World Bank Reprint Series: Number 309 Reynaldo Martorell, Joanne Leslie, and Peter R. Moock &,aracderistics and Dieteriisat of .ld Nutritional tatus in Nepal Reprinted with permission from The American Journal of Clinical Nutition, vol. 39 (January 1984), pp. 74-86. World Bank Reprints No. 269. James R. Follain, Jr., Gill-Chin Lim, and Bertrand Renaud, "Housing Crowding in Developing Countries and Willingness to Pay for Additional Space: The Case of Korea," Joumal of Development Economics No. 270. Bela Balassa, "Policy Responses to Extemal Shocks in Sub-Saharan African Countries," Journal of Policy Modeling No. 271. Jaime de Melo and Sherman Robinson, "Trade Adjustment Policies and Income Distribution in Three Archetype Developing Economies," Journal of Development Economics No. 272. J. B. Knight and R. H. Sabot, "The Role of the Firm in Wage Determination: An African Case Study," Oxford Economic Papers No. 273. William G. Tyler, 'The Anti-Export Bias in Commercial Policies and Export Perfor- mance: Some Evidence from Recent Brazilian Experience," Weltwirtschaftliches Archiv No. 274. Ron Duncan and Emst Lutz, "Penetration of Industrial Country Markets by Agricul- tural Products from Developing Countries," World Development No, 275. Malcolm D. Bale, "Food Prospects in the Developing Countries: A Qualified Optimistic View," Thle American Econo;niL Review (with Ronald C. Duncan) and 'World Agricultural Trade and Food Security: Emerging Patterns and Policy Directions," -\is'Lcoasi International Law Journal (with V. Roy Southworth) No. 2'. Sweder van Wijnbergen, "Interest Rate Management in LDCs," Journal of Monetary Economics No. 277. Oli Havrvlyshyn and Iradj Alikhani, "Is 'There Cause for Export Optmism? An Inquiry into the Existence of a Second Cene,maton of Sujccesshfl Exporters," 1U8'itiiiiiv;a'5c1.fIliclwes Archiv No. 278. ;,-)i Havrylyshyn and Martin I P./ c n Trends in Trade among Developing Countries, " X i Economic FR,Z No'9. `'ancy' Birdsall, "Fertility and Economic C aln"re in Eighteenth and Nineteenth Century Europe: A Comment," Po;nlti ii and i 4 pawn! Revieu No. 280 !iAalter Schaefer-Kehnert and 1T-01 D. Von PR;chke. "Agricultural Credit Policy in Developing Counltrie4.." ransiaku frohm n " b; der uld indwirIsLuat tl Endhrung in en iticihIii; -Len (includes original tJennan text) No. 281. Bela Balassa, "Trade Polic: in Mexico," Vvbrld Dc'rl non! lo. 28Ta. Bela Balissa, "La politica de ccirmercic) exterior de Mexico," ,Con rcln Exterior No. 282. Clive Bell and ShartayaFinan Devarajan "Shnadow Prices for Project Evaluation under Altemnative iv'acroerjonomic Spec,l ficaUi>!. .," Thte nQaatric, jfounr,l of Econo, iL.'s c,. 283. Anne 0. Krueger, "ITrade Policies in Developing Courtries" -indb;ooklci f Inita.'1iional Economics NXo. J,t. Anne 0. Krucocer and Baran Tuncer, 'An Empirical Test of the Infant Industry Argument," Anoric-ci. Econiomic Rc.okw No. 285. Bela Balassa, "Econaomic Policies in Pcrtugal," LECnoMIniI No. 286. F. Bourguignon, G. Michel, and D. Miqueu, "Short-run Rigidities and Long-run Adjustrnents in a Computable General Equilibrium Mvlode! of Income Distribution and Development," Jounal of Development Eco-gmics No. 287. Michael A. Cohen, 'The Challenge of Replicability: Toward a New Paradigm for Urban Shelter in Developing Countries," Regional Development Dialogue No. 288. Hollis B. Chenery, "Interaction between Theory and Observation in Development," IAbrld Development Characteristics and determinants of child nutritional status in Nepal' 2 Reynaldo Martorell,3 PhD, Joanne Leslie, DSc, and Peter R Moock, PhD ABSTRACT Herein, we examine characteristics and determinants of child malnutrition in the districts of Bara and Rautahat of the Terai region of Nepal. The sample studied consists of 510 rural children ranging in age from 3 to 10 yr. The Nepali children were found to have one of the highest reported prevalences of stunting (65% were <90% National study for Health Statistics median height for age). The study children were also I to 1.5 kg lighter when compared to US children of the same height. Fat deposits, as measured by anthropometric variables and Hb levels were also very low. Multiple regression analysis showed that age, district of residence, household income, breast-feeding, and several specific food items were significant predictors of nutritional status. Association with other factors such as caste and parental schooling, were not evident in multiple regressions. Boys were as likely to be malnourished as girls. Prolonged breast- feeding was associated with greater fat stores, but with reduced stature and low Hb values. Both landholdings and household income were found to be positively and significantly associated with almost all -measures of nutritional status, Am J Clin Nuir 1983;39:74-86. KEY WORDS Nutritional status, anthropomnetry, growth, socioeconomic status Introduction children that included anthropometry and Hb determinations (4). The objective herein to report on the anal- Nepal is one of the poorest nations of the ysis of health and nutrition data obtained as world. For example, the repored average part of a World Bank research project un- per capita income was only $140/yr as Of dertaken in the Terai region of Nepal. The 1980 (1). Infant mortality has been esti- data are regional in scope but their richness mated to be about 165 per thousand for the allows for an in-depth characterization of Terai region of Nepal using 1974 child sur- the growth pattern of Nepalese children and vivorship data (2). Other development indi- for the identification of some of the key cators are as follows: adult literacy is 19% determinants of nutritional status in this (1977 estimate), life expectancy at birth is population. 