Consumption of vitamin A-rich foods among 6–23-month-old children in Jigjiga town: a community-based cross-sectional study

Abstract

Background

Vitamin A is a fat-soluble vitamin that has several vital functions in the body. Its deficiency causes several eye disorders. The Somali region of Ethiopia has the lowest vitamin A consumption among children under two years, yet few studies have explored related factors.

Methods

A community-based cross-sectional study was conducted among 522 households in Jijiga town. A multi-stage sampling method was used, and data was analyzed using bivariable and multivariable logistic regression.

Results

Overall, 49.8% (95% CI: 45.3–54.3) of children aged 6–23 months consumed VARF in the 24 hours preceding the survey. VARF consumption was higher among children aged 12–17 months (AOR=2.71; 95% CI: 1.49–4.95) and 18–23 months (AOR=2.02; 95% CI: 1.01–4.03), mothers with secondary or higher education (AOR=2.28–3.48), those exposed to media (AOR=2.98; 95% CI: 1.80–4.92), households with food security (AOR=2.36; 95% CI: 1.30–4.30), and Protestant families (AOR = 6.28; 95% CI: 1.50–26.95).

Conclusions

About half of the children had good intake of vitamin A-rich foods. Educating mothers, promoting nutrition through media, ensuring food security, and promoting age-appropriate feeding practices are key to increasing vitamin A intake.

Introduction

Children aged 6–23 mo belong to a critical life stage when exponential linear growth and cognitive development occur, demanding higher amounts of proteins, carbohydrates, lipids and essential micronutrients compared with early infancy. Parents meet such nutritional demands of their young child mainly through complementary feeding along with continuation of breast feeding until the child is aged ≥2 y.¹ However, different socioeconomic factors such as the high cost of a nutrient-adequate diet, especially in low- and middle-income countries (LMICs), force families to feed their children foods lacking a satisfactory amount of nutrients, such as vitamin A.^2,3

Vitamin A, one of the fat-soluble vitamins, has wide-ranging functions in the body, including triggering signal transmission of neuronal cells in the eye, maintaining differentiation of corneal and conjunctival membranes and defending the body through growth, differentiation and activation of white blood cells.⁴ Depending on the type of vitamin A-rich foods (VARF) they eat, people obtain either preformed vitamin A, from animal sources such as eggs and meat, or pro-vitamin A carotenoids from dark green leafy vegetables and yellow fruits like mangoes and papayas. Pro-vitamin A carotenoids are precursor molecules that need bioconversion into active forms of vitamin A within the intestinal absorptive cell.⁴ Even although they tend to be less biologically available, they are more affordable than animal products. For this reason, they provide most of the vitamin A activity in the diets of economically deprived populations.⁵

For children aged 6–23 mo, the recommended daily safe intake level is 400 µg Retinol Equivalent (RE).⁵ Failing to meet this demand is associated with various health problems, notably eye diseases ranging from night blindness to complete vision loss due to corneal scarring. Vitamin A deficiency (VAD) is the leading cause of preventable blindness in under-five children.⁶

Globally, the consumption of VARF among young children showed only a minimal increase in 2018 compared with 2010, and most progress was seen in upper- and middle-income countries.⁷ In LMICs, just over one-half (55%) of children aged 6–23 mo consume VARF⁸ and, in sub-Saharan Africa, less than one-half (42%) have access to high-quality sources such as animal products.⁹ In Ethiopia, studies have shown inadequate consumption of VARF among children in this age group, with regions like Somali showing particularly high deficiency levels.¹⁰

VAD affects around 190 million preschool children worldwide, mostly in Southeast Asia and Africa.¹¹ In Ethiopia, although clinical forms of VAD, such as night blindness, have decreased, subclinical (serum retinol<0.70 µmol/L) remains a severe public health concern.¹² This deficiency, along with other micronutrient deficiencies, contributes to poor physical and cognitive development in 40–60% of children aged 6–23 mo in LMICs.¹³

Several developing countries, including Ethiopia, have introduced nutrition interventions like biofortification and dietary diversification, which have improved the outcomes for vitamin A levels in young children.^14,15 Ethiopia also implements the Productive Safety Net Program (PSNP) that supports food-insecure households by providing food and cash-for-work opportunities. However, it currently reaches only 9% of urban households.¹⁰

According to the 2019 Ethiopian Demographic and Health Survey (EDHS), Somali has the lowest vitamin A consumption among under-two children in Ethiopia.¹⁰ Moreover, spatial mapping shows the Somali Region as low vitamin A intake clusters in Ethiopia, including the Fafan zone, where Jijiga is located. Despite these alarmingly low rates, there is a notably limited number of localized studies examining the socioeconomic and demographic factors influencing VARF consumption in Jijgjiga town, the capital of Somali Region.

