Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Determinants of nutrient adequacy in a rural area of Bangladesh
Rafiqul Huda Chaudhury
United Nations Department
of Technical Co-operation for Development, Kathmandu, Nepal
INTRODUCTION
To develop an effective preventive programme against malnutrition, it is necessary not only to assess the nature and magnitude of the problem but to identify factors affecting it and their relative importance. The nutritional status of an individual is the outcome of a complex interaction of a broad range of host and environmental factors, the latter encompassing physical, biological, and especially socio-cultural ones. This study was conducted to examine the relative impact of some of these factors, particularly those socio-economic factors known to influence nutrition, as a means of identifying the most important determinants of malnutrition in rural Bangladesh. The decision to study only the effects of socioeconomic factors on nutrient intake was dictated by the consideration that the others can be readily manipulated to improve nutritional status The factors included in the model were per capita income, education, family size, female participation in economic activities and time spent on food preparation, and male participation in economic activities.
HYPOTHESIS
Income
Income is a mirror-image of a household's resources and provides an index of its purchasing capacity. Food that cannot be purchased is never consumed. The economic situation is a potent factor in determining how much and what kind of food will be available. It is, therefore, expected that with the improvement of household income, absolute expenditure on food is likely to go up, as is the calorie and protein intake of the household [9].
Income is considered to be a very important factor determining nutrition, at least in developing countries. Examination of cross-sections of food consumption data in these countries reveals highly significant differences based on income class [12]. The World Bank also underscored the importance of income as the critical determinant of nutrition in developing countries. In 1980, the Bank maintained that the serious and extensive nutritional deficiencies that exist in almost all of these countries are "largely a reflection of poverty, people don't have enough income for food" [24].
In view of these discussions, we would expect to find a positive relationship between income and dietary intake and dietary adequacy of a household. However, it should be borne in mind that income is not always an exogenous variable determining nutrient intake. It may also be an endogenous variable; i.e. income may be determined by level of nutrient intake. For example, one may note two types of nutritional effects on productivity and hence income. These effects are known as the "investment" and the "energy" effects 122]. Inadequate nutrient intake in gestation and the first few years of life may seriously impair physical and mental development. Second, malnourished individuals with a lower energy level will not be able to put enough effort into current activities to raise their period income.
The effect of income is measured here by expenditure on food, which truly reflects a household's income and resources. One important reason for using per capita expenditure on food instead of per capita income is the nonlinear relationship between income and nutrient intake. The elasticity of income with respect to food expenditure is usually less than 1, particularly for higher income groups. It has been shown empirically that as income rises beyond a threshold level, people tend to spend proportionately less on food than on non-food items, particularly consumer durables. Although the proportion of income spent on food declines, absolute expenditure on food is still higher at upper income levels. Therefore it is posited that, with the improvement of a household's income, absolute expenditure on food is also likely to go up, and so is the calorie and protein intake of the household.
Education
The relationship between parental education and the dietary intake of a household is complicated. On the one hand, greater education is associated with greater awareness of the importance of nutrition, the nutrient content of food, and nutrition options from market purchases or from home production. Better-educated parents should be able to provide more nutritious diets at any income level due to their ability to identify the nutrition values of food. Thus, when education represents efficiency in household production, the elasticity of nutrition with respect to education is likely to be positive. On the other hand, education is also a measure of taste. Higher education may also increase the desire to consume status food and non-food items to the detriment of nutrition. If education represents taste, the elasticity of nutrition with respect to education may be positive or negative.
It may be noted further that the effect of education on nutrient intake is likely to be more positive for female than male education, for the following reasons. Female education leads to efficiency in the allocation of food consumption and also, possibly, to an awareness of the importance of nutrition and less regard for status consumption. The positive effect of male education on nutrition is mostly due to the association with expenditure. Education is associated with lower prices paid for by the household, perhaps because marketing ability is enhanced. The ability to identify the nutrition values of foods allows a given expenditure on foods to yield more nutrients.
Family/Household Size
Increased family size may adversely affect the nutritional status of every member of the household because it may be associated with decreased per capita human input. That is, the allocation of food per member is likely to decrease with the increase in the number of household members, which, in turn, may have a negative affect on per capita nutrient intake. Increased household size also implies acceptance of fewer resources in a quality/quantity model of fertility decision.
In some situations, however, as in Bangladesh, family size may itself indicate higher economic status. In this case, larger families may not lead to a reduced nutrient intake by members of the household. One may also note economies of scale in nutrition because of less waste and the possibility of purchasing in bulk associated with larger family size. (One result that does show up frequently in expenditure surveys is that the amount of food purchased rises less than proportionately to household size when per capita income is held constant. This may be considered as an indication of somewhat behavioural economy of scale in food use.) It should be pointed out further that it is not family size per se but the number of adults relative to children that is the crucial factor influencing the intake of members of a household. The lower the dependency ratio, the higher the nutrient intake of the household members.
