Study design, area and period
The institution-based case-control study design was conducted in the Northwest part of Ethiopia. Ethiopia is one of the fast-growing and landlocked countries in the Horn of Africa, and Addis Ababa is the capital city. Ethiopia is bordered by Eritrea, Djibouti, Somalia, Kenya, South Sudan, and Sudan. The country comprises nine regional and two administrative states. Amhara region is one of the largest regional states in Ethiopia and has several sub administration units. Among these, Dangila woreda (the study area) is one of the administrative subunits and located at 487 km away from Addis Ababa and 78 kilometers from Bahir Dar, the capital city of Amhara Region. The woreda has 31 districts, and two of them were a town. The remaining 29 districts are rural (agricultural community) and have six health centers and 31 health posts, and 72 health professionals (Figure 1). Based on the district's annual report, it has an estimated total population of 155,562 in the year 2017/2018. From these, 77562 and 77873 were males and females, respectively (18).
Source population
All lactating mothers visited the Maternal and child health clinic in all the six health centers of Dangila Woreda, during the data collection period.
Study population for cases
All lactating mothers (BMI<18.5 kg/m2) visited the three selected health centers of Dangila Woreda during the data collection period.
Study population for controls
All lactating mothers (BMI ≥18.5 kg/m2) visited the three selected health centers of Dangila Woreda during the data collection period.
Inclusion criteria for cases
All lactating mothers whose body mass index was <18.5 kg/m2 and gave birth before six weeks prior to the study period and able to give oral consent were included in the study.
Inclusion criteria for controls
All lactating mothers whose body mass index was ≥18.5 kg/m2 & gave birth before six weeks prior to the study period and able to give oral consent were included in the study. The controls were unmatched.
Exclusion criteria for cases
All lactating mothers whose body mass index was ≥18.5 kg/m2 & gave birth before six weeks before the study period, and gave oral consent were included in the study.
Exclusion criteria for controls
Those lactating mothers whose body mass index was ≥ 18.5 kg/m2 and had a critical illness, hearing impairment, and physical deformities during the anthropometric measurements were excluded from the study. Lactating mothers whose children age less than six weeks and those lactating mothers whose child's age from 6 weeks to 2 years but non-lactating due to formula feeding were also excluded from the study.
Sample size determination
The sample size was estimated using Epi-info-7 software-based on the assumption of 95% confidence interval, 80% power, control to case ratio of 3:1, frequency of family size in controls 26.02%, adjusted odds ratio of 0.46, and non-response rate of 5% yielding 300 controls and 100 cases (400 study participants).
Sampling procedure
In the study area, there are six health centers. First, the three health centers were selected from the existing six health centers by lottery method. The three selected health centers were called Chara, Abadra, and Gumdry. Secondly, the average monthly number of lactating mothers visiting the Mother and Child Health Clinic & Expanded Programme on Immunization in each selected health centers was calculated. The average number of lactating mothers visiting the Mother and Child Health Clinic & Expanded Programme on Immunization clinic of Chara, Abadra, and Gumdry health centers monthly were 212,186 & 145 respectively. Thirdly, the sample size was allocated proportionally to each of the three selected health centers based on the number of lactating mothers visiting the Mother and Child Health Clinic & Expanded Programme on Immunization during the data collection period. The number of lactating mothers interviewed from Chara, Abadra, and Gumdry health centers was 151, 134, and 104, respectively (Figure 2). Lastly, to select the study participant, every case and every control consecutively were selected. All lactating mothers whose Childs age from 6 weeks to 2 years old who were visiting the Mother and Child Health Clinics during the data collection period were screened using their anthropometric data. Underweight mothers were first identified, and then cases were selected based on the inclusion and exclusion criteria. Cases were lactating mothers diagnosed with underweight (BMI<18.5kg/m2), which can be moderate to severe regardless of the types of malnutrition. As the case was identified, three controls were selected consecutively. Controls were lactating mothers without malnutrition (have anthropometric readings of BMI ≥18.5 kg/m2). The data collectors selected controls and cases from the same Primary Health Care Unit. After measurements, mothers were interviewed based on a validated questionnaire.
Data collection procedure and instrument
In collaboration with expert opinions in the field, the investigators developed the study questionnaire, and the questions were selected based on similar studies in Ethiopia with slight modification (9–11)
The data collector group had four members (three diploma nurses assigned for data collection and one BSC nurse as a field supervisor). They took appropriate training about the data collection instrument. The data collectors were following all the recommended consent forms and steps in the questionnaire. The data were collected through interviewer-administered questionnaires to collect information on the socio-demographic data of lactating mothers, family size, and socioeconomic characteristics like occupation, monthly income, and dietary diversity.
The anthropometric measurement tools were calibrated daily before the measurement by the principal investigator. Weight was taken using a battery-powered digital scale, and the reading was taken to the nearest 0.1kg. Wooden height board was used to measure the height of lactating mothers, and the respondents asked to erect on barefoot while measurement was taken, the reading measurement was taken to the nearest 0.1 cm. Then body mass index Index was computed using the following formula (BMI= weight in kilogram divided by height in meter square for each respondent). All measurement was taken using standard procedures. Those lactating mothers whose BMI<18.5 kg/m2 were counseled on the recommended feeding practices during lactation. The household food insecurity scale established by the Food and Nutrition Technical Assistance was employed to measure the household food security status of the study participants (19). To ensure the accuracy of the data collection, the entire questionnaire was translated into the local Amharic language and translated back into the English language during the analysis.
Study variables
The presence of underweight was the outcome variable of the study, which could be influenced by the independent variables. Age of the child, age of the mother, educational status of the mother, parity, marital status, limited dietary intake, frequency of antenatal care visits, dietary intake, nutritional knowledge, dietary diversity, household food insecurity, place of delivery, family size, age at first pregnancy were the main independent variables of the study.
Data quality control
The clarity, length of time required, content and wording of the questions of the translated questionnaire was checked by data collectors who can read and write the local language as well as by conducting a pretest in the nearby of the study area where the population lives and share similar language and lifestyle. The reliability, validity, consistency, and accuracy of the measurement was assessed and checked by the technical error of measurement. The data collectors were given appropriate training, and the principal investigator was continuously supervising the data collection process to ensure the accuracy and completeness of the information. A pretest of the data collection instrument was conducted in 5% of the sample size in another woreda. Before applying in the main study, all the relevant modifications were considered.
Data analysis
After the completion of the data collection process, all the questionnaires were checked for completeness, clarity, and consistency accordingly, and the data were entered by using EPI-Data 3.02 version statistical software and analyzed by using statistical package for social science version 23 software. Descriptive statistics were computed to calculate the underweight predictors among lactating mothers, and it was analyzed as per educational status, occupational status, and other variables. A bivariate logistic regression was performed, and the variables with p-value <0.2 in the bivariate logistic regression were entered to multivariate logistic regression to control the effect of confounding variables. The result was presented as proportion, the output of logistic regression as an adjusted odds ratio with 95% confidence intervals. P-value < 0.05 was considered as statistically significant.
Operational definition of terms
Dietary diversity: Those lactating mothers who eat 6 and above food groups are considered to be diversified and those lactating mothers who eat below 6 food groups are considered to be not diversified.
Adequate Knowledge: If respondents score ≥ 70% on the knowledge questions (20).
Inadequate knowledge: If respondents score < 70% on the knowledge questions (20).
Lactating mothers: Women who are currently feeding breast milk for their infant/child.
Underweight: When the computed body mass index of lactating mothers < 18.5kg/m2.