Therefore, our study will fill this critical knowledge gap by assessing both the prevalence of VARF consumption and the associated sociodemographic factors among children aged 6–23 mo in Jigjiga town, where the population is predominantly pastoralist, cross-border trade shapes food availability and access to diversified diets remains limited. Understanding these local determinants is essential to designing context-specific public health interventions and nutrition programs tailored to Somali Region.

Methodology

Study design, setting and period

A community-based cross-sectional study was conducted during 3–21 February 2024, in Jijiga town. Jijiga is an administrative town of the Somali regional state, found within the Fafan zone, 628 km east of Addis Ababa, Ethiopia. The projected 2022 mid-year population of the town was 197 966, of whom 102 529 (51.8%) are male and 95 437 (48.2%) are female (Source: Central Statistical Agency (CSA, 2016)). Children aged 6–23 mo account for about 2.2% of the town’s population. Currently, there are four subcities (Karamarda, Dudahidi, Garab’asse and Qordere) and 20 kebeles (the smallest administrative unit in Ethiopia) under Jigjiga city council (Source: Regional Health Bureau).

Sample size determination and sampling technique

The single population proportion formula was used to calculate the sample size by considering the prevalence of vitamin A-rich food consumption as 28.8% from the previous study conducted in Kachabira, southern Ethiopia,¹⁶ at a 95% CI and 5% degree of precision. Assuming a design effect of 1.5 and a non-response rate of 10%, the total sample size for the first objective was determined to be 522.

Out of 20 kebeles under Jijiga city administration, eight kebeles were selected using a stratified multistage random sampling method. For this purpose, each kebele was grouped under the subcity it resides to form four strata, namely, Karamarda, Dudahidi, Qordere and Garab’asse. Then three kebeles from Karamarda (the largest subcity in Jijiga), two kebeles each from Dudahidi and Qordere and one kebele from Garab’asse were selected randomly. The estimated population size of each of the selected kebeles was obtained from the city administration to calculate the sampling population of each kebele using the conversion factor currently used by Somali Regional State for different planning purposes. Finally, the total sample size of the study was distributed to selected kebeles based on the proportion each kebele accounts for in the total sampling population. Study units were identified by using a systematic random sampling method. Whenever two or more eligible children were found in a single household during data collection, one was selected randomly.

Study population and eligibility criteria

The study population included all children aged 6–23 mo who were living with their parents or caregivers in the selected kebeles of Jijiga town and were available during the study period. Children aged 6–23 mo who lived in the selected households were eligible to participate.

Data collection tools and procedure

Data were collected using a structured, pretested, interviewer-administered questionnaire developed by reviewing different studies in the literature.^17–20 The questionnaire was initially prepared in English then translated into local languages (Af-Somali and Amharic). To ensure consistency, it was back-translated into English by two separate individuals who are fluent in the two local languages.

Eight data collectors with a minimum of a diploma certificate in health-related fields, and proficiency in at least two of the local languages (Af-Somali, Amharic and Afan Oromo), and one supervisor with a BSc degree in Public Health, were assigned for the data collection and monitoring processes, respectively.

A brief introductory orientation about the purpose of the study was provided to the participants by the data collectors. Explanations were given on the importance of their involvement. Then mothers/primary caregivers who were volunteers were interviewed using structured and pretested questionnaires. Study participants were contacted in their homes, either in the morning or afternoon.