44 yr (1980 estimate), 9% of the population has access to safe water (1975 estimate), and Methods the average per capita daily supply of food This research was approved by the Bureau of Plan- energy is 2002 cal or 89% of the requirement ning Commission of His Majesty's Government of Ne- (1977 estimate) (1). ' From the Food Research Institute (RM), Stanford Data on health and nutrition in Nepal are University, Stanford, CA 94305; the Economic Devel- limited. The first comprehensive health sur- opment Institute, (JL), Washington, DC 20433; and vey was carried out in 1965 by the Univer- Teachers College (PRM), Columbia University, New sity of Hawaii (3). Dietary but no anthro- York, NY 10027. pomericdat wee clleced n tis lin- IThis paper was prepared as part of World Bank pometic data were collected in this clini- Research Project RPO 672- 1 0, "Health and Rural De- cally oriented study. Recently, the Centers velopment in Nepal." The World Bank, Washington, for Disease Control (CDC) and the United Dc. States Agency for International Develop- 3Autlior to whom requests for reprints should be addressed. ment assisted the government of Nepal in Received May 2, 1983. carrying out a national nutrition survey on Accepted for publication July 26, 1983. 74 The American Journal of Clinical Nutrition 39: JANUARY 1984, pp 74-86 Printed in U.S.A. © 1984 American Society for Clinical Nutrition MALNUTRITION IN NEPAL 75 pal. The data reported herein were obtained from a - ing and the timing of the introduction of solid foods survey of rural households in two of Nepal's 75 admin- was also obtained for each child. Socioeconomic and istrative districts. The districts, Bara and Rautahat, are other data including age and sex structure of the family, located in the Nepal Terai, a flat, fertile area straddling caste, land ownership, agricultural production, and level the border between Nepal and the Indian states of Bihar of schooling were obtained through standardized ques- and Uttar Pradesh. The sample studied is a random tionnaires. Written birth records were not generally sample of 15%c of the households in six panchayats of available for the children. Although mothers were rea- Bara (of 109 panchayats in the district) and six pan- sonably confident in stating the ages of young children, chayats of Rautahat (of 132). The households chosen the field workers did assist some by reference to annual for studv were interviewed on four separate occasions festivals and important local events. beginning in October 1977 as part of a World Bank An attempt was made to reduce the number of Study of education. agricultural productivity, and rural agricultuial production variables by creating a crude development. Each survey varied in content and em- measure of household income. The variable named phasis but was based on essentially the same households 'crop production value' (CRPVL) was computed as and individuals. This report is largely limited to the the sum of the outputs of the paddy and wheat crops data collected during the most recent survey carried out weighted by the respective farmgate prices of these two from December 1980 to January 1981 and includes 510 cereals. Virtually every farm (91%) cultivated paddy in children ranging in age from 3 to 10 yr of age (0 to 7 yr the 1980 growing season, and a smaller number (62%) in 1977, the first survey year). The 1980 to 1981 survey cultivated wheat. Paddy and wheat together accounted emphasized nutrition data collection and included an- for 1.0 ha of land usage that year on the "typical farm" thropometry and hemoglobin determinations as well as out of a total of 1.3 ha available. dietary frequencv questions. Fifty-six food items were included in the dietary The population of the Bara/Rautahat research site is questiorinaire. To reduce the number of variables in composed primarilv of middle- and low-caste Hindu the analyses, the 10 most variable items (eg, largest households from the business, farming, and occupa- variance) were selected. This procedure excludes foods tional castes. Muslims constitute 10% of the population that are consumed either only occasionally (eg, seasonal of Bara and 14% of that of Rautahat, but Muslim fruits) or very frequently (eg, rice). Only six ..f the 10 households were not included in the sample. Household 'high variance' dietary variables were included in the caste was coded as high caste (5.0%), middle caste final regression equations since earlier analyses showed (59.6%), and low caste (35.4%). The caste of a house- that the other four consistently failed to be significant hold, it should be noted, designates its ancestral affilia- predictors of any measure of nutritional status. The tion and occupational status (eg, rniddle caste as 'busi- dietary variables were recoded for the purposes of this ness" and "farming' castes) but does not necessarily analysis into three categories: I = not eaten last week, correspond to actual economic activity or occupational 2 = eaten occasionally last week, 3 = eaten daily. status today. In fact, all of the households were subsist- All data were scanned for outliers and a small per- ence farrm households. centage (less than I %) of cases with inconsistent or Four anthropometnic measures were collected fol- improbable values were deleted. A core sample of 510 lowing standard techniques: height, weight, arm cir- cases was available for the analysis. These 510 cases had cumference, and triceps skinfold. Height was measured complete data for height, weight, age, sex, district, crop to the nearest millimeter with a locally constructed value, and for weaning, dietary, and disease variables. wooden measuring board, and weight was