The Journal of Health, Population and Nutrition icddr,b ISSN: 1606-0997 EISSN: 2072-1315 Vol. 25, Num. 3, 2007, pp. 359-369

Journal of Health, Population and Nutrition, Vol. 25, No. 3, September, 2007, pp. 359-369

Determinants of Persistent Underweight among Children, Aged 6-35 Months, after Huge Economic Development and Improvements in Health Services in Oman

Deena Alasfoor¹ , Pierre Traissac² , Agnès Gartner², Francis Delpeuch²

¹ Department of Nutrition, Ministry of Health, Sultanate of Oman and ²Nutrition Unit, instituted de Recherche pour le Développement (IRD) (WHO Collaborating Centre for Nutrition), B.P. 64501, Montpellier, France
Correspondence and reprint requests should be addressed to: Dr. P. Traissac Nutrition Unit Institut de Recherche pour le Développement (IRD) B.P. 64501 911 avenue Agropolis 34394 Montpellier Cedex 5 France Email: pierre. traissac@ird.fr Fax: +33 4 67 41 63 30

Code Number: hn07047 

ABSTRACT

Over the last decades, health indicators have witnessed major improvements in the Sultanate of Oman. This study was aimed at factors associated with underweight among children in four regions of Oman, as, in 1998, underweight was prevalent among 17.9% of children aged less than five years. A case-control study was conducted in 2002: 190 cases were 6-35-month old children with weight-for-age <-2 z-scores. Controls were individually matched by village of residence, sex, and age. The questionnaire included an­thropometry of children, child-feeding practices, morbidity, anthropometry of mothers, parity, birth-spac­ing, and socioeconomic characteristics. Conditional logistic regression was used for analyses. Birth-weight of <2,500 g was strongly associated with underweight and also were height of mother, low level of educa­tion of mother, bad quality of water in households, diarrhoea of children in the last two weeks, and regular use of infant formula. Factors, such as birth-weight, height of mother, supply of safe water in household, and care for mothers and children were the determinants of persistent underweight after huge economic development and improvements in health services. Further research is also needed to investigate further specific determinants of low birth-weight in the Omani context and try to disentangle emaciation and determinants of linear growth retardation.  

Key words: Birth-weight; Case-control studies; Childcare; Child nutrition disorders; Underweight; Oman

Introduction

In the last three decades, due to oil revenues and policy choices, the Sultanate of Oman has under­gone drastic economic and sociodemographic changes. Between 1975 and 2003, its human development status, as assessed by the human develop­ment index (HDI), experienced the world’s largest observed increase for this period, from 0.493 to 0.780 (1). Oman now ranks 71 among 177 nations in the first quarter of the medium human develop­ment countries. In the same period, most health indicators witnessed major improvements (2). Be­tween 1960 and 2002, the rate of infant mortality dropped from 164% to 11%, and the rate of mortal­ity of children aged less than five years (under-five mortality) went down from 280% to 13% (3). Rate of infant immunization rose from 10% in 1980 to 99% in 2001 (4). Also, during the same period, the prevalence of different forms of undernutrition de­creased markedly. Between 1980 and 1998, under­weight among Omani children aged less than five years decreased from 62.9% to 17.9% and stunting from 20.3% to 10.4% (5,6). Wasting decreased from 12.8% in 1995 to 7.2% in 1998 (6). By the interna­tional standards of the World Health Organization (WHO) for prevalence of child malnutrition (7), level of underweight in Oman is still in the ‘medi­um’ range, even if the health of all Omani children can be monitored due to sufficient resources and adequate heath services.

