Does livestock ownership predict animal-source food consumption frequency among children aged 6–24 months and their mothers in the rural Dale district, southern Ethiopia?

Ethical statement

Ethical approval was obtained from the Hawassa University College of Medicine and Health Sciences Institutional review board, Ethiopia (reference number; IRB/025/10), and from the Norwegian Regional Ethical Committee (REK), Norway (reference number; 2018/90/REK Vest). All the necessary official letters were obtained from the concerned bodies at regional and district levels. The respondent’s signature or finger stamp was obtained before enrolment to signify their willingness to participate in our study. The data has been kept confidential and no personal or household identifiers were used to communicate our findings.

Study setting

This study was conducted in the rural district, in the Sidama regional state of Ethiopia. The Dale district, comprised of 36 rural and two urban kebeles, is one of the 37 districts (previously it was one of the 19 districts) according to the Sidama Region Health Bureau (2022). A kebele is the smallest administrative unit. This study was conducted in seven rural kebeles of the district. The projected population size of the district in 2022 was reported over 350,000 (Sidama Region Health Bureau, 2022). The main town in the district, Yirga Alem, lies 320 km from Addis Ababa. The Dale district has 33 health posts, ten health centres and a hospital called Yirga Alem General Hospital. The people living in the study area are mainly farmers, cultivating maize, ensete (Ensete ventricosum), kale, cabbage, and haricot beans. Coffee, chat (Catha edulis) and different seasonal fruits are grown in the area. There is also a cultural value of keeping livestock such as cows, goats and sheep. Most rural households also have hens.

Study design

A community-based cross-sectional study was conducted to assess animal-source food consumption frequencies, livestock ownership and background information. The study was conducted from August to November 2018.

Study participants

This study included a sample of 851 children aged 6–24 months and caregivers (primarily the mother) out of 971 children aged 0–24 months and caregivers enrolled in a larger study (Household food security and dietary practices in the rural Dale district, southern Ethiopia). Study procedures, including the original sample size estimation, sampling procedures and recruitment are described in a previous article (Behailu et al., 2022). A census, household listing, proportionality according to kebele size, and simple random sampling were performed. This study excluded those younger than 6 months, as exclusive breastfeeding is recommended for that age group, so that 853 children aged 6–24 months remained. Of those, two children had no mother or caregiver eligible for dietary data collection (Fig. 1).

Study variables

The outcome variables in our study were ASFs consumption frequency among children and mothers. ASFs assessed in this study were dairy products (milk, yoghurt, cheese and whey), eggs and meat (flesh meat, organ meat, poultry and fish). Consumption frequency was defined as the usual number of servings per day, week, month and 3 months, irrespective of the portion size served to children and their mothers. Frequencies were stated as ‘never’ if not consumed in the past 3 months, ‘rarely’ if consumed in the past 3 months, but not consumed in the past 1 month, ‘sometimes’ if consumed one to three times in the month prior to this study, ‘often’ if consumed one to six times in the week prior to this study, and ‘always’ if consumed at least once daily. For analysis purposes, the frequencies were assigned with ordered numbers: ‘0’ for ‘never’ and ‘rarely’, ‘1’ for ‘sometimes’, ‘2’ for ‘often’, and ‘3’ for ‘always’.

The independent variables were livestock ownership such as cows, donkeys, oxen, goats or sheep, and hens at household level. Child characteristics included age and gender, information about mothers comprised age and educational status, and household background described household size, and food insecurity status.

Data collection

Data collection procedures and data quality measures are described in our previous article (Kebede et al., 2022), explicitly about methods applied to household food insecurity and wealth index analysis.

