Menu
Menu
Menu
Menu
Background: The Sustainable Development Goal (SDG) 3 aims at reducing neonatal and under-5 mortality to below 12 per 1000 and 25 per 1000 live births, respectively, globally by 2030. Studies have found that initiation of breastfeeding within one hour of birth and continuous breastfeeding for over 12 months can positively impact neonatal and infant health. However, there is evidence that the sex of a child may influence the breastfeeding practices of a mother. Thus, we examined sex inequality in early breastfeeding initiation in sub-Saharan Africa. Materials and methods: Data from Demographic and Health Surveys conducted in 24 sub-Saharan African countries between January 2010 and December 2019 were pooled and analysed. A total of 137,677 women of reproductive age (15-49 years) were considered in this study. Bivariate and multivariable regression analyses were performed, and the results were presented using crude odds ratio (cOR) and adjusted odds ratio (aOR) with statistical significance at a p-value less than 0.05. Results: The highest inequality in early initiation of breastfeeding was reported in Togo with a difference of 5.21% between the female and male children, while the lowest inequality was reported in Guinea with 0.48% difference between the female and male children. A higher odds of breastfeeding within 1 hour was observed among female children [cOR = 1.05; 95% (CI = 1.02-1.09)] compared to male children, and this persisted after controlling for the confounders included in this study [aOR = 1.05; 95%(CI = 1.02-1.08)]. Conclusion: We found higher odds for early breastfeeding initiation of female children compared to male children in sub-Saharan Africa. To reduce breastfeeding initiation inequalities, programmes that educate and encourage early initiation of breastfeeding irrespective of the child sex should be promoted among mothers. Copyright:
This study involved a cross-sectional analysis of DHS data from 24 sub-Saharan African countries. DHS is a nationally representative study conducted in over eighty-five low-and-middle-income countries (LMICs). The survey employed a questionnaire to collect data from respondents on several health indicators such as maternal and child health, men’s health, family planning, fertility, gender-based violence, substance use, Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome (HIV/AIDS), and nutrition [22]. Respondents for the survey were sampled using a two-stage cluster random sampling technique. The study by Aliaga and Ruilin [23] highlights the detailed sampling processes used in the DHS. The present study sample was drawn from the birth recode’s files from all the countries used. A total of 137,677 women aged 15–49 who had complete cases of the studied variables on questions about breastfeeding of the last child they had 5 years preceding the survey were included in the final analysis. Other respondents with incomplete information about the study of interest were dropped from the analysis. We relied on the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement in drafting the manuscript [24]. Sample size distribution by country and survey year are presented in Table 1. The datasets for the DHS can be accessed freely at https://dhsprogram.com/data/available-datasets.cfm. The outcome variable in this study was EIBF. EIBF is defined as the initiation of breastfeeding within the first hour (1 hour) of birth [4, 25]. From the DHS, the respondents were asked when they started to breastfeed their newborn after birth. The responses were documented in “immediately,” “hours,” and “days”. The responses were further re-categorized into EIBF (within 1 hour of birth) and late breastfeeding initiation (More than 1 hour). Similar coding has been used in several studies [9, 26, 27]. The main explanatory variable was the child’s sex. The categorization of this variable was “Male” and “Female”. A study by Sen, Mallick [9] used the same categorization to assess inequalities in EIBF. A total of thirteen (13) covariates were studied. These variables consist of maternal age, age at first birth, assistance at birth, place of residence, maternal educational level, partner educational level, parity, wanted last child, place of delivery, delivery by cesarean section, antenatal care (ANC) visit during pregnancy, wealth index, and media exposure. These variables were not determined a priori; instead, based on parsimony and significant association with EIBF [9, 28–30]. Except for the place of residence and wealth index where the existing DHS coding was used, the remaining covariates were recoded. The other covariates and their recoding include maternal age (15–24, 25–34, and 35 and above); age at first birth (below 20 years and 20 years and above); assistance at birth (unskilled and skilled); maternal educational level (no education, primary, and secondary or higher); partner’s educational level (no education, primary, and secondary or higher); parity (1–3 and 4 and above); wanted last child (wanted and unwanted); place of delivery (home and health facility); delivery by cesarean section (No/Yes); ANC visits during pregnancy (none, less than 4, and 4 or more); and media exposure (No/Yes). Exposure to radio, television, newspaper/magazine was coded as media exposure. Media exposure was derived from these three variables using panel analysis. “Yes” means exposure to mass media while “No” means no exposure to mass media. Data extraction, cleaning, recoding, and analyses were carried out using Stata software version 16.0 (Stata Corporation, College Station, TX, USA). Bar chart was used to show the sex disparities in EIBF by country. Next, the Pearson chi-square test was conducted to determine the relationship between the mother and child’s characteristics and EIBF. After this, two regression models were built to determine the associations between sex of the child, the covariates, and EIBF. Specifically, the first model (bivariate regression) examined the independent associations between sex of the child, each covariate, and EIBF. The second model (multivariable regression) was used to determine the association between a child’s sex and EIBF while controlling for the covariates. The results of the regression analyses were presented in a tabular form using crude odds ratio (cOR) and adjusted odds ratio (aOR) with their respective 95% confidence interval (CIs). Finally, the crude and adjusted results on the association between sex of the child and EIBF were disaggregated by country. Statistical significance at p-value less than 0.05. All the frequency distributions were weighted using the DHS recommended weight of v005/1,000,000 to avoid oversampling and non-response error. The survey Stata command (svy) was used to adjust to the complex sampling structure of the DHS data in the chi-square and regression analyses. The multicollinearity test, which used the variance inflation factor (VIF), revealed no evidence of collinearity amongst the independent variable and covariates. Since the authors of this manuscript did not collect the data, we sought permission from the MEASURE DHS website and access to the data was provided after our intent for the request was assessed and approved on the 10th of January 2021. The DHS surveys are ethically accepted by the ORC Macro Inc. Ethics Committee and the Ethics Boards of partner organizations in different countries, such as the Ministries of Health. The women who were interviewed gave either written or verbal consent during each of the surveys.
N/A