Female Participation in Economic Activities and Time Spent on Food Preparation/Cooking
The relationship between female participation in market activities and nutrient intake on one hand and female involvement in home production and nutrient intake on the other are likely to be contradictory. The former is expected to be negative, while the latter will be positive. It has been often hypothesized that women's participation in market activities will be detrimental to the health and nutrition status of the household, owing to the loss of home food production. Data from rural areas of Laguna province, Philippines [8], clearly indicated the negative effect of the mother's market activities on the overall nutritional status of a household.
However, greater involvement by women in home production is likely to be positively associated with the nutritional adequacy of a household, for three reasons: (i) women spend their time purchasing items rather than using home time to produce them; (ii) a woman working outside the home may resort to non-nutritive convenience foods to the detriment of the health of the family members; (iii) money does not always meet nutritional needs. Even with the same amount of money for food, some families are able to meet their nutritional needs, others exceed theirs, while yet others fail. These variations depend, among other things, on the time used in preparing food and housewives' knowledge and skills. In turn, knowledge and skills determine, to a large extent, how much money and time are needed and how well they are used.
We have so far discussed the ways in which withdrawal of women from home production to market activities may adversely affect the nutritional status of a household. However, this is only one side of the picture. Increasing female involvement in home production does not necessarily improve the nutritional status of the household. There is a point beyond which the degree of increase in dietary quality is reduced as more time is spent on food preparation. It is a fact that heat-labile nutrients are lost the longer some foods are cooked. Similarly, the effect of female participation in market activities on a household's nutrient adequacy is not necessarily negative, and may, in fact, be positive.
Employment outside the home provides women with alternative satisfactions to children, such as companionship, recreation, stimulation, creative ability, and social and economic rewards, which may well compete with raising children and may lead to lower fertility [2] . The lower fertility of a working woman will increase the availability of food grains per child, which may lead to the better nutritional status of a household.
Moreover, working women add substantially to the family income [3] and, in contrast to working men, spend a higher proportion of their income on food, while men spend on nonfood stuffs [10]. This could also improve the nutritional status of a household. In this study, we tried partially to control this conflicting effect of mother's work on the dietary adequacy of a household by adjusting for important related variables, such as income and number of children.
Male Participation in Economic Activities
The relationship between number of hours worked in productive activities and nutrient adequacy of a household is somewhat complex. In a subsistence economy such as that in Bangladesh, family labour, particularly male labour, is important in determining the economic condition of a household. The higher the participation in productive activities, the higher the income of a household, which in turn is likely to increase the nutrient intake of the household. However, work pattern also affects food consumption directly. Increased number of hours worked increases work-calorie requirements, and the poor in Bangladesh tend to work longer hours than the rich [2] . Therefore, in spite of working longer hours, the poor may not be able to meet their nutrient requirements, even though their incomes are raised marginally. This is because the increased income is not sufficient to offset the increased work-calorie requirement.
MATERIALS AND METHODS
The data employed in this study were collected from 572 members of 108 households in the village of Muyiarchar, approximately 220 miles north-east of Dhaka, in Sylhet district. The sample included 50 per cent of the village households selected from various socio-economic groups, classified on the basis of land holding and income. Households were selected from each socio-economic stratum, with a probability equal to their proportion in each group.
The present analysis is based on full 572 members of 108 households in the sample. The data collected included consumption of rice, wheat, fish, milk, meat, pulse, eggs, vegetables, and fruits per person per day, and the sources and costs of food consumed. Together with data on food consumption, we also obtained information on time allocation for every member of the household five years of age or older. This helped in estimating nutritional requirements, based on the type of activities in which the individuals were engaged, together with other considerations.
The data on food consumption were collected from each member of the household (mothers provided the information about children) once a month and those on time allocation twice a month (i.e. every 15 days) over a period of one year (August 1978 to July 1979) by 24-hour recall interview. Time budgets were collected from each eligible respondent of the household in the form of a sequential record of a respondent's activities and their duration for the 24-hour period preceding the interview. Interviewers were locally recruited and were mostly teachers from a nearby elementary school.
Individual food intake was converted into nutrient (calories and protein) intake by employing Indian food composition tables [9] . Average calorie requirements and safe levels of protein intake were estimated for every member of each household to assess whether a person consumed more or less than, or an amount equal to, the average requirement. Nutrient needs were estimated in two phases. In the first we determined the calorie and protein needs of children from birth to age four years, according to the age- and sex-specific recommendations of the joint FAO/WHO/UNU consultation [6]. These recommendations were adopted both to allow for the full potential growth of a child and to provide extra calories and protein to compensate for losses caused by infectious and parasitic diseases widely prevalent in rural Bangladesh. In the second phase we estimated the nutrient requirements for the remaining members of the household, those five years of age and older. These estimates were based on the same recommendations [6] .
The estimation of calorie requirements of persons of five years and above were based on actual information on their rest/sleep and physical activity. This is because human energy requirements consist of two major components: basic energy demands for resting metabolic function and all other energy costs, mainly physical activity.
