Level and Predictors of Minimum Dietary Diversity Among Pregnant Women in Eastern Ethiopia: Evidence From Facility-Based Cross-sectional Survey

Abstract

Low dietary diversity is one of the most serious public health issues in developing countries, resulting in poor nutritional status, particularly vitamin and mineral deficiencies in pregnant women. However, there is insufficient information on the current status of pregnant women’s minimum dietary diversity in Eastern Ethiopia. The main aim of this study is to assess the level and predictors of minimum dietary diversity among pregnant women in Harar Town, Eastern Ethiopia. The study was conducted on 471 women using a health institution-based cross-sectional study design from January to March 2018. A systematic random sampling method was used to select the study participants. A pretested and structured questionnaire was used to collect data on the minimum dietary diversity. A logistic regression model was used to assess the relationship between the outcome variable and the independent variables. A P-value of .05 was used to indicate statistical significance. The proportion of pregnant women who had adequate minimum dietary diversity was 52.7% (95% CI: (47.9%, 57.6)). Urban residence [(AOR = 0.08, 95% CI: (0.02, 0.33)], smaller family size [(AOR = 7.28, 95% CI: (3.25, 16.28)], husband occupation [(AOR = 2.55, 95% CI: (1.39, 4.6)], husband support [(AOR = 3.85; 95% CI: (1.23, 12.02)], having more than 1 dwelling room [(AOR = 5.7, 95% CI: (2.32, 13.8) and medium wealth quantile [AOR = 1.93, 95% CI: (1.13.39)] were associated with adequate minimum dietary diversity. The level of minimum dietary diversity was low in the study area. It was linked to urban residency, smaller family sizes, husband employment, husband support, having more than 1 bedroom, and medium wealth quantile. Efforts should be made to improve husband support, wealth index, husband occupation, and food security status in order to boost mothers’ minimal dietary diversity.

• What do we already know about this topic?

• ● Adequate minimum dietary diversity has a positive effect on pregnancy outcome status

• ● Inadequate minimum dietary diversity leads to undernutrition and poor pregnancy outcome.

• How does your research contribute to the field?

• ● Achieving MDD is a challenging issue for pregnant women in developing country.

• ● Pregnant women had achieving poor minimum dietary diversity status, it needs further intervention and research findings.

• What are your research’s implications toward theory, practice, or policy?

• ● It will help future researchers to work on this related field to develop the existed theory

• ● This finding helps the government to design an appropriate nutrition intervention, policy, strategy and program related pregnant women nutrition.

• ● Pregnant women will achieve MDD for their daily consumption after the intervention has implemented.

Introduction

Dietary diversity (DD) is defined as the sum of different foods or groups of foods consumed over a specific time period, most commonly a day, week, or month. ¹ The DD is a qualitative measure of food consumption that reflects individuals’ access to a variety of foods and serves as a proxy indicator for nutrient adequacy. The dietary diversity score (DDS) quality assessment concept has been tested in a variety of population groups.^1,2

Pregnancy is a condition that develops within a woman’s body as a result of carrying a growing fetus. Extra nutrients are needed at this physiologically unique time to satisfy the needs of both the mother and the fetus. A frequent problem during this time is inadequate food intake, especially of micronutrients. ³ The most fundamental and crucial essential for the survival, well-being, and development of both the current and future generations is proper nourishment for women. The physiological changes that occur throughout pregnancy, immaturity, infancy, lactation, and old age cause variations in dietary demand. Women who are expecting must support not only themselves but also their unborn kid and extended relatives. ⁴

Previously, a reliable indicator called women’s dietary diversity scores (WDDSs), which was based on a 24-hour recall of food consumption across a variety of food groups, could be used to evaluate maternal micronutrient sufficiency and nutrition insecurity. ¹ There are 16 distinct food groups, according to the United Nations’ Food and Agricultural Organization (FAO). ¹ These dietary groups are reorganized into 9 primary food groups in order to evaluate women’s nutrient sufficiency. Grain, tubers, roots, flesh meat, legumes (beans, peas, lentils, and nuts), dairy products, organ meat (liver, kidney, etc.), eggs, vitamin A-rich fruits, dark-green leafy vegetables.

