Birth Patterns of High-risk and Low-risk Childbearing of Bangladeshi Women: A Multivariate Statistical Analysis

Dr. Md. Abdul Mazed Chowdhury1 and Samad Abedin2 1 Associate Professor in Statistics, Department of Accounting and Information Systems, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

ABSTRACT High-risk pregnancy is a critical issue in safe motherhood. Childbearing in Bangladesh is characterized by early start of motherhood, quick progress till the peak age of reproduction and slow progress till the end of childbearing period. This article presents the results of logistic regression analysis of both high-risk and low-risk childbearing concerned with parity. Out of 13 variables, 11 variables are influencing the high-risk childbearing and only two variables are influencing low-risk childbearing. The significant variables which influence high-risk childbearing are: child loss experience, duration of conjugal life, education of women, place of residence, age at first birth, occupation of husband, women's working status, religion, duration of breastfeeding, education of husband, and spousal age difference. On the other hand, the variables influencing low-risk childbearing are age at first marriage and contraceptive use.

INTRODUCTION

Fertility in Bangladesh is high even by the standards of developing countries. Recent evidence suggests that fertility has started to decline in Bangladesh (Amin and others, 1993). The total fertility rate has declined from nearly seven births per woman in 1975 to 3.4 births per woman in 1996 (BFS, 1975; BDHS, 1993-1994). A number of demographers have argued that the mechanism of this steep fertility decline was achieved primarily due to successful family planning (Amin and others, 1990; Cleland et al., 1994; Cleland, 1993; Islam et al., 1998), that succeeded in raising the contraceptive prevalence rate (CPR) from a low level of 8 percent in 1975 to as high as 53 percent in 2000 (Mitra et al., 2001). However, from 1993-1994 the level of fertility appears to be unchanged at a level of 3.3, as indicated by the last two surveys (BDHSs) in Bangladesh in 1996-1997 and 1999-2000, respectively. About half of the population of Bangladesh is females and most of them live in rural areas with low status in the family as well as in the society. Fertility in Bangladesh is high even by the standards of developing countries.

Childbearing may have a risk of a poor or tragic outcome among those who have already had many births (Haaga, 1989). High parity is associated with increased risk of maternal mortality where mothers may be less able to meet the physiological demands of repeated pregnancy (Koenig and others, 1987). In many developing countries about 50 percent of pregnancy terminations occur among the high-risk mothers (Rinehart and Kols, 1984) and the wide choice of family planning methods now available allow health programmes to offer an appropriate technique to avoid each type of high-risk pregnancy and maternal, child and infant mortality. Every pregnancy faces risk - high or low during the childbearing period. In the following section an effort is made to identify the factors that have influence on high-risk and low-risk childbearing according to parity of women. A woman's parity, i.e., fertility, is considered as a dependent variable. The aim should be to identify the variables, which have significant influence on the dependent variables. The logistic regression analysis identifies the variables, which variables influence high-risk childbearing and which variables influence low-risk childbearing.

The analysis of births in the last five years preceding the interview is of interest in this context since it can provide further insights into the mechanisms underlying fertility change (Njogu and Martin, 1991). Childbearing in the human population is a complex phenomenon. Analysis of childbearing (high-risk or low-risk) is even more complex since a number of biological, behavioral and cultural factors are associated with it. In the context of Bangladesh, only a few studies, not all of them nationally representative, have been carried out to examine the effects of various factors on childbearing performance (Islam et al., 1995; Islam, 1999). Therefore, the purpose of the present paper is to identify the factors influencing both high-risk and low-risk childbearing in the context of parity of Bangladeshi women, employing the technique of logistic regression analysis.

