1 Poverty or Patriarchy?

The ability to be well-nourished is one among a relatively small number of centrally important human freedoms (Sen 1992: 44). Malnutrition, or lack of proper nutritional attainment, would imply the absence of freedom to lead a healthy life. Undernutrition, the leading form of malnutrition in the south Asian countries, would further imply the absence of freedom to lead a minimally healthy life. From the perspective of human development, undernutrition demands a higher degree of priority, as it signifies deprivation or deficiency, than overweight/obesity, which indicates excess.

The incidence of undernutrition among women in India is one of the highest in the world. As high as 35.6% of adult women (15-49 years) in India suffer from chronic energy deficiency (CED) indicated by a body mass index (BMI) below 18.5 in 2005-06. This is much higher than the incidence in most countries of sub-Saharan Africa. For instance, of the 23 countries of sub-Saharan Africa, for which comparable data exist, with a singular exception of Eritrea, all other countries have lower incidence of CED than that of India (Deaton and Dreze 2009: 54). Similarly, among 26 countries having comparable data on iron deficiency anaemia, only four countries such as Mali, Senegal, Ghana and Congo have higher incidence than does India (55.3%).1

Why does undernutrition among women in India remain high? Undernutrition is often considered as both an outcome and manifestation of poverty: “being poor almost always means being deprived of full nutritional capabilities” (Osmani 1992: 1). Indeed, the states which remain at the top of the undernutrition pyramid in India, such as Bihar, Chhattisgarh, Jharkhand, Madhya Pradesh and Orissa, also have much higher incidence of poverty (Himanshu 2010).2 Though a broad correspondence between regional patterning of undernutrition and poverty persists in

Economic & Political Weekly

EPW

India, it is mediated by yet another important factor: rigid norms and discriminatory practices against women.

Several studies have brought out that social norms restricting women’s freedom and autonomy and discriminatory practices in the intrahousehold resource allocations against women are relatively intense in the northern region of India, broadly defined, to which some of these states belong (Miller 1981; Chen 1995).3 Analyses on women’s malnutrition suggest that such social norms and discriminatory practices are also responsible, in addition to the usual determinants, for the higher incidence of undernutrition among women in south Asian region in general (Ramalingaswamy et al 1996; Osmani and Bhargava 1998; Osmani and Sen 2003).

These analyses give birth to – by implication – a rather powerful perception: the supposed disadvantage of women in nutritional attainment in the south Asian countries. That is, given the entrenched norms and discriminatory practices against women, the incidence of malnutrition among adult women would be much higher than among men in India. Though this is plausible, it still remains empirically untested given the paucity of data. The availability of representative and comparable data on the aspects of nutrition for men, along with women, at all-India level for the first time enables us to examine the empirical basis of this belief. Apart from the verification of a belief, it is intrinsically important to know how women in India fare vis-à-vis men in nutritional attainment, one of the centrally important aspects of human development.

The questions the paper endeavours to examine are the following. Do adult women in India lag far behind men in nutritional attainment? Where is the gender gap in undernutrition higher: among poor households where hunger coexists with the absence of rigid norms against women or among richer households where the reverse seems to hold good? Does the gender gap in undernutrition vary in line with the varying intensity of norms and practices against women across the states of India? These questions are examined by analysing the National Family Health Survey-3 (2005-06) (NFHS-3, hereafter) unit-level data.

2 Data and Measures

As stated above, the analysis makes use of the NFHS-3, unit-level data, which provide nutritional information for men, along with women, for the first time in India.4 NFHS-3 provides information on two aspects of nutrition, namely, the BMI and iron deficiency anaemia, for 1,16,855 women aged 15-49 years and 70,130 men aged 15-59 years. The survey, carried out during 2005-06 in all the states of India, adopted systematic and uniform sampling techniques and field and clinical procedures (see IIPS and Macro International 2007 for details), and thereby yields comparable estimates on nutritional attainment from a large and representative sample. To ensure uniformity, the analysis is confined to women (1,16,855) and men (64,736) belonging to 15-49 years only.