The minimum dietary diversity for women (MDD-W) indicator is a recent tool that can be used to measure women’s dietary diversity and quality. ⁵ The FAO^5,6 created this dichotomous indicator as a proxy indicator to evaluate women’s vitamin and mineral sufficiency. Using 10 major dietary groups, this indicator calculates the nutritional sufficiency of women. This list includes all starchy staples, beans and peas, nuts and seeds, flesh foods (including organ meat), all dairy products, eggs, other vegetables, other vegetables and fruits high in vitamin A, and other vegetables and fruits. Women who eat from 5 or more food groups are more likely to get the vitamins and minerals they need.

The developing fetus depends on adequate mother nutrition throughout the crucial 1000-day window from conception to a child’s second birthday. A woman will succeed for the rest of her life if she makes sure to eat well during this crucial time. because proper and sufficient nourishment during the first 1000 days of life determines the quality of a child’s future life. Without proper nourishment, the effects will worsen, and the impairment is frequently irreversible. ⁷

Existing data indicates that insufficient nutrition during pregnancy prevents more than 170 million children from growing up to their full potential. ⁷ Every year, 2.6 million children die from malnutrition. Millions of kids also suffer from lasting physical, emotional, mental, and cognitive problems. ⁷

Two billion individuals are thought to be suffering from chronic micronutrient deficiencies worldwide. ⁷ In many poor nations, especially in South Asia and Africa, maternal undernutrition including macro- and micronutrient deficiency is a serious public health offender. The prevalence of micronutrient deficiency is 38% globally, with 36% in the ocean, 27% in Latin America and the Caribbean, 24% in Europe, 36% in Asia, and 39% in Africa, according to several research. ⁸

Around the world, 32 million (38%) pregnant women were anemic, with South Asia having the highest prevalence (52%) of this condition. ⁹ The risk of abortion, fontal brain injury, congenital deformity, stillbirth, and prenatal death is increased in Ethiopia where 22% of pregnant women are anemic and 15% are iodine deficient. ¹⁰

The level of low dietary diversity was varying with time and place globally and locally. Numerous researches carried out in Asian nations found that pregnant women have insufficient dietary diversity score which fails in the range of 1% to 33%.^11-18 Similarly, the level of low dietary diversity varies from 2.4% to 53.9% in African country studies.^19-21 In addition, according to studies done in Ethiopia, pregnant women’s low dietary diversity ranges from 34% to 66.1%.^16,22,23

According to numerous scholars, the main causes of pregnant women’s low dietary diversity are socio-cultural and socioeconomic factors, cultural beliefs, women’s educational status, residence, husband support, husband occupation, household size, wealth index, presence of a dwelling room, food insecurity, maternal age, health service use, food restriction, not taking an additional meal, and skipping a meal.^2,9-12,22

Low dietary diversity during pregnancy has both short- and long-term effects, such as intrauterine growth restriction (IUGR), preterm birth, low birth weight, prenatal and infant mortality, and morbidity.^22–27 There has been an increased emphasis on women’s nutrition during pregnancy as a result of a sharp rise in micronutrient insufficiency over the past 2 decades. The necessity of consuming the right number of micronutrients extends beyond women’s health and wellbeing to the development and health of their offspring.^{28, 29}

According different national and international governments policies and programs, to prevent and control hidden hunger in women and their children, a 3-pronged strategy has been projected and is being used singly or in combination. These strategies are dietary diversity, supplementation, and food fortification for the short-, medium-, and long-term nutritional problems. ⁸

The MDD is essential for women, and it has been found to have the greatest impact on birth outcomes. There is a lack of information regarding the relative magnitude of MDD among pregnant women, despite the fact that several studies on DD in various groups are being carried out globally. Additionally, there is a dearth of studies that evaluate pregnant women’s MDD using this new indication (little dietary diversity for women). Determining the prevalence of MDD and its predictors among pregnant women in Harar Town, Eastern Ethiopia, is the goal of this article.

