Individual and community level factors associated with anemia among children 6-59 months of age in Ethiopia: A further analysis of 2016 Ethiopia demographic and health survey.

BACKGROUND: Anemia is a global public health problem; but its burden is disproportionately borne among children in the African Regions. The 2016 Ethiopia Demographic and Health Survey report showed that the prevalence of anemia among children 6-59 months of age was 57%; far exceeding the national target of 25% set for 2015. Although studies have been conducted in Ethiopia, multilevel analysis has rarely been used to identify factors associated with anemia among children. Therefore, this study aimed to identify individual and community-level factors associated with anemia among children 6-59 months of age by fitting a multilevel logistic regression model.
METHODS: The data was obtained from the 2016 Ethiopia Demographic and Health Survey, conducted from January to June 2016, and downloaded from the website http://www.DHSprogram.com. The sample was taken using two-stage stratified sampling. In stage one, 645 Enumeration Areas and in stage two 28 households per Enumeration Area were selected. A sample of 7790 children 6-59 months of age was included. Data were analyzed using STATA version 14. A multilevel logistic regression model was fitted and an adjusted odds ratio with a 95% confidence interval was obtained. RESULT: From the individual-level factors, anemia was associated most strongly with child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order whereas from the community-level, the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa. Low community-poverty is a protective factor for anemia. The odds of anemia were 0.81 (95% CI: 0.66, 0.99) times lower for children who were living in communities of lower poverty status than children who were living in communities of higher poverty status. Children from Somali and Dire Dawa had 3.38 (95% CI: 3.25, 5.07) and 2.22 (95% CI: 1.42, 3.48) times higher odds of anemia, respectively than children from the Tigray region.
CONCLUSIONS: This study shows that anemia among children 6-59 months of age is affected both by the individual and community level factors. It is better to strengthen the strategies of early detection and management of stunted and underweight children. At the same time, interventions should be strengthened to address maternal anemia, child fever and poverty, specifically targeting regions identified to have a high risk of anemia.

Introduction

Globally, anemia affects around 800 million children and women; of which 42.6% are children [1]. African countries have the highest prevalence (62.3%) followed by South-East Asia (53.8%) and Eastern Mediterranean Region (48.6%) [1]. Furthermore, it is a major public health problem in Sub-Saharan African countries with high national prevalence estimated to be above 40% [2].

According to the 2011 Ethiopia Demographic and Health Survey (EDHS), anemia prevalence among children 6–59 months of age was 44% [3]. In addition, the 2016 EDHS report showed that prevalence of anemia among children 6–59 months of age was 57% [4]. This indicates that anemia is a severe public health problem which increased by 13% within five years period, during a period when the government of Ethiopia undertook efforts such as vitamin ‘A’ supplementation, deworming, and use of fortified foods to reduce anemia through national nutrition programs [5].

Anemia is an important indicator of poor nutrition and health with major consequences of socioeconomic development [6]. Children younger than two years of age with severe anemia are at increased risk of mortality and, even mild forms, which might be corrected cause permanent cognitive damage by decreasing attention span and shortening of memory [7].

Although biochemical and hematological tests exist, hemoglobin concentration in the blood is the most reliable indicator of anemia at the population level [8]. This latter method was used to assess for anemia among children aged 6–59 months in the 2016 Ethiopia Demographic Health Survey (EDHS) [4].

According to the WHO criteria, anemia in children 6–59 months of age is defined as hemoglobin concentration in the blood below 11g/dl [9]. Anemia is said to be a severe public health problem when its prevalence is 40% or more, a moderate public health problem when its prevalence is between 20 and 40% and a mild public health problem when its prevalence is between 5 and 20% in any group [10].

In Ethiopia, previous studies have linked anemia to factors such as child age, child nutritional status, parents’ educational level, and wealth index [11-15]. However, almost all studies used single-level analysis techniques with population groups localized in a specific study area [11, 12, 14, 15]. The Single-level analysis assumes that there is no community effect beyond the characteristics of individuals [16]. That is, the impact of community-level factors on anemia among children aged 6–59 months remains under-studied. Moreover, analyzing hierarchical data like the DHS using single-level analysis leads to incorrect estimation of parameters and standard errors [17].

Using multilevel analysis technique, community-level effects can be identified from individual-level effects [18-21]. However, this approach has rarely been used in Ethiopia to identify factors associated with anemia among children. One study that used multilevel-analysis technique failed to examine the effect of some factors such as poor nutritional status of children (stunting, wasting and underweight), child health-related factors (fever, diarrhea and respiratory infection), maternal anemia, and variables aggregated at the community level [18].

Finally, the aims of this study were:

To identify individual level factors associated with anemia among children 6–59 months of age in Ethiopia.

To identify community level factors associated with anemia among children 6–59 months of age in Ethiopia.