The conceptual framework of our analysis is that of the international conceptual framework of the causes of malnutrition (8). Due to the spe­cific Omani context, our hypotheses were that (a) the factors from the ‘health environment and ser­vices’ (with the exception of water supply which is historically important in Oman) and ‘household food security’ categories of the underlying causes of malnutrition (8) could be considered at a sufficient level (9) and (b) that risk factors of underweight could, thus, be more likely found in the child and maternal care categories (also without excluding the legacy of the predevelopment years via prenatal factors). Despite a few descriptive studies (10,11) or aimed at specific risk factors (10,12), no data were available to assess the relationships between a suf­ficiently large number of factors and the anthropo­metric status of children. Therefore, this study was aimed at assessing the risk factors of underweight among young Omani children. For that purpose, a matched case-control study was conducted among 6-35-month old children residing in four regions of Oman.

Materials and Methods

Study design  

A 1:1 case-control study with individual matching was deemed to be the most efficient for the study (13,14). Cases were defined as underweight chil­dren aged 6-35 months. Controls were non-under­weight children matched by village of residence, sex, and age (within one month).

Study sample  

For a 0.05 first-type error risk and a power of 0.80, assuming a 20% prevalence of exposure among controls, the sample size to detect an odds ratio of 2.0 was computed as 187 case-control pairs (n=374) (calculation performed with the Epitable module in the Epi Info software, version 6.04 (15)).  

Cases were selected from four (Muscat, Dhakhilia, North and South Sharqia) of the 10 health regions in Oman which represent different situations and prevalences of underweight among children aged 0-59 month(s) (respectively 12.8%, 22.7%, 26.8%, and 19.7% in 1998) (6).

Inclusion of case-control pairs  

In each commune, the Community Support Group members screened all children aged less than three years for cases and suitable controls using the ref­erence growth charts and individual screening forms, including relevant inclusion/exclusion in­formation. After verification at the Department of Nutrition of the Ministry of Health in Muscat, the case-control pairs were included based on age, vali­dation of weight-for-age status (by calculation by computer), absence of any visible congenital dis­eases that might affect growth, and accuracy of the matching criteria, as reported by the field teams. All participant mothers gave their free and informed verbal consents.

Variables  

The questionnaire featured items relating to the child himself/herself, his/her mother, father, and the household that are usually assumed to be po­tentially linked to malnutrition of the child. More­over, consanguinity of parents was also considered as it is frequent in Oman (11) and could have an effect on the health of the child.  

Anthropometric measurements were made under standardized conditions according to the recom­mendations of WHO (7). Weight of the child was measured with Uniscale weighing scales (with 100 g precision) (Seca, Birmingham, UK). Calculation of age was based on the date of birth of the child verified on an official health document (which was also used for collection of birth-weight data). Using the Epinut module of the Epi Info software (version 6.04) (15), weight-for-age was computed for each child and expressed in z-scores(WAZ) of the inter­national reference values (16). Underweight was defined as WAZ <-2 z-scores (7). Body mass index (BMI=weight/height² ) was used for assessing the nutritional status of the mother, <18.5 kg/m² and ≥25 kg/m² defining underweight and overweight respectively (7).  

Haemoglobin (Hb) concentration for the child and the mother was measured (HemoCue AB, Angel­holm, Sweden), and Hb status was described us­ing 11 and 12 g/dL cut-off for the child and for the mother respectively (17).  

The current breastfeeding status of the child was considered adequate if breastfed and age was less than 24 months or not breastfed and age was ≥24 months, inadequate otherwise. Complementary feeding was considered timely if age of the child at start was 6-8 months. A child diversity index was computed from 24-hour recall data using four food groups: cereals, vegetables/fruits, dairy products, and proteins.  

A household water-quality index was derived from source of water and method of delivery and storage of water: good for water from pipes directly from well or traditional rural irrigation system ‘falaj’, av­erage if water from wells, falaj, or desalination plant and delivered by vehicle, and ‘bad’ otherwise. A household economic level proxy was computed by correspondence analysis from type of house, owner­ship of house, presence of servants, number of per­sons per room, presence of electricity, refrigerator, computer, and Internet, and number of cars. The first principal component displaying a gradient of household ‘wealth’ was used as a summary index of household wealth. Households were grouped in ‘low’, ‘medium’, and ‘high’ levels according to terciles of this index (18). Based on the Ministry of National Economy methodology and a minimum necessary income of 30 Omani rials per person per month (19), household income ‘poverty gap’ was computed as total declared monthly income - (30 × number of household members). A household dietary diversity index (20) was computed based on usual consumption at least once a day of five food groups: cereals, vegetables, fruits, dairy products, and proteins. After the standardization of measure­ments and implementation of the questionnaire, actual data collection took place simultaneously in the four regions from May to August 2002.