Methods and Materials

Study Settings

This study was conducted in health facilities of Harar Town, Ethiopia’s eastern region. According to the 2007 Census projection, the total population in Harari Town was 2 32 000, with females accounted for 115 230. ¹⁰ The Town has 4 health centers and 7 hospitals (4 government, 2 private, and 1 non-governmental). The region’s antenatal care coverage rate was 75.9%. ³⁰

Study Design and Period

The data collection period was from January to March 2018 using a health institution-based cross-sectional study design

Source Population

The source population consisted of all pregnant women aged 15-49 years who received ANC services in the selected health facilities during the study period.

Study Population

The study population of the study were selected pregnant women aged 15-49 years who received ANC services in the selected health facilities during the study period.

Inclusion Criteria

All pregnant women aged 15-49 years who received ANC services in the selected health facilities were included in the study.

Exclusion Criteria

Women with mental illness, severe illness, or who declined to participate, pregnant women who attended ANCs in Harar Town health facilities during the study period were excluded from the study.

Sample Size Determination Producers

This study has 2 objectives: (1) to determine the level of MDD and (2) to identify predictors of MDD. Based on these objectives, the sample size was calculated.

The sample size for the first objective of this study was calculated using the single population proportions formula, with 66% of prenatal women having DD, ²³ a 95% confidence level, and a 5% margin of error. As a result, the sample size considered for specific objective 1 was 345.

The sample size for the study’s second objective was calculated using a double proportion formula with 3 key independent variables (husband occupation, food insecurity and number of rooms) ¹⁸ considering the following assumptions. These are 95% 2-sided confidence levels, 5% marginal error, 80% power, a 1:1 ratio, and a respective odds ratio for each factor using Epi Info version 7 software. The largest sample size, 428, was chosen after calculating the required sample size for both objectives. Finally, with a 10% non-response rate, the final sample size was 471.

Sampling Methods and Procedures

There are 7 hospitals and 4 health centers in Harar. Two hospitals and 4 health centers were chosen from among these 7 hospitals and 4 health centers based on their antenatal care service provision (ANC). The calculated sample size was allocated proportionally to all nominated health facilities based on the number of ANC attended in the last 2 months of 2017. The review found that 760 and 620 pregnant women were being monitored at ANCs at Hiwot-Fana specialized University hospital and Jegul hospital, respectively. In addition, 120, 280, 440, and 100 pregnant women were being followed up on at ANCs in Aboker, Jenella, Arategna, and Amirnur, respectively. The sample size was then assigned proportionally for each hospital and health center based on their size using the proportional to population size (PPS) formula. The sampling frame was created using the ANC registration book, and subjects were chosen. Then, we used a systematic random sampling method to select the required study participants. Every k-value in each health facility was used to select the final participants (Figure 1).

Figure 1.

Schematic presentation of sampling procedure on dietary diversity and associated factors among pregnant women attending ANC in Harar town public health facilities, Eastern Ethiopia, 2018.

Methods of Data Collection and Quality Control Procedures

The questionnaire for this study was improved and modified for data collection. The English language questionnaire was adapted, modified, and structured based on different literatures.^1,56,22,23 It was interpreted into local languages (Amharic and Afaan-Oromo) and the reliability of the 2 forms was compared. The survey was pre-tested in nearby facilities to ensure its validity, and any necessary adjustments were made.

For the purpose of gathering information, 6 female bilingual diploma nurses were hired from the adjacent Harar town hospital and health center. Information gatherers and supervisors received a day of training on interview techniques, ethical considerations, respecting confidentiality, participant rights, and strategies to avoid underreporting before data collecting got underway. Participants in the study were asked to recall every food they had in the preceding 24 hours on the day of the interview.

The 24-hour dietary recall information was used to create a dietary diversity score. To aid in the memory of all eaten foods and alcoholic beverages, the enumerator probed respondents about the meals they had the day before and the prior evening. The enumerator also looks for the key components of the mixed dish. The prepared list included every consumed food and alcoholic beverage stated by respondents highlighted or crossed off. A great effort was made to avoid including days associated with extraordinary food consumption, such as holidays, fasting days, and special rituals. The responsibility for managing all operations fell to the chief investigator. To compare 2 data cells and minimize some differences, a double data entry was done.

