Social Science & Medicine 73 (2011) 576e585

Contents lists available at ScienceDirect

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

Trends in malnutrition among children in India: Growing inequalities across different economic groups
Praveen Kumar Pathak a, *, Abhishek Singh b
a b

Department of Geography, Shivaji University, Kolhapur, Maharashtra 416004, India Department of Public Health & Mortality Studies, International Institute for Population Sciences, Govandi Station Road, Deonar, Mumbai 400 088, Maharashtra, India

a r t i c l e i n f o
Article history: Available online 6 July 2011 Keywords: Malnutrition Economic inequality Geographical region Principal component analysis Poorerich ratio Concentration index Children India

a b s t r a c t
This paper examines the trends and patterns of economic inequalities with respect to child malnutrition by wealth status of population across major regions and states of India. Data from three rounds of National Family and Health Survey (NFHS) conducted during 1992e2006 were analyzed. The proportion of underweight children (measure of both acute and chronic malnutrition) has been used as a dependent variable. The wealth index is used as proxy for economic status of the population, and was estimated through principal component analysis by employing a set of variables representing durable asset ownership, access to utilities and infrastructure, and housing characteristics of respondents for all the three survey rounds. Bivariate analyses, poor-rich ratio and concentration indices were used to understand the trends in economic inequalities with respect to child malnutrition. Pooled logistic regression models were tted to estimate the adjusted effect of economic status on the likelihood of child malnutrition over time. Results indicate sluggish change coupled with concomitant rise in economic inequalities with respect to child malnutrition in India during 1992e2006. The burden of malnutrition was disproportionately concentrated among poor children. In addition, average decline in malnutrition concealed large economic disparities across space and time. 2011 Elsevier Ltd. All rights reserved.

Introduction Reduction of child malnutrition is one of the prime challenges that India faces. The burden of malnourished children in India is amongst the highest in the world and virtually twice that of SubSaharan African countries. Nearly 60 million Indian children are estimated to be underweight; more than 50 percent suffer from anemia and a similar proportion lacks full immunization (Deaton & Dreze, 2009; FOCUS, 2006; Gragnolati, Shekar, Dasgupta, Bredenkamp, & Lee, 2005). India ranked 96 out of 119 countries in the Global Hunger Index (GHI) developed by the International Food Policy Research Institute (IFPRI) in 2006, and where child malnutrition is concerned, it ranked 117 among 119 countries (Braun, Ruel, & Gulati, 2008). That malnutrition undermines economic growth and perpetuates the vicious circle of poverty has been well established. This occurs due to multiphasic loss, that is, direct loss to productivity because of weak physical status, indirect loss from fragile cognitive development, and loss incurred due to increased health care costs
* Corresponding author. Tel.: 91 8007389928 E-mail addresses: pkp_pathak@rediffmail.com (P.K. Pathak), abhi_iips@yahoo. co.in (A. Singh). 0277-9536/$ e see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2011.06.024

(World Bank, 2006). Generally, there is an inverse relationship between hunger and increasing economic growth. Studies have shown that during the past decade, after the New Economic Policy was introduced in the 1990s, India achieved unprecedented economic growth and has made noteworthy advances in the elds of science, agriculture, medicine, information technology (CSNSI, 2008). Despite these improvements, progress toward reducing the proportion of undernourished children in India has been sluggish (Pathak & Singh, 2009; Svedberg, 2006). The economy grew steadily at an average of 6 percent (Basu & Maertens, 2007) and the per capita income doubled during the post reform period. However, although the Global Hunger Index (GHI) declined between 1981 and 1997, it remained stagnant between 1997 and 2003. This phenomenal stagnation in GHI might be attributed to rising income inequalities and spatial disparities in the country during the same period (Braun et al., 2008). The proportion living in poverty has declined only marginally from 36% in 1992e1993 to 27.5% in 2004e2005 (Economic Survey, 2001e2002; Planning Commission, 2007), and even this decline has been uneven. Recent studies (Ahluwalia, 2002; Ghosh & Chandrasekhar, 2003; Pal & Ghosh, 2007; Sen & Himanshu, 2005) have shown that economic inequalities and regional disparities have grown during the post reform period. During the same period,

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

577

(1992e2006) while the proportion in poverty declined by 24%, the relative prevalence of underweight children, on average, declined more gradually by 12% (Economic Survey, 2001e2002; Planning Commission, 2007). The issue of socioeconomic inequalities in health outcomes in developing countries like India has evoked considerable interest among researchers and social scientists (Carr, 2004; Gwatkin, Bhuiya, & Victora, 2004; Houweling, Kunst, & Mackenbach, 2003; Lawnet, Tinker, Mridha, & Anwar, 2006; Mohanty & Pathak, 2009; Pathak, Singh, & Subramanian, 2010; Poel, Hosseinpoor, Speybroeck, Ourti, & Vega, 2008; Subramanian, Kawachi, & Smith, 2007; Subramanian, Perkins, & Khan, 2009; Subramanian & Smith, 2006; Subramanian, Smith, & Subramanyam, 2006; Zere & McIntyre, 2003; Wagstaff & Watanabe, 2000, Wagstaff, 2002a, 2002b). Studies have shown that the pervasiveness of socioeconomic inequalities in health both between and within countries at any stage of development signicantly retards progress toward the achievement of the Millennium Development Goals (MDGs) such as maternal and child health, universal education, gender equity and combating HIV/AIDS, and especially efforts to meet the rst MDG of halving poverty and hunger. Recent evidence suggests that malnutrition leads to the likelihood of developing chronic diseases, and hence high child morbidity and mortality; it also reduces long term physical development, cognitive skills, and consequently has a negative effect on school enrolment and productivity in later life (GranthamMcGregor et al., 2007; Tarozzi & Mahajan, 2007; Walker et al., 2007; World Bank, 2006). Reduction of malnutrition is thus not only an economic issue but also welfare, social protection and human rights issue. Data shows that undernutrition remains a leading problem in most parts of India, but it is most pronounced in the states of Bihar, Uttar Pradesh, Madhya Pradesh and Rajasthan. In these states, more than half the children are underweight and stunted. Nearly 50% of the children in Orissa, Maharashtra and West Bengal are underweight, while 50% of the children in Assam and Haryana are stunted. States with the lowest percentage of underweight or stunted children are Goa, Kerala, and all the small north-eastern states except Tripura and Meghalaya (Arnold, Nangia, & Kapila, 2004; Mishra, Roy, & Retherford, 2004; Nair, 2007). A multicountry study using Demographic and Health Survey (DHS) data from more than fty developing countries found that the poorest quintile fares worse than better-off groups on nutritional status. On average, stunting is three times more likely among children in the poorest quintile than among those in the wealthiest quintile (Gwatkin et al., 2007). Another study that examined socioeconomic inequality in child nutrition among twenty developing countries found that eighteen countries had statistically signicant inequalities in both stunted and underweight children (Wagstaff & Wanatabe, 2000). Previous studies (Gragnolati et al., 2005; Pande & Yazbeck, 2003) have found signicant differentials in child health and health care utilization by wealth status, gender and geographical regions in India. While efforts to reduce socioeconomic disparities in health outcomes are not lacking, they are rather ineffective. The Public Distribution System (PDS), (one of the worlds largest food security programs for poor households) and Integrated Child Development Services (ICDS) scheme, (the worlds largest early child development program) have been instituted for the last three decades in the country. However, these initiatives have performed far below expectations (Gragnolati et al., 2005). Recently, the government has launched several programs, such as the Mid-Day Meals Program, National Rural Health Mission (NRHM, 2005e2012), National Food Security Mission, National Rural Employment Guarantee Act and others, to attain inclusive growth