Data management and statistical analysis

Data-entry tools, including quality-checks, were developed using the data-entry module of the Epi Info software (version 6.04) (15,21). Data cleaning, management, and computation of derived variables and scores were performed with the Epi Info soft­ware and the SAS software (version 8.2) (22).  

Statistical analyses were performed to take into ac­count the 1:1 case-control design with individual matching on sex, age, and village of residence. To assess the differences between cases and controls, univariate conditional logistic regressions (13,14) with underweight (yes/no) as the response variable were used, the qualitative explanatory variables be­ing considered one at a time. These univariate re­gressions were fitted to assess ‘unadjusted’ effects by likelihood ratio tests and to obtain odds ratios (ORs) using a maximum available sample size for each variable. To account for confounders and/or intermediate factors, variables significantly related to underweight at =0.10 level and/or deemed of importance in the Omani context were retained for inclusion in a multivariate model. This latter analy­sis enabled us to assess the ‘adjusted’ effects (like­lihood ratio tests) and to obtain ORs adjusted for all variables included in the model, based on the pairs for which values of all the retained variables were available (‘complete case analysis’ subsample). The conditional logistic regression models were fit­ted using PROC PHREG in the SAS software (ver­sion 8.2) (23). Unless stated otherwise, confidence intervals (CIs) were computed for 0.95 confidence level.

Results

Screening of all children in 53 villages resulted in 190 pairs (47, 44, 47, and 52 for the Muscat, Dhal­hilia, North Sharqia, and South Sharqia regions res-pectively). For cases and controls, respectively, the mean WAZ was -2.42 (0.42) z-score and -1.01 (0.69) z-score, and the mean age was 24.1 (7.9) months and 23.8 (8.0) months (paired differences 0.95 CI -0.06–0.71). 50.5% of the children were boys.

Univariate analysis of associations between underweight and potential risk factors

The results of univariate analyses showed that un­derweight was linked to variables from the differ­ent groups: prenatal factors, current health of the child, child-feeding practices, and socioeconomic status, but not with factors describing the health status of the mother (percentages are given in Table 1a, b, c relevant ORs and CIs quoted in the text).

Health status of mothers and prenatal factors

The children of the shortest (respectively interme­diate-sized) mothers were more at risk (OR=2.6, CI 1.5-4.4) (respectively OR=2.3, CI 1.4-3.8) of under­weight than children of the tallest mothers. Chil­dren with low birth-weight (<2,500 g) showed an increased risk of underweight (OR=5.8, CI 2.2-15.0) when compared with normal children. Children born from mothers having received iron supple­mentation during pregnancy were more frequent among cases vs control. No risk of underweight was associated with age of mothers at birth, time elapsed since previous child, consanguinity of par­ents, number of pregnancies, mother pregnant or not, and BMI and Hb status of mothers.

Child health and feeding practices  

Children who had diarrhoea (vs not) in the last two weeks were marginally more at risk of underweight (OR=1.6, CI 0.9-3.0). Children whose health was regularly monitored were more numerous among cases than controls. No risk of underweight was as­sociated with a recent episode of acute respiratory infection or with the current Hb status. 

A child eating from a separate plate (vs not) was more at risk of underweight (OR=1.7, CI 1.1-2.8). If the child was ever given formula regularly (vs not) marginally increased the risk of underweight (OR=1.5, CI 0.9-2.3). Underweight was not linked to adequate breastfeeding, age when mother usu­ally stops breastfeeding, the fact that the child’s complementary feeding was timely, and the dietary diversity.