Operational definitions

Consumption of VARF

The consumption of VARF was measured by consumption of the seven food groups within the preceding 24 h. The food groups were (i) eggs; (ii) meat (beef, pork, lamb, chicken); (iii) pumpkin, carrots and squash; (iv) any dark green leafy vegetables; (v) mangoes, papayas and others with vitamin A-rich fruits; (vi) liver, heart and other organs; and (vii) fish or shellfish. Accordingly, when the respondent reported that the child had eaten at least one of these, it was considered ‘yes’; otherwise it was considered as ‘no’ VARF consumption.

Minimum meal frequency

Minimum meal frequency was defined as two and three feedings of solid, semi-solid or soft foods for breastfed infants aged 6–8 mo and children aged 9–23 mo, respectively; and four feedings of solid, semi-solid or soft foods or milk feeds for non-breastfed children aged 6–23 mo, whereby at least one of the four feeds must be a solid, semi-solid or soft feed.¹

Household

A household was defined as individuals who sleep under the same roof and take meals together at least 4 d a week.²¹

Media exposure

Media exposure was assessed as ‘yes’ if the participant had access to at least one of three media sources (i.e. newspaper, radio or television) at least once a week, otherwise ‘no’ if they had no access to any type of media.²⁰

Household food insecurity

Household food insecurity was assessed using the Household Food Insecurity Access Scale (HFIAS), which is adapted from the United States Agency for International Development’s food insecurity assessment tool. This tool measures the level of food insecurity experienced by a household during the past 4 wk (30 d). Each household was asked nine occurrence questions related to food insecurity, and for each ‘YES’ response, a corresponding frequency question was administered.²¹

Data processing and analysis

After data collection, each questionnaire was checked sequentially for completion of the data. The data were coded and double-entered into Epi Data version 3.1 (Epi Data Association, Odense, Denmark), then exported into SPSS version 27.0 (IBM Corp., Armonk, NY, USA) for further analysis. Descriptive statistics were summarized using frequencies, percentages and summary measures.

Bivariable logistic regression analysis was performed to select the variables to be entered into the final multivariable logistic regression model. Explanatory variables with p<0.20 in bivariable logistic regression analysis were included in the multivariable logistic regression analysis model. Before performing the multivariable analysis, multicollinearity among the explanatory variables was checked using the variance inflation factor (VIF) and variables with VIF>2.5 were dropped from the final model. In addition, model goodness-of-fit was checked using the Hosmer–Lemeshow test. Both crude and adjusted ORs, with 95% CI, were estimated to evaluate the presence, strength and direction of associations. Every variable with p<0.05 in the multivariable logistic model was considered as having a statistically significant association.

Results

Demographic and socioeconomic characteristics

The response rate for this study was 94.6%. The mean (±SD) age of the older infants and young children who were enrolled was 14 (±5.3) mo, with 36% of them falling into the 6–11 mo age range. The male-to-female ratio was 1.1. The mothers had a mean (±SD) age of 28.2 (±6.4) y, with more than three-quarters (77.2%) between 20 and 34 y. The majority of the mothers (57.7%) were Somali, and 71.9% reported being Muslim. Most mothers (79.1%) reported being housewives and nearly one-third (31.4%) were unable to read and write. According to the HFIAS, 60.5% of households were food insecure (Table 1).

Table 1.

Demographic and socioeconomic characteristics of children aged 6–23 mo in Jijiga town, Eastern Ethiopia, 2024

Variables (n=494)		Frequency	%
Child’s age (mo)	6–11	176	35.6
	12–17	160	32.4
	18–23	158	32.0
Child’s gender	Male	259	52.4
	Female	235	47.6
Maternal age (y)	<20	20	4.1
	20–34	380	77.2
	35–49	92	18.7
Marital status	Single	13	2.6%
	Married	448	90.7
	Divorced	28	5.7
	Widowed	5	1.0
Ethnicity	Somali	285	57.7
	Amhara	79	16.0
	Oromo	61	12.3
	Other	69	14.0
Religion	Muslim	355	71.9
	Orthodox	104	21.1
	Protestant	29	5.9
	Catholic	6	1.2
Maternal education	Unable to read and write	155	31.4
	Able to write and read	8	1.6
	Primary school	178	36.0
	Secondary school	95	19.2
	College and above	58	11.7
Maternal occupation	Housewife	391	79.1
	Government employee	25	5.1
	Private employee	9	1.8
	Private business owner	69	14.0
Family size	≤5	235	47.9
	>5	256	52.1
Income	<10 000	197	41.2
	10 000–20 000	192	40.2
	>20 000	89	18.6
HFIAS score	Food secure	195	39.5
	Mild food insecurity	69	14.0
	Moderate food insecurity	127	25.7
	Severe food insecurity	103	20.9
Maternal media exposure	No	232	47.0
	Yes	262	53.0

HFIAS: Household Food Insecurity Access Scale.

