July 2004, Volume 54, Issue 7

Original Article

Socioeconomic Differences in Housewife's ability to take Nutritional Care

R. Hakeem  ( Department of Food and Nutrition, RLAK Government College of Home Economics )
F. Asar  ( Department of Food and Nutrition, RLAK Government College of Home Economics )
A. H. Shaikh  ( Department of Botany, Federal Urdu University of Arts, Science and Technology, Karachi. )


Caregiver's knowledge, attitudes, resources and control on resources are important components of model of care presented by UNICEF.1 Women have an important role in determining nutrition of household and societies but due to social injustice they may lack resources or control on resources required to fulfil their roles properly.2 Hupkens et al3 have noted class differences in food rules mothers' impose on their children. Begin et al4 found that while Socioeconomic factors associated with children's height-for-age, caregiver characteristics influence children's nutritional status, even while controlling for the socioeconomic status of the household. Maternal education was found to affect child's nutrition by Bende and McCann5 in Bolivia but not by Kutty6 in India. In Bangladesh, family income, mother's literacy, family size and area of residence exhibited strong and statistically significant associations with night blindness in the 1989 cross- sectional study.7 In Tanzania McCauley noted that health education could not be directed at the women alone because the decision to change behavior had to be sanctioned by the husband in the household.8 Women's power has been observed to be directly associated with child's nutritional status in Zambabwe.9 Women's lack of power could also effect mothers nutrition care practices by inducing depression10 or by effecting her own health related decisions.11,12

Thus not only family income and mother's abilities but also her decision-making power would affect family member's health and nutritional status.8,12,13 In a review of relevant literature from Pakistan, Bharmal14 also, identified unawareness of the mother about healthy behaviors, lack of decision- making power of women as possible causes of malnutrition

In an earlier study conducted in Karachi it was observed that children from higher and middle income group had better nutritional status than those from low-income group did.15 In order to explore whether the three income groups were different in other nutrition-care-related aspects also comparisons were made of various sub-samples of households. The purpose of this study was to explore socio-economic differences in the following aspects of housewives' ability to safeguard food and nutrition security of family titled in this paper as her 'Nutritional Care Potential' (NCP):

1. Nutritional knowledge
2. Belief in health locus of control
3. Decision-making pow


As this study aimed at exploring association between Socio economic status (SES) and nutritional status, of urbanized population, sampling was purposive. A total of 180 families (60 each for low, middle and high-income groups) were recruited only from urban authorized residential areas of Karachi. Plot size was used as the primary variable for selection of families for each of the three groups. Families living in houses built on around 80, 200 and 600 square yards were recruited to represent low middle and high-income groups.


This study was part of a larger project15 that aimed at exploring association between Socio-economic status (SES) and nutritional status, of urbanized population, sampling was purposive. For the project A total of 600 families (200 each for low, middle and high-income groups) were recruited only from urban authorized residential areas of four districts of Karachi division. The households included in this study were a randomized sub-sample of the main cohort. Selection criteria and recruitment procedure for the main sample were as follows:

Selection criteria: The authors and data collectors belonged to the same city, they also had knowledge about general income level of the residents of various localities and plot size was observed to be a feasible factor to recruit families of varying SES. Thus plot size was used as the primary variable for selection of families for each of the three groups. However information about family's monthly income and possession of household assets was also obtained to check the validity of using plot size as criteria for determining the SES of families. Information about housing and household possessions was used to develop as SES scale, and each family was assigned a SES score accordingly. Significant differences in reported income and SES score of the three groups were expected.

In order to control the effect of variability in family composition, stage of family life cycle was controlled. Childless families, families with married children and extended or joint families were not included in the study. Families with pregnant or lactating mothers and children less than 1 year of age were also excluded.

Recruitment procedure: From each district approximately 10% of the urban-legal localities were selected randomly. As valid information regarding exact proportion could not be obtained the proportion of families of varying SES recruited from various localities was not statistically controlled. However an attempt was made to have a balanced representation by recruiting at least 2 families of each income level from each selected locality.

Within the selected localities the blocks of houses having specific size were identified. Data collectors visited these blocks, interviewed families, and families that fulfilled the selection criterion and were also willing to participate were recruited. The families were identified through snow balling rather than statistical randomization.

Data collection

Trained data collectors visited households and the housewives were interviewed about family's socio-demographic characteristics and their own nutritional knowledge, attitudes and decision making.