Socioeconomic status and environment

Children of mothers with no schooling (OR=2.1, CI 1.1-3.7) or primary school (OR=2.0, CI 1.1-3.9) were more at a risk of underweight compared to those whose mothers had secondary education or more. Children whose mothers were engaged in a professional occupation showed a decreased risk of underweight vs children of non-working mothers (OR=0.3, CI 0.1-0.9). Professional occupation of the father was marginally associated with under­weight. If the father was head of the household (vs not), it had a beneficial impact on the weight status of the child (OR=0.6, CI 0.4-1.1). Underweight was not linked to the marital status of the mother, the schooling level of the father, and if the father lived in the household or not.  

Income of the household and poverty gap (the variable retained for the multivariate model) were linked to underweight. For the latter, the children from the poorer households (OR=3.6, CI 1.6-7.9) and the intermediate ones (OR=2.2, CI 1.2-4.1) had an increased risk of underweight when compared with households in the higher-income category. The quality of water in the household was margin­ally associated with underweight, with a risk of un­derweight increased for children from households with ‘bad’ quality of water vs ‘average’ or ‘good’ (OR=1.9 CI 1.0-3.5). Underweight was not associ­ated with size, economic level, and the dietary di­versity of the household, nor if mother (vs father) decided for purchase of foods.

Multivariate analysis of associations between underweight and potential risk factors

The final multivariate analysis was performed on the ‘complete case analysis’ of the subsample (n=354). For comparison purposes (Table 2), unad­justed effects and ORs were also computed for each factor on this (n=354) subsample to assess that the differences stemmed only from the adjustment process and not from some sort of selection bias linked to missing values.

Although their distribution was markedly differ­ent between cases and controls, the variables—iron supplementation to mother and child health regu­larly monitored—were not included in the final multivariate model as they were more likely to be a consequence of an elevated risk of underweight (directly or not) than a potential cause.

After adjustment for one another in the multi­variate model, most variables were still associated with underweight, except father being head of household and income of household expressed as ‘poverty gap’. For some variables, such as low birth-weight vs not, diarrhoea in the last two weeks vs none, or household water-quality index, the as­sociation with underweight was even stronger af­ter adjustment. We specifically studied the quite drastic effect of the multivariate adjustment (un­adjusted OR=5.8 vs adjusted OR=8.6) for the low birth-weight factor by running a series of analyses for which the adjustment was performed one vari­able at a time (data not shown). No single variable was found to explain by itself a sizeable part of this observed inverse confusion effect.

Discussion

The present study specifically considered a number of potential risk factors of underweight from the underlying and immediate categories and mater­nal/prenatal factors reported to be determinants of underweight of children in many developing coun­tries (24,25). Among the underlying determinants, accessibility of health services was considered very good in the Omani context (12) and, thus, was not addressed.  

The study had several limitations. The effect of age and sex on underweight could not be assessed due to the individual matching on these characteristics, but are well-documented in other contexts (26-28). Moreover, in 1995 and 1998, anthropometric status of Omani boys and girls was similar (6). The match­ing also suppressed the estimation of the effect of village of residence. Also a characteristic of the de­sign is the choice of the underweight criteria index which mixes linear growth retardation (stunting defined as height-for-age index <-2 z-score) and/or wasting (weight-for-height index <-2 z-score). In our sample, 42% of the cases were stunted and 44% wasted, vs 10% and 6% respectively for the controls. These two forms of malnutrition usually have specific causes (29,30).

