low vitamin B12 or folate concentrationsand hyperhomocysteinemia were investigated bymultiple logistic regression analysis. Significantdeterminants of haemoglobin and mean corpuscularvolume count were assessed by multiple linear regression analysis. SPSS version 11.0 for windows(SPSS Inc, Chicago) was used for statistical analysis.

Rural and urban middle-class men had been residentin these places for more than one generation, while slumresidents were mostly first generation migrants fromvillages. The majority of the men were Hindu (91%),others were Muslim (Table 1). Rural men weresubsistence farmers and physically very active. Slumresidents were daily-wage workers, and were morefrequent smokers. Urban middle-class men were moreeducated, were wealthier and had sedentary jobs; theywere taller and heavier than the slum and rural residents.

Median plasma vitamin B12 concentration was lowin all 3 groups of men, being lowest in the urban middle-class. Using150 pmol/L as threshold, 67% men had low vitamin B12-concentration (68% rural, 51% slumresidents, 81% urban middle-class). Median red cellfolate concentration was in the normal range in the threegroups (>283 nmol/L), and low folate concemThere wasno significant relationship between plasma vitamin B12and red cell folate concentrations. Plasma vitamin B12concentration was inversely related to plasma tHcyconcentration (r= -0.41, p<0.001), an association that wasindependent of red cell folate concentration. Red cellfolate concentration was inversely related to plasma tHcyconcentration (r= -0.18, p<0.001). Adjusted for age andplace of residence, low vitamin B12 concentrationscontributed 28.4% to the risk of hyperhomocysteinemia(population attributable risk) while low folateconcentrations contributed only 2.2%.

Plasma vitamin B12 concentration decreased andplasma tHcy increased with increasing age (p< 0.01 andp<0.05 respectively). Plasma vitamin B12, tHcy and redcell folate concentrations were not related to body sizemeasurements. Compared to Hindu men, Muslim menhad higher plasma vitamin B12 (148 vs 107 pmol/L),lower plasma tHcy (15.8 vs 19.5 µmol/L) and lower redcell folate concentrations (423.7 vs 505.3 nmol/L) (p<0.01, all), of which difference between plasma vitaminB12 and plasma tHcy became non-significant afteradjusting for non-vegetarian food intake. Highereducation and income were associated with lowerplasma vitamin B12 and higher tHcy concentrations(p<0.05). These relations were not independent of intakeof non-vegetarian foods. Smoking and alcohol habitswere not related to plasma vitamin B12 or tHcy or to redcell folate concentrations.

Only 3 men took vitamin supplements, nonecontaining vitamin B12. Daily energy intakes of thesemen were lower than those recommended by the Indian Council of Medical Research (2425 kcals/day) in all 3groups, and were not related to plasma vitamin B12 ortHcy and red cell folate concentrations. Protein intakeswere comparable to the ICMR recommendation (60g/day). Higher protein intake was associated with higherred cell folate concentration (p<0.05) independent of ageand place of residence.

Food frequency data was available on 424 men. Noneof these men were vegan. Forty one percent rural, 11%slum residents and 44% urban middle-class men werelacto-vegetarians. Non-vegetarian foods were eaten morefrequently by Muslims than Hindus (p<0.001), by thosewho were less educated or poorer (p<0.001), and by theslum residents compared to urban middle class(p<0.001). The portion size of non-vegetarian foods wasusually small (~100 g cooked). Most men ate chicken,fish and eggs and very few ate red meat. There was aprogressive and graded relation between frequency ofconsumption of non-vegetarian foods and plasmavitamin B12 (r= 0.27, p<0.001) and tHcy concentrations(r= -0.26, p<0.001). On univariate analysis (age adjusted)lacto-vegetarians had a 4.3 (95%CI, 2.4, 7.8) times higherrisk of low vitamin B12 concentrations and 4.3 (95%CI,2.4, 7.6) times higher risk of hyperhomocysteinemiacompared to those who ate non-vegetarian foods on atleast alternate days (data not shown). Ninety-three menhad non-vegetarian food frequently (Table 1).Approximately half of the men who ate non-veg foodfrequently had low vitamin B12 concentration andhyperhomocysteinemia. Nine men had hyper-homocysteinemia despite frequent non-vegetarian foodintake, normal circulating vitamin B12, normal folateand plasma creatinine concentrations. Intake ofvegetables, milk and coffee was not related to plasmavitamin B12 and tHcy or red cell folate concentrations.