measured to However, there were some missing values for other within 0.1 kg using a portable beam balance scale. The vanables which resulted in our using just 461 cases in Harpenden skinfold caliper was used to measure triceps the final regression equations. The SPSS software pack. skinfold, and the Zerfas insertion tape to measure arm age was used for all analyses. circumference, both to the nearest mm. Subroutines obtained from the Centers for Disease Control (CDC) Results were used to generate percent of median values and Z scores for weight, height, and weight for height. These Description of sample calculations use data from the National Center for Sample sizes, means, and SDs for height, Health Statistics (NCHS) as the reference or standard (5). Arm muscle area and arm fat area were calculated weight, arm circumference, triceps skinfold, as recommended by Frisancho (6). and Hb are shown in Table 1, separately for Capillary blood was collected by finger prick and boys and girls, and for eight age groups. Age analyzed in the field. Blood was put directly on a slide, 3 includes children 36 to 47 months of age, stirred with a reagent, and the slide inserted into a * 4 reading chamber (method of American Optical Co- age 4 ncludes the range 48 to 59 months pany). Values were not corrected for altitude since the etc. Because the majority of children have Terai. unlike the rest of Nepal, is only slightly above ages that are multiples of 12 months, the sea level. average age of each group is close to the Mothers were asked by trained Nepalese field work- starting month (eg, 36.6 months for group ers about their children's diarrheal and respiratory ill- nesses during the preceding week and about the fre- 36 to 47, 488 months for group 48 to 9, quency of consumption of major foods during the same etc). period. Information about the duration of breast-feed- Z scores for weight for height, weight, and 76 MARTORELL ET AL TABLE I Means and SDs 'or anthropometric measures in nepalese boys and girls (n = 510) Arm circum- Triceps HB Age Sample Ht (cm) Wt (kg) rerence (cm) skinfold (mm) (g/100 ml) size SD x SD x SD x SD x SD yr Males 3 16 87.2 5.9 11.9 1.3 13.9 1.0 3.8 0.8 11.3 1.8 4 20 92.5 7.4 13.0 2.0 14.1 1.0 3.7 0.8 12.2 2.1 5 46 95.7 7.3 13.8 2.2 14.0 1.1 3.4 0.7 11.7 1.7 6 48 101.4 8.0 15.2 2.1 14.0 1.0 3.1 0.6 11.5 1.5 7 43 108.4 6.0 16.4 2.0 14.1 0.8 2.7 0.5 11.9 1.6 8 38 114.3 8.4 18.4 2.8 14.5 1.0 2.6 0.6 11.8 1.8 9 34 116.6 5.2 19.2 2.5 14.8 1.0 2.4 0.4 11.6 1.8 10 14 124.5 7.8 23.2 3.9 16.1 1.2 2.6 0.6 11.5 1.7 Females 3 17 87.4 4.2 11.1 1.4 13.4 0.9 3.9 0.7 10.7 1.6 4 31 92.1 5.9 12.7 1.8 13.8 0.9 3.5 0.6 10.7 1.8 5 58 94.7 8.1 13.3 2.5 13.9 1.0 3.4 0.7 11.0 1.4 6 42 99.9 7.8 14.3 2.4 13.7 0.9 3.0 0.7 11.0 1.7 7 47 107.4 9.4 16.1 3.2 14.3 1.1 2.8 0.6 10.6 1.9 8 24 110.1 8.9 16.8 2.3 14.1 0.6 2.7 0.7 11.1 1.7 9 22 115.6 8.4 18.8 3.1 15.1 1.0 2.6 0.4 11.7 1.5 10 10 123.1 8.4 20.8 3.0 14.9 0.9 2.5 0.4 11.1 2.4 height are shown in Figure 1. The average Z in the case of the NCHS samnple. Relative to score (n = 5 10) was -2.74 for height, -2.26 height, the average arm muscle area of Ne- for weight, and -0.92 for weight for height. palese children is somewhat low in compar- In a well-nourished sample, Z scores for all ison with the NCHS norms but well within three measures would approach zero. There the normal distribution (ie, between the 50th are no statistically significant differences in and 5th percentiles). This is in marked con- Z score values associated with sex. The age trast to arm fat area, in which case Nepalese pattern in Z score values shown in Figure 1 values are substantially below the NCHS 5th seems to indicate progressive stunting until percentile. As would be expected from these about 5 yr of age with the situation stabiliz- relationships, the percentage of arm area ing or slightly improving thereafter. which is muscle is considerably greater in Frisancho (6) has recently published the Nepalese sample than in the NCHS sam- norms for arm circumference, triceps skin- ple. As was the case for other figures, pat- fold, and upper arm fat and muscle areas. terns for girls were similar to those for boys. These data come from the same population Hb values of the Nepalese children are as the NCHS norms for height and weight. very low when compared to US sea level At present, no subroutines are available to norms given by Dallman and Siimes (7). For the public for calculating Z score values. In Nepalese girls, mean values fall below the Figures 2 and 3, the NCHS 50th and 5th 3rd US percentile while for boys they vary percentiles and the Nepalese means are plot- between the 3rd and the 10th percentile. ted for triceps and arm circumference re- Also, Hb values do not rise with age in the spectively for the sample of boys (sample Nepalese sample as is observed in the US sizes as in Table 1). The pattern in girls was population (see Table 1). found to be similar. Means for arm muscle and fat area are Factors influencing nutritional status plotted against mean height for the Nepalese Regressions were carried out for both Z sample in Figure 4. For comparison, the scores and percentages of median values for 50th and 5th percentiles of arm muscle and height, weight, and weight for height. Since fat area are plotted against median heights similar results were obtained for both types MALNUTRITION IN NEPAL 77 0 33 51 104 90 90 62 55 24 SAIPLE SIZES e- WEIGHT FOR HEIGHT I- ~ z- e, -1I"------,-------------------------------- Z SCORES -2 WEIGRT HEIGHT -3 3 4 5 6 7 8 9 10 AGE (YEARS) FIG 1. Anthropometry Z scores in Nepalese children. of dependent variables and since percentages fed for nearly 3 yr on average and solids of median values have been