A balanced diet should include a variety of fruits and vegetables (FV) since they are excellent providers of dietary fiber, vitamins, and minerals, as well as a number of other helpful non-nutrients like flavonoids, plant sterols, and other antioxidants. Eating them contributes to making sure that several of these vital elements are taken in sufficient amounts (1). It is recommended to have one vegetable serving with every meal, and consuming no FV at all is an unhealthy practice, even if there is no consensus recommendation for the ideal amount of servings of FV per day for children older than 6 months (2). To enhance general health and lower the risk of certain non-communicable diseases (NCDs), the World Health Organization (WHO) recommends consuming approximately 400 grams of FV daily as part of a healthy diet low in fat, sugar, and sodium (1).

After 6 months, children undergo rapid growth and development, for which breastmilk alone is inadequate. Thus, it’s fundamental to supplement breastfeeding with nutrient-dense, diversified diets that include fruits and vegetables (3, 4). Zero vegetable or fruit consumption (ZVF), which is one of many elements of minimum dietary diversity (MDD), is a novel metric introduced by the United Nations Children’s Fund (WHO-UNICEF) to assess infant and young child feeding (IYCF). ZVF measures the proportion of children between the ages of 6 and 23 months who have not consumed any vegetables or fruits in the 24 h prior to the query (3).

It was discovered that 90.22% of Ethiopians engaged in poor child feeding practices (5). Although there was a notable 9% rise in minimum dietary diversity from 2011 to 2016 and a 3.4% reduction in the same period, only 4.1% of Ethiopian children aged 6–23 months were fed the minimum permissible diet (6).

With West and Central Africa having the greatest incidence (56.1%) and Latin America and the Caribbean having the lowest (34.5%), the global prevalence of ZVF consumption was 45.7% (3). Like many other low-income countries, Ethiopia has very low FV consumption (7–10). According to one study, 1.5% of Ethiopians consume FV (10), and 69.3% of children in Ethiopia between the ages of 6 and 23 months consume ZVF (11). According to a study done in Southern Ethiopia, 7.2 and 39.1% of the children consumed FV rich in vitamin A, respectively (12), while a study done using nationally representative data showed that 28.2 and 10.3% of children consumed both vitamin A-rich and other FV (13). In 2019, this prevalence rose to 39% (14).

Zero vegetable or fruit consumption during this crucial developing phase has been shown to significantly raise the risk of non-communicable diseases, and WHO-UNICEF has ranked it among the top 10 risk factors for mortality globally (2, 3). Low fruit and vegetable consumption has a major negative influence on public health; globally, it is responsible for 1.7 million (2.8%) deaths and 16 million (1.0%) disability-adjusted life years (DALYs) (15, 16). Another piece of evidence revealed that up to 2.635 million deaths are thought to occur annually worldwide as a result of poor fruit and vegetable consumption (17). In 2017, it was predicted that 3.9 million deaths globally could be attributed to insufficient consumption of fruits and vegetables (2). A lower intake of fruits and vegetables is associated with poor health, an increased risk of obesity (1), and a higher risk of non-communicable diseases such as hypertriglyceridemia, stomach cancer, colorectal cancer, and cardiovascular disorders (1, 9, 15–19). Furthermore, studies revealed that consuming more fruits and green leafy vegetables is linked to a markedly lower risk of type 2 diabetes (20, 21). Furthermore, over half of young Ethiopian children suffer from anemia and micronutrient deficiencies like vitamin A (8). In all, only 13.3% of children in 2011 and 24% in 2016 attained MDD (22), and the nutrition quality of infants and young children in Ethiopia between the ages of 6 and 23 months was impacted by these incorrect complementary feeding methods (5).

Previous studies conducted in Ethiopia revealed inappropriate child feeding practices (5), poor vitamin A-rich food intake (14), and dietary diversity (22, 23) among children was distributed non-randomly across Ethiopia’s regions. Spatial variation of folate nutrition (24) and crop diversity in agroforestry home gardens (25) was observed between farming systems in Ethiopia. A study conducted using the 2019 Ethiopian mini-demographic and health survey data revealed that factors such as middle and rich household wealth quintile, 25–34 maternal age, Muslim religion, having more than four ANC visits during pregnancy, children in the age group of 12–23 months negatively and being married, and health facility delivery positively associated significantly with children’s ZVF consumption (11).

Despite the fact that variations in crop diversity and availability, non-random distribution of dietary diversity, consumption of foods high in vitamin A, and inappropriate feeding practices for infants are indicated, there is no evidence to determine the spatial distribution of ZVF consumption across Ethiopia’s regions. Investigating the spatial distribution of ZVF consumption across Ethiopian regions is advantageous for targeting undernutrition, ensuring proper child feeding practices, and lowering morbidity and mortality from non-communicable diseases in later life.