Data Analysis and Processing Methods

Epi Data version 3.1 software was used to enter, code, and sanitize the data. The data were analyzed using the computer program Statistical Package for Social Science (SPSS) version 24. The sociodemographic characteristics of the respondents, womens healthcare practices, feeding practices, and other maternal health-related characteristics were described using straightforward descriptive statistics such as central tendency measures, variability measures, frequency distributions, and percentages.

The binary logistic regression model was used to analyze the association between the independent variables and the dependent variable using a bivariate regression analysis with 95% confidence intervals (CI) and crude odd ratio. Independent variables having a P-value of .25 or less were included in the multivariable analysis. Multicollinearity was examined using the standard error (SE). Variables with standard errors higher than 2 were thus removed. Hosmer-goodness Lemeshow’s of fit test model coefficient was used to assess the model’s fitness, and the results were non-significant (P = .598). To classify the predictors connected to the outcome variable, a multivariate binary logistic regression analysis model was used. The adjusted odds ratio at 95% confidence interval was obtained. The P-value was chosen as the threshold for statistical significance.

The outcome variable, the MDD-W, was categorized as either adequate or inadequate. The mean DDS was used to classify pregnant women’s DD as adequate (5 food categories out of 10 food groups) or inadequate (5 food groups out of 10 food groups). Each pregnant woman’s MDD Score was calculated based on her diet in the previous 24 hours (5 food groups out of 10 food groups). High MDD was recorded as “1” and low MDD as “0” during the analysis.

According to the Coates et al ³¹ reports, households were classified as food secure or not using the household food insecurity access scale (HFIAS). During analysis, the responses were summed from 0-27 points to determine the HFI status. Then, all “yes” responses were coded “1,” and all “no” responses were coded “0.” Furthermore, during the analysis, HFI was classified as either food secure (coded as “1”) or food insecure (coded as “0”) households.

On the other hand, the family wealth index was analyzed using the principal component analysis (PCA) method, which took into account locally available household assets like having domestic animals (such as cattle, ox, cow, sheep, and goat), durable assets (such as television, refrigerator, telephone, car, living house, and land), productive assets (such as plough, axe, hoe, shovel, and sickle), and housing materials (such as sofa, bed, table, chair, and stove). Furthermore, the WI was divided into 3 terciles: poor, medium, and rich, and coded as “1,” “2,” and “3,” respectively.

Ethical considerations

The Institutional Health Research Ethics Review Committee (IHRERC) of College of Health and Medical Sciences in Haramaya University was granted the Ethical approval. The ethical approval letter was numbered Ref C/AC/R/D/1021/18 and dated 25 December 2018. Before informed consent was obtained, a clear description of the study title, procedure, and duration, possible risks, and benefits of the study was explained for each study participant. Then oral and written consent was taken from each study participant. In addition, they were confirmed that information collected from each study participant was secured and confidential.

Results

Sociodemographic Features of Respondents

This result shows that 457 pregnant women between the ages of 15 and 49 replied to the survey, giving a response rate of 97.03%. The mean (±SD) age of the respondents was 25.7 (±4.7) years, and nearly two-thirds (59.5%) of them were older than 25. City residents made up the majority of participants (83.2%), and Muslims made up 65.6% of the participants. A total of 155 (33.9%) women have earned a college degree or higher. Additionally, 124 (27.1%) of the women’s husbands were employed by the government, while 181 (39.6%) of the women were housewives (Table 1).

Table 1.

Sociodemographic Characteristics of Prenatal Women (n = 457) Attending ANC in Harar Town Public Health Facilities, Eastern Ethiopia, 2018.