Methods and materials

Data source

Data were extracted from the nationally representative 2016 EDHS. The 2016 EDHS is the fourth survey which is implemented by the Central Statistical Agency (CSA) in collaboration with the Ethiopian Ministry of Health under the technical assistance of International Classification of Functioning, Disability, and Health (ICF) through the DHS Program. Ethical approval was obtained from Mekelle University, College of Health Sciences Ethical Review Committee (ERC). Approval to access the 2016 EDHS data set was obtained from DHS Program, after making a request via DHS program website (http://www.DHSprogram.com). The EDHS data has no individual identifiers which could affect the confidentiality of participants and the data was used for analysis purpose only.

Study population

Children 6–59 months of age who were living in selected enumeration areas.

Inclusion and exclusion criteria

The inclusion criteria were children 6–59 months of age who live in the selected enumeration areas (community). And Exclusion criteria were children 6–59 months of age who have no hemoglobin test result.

Study design and sample size

A population-based cross-sectional survey was used to collect the 2016 EDHS data. The 2016 EDHS had used a stratified two-stage cluster sampling design. Stratification was achieved by separating each region into urban and rural areas, yielding 21 sampling strata. In the first stage, 645 Enumeration Areas (EAs) or clusters were selected. Among the selected 645 EAs, 202 were in urban and 443 in rural areas. In the second stage, households were the sampling units and a fixed number of 28 households per each EA were selected. From the total of 10,641 under-five years old children, 9504 were children 6–59 months of age. Data on hemoglobin level from the survey were available for 7790 children.

Definitions of study variables

We assessed the impact of individual and community-level variables on anemia among children 6–59 months of age. We defined anemia in 6–59 month age children as hemoglobin <11 g/dL according to WHO criteria [8]. Individual level variables were: sex, age, birth order, birth weight, religion, number of under-five children, childhood wasting, underweight, stunting, symptoms of acute respiratory infection, child fever and diarrhea, maternal anemia and age, parents’ educational and employment status, wealth index, source of drinking water, and type of toilet facility, whereas community-level variables were: region, community-poverty, community-women education and community- women unemployment.

Wealth index is a composite measure of a household’s cumulative living standard. It was calculated based on household ownership of selected assets such as televisions and bicycles, cars; materials used for the housing construction; source of drinking water; and type of sanitation facilities. It was then generated using principal components analysis and the individual households were placed on a continuous scale of relative wealth. In the EDHS all mothers and children were assigned a standardized wealth index score. It was measured as a composite variable made up of five quintiles as poorest, poorer, middle, richer and richest [4].

Anthropometrics: stunting was defined as height or length for age (HFA) <-2SD (standard deviation), wasting as weight-for-height (WFH) <-2 SD and underweight as weight-for-age (WFA) <-2 SD.

The community-level variables that directly measure the community characteristics in the 2016 EDHS were the place of residence (rural or urban) and region (either of the nine regions or the two administrative cities). We created also other additional variables by aggregating the individual level’s characteristics within their respective clusters. These variables were: community poverty, community women education and community women unemployment.

Community-poverty is the proportion of mothers who reside in poor or poorest households in the community. The aggregate of the poorest or poor individuals can show the overall poverty of the cluster. For this proportion, the median value was calculated as summary statistics and categorized as ‘More poverty’ or ‘less poverty’ based on this national median value.

Community-women education (CWE) is proportion of mothers aged 15–49 with secondary or higher education in the community. The median value was calculated as summary statistics that represent the educational status of women in the cluster and was categorized as ‘High’ or ‘Low’ based on the national median value.

Community-women unemployment status (CWUe) is proportion of mothers aged 15–49 who were not employed in the community in the past twelve months. It was categorized as high if clusters had more than or equal to the national median proportion of unemployed mothers or low otherwise.

Methods of data analysis

Before doing any analysis, sampling weight and normalization were done for the sample in order to ensure the representativeness of the sample to different regions and their place of residence. Data were analyzed by Stata version 13 and a multilevel binary logistic regression model was fitted. Frequencies, percentages, graphs and charts were used to describe categorical variables. The effect of each predictor variable on the outcome variable was checked at a significance level of p≦0.25 independently [22]. Variables that are statistically significant at the bivariate multilevel logistic regression analysis were considered as candidates for multivariable analysis. Accordingly, in the multivariable analysis the following variables were adjusted and controlled: number of <5 children in the household, child age, religion of mother (caretaker), birth order, parents employment status and educational level, maternal age, type of toilet facility, source of drinking water, wealth index, child stunting, wasting, underweight, fever, diarrhea, child deworming, symptoms of acute respiratory infection, maternal anemia, community- women education, place of residence, community-poverty, region and community- women unemployment. Adjusted Odds Ratio (AOR) with 95% Confidence Interval (CI) at a significance level of p<0.05 was estimated. The result of multivariable analysis for individual and community-level factors associated with anemia among children aged 6–59 months is shown in Table 3.