How do we measure malnutrition? The BMI, which measures the weight to squared height (W/H2), below 18.5 is normally referred to as thinness or chronic energy deficiency (CED). By contrast, BMI above 25 and 30 refer respectively to overweight and obesity, which are also indicative of poor nutrition. Similarly, if

96 haemoglobin in the blood is below 12 grams/decilitre for women and 13 grams/decilitre for men, it is taken as an indication of anaemia, in terms of deficiency of iron in the blood. We will examine how women and men fare in these two aspects of malnutrition in India.

3 Gender Gap in Undernutrition

Three broad patterns can be seen from Table 1, which presents the incidence of malnutrition among adult women and men in India. One, around 36% of adult women suffer from CED in India. It has been observed that with close to 40% of women suffering from CED, “the situation can be considered critical in India” (Black et al 2008: 244).5 But the incidence of CED among adult men is no less: it is only marginally lower than among women. Hence, the gender gap in CED remains marginal.6 The gap does not go up, even if we disaggregate the incidence of undernutrition in terms of its severity, from mild/moderate to severe. Spatially, the incidence of CED among women is much higher in rural than urban India – the difference is as high as 15 percentage points. This huge spatial difference appears among men as well.

Table 1: Malnutrition among Adult Women and Men in India (2005-06, %)

All-India Rural Urban Women Men Gap Women Men Gap Women Men Gap

CED* 35.6 34.2 1.4 40.6 38.4 2.2 25.0 26.5 -1.5

Mild/moderate 28.7 28.6 0.1 33.0 32.3 0.7 19.6 21.7 -2.1

Severe 6.9 5.6 1.3 7.6 6.1 1.5 5.4 4.8 0.6

Overweight/obese 12.6 9.3 3.3 7.4 5.6 1.8 23.5 15.9 7.6

No of persons 1,11,781 65,741 – 75,415 42,438 – 36,366 23,303 –

Anaemia Any anaemia 55.3 24.2 31.1 57.4 27.7 29.7 50.9 17.7 33.2

Mild/moderate 53.5 22.9 30.6 55.5 26.1 29.4 49.4 17.0 32.4

Severe 1.8 1.3 0.5 1.9 1.6 0.3 1.5 0.7 0.8

No of persons 1,16,855 64,736 – 79,888 41,963 – 36,976 22,773 –

* Excludes pregnant women and those who gave birth within two months preceding the survey. Source: Computed from NFHS-3 unit-level data.

Two, a much larger proportion of adult women than men are anaemic, in terms of iron deficiency, in India. The difference stands as high as about 30 percentage points. Contrary to CED, the spatial gap in anaemia is smaller among women, though a fair measure of spatial gap exists among men. While the gender gap remains quite high, in absolute terms, in both rural and urban India, it is relatively higher in urban than rural India. The incidence of anaemia is about 10 percentage points lower than the incidence of CED among men, the reverse holds good for women – anaemia is 20 percentage points higher than CED.

Three, CED among urban women is not only marginally lower than that of urban men, it is also substantially lower (by 12 percentage points) than the incidence among rural men as well. By contrast, anaemia among rural men is much lower (by 23 percentage points) than among urban women. These clearly call for going beyond the broad, binary distinction between women and men, as it can possibly cloak certain region-specific patterns.

These broad patterns seem to neither lend support to the widely held notion nor invalidate it either. Instead, they indicate that the gender gap in nutrition is likely to be more complex than generally thought of: women and men are equally disadvantaged in one aspect, whereas women are unequally disadvantaged in

july 16, 2011 vol xlvi no 29

SPECIAL ARTICLE

another aspect of nutrition. These contrasting patterns also pose a rather disquieting question: Why does the gender gap remain quite low in CED and substantially high in iron deficiency anaemia? Before attempting an answer, it is essential to examine whether these broad patterns remain unchanged across the economic and social groups in India. This is important for at least two related reasons.