Plasma vitamin B12 and tHcy and red cell folate concentrations were not related to gastrointestinal symptoms, to H. pylori antibody titre, or to plasma CRPand PAI-1 concentrations (data not shown). Intrinsicfactor antibody results available in 91 randomly selectedmen, was positive in 3 (2 rural and 1 urban middle-class), none of whom had particularly low plasmavitamin B12 or raised plasma tHcy concentrationcompared to the rest of the population.

In a logistic regression model including age, religion,income, education, place of residence, and food habits,low vitamin B12 concentrations were independentlyassociated only with food habits. Lacto-vegetarians were4.4 (95%CI 2.1, 9.4) times more likely to have low plasmavitamin B12 concentration compared to men who atenon-vegetarian foods frequently (Table 3).

In a model including age, place of residence, foodhabits, religion, income, education and plasmacreatinine concentration, hyperhomocysteinemia wasindependently associated with higher plasma creatinineconcentrations, urban middle-class residence and lacto-vegetarian food habits. Urban middle-class residentswere 7.9 times more likely to have hyper-homocysteinemia compared to rural men (95%CI 2.8,22.7), and lacto-vegetarians had 4.3 times higher riskcompared to those who ate non-vegetarian foodsfrequently (95%CI 2.1, 8.9), (data not shown). When lowvitamin B12 and folate concentration were included asindependent variables, both were significantlyassociated with hyperhomocysteinemia (OR 2.9, 95%CI1.7, 4.9 and OR 8.4, 95%CI 1.7, 42.8, respectively). Thisinclusion reduced the strength of association betweenhyperhomocysteinemia and lacto-vegetarianism (OR 3.0,95%CI 1.4, 6.5) but not the place of residence (OR 7.6,95%CI 2.5, 22.6) (Table 3). When we tested this model byincluding plasma vitamin B12, plasma creatinine, andred cell folate concentrations, as continuous variablesthe results were similar.

Low folate concentration was not independentlyassociated with any of the following variables: age, placeof residence, religion, income, total macronutrient intakeand green leafy vegetables intake (data not shown).

Plasma vitamin B12 concentration was directlyrelated to blood haemoglobin concentration (r=0.18,p<0.001) and total leucocyte count (r=0.18, p<0.001) andinversely to mean corpuscular volume (MCV) (r= -0.14,<0.01). Red cell folate concentration was inverselyrelated to MCV (r= -0.20, p<0.001). In a multivariateanalysis, blood haemoglobin concentration wasindependently associated with plasma vitamin B12 andferritin concentrations (partial r= 0.16, and 0.24respectively, p<0.001, for both) but not to red cell folateconcentrations. MCV was independently associatedwith plasma ferritin (partial r=0.22), and plasma vitamin B12 (r= -0.13) and red cell folate (r= -0.21) concentrations(p<0.01, for all). One hundred and eleven of these menwere anaemic (hemoglobin <135 g/L), 31% of these hadmicrocytosis (MCV<80 fL), but only 2% (n=2) hadmacrocytosis (MCV>100 fL). Plasma vitamin B12concentration was low in 65% of those with microcyticanemia, and in 79% of those with normocytic anemia,and in both men with macrocytic anemia. Leucopenia(< 4.5*109/L) and thrombocytopenia (< 140*10/L) were uncommon.

There was a progressive increase in the proportion of men with impaired glucose tolerance, diabetes mellitus,hypertension and hypertriglyceridemia from rural toslum and urban middle class residents. Proportion withlow high-density lipoprotein cholesterol and plasmacreatinine concentrations were similar in three groups of men.

Plasma vitamin B12, tHcy and red cell folateconcentrations were not associated with plasma glucose,total cholesterol, and triglycerides concentrations andblood pressure on continuous analysis. Plasma vitaminB12 was directly (r= 0.14, p<0.05) and tHcyconcentration was inversely (r= -0.13, p<0.05) related toHDL concentration. Men with hyperglycemia (fastingplasma glucose >110 mg/dl or 2h plasma glucose >140mg/dl), hypertension (blood pressure >140/90 mmHg)and low HDL cholesterol concentrations (<35 mg/dl)had higher plasma tHcy concentrations compared totheir normal counterparts, even after adjusting for ageand place of residence. However plasma vitamin B12and red cell folate concentrations were similar in thetwo groups of men.