The interviewers to assess the housewife's nutritional knowledge administered a nutritional knowledge test consisting of 20 Multiple choice questions.

The questions regarding health locus of control were adapted from Health related behavior questionnaire used at the national level for British schoolchildren.16

For determining health locus of control, children were asked to mention their agreement or disagreement with four statements regarding health locus of control. The question were: i) "I Am in charge Of My Health"; ii) "If I Keep Healthy I Have Just Been Lucky"; iii) "If I take care I'll stay healthy", and iv) "Even if I look after myself I can still fall ill". The responses to these questions indicated the extent to which the children considered themselves responsible for their health (internal locus of control). These questions were used to generate a health locus of control score as used by Balding, 1994.16 For the first and the third question where agreement indicated internal control over health, a score of 1 was given for agreement and -1 for disagreement. For the second and the fourth question where agreement indicated external control over health, a score of -1 was given for agreement and 1 for disagreement. For all the four questions 'not sure' was given a score of zero. The sum of all the four scores generated a locus of health control score for each child, which could range from -4 to +4. Lower scores or more negative scores indicated that the child considered external factors to be responsible for their health.

In order to assess housewives power in household decision making a list of 21 common decisions was presented and they were requested to mention who is usually responsible for making the particular decisions. Both, decision having direct or indirect impact on nutritional care were included. Number of decisions made exclusively by the housewife indicated their decision making power.

Date entry and analysis

SPSS for windows, version 7.5 was used for data entry and analysis. For comparing the statistical significance of the differences in three groups where scale level data (e.g. knowledge scores) was available, ANOVA was performed. In order to further explore the differing pairs within the three groups, t-test was used. All the statistical tests were performed at the 95% confidence level.


Socio-demographic characteristics of the sample are given in Table 1. Though all the differences were not statistically significant, house wives from the three income groups differed to varying extent in their nutritional knowledge, health related attitude and decision making power.

Housewives' Attitude Towards Health

Responses to various items of HLC assessment Scale are given in Table 4. It could be noted that in all income groups there was no consistency in responses. Though a vast majority (80-90%) from each income group agreed to the statements 'I am in charge of my health' and 'I would stay healthy if I take care of myself'; more than half and three fourth of them respectively also agreed to the statement 'If I keep healthy, I have just been lucky' and 'Even if I look after myself , I can still fall ill'. This pattern of responses indicates lack of firm belief in internal control over health. Lack of belief in external control is also evident by low total HLC score. The higher score indicated firmer belief in internal health locus of control. At each income

Table 1. Socia-demographic characteristics of housewives.
Parameters Statistics Low Middle High
Age(in completed Years) Mean (SD) 41.3(10) 42.1(10) 42.7(8)
Education (years of formal education) Mean (SD) 12.1 (4) 13.3 (2) 13.0 (2)


% 28 15 13


% 72 85 87
Table 2. Percentage of Housewives who gave correct respones to questions about nutrients at three income levels.
    Low Middle High
    % % %
    n=60 n=60 n=60
1 Calorie is a unit of measuring energy x 42 63 62
2 Protein helps in building muscle and tissues 50 63 67
3 Carbohydrates is found in potatoes and provides us energy 58 66 78
4 Fat is found in butter and provides us energy 92 90 98
5 Calcium is a mineralx 40 67 57
6 Iron is a mineral x 47 62 72
7 Iodine is a mineral 55 58 60
8 Ascorbic acid is a vitamin x 47 65 72
9 Folic acid is a vitamin 36 45 50
10 Thiamine is a vitamin x 27 45 47
x Chi sq. test: differences statistically significant at least at P<0.05 level.

level the mean score was less than one where 4 was the maximum possible score. However, the mean score was highest for the high income group and lowest for the middle income group but the differences were not statistically significant.

Housewives satisfaction with their Decision-Making Power

In high and low income groups only 41.7% and 40.2% house wives were completely satisfied whereas in the middle income group 50.8% of housewives were completely satisfied with their decision making power.

