This study examined the concordance between maternal and child dietary diversity and factors affecting the concordance. The study showed that the proportion of discordant is few, which is 23.9% of the total, 7.1% of mothers and 16.8% of children only. Even though, the Ethiopian government implemented health extension program to educate the community on different health packages including maternal and child feeding practices , the minimum dietary diversity was 24.7% and 34.4% in mothers and children respectively. Even though it calls for further efforts in food security and raising of awareness on the importance of dietary diversity, it is higher than the national figure at 10.8% and the nearby district in Gamo Gofa zone which is 23.3% in children 6–23 months[ 22]. The dominant dietary food groups consumed were grains, roots and tubers 99.4% in mothers and 95.8% in children followed by legumes, pulses and nuts consumption of 70.5% in mothers and 67.8% in children. Consumption of flesh foods is very low at 4.6% in both mothers and children. The possible explanation might be the more accessibility of staple foods of starchy nature  and belief of mothers that young children could not be able to digest flesh foods and the low socioeconomic status made the mothers unable to purchase and fed these relatively costly flesh food groups from the local market for their family. This has an implication on raising public awareness on the benefits of diversified diet.
The study also showed that there is agreement between maternal and child dietary diversity as it was revealed in the Kappa statistics (kappa = 0.43, p < 0.001) which indicates that there is moderate concordance between mother - child dyads in dietary diversity. This is to mean that the more food groups the mothers consumed, the more likely their children achieved their minimum dietary diversity and vice versa. As the mothers’ dietary diversity increased, the percentage of children (6–23 months) meeting this criterion increased dramatically. An increase in the percentage of children reaching the minimum dietary diversity was greater with each successive increase in maternal dietary diversity. Even though there is a dearth of literature on concordance between maternal and child dietary diversity, a related study on maternal and child dietary diversity associations in Bangladesh, Vietnam and Ethiopia showed a fair association between the two. The variation could be attributable to differences in methodology (the current study was conducted using the seven food groups for children and the ten food groups for mothers while the previous study used the seven food groups for mothers and children to enable direct comparisons), study settings, study population dynamics, timing of the study and other related factors.
This study also found the odds of being low concordant was higher for mother-child pairs from rural areas as compared to their urban counterparts. However, of these rural dweller concordants, 77% mother - child pairs did not achieve their minimum dietary diversity score. This showed most rural mothers did not achieve the recommended minimum dietary diversity and also failed to meet their children’s. However, 76.2% of the urban dweller concordants achieved the minimum dietary diversity (high concordant). This result was slightly higher than the study conducted on dietary diversity of Nigerian rural women the majority had low dietary diversity with none being in the high category  but some evidences showed that people who reside in rural were more likely to adopt their traditional food and fed more diverse foods [41, 42]. However, the reverse was true in this study area as the diets were not varied enough. This low dietary diversity of the rural women could be a function of low socioeconomic status of rural women and low awareness on the importance of diversified diet. Because most of them earn low income, and this may lead them to inability to afford food varieties. The low dietary diversity score of the rural mothers and children indicates that they may not meet their micronutrient requirements[2, 43, 44].
The study also revealed that maternal education is a significant predictor of maternal-child dietary diversity concordance where mothers having no formal education had higher odds of being concordant with their children as compared to those who attained secondary and above level education. The finding contrasts with the fact that maternal education enhances diversity both in the mother and child diets [39, 43, 45]. This could be explained by the situation of Ethiopian mothers where majority of uneducated mothers are housewives who could have a better caring opportunity for their children and might feed from the same pot. Interventions targeting such women could improve the micronutrient deficiency among children. However, the association between higher education and better dietary diversity concordance was reflected by the difference in proportion among educated and uneducated mothers in the study. That is, 85.6% of mothers with no formal education did not achieve their minimum dietary diversity (low concordants). On the other side, 41% of mothers who attained secondary and above level of education were at high category of concordance that is relatively higher than the 14.4% who achieved the minimum dietary diversity among women with no education. This result coincides with the study conducted in Bangladesh and Vietnam . Similarly, a study from Zambia on dietary diversity at six months of age also showed that maternal education was positively associated with dietary diversity score[43, 45]. This could be due to maternal knowledge that mothers who were educated take much care of their children and may consume for themselves and feed their children diversified diet compared to those who had no schooling. This suggests that education has positive impact on improving maternal and child DD as educated women are more likely to receive nutrition education which in turn increases the chance of consumption of diversified diet [45, 46].