This study aimed to investigate the spatial distribution of ZVF consumption and its spatial determinants among children aged 6–23 months in Ethiopia.

Our study included a total of 1,489 weighted samples of children aged 6–23 months. More than two-thirds (68.5%) of children’s mothers were aged 25 and above. More than half (55.45%) of the children’s mothers had attended formal education. Almost all (95.44%) of children’s mothers were married. About 57.98% of children were living in a house with more than one under-five children. Majority (71.34%) were living in rural areas of Ethiopia, while about 36.74% were residing in the Oromia region (Table 2).

In order to comprehend regional variation and distribution as well as factors contributing to the observed regional variation, this study looked at the spatial variation of ZVF consumption and its associated factors among children aged 6–23 months in Ethiopia. The majority of the Benishangul Gumuz, Gambela, Somali, and SNNPR areas had a highly clustered intake of ZVF. Additionally, a spatial scan analysis resulted in a total of 120 significant clusters. All explanatory factor taken into account in the final model including mother education, the child’s age (12–13 months), home delivery, having more than one child under the age of five living in the house, and being currently pregnant were significant explanatory factors for the regional variation of ZVF. As a result, we mapped these important variables’ local coefficients.

Accordingly, the spatial distribution of ZVF was non-random in Ethiopia and highly clustered in the entire Somali region, central and southeast Amhara, and western and northwest Afar regions. A previous study also indicated that ZVF intake varies spatially (3). And it is also fair to interpret our result relative to other studies conducted to determine the spatial distribution of vitamin A-rich food consumption. Thus, previous studies conducted in Ethiopia revealed that low consumption of foods rich in vitamin A was significantly clustered in the Dire Dawa, Harari, Somali, Afar, eastern Tigray, and southeast Amhara regions (14, 38–40). A study conducted in Vietnam also revealed a spatial variation of vitamin A-rich foods (41). A study conducted in Nigeria also reported geographical differences in the consumption of fruits (42). Fruit and vegetables are one indicator of MDD, and inadequate dietary diversity was highly clustered in the Amhara, Tigray, and Afar regions (22). Moreover, MDD, which involves fruit and vegetable consumption among children (12), is one indicator of appropriate complementary feeding (43). In Ethiopia, inappropriate child feeding practices had a non-random spatial distribution where high clustering was observed in western Tigray, Afar, western Amhara, and Southern Oromia, as well as eastern SNNPR, and low clustering was seen in Addis Ababa, central Oromia, the southern Amhara region, northern SNNPR, southern Benishangul Gumuz, Harari, Dre-Dawa, and the Northern Somali regions (5).

The observed variation in the consumption of fruits and vegetables can be attributed to multiple factors. The type of food consumed and the concentration of micronutrients in Ethiopia can be greatly influenced by a variety of factors, including variations in farming systems across the regions, affordability, utilization, food preferences, lack of nutrition education and storage facilities, and differences in residential food availability (24, 44–46), as well as by the seasonality of fresh vegetables and fruits that are available (3). In addition, prices are rapidly increasing, both in real terms and when compared to cereal. At the end of 2019, vegetables and fruits were significantly more expensive than 15 years earlier. Especially green leafy vegetables show a significant price rise. There is also a significant spatial price variation in Ethiopia, where vegetable prices are 60 percent more expensive in lowland regions than in the Amhara region, where vegetables are the cheapest. Fruit prices in the lowlands are double the prices in the major producing area, the Southern Nations, Nationalities, and Peoples’ (SNNP) region (47). That is the reason for the lowest prevalence, which was seen in the entire Gambela, the majority of western Oromia, western Benishangul, the majority of SNNPR, and some parts of the northwest Tigray regions, where fruit production is comparatively better as spatial and temporal variation in crop diversity in Ethiopia is known (25). Due to rainfall, temperature, and altitude differences in the country, the eastern part of the country (East Hararghe) with vegetables dominating; Central Oromia Regional State (which produces both fruits and vegetables); SNNPR (Gamo Goffa, Woliata and Sidama zones); and eastern Ethiopia (Dire Dawa and Harari) are also well-known production and supply areas of both FV; some parts of Tigray and Amhara regions are the major fruit and vegetable growing areas of the country (48). Other sociodemographic factor variation across the regions may contribute to the spatial variation of ZVF consumption in Ethiopia (4, 49).