Table 3

Individual and community-level factors associated with anemia among children age 6–59 months selected from the 2016 EDHS.

Variables	Anemia Status		AOR [95%CI]	P-value
	Frequency (%)
	No	Yes
Individual level factors
Child age (months)
6–23	746.5 (28)	1921 (72)	4.45 [3.62, 5.36]	0.000
24–41	1116 (43)	1481 (57)	1.83 [1.59, 2.09]	0.000
42–59	1444 (57)	1082 (43)	1
Wealth index
Poorest	583 (32)	1245 (68)	1.51[1.11, 2.04]	0.035
Poorer	770 (42)	1058 (58)	1.29 [0.98, 1.72]	0.052
Middle	772 (46)	901 (54)	1.11 [.84, 1.46]	0.296
Richer	639 (45)	776 (55)	1.19 [.90, 1.55]	0.106
Richest	542 (52)	504 (48)	1
Birth order
1	602 (44)	760 (55.8)	1
2–3	1053 (43)	1365 (56.5)	1.08[.91, 1.28	0.112
4–5	779 (42)	1100 (58.4)	1.16 [.94, 1.43]	0.062
> = 6	873 (40.8)	1268 (59.2)	1.26 [1.00, 1.61]	0.044
Child underweight
No	2591 (45)	3198. (55)	1
Yes	701 (36)	1262 (64)	1.34 [1.14, 1.57]	0.000
Child stunting
No	2052 (46)	2438 (54)	1
Yes	1223(39)	1953 (61)	1.4 [1.24, 1.59]	0.000
Maternal anemia
No	2563 (48)	2818 (52)	1
Yes	723 (31	1597 (69)	1.42 [1.21, 1.55]	0.000
Child fever
No	2886 (44)	3719 (56)	1
Yes	420 (35)	764 (65)	1.32[1.09, 1.60]	0.004
Community level factors
Community- poverty
More Poverty	1020 (35)	1925 (65)	1
Less poverty	2287 (47)	2558 (53)	0.81[.66, .99]	0.032
Region
Tigray	242 (46)	284 (54)	1
Afar	19 (25)	57 (75)	1.66 [1.11, 2.48]	0.001
Amhara	873 (57)	647 (42)	0.71 [0.52, 0.96]	0.089
Oromia	1170 (34)	2,248 (66)	1.62 [1.16, 2.26]	0.000
Somali	54 (17)	267 (83)	3.38 [3.25, 5.07]	0.000
Benishangul	47 (57)	36 (43)	.58 [.40, .82]	0.000
SNNP	804 (49)	833 (51)	1.00 [0.71, 1.42]	0.113
Gambela	8 (43)	10 (57)	1.20 [0.81, 2.78]	0.052
Harari	5 (33)	10 (67)	1.88 [1.23, 2.88]	0.000
Addis Ababa	76 (51)	72 (49)	1.54[1.01, 2.35]	0.001
Dire Dawa	8 (28)	21 (72)	2.22 [1.42, 3.48]	0.000

Key: 1- reference category, AOR- Adjusted Odds Ratio, CI- Confidence Interval, SNNPR- Southern Nations, Nationalities and People’s Region.

Assuming varying intercepts across communities (clusters) but fixed coefficients, four models were developed. The first one was the null model which is fitted without independent variables; this was used to determine the variance in anemia status between the clusters in the sample. Whereas, model one was adjusted for individual-level factors and used to examine their contribution to the variation of anemia status. Model two was adjusted for community level factors and was used for examining whether the community-level variables explain between-cluster variation on childhood anemia. Model three was developed by combining both the individual and community level variables.

Null model

For individual i in community j, the model can be represented as [17, 23]: Where:

Y is anemia status of ith child in the jth cluster

ϒ = is the intercept; that is the probability of having anemia in the absence of explanatory variables

u = community-level effect; ε error at individual level

Mixed model

This model was derived by mixing both individual and community level factors simultaneously [24]. Where: The term γ is the regression coefficient of the individual-level variable X and γ is the regression coefficient of the community-level variable Zp. X and Zp were individual and community-level explanatory variables respectively. The subscripts i and j represent for the individual level and cluster number respectively.

Intraclass correlation coefficient (ICC)

A measure of within-cluster homogeneity and the proportion of variance due to between-cluster differences. Where: = between cluster (community) variances and = with in cluster (community) variance. The value of in case of standard logistic distribution is 3.29 [25].

The null model showed that there was a significant variation in anemia status among clusters (δ2u0 = 0.76, p-value<0.001). The ICC was 18.77% (95% CI; 0.1598, 0.219), meaning that 18.77% of the total variability in odds of anemia was due to between community differences or attributable to the unobserved factors either at community-level or at individual-level. This indicates that using a multilevel logistic regression model is better for getting valid estimates than single-level logistic regression [17]. The variance which was due to the clustering effect decreased from 18.77% in the null model to 10.03%, 8.04%, and 7.03% in model one, model two and model three, respectively (S1 Annex).