First, these broad patterns might cloak, as seen above, certain patterns specific to economic and social groups as well. Second, it is clear that the incidence of undernutrition among women from the poor and disadvantaged social groups in India remains unequally high (Jose and Navaneetham 2008). But, there is also some agreement that these disadvantaged groups in India do not strictly adhere to the rigid norms and practices against women (Bardhan 1993; Dreze and Sen 1995: 158; Miller 1997). This generates a rather unique combination: the coexistence of hunger with less rigid gender norms among the poor or disadvantaged groups and the reverse among the rich or advantaged groups. Hence, assessing the extent of gender gap among the economic and social groups is of some relevance here.

4 Wealth and Social Groups

Table 2 presents the incidence of undernutrition among women and men across the wealth groups in India. Before discussing the results, a note on the construction of wealth groups is in order. NFHS-3 did not collect information on household income or expenditure. Instead, it collected detailed information on a range of household assets and durables. Based on the approach advanced by Filmer and Pritchett (2001) which uses the possession of 33 household durables and assets to assess the wealth status of the households, all the households have been divided into five wealth groups or quintiles. However, a recent study (Mishra and Dilip 2008) shows that the above methodology is insensitive to difference s between both rural-urban and states in India. Hence, we use the improved methodology suggested by Mishra and Dilip (2008).

Table 2: Gender Gap in Undernutrition among Wealth Groups in India (2005-06, %)

* Same as in Table 1. ** indicates ratio of percentages of lowest quintile to that of highest quintile, respectively. Source: Computed from NFHS-3 unit-level data.

Economic & Political Weekly

EPW

For a start, 47% of women from the bottom wealth group suffer from CED, whereas it is only 24% at the top wealth group – a difference of about 23 percentage points or two times. This drastic decline, going along with a rise in the wealth status, gives rise to a stark and graded disparity, which is both unequal and unjust. Similar is the case with men as well, and the difference (20 percentage points) is only marginally lower than among women. Though huge inequalities appear in both rural and urban India, the inequality is marginally larger in urban than rural India. The inequality seems to be higher among urban women (three times and 26 percentage points), followed by urban men (two times and 23 percentage points). The concentration index, which measures the extent of inequality similar to that of Gini coefficient, clearly reinforces the above.

What is the extent of gender gap in CED across these wealth groups? Surprisingly, not only is the gender gap quite marginal among the bottom wealth groups, but also the gap gets reversed in favour of women among the top wealth groups. This is especially so in urban India. In rural India, though the gap does not reverse, it remains rather marginal, and more so among top two quintiles. Thus, the lack of gender gap in CED reveals the unequally high incidence of CED among both women and men from poor wealth groups, followed by a stark inequality or gradient in rural and urban India.

Around 60% of women from the bottom wealth groups are anaemic. Though the incidence declines – by 11 percentage points

– along with a rise in wealth status, the incidence among women from the top wealth groups is not low either, remaining as high as over 50%. Despite much lower levels of anaemia, the extent of decline along with the improvement in the wealth status remains relatively higher among men (nearly 15 percentage points) than the corresponding decline among women. This goes along with a relatively higher inequality among men than among women. Though wealth-based disparity remains in both rural and urban India, it is much lower for women than for men. Indeed, the highest disparity is found among urban women, followed by rural men. Also, the extent of disparity is relatively lower in anaemia than in CED for both women and men in rural and urban India.

Contrary to CED, the gender gap in anaemia goes up, from 28 to 32 percentage points, along with a rise in the wealth status. Surprisingly, men from the bottom wealth groups have much higher incidence of CED than women from not only the top, but also the next two wealth groups. By contrast, the incidence of anaemia among women from the top wealth group remains even higher than the incidence among men from the bottom wealth group. Yet again, significantly a large gender gap in anaemia exists in both rural and urban India, and the gap is larger in latter than the former and higher among top than the bottom wealth groups. These are quite contrary to those found in CED.