The study also showed absence of milking cow was positively associated with low maternal and child dietary diversity concordance. Of the concordant mother-child dyads, 52.7% of mothers who own milking cow fall under high concordance. This shows that the proportion of mothers who own milking cow and achieved minimum dietary diversity were higher than those mothers who did not own. This result goes in line with a related study on dietary diversity, feeding practice and determinants among children 6–23 months in South Ethiopia which showed that mothers who had access to cow milk fed diversified diet two times more than those who had no access. This association implies that availability of a source of food in the household may influence food intake. Evidences suggest that increased availability of fruits, vegetables and snack foods in the home was associated with increased intake of each food, respectively, among pre-school age children . Similarly, availability of milking cow in households leads to high consumption of milk in mother and children that may enhance their dietary diversity.
Moreover, compared to those children who consumed high dietary diversity; those who consumed low dietary diversity had higher odds of being concordant to their respective mothers. This could be due to the high ratio of concordant children who did not achieve the minimum dietary diversity to low discordant than the ratio of high concordant(who achieved the minimum DDS) to low discordant. The result simply showed that the proportion of children who did not achieve the minimum dietary diversity but being concordant with their mothers was high.
On the other hand, mothers who consumed high dietary diversity had lower odds of being concordant with their children than those mothers who consumed low dietary diversity. The possible reason for this is due to high proportions of high discordant children who achieved the minimum DDS and low number of discordant mothers who achieved the minimum DDS. The results were consistent with the existing literature depicting the association of maternal and child diets among preschools and school-aged children and among under 24 months children . However, the finding was inconsistent with findings from a study conducted in Cambodia, Ghana and Haiti DHS analysis which showed children with mothers who consumed low food groups, breastfed children whose mothers consumed more than 5 food groups had higher odds of achieving minimum dietary diversity across the three countries. Mothers’ dietary diversity predicted increases in child diet in some food groups in the studies in these three countries (5). The more food groups the mothers consumed, the more likely their child attained the minimum DD and the more they become high concordant to each other. The study is also consistent with a study compared with children whose mothers consumed less food groups, children whose mothers consumed high food groups were more than twice as likely to achieve the minimum DD in Bangladesh and Vietnam and were 5 times more likely to achieve the minimum DD in Ethiopia which is much higher than the present study (39).
In this study, 16.8% of mothers were positive deviants who buffer their children, and possibly these mothers may have benefited children’s diets. These mothers reduced their own consumption to act as a buffer against low food diversity for their children to protect children from low diversity or imbalance of micronutrient deficiency. In general this study showed that mothers with higher DD have children with higher DD and mothers with lower dietary diversity have children with lower dietary diversity. This suggests that irrespective of children’s breastfeeding status, they consume the same food groups as their mothers. Because maternal DD is strongly associated with child DD, diverse diet should be promoted for both mothers and children during the entire span of the first 1000 days of mothers and children.
This study used primary data and was conducted as community-based research and believed to be representative for similar settings that should be considered as strength. However, the study has few limitations to consider. Though DDS has been validated as a useful tool to assess the likelihood of meeting micronutrient requirements, the maternal and children’s diet was analyzed only qualitatively as quantity was not taken in to account. The study also did not show the strength of association for each food group as only the general food group concordance was shown and it did not consider seasonal variation in DDS. Even though probing technique was used, recall bias could be introduced. Because of low count or proportion of high concordance, factors affecting low concordance were determined. Therefore, caution should be exercised in the interpretation of the findings.