Variable	Category	Frequency	Percentage (%)
Age of women (in years)	15-19	37	8.1
	20-24	148	32.4
	≥25	272	59.5
Residence of women	Urban	380	83.2
	Rural	77	16.8
Religion of women	Muslim	300	65.6
	Orthodox	107	23.4
	Protestant	40	8.8
	Catholic	9	2
	Other*	1	0.2
Women educational	No formal education	68	14.9
Level	Grade 1-8	115	25.2
	Grade 9-12	119	26
	College and above	155	33.9
Women occupation	Housewife	181	39.6
	Merchant	45	9.8
	Daily laborer	45	9.8
	Government employment	115	25.2
	Self-employed	52	11.4
	Others**	19	4.2
Husband occupation	Merchant	100	21.9
	Farmer	53	11.6
	Government employment	124	27.1
	Daily laborer	98	21.4
	self-employed	74	16.2
	Others***	8	1.8
Family size	<3	286	62.6
	04-May	105	23
	>5	66	14.4
Heard about food restriction	Yes	80	17.5
	No	377	82.5
Type of restriction food(n = 80)	Porridge	29	36.25
	Banana	9	11.25
	Butter	4	5
	Raw meat	15	18.75
	Others****	6	8.75
Had dwelling room	Yes	392	85.8
	No	65	14.2
Number of doweling rooms (n = 392)	1	339	86.4
	>1	53	13.6
Husband support	Yes	426	93.2
	No	31	6.8
Family wealth index	Poor	179	39.1
	Medium	141	30.9
	Rich	137	30
Method of husband support	By giving money	314	73.7
	By purchasing food	20	4.69
	By farming	29	6.8
	By farming and giving money	13	3.08
	By giving money and purchasing food	50	11.73

Note. Others^* = traditional; ^** = students, unemployed; ^*** = students, unemployed; ^**** = eggs, tomato.

More than three-fifths of women (63%) said they had fewer than 3 close relatives. The vast majority of expectant mothers 377 (82.5%) were not aware of any culturally taboo foods. Twenty-nine (36.25%) of those who were aware that some meals were culturally taboo during pregnancy said they avoided porridge because of the challenges of having a macrosomic baby and weight gain. About 392 respondents, or 85.8%, reported having a living room, and 339 (86.4%) reported having just 1. Four hundred twenty-six (93.2%) women were supported financially by their husbands, and 314 (73.7%) of the men supported their spouses financially. According to the family wealth index, respectively, 30.9% of households were medium-wealth, 30.9% were rich, and 39.1% were poor (Table 1).

Women Feeding Practice and Healthcare-Related Characteristics

Two hundred forty-six (53.8%) of all respondents had 2 to 3 kids, and 273 (59.7%) had their first prenatal care (ANC) appointments. About 397 (86.9%) of the pregnant women made food purchases. A 80.1 percent of the 366 women said they ate 3 or more times daily. Vegetables were consumed once or twice a day by more than half of the respondents (240), whereas fruits were ingested once or twice a day by 178 respondents (38.9%). In terms of women’s medical issues, 22.5% said they had illnesses during their pregnancies, and 86.4% said they had to skip meals as a result. Dyspepsia was reported by 97 people (94.3%), diabetes by 3, and hypertension by 2 people (1.9%). The majority of the pregnant women who took part in the study (82.7%) had sufficient knowledge of nutrition (Table 2).

Table 2.

Women Health and Feeding Practice-Related Characteristics of Pregnant Women (n = 457) Attending ANC in Harar Town Public Health Facilities, Eastern Ethiopia, 2018.

Variable	Category	Frequency	Percentage
Number of children	No child	122	26.7
	01-Feb	246	53.8
	>3	89	19.5
ANC visit	First visit	273	59.7
	Second visit	173	37.9
	Third or more visit	11	2.4
Source of food	Purchased food items	397	86.9
	Own production	48	10.5
	Purchased and own production	6	1.3
	Other	6	1.3
Number of meals per day	<2	35	7.7
	2	56	12.2
	3+	366	80.1
Number of vegetables consumed per day	Not consumed	51	11.2
	01-Feb	240	52.5
	3+	18	3.9
	Sometimes	148	32.4
Number of fruits consumed per day	Not consumed	92	20.1
	01-Feb	178	38.9
	3 +	4	0.9
	Sometimes	183	40
Women knowledge of nutrition	Poor	79	17.3
	Good	378	82.7
Ongoing medical condition /disease	Yes	103	22.5
	No	354	77.5
Skipping meals due to disease (n = 103)	Yes	89	86.4
	No	14	13.6
Type of disease reported (n = 103)	Dyspepsia	97	94.3
	Diabetes mellitus	3	2.9
	Hypertension	2	1.9
	Others	1	0.9

Household Food Security Status

In this finding, 79% of respondents to this poll reported having food security, compared to 21% who did not. A high MDD was indicated by 39.4% of respondents who said they were food secure. Contrarily, 13.1% of respondents with significant dietary diversity reported light (1%), moderate (5%), or severe (15%) food insecurity (Figure 2).