Patterns almost parallel to the above appear among social groups as well (Table 3, p 98). The incidence of CED remains as high as 47% among tribal women, though this is a bit lower than the incidence among the bottom wealth group. Expectedly, a (so-called) rise in the social status goes along with a reasonable decline (17 percentage points), which is much lower than the decline associated with the rise in the wealth status. This implies that social and wealth statuses are associated with stark disparity in the incidence of CED among women. But, the extent of such disparity is much larger between the wealth than between the social groups. These clearly hold good for men as well. Here too, the extent of gender gap in CED is lower across all the social groups, and the largest gap (only 5 percentage points) is found among the tribals.

Table 3: Gender Gap in Undernutrition among Social Groups (2005-06, %)

* Same as in Table 1. ** indicates the ratio of ST percentages to corresponding percentages of others. Source: Computed from NFHS-3 unit-level data.

Closer to 70% of tribal women in India are anaemic, which is 7 percentage points larger than that found among the bottom wealth group. Yet again, social differentiation coexists with a large difference (about 17 percentage points between the tribals and others), which is higher than the difference (10 percentage points) between the top and bottom wealth groups. However, despite this huge difference between the social groups, the incidence of anaemia remains quite high among all social groups. Interestingly, not only is the incidence of anaemia among men lower than the incidence of CED, the extent of decline is even larger than that of CED. This leads to a substantial gender gap in anaemia across all social groups. In fact, the extent of gender gap in anaemia among tribals and scheduled castes is marginally higher than that of the bottom two wealth groups.

Can we infer from these broad patterns whether the gender gap in undernutrition is larger among richer households and advantaged social groups or vice versa? The answer is far from sure. Not only does the gender gap in CED remain lower across all wealth and social groups, but also marginal among richer wealth groups and advantaged social groups. Conversely, a substantially large gender gap in anaemia appears across all wealth and social groups, and the gap is marginally higher among the richer and advantaged social groups. Moreover, the incidence of CED remains higher among wealth groups when compared to social groups, and the inter-group disparity is also higher among the former than the latter. Contrary to this is the incidence of anaemia, which remains higher among social groups than wealth groups, though inter-group disparity remains almost same among wealth and social groups.

In an interesting recent paper, Sen et al (2009) bring out an important finding which has a direct bearing on our analysis, and hence, needs a brief discussion. With the help of a new methodology, which singles out the influence of gender from class, they establish an entrenched gender hierarchy, independent of class, in non-treatment for long-term ailments in Koppal district, Karnataka: the poorest women at the bottom, preceded by poor women, non-poor women, and then come the poorest men, who are preceded by poor men and non-poor men at the top (ibid: 406-o7). Since this finding has some relevance to our attempt, we would like to see whether a similar hierarchy emerges in the aspects of undernutrition in India as well. Table 4 presents the results.

Before discussing the results, a methodological note is in order. Sen et al (2009: 404) have regrouped the five class (income) groups into three to make the discussion simpler. We have carried out a post hoc analysis, which suggests that the mean differences between these five wealth groups are statistically significant. Hence, instead of regrouping them into three, we present the nutritional performance of women and men from these five wealth groups as such to see whether any hierarchy exists.

Two contrasting patterns emerge from Table 4. A definite gender-based hierarchy emerges in the incidence of anaemia: with much higher incidence, women, irrespective of the wealth differences, remain at the top. Though a hierarchy based on wealth does seem to emerge among both women and men, it is

Table 4: Undernutrition and Gender Hierarchy in India

CED* % Anaemia %

Poorest women 47.1 Poorest women 61.2

Poorest men 44.1 Poor women 58.2

Poor women 42.2 Non-poor women 56.3

Poor men 39.8 Richer women 53.1

Non-poor women 37.7 Richest women 50.0

Non-poor men 36.3 Poorest men 32.8

Richer women 32.1 Poor men 26.7

Richer men 30.6 Non-poor men 23.7

Richest men 23.9 Richer men 22.3

Richest women 23.7 Richest men 18.0

* Same as in Table 1. Source: Computed from NFHS-3 unit-level data.

