Comparing level of food insecurity between households with and without home gardening practices in Zege, Amhara region, North West Ethiopia: Community based study

Background Globally, close to 1 billion people suffer from hunger and food insecurity. Evidence showed that prevalence of household food insecurity in Ethiopia is ranged from 25.5%-75.8%. Home gardening is one way to alleviate food insecurity. Hence, the study aimed to determine level of food insecurity and its associated factors between home gardening and non-home gardening household in Zegie, North west Ethiopia. Methods Community-based study was conducted from February 10th-March 10th/2020. A total of 648 samples were included. First, 2142 total households who have 6–59 months of age children in the area identified and registered. Then, households categorized in to home garden practicing (1433) and non-home garden practicing (709). The calculated sample size, 324 for each group were selected using simple random sampling technique. Results The overall prevalence of food insecurity was 38.1% (95% CI: 34.29–42.11%). Food insecurity was significantly higher in non-home gardening groups than their counter parts 45.5% (95% CI: 39.80–51.20%). Having primary education and above (AO = 1.89, 95% CI: 1.25–2.86%), wealth index; 2nd quantile (AOR = 0.44, 95% CI: 0.25–0.85%), 3rd quantile (AOR = 0.32, 95% CI: 0.17–0.62%) and 4th quantile (AOR = 0.27, 95% CI: 0.15–0.54%), dietary diversity (AOR = 1.83, 95% CI: 1.15–2.92%) and home garden practices (AOR = 1.57, 95% CI: 1.06–2.32%) were variables significantly associated with food insecurity. Conclusion Food insecurity in non-home garden practicing households is higher than practicing households. The local agriculture sector needs to emphasis and empowered households on home gardening practices to realize food security. Community based comparative cross-sectional study was conducted in Zegie rural satellite town of Bahir Dar city administration, Amhara Regional State, Ethiopia. The study period was from February 10th to March 10th/2020. Zegie rural satellite town is found at a distance of 600 km away from Addis Ababa, capital city of Ethiopia and 32 km from Bahir Dar in the northwest direction of the country. Zegie peninsula is one of the religious area found in the region and households did not have enough farm lands for production of diversified crops [27]. Based on 2019 population projection given from Bahir Dar city administration, the total population size of Zegie was 10,083 (4,041 males and 6,042 females). For administrative purpose Zegie is divided in to three kebeles (the smallest administrative units of the government). There were 2,142 total households in the town and from theses 709 households practised home gardening while the remaining 1,433 were not. All mothers who have 6–59 months of age children in the household were considered as the source population of the study. This happened due to the fact that the study has other objective that determined the nutritional status of children between the two populations. Selected mothers from the two population groups (households who practised home gardening and who did not) were considered as study population. In the home gardening practised population groups those households started their home gardening practice for at least 6 months were included in the study. The sample size of the study was determined using double population proportion formula by considering the following assumptions; 95% confidence level, 80% power of the study, P1 and P2 the prevalence of stunting in home gardening and non-home gardening populations, respectively. Where, n = Sample size for each group Z1 = 1.96 for 95% confidence level, Z2 = 0.84 for 80% power of study P=P1+P22=0.41+0.5252=0.47 P1 = prevalence of stunting in under five children with home gardening practiced households (41%) from previous study [28] and P2 = prevalence of stunting in under five children from households without home gardening practiced (52.5%) [29]. Stunting was considered to estimate the sample size since it is perceived that the nutritional status of the community can be well explained by it than other indices. Other reason for the consideration of stunting for the sample size estimation there was other objective addressed by this study which was to determine the nutritional status of under five children in the study area. Having the above given conditions and 10% non-response rate, the calculated sample size was 648 paired child-mother/care givers. The estimated sample size was checked for its sufficiency by comparing with sample sizes that estimated by considering other factors. First 2142 total households who have 6–59 months of age children in the town were identified and registered at the health post level. This was done to address the previously stated objective, that is to determine under five children nutritional status. Then these households were grouped in to home gardening (1433) and non-home gardening (709). Then, the calculated sample size (648), 324 for each group were selected using simple random sampling (computer generating method) technique. Food security/insecurity Socio-demographic variables (marital status, maternal education, paternal education, family size, household head, occupational status), wealth index. Is defined as households who cultivate at least one kind of fruit and vegetable in their yard or compound. Defined as proportion of households who receive 4 or more food groups from the 7 food groups consumed over 24 hours [30]. Defined as proportion of households who receive 3 or less food groups from the 7 food