Proportional change in variance (PCV)

Calculated with reference to the null model to see relative contribution of factors to explain variation in childhood anemia. Where: is between community variance in the null model; is between community variance in the consecutive models [17].

Model three had the highest PCV which is 67%. This shows that 67% of the variance in the anemia status among children was due to the simultaneous effect of both individual and community-level factors found in the model (S1 Annex).

Model diagnostics and adequacy checking

Multicollinearity was checked by using a variation inflation factor (VIF) with a cut-off point of 10 [26]. It was checked for the independent variables in the final model and the VIF was found to range from 1.2 to 4.2 with mean VIF of 2.1. This shows that multicollinearity might not be a problem.

Interaction between variables was checked for those variables found significant at the final model. As a result, there were significant interactions between these variable (p<0.05). However, as we examined the interaction effect by fitting regression models that contained interaction terms yields no significant (p>0.05) interaction effect. Model selection was carried out by using Akaike information criteria (AIC). AIC values for each model were compared and the model with the lowest value of AIC was considered as a better explanatory model [25]. Accordingly, model 4 with AIC value of 8281 was selected as the best model for explaining anemia status among children aged 6–59 months in Ethiopia (S1 Annex). Model accuracy was checked by the area under the curve. The Receiver Operating Characteristics curve (ROC-curve) provides a measure of the model’s ability to discriminate between those subjects who experience the outcome of interest versus those who do not [27]. The area under the ROC of this model was 0.7376; which means the ROC curve accuracy for outcome variable (anemia) was 74% (Fig 1).

Fig 1

Receiver operating characteristics curve for anemia of included children age 6–59 months selected from the 2016 EDHS (n = 7790).

Results

Individual-level characteristics of study subjects

From the total of 10,641 under-five years’ old children, 9504 were children 6–59 months of age. Data on hemoglobin test result from the survey were available for 7790 children. As a result, 1714 children aged 6–59 month were excluded from the study due to missing data of hemoglobin test result. In addition, the variables of dietary intake and child feeding practices were not included due to missing value. These variables were missing for nearly half of observations.

Seven thousand seven hundred ninety (7790) children 6–59 months of age were included in this study. Above half (52%) of the children were male, and 34.2%, 33.3% and 32.2% where in the age category of 6–23, 24–41 and 42–59 months of age, respectively with mean ± SD (standard deviation) of 32 ±15 months. The prevalence of anemia was 57.6% with a median hemoglobin concentration of 10.7 (IQR: 9.6–11.6).

Almost all of the mothers (95%) were living with their respective partners and most of them were Muslims (40%) followed by Orthodox Christians (34%). Nearly half of the mothers (48.7%) were 20–29 years of age. The proportion of no formal education among the children’s mothers (67%) was higher than their fathers (48.5%) (Table 1). About one-fifth (23.5%) and 13.4% of respondents fall within the poorest and richest wealth index quintiles, respectively (Fig 2).

Table 1

Individual-level characteristics of included children age 6–59 months selected from the 2016 EDHS.

Variable	Frequency (weighted)	Percentage (%)
Child Sex
Male	4039	51.8
Female	3751	48.2
Child age (months)
6–23	2667	34.2
24–41	2597	33.3
42–59	2526	32.4
Birth weight
Average	3276	42.1
Smaller than average	2001	25.7
Larger than average	2513	32.3
Birth order
1	1362	17.5
2–3	2417	31
4–5	1870	24
> = 6	2141	27.5
Religion
Orthodox	2677	34.4
Protestant	1722	22.1
Muslim	3153	40.5
Other	238	3.1
Maternal age
15–19	208	2.7
20–29	3796	48.7
30–39	3065	39.3
40–49	721	9.3
Maternal educational level
No education	5224	67.1
Primary	2091	26.8
Secondary	318	4.1
Higher	157	2
Husband educational level
No education	3582	48.5
Primary	3008	40.7
Secondary	526	7.1
Higher	275	3.7
Variable	Frequency (weighted)1	percentage
Maternal employment status
No	4248	54.5
Yes	3542	45.5
Husband employment status
No	599	8.1
Yes	6792	91.9
Source of drinking water
Improved	4320	55.5
Unimproved	3470	44.5
Types of toilet facility
Improved	719	9.2
Unimproved	4221	54.2
No facility	2850	36.6
Child wasting
No	7049	90.6
Yes	731	9.4
Child underweight
No	5789	74.7
Yes	1964	25.3
Child stunting
No	4490	59
Yes	3175	41
Maternal anemia
No	5381	70
Yes	2320	30
Child deworming
No	6787	87
Yes	1003	13
Child diarrhea
No	6788	87.14
Yes	1002	12.86
Child fever
No	6605	84.79
Yes	1185	15.21
Symptoms of acute respiratory infection
No	6157	79
Yes	1633	21

1 Weighting variable(wgt) = women individual sample weight (v005)/106; normalization variable(w) = un-weighted sample/weighted sample *wgt.