Obstetrics, morbidity and healthcare-related characteristics

Among the participants, 69.2% of mothers had at least four antenatal care visits. However, nearly two-thirds (64.6%) of them did not have any postnatal care follow-up appointments at all. About three-quarters (74.7%) of the mothers had between one and four children. In terms of child health, 90.7% of children were vaccinated. Additionally, 46% of the children experienced illness in the 2 wk preceding the survey, with fever (14.2%) being the most commonly reported illness, followed by cough (12.1%).

Child feeding characteristics

Of the 494 children who were involved in this study, 91.5% began complementary feeding and 61.9% are currently breast fed. Of those who had already started complementary feeding, only two-thirds (67.5%) followed the recommendation to start at 6 mo of age (Figure 1).

Figure 1.

Proportion of children who met the minimum meal frequency as per the WHO standard among those who had already started complementary feeding classified by age and breast-feeding status, 2024. BF: breast fed; NBF: non-breast fed.

Consumption of VARF

As reported by their mothers or caregivers using the 24-h dietary recall method, 49.8% of children aged 6–23 mo consumed VARF (95% CI 45.3 to 54.3%). The most common type was vitamin A-rich fruits (27.3%, 95% CI 23.4 to 31.5%) and the least common was organ meats (2.2%, 95% CI 1.1 to 3.9%) (Table 2).

Table 2.

Proportion of 6–23-mo-old children who consumed different vitamin A-rich foods in the last 24 h preceding the survey in the study area, 2024

Food item	Frequency	%
Eggs	108	21.9
Meat (beef, pork, lamb, chicken)	80	16.2
Pumpkin, carrots and squash	124	25.1
Any dark green leafy vegetables	70	14.2
Mangoes, papayas and other vitamin A-rich fruits	135	27.3
Liver, heart and other organs	11	2.2
Fish or shellfish	15	3.0

Factors associated with consumption of VARF

Socioeconomic, demographic, obstetric and child health-related variables with p<0.2 in the bivariable logistic regression were entered into a multivariable logistic regression model to determine which factors are significantly associated (i.e. p< 0.05) with the consumption of VARF among children aged 6–23 mo. The study found that maternal religion, educational status, media exposure, household food security and the age of the child were significantly associated with the consumption of VARF.

Children from food-secure households were 2.4 (adjusted OR [AOR]=2.36; 95% CI 1.30 to 4.30) times more likely to consume VARF compared with children from food-insecure households. The odds of consuming VARF were 2.3 times higher among children whose mothers had secondary education (AOR=2.28; 95% CI 1.11 to 4.70) and 3.5 times higher among those whose mothers had education above secondary education (AOR=3.48; 95% CI 1.24 to 9.83) compared with children whose mothers were illiterate. Children whose mothers had access to the media had threefold higher odds of consuming VARF (AOR=2.98; 95% CI 1.80 to 4.92) than those children whose mothers did not have any access to media. Children aged 12–17 mo were 2.7 times more likely (AOR=2.71; 95% CI 1.49 to 4.95) and those aged 18–23 mo were two times more likely (AOR=2.02; 95% CI 1.01 to 4.03) to consume VARF compared with those in the 6–11 mo age group. Additionally, children from Protestant families were 6.4 (AOR=6.28; 95% CI 1.50 to 26.95) times more likely to consume VARF than their counterparts (Table 3).

Table 3.