5 Behind the Gender Gap

Before discussing the possible reasons for gender gap in anaemia, it may be useful to examine whether these contrasting patterns are specific to India alone. Interestingly, these same contrasting patterns do appear from other regions of the world as well. For instance, the gender gap in CED seems to be small not only in south/south-east Asia, but also in sub-Saharan Africa and Latin America (Nube and Boom 2003: 525). By contrast, the incidence of anaemia is much higher among women than among men, leading to a large gender gap in almost all regions of the world, though the prevalence rates are higher in south Asia, which is partially included in the regional group referred to as south-east Asia-II (Table 6).

Table 6: Gender Gap in Anaemia across the Regions of the World (Figures in %)

It, thus, appears that the two contrasting patterns, including the large gender gap in iron deficiency anaemia, are not exclusively the problems of India. In a sense, these broad patterns convey that India is no different from the global patterns on gender gap in

Economic & Political Weekly

EPW

undernutrition. Instead, what makes India noteworthy is the much higher levels of undernutrition among women. As we have seen already, the incidence of CED is equally high among both women and men, whereas the incidence of anaemia is unequally high among women. Hence, identifying the potential causes for the higher incidence of anaemia among women becomes important.

Iron deficiency anaemia is a combined outcome of a number of related, mutually reinforcing factors, such as (Denic and Agarwal 2007: 607):

(1)“a low bioavailability of iron” in plants; (2) “a low intake of meat”, which has a higher content and bioavailability of iron than plants;

Broadly, these factors can be grouped into three, such as

These are contrary to the claim that “menstruation is the dominant reason for a higher frequency of iron deficiency in women in comparison with men” (Denic and Agarwal 2007: 608). Similarly, Harvey et al (2005: 562) argue that “menstrual blood loss is the most significant factor affecting the iron status” of women. More recent studies, however, take a rather balanced view. For instance, Clark (2008: 131) argues that “key factors contributing to iron deficiency anaemia in the female population stem from a combination of menstrual blood losses and poor iron intake”. Whittaker (2008) attributes a number of causes for the iron deficiency, which also include pregnancy and blood loss due to menstruation (p 709).

Notwithstanding these competing views, it is reasonable to suggest that physiological/reproductive factors, especially blood loss due to menstruation and pregnancies, play an important role in the higher incidence of iron deficiency anaemia among adult women. Of these, heavy menstrual blood loss assumes significanc e, as it is not only one of the important factors for the anaemia among adult women (Harvey et al 2005), but also remains mostly untreated due to cultural barriers or ignorance or both. Thus, physiological/reproductive factors are responsible for at least a part of the gender gap in iron deficiency anaemia in India, as also in other countries.

However, physiological/reproductive factors, which are applicable to all adult women in the world, are only a part of the problem, and therefore, cannot explain such a high incidence among women in India. Herein, the poor bioavailability of iron, which is related to the kind of foods consumed, assumes relevance. At the risk of simplification, it can be summarised that cereal-based, vegetarian diet has relatively low iron content (Baynes and Bothwell 1990; Denic and Agarwal 2007). Absorption of iron from this iron-deficient diet is significantly inhibited by a combination of factors. The major factors are presence of large amounts of phytates and polyphenols, overcooking of vegetables and consumption of milk along with or immediately after the main meal, among others (Cook 1990; Denic and Agarwal 2007; Thankachan et al 2007).

Table 7: Consumption of Food Items among Women and Men in India (2005-06, %)

Occa and Week refer, respectively, to occasionally and weekly. Source: Computed from NFHS-3 unit-level data.