with social justice. These programs aim at improving the economic, health and nutritional status of the population, especially that of marginalized poor women and children from rural areas. However, there is hardly any study that documents the progress and achievements of these programs in relation to the agenda of inclusive growth in the country. The studies mentioned earlier either did not include India in their analysis or did not analyze data on child malnutrition (Gwatkin et al. 2007; Wagstaff & Wanatabe, 2000). Though Gragnolati et al. (2005) did attempt to analyze economic inequalities with respect to various indicators of maternal and child health in India, they did not analyze the nutritional aspect in detail. Further, they used data from NFHS I and NFHS II to understand the trends in child malnutrition. Therefore, we nd that there is a dearth of studies that examine the recent trends and patterns of economic inequalities with respect to child malnutrition in India. Given the high prevalence of child malnutrition in India, it is important to understand the economic gradient in child malnutrition and the changes in the same during the period 1992e2006. Economic reforms were launched in India during the early nineties. The present analysis examines the changes in the economic gradient in child malnutrition during the period of economic reforms. Further, there is an immediate need for such studies as they not only throw light on the magnitude and complexity of the problem of child malnutrition across space and time, but also assist policy makers and program managers to focus targeted interventions on those who are in real need. In this study, we use three rounds of NFHS data to examine the trends and patterns of economic inequalities with respect to child malnutrition in major geographic regions and states of India. Mackenbach and Kunst (1997) have suggested a framework for measuring the magnitude of socioeconomic inequalities in health, which was developed in the context of the efforts of the World Health Organization (WHO) European region to monitor changes over time. Within this framework, we describe the variations in child malnutrition and then summarize the observed variations into single measures to facilitate comparisons over time and between populations. As suggested by Mackenbach and Kunst (1997), we compute the measures of total impact such as the ratio of child malnutrition among those in the poorest quintile to those in the richest quintile and the concentration index. One drawback with the measures of total impact is that they only take into account inequalities between the two economic groups that are compared and ignore the full range of malnutrition differences. We examine the changes in economic inequalities with respect to child malnutrition during 1992e2006 after adjusting for important socioeconomic and demographic factors affecting child malnutrition. Further, given the strong interest in reduction of average rates of malnutrition in national and state development goals and targets, we test the association between average level of malnutrition rates and economic related inequality in malnutrition to decipher how various geographical regions and states of India compare among themselves in average malnutrition rates and economic related inequalities. Data and methods Data The present study uses data from the three rounds of the National Family Health Survey (NFHS) conducted during 1992e1993 (NFHS-1), 1998e1999 (NFHS-2) and 2005e2006 (NFHS-3) respectively. NFHS is a nationally representative, large scale, repeated cross sectional survey in representative samples of households throughout India. The principal objective of NFHS is to

578

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

provide state and national level estimates on fertility, mortality, family planning, HIV-related knowledge, and on important aspects of nutrition, health and health care. The survey provides state and national level estimates of demographic and health parameters as well as data on various socioeconomic and program dimensions, which are critical for implementing the desired changes in demographic and health parameters. Ethical approval was provided by the International Institute for Population Sciences, Mumbai, India. The survey adopted a two-stage sample design in most rural areas and a three-stage sample design in most urban areas. In rural areas, the villages were selected at the rst stage using a Probability Proportional to Size (PPS) sampling scheme. The required number of households was selected at the second stage using systematic sampling. In urban areas, blocks were selected at the rst stage, census enumeration blocks (CEB) containing approximately 150e200 households were selected at the second stage, and the required number of households were selected at the third stage using systematic sampling technique (for details regarding sampling, see IIPS & ORC Macro, 2007). A similar sampling scheme was adopted in all the three rounds of NFHS. More than 90,000 households were interviewed in each round of the NFHS. So, the different rounds of NFHS provide sufciently large sample sizes to carry out analysis at the national as well as the state level. The data were collected using similar interview schedules in the three rounds of the NFHS. The household and eligible female informant response rates were consistently above 90% in all three NFHS rounds. There were only small variations in the household and eligible informant response rates across different states of the country. To make the estimates representative and comparable across the three survey rounds, and to account for the multi-stage sampling design adopted in the three rounds of NFHS, we used appropriate weights in the analysis. The details of the sampling weights are given in NFHS reports of the various rounds (IIPS, 1995; IIPS & ORC Macro, 2000, 2007). Measures By virtue of similar sampling design, the estimates made in the three survey rounds are comparable (Mishra et al., 2004; Ram & Roy, 2004). Information on anthropometric indicators for children below three years of age has been used uniformly in the analyses to measure the nutritional status of children in all three rounds of NFHS. NFHS provides information about weight-for-age, weightfor-height (not available in a few states in NFHS-1) and height-forage (not available in a few states in NFHS-1) for all three survey rounds. But the height of children could not be measured in the states of Andhra Pradesh, West Bengal, Himachal Pradesh, Madhya Pradesh, and Tamil Nadu due to unavailability of measuring instruments during the rst phase of NFHS-1. Therefore, any allIndia estimate using height of children may be biased. In order to generate robust all-India estimates over time and space, we used weight-for-age, underweight as our preferred indicator for measurement of the nutritional status of children. This approach has been backed by studies that argue that weight-for-age, underweight deserves special attention as it is a comprehensive indicator of child nutritional status, incorporating both stunting and wasting (Deaton & Dreze, 2009). Underweight is a measure of protein-energy undernutrition. It is used to describe children who have a weight-for-age measurement less than two standard deviations below the median value of United States National Centre for Health Statistics (US NCHS) international reference population as recommended by the World Health Organization (Dibley, Goldsby, Staehling, & Trowbridge, 1987; Dibley, Staehling, Neiburg, & Trowbridge, 1987). We estimated underweight children based on US NCHS reference population uniformly

