Sustainable under nutrition reduction program and dietary diversity among children’s aged 6–23 months, Northwest Ethiopia: Comparative cross-sectional study

A community- based comparative cross-sectional study was conducted in three selected districts of West Gojjam zone from February 29 to April 20, 2019. West Gojjam is one of the administrative zones in Amhara region, North West Ethiopia. It is located 567 km from Addis Ababa, the capital of Ethiopia, and has 16 districts and 444 kebeles. SURE and Save the children programs which were working to strengthen existing efforts in the country covered four of the 16 districts each, that is eight districts. A total of 117,673 mothers who had young children aged 6 to 23 months lived in the SURE covered (Yilimanadennsa) and the uncovered (Bahir-Dar Zuriya and Debub Achffer districts) which largely dependent on agriculture. The zone had 3 hospitals, 104 health centers, and 391 health posts that providing health services including maternal and child health care.

All infants and young children aged 6–23 months and their mothers who had lived for at least 6 months in the area participated in the study. As this was a comparative cross-sectional study, the minimum sample size was determined by using the double population proportion formula with the assumptions exposed (intervention applied) and unexposed (intervention not applied) groups. To estimate the minimum sample size, a dietary diversity proportion (13%) was taken as p2 from a previous study [18]. However, since there has been no previous finding for the intervention group, the assumption that intervention increases the proportion of dietary diversity by 15% p1 yielded 28%. The final sample size was calculated using the Epi Info software with the assumption of a 95% confidence interval, 80% power, 1:1 ratio of exposed to unexposed, 3% design effect, and 10% non-response rate. Therefore, the final minimum adequate sample size was 832. A multistage stratified sampling and the simple random sampling technique was employed to select study participants in West Gojjam zone. Initially, districts were categorized as SURE program covered and uncovered. Three districts, one covered and two uncovered were selected using the lottery method for the study. The three selected districts had a total of 85 kebeles (35 covered and 50 uncovered).Out of the 85kebeles, seven in SURE covered and ten in uncovered districts were selected using the lottery method. Participants were proportionally assigned to each kebele using the community-based demographic and health related information registration book of health extension workers. Finally, mother to child pairs were selected from each keble using the simple random sampling methods after giving codes to each household which had young children aged 6 to 23 months. If there were more than one children in the households, we selected the index child by the lottery method.

Data was collected through a face to face interview, using a structured and pre-tested questionnaire.

In order to maintain the quality of data, 2 days training was given to data collectors and supervisors by the principal investigator. A 5% pretest was conducted in non selected districts, and the questionnaire was initially prepared in English and translated to Amharic and retranslated to English by language and public health experts to guarantee consistency. On-site supervision was performed, and each copy of the questionnaire was checked for completeness and accuracy before data entry; 17 clinical nurses and six BSc graduate nursing or public health field supervisors were involved in the data collection process.

Dietary diversity practice was collected and calculated as the sum of the number of different food groups consumed by the child in the 24 h prior to the assessment. The list of food groups included, grains, roots, and tubers; legumes and nuts; dairy products (milk, yogurt, cheese); flesh foods (meat, fish, poultry and liver/organ meats); eggs; vitamin-A rich fruits and vegetables and other fruits and vegetables. Finally, if respondents consumed four or more of the food groups, they were considered as having adequate dietary diversity [2].

Household wealth index adopted from EDHS 2011 was determined using the Principal Component Analysis (PCA) by considering household assets, such as livestock, type of house, durable assets and productive assets. First, variables coded between 0 and 1 were entered and analyzed using PCA; then variables with commonality values of greater than 0.5 were used to produce factor scores. Finally, the factor scores were summed and ranked as “poor”, “medium” and “rich”.

Food insecurity was measured using the FANTA (Food and Nutrition Technical Assistance Tool) household food insecurity access scale (HFIAS) [21]. It consisted of nine “occurrence questions” that represented a generally increasing level of severity of food insecurity (access) and nine “frequency-of-occurrence” questions that asked as a follow-up to each occurrence question to determine how often the condition occurred. The frequency-of-occurrence question was skipped if respondents reported when the condition described in the corresponding occurrence question was not experienced in the previous 4 weeks (30 days). Finally, individuals were considered as food secure, if they said “no” to all items or just experienced worry but rarely; mildly food insecure households were those who were defined sometimes or often worried about not having enough food and/or unable to eat favorite foods and/ or rarely ate a more monotonous diet than desired. Households that reported they rarely or sometimes ate more monotonous diets than desired sometimes or often and/or had started to cut back on quantity by reducing the size of meals or the number of meals were coded as moderately food insecure.