used more often were not introduced until about 18 months. in previous studies, results for the Z score Eighteen percent of the children were re- regressions are not reported here. Variable ported to have had diarrhea the previous definitions, sample sizes, and mean and SDs week and 41% were said to have had respi- for dependent, background, dietary, and ratory infections. morbidity variables used in the multiple Most of the independent variables are sig- regressions are given in Table 2. Mean per- nificantly related to one or more of the centages of median values are 88 for height, dependent variables in the expected direc- 73 for weight, and 93 for weight for height. tion as shown in Table 3. Greater values for Families had access to very limited amounts the variables LAND and CRPVL are asso- of land and their total agricultural produc- ciated with better growth as measured by all tion was very low. Only 26% of fathers and of the anthropometric variables except tri- 7% of mothers had ever been to school. ceps skinfold and fat area. Increased parental Mothers reported that children were breast- age and schooling are associated with im- 78 MARTORELL ET AL NCHS 50th PERCENTILE 20 - 18 - /NCHS 5th PERCENTILE Z 1NEPAL MEAN 16 14 12 3 4 5 6 7 8 9 1O AGE (YEARS) FIG 2. Arm circumference: males. 15 10 NCHS 50th PERCENTILE NCHS 5th PERCENTILE 5 NEPAL MEAN 0 3 4 5 6 '7 8 9 10 AGE (YEARS) FIG 3. Triceps skinfold: males. proved growth, and caste is related to growth itive association with triceps and fat area, in the expected direction, with high caste there is no association for respiratory ill- status being associated with greater fatness, nesses. A history of prolonged breast-feeding and low caste with smaller heights and is associated with shorter heights and weights. No significant associations with months of exclusive breast-feeding (ie, late diarrhea are observed, and except for a pos- introduction of solids) is associated with MALNUTRITION IN NEPAL 79 -1-- I - I - I I-' NCHS MUSCLE, 20 50Sth PERCENTiLE * NEPAL MUSCLE, ~"MEAN 15NCHS MUSCLE, - -. 1 5 5th PERCENTILE w 10 NCHS FAT, 50th PERCENTILE 5 p * pNCHS FAT, 5 (3.5) (45 th PERCENTILE v0 - - -o- - -7-.0 NEPAL FAT, MEAN 0 (3.1)(4.1)15.1) (6.1) (7.O)(U.fl(9.O) (10.3) so 90 100 110 120 130 140 150 HEIGHT (cm) FIG 4. Arm muscle and arm fat area plotted on height for males (ages in parentheses). TABLE 2 Variable definitions, means, and SDs for variables used in multiple regressions (n = 461) Vanable description Mean SD Dependent variables HA Percent -of NCHS median ht for age 88.47 6.43 WA Percent of NCHS median wt for age 73.22 11.89 WH Percent of NCHS median wt for ht for age 92.97 8.44 AC Arm circumference (cm) 14.19 1.07 TRICEPS Triceps skinfold (mm) 3.04 0.75 MAREA Muscle area (mm2)/I0 138.41 231.60 FAREA Fat area (mm2)/l0 20.54 51.72 HB (g/dl) 11.32 1.77 Background variables AGE Age of child (mo) 76.67 22.41 AGESQ Age of child squared 6,379.18 3,578.17 SEX I = male, 2 = female 1.49 0.50 DISTRICT I = Bara, 2 Rautahat 1.56 0.50 CRPVL Value of crop output (100 Rs) 23.96 32.92 LAND Land Owned (Bighas)* 2.07 3.78 MOAGE Age of mother (yr) 33.80 8.11 HICST High caste (I = high, 0 = others) 0.05 0.23 LWCST Low caste (I low, 0 = others) 0.36 0.48 MOSCH Mother ever been to school (I = yes, 0 = no) 0.07 0.25 FASCH Father ever been to school (I = yes, 0 = no) 0.26 0.44 Dietary variables MBF Months of breast-feeding 34.19 11.26 MEBF Months of exclusive breast-feeding 18.20 7.73 BRICE Beaten rice last wk (I = none, 2 some, 3 = daily) 1.83 0.69 ROTI Roti last week (I = none, 2 = some, 3 = daily) 1.52 0.61 PPEAS Pigeon peas last wk (I = none, 2 = some, 3 = daily) 1.34 0.64 SWPOT Sweet potatoes last wk (I = none, 2 = some, 3 = daily) 2.15 0.61 Sugar Sugar or molasses last wk (I = none, 2 = some, 3 = daily) 1.87 0.74 TURNIP Turnips last week (1 = none, 2 = some, 3 = daily) 1.93 0.63 Morbidity variables RESP Resriratory infection last wk (1 yes, 0 = no) 0.41 0.49 DIARR Diarrhea last wk (I = yes, 0 = no) 0.18 0.39 * I Bigha = 1.6 acres = 0.6 ha. 0 TABLE 3 Significant correlations between dependent and independent variables (n = 46 1) Ha Wa Wh AC TRICEPS MAREA Vanables (% ormedian (% of median (% of median (arm circum- (Iriceps (munscle FAREA HB ht) WI) W/H) fercene) skinfbId) area) (fat arca) Background AGE -0.297* -0.226* 0.356* .0.559* 0.470* -0.432* 0.089t SEX -0.077t -0.089t -0.220* DISTRICT -0.1 12$ -0.137t -0.258* -0.252* CRPVL Crop output value 0.157* 0.169* 0.214* 0.205* 0.083t 0.094t LAND Land 0.142* 0135* 0.199* 0.201* MOAGE Mother's age 0.176* -0.101t 0.196* 0.080t HICST High caste 0.109t 0.109t LWCST Low caste -0.092t -0.1 16f > MOSCH School (mother) 0.080t 0.088t 0.103t 0.099t 0.084t W FASCH School (father) 0.078t 0.078t 0.093t C.090t O ri, Dietary MBF Mos. breast fed -0.096t MEBF Mos. only breast 0.107t 0.154* 0.139* -0.106t i BRICE Beaten rice 0.098t 0.137t 0.087t 0.130t 0.105* 0.108t 0.139+ > ROTI Roti 0.089t 0.083t r PPEAS Pigeon peas 0.080t 0.102t 0.135f 0.1 19f 0.102t SWPOT Sweet potatoes 0.077t SUGAR Sugar/molasses 0. 124t 0.1 15f 0.155* 0. 