Figure 2.

Household food security levels status and high dietary diversity among the study participants (n = 457) attending ANC in Harar town public health facilities, Eastern Ethiopia, 2018.

Level of MDD in Pregnant Women

In this study, a high level of MDD was present in 241 (52.7%) of the pregnant mothers. All mothers said they consumed an all-starch staple, and 63.9% said they consumed legumes, Injera with Shiro made from grains like teff, bread made from the nutrient-rich grain teff, and a regional stew based on legumes. Dairy, meat, poultry, and fish accounted for 42.0%, meat, poultry, and fish accounted for 36.5%, and eggs accounted for 20.4% of the animal products consumed by women. About 36.5 eaten additional nutritionally dense fruits and vegetables, and nearly all respondents (95.2%) used other veggies in food preparation. The least consumed fruit in this survey was other fruits (2.8% of respondents; Figure 3).

Figure 3.

Food groups of dietary diversity of pregnant women (n = 457) attending ANC in Harar town public health facilities, Eastern Ethiopia, 2018.

Predictors of MDD Among Prenatal Women

In a bivariate regression model (P = .25), the following variables were selected: residence, women’s educational level, husband occupation, husband support, family size, family wealth index, number of dwelling rooms, current ANC visit, household food security status, source of food, and food restriction. These variables were also taken into account in a multivariable logistic regression analysis model. At the 5% level of significance, it was discovered that the home, husband’s occupation, family size, having more living rooms, husband support, family wealth index, and food insecurity in the home were statistically significant predictors of high MDD (Table 3).

Table 3.

Factors Associated With Dietary Diversity Among Pregnant Women (n = 457) Attending ANC in Harar Town Public Health Facilities, Eastern Ethiopia, 2018.

Variables	Minimum dietary diversity		COR (95% CI)	AOR (95% CI)
	High (%)	Low (%)
Residence of women
Rural	62 (80.1)	15 (19.5)	1.00	1.00
Urban	179 (47.1)	201 (52.9)	0.21 (0.11, 0.39) **	0.08 (0.02, 0.35) **
Educational status of women
Non formal	44 (64.7)	24 (35.3)	1.00	1.00
Formal	197 (50.6)	192 (49.4)	0.56 (0.32, 0.96) **	0.94 (0.42, 2.13)
Husband occupation
Non business	172 (48.2)	185 (51.8)	1.00	1.00
Business	69 (69.0)	31 (31.0)	2.39 (1.49, 3.83) **	2.55 (1.39, 4.6) **
Family size
<3	165 (57.5)	121 (42.3)	1.85 (1.07, 3.18) **	7.28 (3.25, 16.28) **
4-5	48 (45.7)	57 (54.3)	1.14 (0.61, 2.12) **	1.84 (0.78, 4.31)
>5	28 (42.4)	38 (57.6)	1.00	01:00
Heard about food restriction
Yes	49 (61.2)	31 (38.8)	1.00	1.00
No	192 (50.9)	185 (49.1)	0.65 (0.4, 1.07) **	1.58 (0.82, 3.03)
Had dwelling room
1 doweling room	173 (51.0)	166 (49.0)	1.00	1.00
>1 dwelling room	44 (83.0)	9 (17.0)	4.69 (2.22, 9.91) **	5.7 (2.32, 13.98) **
Husband support
Yes	231 (54.2)	195 (45.8)	2.48 (1.14, 5.4) **	3.85 (1.23, 12.02) **
No	10 (32.3)	21 (67.7)	1.00	1.00
Current ANC follow-up
First visit	124 (45.4)	149 (54.6)	1.00	1.00
Second visit	109 (63.0)	64 (37.0)	2.04 (1.38, 3.02) **	1.39 (0.82, 2.34)
≥Third visit	8 (72.7)	3 (27.3)	3.20 (0.83, 12.33) **	2.64 (0.5, 13.98)
Source of food
Own production	35 (72.9)	13 (27.1)	2.65 (1.36, 5.16) **	0.58 (0.15, 2.15)
Others	206 (50.4)	203 (49.6)	1.00
Household food insecurity
Food secure	180 (49.6)	183 (50.4)	0.53 (0.33, 0.85) **	0.37 (0.17, 0.77) **
Food insecure	61 (64.9)	33 (35.1)	1.00	1.00
Family wealth index
Poor	79 (44.1)	100 (55.9)	1.00	1.00
Medium	78 (55.3)	63 (44.7)	1.56 (1.0, 2.44) **	1.93 (1.1, 3.39) **
Rich	84 (61.3)	53 (38.7)	2.0 (1.27, 3.15) **	2.12 (1.18, 3.81) **

Note. CI = confidence interval; COR = crude odds ratio; AOR = adjusted odds ratio; ANC = antenatal care, 1.00 = reference category.

The bold text in the table indicated the significant and highly significant factors of the study.

^*P < .05.

^**P ≤ .001.

Antenatal women in cities were 92% less likely to have high DD compared to those who lived in rural areas [(AOR = 0.08, 95% CI: 0.02, 0.35)]. In comparison to husbands without a job, those with business-related employment had 2.55 times more frequently MDD [(AOR = 2.55, 95% CI: 1.39, 4.6)]. In comparison to women with more than 3 family members, those with less than 3 had more than 7 times higher MDD [(AOR = 7.28, 95% CI: 3.28, 16.28)]. Nearly 6 times as many women with MDD had more than 1 living room compared to just 1 room [(AOR = 5.7, 95% CI: 2.32, 13.98)] (Table 3).

Women who had husband support during pregnancy were nearly 4 times compared to women who did not receive husband assistance during pregnancy (AOR = 3.85, 95% CI: 1.23, 12.02). Access to food during pregnancy was associated with a 63% lower risk of high MDD [(AOR = 0.37; 95% CI: 0.17, 0.77]. The family wealth index is another element connected to pregnancy in this study. High MDD was 1.93 and 2.12 times more likely to be present for people with medium and highest wealth indices, respectively [AOR = 1.93; 95% CI (1.1, 3.39), AOR = 2.12; 95% CI (1.18, 3.81; Table 3).

Discussion

In the current study, the level of high MDD among pregnant women was 52.7% (95% CI: 47.9, 57.6). Women’s MDD was significantly correlated with factors such as residence, having a husband who worked in the business world, having less than 3 family members, having more than 1 bedroom, having the support of the husband, being in a household where food security was a priority, and having a rich or middle-class family.

The level of MDD in this study is consistent with the level in Ghana (50%). ²¹ In comparison, it is lower in Nepal (74.4%), ¹² Malaysia (87.3%), ¹⁵ India (77%), ¹⁷ Kenya (60.6%), ²⁰ and Wondogenet, Ethiopia (66%). ²² This disparity could be attributed to geographic differences that affect food accessibility, socioeconomic status, seasonal and temporal factors, and methodological differences. Differences in outcome variable tool of measurement could also explain the observed difference, as the current study employs the recently endorsed indicator, which is a dichotomous indicator.

In the current study, however, all respondents (100%) consumed all starchy staple foods. This is much higher than a study reported from India (82.5%) ¹⁷ and much closer to study conducted in Kenya (99.2%). ²⁰ However, this is significantly higher than the study conducted in Wondogenet, Ethiopia (41.8%). ²² This inconsistency could be attributed to differences in the study participants’ cultural diets as well as geographic variation in cereal growth.