There was no significant relationship between plasmavitamin B12, red cell folate and tHcy concentrations andECG abnormalities and intima media thickness (datanot shown).

DISCUSSION

Our results demonstrate widespread low plasmavitamin B12 concentration in a community based studyof rural and urban middle-aged Indian men in Pune,Maharashtra, India, confirming our previous findingsof a clinic based study of patients with or withoutcoronary heart disease and diabetes.5 Low folateconcentration was relatively rare. Hyper-homocysteinemia was very common and more ascribableto low vitamin B12 concentration (attributable risk 28%)than to low folate concentration (2%). Lower plasmavitamin B12 concentration was associated not only withhigher plasma tHcy concentration but also with lowerblood haemoglobin concentration and higher MCV. Thissuggests that vitamin B12 concentration had metabolicand hematological consequences, although macrocyticanemia was rare, probably because of adequate folateand low ferritin concentration. Low vitamin B12concentrations were partly explained by vegetarianism.Hyperhomocysteinemia was independently related tovegetarianism and to urban middle-class residence.These findings represent the first community-basedstudy of vitamin B12 and homocysteine status inmainland India.

In the CRISIS study we sampled men from villages,urban slums and urban middle-class to reflectepidemiological, socio-economic and nutritionaltransition in India. Approximately half the rural andslum residents and 81% of urban middle-class residentshad a low vitamin B12 concentration which waspartially explained by low dietary intake. A third of allmen were lacto-vegetarian and only half of these drankmilk regularly, thus excluding important dietary sourcesof vitamin B12. Vegetarianism was 4 times more commonin the urban middle-class (44%) than in the slumresidents (11%). Indians living abroad also have lowvitamin B12 concentration ascribable to low dietaryintake due to vegetarian food habits.7 One morecontributory factor could be impaired food vitamin B12absorption.17 Gastrointestinal symptoms and presenceof H. pylori antibodies were not related to low vitaminB12 concentration in our study. However, this possibilitydeserves formal investigation. Tropical sprue is also anunlikely cause because it usually causes folate deficiency18, which is in contrast to our data. Antibodies to intrinsicfactor were present in less than 5% of subjects, suggestingthat pernicious anemia is not a common cause. Theultimate source of vitamin B12 in nature is microbes.The association of higher education and income withlower vitamin B12 concentration could reflect a lack ofmicrobial vitamin B12 from ingestion of contaminatedfood and water, as well as recycled colonic bacteria.19

Hyperhomocysteinemia was very common in thesemen and the median plasma tHcy concentration wastwice that reported in White-Caucasian populations.20,21The aetiology of hyperhomocysteinemia appearsmultifactorial, with contributions from diminished renalfunction (higher plasma creatinine but still withinnormal range), vitamin B12 deficiency associated withvegetarianism, folate deficiency and urban residence.Contribution of vegetarianism and low vitamin B12concentration was much more important than that oflow folate concentration. Other known determinants oftHcy, such as smoking, coffee drinking and high alcoholintake20,21 were not associated with hyper-homocysteinemia in this population. Occurrence ofhyperhomocysteinemia in a substantial number of meneating non-vegetarian food could either be due to smallintake of these foods, rarity of red meat consumption ordue to reduced absorption of food vitamin B12. Thesubstantial contribution of urban middle class residenceto hyperhomocysteinemia, independent of dietary habitsand blood vitamin status remains unexplained andneeds further investigation. Urban middle classrepresent the leading edge of epidemiologic andnutritional transition, and differ from the rural and slumresidents in a number of social, economic, behaviouraland other lifestyle factors. Some or all of these mightcombine to increase the risk of hyperhomocysteinemia.