Cereal-based vegetarian diet is not only deficient in iron, but also leads to inefficient absorption of iron found in it. Though tea and coffee emerge as the prominent inhibitors of iron – if tea is eliminated, absorption of iron will increase threefold (Cook 1990: 303), phytates found in many cereals, nuts and legumes also significantl y reduce the iron absorption (Baynes and Bothwell 1990). Similarly, calcium interferes with iron absorption and addition of calcium to the diet may even induce iron deficiency – indeed it is suggested that menstruating and pregnant women should try to restrict calcium intake with main meals (Hallberg 1998).

These factors are quite relevant to the food habits of Indian households. Not only is our diet predominantly vegetarian in nature, but also largely cereal-dominated (Deaton and Dreze 2009). This iron-deficient food, coupled with poor absorption could be one of the reasons for the high incidence of anaemia among women from all wealth groups (ibid: 61). However, the combined problems of low iron content and poor absorption of iron would be equally applicable for men as well. Then, why is the gender gap in anaemia so large? Does it indicate that women are discriminated against in the intrahousehold food allocation, especially in the food items which are rich sources of iron?

It appears from Table 7 that the diet of women and men in India is largely vegetarian in nature, as large percentages of women (64.6%) and men (59%) do not eat fish, chicken/meat on a regular basis (that is, either weekly or daily). Lower still are the proportions of women and men consuming fruits (60% and 53%, respectively). By contrast, substantially large proportions of women and men eat pulses/beans (90% and 91%, respectively) and leafy vegetables (93% and 94%, respectively) on a regular basis. Consumption of milk appears in between – 55% of women and 67% of men consume milk/curd regularly.

The question here is: is there a gender gap in the consumption of these food items? Barring the milk/curd and fruits, there seems to be no significant gender gap in the regular consumption of other food items. In these two items, by contrast, the differences are rather large – about 12 percentage points in the consumption of milk/curd and about 8 percentage points in fruits.7 Interestingly, the consumption of milk/curd and fruits, where the gender gap is high, tends to go with a large decline in anaemia among women. For instance, consumption of milk/curd from never to daily goes with about 9 percentage points decline in the incidence of anaemia. The extent of decline is larger for fruits (about 16 percentage points). The largest decline emerges in

the combined consumption of milk/curd and fruits (19 percentage points).

It is likely that the gender gap in anaemia might be related to the gender gap in the regular consumption of fruits. But the decline in anaemia found along with the increase in the frequency of consumption of milk/curd seems surprising given the above discussion which identifies cal

cium as one of the inhibitors of iron absorption. Equally surprising is the increase in the incidence of anaemia along with an increase in the frequency of consumption of green leafy vegetables among women. Contrary to our expectation, increase in the frequency of consumption of fish, chicken/meat does not seem to go with the largest decline.8 These patterns seem to suggest that the relationship between intake of iron -rich foods and the incidence of anaemia is far from simple.

While regular consumption of fruits might bring a significant decline on the incidence of anaemia, whether it is possible to afford to, if not feasible to have, such a regular consumption especially among poor households is an open question. Table 8 shows that about over 75% and 70% of women from bottom two quintiles do not consume fruits regularly. By contrast, 45% and 58% of women from top two quintiles consume fruits regularly. Thus, though it is desirable and necessary, it is much beyond the reach of poor women to consume these food items on a regular basis.

The other issue is about the incidence of anaemia among women who consume these food items regularly. It appears that

6 Concluding Remarks

Is there a huge gender gap in the incidence of undernutrition among adults in India? The analysis suggests that a huge gender gap in iron deficiency anaemia coexists with an absence of a gender gap in CED. These contrasting patterns indicate that the gender gap in nutrition is likely to be more complex than generally assumed. This complexity is reinforced further by a possible wealth-based hierarchy in CED and a gender-based hierarchy in anaemia. These contrasting patterns, which are not specific to India alone, raise a number of questions. Why does the gender gap remain quite low in one aspect of nutrition and substantially high in yet another aspect? What specifically prevents the richer women and men from higher levels of CED? Similarly, what does collectively expose all women, rich and poor, to such higher levels of anaemia?