Data were entered into EPI INFO version 7 and analyzed using the Statistical Package for Social Sciences (SPSS) version 20. Descriptive statistics, including frequencies and proportions were used to summarize the variables. A binary logistic regression model was fitted to identify factors associated with dietary diversity practices. Variables with P–values of < 0.2 in the bi-variable analysis were entered in to the multivariable analysis to control possible effects of confounders. The Adjusted Odds Ratio (AOR) with a 95% of confidence interval was used to examine the strength of associations, and a P– values ≤0.05 was used to declare statistical significance in the multivariable analysis.

A total of 832 mother-child pair participated in the study with a 100% response rate. Out of these 416 (50%) were from SURE program covered district. Over half, 54.6 and 61.5% of the respondents from SURE covered and uncovered areas, respectively, were in the age range of 25–34 years with the mean (±SD) of 28.75(±5.935) years. 248 (59.5%) of the mothers in the SURE covered and 251(60.3%) of uncovered districts were unable to read and write; 228 (54.8%) children aged 6–23 months in the covered and 259(62.3%) in the uncovered districts were second to fifth birth order. Nearly one third of the respondents in both groups had poor family wealth status (Table 1).

Almost all of the participants (92% of SURE covered and 97% of uncovered) had at least one ANC service, while 263(63.2%) of mothers under SURE and 243(58.4%) out of SURE had PNC services; 348 (83.7%) children under and 282 (67.8%) out of SURE took Vitamin A supplementations (Tables 2 and 3).

Two hundred forty (57.57%) of SURE covered and 296(71.2%) of non covered caregivers had good knowledge about dietary diversity and child feeding (Fig. 1).

Of the mothers,92.3% in SURE covered and 91.8% in the uncovered districts had information about child feeding; 88.3% of the mothers in the covered and 91.7% in the uncovered districts obtained the information during their ANC follow ups,while 207(49.8%) and 80(19.2%) in the covered and uncovered districts, respectively, gathered information from HEW/AEW at food demonstration programs. In the covered and uncovered districts, 360(60.5%) and 410(98.6%) of the households were food secure respectively (Fig. 2).

The overall proportion of children who received adequate dietary diversity in the study area was 29.9% at (95% CI: 27–33%). The Pearson chi2-square test illustrated that there was a statistically significant adequate dietary diversity difference in the SURE program covered and uncovered districts (pr = 0.028). Lower proportion of adequate dietary diversity was noted among children in the uncovered districts 26.4% (95%CI: 22–31%) than covered district which was 33.4% (95%CI: 29–38%) (Fig. 3).

The result of the multivariate analysis revealed that SURE coverage of districts, ANC visits, PNC follows ups, GMP, Vit A supplementation, child food demonstration, and HEW/ AEW household visits were found to be statistically and independently associated with adequate dietary diversity in the study area.

Children who were covered by SURE were 2.5times [AOR = 2.55,95%CI:1.63, 3.97] more likely to have adequate dietary diversity compared to those who were not covered, and children who had GMP were 2.7 times [AOR = 2.74,95%CI:1.80, 4.18] more likely to take diversified diets than those who had not. Children whose mothers had PNC follow ups during their last childbirth were 2.1times [AOR = 2.13, 95%CI: 1.38, 3.28] more likely to have adequate dietary diversity than their counterparts. Also, children who received vitamin A supplementation were almost 2 times [AOR = 2.10, 95%CI: 1.22, 3.61] more likely to have diversified diets than those who didn’t receive. Children whose mothers had received food demonstration trainings were 1.9 times [AOR = 1.88, 95% CI: 1.19, 2.96] more likely to have adequate dietary diversity compared to those whose mothers didn’t receive trainings. Mothers whose houses were visited by HE/AEWs were 2 times [AOR = 2.00, 95%CI: 1.25, 3.21] more likely to have adequate dietary diversity than those whose mothers’ houses were not visited (Table 4).

That we compared the impact of SURE by using the program covered districts as intervention and the uncovered ones as control through a comparative study design is perhaps the strength of the work. On the other hand, the fact that the measurements of dietary diversity were prone to recall bias might be the weakness although data collectors did their best to probe mothers to recall the food items they gave to their index child. Besides, the study could not claim to be free from social desirability bias in the measurement of dietary information.