Hyperhomocysteinemia in some men with normalcirculating vitamin B12 concentration suggests acontribution of additional mechanisms, for example,associated deficiency of other relevant nutrients such asfolate (rare in our population), vitamin B2, vitamin B6 orbetaine or its precursors.22 Another possibility is defectivetransport into the cells. Migrant Indians in the US had higher plasma tHcy concentrations compared to whiteCaucasians at equivalent plasma vitamin B12concentrations. This was only partly explained by lowvitamin B6 concentrations.8

In addition to hyperhomocysteinemia, low plasmavitamin B12 concentration was associated with lowerblood haemoglobin concentration and larger red cellvolume (in the ‘normal’ range), though macrocyticanaemia was rare. This may be due to adequate folateconcentration and low iron status (ferritin), both of whichprevent macrocytosis. In clinical practice lack of‘macrocytosis’ is usually interpreted as indicatingnormal vitamin B12 status, this may have resulted inunderdiagnosis of low vitamin B12 status in India. Astudy in migrant Indians in the UK23 as well as a recentstudy in Delhi24 reported that dietary deficiency ofvitamin B12 was a more common cause of megaloblasticanaemia than was pernicious anaemia. We havepreviously demonstrated high circulating methylmalonic acid concentrations in middle-aged men andwomen,5 a specific indication of impaired action ofvitamin B12.

Hyperhomocysteinemia was associated with somecomponents of the metabolic syndrome i.e.hyperglycemia, hypertension and low levels of plasmaHDL-cholesterol concentration but not with HOMAinsulin resistance, inflammatory and prothromboticmarkers. It was also not related to intima media thicknessor ECG abnormalities in this cross sectional study. Astudy in migrant Indians in the US showed anassociation between plasma tHcy concentration andinsulin resistance.8 Recently high plasma tHcyconcentration has been shown also to predict incidentdiabetes (both gestational and post-gestational).25,26 InIndians as yet no study has demonstrated an associationbetween hyperhomocysteinemia and cardiovasculardisease. These associations need to be tested inprospective studies. Studies in Pune have demonstratedan association between high plasma maternal tHcyconcentration and small for gestational age babies.4 Inmigrant Indians in the UK, vegetarian habits and lowvitamin B12 concentration have also been associatedwith increased risk of tuberculosis.27

Thus, low plasma vitamin B12 concentration andhyperhomocysteinemia are common in middle-agedIndian men in Pune, Maharashtra. Despite repeateddemonstration of low plasma vitamin B12 concentrationin Indians for over 50 years there is poor appreciation ofthe problem by both the medical profession and thepolicy makers. Vegetarianism and urban middle classresidence are important aetiological factors.Vegetarianism in India is multigenerational, lifelong andbased on religious and cultural beliefs. Further researchis necessary to improve our understanding in this areaand to investigate novel ways to treat and prevent the problem.

Acknowledgements

The study was funded by the Nestle Foundation,Lausanne, Switzerland. We thank Dr K. Shelgikar, Dr. AKolhatkar, Dr N Joshi, Dr S Gandhi, Dr A Khadilkar, DrM Chinchawadkar, Mr C Joglekar, Mr MG Sayyad, Msganpule, Dr N Thuse, Ms Kulkarni, Mr. S Yenge, Mr. TDeokar, Mr. S Chaugule, Mr. A Bhalerao, Mr. V Solat, Mr.Kadam, Mr Gaikwad for their invaluable contribution.We are grateful to Prof. Michiel Bots for his help invalidating our Intima media thickness data.

REFERENCES

1.Review. Homocysteine and risk of ischemic heart diseaseand stroke: a meta-analysis. The Homocysteine Studies Collaboration. JAMA 2002;288:2015–22.

2.McCully KS. Vascular pathology of homocysteinemia:implications for the pathogenesis of arteriosclerosis. Am JPathol 1969;56:111–28.

3.Ueland PM, Vollset SE. Homocysteine and folate inpregnancy. Clin Chem 2004;50:1293-5.

4.Yajnik CS, Deshpande SS, Panchanadikar AV, Naik SS,Deshpande JA, Coyaji KJ, Fall CHD, Refsum H. Maternaltotal homocysteine concentration and neonatal size in India.Asia Pac J Clin Nutr 2005;14:179-81

5.Quadri P, Fragiacomo C, Pezzati R, Zanda E, Forloni G,Tettamanti M, Lucca U. Homocysteine, folate, and vitaminB12 in mild cognitive impairment, Alzheimer disease andvascular dementia1-3. Am J Clin Nutr 2004;80:114-22.