Table 3. Housewives' Knowledge about sources of nutrients.
Q=Which of the following Answers! Low Middle High
is Good source of ...   n=60 n=60 n=60
    % % %
CALCTUM Milk 83 85 90
  No Response 0 2 0
  Potato 5 0 3
  Tomato 12 13 7
IROW Liver 32 38 52
  No Response 3 2 0
  Fish 23 15 2
  Apple 37 46 47
VITAMIN C A Orange 88 98 97
  No Response 2 2 0
  Apple 5 0 0
  Banana 5 0 3
VITAMIN A Milk 42 44 50
  No Response 17 8 12
  Rice 20 20 25
  Banana 22 28 13
VITAMIN B1 Wheat 52 66 72
  No Response 42 12 15
  Apple 5 12 3
  Banana 2 12 3
PROTEIN Egg 83 90 90
  No Response 7 2 0
  Rice 5 3 5
  Soup 5 5 5
IODINE Fish 87 93 93
  No Response 7 2 2
  Rice 2 2 5
  Orange 5 3 0
VITAMIN D Fish oil 38 41 57
  No Response 23 8 8
  Spinach 12 23 7
  Oats 27 28 28
HARMFUL FATS Ghee 98 89 95
  No Response 0 3 2
  Vegetable Oil 0 3 2
  Fisk oil 2 5 2
CARBOHYDRATES Banana 52 66 60
  No Response 10 2 7
  Spinach 12 16 13
  Grapefruit 27 16 20
        !correct answer in bold and underlined
ä Chi.Sq. Test (between right and wrong answers); difference statistically significant at P<0.05 level
Table 4. Housewives ciews about control on health.
  Low Middle High
  % % %
I am in charge of my health      


92 93 95


5   3

Not sure

3 7 2
If I keep healthy. I have just been lucky      


83 90 92


10 8 3

Not sure

7 2 5
If I take care of myself. I'll stay healthy      


80 73 88


8 8  

Not sure

12 18 12
Even if I look ofter myself. I can still fall ill      


57 57 55


20 20 25

Not sure

23 23 20
Mean helth locus of control score 0.48 0.4 0.62
(max possible = 4 )      

Though the housewives from middle income group appear to be more satisfied than the two other groups, the difference was not statistically significant.

Overall Nutrition Care Potential of Housewives

In order to have an overall estimate of housewives nutrition care potential, the scores on three components were ranked and sum of ranks was calculated to generate NCP score. The three groups differed in their NCP scores but the difference was statistically significant only between low and high-income group (P=0.015).

In order to explore association between education and components of nutrition care potential within each SES group the housewives were divided into two groups according to educational level. Those at or above the 50th percentile were labeled as 'more educated' and others as 'less educated'. Mean scores for Decision making power, and nutrition knowledge of the less and more educated groups within each income level were compared. DMP score of more educated group was higher than the less educated group at both low and middle income groups but the difference was statistically significant for middle income only (Figure 1). Nutrition knowledge score of more educated group was higher than the less educated group at all three income groups but the difference was statistically significant for low income only. Nutrition Care potential score also of more educated group was higher than the less educated group at all three income groups and again the difference was statistically significant for low income only (Figure 2). Within each income group working housewives appear to have higher nutrition care potential (Table 6). The pattern of differences though consistent was not statistically significant in most cases.


The results of this study document socioeconomic difference in nutrition care potential. Though nutrition-care potential of housewives was not found to be commendable at any income level, the fact that it is low at lower income levels indicates that poor nutritional status at lower income level is not because of income only. Provision of balanced diet at low income levels is possible only when all the resources of the family are very carefully distributed among various items of expenditures and the food money is used judiciously to provide adequate diet to all family members. If we assume that the housewife has major responsibility for family's nutritional status, she must have ability and authority to exercise these controls. Reports of improvement in nutrition and health without escalation of income level are not non-existent.17 Even in developed countries overall development is attributed to input in nutrition education and women empowerment.18

