A community-based cross-sectional study was employed from December 17, 2021, to January 16, 2022, at Waliso town in the Southwest Shewa zone, Oromia region, Central Ethiopia, which is 114 km southwest of Addis Ababa. Waliso town is located at 8°32° N and 37°58° E and is 2063 m above sea level. Its climate is Wiena Dega (temperate), and it rains three times a year. The main types of crops produced in the district surrounding the town are teff, wheat, maize, and inset. From the projection of the 2007 CSA report, Waliso town has 67,496 inhabitants, of which 33,643 were males and 33,853 were females in 2021, which is projected from the Ethiopian census conducted in 2007, and 14,673 households.

The town is classified into four administrative kebeles (lowest administrative units). The majority of Waliso town residents found food for their families by purchasing it from the market. While the majority of households are unemployed, wage laborers and self-employed individuals earn less to cope with the current high inflation in the country.

Source population

All households in Waliso Town.

Study population

All randomly selected households in Waliso town and household members aged > 18 years who were responsible for food purchase and preparation, preferably women from households in selected kebeles of Waliso town, were interviewed.

Sample size determination

The sample size was determined by using the single population proportion formula.

$$\mathrm=\frac/2\right)}^\mathrm\left(1-\mathrm\right)}}^}=\frac^0.376\left(0.624\right)}=361$$

Assumptions

P = 37.6% (prevalence of household food insecurity among households in Woliata Sodo town) [23], Z/2 = critical value at 95% confidence level of certainty (1.96), d = margin of error (5%). Adding a 10% nonresponse rate, the final sample size (n) was 397.

The sample size for the second objective: The sample size for the second objective was calculated using EpiInfo7 software and by using three significantly associated variable percentages (P) and their respective adjusted odds ratios. Assuming power is 80% and Z/2 95% CI, then a 10% no response rate is added. Finally, the sample sizes found were sample size one (S1) = 346; sample size two (S2) = 382; and sample size three (S3) = 392, but all three sample sizes calculated for the second objective were less than the sample size calculated for the first objective, i.e., 397. Therefore, the sample size calculated for the first objective was also used for the second objective because it was the largest.

Sampling procedure and technique

All households that lived in Waliso town for more than six months were included in this study. There is a registration book containing the list of all households in each kebele, which is updated every six months and used for urban health extension program (family health) service provision based on their health needs, and the study participants for this study were selected by simple random sampling from this family health register in all kebeles in Waliso town. A computer program was used to select the needed sample population, and proportional allocation was employed for each kebele. After selecting the required sample by simple random sampling from the register, the data were collected house to house (Fig. 1).

Fig. 1

Sampling procedure used to assess the prevalence of household food insecurity and associated factors among residents of Waliso town, Oromia, Ethiopia, 2021/2022. Waliso town has four kebeles (the lowest administrative units), i.e. Ayetu, Burka, Ejersa, and Hora

Data collection tools and techniques

A structured, interviewer-administered questionnaire translated into the regional language “Afaan Oromoo” and translated back to English by a language expert was adapted from the Household Food Insecurity Access Scale (HFIAS) version 3 from Food and Nutrition Technical Assistance (FANTA) and has nine occurrence questions that represent a generally increasing level of food insecurity (access) and has three domains. The first question is related to anxiety and uncertainty about household food supplies. The next three questions are about insufficient quality (variety and preferences of the types of food). The remaining five are questions about access to insufficient food intake and its physical consequences, and nine “frequency-of-occurrence” questions are asked as a follow-up to each occurrence question to determine how often the condition occurs [24].

Sociodemographic and wealth index questions were adapted from the Ethiopian Demographic and Health Survey. The household wealth index (HWI) questionnaire was constructed from household asset ownership questions that were adapted from the 2016 EDHS (Ethiopian Demographic and Health Survey) questionnaire [25] and categorized into three wealth categories: low wealth index (poor), medium wealth, and high wealth index (rich).

The data collection was performed by trained urban health extension professionals; two BSc health science professionals supervised the data collection activities, and the investigators monitored the overall data collection activities and procedures. The data collectors and supervisors were trained for two days on study variables, data collection, and interview techniques to minimize interviewer bias. The questionnaire was pretested on 5% of the sample (20 households) in the adjacent town of Waliso in Obi town. Then, the questionnaire was modified based on the pretest results; repetitive ideas and ambiguous questionnaires were corrected, and the modified questionnaire was used for the final data collection. Even if data collection overlapped with the festive period (Christmas), we interrupted the data collection process for five days during the Christmas celebration, and all dietary histories linked with the festive were excluded during data collection.

Data processing and analysis

After data collection, the questionnaire was checked for completeness and consistency. The data entry template was prepared and double-entered using Epi Info 7. The data were cleaned by limiting the values that would be entered, providing skipping patterns, and, after data entry, by using simple frequency, tabulating variable frequency, and sorting ascending and descending.

Then, the data were exported into SPSS version 20 for analysis. The household wealth index was determined by principal component analysis using asset ownership questions, easy-to-collect data on a household's ownership of selected assets, materials used for housing construction, and types of water access and sanitation facilities and categorized into poor, medium, and rich. First, univariable analysis was conducted using descriptive summaries such as frequencies, percentages, mean, standard deviations, and prevalence. Bivariable analyses were performed to identify candidate variables for a multivariable logistic regression model. Normality was checked for continuous variables using a histogram and Q–Q plot, and HH monthly income and HH monthly expenditure were not normally distributed, while HH age was normally distributed. Model fitness was checked using the Hosmer and Lemeshow test, and it was not significant (p > 0.05) except for the model being fit.

All explanatory variables that were associated with the outcome variable at a P value of 0.25 were considered candidates for multivariable logistic regression. Multicollinearity between different explanatory variables was checked using tolerance (tolerance 0.2) and the variable inflation factor (VIF > 5) [26]. Two independent variables, i.e., monthly household income (tolerance = 0.13, VIF = 7.53) and monthly household food expenditure (tolerance = 0.13, VIF = 7.56), were excluded from multivariable analysis with multicollinearity.

The association between the outcome variable and predictors was determined, and P < 0.05 was used to declare statistical significance. Adjusted odds ratios and 95% confidence intervals were used to measure the strength of the association.

Study variablesDependent variable

Household food security status: secure, insecure.

Independent variables

Household food sources and access: self-production, market purchase, donation, or transfer of money or food.

Sociodemographic Factors and Economic Status: Age, sex, marital status, occupation, educational status of the household head, family size and the number of dependents in the household, household monthly income, household monthly food expenditure, and household wealth index.

Alcohol consumption and substance use: alcohol consumption, chewing chat.

Operational definitions

HFIAS Score: Replies on the nine HFIAS questions will have a minimum of zero and a maximum score of 27 (0–27), and HFIAS scores 0–1 were considered food secure, while those with HFIAS scores 2–27 were considered food insecure. Households’ food insecurity was also categorized according to their score, with HFIAS scores of 2–7, 8–14, and 15–27 indicating mild, moderate, and severe food insecurity, respectively [27]. Accordingly,

Food secure: HFIAS scores 0–1 were considered food secure.

Food insecure: HFIAS scores 2–27 were considered food insecure.