Tribal Populations of Maharashtra Cormic Index

8/8/2019 Tribal Populations of Maharashtra Cormic Index

1/9

Ma1 J Nutr 12(1): 23-31,2006

Assessment of Nutritional Status through Body Mass Indexam ong Adult Males of 7 Tribal Populations of Maharashtra,

IndiaDipak K Adak, Rajesh K Gautam & Ajay K Gharami

Anthropolo gical Su rv ey of India, c/o Depa rtm ent of Anthropolo gy, Dr .H.S. Gour Universi ty,Saugor-470003, Madhya Pradesh, India

ABSTRACT

This paper deals with the assessment of nutritional status through body mass

index (BMI) in adult males of 7 tribal populations of the Maharashtra state ofIndia. Anthropometric data collected by the Anthropological Survey of Indiaon 600 adult males aged 18-60 years were used in this analysis. Majority ofthem were thin and lean with medium to short stature. The mean BMI wasfound to be highest among the Gond (18.33 kg/m2) followed by Korku (18.30kg/m2) and Mahadeokoli (18.17 kg/m2). There were significant variations inthe mean BMI among the tribal populations, ranging between 16.82 kg/m2 and18.33 kg/m2. The prevalence of Chronic Energy Deficiency (CED) was highestamong the Warli, among whom 30.0% were found to be suffering from severeCED, 32.0% moderate CED and 26.0% mild CED. A greater prevalence (36.0%)

of normal status of nutrition was found among the Korku. The linear regres-sion coefficient (b f standard error) of BMI on Cormic Index or CI (SH/H) forthese tribal groups was 13.68 f 3.3 (t=4.14, p < 0.000), and the correlationcoefficient (r f standard error) was 0.16 f 0.02. The differences in BMI acrossthe tribes were significant even after allowing for the Cormic Index. It can besaid that the majority of the tribal populations of Maharashtra were lean. Thehigh proportion of individuals with CED corroborates their low social and

-. economic status.

INTRODUCTION

There are many measures to assessthe nutritional status of a population.Body mass index (BMI) is one of them.Anthropometry is considered to be animportant tool for assessing nutritionalstatus of individuals or of the community.Hence, measurements like stature, sittingheight, weight and indices based on thesemeasurements developed by differentscholars have been extensively used todefine the extent of malnutrition. Bodymass index (BMI) expressed as ratio of

weight to height squared can be a goodparameter to grade chronic energy defi-ciency (CED) in adults (Naidu e t al., 1991).There are many studies based on thisaspect (for example Ferro-Luzzi et al.,1991; Khongsdier, 2001). Inadequacies innutritional intake or under-nutrition canbe considered as a major source of manyadverse effects on the growth and healthof individuals (Gordon et al., 1968).Knowledge of the nutritional status of acommunity is necessary to have a compre-hensive idea about its developmentprocess, as under-nutrition is one of the

Correspondence author: Dr Dipak Kumar Adak, email address: [email protected]


2/9

Adak DK, Gautam RK b Gharami AK

major health problems in developingcountries. It is reported that the basiccauses of under-nutrition and infections indeveloping countries are poverty, poorhygienic conditions and little access to pre-ventive health care (Mitra, 1985; WHO,1990). Hence, assessment of the nutritionalstatus of a population has attracted theattention of not only the nutritionists andother biological scientists, but also econo-mists and other social scientists with aview to understanding the health andsocio-economic status of the population(Osmani, 1992). Literature on BMI of adult

Indians is limited to certain geographicalareas or populations. Noteworthy amongthem are the study of BMI among theNorth-East Indian (Khongsdier, 2001) andSouth Indian populations (Ferro-Luzzi e tal., 1991). However, little is known aboutthe BMI of tribal populations. The presentstudy is an attempt to assess the nutrition-al status through body mass index amongadult males of seven tribal populations in

the state of Maharashtra, located in theNorth-Western part of India.Approximately 10% of the popula-

tion of Maharashtra (about 85 million)belongs to tribal population groups. Thesegroups remain isolated, living in forestsand hilly areas. Majority of them havepoor health status (Kate, 2000).