1 50* TURNIP Turnips Morbidity RESP Respiratory DIARR Diarrhea -0.078t 4 p < 0.00 1. tp<0.O5. fp <0.01. TABLE 4 Determinants ofanthropometry and Hb (n = 461) HA WA WH AC TRICEPS MAREA FAREA Hh Vanabics b b b ( b b b t b t Background Age -0.216 2.64* -0.171 1.18 -0.118 1.10 -0.025 1.98t -0.048 6.20t -2.783 1.08 -3.251 5.65t 0.007 0.29 AGESQ 0.001 2.51t 0.000 0.03 0.000 0.27 0.000 3.29t 0.000 3.81t 0.047 2.92* 0.014 3.95t -0.000 0.16 SEX -0.460 0.77 -0.094 0.09 0.672 0.86 -0.039 0.43 0.066 1.18 -11.882 0.63 3.472 0.83 -0.774 4.70 DISTRICT -1.058 1.49 -2442 1.9St -0.736 0.79 -0.112 1.03 -0.330 4.96f -2.459 0.11 -22.785 4.58t 0.182 0.93 CRPVL 0.015 1.34 0 035 1.76 -0.016 1.10 0.005 2.94* 0.000 0.63 1.004 2.86* 0.011 1.39 0.003 1.07 MOAGE 0.088 2.33t 60.193 2.76t 0.046 0.89 0.012 2.03t 0.002 0.64 2.476 1.99t 0.315 1.14 0.012 1.09 HICST -1.341 0.90 - 1.569 0.59 0.621 0.32 -0.132 0.58 0.208 1.48 -43.000 0.91 11.823 1.13 0.154 0.38 i LWCST -0.690 1.06 -1.499 1.30 -0.085 0.10 0.045 0.45 0.040 0.65 4.881 0.24 2.932 0.64 -0.124 0.69 > MOSCH 1.389 0.94 3.494 1.33 1.053 0.54 0.364 1.60 0.123 0.89 70.064 1.50 11.574 1.11 0.166 0.41 Z FASCH 0.469 0.58 -0.395 0.27 -1.621 1.52 -0.006 0.05 -0.127 1.66 7.197 0.28 -8.585 1.50 0.013 0.05 Dietary MBF -0.068 2.44t -0.059 1.19 0.056 1.53 0.003 0.79 0.005 1.75 0.432 0.49 0.338 1.71 -0.003 0.34 0 MEBF -0.026 -.62 -0.019 0.26 0.058 1.07 -0.006 0.87 0.008 1.98t -1.624 1.25 0.418 1.44 -0.025 2.18t Z BRICE 0.202 0.41 0.942 1.07 0.639 0.99 0.041 0.55 0.031 0.67 7.008 0.45 3.113 0.89 0.082 0.60 z ROTI -1.055 2.07t -1.170 1.30 0.692 1.04 0.121 1.54 0.046 0.95 22.538 1.40 4.323 1.21 0.134 0.96 z PPEAS -0.121 0.23 0.943 0.99 1.330 1.90 0.132 1.60 0.068 1.36 23.234 1.38 6.536 1.74 0.059 0.40 m SWPOT -0.631 1.21' -1.061 1.15 -0.217 0.32 -0.005 0.05 0.032 0.65 -2.800 0.17 1.080 0.46 0.053 0.37 > SUGAR 0.923 1.91 0.646 0.75 -0.663 1.05 -0.017 0.23 -0.031 0.68 0.046 0.00 -2.516 0.74 -0.003 0.00 r TURNIP -0.240 0.48 -1.147 1.31 -0.980 1.51 *-0.225 2.96t -0.029 0.62 -44.384 2.85* -4.980 1.43 -0.048 0.35 Morbidity RESP 0.522 0.82 1.448 1.29 0.515 0.62 0.161 1.65 0.000 0.00 36.310 1.82 2.399 0.54 -0.254 1.45 DIARR -0.311 0.39 -0.633 0.45 0.009 0.00 -0.007 0.05 -0.062 0.83 . 5.038 0.20 -4.463 0,80 -0.060 0.28 Constant 100.033 87.209 95.832 14.059 5.387 1,216.36 350.932 11.563 0.094 0.160' 0.098 0.030 0.415 0.306 0.308 0.092 *p <0.0I. tp<0.05. * p < 0.001. Two-tailed tests. CI 82 MARTORELL E-' AL greater fatness (increased triceps, fat area, larger arm circumferences and arm muscle and weight for height) and with lower Hb areas, but lower Hb levels. values. More frequent consumption of CRPVL is one of the variables in Tables foods, particularly beaten rice, roti, pigeon 3 and 4 that show consistent associations peas, and sugar, is associated with better with the anthropometric variables and Hb. growth. Table 5 presents additional analyses and Multiple regression equations are shown shows how the distribution of children class- in Table 4 for all of the anthropometric ified as having "poor nutritional status" var- variables and for hemoglobin. All tests of ied by terciles of CRPVL. Table 5 consists significance given are two-tailed. The varia- of five widely used indicators of nutritional ble LAND is highly correlated with CRPVL status: stunting, wasting, the "color codes" (r = 0.820, N = 461, p < 0.001). Since both of arm circumference, the G6mez classifi- variables were thought to measure the same cations of weight for age, and anemia clas- construct and since CPRVL was found to be sifications. (See Table 5 for precise defini- related more consistently to the anthropo- tions of all of the indicators.) A very large metric variables (see Table 3), LAND has percentage of the sample is stunted (331 of been deleted from the list of independent 510 cases or 64.9%); however, the preva- variables. The results in Table 4 indicate lence of stunting significantly decreases as that age is a significant predictor of the CRPVL increases. Wasting is seen to exist anthropometric variables. The fact that age for 4. 1% of the sample, but whereas it is squared is positive and statistically signifi- 6.3% for those in the low tercile of CRPVL, cant in a number of the regressions suggests it diminishes to only 0.6% in the upper that the relationship between age and the tercile. While 20.6% of the children whose anthropometric variables is not linear but families had low CRPVL values fall into U-shaped as was evident for height in Figure yellow and red categonres of arm circumfer- 1. DISTRICT is negatively associated with ence, onlv 14.9% do so in the case of those all of the anthropometric variables, suggest- whose families were in the high CRPVL ing that children from Bara are "better off" group. The percentage of children classified in ways not captured by the included back- as either moderately or severely malnour- ground measures. The sex of the child is not ished according to the G6mez categories of an important predictor of any of the anthro- weight for age is 66.4% for the bottom ter- pometric variables: boys, however, are seen cile, 64.2% for the middle tercile, and 54.6% to have significantly greater Hb values than for the top tercile of CRPVL. Finally, 36.1% do girls. The associations observed previ- of the children show some degree of anemia, ously in the table of simple correlations (Ta- but no relationship with CRPVL is evident. ble 3) for caste (HICST and LWCST) and schooling (FASCH and MOSCH) are not Discussion statistically significant in the multiple regres- sions. A longer duration of breast feeding The extent of malnutrition found in this (MBF) is significantly associated with re- study is among the highest that has been duced stature (HA) and there is also a tend- reported outside of extraordinary depriva- ency for positive associations with fatness tions occurring during events such as wars (weight for height, TRICEPS and FAREA). and famines. The results of the 1980 to 1981 This pattern is consistent with negative as- survey were that 65% of the children were sociations between stature and fatness in the stunted, 4% were wasted, and 12% were sample (HA and TRICEPS: r = -0.18, p < G6mez 3 (less than 60% of weight for age). 