While statistical analysis suggests the positive and significant role of wealth status in the incidence of CED among both women and men (results are not reported here), whether wealth status alone is responsible for the absence of gender gap in CED is not clear. Nonetheless, it may be relevant to mention here that the CED status of an individual would depend on the intake of nutrients (input), its transformation into body mass (absorption), and energy expended (output). Equal incidence of CED might result even with differential intake of nutrients, if either energy expended or the efficiency of absorption differs. Given the gendered division of responsibilities and labour, it is likely that men’s work involves more expenditure of muscular energy than do women’s work. This might partly account for similar CED status even with differential food intake.10 Hence, the absence of gender gap in CED does not imply that there is no bias or discrimination against women: the absence of gap might persist despite the presence of gender gap in the intake of nutrients.

Physiological/reproductive factors tend to play an important role for the higher incidence of anaemia among women. They are partly responsible for the higher incidence of anaemia among women from all wealth groups in India. However, they are complemented by an iron-deficient diet (largely vegetarian, cerealbased and less-diversified) and poor absorption of iron from such iron-deficient diet. To follow a balanced, diversified diet is a desirable option, but is it feasible for the poor women, among whom the incidence of anaemia is the highest, is an open question. Herein, the fortification of commonly eaten, less-expensive staple foods appears to be an effective option.

On the face of it, the absence of gender gap in CED might look like a progressive phenomenon. But such an absence is an outcome of an equally higher incidence of CED among men does not augur well, for it implies that adult women and men in India fare equally poor in one of the central aspects of human development. Hidden behind such generic pattern is a stark socio-economic inequity – unequally large levels of CED among women and men from poor households and disadvantaged social groups in India. Added to these are the twin facts that 77% of women with CED and 73% of men with CED in India live in villages and CED has gone up, though marginally, among poor women between 1998-99 and 2005-06, a period of higher rates of economic growth in India. Hence, addressing the huge socio-economic inequity in undernutrition is as important as addressing the higher incidence of iron deficiency anaemia among women. These call for progressive public policies rather than merely hoping economic growth to trickle down and do wonders.

Notes

1 This is based on the table generated with the help of Measure DHS STAT compiler (available at www.measuredhs.com, accessed on July 7 2010), on the incidence of iron deficiency anaemia among countries where the DHS survey was carried out between 2003-04 and 2007.

2 These five states account for nearly 28% of the undernourished adult women in India. Uttar Pradesh, yet another poor state, contributes 15%. These six states, along with West Bengal, are home to more than 50% of undernourished adult women in India.

3 Note, however, that some of the states where undernutrition is high, namely, Bihar and Orissa, belong to the eastern region of India, which, in terms of regional patterning of cultural norms against women, has been placed in the middle: neither intense as in the north/north-west, nor liberal as in the south. This not only raises the question why do some of the north-western states, despite rigid cultural norms against women, perform better on women’s nutrition than the eastern states of India, but also cautions against taking a generalised view on the regional patterning of culture in India.

5 Black et al (2008) made this observation based on the estimates from NFHS-2 (1998-99) data. The observation still holds good, as the decline in CED between 1998-99 and 2005-06, a period of higher rates of economic growth in India, has been only marginal. See, in this regard, Jose and Navaneetham (2008).

6 The absence of gender gap in CED also appears from the analysis of National Nutrition Monitoring Bureau data, which is collected only from nine Indian states. See in this regard, Deaton and Dreze (2009: 52, Table 12).

Economic & Political Weekly

EPW

7 While this may be the case specific to the NFHS-3 data, it is important to bear in mind that the NFHS-3 data was not particularly aimed to capture the gender discrimination in food consumption. After all, it is not only difficult to capture this complex process through survey-based methods, but also such an attempt clearly requires intensive, time-consuming and qualitative approaches.

