Determinants of early initiation of breastfeeding in Amibara district, Northeastern Ethiopia: a community based cross-sectional study

A community based cross sectional study was conducted in Amibara district during April, 2015. Amibara district is one of the six districts of Zone Three, which is located 250 kms northeast of Addis Ababa. Amibara district is the main livestock productive area, which is bordered on the south by Awash Fentale, on the west by the Awash River, on the northwest by the Administrative Zone Five, on the north by Gewane, on the east by the Somali Region and on the southeast by Oromia region. It is found in Awsa and Gewane agro-pastoral zone. This agro-pastoral zone is characterized by irrigated and rain-fed crop production combined with livestock rearing. Cotton and irrigation based crop production are mainly practiced in this district.

There are eighteen kebeles (the smallest administrative units next to district in Ethiopia) under Amibara district; three urban and fifteen rural kebeles. Based on the 2007 Census population projection, the district has a total population of 76, 649. Of these 37.7 % are urban inhabitants, 51.96 % agro-pastoralists and 10.34 % pastoralists. Children under 2 years of age are estimated to be 2.5 % of the total population and average family size of the households to be 4.5 people. There are three health centers and eighteen health posts in the district [9].

A total of 407 mother-child pairs were calculated using a single population proportion formula;

Where:

n = required sample size

\( z\frac{\alpha }{2} \) = critical value for normal distribution at 95 % confidence level which equals to 1.96 (z value at alpha 0.05)

p = prevalence of early initiation of breastfeeding

d = an absolute precision (d = 0.05)

Assumptions:

5 % margin of error, 95 % confidence level, 59.6 % prevalence of early initiation of breastfeeding in Afar region of Ethiopia [10], and 10 % for non-response.

The sampling procedure was started from the stratification of the fifteen kebeles as rural and the three kebeles as urban. Out of the eighteen kebeles, two urban and three rural kebeles were randomly selected using lottery method. About 38 % (37.7 %) of mothers of children aged less than 24 months live in urban areas and the remaining 62.3 % live in rural areas. Therefore, based on population proportion, 254 were taken from rural areas and 153 from urban areas.

Then the sample size was proportionally allocated to each selected kebeles. At the time of survey, from each household unit one eligible mother who had a biological child aged less than 24 months was selected. If there were more than one mother with children aged less than 24 months in one household unit, one mother with the youngest child was selected. From mothers who had two children aged less than 24 months, the youngest child was selected as the reference. If mothers had twin children aged less than 24 months, one child was selected by lottery.

Data were collected using a pre-tested, structured and interviewer administered questionnaire adapted from the Ethiopian Demographic and Health Survey [10] and the National Nutrition Survey Questionnaire [11]. The adapted questionnaire was modified and contextualized to fit the local situation and the research objective. The questionnaire was prepared first in English and translated in to Qafar’af, then back to English to check for consistency. The Qafar’af version of the questionnaire was used to collect the data. The data was collected by five preparatory school graduates. The data collectors and the supervisors (two BSc nurses) were trained for 2 days by the principal investigator on the study instrument, consent form, how to interview and data collection procedure.

In this study the dependent variable was ‘early initiation of breastfeeding practice’ among mothers of children aged less than 24 months. Early initiation of breastfeeding is the initiation of breastfeeding within 1 h of birth [3]. Initiating breastfeeding within 1 h of birth was coded as ‘1’, while initiating breastfeeding after 1 h of birth was coded as ‘0’ for regression analysis. The independent variables were maternal characteristics (age, religion, ethnicity, educational status, marital status, parity), household characteristics (area of residence either rural or urban, household head, family size), paternal educational status, sex of the index child, birth order of the index child, antenatal care utilization, place of delivery, mode of delivery, postnatal care utilization, prelacteal feeding and colostrum avoidance practices.

Preparatory school graduates who can speak the local language were recruited as data collectors. The questionnaire was pretested on 21 mothers (5 % of the sample size) having at least one child aged less than 24 months before the initiation of the study. These mothers were sampled from two kebeles which were not included in the research. The pretest was done to ensure clarity, wordings, logical sequence and skip patterns of the questions. Then the pretest amendments on the questionnaire were made accordingly. The supervisors checked the day to day activities of data collectors regarding the completion of questionnaires, clarity of responses and proper coding of the responses.

The data were checked for completeness and consistencies. Data were also cleaned, coded and entered into EpiData version 3.02, then exported to SPSS version 20 statistical package for analysis.

The study was approved by the Ethical Review Committee (ERC) of Samara University. An official letter was written from Samara University to Amibara District Administration Office. Then permission and support letter was written to each selected kebeles. The participants enrolled in the study were informed about the study objectives, expected outcomes, benefits and the risks associated with it. A signed written consent was taken from the participants before the interview. Illiterate mothers were consented with their thumb print. Confidentiality of responses was maintained throughout the study.

A total of 403 mother-child pairs were included in the study with a response rate of 99 %. About 75 % of the mothers were in the age group of 20–34 years. More than half of mothers (56.8 %) were Muslims, 72.5 % were currently married and 37 (9.2 %) of mothers never married (single). Only 22.8 % of mothers were household heads, and nearly 57 % had five and more family members (Table 1). Regarding maternal and child health (MCH) service utilization, the majority of mothers had attended antenatal care visits (75.9 %), 86.4 % of mothers had spontaneous vaginal delivery and 65 % had attended postnatal care visits (Table 2).

Three hundred eighty one (94.5 %) of the respondents had ever breastfed their index child. Of those who had ever breastfed, 151, 39.6 % (95 % CI 35.0 %, 45.0 %) mothers initiated breastfeeding within 1 h after birth.

Binary logistic regression analysis showed that residence, household headship, maternal educational and marital status, parity, mode of delivery and birth order were statistically associated with early initiation of breastfeeding at p-value < 0.05 (Table 3). However, maternal age, ethnicity, religion, family size, prelacteal feeding and husband educational status were not significantly associated with early initiation of breastfeeding at p-value < 0.3 in the binary model. Thus, these variables were not included in multivariable regression model.

In multivariable logistic regression analysis living in urban places, maternal attendances of formal education, giving birth by caesarean section and birth order remained statistically significant factors associated with early initiation of breastfeeding (Table 3). Mothers living in urban areas were 3.8 times (AOR 3.8; 95 % CI 2.32, 6.06) more likely to initiate breastfeeding within 1 h of birth. Compared to mothers who had no formal education, mothers who attended formal education were two times (AOR 2.0; 95 % CI 1.21, 3.46) more likely to initiate breastfeeding within 1 h of birth. Mothers who had delivered their child by caesarean section (AOR 0.46; 95 % CI 0.22, 0.97) were less likely to initiate breastfeeding within 1 h of birth as compared to those who gave birth vaginally. Mothers with two or three children were less likely (AOR 0.59; 95 % CI 0.35, 0.99) to breastfeed their child within 1 h of birth compared to mothers with four or more children.