Our study revealed that mother education was a spatially significant predictor variable for ZVF consumption. It had a significant negative effect almost across all regions that ranged between-0.488 and-0.191. This indicates a 1% increase in the proportion of women with formal education, which decreases ZVF intake by a range of 19.1 to 48.8% among children aged 6–23 months. This finding is supported by previous reports (3, 18). Other studies also revealed that mother education was positively associated with different micronutrient rich food intakes (14, 40), dietary diversity (22, 23), and the minimum acceptable diet intake (6). The highest effect was seen in less developed regions and this might be due to regional inequalities in educational achievement and opportunities among women (49). Our finding is in contrast to the previous multilevel analysis (11). This might be due to the effect of space and spatial relationships in our analysis (50).

Children’s age (12–23 months) was another significant spatial predictor for ZVF intake. It had a spatially negative and positive effect that ranged between-0.267 and 0.253. A 1% increase in the proportion of children aged 12–23 decreases the prevalence of ZVF intake by 26.7 to 20% in northwest Benishangul, western and northwest Amhara, Dire Dawa, Hareri, some parts of eastern Somali, some parts of northeastern Oromia, and the southern tip of the Afar regions. This finding is consistent with the results of previous studies (3, 11). However, our result also indicated that a 1% increase in the proportion of children aged 12–23 increases ZVF intake by a range of 2.3 to 25.3% in southeastern SNNPR and southern Oromia. This might be due to the lack of availability of fruit and vegetables in these specific areas (48) or decreased maternal care and attention given to children as they grew up (51).

The effect of home delivery had varying statistical significance for ZVF consumption among children aged 6–23 months in Ethiopia. It’s the coefficients that vary spatially between-0.02 and 0.512. However, in the significant areas (entire Benishangul, the majority of Amhara, southern Afar, some parts of northwest and central Oromia, and Addis Ababa), the effect size ranges between 0.208 and 0.512. This indicates that a 1% increase in the proportion of children delivered at home in these areas can increase ZVF intake by a range of 20.8 to 51.2%. Another study also supports this evidence (52). This might be due to the reduced effect of home delivery on good infant and young child feeding practice (53–55), since health facility delivery creates opportunities for advice and encouragement that foster child feeding practice. However, our finding is in contrast to the previous report (11). This might be the reason why spatial analysis is more advantageous over other statistical analysis methods since it integrates space and spatial relationships (50).

The presence of more than one under five children in the house had a varying negative effect in different parts of Ethiopia on ZVF intake among children aged 6–23 months in Ethiopia. It’s the coefficients that vary spatially between-0.494 and-0.024. However, in the most significant areas, the effect size ranges between-0.494 and-0.289. This indicates a 1% increase in the proportion of children living with more than one under five children in the house, which decreases ZVF intake by a range of 28.9 to 49.4%. A previous study revealed that having three or more under-five children in the house increases undernutrition status (56). However, this significant effect was seen in Addis Ababa, central Oromia, southern Amhara and afar, some parts of northeast SNNPR, and southeastern Benishangul, where a better availability of fruits and vegetables is indicated (48). Thus, if the number of under-five children in the house increases, mothers may give better attention to caring for their children.

The relationship between mother’s current pregnancy status and the prevalence ZVF intake was spatially different and both negative and positive which ranges between-0.085 and 0.226. In the significant areas, a 1% increase in the proportion of pregnant women increases ZVF consumption by a range of 11.2 to 22.6%. Mothers may worry more about their pregnancies than their children, which could be the cause of the positive effect of pregnancy on ZVF consumption.

Significant geographic variation in ZVF consumption was found in this study throughout Ethiopia’s regions. Significant explanatory factors for the spatial variation in ZVF consumption among children aged 6–23 months were the formal education level of the mother, wealth status, the child’s age, place of delivery, number of under-five children living in the home, and pregnancy status of the woman. Our research shows that improving children’s consumption of fruits and vegetables in hotspot areas is a critical requirement. Therefore, in order to lower children’s rising ZVF consumption rate, boosting household wealth status and expanding access to education should take precedence. Consequently, rather than implementing services randomly in hotspot regions, area-based interventional strategies that take spatial explanatory factors into account would make regional nutritional and educational interventions more cost-effective. Lastly, our findings should encourage public health authorities to develop targeted nutritional education.

Publicly available datasets were analyzed in this study. This data can be found here: http://www.dhsprogram.com.

Ethical approval was not required for the studies involving humans because this study was conducted using the 2019 Ethiopian demographic and health survey datasets. To use the datasets we granted a permission from the MEASURES DHS program and the authorization letter is attached with files. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements because this study is a secondary data analysis. Thus, it is not applicable.

WT: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. MS: Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing. HG: Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing. ZG: Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing. LA: Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing. TT: Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing.