Factors Associated With Nutritional Status Of Infants And Young ...

Study setting and design

This community-based cross-sectional study was conducted in Filtu town, which is the capital town of Liben Zone of the Somali Region. The total population in Filtu woreda was 130, 912, of which 4960 reside in filtu town [14]. The climate in Filtu is arid and the rainfall pattern is bimodal with low annual rainfall [15]. Most residents are dependent on animal products (their own or from the market) for consumption. Likewise, most of their income depends on animals, either selling their products or depending on them as a means of income generation like fetching water. A few members of the community were merchants, selling materials from the border of the country. As a whole, the communities have ration distribution from the government through the “work for food” program within their kebeles (smallest administrative unit).

Sample size and sampling procedure

The sample size was calculated based on a single population proportion formula. Prevalence of malnutrition in the specific age group (6–23 months) of Somali region was used to calculate the sample size. The prevalence of underweight, stunting and wasting were 24.2, 22.3 and 17.8 %, respectively, reported by the 2011 EDHS for the Somali region [13]. The confidence level 95 % (a = 0.05) and margin of error (d = 0.05) were considered. The largest sample size was taken after sample size was calculated based on the three indicators of under nutrition. A sample size of 276, calculated based on the prevalence of underweight, was found to be the largest and was taken as a sample size for this study. Since the source population was less than 10,000, a finite population correction formula was used and 5 % non-response rate was added to get the final sample size. The total numbers of eligible study subjects in the town were 781 infants and young children.

So considering, N = 781, no = 276

Finite population correction formula \( n=\frac{no}{1+\frac{no}{N}} \)

where n0 is the total number of the target population (source population). Therefore, after adjusting the initial calculated sample size and adding 5 % of non- response rate, the final sample size was 214.

There are three kebeles in the study area and infants and young children from all the three kebeles were included in the study. All households with infants and young children aged between 6–23 months were registered, and the number for the sample population was allocated according to size of the kebeles. Simple random sampling was employed to select households within each kebele. In those households having more than one infant or young child between 6 and23 months, the index child was selected by the lottery method.

Data collection procedures

A structured questionnaire adapted from the Ethiopian Demographic and Health Survey (EDHS) was used to collect socio-demographic and other relevant child and mother related information. Besides to the EDHS questionnaire, we included additional questions on the questionnaire for this study based on the study objective. Feeding practices were assessed using a qualitative 24-h dietary recall method. The three data collectors were certified at the diploma level in nursing and spoke the local language fluently. The data collectors were trained on data collection techniques for two days including practical work. Data collectors interviewed each mother individually using the Somalifa language version of the questionnaire.

Measurements

Anthropometric measurements (weight and length) were taken for all children by the principal investigator with an assistant. Standard anthropometric measurement procedures were used as outlined in the measurement guide prepared by the Food and Nutrition Technical Assistance (FANTA) project [16].

The family’s food security status was measured with the household hunger scale and categorized as food secure or food insecure. The household hunger scale is most appropriate to use in areas of substantial food insecurity. Households that scored as “no hunger” to little hunger in the household (score of 0–1) were classified as food secure and moderate and severe hunger in the household were categorized as food insecure (score of 2–6) [17]. Meal frequency and dietary diversity were assessed by 24 h recall. Minimum meal frequency was fulfilled if food was received 2 to 3 times per day at 6 to 8 months of age, 3–4 times per day at age 9–11 months and 3–4 times at age 12 to 24 months, with additional nutritious snacks offered 1–2 times per day between meals in the last 24 h. Minimum dietary diversity was fulfilled if a child had received foods from 4 or more food groups from the seven WHO food groups in the last 24 hrs [18].

Bottle feeding practices were measured by a 24-hour recall as recommended by WHO [18] and the questions was asked as “Did [Child Name] drink anything from a bottle with a nipple yesterday during the day or night time?” Similarly breastfeeding practice was assessed by asking the mother to recall whether her child had breastfed in the last 24 h or not and it was asked as “Did [Child Name] breastfeed yesterday during the day or night time? Wealth index was constructed using household asset data via a principal components analysis to categorize individuals into wealth tertiles (low, medium, high).

Pre test

The questionnaire was pre-tested to assess for clarity, understandability and completeness in communities which have similar geographic setting and socio demographic profile as the study area. During the pre-test, problems on the order, response options and difficult sentence constructions were identified. Based on the findings of the pre-test, rearrangement of sequence and change of wording of questions was made as needed.

Data analysis

Anthropometric data were standardized for age using WHO Anthro v. 3.2.2. The data analysis was performed using SPSS version 20. Descriptive statistics (mean and standard deviation) were calculated for continuous variables. The nutritional status indicators, weight-for-length (WLZ), length-for-age (LAZ) and weight-for-age (WAZ) were compared with reference data from World Health Organization standards [3]. Children below-2 standard deviations (−2SD) of the WHO median for WLZ, LAZ, and WAZ were considered wasted, stunted or underweight respectively.

Univariable analysis (bivariate logistic regression) was carried out between the predictor and outcome variables. Using significant variables at p value 0.05 from the binary logistic regression models, a multivariable logistic regression model was fitted to identify the independent predictors of nutritional status (measured as wasting, underweight and stunting). The strength of association was measured by odds ratios with 95 % confidence intervals. Both the Crude (COR) and Adjusted Odds Ratios (AOR) are reported. Variables with p < 0.05 in the multivariable logistic regression model were considered as associated factors.

Ethical consideration

Ethical clearance was obtained from Hawassa University Institutional Review Board. After explaining the study procedures for the study subjects, verbal consents were obtained from each mother/caregiver.