8 Of the two types of iron, such as heme and nonheme, heme iron is mainly found in the meat and constitutes up to 15% of the dietary iron (Hallberg 2001). Additionally, meat can enhance the absorption of non-heme iron, which constitutes the rest of the dietary iron (Baynes and Bothwell 1990).

9 See in this regard Mannar and Gallego (2002) and Mannar and Sankar (2004). 10 I am thankful to the referee for highlighting this important aspect.

References

Bardhan, K (1993): “Social Class and Gender in India: The Structure of Differences in the Condition of Women” in A W Clark (ed.), Gender and Political Economy: Explorations of South Asian Systems (New Delhi: Oxford University Press), 146-78.

Baynes, R D and T H Bothwell (1990): “Iron Deficiency”, Annual Review of Nutrition, 10: 133-48.

Black, Robert E, L H Allen, Z A Bhutta, L E Caulfield, M de Onis, M Ezzati, C Mathers and J Rivera for the Maternal and Child Undernutrition Study Group (2008): “Maternal and Child Undernutrition: Global and Regional Exposures and Health Consequences”, Lancet, 371(9608): 243-60.

Chen, M (1995): “A Matter of Survival: Women’s Right to Employment in India and Bangladesh” in M Nussbaum and J Glover (ed.), Women, Culture and Development: A Study of Human Capabilities (New Delhi: Oxford University Press), 37-57.

Clark, Susan F (2008): “Iron Deficiency Anaemia”, Nutrition in Clinical Practice, 23(2): 128-41.

Cook, James D (1990): “Adaptation in Iron Metabolism”, American Journal of Clinical Nutrition, 51(2): 301-08.

– (1997): “Social Class, Gender and Intrahousehold Food Allocations to Children in South Asia”, Social Science and Medicine, 44(11): 1685-95.

Mishra, Udaya S and T R Dilip (2008): “Reflections on Wealth Quintile Distribution and Health Outcomes”, Economic & Political Weekly, 43(48): 77-82.

Nube, M and G J M van den Boom (2003): “Gender and Adult Undernutrition in Developing Countries”, Annals of Human Biology, 30(5): 520-37.

Osmani, S R (1992): “Introduction” in S R Osmani (ed.), Nutrition and Poverty (New Delhi: Oxford University Press), 1-15.

Osmani, Siddiq and Alok Bhargava (1998): “Health and Nutrition in Emerging Asia”, Asian Development Review, 16(1): 31-71.

Osmani, Siddiq and Amartya Sen (2003): “The Hidden Penalties of Gender Inequality: Fetal Origins of Ill-health”, Economics and Human Biology, 1(1): 105-21.

Ramalingaswami, Vulimiri, Urban Jonsson and Jon Rohde (1996): “The Asian Enigma” in UNICEF, The Progress of Nations (New York: UNICEF) 11-17.

Sen, Amartya (1992): Inequality Reexamined (New Delhi: Oxford University Press).

Sen, Gita, Aditi Iyer and Chandan Mukherjee (2009): “A Methodology to Analyse the Intersections of Social Inequalities in Health”, Journal of Human Development and Capabilities, 10(3): 397-415.

Stoltzfus, Rebecca J (2003): “Iron Deficiency: Global Prevalence and Consequences”, Food and Nutrition Bulletin, 22(4, supplement): S99-S103.

Thankachan, Prashanth, Sumithra Muthayya, Thomas Walczyk, Anura V Kurpad and Richard F Hurrell (2007): “An Analysis of the Aetiology of Anaemia and Iron Deficiency in Young Women of Low Socioeconomic Status in Bangalore, India”, Food and Nutrition Bulletin, 28(3): 328-36.

Whittaker, P (2008): “Iron Deficiency and Its Prevention” in International Encyclopedia of Public Health: Volume-III (New York: Elsevier), 707-10.

Zimmermann, Michael B and Richard F Hurrel (2007): “Nutritional Iron Deficiency”, Lancet, 370(9586): 511-20.

july 16, 2011 vol xlvi no 29