Table 5. Percentage of housewices who had a major role in making particular decisions at three income levels.
  Decisions Income Level
  Low Middle High
  n=60 n=60 n=60
  % % %
Decision directly effecting outritional care      
1. How much food should be given to carious family members? 90 97 93
2. How the food should be cooked? x 83 95 78
3. How the food should be served? x 83 98 80
4. What type of food should be bought? 45 52 57
5. What type of food should be cooked in home? 63 80 68
6. What type of fruits should be bought 18 23 28
7. Who should cook the food? x 73 92 78
8. Who should get the best quality of food? x 58 85 55
9. Whether to eat out or not? 23 17 22
10. Whether to eat the leftover food or not 72 80 75
Decision indirectly effecting nutritional care      
11. How much money could be spent on servants? 67 68 63
12. How much money should be spent on house? 30 27 33
13. What type of clothes should be worn? x 68 88 82
14. What tupe of medical treatment should children get? 33 30 43
15. Whether to go to a doctor or not? 35 30 47
16. Whether to go for outings or not? 27 22 15
17. Whether to arrange a function at home or not? 37 28 20
18. Whether to buy news papers and magazines or not? 20 23 23
19. Whether to go to any wedding or not? 43 38 32
20. Which schools should children go to ? 47 47 43
21. What should you do with your spare time? x 83 97 85
Table 6. House wife's occupation and her nutrition care potential score.
SES Occupation Components on NCPS Total
Mean SD Mean SD Mean SD Mean SD
LOW Housewives 10.93 11.47 0.47 14.35
  n=43 4.24 3.98 1.12 5.47
  Working 11.18 12.24 0.53 15.76
  n=17 5.48 4.47 1.23 5.83
MIDDLE Housewives 11.82 x 13.22 0.39 16.35
  n=51 2.41 3.44 1.25 5.55
  Working 14.11 13.22 0.44 17.44
  N=9 4.57 3.60 1.01 5.70
HIGH Housewives 11.38 13.77 0.62 17.00
  n=52 4.25 3.98 1.05 5.22
  Working 10.13 16.75 0.63 18.50
  n=8 4.32 2.96 0.92 5.71
1DMPS = Decision Making Power Score
2NKS = Nutrition Knowledge Score
3HLCS = Health Locus Of Control Score
4NCPS = Nutrition Care Potential Score
P=0.024 t-test
P=0.038 t-test

The results of this study also indicate a need for updating and enrichment of school and medical colleges' curriculum in terms of nutrition education. As it is evident from the results that even at middle and high income where the education level was higher, the very basic facts of nutrition were not answered correctly by all. The reason could be that nutrition is not included in general science curriculum in enough detail. Further, as dietitians are hardly available to vast majority, family physicians are the main source of their nutritional guidance; and regrettably nutrition education is not imparted in medical colleges also in sufficient depth and nutrition knowledge of physicians themselves has been found to be insufficient.19,20

As nutrition education at graduate or post-graduate level also is very deficient, number of nutrition experts in the country is very low. Due to lack of accreditation of nutrition courses and any register of qualified nutritionists, it is currently an open field for anyone with vested interest in nutrition. Nutrition information disseminated through local mass media is also in most cases not very effective, as it is either provided by non-nutritionists or is poorly translated from foreign books and magazines.

Besides having insufficient knowledge, at all income levels there was lack of firm belief in self-control on health. In a previous study in which the same tool was used to assess HLC score of a group of rural and urban Pakistani children, the relative difference in income groups was similar but the mean scores for the rural, low income urban and affluent urban group were relatively higher (0.53, 0.79 and 0.1.28 respectively).21 If the housewives do not believe in possibility of controlling health, no amount of knowledge or power would encourage them to modify dietary practices according to scientific knowledge. This trend needs to be ameliorated through formal and informal education. Practical experimentation and observation to appreciate the impact of nutrition rather than provision of theoretical information alone could be of greater help. Role of religious leaders could also be important in this regard.

The observation that within each income group the more educated ones had higher NCP indicates that at least some change could be expected by providing formal education opportunities to all. Compulsory school education could help women break the barriers to empowerment to some extent.

While the results of this study evidence socioeconomic differences and similarities in nutrition care potential of housewives, they also have implications for policy makers in relation to similar opportunities for education at various income levels, provision of nutrition education and contents of the curriculum. In conclusion, if all females have similar opportunities for good quality education, the effects of financial resources on family's health could be minimized.

Financial support

ISFE Switzerland provided financial support for the research project of which this study is a part.


1. Engle PL, Menon P, Haddad L. Care and nutrition: concepts and measurement. Washington, DC: International Food Policy Research Institute; 1997.

2. ACCSCN Nutrition Policy Papers. Women and nutrition, including papers by McGuire J, Popkin B, Chatterjee M, et al. October 1990. (SOA No. 6).

3. Hupkens CL, Knibbe RA, Van Otterloo AH, et al.. Class differences in the food rules mothers impose on their children: a cross-national study. Soc Sci Med 1998;47:1331-9.