dard techniques (Martin and Saller, 1956).Therefore, it can be well assumed that theaccuracy of the data is properly taken careof. Subjects were not chosen on the basis ofbodily structure and proportion. Effortswere also made to exclude closely relatedindividuals like brothers, fathers, sons andthose with any kind of physical deformi-ties. Therefore, the samples were free fromany selection bias. For convenience only,adult males who looked apparently activeand healthy (not suffering from any appar-ently visible infection or disorder) wereconsidered in the sample. In this study,

body weight (W), height (H) and sittingheight (SH) were taken to calculate cormicindex (SH/H) and BMI (W/H2). Both MS-EXCEL and SPSS software were used toanalyse the data. Cormic index and bodymass index were calculated for each indi-vidual followed by calculation of centraltendency viz. mean and deviation of eachmeasurement among each tribe. FurtherANOVA-Test or one-way analysis of

variance was performed. For screening ofthe CED groups, the value of 18.5 wastaken as a cut-off point following James e tal. (1988), Ferro-Luzzi et al. (1991), andKhongsdier (2001). Regression analysiswas done to find the correlation betweenBMI and CI.

RESULTSMATERIALS AN D METHODS

The study sample is based on basicanthropometric data collected on adultmales aged 18-60 years by theAnthropological Survey of India (Basu etal., 1994). In the present study, data werecollected on 600 individuals from 7 tribalgroups to assess their nutritional status.These tribal groups are: 1. Andh, 2. Bhil,3. Gond, 4. Kathodi, 5. Korku,

6. MahadeoKoli and 7. Warli. Anthropo-metric data were collected by trainedphysical anthropologists of the Anthro-pological Survey of India, following stan-

Mean and SD values for age, height,sitting height, weight, cormic index andbody mass index for each tribal group areshown in Table 1. Average age of samplepopulation varies between 31.52 yearsamong Bhil and 37.53 years among Korku,while it is 33.04 years for total population.In stature, Bhil are the tallest with 164.41cm average height. Their sitting heightwas also found to be greater (83.79 cm).

Side by side, sitting height and weight wasalso higher among them than other tribalgroups. However, Warli are shorter instature as well as in sitting height as


3/9

Body Mass Index among Adult Males of7 Tribal Populations of Maharashtra, India 25

Table 1. Mean and standard deviations of age, anthropometric measurements and indices

Tribal Groups N *ge Height (cm) Sitting Weig ht (kg) Cormic Index BMI 1

height (cm)

1 Andh2 Bhil

3 Gond

4 Kathodi

5 Korku

6 Mahadeokoli

7 Warli

Total

compared to the other 6 tribal groups.Average weight of Warli was also found tobe comparatively lower. The variation inmean BMI between populations washighly significant, ranging between 16.82among Warli and 18.33 among Gond.Mean value of BMI was highest amongGond followed by Korku, Mahadeokoliand so on, as evident from bar diagram

(Figure 1). The mean BMI for the total sam-ple is 17.86k1.89, which is slightly greaterthan the estimation of Ferro-Luzzi et al.(1991) for Indian males (17.7f2.1). Themean cormic index (CI) or proportion of

sitting height to stature was found to varybetween 0.499k0.014 among Andh and0.499f0.016 among Warli to 0.509f0.011among Bhil.

It is apparent from Table 2 that theprevalence of Chronic Energy Deficiency(CED) is highest in the Warli, amongwhom 30.0 percent of the population fallsunder the severe grade of CEDI 32.0

percent under the moderate grade of CEDand 26.0 percent under the mild grade ofCED. A greater percentage of individualswith normal statu s of nutr ition wasobserved among Korku (36.0 percent). All

Population groups

Figure 1. Mean BMI among different tribal groups of Maharashtra, India


4/9

26 Adak DK, GautamRK & Gharami A K

Table 2. Percentage distribution of population as per chronic energy deficiency level

Percentage dist ribu tion of population as per chronic

energy deficiency level

CED CED CED Low Normal Obese TotalTribal Grade I11 Grade I1 Grade 1 W ei gh t (20.0- Grade I