Health of mother and prenatal factors  

From our data, the factor that was more strongly associated with underweight was low birth-weight; as its effect was adjusted for all other covariates, in­cluding height of mother, it could be interpreted as an intermediate factor for other prenatal factors pertaining to maternal nutritional status. The ef­fect on child nutritional status of the physical ma­ternal and child prenatal characteristics, mediated through birth-weight, is well-documented (31) and was also recently reported in Oman (32). Mater­nal nutrition during pregnancy requires major at­tention, and iron supplementation for pregnant women, thought to improve birth outcome (33), was introduced in Oman in 1986; however, little is known about its compliance and impact. The ob­served higher prevalence of iron supplementation among mothers of cases was deemed more likely to be a consequence of an elevated risk of under­weight (directly or not) than a potential cause. In our study, height of mother was also associated with underweight, independently of other covariates, in­cluding birth-weight and socioeconomic variables. Apart from the effect of possible genetic factors, this association with underweight may be related to the interpretation of adult height as a marker of social class distinctions (34) not entirely taken into account by the socioeconomic variables.  

In fact, maternal height and low birth-weight are central in the intergenerational cycle of malnutri­tion (7) and can only be acted upon in the long term. The prevalence of low birth-weight in Oman remained around 8% from 1985 to 2002 (4), and in 1992, the average height of mothers was 155 cm (11). Height of mother was already reported as a possible cause of malnutrition through the inter­generational cycle when the prevalence of under­weight was higher than expected from the development level (35).

Environment, care, and health of children  

In our study, the quality of household water was strongly associated with underweight as observed in other developing countries, e.g. Brazil (36). His­torically, management of water resources is impor­tant in Oman where rain is scarce in most parts of the country: many villages or oasis still depend on the sophisticated ‘falaj’ underground irrigation sys­tem which dates back a thousand years. Also, con­trary to nearly every other country, the coverage of water supply (37% in 1990 and 39% in 2000) in Oman is lower than the sanitation coverage (84% and 92% respectively) (37).  

Diarrhoea was associated with underweight. How­ever, it could be assumed that having diarrhoea could not only be a cause but also a consequence of being underweight. The occurrence of diarrhoea could be due to the use of bad-quality water, for example when using infant formula; however, because of adjustment for the quality of water in our study, the occurrence of diarrhoea could more likely be a marker of the ‘care for mothers and children’ category of underlying causes (38). Also in this category, feeding practices, such as regular use of infant formula or child-eating from a sepa­rate plate, were linked to underweight. This is in line with the findings that, in three of the regions sampled by our study, the variety, quality, and ad­equacy of complementary foods given to children are affected by lack of nutritional awareness among mothers in Oman (39).

Socioeconomic environment

Schooling level of the mother and whether or not she was engaged in a professional occupation, that we found marginally associated to underweight, could partly act through prenatal factors as shown in another context in Congo (29), but may also be understood as a caring factor. In numerous developing countries, schooling of the mother and eco­nomic level of the household or income remained risk factors for child malnutrition (26,29,36,40,41). In Indonesia, another context of recent impressive public-health gain and decrease of underweight in children, maternal education, and household economic status have continued to be very strong predictors of the nutritional outcomes of children (42). Nevertheless, in our study, after adjustment, the household poverty gap was not associated with underweight. Assuming that income of the household could reflect household food security in Oman, the result underlines that this category of underlying causes is probably no longer relevant in the country. On the contrary, despite a good score on the life expectancy and gross domestic product GDP components of the human develop­ment indicators (1), Oman did not score so well on the education component due to an literacy rate of only 74.4% and a combined gross enrollment ratio for primary, secondary, and tertiary schools of 63%. There is also clearly a gender issue as the lit­eracy rate is 65.4% for females as opposed to 82% for males (1).  

In the Omani context, some determinants from the international conceptual model of the causes of malnutrition are probably no longer relevant, such as household food security or accessibility of health services. Some others are still pertinent, such as supply of safe water at the household level, and several factors relating to care for mothers and chil­dren, including education of mothers, prenatal nu­tritional status, and complementary child-feeding practices. Some of these factors, such as maternal nutrition during pregnancy, could be acted upon in the relatively short term. For some others, the solution is rather in the mid-term range, such as supply of safe water, education of mother, or even a distant future, such as height of mothers, which will at best improve only after several iterations of the intergenerational cycle. Further research is also needed to investigate further specific determinants of low birth-weight in the Omani context and try to disentangle emaciation and determinants of lin­ear growth retardation. The research should not underestimate the importance of sociocultural fac­tors, such as women’s education and status.