Fig 2

Wealth index characteristics of children age 6–59 months selected from the 2016 EDHS (n = 7790).

Nearly half of the mothers (45.5%) and 92% of husbands were employed in the last twelve months prior to the survey. A quarter (25%) and more than two-fifths of children (41%) were underweight and stunted, respectively. One-third of mothers (30%) were anemic. Fifteen percent of children had fever two weeks prior to the survey. Above half (54.2%) of households had an unimproved type of toilet facility and about 55.5% of households had improved water sources (Table 1).

Community-level characteristics of study subjects

About nine in ten (89.9%) of the respondents were rural area residents. Most of the respondents were from Oromia (43.9%) followed by SNNP (21.1%) and Amhara (19.5%) region. Nearly half (47.4%) of the respondents were living in communities with a high proportion of women unemployment. The above half (62.2%) of mothers were from communities with lower poverty status. Above half of mothers (52.5%) were from communities with a low proportion of women education (Table 2).

Table 2

Community-level characteristics of included children age 6–59 months selected from the 2016 EDHS.

Variables	Frequency (weighted)	Percentage (%)
Region
Tigray	526	6.8
Afar	76	0.98
Amhara	1520	19.5
Oromia	3418	43.9
SNNP1	1637	21.1
Benishangul Gumuz	83	1.1
Gambela	18	0.23
Somali	321	4.11
Harari	15	0.2
Addis Abeba	148	1.9
Dire Dawa	29	0.37
Place of residence
Rural	7002	89.9
Urban	788	10.1
Community-women education2
Low	4092	52.5
High	3698	47.5
Community- women unemployment3
Low	4098	52.6
High	3692	47.4
Community -poverty4
Low	4098	62.2
High	2945	37.8

1 Sothern Nations and Peoples of Ethiopia.

2 Proportion of mothers aged 15–49 with secondary or higher education in the community.

3 Proportion of mothers aged 15–49 who were not employed in the community in the past twelve months.

4 Proportion of mothers who reside in poor or poorest households in the community.

Distribution of anemia by individual and community level factors

Under this subtitle, the distribution of anemia by the factors which had significant association with anemia (Table 3) among children aged 6–59 months was elaborated. The highest proportion of anemia (72%) was observed in children 6–23 months old as opposed to the lowest proportion in children 42–59 months old (43%) (Fig 3).

Fig 3

Anemia distribution by age of children in Ethiopia: A multilevel analysis using EDHS 2016 (n = 7790).

The highest proportion of anemia among children 6–59 months of age was observed in Somali (83%) and Afar (75%) as opposed to the lowest percentage recorded in Amhara (42%) (Fig 4).

Fig 4

Percentage distribution of anemia among children 6–59 months by Region of respondents in Ethiopia: A multilevel analysis using EDHS 2016 (n = 7790).

Children from anemic mothers (69%) had higher prevalence of anemia as compared to children whose mothers (caretakers) were not anemic (Fig 5).

Fig 5

Distribution of anemia prevalence among children 6–59 months in Ethiopia: A multilevel analysis using EDHS 2016 (n = 7790).

Individual and community-level factors associated with anemia

From the individual-level factors, anemia was most strongly associated with child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order whereas from the community-level, the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa regions (Table 3).

The odds of anemia were 4.45 (95%CI; 3.62, 5.36) and 1.83 (95% CI: 1.59, 2.09) times higher for children 6–23 and 24–41 months of age than children at age 42–59 months, respectively. Children from the poorest families had 1.51 (95% CI: 1.11, 2.04) times higher odds of anemia than children from the richest families. The odds of anemia were 1.26 (95% CI: 1.00, 1.61) times higher for children with birth order six and above than first-order children (Table 3).

Underweight children had 1.34 (95% CI: 1.14, 1.57) times higher odds of anemia than children who were not underweight. The odds of anemia were 1.40 (95% CI: 1.24, 1.59) times higher among children who were stunted than children who are not stunted. Maternal anemia was positively associated with childhood anemia. Children whose mothers were anemic had 1.42 (95% CI: 1.21, 1.55) times higher odds of anemia than children from non-anemic mothers. The odds of anemia were 1.32 (95% CI: 1.09, 1.60) times higher for children who suffered a fever two weeks prior to the survey than children with no fever (Table 3).

The region was a significant predictor of anemia among children. Children from Somali, Dire Dawa and Harari had 3.38 (95% CI: 3.25, 5.07), 2.22 (95% CI: 1.42, 3.48) and 1.88 (95% CI: 1.23, 2.88) times higher odds of anemia than children, respectively than children from Tigray region. The odds of anemia were 0.71 (95% CI: 0.52, 0.96) and 0.58 (95% CI: 0.40, 0.82) times lower among children who were living in Amhara and Benishangul than children, respectively than children from Tigray region. Children who were living in communities of less poverty status had 0.81 (95% CI: 0.66, 0.99) times lower odds of anemia than others (Table 3).

