The aetiology of low birthweight in a rural area of Tanzania

A study conducted at Sumve District Hospital in northwestern Tanzanias Mwanza region reviewed the 3162 births (2929 singleton and 233 multiple births) at this rural facility in 1989 and 1990. Within 48 hours of delivery mothers completed a questionnaire on socioeconomic factors and their medical and obstetric history; midwives provided information on the delivery and weighed the infants within half an hour of birth. Mean birth weight of the 2926 live births was 3015 grams; 14.6% of these infants were low birth weight (LBW under 2500 grams). The perinatal mortality rate was 96/1000 births (90/1000 in singletons and 165/1000 in multiple births). Stillbirths accounted for 81% of the perinatal mortality and 19% of deaths occurred in the early neonatal period. The local curve of birth weight-for-gestational age was comparable to that in other East African populations. There were no systematic seasonal differences in LBW. Multiple regression analysis adjusted for socioeconomic and demographic factors revealed significant associations between LBW and nulliparity (adjusted odds ratio [OR] 3.32) self-reported vomiting (OR 1.44) enlarged spleen (OR 1.84) and female sex of the infant (OR 1.38). Among multiparous women previous perinatal mortality was significantly associated with LBW (adjusted OR 1.69) while preterm birth (under 37 weeks gestation) was correlated with nulliparity (OR 2.16) and enlarged spleen (OR 2.35). Small-at-term (LBW plus gestational age over 37 weeks) was significantly associated with nulliparity (adjusted OR 3.59) and female sex of the infant (OR 1.45). Although future studies are needed to explore the effect of hard physical work maternal morbidity and antenatal care on pregnancy outcomes the only modifiable factor associated with either preterm birth or small-for-date in the present study was malaria-associated spleen enlargement.