MATERIALS AND METHODS

The study will utilize data from the 1999-2000 Bangladesh Demographic and Health Survey (BDHS) which employed a nationally representative, two-stage sample that was selected from the master sample maintained by the Bangladesh Bureau of Statistics (BBS) for the implementation of surveys before the next surveys (2001). A total of 10,268 households were selected for the sample, of which 9,854 were successfully interviewed. Among these households, 10,885 women were identified as eligible and interviews were completed for 10,544 or 97% of them. The multivariate logistic regression analysis provides a powerful statistical technique for identifying high-risk and low-risk childbearing with respect to several socio-economic and demographic variables, simultaneously. In this paper, high-risk and low-risk childbearing is concerned with births in the last five years and have been treated as "1" and "0", respectively.

In the 1999-2000 BDHS, a number of socio-economic and demographic variables are available. In the present analysis, the impact of 13 explanatory variables on parity is examined. The correlates are seven socio-economic characteristics: education of women, education of husband, working status, place of residence, contraception, religion of women and occupation of husband; and six demographic characteristics: duration of conjugal life, spousal age difference, age at first marriage, age at first birth, child loss experience, and duration of breastfeeding. Finally, we have considered here two groups viz., high-risk and low-risk childbearing with respect to parity treated as "1" and "0", respectively. Some socio-economic and demographic variables to identify the high-risk and low-risk childbearing are given below:
Births in the last five years precede the survey, i.e., fertility is considered as a dependent variable. The aim should be in the identifying of the variables, which have significant influence on the dependent variable. The explanatory variables considered in this model are as follows:

RESULTS AND DISCUSSION

Table 1.2 shows the percentages of births in the five years preceding the interview
that fall into different child survival risk categories, as well as the distribution of all currently married women across these categories. Over the last several decades demographers have identified high-risk fertility behavior by the following risk category.

The result of logistic regression analysis is shown in Table 1.3. The regression coefficients in the model shown in Table 1.3 are statistically significant at different levels. For example, the regression coefficient of X10 (child loss experience) is significant at 1% level of significance while the regression coefficient of X6 (duration of breastfeeding) is significant at 5% level of significance. The third column of Table 1.3 shows the odd ratio. For example, the odds ratio of education of women is 1.922 and indicates that the risk of childbearing will be 1.922 times higher for those mothers who have no education than those mothers who have some education. Similarly, the odds ratio of child loss experience is 16.977 which indicates that the risk of childbearing will be 16.977 times higher for those mothers who have child loss than those mothers who have no child loss.

Therefore, the most important significant variables that influence high-risk childbearing are place of residence, religion, age at first birth, education of women, duration of breastfeeding practices, duration of conjugal life, women's working status, child loss experience, and occupation of husband, education of partner, and spousal age difference. On the other hand, the variables that influence low-risk childbearing mothers are surviving children, contraceptive use, and age at first marriage. Table 1.3 suggests that the most highly significant variable is child loss experience (odds ratio 16.977) and the next significant variable is duration of conjugal life (odds ratio 6.523). The analysis further indicates that child loss experience will affect childbearing pattern. Age at first birth is also an important correlate of a high-risk childbearing pattern. The higher the infant and child mortality in a community; the lower the age at marriage.

Thus, improved child survival may help to motivate mothers to prolong birth spacing by practices of breastfeeding and contraceptive use. Education of women is also an important determinant of high-risk childbearing. Duration of breastfeeding is found to have a significant direct negative effect on fertility. Encouraging women to breastfeed their children for a relatively longer duration may also contribute to a reduction in fertility. The total effect of female education on fertility is found to be negative. Education may provide better employment opportunities outside home and providing education to females, especially at the secondary and higher levels can rise age at first marriage and age at first birth. Another important factor of high-risk childbearing is women's work status. Most women in Bangladesh work at home as housewives, for example, cooking, maintaining home, taking care of children and so on.