Groups (Severe) (Moderate) (M ild) Normal 24.9) (25.0-< 16.0 (16.0- (17.0- (18.5- 29.99)

16.99) 18.49) 19.99)

1 Andh 12.0 26.02 Bhil 13.0 10.0

3 Gond 13.0 9.04 Kathodi 12.0 34.05 Korku 8.0 16.06 Mahadeokoli 11.0 15.07 Warli 30.0 32.0

Total 13.5 16.3

seven tribes along with their proportionatedistribution in different grades of CED areshown in the 100% bar diagram in Figure2. Cumulative distribution of adult BMIamong different tribal groups is shown inFigure 3. It reveals that 54 to 90 per cent ofthe respondents were chronic energydeficient with BMI value below 18.5. Thedifferences between the proportions ofsubjects with a low BMI were also striking.

To find out the difference in meanBMI between pairs of studied tribalgroups, the ANOVA test was applied,results of which are furnished in Table 3. Itappears that in most of the cases, thereexists no significant difference, exceptingthe pairs Andh and Gond, Andh andKorku and Mahadeokoli and Warli. In thesame way when the same test was appliedto find the difference of mean CI betweenthe groups, all the results of the testshowed insignificant difference.

The regression coefficient of BMI onCI was calculated and the results arefurnished in Table 4, and in bivariatescattered plot diagram (Figure 4). Theregression coefficient was found to be

significant (p>0.05) among Bhil, Kathodiand Mahadeokoli. It is apparent from thebivariate plot diagram (Figure 4) that thereis a positive correlation between BMI andCI. Side by side, BMI is dependent on CIand increases with rise in CI. On the basisof bivariate scattered plot diagram thestudied tribal groups can be categorisedinto two different groups. Andh, Kathodiand Warli form the first group, charac-terised by low mean BMI and CI, while thesecond group is formed by Korku, Gond,Mahadeokoli and Warli with considerablyhigher values of mean BMI and CI.

Bivariate correlation values amongthree anthropometric measurements andtwo indices are furnished in Table 5. Thereis significant correlation (at the 0.01 level)between stature and sitting height, andstature and weight; whereas an inversecorrelation was observed between statureand cormic index, and stature and BMIamong studied groups. Significant correla-tion was also found between BMI andsitting height, BMI and weight, and BMIand CI.


5/9

Body Mass lndex among Adult Males of7 Tribal Populations of Maharashtra, India 27

H C E D G rade II (Moderate)

O C E D Grade Ill (Severe)

Figure 2.100% Bar-diagram showing proportional distribution of each tribal populationin different grades of chronic energy deficiency.

Table 3. One way analysis of variance test for BMI

Tribal Groups 1 Andh Bhil Gond Kathodi Korku Mahadeokoli Warli

Andh

Bhil

Gond

Kathodi

Korku

Mahadeokoli

Warli

+ p0.05


6/9

28 Adak DK, Gautam RK & Gharami AK

100 -

80 -

150.-

C ,

m 60 -

Q0Q

IC 40 -0

s

13 14 15 16 17 18 19 20 21 22 23 24 25 26

BMI

Figure 3. Cumulative frequency distribution of adult BMI among 7 tribal populcltions otMaharashtra India

1 A Andh 1, 7 0 1 A Kathodi

A Warli1 6 8 I I I I I I I I0.498 0 500 0.501 0.502 0.504 0 505 0 507 0 508

Cormic Index (CI)Figure 4. Bivariate Scattered diagram of the populations with Cormic Index and BodyMass Index values


7/9

Body Mass Index among Adult Males of7 Tribal Populations of Maharashtra, India 29

Table 4. Regression coefficient and F statistics of BMI on CI

Tribal Groups CoefFcients of Regression F Statistics

R 72 P SE t- value F Change p-value

Andh 0.176 0.031 16.793 13.545 1.240 1.53 0.221Bhil 0.138 0.019 22.594 11.550 1.956 3.82 0.052Gond 0.103 0.011 4.613 4.504 1.024 1.04 0.308Kathodi 0.289 0.084 38.199 18.241 2.094 4.38 0.042Korku 0.176 0.031 17.762 14.379 1.235 1.52 0.223Mahadeokoli 0.370 0.137 43.802 11.123 3.938 15.50 0.000Warli 0.230 0.053 20.258 12.392 1.635 2.67 0.109