Bivariable and multivariable logistic regression analysis of factors associated with the consumption of vitamin A-rich foods among children aged 6–23 mo in Jijiga town, eastern Ethiopia, 2024

Variable (n=494)	Category	COR (95% CI)	AOR (95% CI)	p
Child’s age (mo)	6–11	1	1
	12–17	2.83 (1.82–4.41)	2.71 (1.49–4.95)	0.001
	18–23	2.25 (1.45–3.50)	2.02 (1.01–4.03)	0.047
Ethnicity	Somali	1	1
	Amhara	4.13 (2.38–7.19)	2.32 (0.76–7.07)
	Oromo	1.66 (0.95–2.89)	2.07 (0.84–5.08)
	Other	2.33 (1.36–3.99)	1.67 (0.67–4.19)
Religion	Muslim	1	1
	Orthodox	2.96 (1.87–4.73)	0.95 (0.38–2.38)
	Protestant	5.30 (2.11–13.3)	6.36 (1.50–26.95)	0.012
	Catholic	1.38 (0.28–6.94)	0.74 (0.08–6.61)
Educational status	Unable to write and read	1	1
	Able to write and read	2.44 (0.59–10.2)	1.99 (0.34–11.6)
	Primary education (1–8th grade)	2.56 (1.62–4.02)	1.66 (0.92–3.00)
	Secondary education (9–12th grade)	5.05 (2.91–8.76)	2.28 (1.11–4.70)	0.025
	College and above	6.42 (3.28–12.5)	3.48 (1.24–9.83)	0.018
Occupation	Housewife	1	1
	Government employee	1.67 (0.73–3.81)	0.52 (0.15–1.78)
	Private employee	2.23 (0.55–9.03)	1.00 (0.17–5.76)
	Private business owner	1.54 (0.92–2.58)	1.19 (0.59–2.42)
Family size	<5	1	1
	≥5	0.63 (0.44–0.89)	0.88 (0.48–1.63)
Income (ETB)	<10 000	1	1
	10 000–20 000	1.62 (1.09–2.42)	1.24 (0.72–2.16)
	>20 000	1.65 (0.99–2.73)	1.27 (0.54–3.01)
Food security (HFIAS)	Food insecure	1	1
	Food secure	2.36 (1.63–3.42)	2.36 (1.30–4.30)	0.005
Place of birth	Home	1	1
	Institution	4.34 (1.86–10.2)	1.98 (0.69–5.72)
Number of ANC appointments	None	1	1
	1–4	1.04 (0.54–2.02)	0.53 (0.23–1.26)
	≥4	2.37 (1.42–3.97)	0.72 (0.34–1.51)
Number of PNC appointments	None	1	1
	1–4	0.69 (0.45–1.04)	0.79 (0.36–1.73)
	≥4	0.70 (0.36–1.37)	1.17 (0.65–2.08)
Current pregnancy	No	1	1
	Yes	1.49 (0.86–2.61)	1.47 (0.71–3.04)
Media exposure	No	1	1
	Yes	3.93 (2.70–5.72)	2.98 (1.80–4.92)	<0.001
Fever	No	1	1
	Yes	0.63 (0.38–1.05)	0.54 (0.28–1.04)
Birth order	No	1
	Yes	0.52 (0.34–0.78)	1.03 (0.54–1.96)
Breast-feeding status	Not breast feeding	1	1
	<8 mo	1.17 (0.68–2.01)	0.91 (0.43–1.96)
	≥8 mo	0.62 (0.42–0.91)	0.87 (0.46–1.62)
Starting age of complementary feeding	<6 mo	1	1
	At 6 mo	2.92 (1.16–7.34)	0.97 (0.33–2.96)
	>6 mo	2.10 (1.37–3.22)	0.71 (0.22–2.26)
Vaccination	No	1	1
	Yes	1.79 (0.96–3.35)	0.80 (0.33–1.96)

ANC: antenatal care; AOR: adjusted OR; COR: crude OR; ETB: Ethiopian Birr; HFIAS: Household Food Insecurity Access Scale; PNC: postnatal care.