0.001; HA and FAREA: r = -0.07, p < Results of earlier surveys in the same house- 0.10). Similarly, the longer the duration of holds also indicate that the rates of malnu- exclusive breast-feeding (MEBF), the greater trition were very high (8, 9). A national are the fat stores and the lower are the Hb nutrition survey coordinated by the CDC (4) values. Some of the dietary variables are also reported high rates of malnutrition in a significant predictors and children reporting sample of Terai children 6 months to 6 yr respiratory illnesses the previous week have of age; 45% of the CDC sample was found MALNUTRITION IN NEPAL 83 - . -oto be stunted and 9% to be wasted. The c E ' difference in age between the two samples Al may account for the differences in the degree L .c2 . . <' of stunting and wasting. a r The CDC-coordinated survey also in- va 8 cluded a Nepal urban population for com- n m o X en < = parison. Twenty-nine percent of this sample was found to be stunted, and 2% was found c o to be wasted (4). Similarly, Farquharson (10) %- CO> C, a, > ` - ,n r measured a rural 'well-to-do" group chosen Nn C) from among children attending private clin- I VI ics or schools and from among the children o o C w ^ A of health workers. Visual inspection of the z = V - '-I vi growth curves prepared by Farquharson (10) - O shows-that relatively affluent rural children VI °grow better than do their poor counterparts. °E Means for the well-to-do were plotted be- = - - = m- tween the 10th and the 50th percentile of W * ' £ A the. British standards, while those for the S Al poorer children fell far below the 10th per- e A v ( centile. Few details were provided by Brink o >q er- 4 ,o a, -J :: et al (4) and Farquharson (10) on the char- M.; v a acteristics of the well-to-do samples, and it is difficult to assess whether their results r co C °°°-Erepresent the genetic potential of the Ne- - z Br 3palese. Studies from northern India, for ex- 2the same anthropometric characteristics as British and American children (11-13). V I A < The Hb values of the Bara/Rautahat sam- > u - 3ple were very low. Equally low values have O EV been reported for Nepalese children by - - Brink et al (4), as well as for Indian children mation in Nepal (1 (14-d7). -C It is difficult to obtain accurate age infor- i s £ >2 >mation in Nepal (19). Inspection of the data E shows that 83% of the children have re- .c =- - - ported ages that were multiples of 12 VI months. There is clearly an excessive tend- - - ency toward "heaping" around completed i. s, years since, if age were randomly distributed, , i' ° only 8% of the children would be expected :1 ato have ages divisible by 12. Mothers are Gbelieved to have rounded upward when O- v 6 asked their children's ages. The consequence . z 06 - 0-bo C- a, . OC V C¢ , °>1 VI of this is that many of the analyses have < compared younger Nepalese children to VI 3older children in the NCHS sample and thus o q overestimated the magnitude of retardation , eA V V in age-dependent measures such as height n E and arm muscle area. As the average age of 62 'P 3othe study sample is underestimated by no H U X more than 6 months, however, the bias in- 84 MARTORELL ET AL troduced is small. Indeed, analyses of vari- ciated with caste in poor rural families of ance showed no difference in height, weight, Nepal. On the other hand, two Indian stud- or weight for height between "heapers" and ies in the Punjab found caste to be a signif- "nonheapers." Also, the two groups were icant predictor of child nutritional status evenly distributed across sex and district (22, 23). No differences were found between lines and showed no socioeconomic differ- boys and girls in the degree of growth retar- ences (eg, CRPVL, LAND). dation relative to the NCHS norms. Differ- Unlike malnourished children from Latin ential treatment favoring boys over girls may America (18, 19), Nepalese children are not not be as marked in Nepal as reported for able to maintain normal weight to height India (23). relationships but are instead 1 to 1.5 kg Nepalese babies are breast-fed for long lighter when compared with well-nourished periods of time. Farquharson (10) states that children of the sarrie height. Relative to most are breast-fed for at least 2 yr and height, arm muscle area in the Nepalese frequently longer. Brown et al (21) found sample is smaller than in the NCHS sample, that nearly all children were breast-fed for while muscle areas in Latin American chil- more than 18 months, and over half were dren are normal once size is taken into breast-fed for more than 2 yr. The duration account (19, 20). Nepalese and Guatemalan found in this study was 34 months, which children show markedly lower arm fat areas does not seem unreasonable in view of pre- when compared to NCHS children of the vious findings. Solids tend to be introduced same size (20); however, Brazilian children to the diets of Nepalese infants at later ages show fat areas closer to US norms (19). We than in many areas of the world (10, 21). hypothesize that the growth patterns in Ne- However, the finding in this study that the pal, Guatemala, and Brazil reflect different average duration of exclusive breast. feeding degrees of severity of the same underlying is 18 months is probably an exaggeration. response to protein-energy malnutrition. Perhaps, the question was not properly The severity of protein-energy malnutrition understood and instead the results refer to is greater in Nepal than in Guatemala and the timing of the full establishment of the Brazil, and Latin American children may be regular adult diet. The variable MEBF able to cope with the stress of malnutrition (months of exclusive breast-feeding) was by growing less in body size but without found to be positively associated with greater altering weight/height ratios. The situation fatness (ie, skinfold, fat area, and wei!'ht for in Nepal may reflect more severe conditions height). MEBF was negatively assc?iated, which demand more powerful adaptive however, with the level of Hb, a finding that mechanisms. In Nepal, growth in height is might suggest a difficulty in recovering good affected to an even greater degree, and iron status in this population, even several weight for height ratios are diminished as years after the introduction of food solids. well. The total length of breast-feeding (MBF) was The data presented here are also inform- not found to be related to physical growth ative of the correlates of nutritional status measures in bivariate fashion but, after con- measures in Nepal. Caste and sex were two trolling for all other independent factors, the such variables investigated. High caste was duration of breast-feeding is associated with associated with increased skinfolds and fat a greater degree of stunting and with larger areas, and low caste was associated with skinfolds and fat areas. This is a population reduced heights and weights (Table 3). Once reporting universal and prolonged breast- other socioeconomic indicators were taken feeding, and the association with stunting into account, however, all associations with should be interpreted with caution. caste lost their statistical significance (Table The analysis of previous survey data 4). Brink et al (4) found caste to be essen- pointed to a significant negative association tiallv unrelated to nutritional status in the between infectious disease and nutritional Nepal national nutrition survey. Also, status, particularly between diarrhea and low Brown et al (21) reported there were no weight for height (8). This same association marked differences in nutrient intakes asso- was not found in the analysis of the 1980 to MALNUTRITION IN NEPAL 85 1981 data. The fact that the children in the References 1980 to 1981 survey were past the age of 1. World Bank. World Development Repor. New peak diarrheal illness, combined with the York, NY: Oxford University Press. 1982. fact that the diarrhea data referred only to 2. Thapa S, Rutherford RD. Infant monality esti- the previous week, probably explains the mates based on the 1976 Nepal fertility survey. finding of no association between diarrhea Popul Stud 1982;36:61-80. and nutritional status. Findings with respect 3. Worth RM, Shah NK. Nepal health survey, 1965- o u r r 1966. Honolulu, HI: University of Hawaii Press, to upper respira.tory Illness were inconisist- 1969. ent. 4. Brink EW. Khan IH, Splitter JL, Staehling NW, The usual diet of the Nepalese Terai pop- Lane JM, Nichaman MZ. Nutritiocal status of ulation consists primarily of rice, dhal children in Nepal. WHO Bull 1976;54:311-18. (cracked legumes cooked to a soup-like c>n- 5. Hamill PVV, Drizd TA, Johnson CL, Reed RB, k lRoche AF, Moore WM. Physical growth: National sistency), a vegetable curry (made predomi- Center for Health Statistics percentiles. Am J Clin nantly of potatoes and a variety of coarse Nutr 1979;32:607-29. spinaches), small amounts of milk, and oc- 6. Frisancho AR. New norms of upper limb fat and casional fruits and vegetables (10, 21). Many muscle areas for assessment of nutritional status. families follow the Hindu prohibition Am J Clin Nutr 1981;34:2540-5. 7. Dallman PR, Siimes MA. Iron deficiency in in- against meat. Dietary data in this study were fancy and childhood. Washington DC: The Nutri- limited to the frequency of consumption of tion Foundation, 1979. (Report for the Interna- particular food items.The 10 most ?ariable tional Nutritional Anemia Consultative Group.) dietary variables were considered in the 8. Leslie J, Baidya BG, Nandwani K. Prevalence and analyses and it was found that a number of ccorrelates of childhood malnutrition in the Terai analsesandit ws fundthata nmbe of region of Nepal. Washington, DC: The World them were consistently related to nutritional Bank, 1981. (Population and Human Resources status. However, most of these associations Division discussion paper no 81-35.) were considerably attenuated after control- 9. Leslie J. Family characteristics and the nutritional ling for other factors in multivariate anal- status of children in the Terai region of Nepal. Baltimore, MD: Johns Hopkins University, School yses. Some variables were negatively related of Hygiene and Public Health, MSc thesis, :978. to growth in the multivariate analyses: roti 10. Farquharson SM. Growth pattems and nutrition (with height) and turnips (with arm circum- in Nepali children. Arch Dis Child 1976;51:3-12. ference and muscle area). While never Sig- 1. Udani PM. Physical growth of children in different socioec6nomic groups in Bombay. Ind j Child nificant, associations involving sweet pota- Health 1963;1 2:593-61 1. toes were often negative. Sweet potatoes and 12. Banik NDD, Nayar S, Krishna R, Raj L. The effect turnips are known to be foods that are con- of nutrition on growth of pre-school children in sumed more frequently by poorer families. different commnunities in Delhi. Ind Pediatr 1972; Also, the code of Manu forbids the higher 9:460-6. 13. Currimbhoy Z. Growth and development of Bom- castes from eating turnips (24). bay children. Ind J Child Health 1963;12:627-51. Finally, socioeconomic variables (CRPVL 14. Rao KV, Radhaiah G, Raju SVS. Association of and LAND) were found to be significant growth status and the prevalence of anemia in pre- determinants of nutritional status. Studies school children. Ind J Med Res 1980;71:237-46. in India (15, 22, 23) and Latin America (25) 15. Devadas RP, Rajalakshmi R, Kaveri R. Influence of family income and parent's education on the have found that income, for which CRPVL nutritional status of pre-school children. Ind J Nutr is a proxy in this study, is one of the most Diet 1980;17:237-44. significant predictors of nutritional status. 