Total 0.167 0.028 13.680 3.300 4.141 17.14 0.000

Table 5. Bivariate correlations between stature, sitting height, weight, cormic index andbody mass index

Stature Sitting Height We igh t Cormic Index BMI(cm) (cm) (kg)

Stature (cm) 1.000 0.569** 0.558** -0.134"" -0.033Sitting Height (cm) 1 OOO 0.428** 0.731** 0.116**Weight (kg) 1.000 0.059 0.809**Cormic Index 1.000 0.167**BMI 1 OO O

**Correlation is significant a t the 0.01 level (2-tailed).

DISCUSSION

It is clear from the above findings

that the tribes of Maharashtra state ofIndia are thin and lean with medium toshort stature. Majority of them fall into thecategory of chronic energy deficientgrades, using the BMI value 18.5 as thecut-off point. Altogether 69 percent of therespondents fall below the level of 18.5. Inthis context, it can be said that the tribes ofMaharashtra have had poor subsistence oflivelihood. A large proportion of theirfoodstuffs are coarse grains, and forestproduces like roots, shoots, leaves, berries,seeds, tender bamboo sticks, mushrooms,flowers, fruits, nuts etc., which lack essen-tial nutrients. Availability of these food-

stuffs is also affected by seasonal variabili-ty, rainfall, monsoon and other environ-mental ups and downs. In such an adverse

condition, tribes are used to facing starva-tion or partial starvation. For generations,they have been surviving in such condi-tions and may adapt to survive with defi-ciency of essential energy required fornormal activity by a normal person.Besides poor subsistence of livelihood,alcoholism is also common among tribals,which may be one of the causes associatedwith low BMI. Ferro-Luzzi et al. (1991)proposed that BMI alone is sufficient todefine C ED in adults, irrespective ofenergy turnover as they suggested earlier(James et al., 1988). Furthermore, theirfindings on the distribution of BMI accord-


8/9

30 Adak DK, Gautam RK & Gharami AK

ing to various grades of CED in a southIndian population seem to be corroboratedby the present result, i.e. a large propor-tion of individuals with CED belong tograde I CED. They have suggested that themajority of the rural populations in Indiaare likely to be undernourished if grade ICED is associated with increased risk ofmorbidity and mortality.

Quetlet's Index, or BMI, is widelyused as a measure of fatness, or the nutri-tional status of popula tions in bothdeveloped and developing countries(Khongsdier, 2001). Recent studies have,however, questioned the validity of BMIas an indicator of fatness (Frankenfield etal., 2001; Kyle et al., 2003) because it lacksspecificity in terms of the variation in bodycomposition, and the confounding effectsof various factors such as age, sex, bodyshape and ethnicity (Norgan, 1994; Gurriciet al., 1998; Wagner and Heyward, 2000;Prentice and Jebb, 2001). It has also beensuggested that body fat compositionvaries considerably between ethnic groups(Norgan, 1994; Gallagher et al., 2000).

Norgan (1994) was of the opinionthat BMI is correlated with sitting height,or BMI is lower in those populations withhigher sitting height. It indicates that thereexists an inverse relationship between BMIand CI, but in the present study it wasfound that none of the 7 tribes had thisinverse relationship. Further, the differ-ences in means of CI between studiedtribal groups were insignificant. Thepresent findings thus do not support theview of Norgan (1994). According toKhongsdier (2001) BMI is largely inde-pendent of e thnic or genetic variation; itscorrelation with CI may have certainimplications as the la tter may be subject to

both genetic and environmentalinfluences. So the differences in means ofBMI between ethnic groups in the presentstudy (Table 3) may not only be due tonutrition, but also due to the low degreeimpact of other environmental and geneticfactors, as BMI is determined by both

genetic as well as environmental factors,whereas CI is determined more by genet-ics and less by environment. The regres-sion analysis (Table 4) also supports thatBMI is less dependent on CI as r-valuevaries from 0.103 to 0.370. However, it canbe surmised that improvement in thesocio-economic conditions of these tribesmay lead to an increase in their BMI. Thepresent study also indicates a need foreffective implementation of nutritionprogrammes among the studied popula-tions. Other information like dietaryintake, morbidity and health studiesshould also be collected from such tribalgroups.