Aknowledgements

References

1. United Nations Development Programme. Human development report 2005; international cooperation at a crossroads: aid, trade and security in an unequal world. New York, NY: United Nations Development Programme, 2005. 372 p.
2. Oman. Ministry of Health. Communicable diseases control in the Sultanate of Oman. Muscat: Ministry of Health, 2004. (ttp://www.moh.gov.om/comm.htm, accessed on 3 November 2005).
3. Bellamy C. The state of the world’s children 2003. New York, NY: United Nations Children’s Fund, 2003.123 p.
4. Oman. Ministry of Health. Health indicators related to maternal health care services (ante- and post-natal care). Muscat: Ministry of Health, 2004. (http://www.moh.gov.om/mohstat.php, accessed on 3 November 2005).
5. Amine EK. Oman nutrition status survey. Abu Dhabi: United Nations Children’s Fund, 1980.
6. World Health Organization. WHO global database on child growth and malnutrition. Geneva: World Health Organization, 2005. (http://www.who.int/nut­growthdb, accessed on 3 November 2005).
7. World Health Organization. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. Geneva: World Health Or­ganization, 1995. 462 p. (Technical report series no. 854).
8. Engle P. The care initiative: assessment analysis and action to improve care for nutrition. New York, NY: United Nations Children’s Fund, 1997. 72 p.
9. Smith LC, Haddad L. Explaining child malnutrition in developing countries: a cross-country analysis. Washington, DC: International Food Policy Research Institute, 2000. 112 p.
10. Alasfoor D, Rawas S, Al-Farsi Y, Alshishtawi M. Na­tional study on the role of care in the nutritional sta­tus of children under 2 years old in Oman. Muscat: Ministry of Health, 2000. 30 p.
11. Musaiger AO. Study of health and nutritional status of Omani families. Muscat: United Nations Children’s Fund, 1992.
12. Suleiman AJM, Al-Ryiami A, Farid SM. Oman familiy health survey 1995. J Trop Peditr 2001;47(Suppl 1): 1-33.
13. Rothman KJ, Greenland S. Modern epidemiology. 2d ed. Philadelphia, PA: Lippincott, Williams & Wilkins, 1998. 737 p.
14. Breslow NE, Day NE. Statistical methods in cancer research. V. 1. The analysis of case-control studies. Lyon: International Agency for Research on Cancer, 1993. 350 p.
15. Dean AG, Dean JA, Coulombier D, Brendel KA, Smith DC, Burton AH et al. Epi Info, version 6: a word pro­cessing, database, and statistics program for epidemiology on microcomputers. Atlanta, GA: Centers for Disease Control and Prevention, 1995.
16. World Health Organization. Measuring change in nutritional status: guidelines for assessing the nutri­tional impact of supplementary feeding programme. Geneva: World Health Organization, 1983. 101 p.
17. World Health Organization. Iron deficiency anaemia: assessment prevention and control. A guide for pro­gramme managers. Geneva: World Health Organiza­tion, 2001. 114 p.
18. Delpeuch F, Cornu A, Massamba JP, Traissac P, Maire B. Is body mass index sensitively related to socio-eco­nomic status and to economic adjustment? A case study from the Congo. Eur J Clin Nutr 1994;48(Suppl 3):S141-7.
19. Oman. Ministry of National Economy. [Reports on the results of the household expenditure and income survey, v. 2]. Muscat: Ministry of National Economy, 2000. [Arabic].
20. Hoddinott J, Yohannes Y. Dietary diversity as a household food security indicator. Washington, DC: Food and Nutrition Technical Assistance, Academy for Educational Development, 2002. 4 p. (Technical note no. 4).