16. Uberoi IS, DeSweemer C, Taylor CE. A study of The amount of land available to the family anemia among rural Punjabi children. Ind J Med Res 1972-,60.793-9. is fnother important variable predictive of 17. Ghai OP, Chaudhuri SN, Kapur S, Jaiswal VN, nutritional status, as has been shown in Sinclair S. Nutritional assessment of pre-school other studies as well (26-29). These findings children of a rural area. Ind J Med Res reinforce the views that protein-energy mal- 1 970;58:1621-8. nutrition is a problem of poverty and suggest 18. Yarbrough C, Habicht J-P, Malina RM, Lechtig A, Klein RE. Length and weight in rural Guatemalan that substantial improvement is likely to Ladino children: birth to seven years of age. Am J come from raising the standard of living. E Phys Anthropol 1975;42:439-48. 86 MARTORELL ET AL 19. Leslie J. Child malnutrition and diarrhea: a longi- age JR, Lubawski JL. Food and people. 3rd ed. tudinal study from northeast Brazil. Baltimore, New York, NY: John Wiley and Sons, 1979. MD: Johns Hopkins School of Hygiene and Public 25. Valverde V, Mejia-Pivaral V, Delgado H, Belizan Health, DSc thesis, 1982. J, Klein RE, Martorell R. Income and growth 20. Martorell R, Yarbrough C. Lechtig A. Delgado H, retardation in poor families with similar living con- Klein RE. Upper arm anthropometric indicators of ditions in rural Guatemala. Ecol Food Nutr nutritional status. Am J Clin Nutr 1976;29:46-53. 1981; 10:241-8. 21. Brown ML, Worth RM, Shah NK. Food habits 26. Desai P, Standard KL, Miall WE. Socioeconomic and food intake in Nepal. Trop Geogr Med and cultural influences on child growth in rural 1968;20:217-24. Jamaica. J Biosci 1970:2:133-43. 22. Kielmann NS, Kielmann AA, Grewal GK. Etio- 27. Rawson IG, Valverde V. The etiology of malnutri- logical determinants of protein calorie undernutri- tion among preschool children in Costa Rica. En- tion in a rural child community. Acta Tropica viron Child Health 1976;22:12-17. 1976;33:342-68. 28. Valverde V, Martorell R, Mejia-Pivaral V, et al. 23. Levinson FJ. Morinda: an economic analysis of Relationship between family land availability and malnutrition among young children in rural India. nutritional status. Ecol Food Nutr 1977:6:1-7. Cambridge, MA: Cornell-MIT International Nutri- 29. McDowell I, Hoorweg J. Environmental risk fac- tion Policy Series, 1974. tors in outpatient recovery from malnutrition. Ecol 24. Lowenberg ME, Todhunter EN, Wilson ED, Sav- Food Nutr 1977:6:23-30. The American Society for Clinical Nutrition does not endorse any commercial program or product. No. 289. J. B. Knight and R. H. Sabot, "Educationial Expansion and the Kuznets Effect," The American Economic Review No. 290. Malcolm D. Bale and Urich Koester, 'Maginot Line of European Farmn Policies," The TA.brld Economy No. 291. Danny M. Leipziger, "Lending versus Giving: The Economics of Foreign Assistance," irld Development No. 292. Gregory K. Ingram, "Land in Perspective: Its Role in the Structure of Cities," World Congress on Land Policy, 1980 No. 293. Rakesh Mohan and Rodrigo Villamizar, "The Evolution of Land Values in the Context of Rapid Urban Growth: A Case Study of Bogota and Cali, Colombia," World Congress on Land Policy, 1980 No. 294. Barend A. de Vries, "International Ramifications of the External Debt Situation," The AMEX Bank Review Special Papers No. 295. Rakesh Mohan, "The Morphology of Urbanisation in India," Economic and Political Weekly No. 296. Dean T. jamison and Peter R. Moock, "Farmer Education and Farm Efficiency in Nepal: The Role of Schooling, Extension Services, and Cognitive Slklls," World Development No. 297. Sweder van Wijnbergen, "The 'Dutch Disease': A Disease after All?" Tlhe Economic jounial No. 298. Ame Drud and Wafik M. Grais, "Macroeconomic Adjustment in Thailand: Demand ianagerment and Suppiy Conditions," Journal of Policy Modeling No. 299. Shujiir Urata, "Factor Inputs and Japanese Manufacturing Trade Structure," The Reoicw of Lcoioniuts and Statistics No. 300. Dipak lazurdar, "The Ruiral-Urban Wage Gap Migration and the Working of Urban Labor Market: An interprcEl*tioii Based on a Study of the Workers of Bombay City," hndian Economic Revie No. 301. Gershon Feder and Roger Slade, "Contact Farmer Selection and Extension Visits: The Training and Visit Extension System in Haryana, India," Qunrterlt, Jonrial of Itterna- tional.Agricid!ture No. 302. James Hanson and Jaime de Melo, "The Uruguayan Experie.nce with Liberalization and Stabilization, 1974-1981," Journal of Interamerican Studies and World Affairs No. 3i,-. Nancy Birdsall and Dean T. Jan-iison, "Income and Other Factors Influencing Fertility in China," f.?1pda!iv;i and Developmen t Rev'iew Ne. 304. Graham Donaldson,"Food Slcunity and the Role of the Grain Trade," American Journal Of Agricultural Economics No. 305. William F. Steel and Yasucki Takagi, "Small Enterprise Development and the Wr.-,o menl- Output Trade-Off," Oxford Economic Papers No. 306. Oli Ha-vi'lyshyn and Engin Civan, "Intra-Industry Trade and the Stage of Develop- ment: A Regression Analysis of Industrial and Developing Countries," Intra-Industry Trade: Empirical and Methodological Aspects No. 307. Mateen Thobani, "A Nested Logit Model of Travel Mode to Work and Auto Ownership," Journal of Urban Economics No. 308. Johannes Bisschop and Alexander Meeraus, "On the Development of a General Algebraic Modeling System in a Strategic Planning Environment," Mathematical Programming Study Issues of the World Bank Reprint Series are available free of charge from the address on the bottom of the back cover.