for all three rounds of NFHS due to the unavailability of the new WHO reference population (WHO Multicenter Growth Reference Study Group, 2006) in NFHS-1 and NFHS-2 datasets. Children (below three years of age) whose weight-for-age is below minus two standard deviations from the median reference population are classied as underweight or malnourished. These terms have been used interchangeably in the paper. We also note that, on an average, the US NCHS standards overestimate the prevalence of undernutrition in India by three to ve percentage points during 1992e2006 as compared to the new WHO standards of 2006. The analytical sample size of the present study is restricted to 27875, 24989 and 24960 children below 35 months of age after excluding missing and agged cases, which comprise around 20%, 24% and 15% of the total eligible sample respectively in NFHS-1 (1992e1993), NFHS-2 and NFHS-3 (2005e2006). In the absence of direct data on income or expenditure in household sample surveys like NFHS, the wealth index was based on the ownership of household assets is widely used as a proxy for assessing the economic status of the households (Filmer & Pritchett, 2001; Gwatkin et al., 2007; Howe, Hargreaves, Gabrysch, & Huttly, 2009; Johnson & Bradley, 2008; Montgomery, Gragnolati, Burke, & Paredes, 2000; Rutstein, 2008; Rutstein & Johnson, 2004; Vyas & Kumaranayake, 2006). Moreover, studies have noted that the wealth index is an indicator of the level of wealth that is consistent with expenditure and income measure, and widely tested in a large number of developing countries to examine economic inequalities in household income, including India (IIPS & Macro International, 2007; Rutstein, 1999; Rutstein et al., 2000). In the present study, the wealth index is estimated by principal component analysis using a set of durable asset ownership, access to utilities and infrastructure, and housing characteristics variables available for all three rounds of NFHS (see Appendix 1). This has been done to make the wealth index comparable over the years. The categories of the wealth index are constructed in such a way that each category consists of 20% of the surveyed population. The wealth index and its ve categories are already given in the NFHS-3 dataset. However, we constructed the wealth index and categorized them into ve groups for all the three rounds of NFHS using the exact methodology followed in NFHS-3. The wealth quintiles were determined based on all-India sample data and checked for reliability and consistency. These wealth quintiles were found to capture the economic differentials in the outcome variables in each round of NFHS. The states of India are at different stages of socioeconomic and demographic development. The different geographic regions, comprising a group of states, are also found to vary signicantly on various indicators of social and economic development; southern India is much more economically and demographically developed than northern India (Bhat & Zavier, 1999; Bose, 1991; Dyson & Moore, 1983; Kurian, 2000). So, a country level analysis that does not take into consideration regional variations may not represent the trends in child malnutrition. Therefore, we present the initial descriptive results separately for the six geographic regions to decipher the socioeconomic and demographic variations across the different regions of India. In order to facilitate regional level analyses, India is divided into six broad geographical regions namely North, Central, East, North-East, West and South (see Table 2 for location of states by region). There is a growing body of research that shows that besides wealth, various other socio-demographic variables have a signicant impact on child malnutrition. So, we adjusted for some important socio-demographic variables in our analysis. The sociodemographic variables used in the regression model are state of residence, maternal and paternal education, type of area of residence (in terms of urban-rural categories), age of child, sex of child,

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

579

birth order, preceding birth interval, mothers age at birth of the index child, breastfeeding status, social group (measured in terms of whether the mother belongs to scheduled castes/tribes, other backward castes, or other castes), and religion (measured in terms of Hindu, Muslim and others). Statistical procedures Cross-tabulations, poor-rich ratio (poorest quintile/richest wealth quintile) and the concentration index (CI) are used to understand the trends with respect to economic inequalities in child malnutrition. Bivariate analyses are carried out to understand the differentials in child malnutrition by wealth quintiles across major geographic regions and states in India during the period, 1992e2006. One-way Analysis of Variance (ANOVA) was performed to check whether the differences in child malnutrition by wealth (at the all-India level) and region of residence were significant in the bivariate analysis. In the second step, again one-way ANOVA was used to test for signicant differences in child malnutrition by wealth quintile in each of the six geographic regions. The poorerich ratio, dened as the ratio between the percentage of underweight children among the poorest and the percentage of underweight children among the richest wealth quintile, is used to summarize the economic inequalities in the distribution of child malnutrition. If the poorerich ratio has a value of 1, it indicates that the poorest and richest experience malnutrition equally; and if the poorerich ratio is greater than 1, the poorest are more likely to suffer from malnutrition. The concentration curves and concentration indices are used to measure the overall inequalities in malnutrition among the wealth quintiles (Kakwani et al., 1997;). A concentration index for child malnutrition results from a concentration curve. This curve plots the cumulative proportion of children, ranked by socioeconomic status x, against the cumulative proportion of underweight children y. If all children, irrespective of their economic status x, have the same y, the concentration curve would coincide with the diagonal (line of equality). The concentration curve lies above the diagonal if y is larger among the poorer children and vice versa. The greater the distance of the curve from the diagonal, the higher

the economic inequality. A concentration index is a measure of this inequality and is dened as twice the area between the concentration curve and the diagonal, and it varies between 1 and 1 (Wagstaff & Doorslaer, 2004). The convention is that if the concentration curve lies above the equality line (indicating disproportionate distribution of a specic type of disbenet among more deprived groups) it takes values between 0 and 1, so values closer to 1 in the following analysis indicate greater inequality in this sense. We have estimated the concentration index for all the three rounds of NFHS using the aforementioned methodology. As the sampling design of the three rounds of NFHS (Mishra et al., 2004; Ram & Roy, 2004), is comparable, many earlier studies have pooled the different rounds of DHS/ NFHS datasets to observe changes over time (Kandala, Fahrmeir, Klasen, & Priebe, 2009; Mishra et al., 2004). In this study, three rounds of NFHS datasets have been pooled and four dummy variables, that is, poorer, middle, richer and richest representing the wealth quintiles interact with a set of dummy variables denoting historical periods. In order to measure the economic inequalities in the risk of underweight among children over time, we t a pooled binary logistic regression model while adjusting for the important socioeconomic, demographic variables and region of residence. The results of this analysis have been presented as a set of predicted probabilities of being underweight for individuals grouped into ve categories dened by wealth quintiles (Table 3). The predicted probabilities are based on terms in the logistic regression model relating to interactions between year and wealth status. The Wald test was used to test the signicance of the interaction between wealth quintile and survey year. Maps have been generated using Arc GIS software package (Arc Map/Arc Catalog) (ESRI, 2009). Results Trends in child malnutrition The results indicate that the prevalence of child malnutrition in India, on an average, has declined by six percentage points (from 53% to 47% during 1992e1998) and then stagnated, reaching 46 % in 2005e2006 (only 0.6 % decline during 1998e2005).

Fig. 1. Spatial pattern in prevalence of underweight children (weight-for-age <2 S.D., 0e35 months) across major states, India, 1992e1993, 1998e1999 & 2005e2006.

580

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

Table 1 Percent of children below three years of age who were underweight by economic status of their households in different geographic regions of India, 1992e2006.

Data presented in Table 1 reect the trends and differentials in child malnutrition by economic status of household across the major geographic regions of India during 1992e2006. In spite of the consistent decline in the prevalence of child malnutrition during 1992e2006, the children from the poorest wealth quintile fared worse than their richest counterparts. The differences in the prevalence of child malnutrition by various categories of wealth quintile and region of residence were highly signicant in each of the three rounds of NFHS (as obtained in one-way ANOVA). Overall, the results indicate that while the levels of child malnutrition were much higher among children from the poorest quintile compared to those in the richest quintile, the decline in child malnutrition during 1992e2006 was much lower among

Middle

NFHS-1 (1992e1993)

Poorer

Poorest

Indiab Northc,d Centralc,d Eastc,d Westc,d Southc,d North-East

Region

62.3 49.8 62.2 66.4 65.8 58.7 58.5

58.5 46.5 61.0 61.7 61.4 54.4 50.5

53.5 50.3 59.5 59.5 50.6 47.1 45.2

43.3 41.5 51.6 47.2 48.2 33.4 24.7

Trends and differentials in child malnutrition by economic status

Percentage point change per year calculated as the difference between rst and last data points and divided by the number of years (15) between the two surveys. Differences in child malnutrition by wealth quintile signicant at p < 0.01. c Differences in child malnutrition by region of residence were signicant at p < 0.01 in all the three rounds of NFHS except in one round where the southern region was no different from the northern region in NFHS I. d Richest and Richer quintiles were signicantly different from the poorest quintile in all the six regions of India.
b