Discussion

This study aimed to identify individual and community-level factors associated with anemia among children aged 6–59 months. We found that anemia among children aged 6–59 was most strongly associated with individual-level factors such as child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order, whereas from the community-level the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa. This was also supported by the observed heterogeneity in odds of anemia between communities.

Child age was negatively associated with anemia in which the odds of anemia decreased as the age of child increased. Children 6–23 months old had higher odds of anemia as compared to children 42–59 months old. This result is in line with some previous studies done in Bangladesh and Ethiopia [13, 18, 28]. This might be due to children experiencing intense growth and development in the first 2 years of life, resulting in a high demand for iron [29]. Additionally, complementary foods are often initially rejected by the infant, thereby exacerbating the risk of anemia.

Children with six and above birth order had higher odds of anemia than first-order children. This result is similar to findings from a prior study done in New Delhi, India [30]. This could be due to increasing birth order might relate to maternal depletion of iron; as the finding of this study showed that maternal anemia leads to child anemia.

Children from the poorest households had higher odds of anemia than children from richest households. This finding is in agreement with what was reported in Bangladesh, Malawi and Ethiopia [11, 21, 28]. This could be explained as poor families are less likely to afford adequate and diversified foods and access to health care which leads to poor child health outcomes.

The odds of anemia were higher for children from anemic mothers than non-anemic mothers. This is supported by studies conducted in India and Ethiopia [13, 31]. The reasons may be mothers and children share common home environments, socioeconomic, and dietary conditions. Moreover, maternal iron deficiency is associated with low birth weight; even children born with adequate weight have reduced iron reserves when their mothers are anemic [32].

This study also revealed that, being underweight or stunted was positively associated with anemia. The association between these anthropometric indices and anemia has been observed in other studies [13, 15, 21, 28]. The possible explanation could be stunting, underweight and anemia are all caused by malnutrition, and thus follow a similar causal pathway that is; feeding children less than four times a day and low dietary diversity. Another explanation could be nutritional inadequacies may impair immunity with a repeated infection which, in turn, depletes iron stores.

The presence of fever in the last two weeks prior to the survey was found to be a significant determinant of anemia. This finding confirms the findings from Indonesia, Burma, Nigeria and Malawi [21, 33–35]. This could explain as fever is a symptom of acute febrile illness such as malaria; which might cause red blood cell destruction. Inflammation also decreases red blood cell production [36]. Another explanation could be that sick children are known to have poorer appetites; hence a lower dietary intake.

This study found that parents’ educational level had no significant association with anemia among children aged 6–59 months. This finding is in contrast with several other studies where an association was observed, especially with maternal educational level [18, 28, 35]. The possible explanation could be controlling socio-economic factors such as parents’ employment, wealth-index, and the composite factors like community poverty in the model may have a more pronounced effect in determining childhood anemia than educational level.

Region was found to be significantly associated with anemia. This finding is supported by other studies from Ghana and Ethiopia [18, 37]. Children from Somali, Dire Dawa, Harari and Afar regions had higher odds of anemia than children from Tigray. This could be because of differences in living standards, socioeconomic status and cultural norms regarding feeding habits among regions. People who are living in Afar and Somalia make their living from livestock production and the main daily meal is milk. A previous study documented that milk reduces the bioavailability of iron which leads to anemia [38]. Another finding of this study was the significant association of community poverty status and anemia among children. Children who live in communities with high poverty status had higher odds of anemia than others. This significant association might be due to less access to health care, fewer job opportunities and lack of other social services within the community which resulted in lower-income.

Limitations

The limitations of this study include: not controlling variables of dietary intake and child feeding practices due to missing values. Other possible child-related explanatory variables such as parasitic infection and chronic illness were not included in the analysis because these variables are not in the EDHS data. The data based on self-reporting are limited by recall and misclassification biases, and only children living at the time of the survey were included. In addition, the community variability in the combined model was 7.03%. This indicates that there are still other variables that are not controlled. These variables could be parasitic infection, immunization history, chronic infections, dietary intake and child feeding practices.

Conclusion

We found that anemia among children aged 6–59 was most strongly associated with individual-level factors such as child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order whereas from the community-level the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa region. Interventions like community-based screening for early detection and management of stunted and underweight children should strengthen to reduce anemia. As our model predicts, wealth index has additional contribution to anemia in Ethiopian children 6–59 months. This suggests that interventions targeted to the improvement of Economic subsidy may contribute to a reduction in childhood anemia and its devastating complications. In addition, special attention should be given to children less than two years of age. Similarly, priority should be given to regions such as Somali, Harari and Afar during the implementation of interventions to reduce anemia.