The logistic regression equations for both low-risk and high-risk childbearing are-
₀ = 0.216 - 0.393_x1 - 0.635_x2 - 0.215_x3 - 0.273_x4 + 0.084_x5 - 0.023_x6 + 0.223_x7 - 0.2.034_x8 - 0.010_x9 - 9.911_x10 + 0.714_x11 + 0.313x12 - 0.031x13

₁ = - 1.390 + 0.396_x1 + 0.653_x2 + 0.198_x3 + 0.323_x4 - 0.132_x5 + 0.096_x6 - 0.197_x7 + 1.875_x8 + 0.040_x19 + 9.930_x10 + 0.306_x11 + 0.302_x12 + 0.040_x13

CONCLUSION

Some findings of this study deserve consideration from the viewpoint of their policy implications. Child loss experience is found to have a most significant positive effect on fertility in Bangladesh, which means that mothers who have experienced child loss are found to have more births. Mothers always try to replace their dead children as early as possible. Such behavior is a result of social fear about the survival of children. Maternal mortality is also high in Bangladesh. Therefore, it is essential to strengthen maternal and child health care activities in order to reduce the level of child and maternal mortality. Great attention should also be given to the delivery of family planning services to women, particularly younger ones, and to provide them with motivational messages about the health benefits of fewer children. Also lengthening the birth interval can reduce fertility in both urban and rural areas of the country. Since the fertility of urban women is much lower than that of rural women, increasing urbanization will hasten the current trend in fertility reduction. Perhaps later age at marriage will be another observable effect of increased female status, which would have a subsequent fertility-reducing effect. Therefore, the encouragement of early marriage should be stopped, especially for adolescents.

The present study is a modest attempt to categorise the variables into two ways and to analyze the high-risk and low-risk childbearing pattern in the context of prevailing socio-economic conditions. There is no doubt about the contribution of childbearing pattern in regulating population growth of the country and policy makers and planners will show a congenial and judicious path for the development of Bangladesh.

The total effect of female education on fertility is found to be negative. Education may provide better employment opportunities outside the home, and age at marriage and age at first birth can be raised by providing education to females, especially at the secondary and higher levels.

Based on the findings of this study, it may be suggested that attention should be focused on the need for providing educational facilities for Muslim women in rural areas in order to depress the level of fertility in Bangladesh.

REFERENCES

1. Ahmed, B. (1981). "Differential fertility in Bangladesh: A path analysis". Social Biology, 28 (1-2): 102 - 110.
2. Amin, R., J. Chowdhury, A. U. Ahmed, R. B. Idyll and M. Kabir (1993). "Reproductive Change in Bangladesh: evidence from recent data", Asia-Pacific Population Journal, 8(4): 39 - 58.
3. Amin, R., S. Becken and J. Chowdhury (1990). "Recent evidence on trends and differentials in Bangladesh Fertility", Journal of Bio-social Science, 22: 225 - 230.
4. BFS, (1975). "Bangladesh Fertility Survey, 1975". Secondary Analysis, National Institute of Population Research and Training, Dhaka, Bangladesh.
5. BDHS, (1993-1994). "Bangladesh Demographic and Health Survey 1993-1994". National Institute for Population Research and Training. Dhaka, Bangladesh. Mitra and Associate and Macro International Inc. Calverton, Maryland, USA, 1993.
6. BDHS, (1996-1997). "Bangladesh Demographic and Health Survey 1996-1997". National Institute for Population Research and Training. Dhaka, Bangladesh. Mitra and Associate and Macro International Inc. Calverton, Maryland, USA, 1996.
7. BDHS, (1999-2000). "Bangladesh Demographic and Health Survey 1999-2000". National Institute for Population Research and Training. Dhaka, Bangladesh. Mitra and Associate and Macro International Inc. Calverton, Maryland, USA, 1999.
8. Cleland, J.. J.F. Phillips, S. Amin and G. M. Kamal (1994). "The Determinants of reproductive change in Bangladesh: Success in a challenging environment", Washington, D. C., World Bank.
9. Cleland, J. (19930. "Equality, Security and Fertility: a reduction to Thomas", Population Studies, 47(2): 345 - 352.
10. Haaga, J. (1989). "Mechanisms for the Association of Maternal Age, Parity and Birth Spacing with Infant Health", in A.M. Pamell, id., contraceptive use and Controlled Fertility: Health Issues for Women and Children, National Academy Press, Washington, D.C.
11. Hobcraft, J., J.W. McDonal and S. Rutstein (1983). "Child-spacing effects on infant and early child mortality". Population Index, 49: 585-618.
12. Hosmer, D. W., B. Jovanovic and S. Lemeshow (1989). "Best Subsets. Logistic Regression". Submitted for Publication.
13. Islam, M. M. (1999). "Adolescent childbearing in Bangladesh". Asia-Pacific Population Journal, Vol. 14, No. 3: 73 - 87.
14. Islam, M. M., A. A. Mamun and R. Bairagi (1998). "Fertility and its proximate determinants in Bangladesh". Evidence from the 1993 - 94 Demographic and Health Survey, Asia-Pacific Population Journal, Vol. 13, No. 3.
15. Islam, S. M. S. and H. T. A. Khan (1985). "Influences of selected socio-economic and demographic variables on fertility in Bangladesh". Asia-Pacific Population Journal, Vol. 10(2)" 51 - 63.
16. Islam, S.M.S. and H.T.A. Khan (1991). "Effects of selected socio-economic and demographic factors on fertility: A path analysis", Asian Profile, 19 (6): 561-574.
17. Koenig, M. A., J. F. Phillips and R. S. Simmons (1987). "Trends in family size preference and contraceptive use in Matlab, Bangladesh". Studies in Family Planning, 18(3), pp. 117 - 127.
18. Mahalanobis, P.C. (1963). "On the Generalized Distance in Statistics". Proceedings of the National Institute of Science, India. 12: 49-55.
19. Mahmud, M. and M. Mazharul Islam (1999). "Health status of adolescent mothers and their child: evidence from Bangladesh Fertility Survey 1989". Social Science Review 16(1).
20. Mitra, S. N., Ahmed Al-Sabir, A. R., T. Saha and S. Kumar (2001). "Bangladesh Demographic and Health Survey 1999-2000 (Dhaka, Bangladesh, Mitra and Associates).
21. Mturi, A.J. and S.L. Curtis (1995). "The determinants of infant and child mortality in Tanzania". Health Polocy and Planning, 10: 384-394.
22. Njogu, W. and T.C. Martin (1991). "Fertility decline in Kenya: the role of timing and spacing births". Proceedings of the Demographic and Health Surveys World Conference, August, 5-7, Washinton D.C. Proceedings, volume III: 1883-1901.
23. Noble, J., J. Cover and M. Yanagishita (1996). "The world's youth 1996". (Wall Chart) (Eng, Fre, Spa) Washinton, D.C., Population Reference Bureau, 2 P.
24. Population Reports (J. 1984). "Family Planning Programs". Population Information Program, The John Hopkins University, Hampton House. 624 North Broadway. Baltimore, Maryland 21205, USA.
25. Preston, S.H. (1978). "Introduction in S.H. Preston (ed.)". The effects of infant and child mortality. Academy Press, New York, p. 1-20.
26. Rindfuss, R. R., L. L. Bumpass and J. A. Palmore (1987). "Analyzing fertility histories: Do restrictions bias results?" Demography, 24: 113-122.
27. Rinehart, W. and A. Kols (1984). "Healthier mothers and children thtough family planning". Population Reports, Series J, No. 27. Baltimore, Johns Hopkins School of Public Health, Population International Program, May - June. 38p.
28. Rodriguez, G. and J.N. Hobcraft (1980). "Illustrative life table analysis of birth intervals in Colombia". WFS Scientific Report No. 16, International Statistical Institute, Voorburg, Netherlands.
29. Shafiqul, I. (1995). "Influences of Selected Socio-economic and Demographic Variables on Fertility in Bangladesh". Asia-Pacific Population Journal. Vol. 10, No. 2, pp. 51-63.
30. Srinivasan, K. (1972). "Analytical models for two types of birth intervals with applications to Indian population. St. Joshep's Press. Trivandrum, India.
31. Trussel, I. and A.R. Pebley (1984). "The impact of family planning program, infant, child and maternal mortality". World Bank Staff Working Paper.