Prevalence of underweight by wealth quintile (in %)

Prevalence of child malnutrition varied considerably across the different states of India (Fig. 1). During 1992e1993, majority of the Indian states, except Kerala, were suffering from this syndrome. The major problem states were Chhattisgarh, Bihar, Madhya Pradesh, Uttar Pradesh, West Bengal, Jharkhand, Orissa, Uttaranchal, Maharashtra and Karnataka, where more than 50% of the children were malnourished. Nearly 50%e60% of the children from most of the states in the central, eastern, and western parts of India were malnourished. A modest change in the spatial prole of underweight children in India was observed during 1992e1998 (Fig. 1). Many of the southern, northern, eastern and north-eastern states registered a noticeable decline in child malnutrition during this period; the levels of malnutrition declined from 50%e59 % in 1992e93 to 40%e49% in the states of Karnataka, Uttaranchal and West Bengal in 1998e99. On the other hand, the prevalence of malnutrition declined from 40%e49% in 1992e93 to 30%e39% in 1998e99 in the states of Tamil Nadu, Andhra Pradesh, Jammu and Kashmir, and the north-eastern states. However, the states of Chhattisgarh, Bihar, Madhya Pradesh, Uttar Pradesh, Rajasthan, Orissa and Maharashtra continued to remain the problem states in the country with more than 50% of their children in the underweight category. In fact, the problem of child malnutrition was aggravated in the state of Rajasthan during 1992e1998. During 1998e2006, the spatial mosaic of child malnutrition in India had undergone noticeable change despite stagnation at the national level (Fig. 1). Several states in the western, central, eastern and northern parts of India were able to reduce the burden of underweight children from 50%e59% to 40%e49% during 1998e2006. This includes the states of Uttar Pradesh, Orissa, Rajasthan, and Maharashtra. The prevalence of malnourished children declined from 40%e49% to 30%e39% in the states of Himachal Pradesh and Uttaranchal. However, Madhya Pradesh, Jharkhand, Bihar, and Chhattisgarh had more than 50% of malnourished children. The problem of malnutrition deteriorated further in Madhya Pradesh, Haryana and Kerala during 1998e2006. These results clearly suggest growing interstate differentials with respect to child malnutrition in India. We also nd a substantial and signicant regional divide in the prevalence of child malnutrition. The central and eastern regions had the highest average levels of child malnutrition in the country. Though the northern region showed the smallest reduction in child malnutrition, the central and eastern regions also showed relatively small reductions compared to the other three regions (Table 1). The ndings clearly suggest that the central and eastern regions did not achieve appreciable reduction in child malnutrition and continued to remain the major areas with child malnutrition during 1992e2006. Notably, these are the regions that are relatively poor on economic and demographic parameters (Bhat & Zavier, 1999; Bose, 1991; Dyson & Moore, 1983; Kurian, 2000).

%a point change per year

Average underweight

Prevalence of underweight by wealth quintile (in %)

Richest Richer Middle Poorer Poorest Richest Richer Middle Poorer Poorest Richest Richer

NFHS-3 (2005e2006)

Average underweight

Prevalence of underweight by wealth quintile (in %)

NFHS-2 (1998e1999)

Average underweight

34.0 32.0 43.8 33.2 32.9 28.5 16.5

52.8 43.5 58.3 59.6 50.7 45.6 47.2

60.2 58.6 61.6 62.0 64.8 51.8 39.9

52.9 50.4 57.5 53.3 58.3 46.7 40.5

44.8 46.0 47.6 48.4 47.8 40.1 32.0

36.6 38.5 42.6 35.5 43.4 29.4 22.6

25.4 23.7 33.3 22.9 30.4 18.1 22.3

47.1 42.9 53.5 53.0 48.4 37.7 34.7

58.5 52.3 58.0 61.5 61.6 53.3 46.6

48.4 45.5 51.8 48.6 51.6 43.7 43.0

39.6 38.9 44.8 37.8 43.9 35.9 29.4

32.4 31.7 35.9 32.3 37.7 27.8 16.3

24.8 20.9 30.4 25.0 25.6 21.2 31.8

46.5 39.5 51.4 52.8 43.4 36.8 38.7

0.42 0.27 0.46 0.45 0.49 0.59 0.57

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

581

Table 2 Poor-rich ratio and concentration index depicting trends in economic inequalities with respect to child malnutrition (weight-for-age <2 S.D.) across geographic regions and states, India, 1992e2006. 1992e1993 PR-ratioa India North Jammu and Kashmir Himachal Pradesh Punjab Uttaranchal Haryana Delhi Rajasthan Central Chhattisgarh Madhya Pradesh Uttar Pradesh East Bihar Jharkhand Orissa West Bengal West Goa Gujarat Maharashtra South Andhra Pradesh Karnataka Kerala Tamil Nadu North east
a b c

1998e1999 CIb 0.10 0.08 0.12 0.08 0.11 0.08 0.17 0.10 0.00 0.04 0.07 0.06 0.04 0.06 0.04 0.06 0.08 0.11 0.12 0.20 0.12 0.13 0.14 0.12 0.10 0.16 0.14 0.13 (SE)c 0.003 0.011 0.056 0.041 0.022 0.032 0.027 0.032 0.018 0.005 0.017 0.01 0.006 0.006 0.008 0.024 0.017 0.01 0.009 0.129 0.016 0.011 0.008 0.013 0.014 0.032 0.015 0.017 PR-ratioa 2.4 2.5 2.9 2.9 3.4 5.0 2.4 1.7 1.9 1.9 2.1 2.6 1.5 2.7 2.3 3.3 2.8 2.8 2.1 2.9 2.2 2.0 2.9 2.1 3.2 3.1 4.0 1.8 CIb 0.14 0.15 0.19 0.12 0.21 0.13 0.15 0.13 0.08 0.09 0.06 0.11 0.08 0.10 0.06 0.11 0.11 0.14 0.13 0.18 0.16 0.12 0.16 0.14 0.17 0.14 0.16 0.10 (SE)
c

2005e2006 PR-ratioa 2.4 2.5 3.1 3.3 3.1 3.1 2.2 1.1 2.1 1.9 3.1 1.5 2.1 2.5 2.0 3.9 1.8 3.5 2.4 4.1 1.9 2.8 2.5 2.6 2.1 2.6 3.1 1.5 CIb 0.14 0.15 0.16 0.11 0.22 0.20 0.11 0.15 0.10 0.08 0.10 0.05 0.10 0.11 0.08 0.08 0.13 0.16 0.15 0.09 0.10 0.18 0.15 0.13 0.12 0.16 0.15 0.12 (SE)
c

1.8 1.6 2.4 1.9 1.8 1.7 2.7 2.1 1.1 1.4 1.6 1.4 1.4 2.0 1.5 3.5 2.8 2.9 2.0 3.2 1.9 2.1 2.1 1.8 2.0 2.3 2.1 3.5

0.004 0.011 0.05 0.052 0.035 0.032 0.034 0.041 0.014 0.007 0.023 0.012 0.009 0.007 0.011 0.021 0.016 0.012 0.01 0.172 0.017 0.011 0.01 0.017 0.018 0.034 0.018 0.028