To arrive at an estimation of caloric need for physical activities and rest/sleep, physical activities were grouped into three categories, light, moderate, and heavy, on the basis of actual time in minutes for each of these activities per day per person. We then calculated the energy required per person per day for each of these kinds of activities for males and females separately, as per recommendations of the FAO/WHO/UNU joint report [6] . To this was added the energy needed for sleep/rest (basal metabolic rate) per person per minute, estimated on the basis of the individual's actual weight.
A final adjustment was made to allow sufficient energy for pregnancy and lactation; the extra allowance was 285 calories per day for a pregnant woman. In Bangladesh, breast-feeding commonly extends beyond the second year of life and may continue as long as four years [4]. On the basis of actual milk production of lactating mothers ob- from rural areas of Bangladesh, we added 500 calories for mothers with children two years old or younger.
The safe allowance for protein for a person age five years and above was made on the basis of FAO/WHO/UNU recommendations [6] . This was calculated by multiplying the average safe protein allowance per kilogram of body weight in a particular age/sex group by the average weight in kilograms of a person in the corresponding age/sex group. In addition to this, allowance was made for pregnancy and lactation, a pregnant women being allotted 5.6 g of protein. We added 17 g of protein (per day per person) for mothers with children one year old or younger and 11 g for mothers with children between the ages of one and two years. All protein estimates were corrected for 70 per cent utilization.
Dependent Variable
The major dependent variable used in this study was the mean nutrient adequacy ratio (NAR) of a household. This is the ratio of the total nutrient intake of a household divided by its total nutrient requirement, times 100. This ratio is derived by summing the calorie and protein intake and nutrient need (i.e. recommended daily allowance of calorie and protein) for every member of the household separately, and then dividing the total of the former (i.e. intake) by the latter (i.e. need) times 100. The mean nutrient adequacy ratio measures whether a household is consuming more or less than, or the equivalent of, its need.
Overall Nutrition Situation of the Study Village
According to our measurement of nutrient adequacy, 40 per cent of the households in the study village failed to meet calorie requirements. This is an overall percentage only, and conceals variations in nutrient adequacy among households. For example, if the nutrient adequacy of a household is examined by its level of income utilizing data of the present study, 100 per cent of the households with a per capita monthly income of less than or equal to 148 take fail to meet the calorie requirement. The corresponding figure among households with a per capita monthly income of 149 to 225 take is 54 per cent. These two income groups acounted for 45 per cent of the households in the study village. However, 95 per cent of the households appeared to meet their protein requirements, though in reality the situation may be different. We mentioned earlier that 40 per cent of the households could not meet their calorie requirements, and these households may, for all practical purposes, be considered deficient in protein intake. Inadequate energy intake reduces the utilization of protein added to the diet.
The average calorie and protein intake per day per person in the study village was 1,829 kcal and 58 9. Compared to the national level, the average calorie intake was slightly lower, while protein intake was almost the same. According to the nutrition survey of 1981/1982 [18], the average calorie and protein intake per day per person was 1,943 kcal and 48.4 g respectively. Cereals, particularly rice, were the dominant source of calories and protein, with rice accounting for 82 and 53 per cent of calorie and protein intake respectively. This finding was corroborated by other nutrition studies in Bangladesh [16-18] .
Fish and meat were the second important sources of protein in the village, accounting for 30 per cent of total protein intake; the other noteworthy sources were pulses and wheat. Fish, meat, and wheat contributed 10 per cent of total calorie intake. These findings, showing the overwhelming dominance of cereal over animal sources, clearly reflect an imbalance in the diet of rural households
Determinants of Nutrient Adequacy: Estimates of the Basic Model for Household Demand for Nutritional Inputs
We have so far discussed the overall nutritional situation of the study village. We now examine the socio-economic and demographic factors, as posited earlier, affecting the nutrient adequacy of a household. Multiple-regression analysis was employed to measure the net effect of each of the variables in the model on dietary adequacy, and in this case could be expected to identify the effects of each of the contributory variables in the equation, that is, holding other factors constant. The functional form of the relationships is expressed in the following equation:
, where
A = average nutrient adequacy
ratio of a family;
E = effect of xs;
n = calorie/protein adequacy ratio;
Y = family income;
S = family size; and
xs = a vector of household socio-economic variables,
including education determining the taste
The reasons for employing simple linear functions for these estimates over other complicated functions are as follows: (i) the implied specification of the underlying preference function could be arbitrary; (ii) the data employed here, although significant in many respects, may not be precise enough to permit discrimination with much confidence among higher properties of functional forms; and (iii) ordinary least-squares (OLS) estimation methods stand up against specification errors better than most simultaneous methods [13] .
Multiple regression is not per se a technique for policy selection but rather a useful means of providing insights to facilitate and improve upon such selection. It provides important guidance toward relative selection of policy instruments.