The current study discovered that the residence of pregnant women was substantially related to MDD in terms of predictors. Compared to pregnant women in rural settings, city dwellers had a 92% lower risk of MDD. The findings of the Ghana research ²⁴ are in conflict with this finding. This variation might be explained by variations in the research environment as well as sociocultural variations between the 2 study populations. Another theory is that people who live in rural areas can grow their own food, and that these areas are also where produce is mass-produced for the market and for individual use.

In this study, it was discovered that the MDD and family size were substantially correlated. The probability of having a greater MDD were more than 7 times higher for antenatal women with fewer than 3 family members compared to those with more than 3 family members. This result goes against research from Bangladesh. ¹⁸ Due to the higher number of dependents in large families compared to small families, the opposite result may be related to the socioeconomic position of the households.

It has been proposed that the MDD of husbands’ pregnant spouses is related to their line of work. Mothers with business-related husbands had more than twice as much MDD as those without such a husband. This result is analogous to studies done in Bangladesh. ¹⁸ This might be the case since men who work in business typically earn more money and can afford to purchase a wide range of nutrient-dense foods.

In this study, women who received support from their husbands had more than twice as much MDD as mothers who did not receive support from their husbands. A Pakistani study ¹¹ found that pregnant women who received a weekly budget from their husbands had 4 times the odds of having a high MDD index compared to women who did not. One possible explanation for this is that women who are supported by their husbands are more likely to be able to buy food for themselves and may choose to buy more varied and diverse foods.

Women who lived in food secure households had a lower risk of developing low MDD. Food insecure pregnant women were 63% less likely to have high MDD. This finding contradicts the findings of the Bangladesh study. ¹⁸ Even if participants were food secure in this study, they may not have used a diverse feeding style, which requires further investigation.

In terms of the study participants’ household wealth index, pregnant women from average and high wealth indices had more than 2 and nearly 2 times higher MDD than their counterparts, respectively. This finding was consistent with previous research from India ¹⁷ and Ghana. ² The most likely explanation is that increased income is associated with increased purchasing power, allowing households to purchase more diverse foods and thus promote dietary diversity.

Pregnant women who had more than 1 living room were nearly 6 times more likely to have high MDD. This is consistent with the research conducted in Bangladesh. ¹⁸ This could be related to the fact that more living rooms lead to an increase in their socioeconomic status by renting those rooms and earning extra money to buy different food groups. Furthermore, a higher income allows families to buy a wider variety of foods.

As a result, the Ethiopian government, Harari Regional Health Office, and other stakeholders should provide food vouchers to assist women in purchasing more diverse foods, encourage local communities to grow their own foods, raise awareness of the importance of DD in pregnancy, and encourage mothers to spend what little money they have on a better diet, as well as educate husbands on the importance of purchasing diverse foods for their wives. Finally, because DD is multifactorial, more research is needed to identify other independent variables among antenatal women.

Ethiopia’s government should make plans to reduce undernutrition, particularly micronutrient deficiencies caused by low MDD among pregnant women in the country. This research will help the country, specifically the study area, prepare special nutrition education and counseling (NEC) on dietary diversity for pregnant women during ANC follow-up. It will also help policymakers, administrators, other concerned organizations, and stakeholders in the study area pay more attention to low MDD for pregnant mothers. This helps to reduce inadequate MDD during pregnancy and adverse birth outcomes. Furthermore, this study may be useful as a reference for other researchers conducting similar research in the future.

The strength of the study was collecting the data for this study in a secure and private room to get appropriate data, using appropriate model of analysis and report writing system. Concerning the study’s limitations, the current study may have an inherent limitation, namely, recall bias, because some of the questions asked about events that occurred between 24 hours and 1 month ago. This bias was reduced by asking study participants about an event that had already occurred. Reporting bias can be other limitations of the study. It was minimized by probing about the detail of the question.

Conclusions

The results of this study showed that only half of pregnant women met the MDD score. This indicated that the remaining half of pregnant women were not achieved MDD in the study area. Regarding the factors associated with MDD, the following factors were contributory variables. Women who had a higher MDD score were associated with husbands who had business owners, had made more money, had fewer families, had higher household wealth indices, had assistance from their spouses, had food security at home, and had lived in a rural location.