Statistical analysis

The data was analyzed using SPSS version 25 (SPSS Inc., Chicago, IL, USA). Frequencies, percentages and means were used to describe continuous and categorical variables. Ordinal logistic regression (OLR) analysis was used to explain the association between the outcome variable, which was a composition variable including dairy, egg and any meat consumption, and predictor variables. Before running the multivariable ordinal logistic regression, bivariable ordinal logistic regression was conducted. Variables with a P value < 0.2 in the bivariable ordinal regression were included in the multivariable ordinal logistic regression. Multi-collinearity was also checked to rule out any association between the independent variables before running the adjusted ordinal logistic regression. The unadjusted odds ratio (uOR) and the adjusted odds ratio (aOR), with the 95% confidence intervals (CI), were used to explain associations between the outcome and the exposure variables. P value < 0.05 was used to determine the level of statistical significance.

Background characteristics

In this study, a total of 851 child-mother pairs were analyzed. Among the children, 50.9% (433) were females and the mean age in months was 15.4 (95% CI [15.1–15.8]). The mean age of the mothers was 26.9 years (95% CI [26.5–27.2]). About half of the mothers, 45.1% (384) had attended primary school. The mean household size was 4.8 persons (95% CI [4.7–4.9]), with 51.5% (438) of households having less than five members. Approximately 54% (459) of the households experienced moderate to severe food insecurity (Table 1). When it comes to livestock ownership, 74.1% (631) of the households owned cows, while 25% (213) owned goats or sheep. Around one-fifth, 20.7% (176) of the households kept all types of livestock including sheep/goats, cows, and hens, whereas 17.4% (148) did not have any of these animals. Oxen and donkeys were found in only a few households; specifically, oxen were kept by 2.9% (25) of the households and donkeys by 6.5% (55) of the households (Table 2).

Animal-source food consumption frequencies among children aged 6–24 months

More than 85% (745) of the children had consumed dairy products daily or up to six times per week, and 70 children had never consumed dairy products during the past month. Eggs were served one to six times per week for 36.5% (311) of the children, while 17% (146) of the children had not eaten eat eggs during the past month. No child consumed meat on a daily basis. Almost three-fourths, 73.8% (628), had not eaten meat in the past month. Out of the 851 children in total, 6.6% (56) had not consumed any ASFs during the past month (Table 3). In addition to this, 50 (28 female and 22 male) children aged 6 to 9 months were not getting any food other than breast milk. The distribution of ASF consumption frequency among children by each livestock of ownership was provided in Table S1.

Animal-source food consumption frequencies among mothers

The dairy product consumption of mothers during the past month was 96% (817) of the 851 mothers (Table 4). More than 50% (430) of mothers had not eaten eggs during the past month and 31.3% (266) of the mothers had eaten eggs one to three times per month. Overall, 2.2% (19) of them did not consume any ASFs during the month prior to our survey (Table 4). The distribution of ASF consumption frequency among mothers by each livestock of ownership was provided in Table S2.

Comparison of animal-source food consumption between children (male and female) and mothers

Animal-source food consumption among children and mothers was varied. While meat and egg consumption was highest among children, dairy product consumption was highest among mothers. ASFs were almost equally consumed among male and female children aged 6–24 months (Table S3).

Animal-source food consumption among child-mother pairs at household level

Animal source food consumption at household level was also described according to child-mother pairs. Households in which only the child, only the mother or both, consumed or did not consume ASFs during the month prior to this study were described. In 13% (111) of the households both the child and the mother had consumed meat, while this was 47% (400) for eggs and 89% (757) for dairy products. Children ate meat and eggs a little more frequently than their mothers, while mothers ate dairy products slightly more frequently than their children (Table S4).

Associations between livestock ownership and animal-source food consumption frequencies among children aged 6–24 months

The frequency of dairy products consumption was 1.8 times higher among children in households owning cows than among those which did not (aOR = 1.8; 95% CI [1.3–2.5]). In contrast, the frequency of meat consumption was lower among children in households keeping cows (aOR = 0.66; 95% CI [0.45–0.95]), compared with those that did not have cows. Goat or sheep ownership was associated with a 1.7 times greater frequency of egg consumption among children, than for households without goats or sheep (aOR = 1.7; 95% CI [1.2–2.3]). Furthermore, owning goats or sheep increased the child’s dairy product consumption frequency more than twofold (aOR = 2.3; 95% CI [1.6–3.3]). Egg consumption frequency among children was 3.5 times higher among households that kept hens than among those which did not (aOR = 3.5; 95% CI [2.6–4.8]) (Table 5).