Note: This is a revised version of the paperpresented in the National Seminar on"Tribal Health in India: Perspectives andChallenges", organised by the Departmentof Anthropology, Pondicherry University,Pondicherry, India, 14 - 15 March, 2005.

REFERENCES

Basu A, Mukherjee DP, Dutta PC, BoseDK, Basu MP, Ghosh GC, Kumar GD& Huq F (1994). All IndiaAnthropometric Survey: North ZoneBasic Anthropometric Data, Vol. I:Madhya Pradesh. Calcutta, Anthro-pological Survey of India.

Ferro-Luzzi A, Sette S, Franklin M & JamesW PT (1991). A Simplified approachof assessing adult chronic energydeficiency. Eur J Clin Nu tr 46:173-186.

Frankenfield DC, Rowe WA, Cooney RN,Smith JS & Becker D (2001). Limits of

body mass index to detect obesityand predict body composition.Nutrition 17: 26-30.

Gallagher D, Heymsfield SB, Heo M, JebbSA, Murgatroyd PR & Sakamoto Y(2000). Healthy percentage body fat


9/9

Body Mass Index among Adult Males of7 Tribal Populations of M aharashtra, India 31

ranges: an approach for developingguidelines based on body massindex. A m J Clin Nu tr 72: 694 -701.

Gordon JE, Ascoli W, Mata LJ, GuzmanMA & Scrimshaw NS (1968).Nutrition an d infection field study inGuatemalan villages 1959-1964,Acute diarheal disease and nutritiondisorder in general disease inci-dence. Archives Environ Hlth 16: 424-437.

Gurrici S, Hartriyanti Y, Hautvast JG &Deurenberg P (1998). Relationshipbetween body fat and body massindex: Difference between Indo-nesians and Dutch Caucasians. Eur JClin Nu tr 52: 779-783.

James WPT, Ferro-Luzzi A & Waterlow JC(1988). Definition of chronic energydeficiency in adults. Eur J Clin Nutr42:969-981.

Kate SL (2000). Health problems of TribalPopulation Groups from the State ofMaharashtra. httv:/ /sickle.bwh.har-vard.edu/ india scd.htm1. Accessedon 29 March 2006.

Khongsdier R (2001). Body mass index ofadult males in 12 populations ofNortheast India. Annals of HumanBiology 28:374-383.

Kyle UG, Schutz Y, Duper tu is YM &

Pichard C (2003). Body compositioninterpretation: contribution of the fatfree mass index and the body fatmass index. Nutrition 19: 597-604.

Martin R & Saller K (1956). Lehrbuch derAnthropologie. Vol. 3, Stuttgart, G.Fisher.

Mitra A (1985). The nutrition situation inIndia. In: Nutrition and Development.M Biswas and P Andersen (eds.)Oxford, Oxford University Press,p p 142-162.

Naidu AN, Neela J & Rao NP (1991).Maternal body mass index and birthweight. Nutrition News, NationalInstitute of Nutrition, Hyderabad.

Norgan NG (1994). Relative sitting heightand the interpretation of body massindex. Annals of Human Biology21:79-82.

Osmani SR (eds) (1992). Nutr ition an dPoverty. Oxford: Clarendon Press.

Prentice AM & Jebb SA (2001). Beyondbody mass index. Obes Rev 2: 141-147.

Wagner DR and Heyward VH (2000).Measures of body composition inblacks and whites: a comparativereview. A m 1 Clin Nutr 71: 1392-1402.

WHO (1990). Diet, Nutrition, an d thePrevention of Chronic Disease.Report of a Joint WHO/FAO ExpertConsultation. WHO TechnicalReport Series, No. 916. World HealthOrganization, Ge-neva.

Documents

Tribal Populations of Maharashtra Cormic Index