Discussion

The study has revealed that the consumption of diets rich in vitamin A among 6–23-mo-old children in Jijiga town is 49.8%. This figure is higher than findings from Butajira (37.8%), Kemba Woreda (24.7%), rural part of Sidama region (9%), Abyi-Adi town (26.3%) and EDHS 2019 (38.%).^22–26 This discrepancy might be due to differences in study setting, agro-ecological zones and dietary habits. Because the study was conducted in an urban setting, higher intake is expected, as city residents have better access to a variety of food groups that contain vitamin A.²⁷ In addition, the high female literacy rate of the urban population is expected to have a positive impact on the quality of foods served for children.¹⁰ This study found that children whose mothers have a higher educational status (i.e. secondary and above) are more likely to take VARF than children whose mothers are not able to write and read. This is in line with the findings of prior studies done in Ethiopia.^17,27 This might be because educated mothers are more likely to implement proper complementary feeding as taught by health professionals.^28,29

We found that maternal media exposure is one of the positive predictors of VARF consumption among children aged 6–23 mo in this study.²⁶ This finding is in line with the results of multilevel analyses of DHS data at both the national and East African levels.^17,30 This may be explained by the fact that health and nutrition information disseminated through the media is presented in a way that is easy and simple to understand, and can reach many people efficiently.

Children in the age groups of 12–17 and 18–23 mo have higher odds of consuming VARF, a finding also supported by findings from both the 2019 EDHS and studies conducted in southern Ethiopia.^16,17 This is likely because older children have an increased exposure to family meals and are better able to consume a variety of solid foods that provide vitamin A.³¹

Household food security status is another factor affecting the intake of vitamin A-rich diets among children aged 6–23 mo in this study. Children from food-secure households had more than twofold higher consumption of vitamin A-rich diets compared with children from food-insecure households. A similar study conducted in eight zones in four Ethiopian regions (Afar, Amhara, Benishangul-Gumuz and Tigray) revealed that households that had received food and cash assistance from the PSNP were more likely to eat animal-source foods.³² This could be because food-secure households are better able to access and afford a variety of nutrient-rich foods.³³

Based on our findings, children from Protestant families were found to have a higher intake of VARF than those from other religious groups. Similar findings were reported in a study based on data from the EDHS between 2005 and 2016 and the 2019 Mini DHS.³⁴ One possible explanation is that Protestant communities in Ethiopia usually do not follow strict fasting practices that restrict the intake of animal-source foods. The cause behind this association is a topic for further investigation.

Conclusion

The current study revealed that almost one-half of the children aged 6–23 mo who live in Jijiga town have a good intake of vitamin A-rich diets. Higher maternal education, media exposure, household food security, older child age (>11 mo) and being from protestant families are factors associated with good consumption of VARF. Making health and nutrition information more available through the media and empowering mothers through education are crucial for improving vitamin A-rich food intake in young children. Along with this, teaching parents about age-appropriate complementary feeding practices and ensuring household food security are important. Moreover, engaging religious leaders while delivering culturally sensitive health education and nutrition messages that align with religious beliefs can improve community acceptance and participation.

Limitations of the study

One of the limitations of this study is the use of a dichotomous outcome variable to measure vitamin A-rich food consumption. This approach is consistent with national survey methods of Ethiopia, such as the EDHS. However, it may not capture the quantity, frequency or portion size of the foods consumed, which limits a more detailed understanding of dietary adequacy.

Author contributions

Biruk Fantahun (Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing—original draft, Writing—review & editing), Sinetibeb Mesfin (Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing—original draft, Writing—review & editing), Dawit Abebe (Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing—original draft, Writing—review & editing), Berhe G/Michael (Methodology, Supervision, Writing—review & editing), and Adisu Birhanu (Methodology, Supervision, Writing—review & editing).

Funding

The authors declare that no financial support was received for the research, authorship and/or publication of this article.

Competing interests

The authors report there are no conflicts of interest associated with this work.

Ethical approval

Before starting the data collection process, the study protocol was submitted and approved by the Institutional Research Ethics Review Committee (IRERC) of the College of Health and Medical Sciences, Haramaya University, on 23 December 2023 (Ref. No. IHRERC/173/2023). Official letters of cooperation were submitted to Jigiga Health Bureau and other concerned bodies to obtain their support and consent for facilitating the study. Informed, voluntary, written and signed consent was obtained from all respondents before participation. Participants' privacy and confidentiality were assured by excluding names and identifiers from the questionnaire.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request. The raw data are not publicly available due to institutional restrictions but can be shared without undue reservation upon request.