0.003 0.013 0.062 0.068 0.038 0.052 0.029 0.070 0.016 0.007 0.022 0.011 0.008 0.006 0.008 0.015 0.020 0.014 0.012 0.011 0.018 0.017 0.013 0.021 0.023 0.047 0.026 0.023

Poorerich ratio. Concentration index. Standard error. Table 3 Predicted probability of being underweight from logistic regression analysis for different categories of wealth quintile, India, 1992e2006. Interaction effects: wealth status and time 1992e1993 Poorest Poorer Middle Richer Richest 1998e1999 Poorest Poorer Middle Richer Richest 2005e2006 Poorest Poorer Middle Richer Richest Change 1992e2006 Poorest Poorer Middle Richer Richest Predicted probabilitya, 0.617 0.560 0.500 0.389 0.285 0.595 0.499 0.406 0.323 0.211 0.508 0.308 0.289 0.213 0.151 0.109 0.180 0.211 0.176 0.134 95% C.I.
b

children belonging to the poorest quintile compared to those in the richest quintile. The decline in malnutrition ranged between 6% in the poorest and 27% in the richest category of the wealth quintile during 1992e2006 (Fig. 2). The decline in the prevalence of child malnutrition during the period 1992e1998 was relatively faster than during the period 1998e2006, cutting across various categories of the wealth quintile. The decline in the rate of malnutrition among the poor was slower than the decline among the better-off during 1992e2006 indicative of the rising inequality that was disadvantageous to the poor. A similar pattern was observed across different geographical regions in India during this period. The prevalence of child malnutrition varied considerably and signicantly by the wealth status of households across the different states and geographic regions of India. For instance, in 2005e2006 in Tamil Nadu, nearly 49% of the children among the poorest quintile were malnourished whereas only 16% of children from the richest quintile were malnourished (Appendix 2) On the other hand, in Madhya Pradesh, 66% of the children in the poorest quintile were malnourished compared to 45% in the richest quintile. Bihar was not much behind Madhya Pradesh in this regard. In the southern region, 53% of the children belonging to the poorest quintile were underweight compared to only 21% among those belonging to the richest quintile. In the eastern region, the prevalence of child malnutrition ranged between 62% among the poorest to 25% among the richest. A similar pattern was observed in the central region. The other two rounds of NFHS also depict patterns that were similar to patterns observed in the third round of NFHS in 2005e2006.

(0.604e0.630) (0.546e0.573) (0.485e0.513) (0.373e0.400) (0.269e0.296) (0.581e0.608) (0.485e0.513) (0.391e0.419) (0.308e0.335) (0.196e0.222) (0.495e0.521) (0.367e0.393) (0.276e0.302) (0.200e-0.225) (0.138e0.161)

a Predicted probabilities adjusted for state of residence, maternal and paternal education, type of area of residence, age of child, sex of child, birth order, preceding birth interval, mothers age at birth, breastfeeding status, social group, and religion. b All the predicted probabilities were signicantly different at p < 0.01.

582

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

Fig. 2. Change in proportion of underweight children (weight-for-age <2 S.D., 0e35 months) by the different categories of wealth quintile, India, 1992e2006.

Inequality analysis Economic inequalities with respect to child malnutrition refer to the degree to which malnutrition rates differ between more and less economically advantaged groups. In this study, economic inequality with respect to child malnutrition is measured using the poorerich ratio (poorest/richest wealth quintile) and concentration index. The concentration index is plotted through maps across all the states of India over three rounds of NFHS to examine the changing spatial pattern of economic inequality in relation to the prevalence of child malnutrition (Fig. 3). This reveals that states with higher prevalence of malnutrition had low economic inequality as compared to states with low prevalence of malnutrition cutting across time periods. For example, it may be noted from Fig. 3 that southern, western and northern states with relatively low prevalence of malnutrition had higher economic inequalities as seen from the concentration index during 1992e2006, compared to central and eastern states which had higher levels of malnutrition. An increasing trend was observed in the poorerich ratio while the concentration index declined (Table 2) over the three NFHS rounds in India. The poorerich ratio increased from 1.8 to 2.4 during 1992e1998 and then stagnated at 2.4 during 1998e2006. The concentration index decreased from 0.10 to 0.14 during 1992e1998 and remained at a plateau (0.14) during 1998e2006, suggesting that while the levels of child malnutrition were falling,

economic inequalities with respect to child malnutrition were increasing, thereby being disadvantageous to the poor. Economic inequalities with respect to child malnutrition varied considerably across the different states and geographic regions in India (Table 2). The economically and demographically progressive states in the southern and western parts of India recorded relatively higher economic inequalities with respect to child malnutrition than their counterparts in the central and eastern parts. For instance, in Tamil Nadu, rising economic inequalities during 1992e2006 were indicated by the increasing poorerich ratio (from 2.1 to 3.1) and declining concentration index (from 0.14 to 0.15). On the other hand, in a demographically and economically laggard state like Madhya Pradesh, economic inequalities with respect to child malnutrition have been relatively low, though they increased slightly from 1.4 to 1.5 (poorerich ratio) during 1992e2006. It is also important to examine the trends in economic inequalities with respect to child malnutrition among the geographic regions/states in which child malnutrition registered a steep decline, and in regions/states where marginal declines were observed (Appendix 3). We found that geographic regions/states that achieved higher declines in child malnutrition also registered higher increase in economic inequalities with respect to child malnutrition during 1992e2006. For example, the southern region and states of Punjab, Tamil Nadu, Jammu and Kashmir, Uttaranchal, Maharashtra, Andhra Pradesh, Himachal Pradesh, and West Bengal where the relative change in prevalence of malnutrition was more than 20%, inequalities measured in terms of concentration index remained exceptionally high during 1992e2006. On the other hand, in other geographic regions/states where the relative change in prevalence of malnutrition was less than 20%, inequalities remained low.

Association between average malnutrition and economic inequality We also examined the association between average malnutrition rates and economic inequality with respect to child malnutrition measured by concentration index. This was used to comprehend how various states of India compare in average malnutrition rates and economic inequalities with respect to child malnutrition during 1992e2006. Fig. 4 plots the prevalence of average underweight children on the x-axis and negative values of

Fig. 3. Spatial pattern of concentration index across major states, India, 1992e1993, 1998e1999 & 2005e2006.

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

583

concentration index on the y-axis. Data presented in Fig. 4 shows a negative association between the concentration index and prevalence of underweight children in each of the three survey rounds. The ndings indicate that the magnitude of socioeconomic inequalities are relatively higher among states where average malnutrition rates are low, implying that states which have achieved relatively substantial reduction in child malnutrition (low average malnutrition rates) have not been as successful in reducing economic inequalities (high values of CI) in undernourishment over time i.e. the benet of reduction in the burden of child undernutrition has largely bypassed the children from the poorest wealth quintile, and were more to the advantage of children from the richest wealth quintile during 1992e2006. For instance, Punjab, Tamil Nadu, Maharashtra, Kerala observed higher economic inequalities (although the average malnutrition rates were low) with respect to child malnutrition than Madhya Pradesh, Bihar, Jharkhand, Chattisgarh where average malnutrition rates were high. This nding was consistent across survey rounds. Multivariate analysis Earlier studies on child malnutrition had documented important socioeconomic, demographic and residence related variables that affected child malnutrition. In order to examine the magnitude of change in child malnutrition among children belonging to the ve categories of the wealth quintiles over the period 1992e2006, having adjusted the results for important socioeconomic and demographic characteristics, we ran a binary logistic regression

model after pooling datasets from three rounds of NFHS. The addition of two-way interactions between the poorer, middle, richer, richest with three variables reecting the historical period was statistically signicant suggesting that economic inequalities with respect to child malnutrition have changed over time. The predicted probabilities presented in Table 3 suggest that, for children belonging to the poorest quintile, the probability of being underweight declined from 0.617 in 1992e1993 to 0.508 in 2005e2006, a decline of about 18 percentage points. Over the same period, the proportion of underweight children among the richest quintile declined from 0.285 in 1992e93 to 0.151 in 2005e06, a decline of about 47 percentage points. The decline in underweight children among the poorer and richer quintiles during 1992e2006 was 32% and 45% respectively. The regression result clearly suggests that the disparities between the poor and the rich with respect to child malnutrition have widened during 1992e2006. The improvement in the nutritional status of children in India has been disproportionate as it accrued generally to the richest quintile, while the children belonging to the poorest quintile have benetted the least. Discussion and conclusion This paper examines the trends and patterns of economic inequalities with respect to child malnutrition across major geographical regions and states of India during 1992e2006. The study indicates that though the average prevalence of underweight children in India declined from 53% to 46% during 1992e2006, it fails to divulge the enormous economic disparities in child malnutrition across the geographical regions and states of India. Evidence reveals disproportionate concentration of child malnutrition among the poor across the geographical regions and states of India. Results clearly suggest an increase in disparity between the poor and the rich with respect to child malnutrition during 1992e2006. Economic inequalities with respect to child malnutrition varied considerably across different states and the pace of change was different too; the developed states in the southern and western parts of India recorded higher economic inequalities with respect to child malnutrition compared to their northern counterparts. Previous studies have noted large economic inequalities in the health status of children and utilization of ICDS services (IIPS & ORC Macro, 2007; Nandy, Irving, Gordon, Subramanian, & Smith, 2005). For example, Nandy et al.s (2005) studies have found that children from the poorest households in India had a higher risk of being undernourished with multiple anthropometric failures than their rich counterparts. Similarly, the utilization of any ICDS service varies sharply by economic status of households; for instance, in 2005e2006, 37% of the children from the poorest quintile received any ICDS services compared to only 16% of the children from the richest quintile (IIPS & ORC Macro, 2007). Again, during the same period, only 24% of the children from the poorest quintile received complete immunization compared to 71% from the richest quintile. These inequalities in utilization of ICDS services contribute toward the inferior health status of mother and child, and offers limited access to health facilities, particularly among the poor. This leads to inappropriate physical growth, poor nutritional status and frequent bouts of infectious diseases like diarrhea that further exacerbate the risk of malnutrition, particularly among children belonging to the lower socioeconomic groups. Earlier studies have found poverty as one of the determinants of child malnutrition (Bhat & Zavier, 1999; Nair, 2007; Pathak & Singh, 2009). Poverty estimates for 1992e1993 suggest that 40% of the population of the economically backward states like Bihar (including Jharkhand), Uttar Pradesh (including Uttaranchal),

Fig. 4. Average underweight children (weight-for-age <2 S.D., 0e35 months) versus concentration index across states, India, 1992e2006.

584

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585

Madhya Pradesh (including Chhattisgarh), Orissa and Assam from the central and eastern parts of India were living below the poverty line (Economic Survey, 2001e2002). As found in earlier studies, our results also suggest that these were the states that also recorded the highest prevalence of underweight children (more than 50%) in the country in 1992e1993. Recent poverty estimates released in 2004e2005 and our results also indicate a positive association between the levels of poverty and child malnutrition. Regional variations in the prevalence of child malnutrition can be partially attributed to the variations in access to ICDS program across different regions. The access to the ICDS program was found to be poorest in the states which reported the worst nutritional indicators (Deolalikar, 2004; Gragnolati et al., 2005). The utilization of ICDS varied noticeably from 10% to 90% across the different states of the country; states with a better performance in terms of child malnutrition also reported higher utilization of ICDS and vice versa. The highest utilization of ICDS was recorded in the North-eastern states. Our results suggest that there was indeed a noticeable decline in economic inequalities with respect to childhood malnutrition in the North-eastern states during 1992e2006. Not only did the focus and utilization of the ICDS vary across the better and poorer states, but the budgetary allocations to tackle the nutritional problem also varied across these states (Dasgupta, Lokshin, Gragnolati, & Ivaschenko, 2005). Signicant variations were also found in public expenditure on nutritional programs across these states. Poor performing states like Bihar, Uttar Pradesh, Madhya Pradesh, and Rajasthan were found to spend only Rs.30/- to Rs. 50/- on nutritional programs per child. On the contrary, better performing states like Gujarat, Punjab, Haryana, and Tamil Nadu were found to spend Rs. 90/- to Rs. 170/- per child (Deolalikar, 2004). This clearly suggests that the states reporting lower levels of child malnutrition were also likely to record three to ve times more expenditure on nutritional programs compared to the states that reported higher levels of child malnutrition. The investment per child in the North-eastern states was of the magnitude of Rs. 500, the highest among the six geographic regions. The largest decline in the risk of malnutrition was observed in the middle wealth quintile, while the least decline occurred in the poorest quintile in India and across the states. This may be attributed to the striking improvement in the economic status of the population, as measured by the wealth index, in the middle wealth quintile, while no change or only a marginal one was observed among the poorest and richest wealth quintile during 1992e2006. The largest decline in the risk of malnutrition among the middle wealth quintile also supports the argument toward the critical role of improvement in the economic status of households in the middle wealth quintile, relative to the poorest and richest wealth quintile in India and across states during 1992e2006. An important nding that emerges from this study is the negative association between average malnutrition rates and economic inequalities with respect to child malnutrition (measured by concentration indices) indicating that states that have achieved relatively substantial reduction in average levels of child malnutrition have been less successful in reducing economic inequalities in undernourishment over time i.e. the benet of reduction in the burden of child undernutrition has largely bypassed the children from poorest wealth quintile chiey to the advantage of children from richest wealth quintile during 1992e2006. These ndings are consistent with those of earlier studies that also document a negative association between average malnutrition rates and socioeconomic inequalities in child malnutrition (Wagstaff & Wanatabe, 2000; Zere & Mcintyre, 2003). These ndings may guide several health policy lessons. For instance, any unidimensional health policy that focuses solely on reducing average malnutrition rates among high prevalence states