21. Bennett S, Myatt M, Jolley D, Radalowicz A. Data management for surveys and trials. A practical prim­er using Epi Info. Llanidloes, Powys: Brixton Books, 1996. 98 p.
22. SAS Institute Inc. The SAS system, version 8. Cary, NC: SAS Institute Inc., 1999.
23. Stokes ME, Davis CS, Koch GG. Categorical data analysis using the SAS system. Cary, NC: SAS Institute Inc., 2000. 626 p.
24. Mozumder AB, Barkat EK, Kane TT, Levin A, Ahmed S. The effect of birth interval on malnutrition in Bangladeshi infants and young children. J Biosoc Sci 2000;32:289-300.
25. Orozco M, Martinez H, Reyes H, Guiscafre H. A scale without anthropometric measurements can be used to identify low weight-for-age in children less than five years old. J Nutr 1998;128:2363-8.
26. Mishra VK, Retherford RD. Women’s education can improve child nutrition in India. Natl Fam Health Surv Bull 2000;15:1-4.
27. Ray SK, Haldar A, Biswas B, Misra R, Kumar S. Epidemiology of undernutrition. Indian J Pediatr 2001;68:1025-30.
28. Bloss E, Wainaina F, Bailey RC. Prevalence and pre­dictors of underweight, stunting, and wasting among children aged 5 and under in western Kenya. J Trop Pediatr 2004;50:260-70.
29. Delpeuch F, Traissac P, Martin-Prevel Y, Massamba JP, Maire B. Economic crisis and malnutrition: socio­economic determinants of anthropometric status of preschool children and their mothers in an African urban area. Public Health Nutr 2000;3:39-47.
30. Frongillo EA, Jr., Hanson KM. Determinants of vari­ability among nations in child growth. Ann Hum Biol 1995;22:395-411.
31. Barker DJP. Mothers, babies, and health in later life. 2d ed. New York, NY: Livingstone, 1998. 217 p.
32. Kurup PJ, Khandekar R. Low birth weight as a deter­minant of protein energy malnutrition in “0-5 years” Omani children of South Batinah region, Oman. Saudi Med J 2004;25:1091-6.
33. Ramakrishnan U. Nutrition and low birth weight: from research to practice. Am J Clin Nutr 2004;79:17-21.
34. Tanner J. Growth as a mirror of the condition of so­ciety: secular trends and class distinctions. In: Demir­jian A, Brault Dubuc M, editors. Human growth: a multidisciplinary review. London: Taylor & Francis, 1986:3-34.
35. Mason JB. Update on the nutrition situation. SCN News 1989;4:1-4.
36. Marins VM, Almeida RM. Undernutrition prevalence and social determinants in children aged 0-59 months, Niteroi, Brazil. Ann Hum Biol 2002;29:609-18.
37. World Health Organization. Global water supply and sanitation assessment 2000 report. Geneva: World Health Organization, 2000. 80 p.
38. Martin-Prevel Y. [“Care” and public nutrition]. Sante 2002:12:86-93.
39. United Nations Children’s Fund. Summary of mid-term reviews and major evaluations of country pro­grammes. Middle East and North Africa region. New York, NY: United Nations Children’s Fund, 2003. 15 p. (http://www.unicef.org/about/execboard/files/0-PL28.pdf, accessed on 3 November 2005).
40. Fletcher PD, Simeon DT, Grantham-McGregor SM. Risk indicators of childhood undernutrition in Kings­ton, Jamaica. Trans R Soc Trop Med Hyg 1992;86:566-9.
41. de Souza AC, Peterson KE, Cufino E, Gardner J, Craveiro MV, Ascherio A. Relationship between health services, socioeconomic variables and inadequate weight gain among Brazilian children. Bull World Health Organ 1999;77:895-905.
42. Waters H, Saadah F, Surbakti S, Heywood P. Weight-for-age malnutrition in Indonesian children, 1992-1999. Int J Epidemiol 2004;33:589-95.

© 2007 ICDDR,B: Centre for Health and Population Research

The following images related to this document are available:

Photo images