groups consumed over 24 hours [31] Defined as households experiences none of the food insecurity (access) conditions, or just experience worry, but rarely otherwise food insecure households [13]. Is defend as the physical size of the farm, primarily in terms of hectares of operated land [32]. A woman said to be decision maker if she participated lonely or and jointly 5 and above from 10 decision making related questions [33]. Different types of tools and measurements were implemented to collect the required data. Structured interviewer administered questionnaire was developed by reviewing different literature. The questionnaire was developed in English and translated to the local language (Amharic) and back to English to check its consistency. The questionnaire has sections like socio-demographic, and/or socio-economic characteristics, nutrition related, wash related, health related factors and anthropometric measurements. After households who have 6–59 months of age children selected from health posts, then data collectors went to the house for interview. Four clinical nurses and two health officers were assigned for data collection and supervisory respectively. A 24-hour recall method (from sun rise to sun rise) was used to assess dietary diversity practices. It was based on the mother’s recall of foods given to her child in the previous 24 hours prior to the interview date. Then, minimum dietary diversity was estimated using information collected from the 24 hours dietary recall. Minimum dietary diversity was fulfilled if a child had received four or more food groups from the seven WHO food groups in the last 24 hours preceding the survey. Seven food groups included were grains, roots, and tubers; legumes and nuts; dairy products (milk, yogurt, and cheese); flesh foods (meat, fish, poultry, and liver/organ meats); eggs; vitamin rich fruits and vegetables; and other fruits and vegetables [30]. Household food-security (access) information was collected by using the questionnaire adopted from the Household Food Insecurity Access Scale, which was developed by the Food and Nutrition Technical Assistance project. This instrument consists of nine questions that measure uncertainty on obtaining food, limited access to high-quality foods, and reduction in food quantity in the past 4 weeks. The precoded options were never (0 points), rarely (once or twice in the past 4 weeks; 1 point), sometimes (three to ten times in the past 4 weeks; 2 points), and often (more than ten times in the past 4 weeks; 3 points). Scores for answers to these questions were summed (0–27), and thus a household experiences none of the food insecurity conditions, or just experiences worry, but rarely categorized as food secure otherwise food insecure household [13]. Wealth index of the households was determined using the Principal Component Analysis (PCA). Communality value > 0.5, KMO (sampling adequacy) with P-value > 0.05 and complex structure factor (Eugene value) greater than 1 was considered. Quintiles of the wealth score was created to categorize households as poorest (1st quantile), poor (2nd quantile), medium (3rd quantile), rich (4th quantile) and richest (5th quantile). To maintain the quality of data, first, standardized data collection tools were adopted from published sources and contextualized to the local study area. Pretest was done on 5% of the total sample size (26) other than study sites with similar characteristics. Weighing scale was calibrated before each measurement using known weight and all anthropometric measurements were taken twice, and the average of the two measurements were calculated and recorded. Two days training was given for data collectors and supervisors prior to the actual data collection time on the selection procedure of study participants, purpose of the study, on the steps how they can give the necessary information for the participants when they start data collection. The supervisor and principal investigator were supervised and checked the completeness and quality of data daily. During data collection, questionnaires were reviewed and checked for completeness by the supervisor and principal investigator and the necessary feedback was offered to the data collectors in the next morning. Then the data obtained from the study population were entered, and cleaned for missing value by the investigator. The collected data was coded, entered and cleaned using Epi data version 3.02 and exported to SPSS version 23 for analysis. Descriptive statistics like frequency, percentage and mean were carried out for different variables. The association between two populations was cheeked using chi square test. Bi variable logistic regression analysis was used to know the crude association between each independent variables and outcome variable (stunting and wasting) and crude odds ratio was taken. Then variables which were associated with the dependent variable in bi-variable analysis with p-value < 0.2 were included in the models of multivariable logistic regression analysis with backward likelihood ratio approach. The Hosmer–Lemeshow test was performed for model fitness in the final model, and p-value > 0.05 was considered as a good fit. Anthropometric data were converted in to indices and indicators using WHO Anthro software. Having p-value less than 0.05 in multivariable logistic regression analysis was used to conclude the presence of statistically significant association between different predictor variables with outcome variable (stunting and wasting). The strength of statistically association was measured by adjusted odds ratio at 95% confidence level.