4. Begin F, Frongillo EA Jr, Delisle H. Caregiver behaviors and resources influence child height-for-age in rural Chad. J Nutr 1999;129:680-6.

5. Bende DE, McCann MF. The influence of maternal intergenerational education on health behaviors of women in peri-urban Bolivia. Soc Sci Med 2000;50:1189-96.

6. Kutty VR. Women's education and its influence on attitudes to aspects of child-care in a village community in Kerala. Soc Sci Med 1989;29:1299-303.

7. Hussain A, Kuale G, Ali K, et al. Determinants of night blindness in Bangladesh. Int J Epidemiol 1993;22:1119-26.

8. McCauley AP, West S, Lynch M. Household decisions among the Gogo people of Tanzania: determining the roles of men, women and the community in implementing a trachoma prevention program. Soc Sci Med 1992;34:817-24.

9. Hindin MJ. Women's power and anthropometric status in Zimbabwe. Soc Sci Med 2000;51:1517-28.

10. Deimling GT, Smerglia VL, Schaefer ML. The impact of family environment and decision-making satisfaction on caregiver depression: a path analytic model. J Aging Health 2001;13:47-71.

11. Bloom SS, Wypij D, Das Gupta M. Dimensions of women's autonomy and the influence on maternal health care utilization in a north Indian city. Demography 2001;38:67-78.

12. Nath DC, Land KC, Goswami G. Effects of the status of women on the first-birth interval in Indian urban society. J Biosoc Sci 1999;31:55-69.

13. Obrist van Eeuwijk B. Urban health and household decision-making. Trop Med Int Health 1999;4:77-8.

14. Bharmal FY. Inequity and health. Is malnutrition really caused by poor nutrition? J Pak Med Assoc 2000;50:273-5.

15. Hakeem R. Socio-economic differences in height and body mass index of children and adults living in urban areas of Karachi, Pakistan. Eur J Clin Nutr 2001;55:400-6.

16. Balding J. Young people in 1993. Exeter: Health Education Unit, University of Exeter:1994.

17. Ratcliffe J. Social justice and the demographic transition: lessons from India's Kerala State. Int J Health Serv 1978;8:123-44.

18. Lithell UB. Background factors in the decline of infant mortality in Sweden 1800-1930. Scand J Soc Med Suppl 1991;46:53-61.

19. Akhtar T, Rehman Z, Godwin P. Nutrition situation in Pakistan gap in medical education. Islamabad: National Institute of Health and UNICEF. National Institute of Health, 1988.

20. Nizami SQ, Bhutta ZA. Self-reported feeding advice by physicians for common childhood illnesses. J Pak Med Assoc 1999;49:298-301.

21. Hakeem R, Thomas J, Badruddin SH. Urbanisation and health related knowledge and attitudes of South Asian children. J Pak Med Assoc 2001;51:437-43.

(Erratum in: J Pak Med Assoc 2002;52:51.


Objective: To explore socio-economic differences in 'Nutritional Care Potential' (NCP) of housewives belonging to three distinctively different income groups living in urban areas of Karachi.

Methods:Data was collected from families living in small, medium and large sized houses located in the authorized urban residential areas of Karachi. A total of 180 housewives (60 each for low, middle and high income groups) were interviewed. Trained data collectors visited the households, interviewed the housewives about family's socio-demographic characteristics and their own nutritional knowledge, health locus of control and decision making.

Results: The total NCP scores increased with income level (Low = 14.8+5.6; Middle 16.58+5.5; High = 17.28+5.3) but the difference was statistically significant only between low and high income groups (t-test P = 0.015). The mean nutrition knowledge score of low income group was lower (mean score = 11.7+4.1) and significantly different from both middle (mean score = 13.5+3.4 t-text P = 0.013) and high income group (mean score = 14.2+4.0, t-test P = 0.001). All the three income groups had firmer belief in internal health locus of control (mean score less than one out of a total of 4). Though, the mean score was highest for the high income group and lowest for the middle income group but the difference were not statistically significant. There were no marked differences in decision making power of the three groups.

Conclusion:The results of this study document socioeconomic difference in nutrition care potential. Though nutrition care potential of housewives was not found to be commendable at any income level, the fact that it is low at lower income levels indicates that poor nutritional status at lower income level is not because of income only (JPMA 54:382;2004).

Journal of the Pakistan Medical Association has agreed to receive and publish manuscripts in accordance with the principles of the following committees: