Low Dietary Diversity Score is the Cause of Underweight Among Expectant Mother from Rural Ethiopia: A Community Based Cross-Sectional Study


 Background: Maternal underweight is a challenging public health issues globally affecting women in most developing countries and consequently affect their newborns. Such information however is scantly in the study area and thus we examined the contribution of low dietary diversity score towards underweight among expectant mothers from rural Ethiopia.Methods: This study was based on data extracted from a large data set gathered from three rural villages from September 2017 to June 2018 in a district of North-Shewa Zone of Amhara Regional State which is 220 km north of Addis Ababa, the capital city of Ethiopia. Data on socio demographic characteristics food groups consumed and Mid-Upper Arm Circumference (MUAC) measurement were taken for all expectant mothers by trained health extension workers in accordance with relevant ethical guidelines and regulations. The collected data were entered into Epi-data software version 3.02 and analyzed using SPSS version 20. Statistical significance was affirmed at a p-value ≤ 0.05. Result: Of the 400 consented participants, 380 of them had adequate information and thus analyzed for their variety of food groups consumption and their MUAC measurement outcomes. The magnitude of underweight was 18.2% and the mean (±SD) dietary diversity score was 4.45(±1.32). interestingly, the relationship between poor dietary diversity score and underweight of the participants was significant (AOR= 2.4; 95% CI=1.05 to 5.3). Conclusion: Our study revealed a significant relationship between low dietary diversity score and underweight of expectant mothers. it is thus, essential to improve the dietary diversity and nutritional status of pregnant women through strengthening the ongoing productive safety net program and empower women to earn more income so that mothers would purchase variety of foods and ultimately break the intergeneration cycle of malnutrition.

their variety of food groups consumption and their MUAC measurement outcomes. The magnitude of underweight was 18.2% and the mean (±SD) dietary diversity score was 4.45(±1.32). interestingly, the relationship between poor dietary diversity score and underweight of the participants was signi cant (AOR= 2.4; 95% CI=1.05 to 5.3).
Conclusion: Our study revealed a signi cant relationship between low dietary diversity score and underweight of expectant mothers. it is thus, essential to improve the dietary diversity and nutritional status of pregnant women through strengthening the ongoing productive safety net program and empower women to earn more income so that mothers would purchase variety of foods and ultimately break the intergeneration cycle of malnutrition.

Background
Maternal underweight is a challenging public health issues globally affecting women in most developing countries including Ethiopia and responsible for the death of mothers and children (1)(2)(3)(4). Because of the grave consequences of malnutrition women are advised to take more diversi ed food to get adequate energy and multiple nutrients to satisfy their demand as well as their offspring. Mothers in most developing countries however, do not adhere to the nutritional advices provided to them during their antenatal follow up due to several factors that include limited access to and consumption of micronutrient rich foods, frequent illnesses, teenage pregnancy with narrow birth intervals and high fertility rates (1,6).
The solution requires effective multisectoral interventions that include health, social welfare, education, water, sanitation, in addition to consumption of balanced diet which contributes to the wellbeing of the mother as well as her offspring (7). In line with this, the government of Ethiopia has developed a revised national nutrition program (NNP) in 2016 to address the burden of malnutrition in pregnant and lactating women (8).
Despite the implementation of the revised NNP initiative, maternal underweight remained a pressing problem according to the 2016 Ethiopian demographic health survey which documented 22.4% of women in reproductive ages (15-49 years) were undernourished (9).
Other recent studies done on pregnant women indicated that poor dietary scores in the same age and sex groups are documented to be another important problem that have a strong positive relationship between appropriate maternal nutritional status and birth outcomes and ranged from 24.2%-74.6% (15)(16)(17).
Nonetheless, such information for expectant mothers from rural north Shewa is scanty and therefore, examined the association of low dietary diversity score with the nutritional status of pregnant women based on a large data set collected in a rural community to generate evidence that can be used to design appropriate interventions for the villages as well as for the region and the country.

Study Area and Population
This study was extracted from a large data set gathered in a Community based cross-sectional study conducted among expectant mothers from September, 2017 to June, 2018 in a rural district of North-Shewa Zone of Amhara Regional State which is 220 km north of Addis Ababa, the capital city of Ethiopia [Master Thesis, 2018] (18).
A total of 400 expectant mothers from ve villages were randomly selected in accordance with relevant ethical guidelines and regulations in the initial mother document and allocated proportionally to each of the villages based on the constituted sampling frame obtained from health posts.
Data on socio demographic characteristics food groups consumed (nine food groups as displayed in Table 1) and MUAC measurement were taken for all expectant mothers by trained health extension workers. Maternal MUAC was measured using non stretchable tape (MUAC) and recorded as normal or abnormal when MUAC >22cm and <22cm, respectively (19). Dietary diversity scores (DDS) was formed by summing up the number of food groups consumed over a 24-hour period by an individual and labeled as high and low DDS when the food groups consumed were more than ve and below ve, respectively as recommended by FAO and FHI 2016 (20). The collected data were entered into Epi-data software version 3.02 and analyzed using SPSS version 20 (21)(22).

Results
Of the 400 recruited mothers, only 380 of them who had complete information were included in the analysis and the rest were excluded. The mean (±SD) age was 22.5 ± 4.3 years indicating younger age groups.

