Determinants associated with preterm births in Uganda: A cross-sectional study

Background Preterm births affect households’ incomes through direct and indirect expenditures associated with low productivity and the actual loss of employment in many cases. We studied the determinants of preterm birth in Uganda as one of the major contributors to neonatal morbidity and mortality, leading to households’ economic losses. Methods We used a cross-sectional research design based on the most recent Uganda Demographic Health Survey of 2016. The sample contained 1,537 women aged 15-49 years. The variable selection process was guided by categorization of the variables into; socio-demographic, reproductive history, and gestational birth characteristics. The study adopted two means of analysis. The logistic regression model to determine variables of preterm birth between 22 – 36 weeks and normal delivery period. Then the multinomial logistic regression model to determine how two preterm birth categories (22 – 32 weeks and 33 – 36 weeks) relate with the normal delivery period. Results Belonging to the poorest quintile (AOR2.09, 95% CI (1.69-2.57)) and attending antenatal care less than four times (AOR1.41, 95% CI (1.20-1.66)) had the highest odds ratios for the logistic regression model. Whereas the multinomial logistic regression model; for the 22-32 weeks category, belonging to the poorest quintile (RRR2.43, 95% CI(1.45-4.08)), attending antenatal care less than four times (RRR2.44, 95% CI (1.63-3.64)), had the highest relative risk ratios. For the 33-36 weeks category; belonging to a poorest quintile (RRR2.03, 95% CI (1.62-2.53)), having had less than four antenatal visits (RRR1.29, 95% CI (1.09-1.54)), and unwanted pregnancy (RRR1.22, 95% CI (1.03-1.45)), had the highest relative risk ratios. Conclusion Attending antenatal care for less than four times and belonging to the poorest quintile are common risk factors related to preterm birth. We therefore recommend that these receive utmost attention from the policy makers and implementers.