Random effects estimates of anemia for included children age 6–59 months selected from the 2016 (n = 7790).

(DOCX)

Click here for additional data file.

Child dataset final.

(DTA)

Click here for additional data file.

(PDF)

Click here for additional data file.

14 Jul 2020

PONE-D-20-09803

INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA AMONG CHILDREN 6 - 59 MONTHS OF AGE IN ETHIOPIA: A FURTHER ANALYSIS OF 2016 ETHIOPIA DEMOGRAPHIC AND HEALTH SURVEY

PLOS ONE

Dear Dr. Gebremeskel,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 28 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Aamer Imdad

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include a copy of Table 3 which you refer to in your text

3. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

Additional Editor Comments (if provided):

Please review the comments from reviewer 1. We dont need to change the language from past to present tense.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA AMONG CHILDREN 6 - 59 MONTHS OF AGE IN ETHIOPIA: A FURTHER ANALYSIS OF 2016 ETHIOPIA DEMOGRAPHIC AND HEALTH SURVEY

General comments:

The information you have analyzed and interpreted is valuable. I have made suggestions to improve the organization and presentation of your data.

Cut down on stating data and use the space to provide more interpretation. For example, use more graphics (ex. bar graph) showing the distribution of anemia in your strongest predictors.

Review

Abstract

Background:

Result

- Consider wording the results as Anemia was associated most strongly with ….

Conclusions

- Consider using present tense throughout the paper (e.g. “shows” not showed)

- Malnutrition is not specifically discussed in your results, why is it one of the main conclusion?

Introduction

- “below” not “bellow”

- Childhood anemia is defined by age group and gender. I disagree with your definition. Anemia in children 6-59 months specifically is defined by WHO as <11g/dL hgb. Maybe that is what you meant? I would suggest stating in your methods that you defined anemia as hemoglobin <11g/dL based on xxxx definition from the WHO. For

- I would suggest starting with sentence on prevalence of anemia in children worldwide versus in Ethiopia, why it is important/what the complications are, then why hemoglobin is used as a measure of anemia, then the WHO definition of anemia in children 6-59 months of age, and then discuss more details on Ethiopian anemia classification and problems specific to anemia.

- Specifically state the aims. The aims of this study are to 1), 2), 3)….

Methods

- Where are your inclusion and exclusion criteria? What is your study population?

- I would remove all formulas from your methods unless they are novel or controversial methods of analysis.

-

Data Source

- I would again spell out EDHS the first time in this section, but that is up to you.

- Which MOH? Specify Ethiopian Ministry of Health.

- Yes, we need to know that you have permission to access the EDHS database, but I don’t think we need to know how you obtained permission.

- “through” not “thorough”

Study Design and Sample Size

- The cross sectional survey you describe is the original method of the EDHS data collection? Or are we talking about your study?? If this is your study, then very briefly describe the methods of the EDHS in the first paragraph.

- We need your inclusion and exclusion criteria in order to understand why it is important that 9504 children were 6-59 months of age.

- Revise grammar for the sentence “the 2016 EDHS was used…”

- Please explain the rationale for the use of clusters if this is your selection methodology

-

Study Variables (Maybe a better title is Definitions?)

- What do you mean the variables were selected from the literature?

- I would not go to the extent of telling us if your variables are dichotomous.

- Perhaps simplify this paragraph by first stating that “We assessed the impact of individual and community-level variables on our primary outcome of anemia. We defined anemia as … (see below). Individual level variables were …. Community-level variables were …..”

- “We defined anemia in 6-59 month children as <11 g/dL according to WHO criteria (reference).” Please see earlier comments. Remember that unless you specify, I don’t even know that you are discussing children 6-59 months only.

- Definitions are important. Consider using full sentences such as, “Wealth Index is a composite measure…” instead of : .

- Consider combining anthropometrics into a paragraph “Anthropometrics: Stunting was defined as height or length for age, HFA) <-2, wasting as ….

- What about children without a mother? Are they included? I wonder if some of the “mothers” are other relatives who are the child’s guardian.

- What do you mean by “poor” or “poorest” etc.? Are these economic quintiles? Please define objectively this subjective term. This makes it difficult for me to interpret Figure 2.

Methods of data analysis

- I would cut this section down to two paragraphs. The explanation of your model also needs to be condensed to one or two paragraphs. You may consider sharing your variable selection as an Appendix or supplement.

Results

- How many children were included?

- Please first state how many children you included and basic demographics (age, gender). Then state prevalence of anemia. Then go on to other variables.

- The sentence tells me that children were living with their partners. Surely this is not what is meant. Consider using “caretakers” to clarify when you discuss characteristics of the caretaker or mother.

- I would be more interested in seeing anemia sorted by economic status than a pie chart only of economic status. I think seeing child anemia sorted by maternal anemia (what is the definition of this, by the way?) would also provide useful information, as maternal anemia is a known risk factor for childhood anemia especially for the youngest.