Consumption of the Modi ed Dietary Diversity Score (MDDS)
Grain, white roots, and tubers were amongst the commonly consumed grains by all expectant mothers predominantly in the form of breads, pasta, macaroni, porridge, rice. The second most commonly consumed meals as their integral part of the main dish for the vast majority (98.7%) was vegetables. Consumption of pulses and other vitamin A-rich fruits/vegetables was also reported in 66.3% and 55.3% of them, respectively. Milk and milk products (dairy), and other dark green leafy vegetables were consumed in 46.6% and 29.7%, respectively. Meat, poultry, and sh were consumed in 27.6%. the details of food groups are displayed in Table 1. As shown in Table 2, the proportion of mothers who consumed more than ve food groups (high dietary diversity score) was nearly 43.0% while those who consumed ve or less (low dietary diversity score) were close to 57.0%. The mean DDS was 4.45±1.32. 95%CI=0.03 to 0.95). Nevertheless, this association when further analyzed, its signi cance was lost in the multivariate regression analysis.
As expected on the other hand, DDS was signi cantly associated with the nutritional status of the expectant mothers in the multivariate analysis (AOR= .4; 95% CI=1.05 to 5.3).

Discussions
The association of low dietary diversity score with the underweight status of expectant mother from rural Ethiopia was examined and showed the overall proportion of underweight to be less than a quarter and the majorities to have low DDS. Nonetheless, when the proportion of underweight is disaggregated against age, the most affected age group was those in the age group between 15 and 25 years. In the same breath, the proportion of underweight was signi cantly higher among mothers with low DDS indicating low dietary diversity is a signi cant factor associated with the nutritional status of mothers.
The magnitude of underweight observed in the present study is concordant with Northwest Ethiopia (23), Sri Lanka (24) and Kenya (25)  On the other hand, when compared with the Indian (61.5%) and South West Ethiopia (44.9%), our nding is expressively lower than the nding documented for India (26) and south west Ethiopia (27).
The difference might be due to that the study in India exclusively included participants from lower social classes. Further, the discrepancies might be attributed to the difference in stage of pregnancy, nutritional practice, a seasonal variation on food consumption and the use of different cut off points for mid-upper arm-circumference measurement in different studies.
The proportion of mothers with high DDS in our study is better than that of Pakistani report (80.4 vs 5.0) which documented only 5% of mothers to have high DDS (28).
In contrast, when compared with the Kenyan study our nding is lower (80.4 vs 92.4) (24). Such variations are expected due to the differences in the sociodemographic status and to some extent to the cultural reasons of the participants. The mean DDS(±SD) observed in our study is closer to the ndings of pregnant women reported mothers in Southwest Ethiopia (4.6 vs 4.3) (30), Bangladesh (4.6 vs 4.38) (31) and higher than the study nding from central Ethiopia (4.6 vs 3.5) (32). On the other hand, compared with the Kenyan which reported a mean DDS of 6.8 (33) and South Ethiopian mean DDS of 7.0 (34), the present gure is however lower than the above-mentioned ndings possibly due to their differences in socio-socioeconomic status, and cultural/seasonal variations.
As in many developing countries, this study showed that plant-based foods are the major sources of both macro and micronutrient for the vast majority of the mothers and were consumed in the form of breads, pasta, macaroni, porridge, rice followed by vegetables. While consumption of animal-based proteins which are good sources of nutrients is markedly low and consequently increased the risk of nutritional de ciencies. Similar ndings were reported in Kenya, southwest Ethiopia and Eastern Ethiopia (12,30,32,(33)(34)(35)(36).
In this study, less than a quarter of the expectant mothers were underweight based on the suggested cut off point of MUAC measurement of less than or equal to 22cm and most of the underweight cases were among the younger age group and our nding is concordant with some studies conducted in the country and elsewhere. For instance, in Northwest Ethiopia, Sri Lanka and Kenya (23)(24)(25)(26)(27)30), the magnitude of underweight among same groups was reported to be 14.4%, 15% and 19.3%, respectively. When compared with the Indian study (26), our nding is markedly lower (18.7 vs 61.5) and this difference is probably attributed to cultural and economic factors as most of the cases in India were from the lower social classes.
As expected, the odds of underweight were 4 times higher among mothers with poor DDS than their referent groups and our nding is concordant with some previous pocket studies conducted locally (30,36). It is eminent that when women consume less food groups, particularly low consumption of animalbased proteins which are good sources of nutrients, there would be an increased risk for malnutrition and similar nding is reported in west Ethiopia (30,36).

Strength and limitations
The strength of this study is that we used rural community that had no data collection fatigue and validated tool. Nonetheless, the study had some limitations emanated from the nature of study design which might led to recall bias.

Conclusion
This study demonstrated that low dietary diversity score is signi cantly associated with the observed level of maternal underweight. In order to improve the situations, it essential that all development sectors need to strengthen the current ongoing productive safety net program and empower women to earn more income that would enable them to purchase variety of foods to improve their nutritional status and ultimately break the intergeneration cycle of malnutrition. Other than this, mothers need to be counselled on the importance of the nutritional value of different food groups and its use. Ethical clearance and approval were obtained from Ethical approval committee of Debre Berhan University as well as from all relevant sectors of the zonal health o ces of the Amhara Regional state. Each study participant was adequately informed about the purpose, methods, and anticipated bene ts of the study by the data collectors. Respondents who were available and signed the written consent form at the time of data collection were interviewed without writing their names on the questionnaire, though code was used, to ensure con dentiality. In addition, underweight mothers with low DDS were counseled and linked to the PSNP through the zonal o ces. All data collection methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication
Not applicable.