at the cost of overlooking the socioeconomic distribution of malnutrition among low prevalence states might further escalate the disproportionate burden of child malnutrition among the poor and the marginalized in the low prevalence states. This will deepen the poorerich divide and increase socioeconomic inequities in nutrition and health besides destabilizing the efforts to reduce average malnutrition rates in India. With the new thrust on nutrition and child health care programs, especially in the central, eastern, northern, and north-eastern regions under the National Rural Health Mission (NRHM, 2005e2012) (MOHFW, 2005), so in areas with highest malnutrition average malnutrition rates may improve in the near future in India. Findings reported here suggest that the prevalence of underweight is disproportionately higher among the poor and that disadvantaged groups in the low prevalence regions also need special attention. Policies should focus more on arresting growing economic inequalities in child malnutrition than just accelerating the decline in average malnutrition rates. To conclude, we summarize the three key messages that emerge from our analysis. First, the burden of child malnutrition remains disproportionately higher among the poor irrespective of the state or region of residence. Second, despite huge efforts on the part of the Government of India and various state governments to reduce the rich-poor gap in child malnutrition, it is clear that the disparities between the rich and the poor with respect to child malnutrition have widened during 1992e2006. Third, any attempt to curb average malnutrition among children may not automatically reduce socioeconomic inequalities associated with child malnutrition as demonstrated by the strong negative association between average underweight rates and the concentration index across Indian states over time. Therefore, an approach is warranted that simultaneously regulates and addresses socioeconomic inequalities, as well as reducing high average malnutrition rates among children in India. Acknowledgment An earlier version of this paper was presented at the Harvard Centre for Population and Development Studies (HCPDS), Harvard University (7th May, 2009) and at the Annual Meeting of the Population Association of America (PAA) held in Detroit in 2009. The comments of the participants helped in the revision of this paper. We thank Prof. David Canning and Sebastian Linnemayr from the Harvard University, and Prof. T.K. Roy, Prof. M. Guruswamy and Dr. S.K. Mohanty from the International Institute for Population Sciences, Mumbai, India for their useful comments. We also acknowledge the editor and four anonymous reviewers whose suggestions helped improve the paper. Appendix. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.socscimed.2011.06.024. References
Ahluwalia, M. S. (2002). State level performance under economic reforms in India. In A. Krueger (Ed.), Economic policy reforms and the Indian economy (pp. 91e128). Chicago: University of Chicago Press. Arnold, F., Nangia, P., & Kapila, U. (2004). Indicators of nutrition for women and children in India: current status and programme recommendations. In S. Irudaya Rajan, & K. S. James (Eds.), Demographic change, health inequality and human development in India (pp. 141e164). Hyderabad, India: Centre for Economic and Social Studies. Basu, K., & Maertens, A. (2007). The pattern and causes of economic growth in India. Oxford Review of Economic Policy, 23, 143e167. Bhat, P. N. M., & Zavier, F. (1999). Findings of national family health survey: regional analysis. Economic and Political Weekly, 34(42/43), 3008e3032.

P.K. Pathak, A. Singh / Social Science & Medicine 73 (2011) 576e585 Bose, A. (1991). Demographic diversity of India-1991 census, state and district level data. Delhi, India: B.R. Publishing Corporation. Braun, J. V., Ruel, M., & Gulati, A. (2008). Accelerating progress toward reducing child malnutrition in India e A concept for action. Washington, D.C. USA: International Food Policy Research Institute (IFPRI). Carr, D. (2004). Improving the health of the worlds poorest people. In. Health bulletin, Vol. 1. Washington DC: Population Reference Bureau. Coalition for Sustainable Nutrition Security for India (CSNSI). (2008). Overcoming the curse of malnutrition in India: A leadership agenda for action. New Delhi: Secretariat, Safdarjung Enclave. Dasgupta, M., Lokshin, M., Gragnolati, M., & Ivaschenko, O. (2005). Improving child nutrition outcomes in India: can the integrated child development services program be more effective. World Bank Policy Research Working Paper, 3647, 1e25. Deaton, A., & Dreze, J. (2009). Nutrition in India: facts and interpretation. Economic & Political Weekly, XLIV(7), 42e65. Deolalikar, A. B. (2004). Attaining the millennium development goals in India. Washington DC: The World Bank, Human Development Unit, South Asia Region (Mimeo). Dibley, M. J., Goldsby, J. B., Staehling, N. W., & Trowbridge, F. L. (1987). Development of normalized curves for the international growth reference: historical and technical considerations. American Journal of Clinical Nutrition, 46(5), 736e748. Dibley, M. J., Staehling, N. W., Neiburg, P., & Trowbridge, F. L. (1987). Interpretation of Z-score anthropometric indicators derived from the international growth reference. American Journal of Clinical Nutrition, 46(5), 749e762. Dyson, M., & Moore, M. (1983). On kinship structure, female autonomy, and demographic behaviour in India. Population and Development Review, 9(1), 35e60. Economic Survey. (2001e2002). Ministry of Finance, Government of India. http:// indiabudget.nic.in/es2001-02/chapt2002/chap102.pdf Accessed 20.03.09. ESRI. (2009). Arc GIS 9.3. Redlands, CA, USA: The ESRI. Filmer, D., & Pritchett, L. H. (2001). Estimating wealth effects without expenditure data- or tears: an application to educational enrollments in states of India. Demography, 38(1), 115e132. FOCUS (Focus On Children Under Six). (2006). Citizens initiative for the rights of children under six, New Delhi. http://www.righttofoodindia.org/data/rtf06focus reportabridged.pdf Accessed 15.02.09. Ghosh, J., & Chandrasekhar, C. P. (2003). Per capita income growth in the states of India. http://www.thehindubusinessline.com/2003/08/12/stories/2003081200 110900.htm Accessed 13.04.09. Gragnolati, M., Shekar, M., Dasgupta, M., Bredenkamp, C., & Lee, Y. (2005). Indias undernourished children: a call for reform and action. In. Health Nutrition, and Population (HNP) Discussion paper. Washington D.C: World Bank. Grantham-McGregor, S., Cheung, Y. B., Cueto, S., Glewwe, P., Richter, L., & Strupp, B. (2007). Child development in developing countries. 1. Developmental potential in the rst 5 years for children in developing countries. Lancet, 369, 60e70. Gwatkin, D. R., Rutstein, S., Johnson, K., Suliman, E., Wagstaff, A., & Amouzou, A. (2007). Socio-economic differences in health, nutrition, and population. Washington DC, USA: World Bank. Gwatkin, D. R., Bhuiya, A., & Victora, C. (2004). Making health systems more equitable. Lancet, 364, 1273e1280. Houweling, T. A. J., Kunst, A. E., & Mackenbach, J. P. (2003). Measuring health inequalities among children in developing countries: does the choice of the indicator of economic status matter? International Journal of Equity in Health, 2(8),1e12. Howe, L. D., Hargreaves, J. R., Gabrysch, S., & Huttly, S. R. A. (2009). Is the wealth index a proxy for consumption expenditure? A systematic review. Journal of Epidemiology and Community Health, 63, 871e877. International Institute for Population Sciences (IIPS)ORC Macro (2007). National Family Health Survey (NFHS - 3), 2005e06: India, Vol. I. Mumbai: IIPS. International Institute for Population Sciences (IIPS)ORC Macro (2000). National Family Health Survey (NFHS - 2), 1998e99: India. Mumbai: IIPS. International Institute for Population Sciences (IIPS). (1995). National Family Health Survey (NFHS - 1), 1992e93: India. Mumbai: IIPS. Johnson, K., & Bradley, S. (2008). Trends in population and health outcomes: Further analysis of the 2006 Nepal demographic and health survey. Calverton, Maryland, USA: Macro International Inc. Kakwani, N., Wagstaff, A., & Doorslaer, V. E. (1997). Socioeconomic inequalities in health: measurement, computation and statistical inference. Journal of Econometrics, 77(1), 87e104. Kandala, N. B., Fahrmeir, L., Klasen, S., & Priebe, J. (2009). Geo-additive models of child undernutrition in three sub-Saharan African countries. Population, Space and Place, 15(5), 461e473. Kurian, N. J. (2000). Widening regional disparities in India - some indicators. Economic and Political Weekly, 35(7), 538e550. Lawn, J. E., Tinker, A., Mridha, M. K., & Anwar, I. (2006). Where is maternal and child health now? Lancet, 368(9546), 1474e1477. Mackenbach, J. P., & Kunst, A. E. (1997). Measuring the magnitude of socioeconomic inequalities in health: an overview of available measures illustrated with two examples from Europe. Social Science and Medicine, 44(6), 757e771. Ministry of Health & Family Welfare (MOHFW). (2005). National Rural Health Mission (2005e2012). New Delhi: Department of Family Welfare, Government of India. Mishra, V. K., Roy, T. K., & Retherford, R. D. (2004). Sex differentials in childhood feeding, health care, and nutritional status in India. Population and Development Review, 30(2), 269e295.

