Send to

Choose Destination
See comment in PubMed Commons below
J Biosoc Sci. 1998 Jul;30(3):333-48.

Infant and child mortality in Bangladesh: age-specific effects of previous child's death.

Author information

  • 1Health and Population Surveillance Programme, International Centre for Diarrhoeal Disease Research, Dhaka, Bangladesh.


This study examines whether mortality of two adjacent siblings in families is age-specific and is modified by the MCH-FP programme and fertility and mortality declines in Matlab, Bangladesh, using data for singleton births during 1977-78, 1985-86 and 1989-90 in the treatment (MCH-FP) and comparison areas. Logistic regression was used to estimate the net effects of survival status of elder siblings on mortality of younger siblings in the neonatal, postneonatal and toddler periods, controlling for birth order, previous birth interval, maternal age, education and religion, household possession of valuable items and sex of the child. Odds of neonatal and postneonatal deaths of younger siblings were found to be higher if the elder sibling had died at the same age than if the sibling had survived infancy. Toddler mortality was lower if the elder sibling had died in infancy. The association between two siblings'mortality risks did not decline over time in either area. The results suggest that a family history of child deaths by age is important to identify when subsequent infants would be at a higher risk of dying.


This study examined the patterns of 2 adjacent siblings' mortality risk at different ages in 3 birth cohorts in treatment and comparison areas of Matlab, Bangladesh. Data were obtained from records of the Matlab demographic surveillance system and Matlab Household Socioeconomic Censuses in 1974 and 1982. The sample included singleton births occurring in 1977-78, 1985-86, and 1989-90. Findings indicate that the odds of neonatal and postneonatal deaths of index children were higher, if an elder sibling had died at the same age. Toddler mortality was lower, if the older sibling had died as an infant. In the 2 later cohorts, the odds of neonatal deaths were very high among cases where the survival status of the older sibling was unknown. The odds of neonatal death were significantly higher, if the elder sibling died in the neonatal period rather than as a toddler. A short preceding birth interval and higher birth order were associated with higher odds of neonatal deaths. In all 3 cohorts, the odds of postneonatal mortality were significantly higher if the elder sibling had died in the postneonatal period rather than as a toddler. Infant death of an older sibling reduced the odds of a toddler death in the 1985-86 cohort. Program factors did not reduce siblings' shared mortality risk. The odds of dying were lower in later cohorts. Mortality decline was faster in the toddler group. Findings suggest family factors associated with infant mortality risk were not associated with the absolute level of mortality. Higher order births had higher mortality risk at all ages. The age pattern of sibling mortality risk varied widely by all factors.

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Loading ...
    Support Center