6.Refsum H, Yajnik CS, Gadkari M, et al.Hyperhomocysteinemia and elevated methylmalonic acidindicate a high prevalence of cobalamin deficiency in AsianIndians. Am J Clin Nutr 2001;74:233–41.

7.Misra A, Vikram NK, Pandey RM, et al.Hyperhomocysteinemia and low intakes of folic acid andvitamin B12 in urban North India. Eur J Clin Nutr 2002;41:68-77.

8.Chambers JC, Obeid OA, Refsum H, et al. Plasmahomocysteine concentrations and risk of coronary heartdisease in UK Indian Asian and European men. Lancet2000;355:523–7.

9.Chandalia M, Abate N, Cabo-Chan AV JR., Devraj S, Jialal I,Grundy SM. Hyperhomocysteinemia in Asian Indians livingin the United States. J Clin Endocrinol Metab 2003;88:1089-95.

10.Selhub J, Jacques PF, Bostom AG, Wilson PW, Rosenberg IH.Relationship between plasma homocysteine and vitaminstatus in the Framingham study population. Impact of folicacid fortification. Public Health Rev 2000;28:117–45.

11.Banerjee DK, Chatterjee JB. Serum vitamin B12 invegetarians. Br Med J 1960;2:992-4.

12.Bhat DS, Yajnik CS, Sayyad MG, Raut KN, Lubree HG, RegeSS, Chougule SD, Shetty PS, Kurpad AV. Body fatmeasurement in Indian men: Comparison of three methodsbased on a two-compartment model. Int J Obes (Lond)2005;29:842-8.

13.Nutritive values of some Indian food preparations.Department of Biometry and Nutrition. MACS ResearchInstitute, Pune; 1992.

14.Gopalan C, Rama Sastri BV, Balasubramanian SC. Nutritivevalue of Indian foods, Revised ed. NIN, ICMR,Hyderabad;1989.

15.Rose GA, Blackburn H, Gillum RF, Prineas RJ. CardiovascularSurvey Methods, 2nd ed. WHO, Geneva;1982.

16.Wallac TM, Levy JC, Matthews DR. Use and Abuse of HOMAModeling Diabetes Care 2004;27:1487–95.

17.Westergaard H. Tropical Sprue. Curr Treat Options Gastroenterol2004;7:7-11.

18.Cook GC, Aetiology and pathogenesis of postinfectivetropical malabsorption (tropical sprue). Lancet 1984:721-3.19.Pathophysiology, Pathophysiologic principles of nutrition;470-72.

20.Jacques PF, Bostom AG, Wilson PW, Rich S, Rosenberg IH,Selhub J. Determinants of plasma total homocysteineconcentration in the Framingham Offspring cohort. Am JClin Nutr 2001;73:613-21.

21.Nygard O, Refsum H, Ueland PM, Vollset SE. Major lifestyledeterminants of plasma total homocysteine distribution: theHordaland Homocysteine Study. Am J Clin Nutr 1998;67:263-70.

22.Ueland PM, Holm PI, Hustad S. Betaine: a key modulator of one-carbon metabolism and homocysteine status. Clin ChemLab Med 2005;43:1069-75. Review

23.Britt RP, Hollis Y, Keil JE, Weinrich M, Keil BW. Anaemia inAsians in London. Postgrad Med J 1983;59:645-7.

24.Khanduri U, Sharma A, Joshi A. Occult cobalamin andfolate deficiency in Indians. Natl Med J India 2005;18:182-3.

25.Seghieri G, Breschi MC, Anichini R, Bellis De A, Lorenzo A,Maida F, Franconi F. Serum homocysteine levels are increasedin women with Gestational Diabetes Mellitus Metabolism2003;52:720-23.

26.Cho NH, Lim Soo, Jang HC, Park HK, Metzger BE. Elevatedhomocysteine as a risk factor for the development of diabetesin women with a previous history of Gestational DiabetesMellitus, A 4-year prospective study. Diabetes Care2005;28:2750-55.

27.Chanarin I, Stephenson E. Vegetarian diet and cobalamindeficiency: their association with tuberculosis. J Clin Pathol1988;41:759-62.