Associations between livestock ownership and animal-source food consumption frequencies among mothers

The ownership of livestock, specifically goats or sheep and hens, was associated with higher frequencies of dairy, egg and meat consumption frequency among mothers. Cow ownership was specifically associated with increased dairy products consumption frequency. While goat or sheep ownership was associated with all ASFs consumption frequency, cow ownership was solely associated with dairy product consumption frequency (aOR = 3.0; 95% CI [2.2–4.2]). The frequency of dairy product consumption among mothers was also greater in households keeping goats or sheep (aOR = 1.9; 95% CI [1.3–2.6]) than among mothers in households without these animals. Regarding egg consumption, there was a 2.3 and 2.0 greater frequency among mothers in households with goats or sheep (aOR = 2.3; 95% CI [1.6–3.1]) and hens (aOR = 2; 95% CI [1.5–2.7]), respectively. Having hens was also associated with meat and egg consumption frequencies among mothers. Meat consumption frequency among mothers was 1.8 times (aOR = 1.8; 95% CI [1.3–2.5]) higher in households that had goats or sheep than in those that did not. Likewise, ownership of hens was associated with a 1.5 increased frequency of meat consumption among mothers (aOR = 1.5; 95% CI [1.1–2.2]) (Table 6). In general, these findings suggest that having access to livestock may contribute to a higher intake of ASFs in rural households.

Socio demographic variables associated with animal-source food consumption among children and mothers

Child age, household size and educational status of the mother were associated with ASFs consumption frequencies among children and mothers. Child age was positively associated with dairy (aOR = 1.05; 95% CI [1.03–1.08]), egg (aOR = 1.07; 95% CI [1.04–1.09]) and meat (aOR = 1.06; 95% CI [1.03–1.09]) consumption among children. It was also positively associated with egg (aOR = 1.04; 95% CI [1.01–1.06]) and meat (aOR = 1.05; 95% CI [1.02–1.07]) consumption among mothers. Secondary school attendance of the mother was positively associated with meat consumption among children (aOR = 1.9; 95% CI [1.3–3.0]), and with dairy (aOR = 2.1; 95% CI [1.4–3.2]) and meat (aOR = 1.8; 95% CI [1.1–3.1]) consumption among mothers. In addition, primary education of the mother was positively associated with all animal-source food consumption among mothers. However, household size was negatively associated with dairy consumption among children (aOR = 0.84; 95% CI [0.8–0.9]) (Table 5), and with dairy (aOR = 0.85; 95% CI [0.8–0.9]) and egg (aOR = 0.8; 95% CI [0.67–0.85]) consumption among mothers (Table 6).

Strengths and limitations of our study

Our study had several strengths: it was a community-based study using a random sampling technique, collecting a comprehensive set of data at the child, mother and household level, and using the ordinal logistic regression model to describe associations between the outcome and exposure variables. Checking independent associations between each livestock type with each ASF is also an additional strength, since this helps to address indirect associations between livestock ownership and consumption of specific ASFs. Moreover, assessing the food frequency over one month helps to address food items with infrequent consumption. On the other hand, the limitations of this study include the inherent limitations of the cross-sectional design: one can never rule out reverse causality. Between the mapping and the actual cross-sectional study there was replacement of households due to people leaving their residence prior to the data collection. Recall bias for self-reported data, such as age and food frequencies, could be a source of bias. To minimise the recall bias, however, we used the strategies described under data quality control measures in the method section. Survivor bias might also exist, since our samples were mother-child pairs, and mothers who had lost their children were not included. We did not assess the food portions consumed, the number of livestock and to whom they belonged (the mother or the father), which is also a limitation.