J. biosoc. Sci. (1992) 24, 427^132
EFFECT OF CHILD SURVIVAL ON CONTRACEPTIVE USE IN BANGLADESH A. I. CHOWDHURY, VINCENT FAUVEAU AND K. M. A. AZIZ International Centre for Diarrhoea! Disease Research, Dhaka, Bangladesh
Summary. A study in Bangladesh showed that couples who lost a child often stopped practising contraception in order to have another child. Logistic regression analysis revealed that contraceptive continuation was related to maternal age, parity, husband's education and the sex of the last child. Introduction
The contraceptive practices of couples whose last child died are different from those whose last child survived. Information about the way in which child mortality affects fertility is particularly important for population policy, because childhood mortality reduction is not only a goal in itself but may also represent a way to reduce fertility (Cochrane & Zachariah, 1984). Child mortality may affect fertility in three ways; the death of a breast-fed child reduces the lactation period and thus may reduce the postpartum amenorrhoeic period, whereas infant survival and prolonged lactation should increase average pregnancy intervals (Ware, 1977). However, women who have experienced previous infant loss may be at higher risk of subsequent perinatal and neonatal loss than women whose children have survived (Butler & Bonham, 1963). A decline in child mortality may alter community fertility norms, and modify social and economic structures designed for high mortality regimes (Freedman, 1963; Beaver, 1975). At the community level the continuation of contraception comes into conflict with deeply established pro-fertility norms embedded in age-old traditions. This paper examines the changes in couples' contraceptive behaviour as a result of infant or child mortality over a 24-month period. Differential contraceptive use was also examined in relation to mother's age, parity, husband's education and sex of the last child. Materials and methods The Matlab Family Planning Health Services Project
The data are from the Matlab upazila study of a rural community in Bangladesh, which has been extensively reported (Bhatia et at., 1980; Phillips et at., 1984). This study was concerned with the 2111 women who were present in three surveys in 1977, 1978 and 1979. Of these women, 414 were using contraceptives in the baseline survey in December 1977. The project monitored their changing contraceptive practices 1 and 2 years later and relates changes to the survival status of their last child. 427
A. I. Chowdhury, V. Fauveau and K. M. A. Aziz
428
Results Table 1 presents the data on women not using contraception at the beginning of the study, together with the percentage using it 12 and 24 months later relative to the survival status of their last child. The table shows that after 1 year 23% of those who had not lost their last child commenced using contraception while only 14-8% did so among those who had lost their last child (x2 = 37-9, p < 0001). Only 4-7% of the 192 women who had no child in the prior 12 months used contraception. The use of Table 1. Percentage accepting contraception among baseline non-users by the survival status of their last child % accepting Last child (N) Demographic characteristics
After 12 months
After 24 months
Survived
Died
Survived
Died
Survived
Died
Mother's age
EFFECT OF CHILD SURVIVAL ON CONTRACEPTIVE USE IN BANGLADESH A. I. CHOWDHURY, VINCENT FAUVEAU AND K. M. A. AZIZ International Centre for Diarrhoea! Disease Research, Dhaka, Bangladesh
Summary. A study in Bangladesh showed that couples who lost a child often stopped practising contraception in order to have another child. Logistic regression analysis revealed that contraceptive continuation was related to maternal age, parity, husband's education and the sex of the last child. Introduction
The contraceptive practices of couples whose last child died are different from those whose last child survived. Information about the way in which child mortality affects fertility is particularly important for population policy, because childhood mortality reduction is not only a goal in itself but may also represent a way to reduce fertility (Cochrane & Zachariah, 1984). Child mortality may affect fertility in three ways; the death of a breast-fed child reduces the lactation period and thus may reduce the postpartum amenorrhoeic period, whereas infant survival and prolonged lactation should increase average pregnancy intervals (Ware, 1977). However, women who have experienced previous infant loss may be at higher risk of subsequent perinatal and neonatal loss than women whose children have survived (Butler & Bonham, 1963). A decline in child mortality may alter community fertility norms, and modify social and economic structures designed for high mortality regimes (Freedman, 1963; Beaver, 1975). At the community level the continuation of contraception comes into conflict with deeply established pro-fertility norms embedded in age-old traditions. This paper examines the changes in couples' contraceptive behaviour as a result of infant or child mortality over a 24-month period. Differential contraceptive use was also examined in relation to mother's age, parity, husband's education and sex of the last child. Materials and methods The Matlab Family Planning Health Services Project
The data are from the Matlab upazila study of a rural community in Bangladesh, which has been extensively reported (Bhatia et at., 1980; Phillips et at., 1984). This study was concerned with the 2111 women who were present in three surveys in 1977, 1978 and 1979. Of these women, 414 were using contraceptives in the baseline survey in December 1977. The project monitored their changing contraceptive practices 1 and 2 years later and relates changes to the survival status of their last child. 427
A. I. Chowdhury, V. Fauveau and K. M. A. Aziz
428
Results Table 1 presents the data on women not using contraception at the beginning of the study, together with the percentage using it 12 and 24 months later relative to the survival status of their last child. The table shows that after 1 year 23% of those who had not lost their last child commenced using contraception while only 14-8% did so among those who had lost their last child (x2 = 37-9, p < 0001). Only 4-7% of the 192 women who had no child in the prior 12 months used contraception. The use of Table 1. Percentage accepting contraception among baseline non-users by the survival status of their last child % accepting Last child (N) Demographic characteristics
After 12 months
After 24 months
Survived
Died
Survived
Died
Survived
Died
Mother's age