585

Mohanty, S. K., & Pathak, P. K. (2009). Rich-poor gap in utilization of reproductive and child health services in India, 1992e2005. Journal of Biosocial Science, 41(3), 381e398. Montgomery, M. R., Gragnolati, M., Burke, K. A., & Paredes, E. (2000). Measuring living standards with proxy variables. Demography, 37(2), 155e174. Nair, K. R. G. (15 September, 2007). Malnourishment among children in India: a regional analysis. Economic and Political Weekly, 3797e3803. Nandy, S., Irving, M., Gordon, D., Subramanian, S. V., & Smith, G. D. (2005). Poverty, child undernutrition and morbidity: new evidence from India. Bulletin of the World Health Organization, 83(3), 210e216. Pande, R., & Yazbeck, A. S. (2003). Whats in a country average? Wealth, gender and regional inequalities in immunization in India. Social Science and Medicine, 57(11), 2075e2088. Pal, P., & Ghosh, J. (2007). Inequality in India-A survey of recent trends. In. DESA working paper, No. 45. http://www.un.org/esa/desa/papers/2007/wp45_2007. pdf Accessed 12.02.09. Pathak, P. K., Singh, A., & Subramanian, S. V. (2010). Economic inequalities in maternal health care: prenatal care and skilled birth attendance in India, 1992e2006. PLoS ONE, 5(10), e13593. doi:10.1371/journal.pone.0013593. Pathak, P. K., & Singh, A. (2009). Geographical variation in poverty and child malnutrition in India. In K. K. Singh, R. C. Yadava, & A. Pandey (Eds.), Population, poverty and health: Analytical approaches (pp. 183e206). New Delhi, India: Hindustan Publishing Corporation. Planning Commission, Government of India. (2007). Poverty estimates for 2004e2005. http://planningcommission.nic.in/news/prmar07.pdf Accessed 13.03.09. Poel, E. V., Hosseinpoor, A. R., Speybroeck, N., Ourti, T. V., & Vega, J. (2008). Socioeconomic inequality in malnutrition in developing countries. Bulletin of World Health Organisation, 86(4), 282e291. Ram, F., & Roy, T. K. (2004). Comparability issues in large sample surveys-some observations. In T. K. Roy, M. Guruswamy, & P. Arokiaswamy (Eds.), Population, health and development in India-changing perspectives (pp. 40e56). New Delhi, India: International Institute for Population Sciences, Mumbai. Rawat Publications. Rutstein, S. O. (2008). The DHS wealth index: approaches to rural and urban areas. In. DHS working paper, No. 60. Calverton, MD: Macro International Inc., USA. Rutstein, S. O., & Johnson, K. (2004). DHS comparative report 6: The DHS wealth index. Calverton, MD: ORC Macro; Measure DHS, USA. Rutstein, S. O., Johnson, K., and Gwatkin, D. R. (2000). Poverty, health inequality, and its health and demographic effects. In: Paper presented at the 2000 Annual Population Association of America (PAA) Los Angeles, California, USA. Rutstein, S. O. (1999). Wealth versus expenditure: Comparison between the DHS wealth index and household expenditure in four departments of Guatemala. Calverton, Maryland: ORC Macro. Sen, A., & Himanshu. (2005). Poverty and inequality in India: getting closer to the truth. In A. Deaton, & V. Kozel (Eds.), Data and dogma: The great Indian poverty debate. New Delhi: Macmillan. Subramanian, S. V., Perkins, J. M., & Khan, K. T. (2009). Do burdens of underweight and overweight coexist among lower socioeconomic groups in India? American Journal of Clinical Nutrition, 90(2), 369e376. Subramanian, S. V., Kawachi, I., & Smith, G. D. (2007). Income inequality and double burden of under and over nutrition in India. Journal of Epidemiology and Community Health, 61, 802e809. Subramanian, S. V., & Smith, G. D. (2006). Patterns, distribution, and determinants of under- and overnutrition: a population-based study of women in India. American Journal of Clinical Nutrition, 84(3), 633e640. Subramanian, S. V., Smith, G. D., & Subramanyam, M. (2006). Indigenous health and socioeconomic status in India. PLoS Med, 3, e421. Svedberg, P. (2006). Declining child malnutrition: a reassessment. International Journal of Epidemiology, 35(5), 1336e1346. Tarozzi, A., & Mahajan, A. (2007). Child nutrition in India in the nineties. Economic Development and Cultural Change, 55(3), 441e486. Vyas, S., & Kumaranayake, L. (2006). Constructing socio-economic status indices: how to use principal component analysis. Health Policy and Planning, 21(6), 459e468. Wagstaff, A., & Doorslaer, V. E. (2004). Overall versus socioeconomic health inequality: a measurement framework and two empirical illustrations. Health Economics, 13(3), 297e301. Wagstaff, A. (2002b). Poverty and health sector inequalities. Bulletin of World Health Organization, 80(2), 97e105. Wagstaff, A., & Watanabe, N. (2000). Socioeconomic inequalities in child malnutrition in developing world. In. Policy research working paper, Vol. 2434. Washington D.C: World Bank. Wagstaff, A. (2002a). Caring more about the poor: Inequality aversion, health inequalities, and health achievements. Washington D.C: World Bank. Walker, S. P., Wachs, T. D., Gardner, J. M., Lozoff, B., Wasserman, G. A., Pollitt, E., et al. (2007). Child development in developing countries. 2. Child developmental-risk factors for adverse outcomes in developing countries. Lancet, 369, 145e157. World Bank. (2006). Repositioning nutrition as central to development: A strategy for large scale action. Washington DC, USA: The International Bank for Reconstruction and Development. World Health Organization Multicenter Growth Reference Study Group. (2006). WHO child growth standards: Length/height-for-age, weight-for-age, weight-forlength, weight-for-height and body mass index-for age e methods and development. Geneva: World Health Organization. Zere, E., & Mcintyre, C. (2003). Inequities in under-ve child malnutrition in South Africa. International Journal for Equity in Health, 2(1), 1e10.