- Table 2. I think this should be Table 1. I would not present the current Table 1. Good to have this table (2), but some of this information can be presented as supplemental data. Please indicate the units, for example child age (months). Please consider labeling this as something like “Table x. Demographic characteristics of included children age 6-59 months selected from the 2016 EDHS.”

- Why does birth order 4-5 matter more than >=6?

- Community Level characteristics table: Please note this is mislabeled as Table 1. Please also see above suggestion for the title.

Individual and community-level factors associated with anemia

- Consider presenting the odds ratios in Table 1 next to the demographics. This would condense a lot of information into less space. Alternatively, cut down your demographics to the essential for Table 1 (currently labelled 2) and then keep Table 4 as is. Then present some of your data in graphs as well.

- Consider instead of simply stating all the OR, giving some context of importance. For example, “the strongest odds of anemia occurred with x and y (data).” Or “anemia was most strongly associated with ….”

- Table 4: I would remove crude OR and present the adjusted OR with p value in the next column. Also, see above to cut down on demographics table as much of the same information is repeated here. Please explain what you adjusted for. Tell me more about the results of your multivariate model? Which variables made it into your model? This is not clear.

Discussion

- “This study aimed to ….” (not was aimed to)

- “This finding indicated that” is too specific to one finding when you are discussing an entire study. Consider wording like: “We found that anemia was most strongly associated with … “ or “We found that … were the strongest predictors of anemia followed by…”

- Consider making a figure using a flowchart of the strongest predictors of anemia (from your model) with the adjusted OR or other information along the arrows that lead to anemia. This would be an excellent visual representation of your most important conclusion.

- What about confounders or collinearity? How confident are you that those factors did not influence your model?

Conclusion

- I would emphasize here the strongest of all predictors first.

- I think going from economic quartile as a risk factor to suggesting food subsidies is a big stretch. Perhaps say something like “our model predicts that … are most strongly associated with anemia in Ethiopian children 6-59 months with additional contribution from …. This suggests that interventions targeted to the improvement of …. may contribute to a reduction in childhood anemia and its devastating complications.”

Reviewer #2: This study addressed an important question. The study question is clearly defined and methods are described in detail and the interpetation of results is appropriate. This study will be an important contribution to the existing literature about risk of anemia in young children.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

12 Aug 2020

Responses to reviewers

Reviewer #1

Dear reviewer, thank you for your thorough review of our manuscript. Your comments are constructive and very good lessons for us. We have tried to put our responses to your constructive comments and questions. The responses are put immediately after the questions, suggestions or comments.

General comments:

The information you have analyzed and interpreted is valuable. I have made suggestions to improve the organization and presentation of your data.

1. Cut down on stating data and use the space to provide more interpretation. For example, use more graphics (ex. bar graph) showing the distribution of anemia in your strongest predictors

� As per the comment, we presented the distribution of anemia by the factors found to have significant association using pie chart and bar graphs.

Review

2.Abstract

Background:

Result

2.1. Consider wording the results as Anemia was associated most strongly with ….

� The result part of the abstract is reviewed and as per the comments given, we rewrite it as: Anemia was associated most strongly with child age, stunting, underweight, fever, maternal anemia, low poverty-community and Region. The odds of anemia were 0.81 AOR=0.81; 95% CI: 0.66, 0.99) times lower for children who were living in communities of lower poverty status than children who were living in communities of higher poverty status. Children from Somali and Dire Dawa had 3.38 (AOR=3.38; 95% CI: 3.25, 5.07) and 2.22 (AOR=2.22; 95% CI: 1.42, 3.48) times higher odds of anemia, respectively than children from the Tigray Region.

Conclusions

2.2. Consider using present tense throughout the paper (e.g. “shows” not showed)

� The conclusion is reviewed not only in the abstract but also in the main body, and based on your comment we rewrite it as: This study shows that childhood anemia is affected both by the individual and community level factors. It is better to strengthen the strategies of early detection and management of stunted and underweight children. At the same time, interventions should be strengthened to address maternal anemia, child fever and poverty, specifically targeting regions identified to have a high risk of anemia.

2.3. Malnutrition is not specifically discussed in your results, why is it one of the main conclusion? is reviewed not only in the abstract but also in the main body

� Yes. Malnutrition itself is not discussed as our result. However, we found that child stunting and underweight had strong association with anemia, and these variables are some forms of Malnutrition in children. But now your comment reminds us it is better to conclude based on our result i.e stunting and underweight. So, we rewrite the conclusion as: This study shows that childhood anemia is affected both by the individual and community level factors. It is better to strengthen the strategies of early detection and management of stunted and underweight children. At the same time, interventions should be strengthened to address maternal anemia, child fever, food insecurity and poverty, specifically targeting regions identified to have a high risk of anemia.

3. Introduction

3.1. “below” not “bellow”

� The word bellow is deleted. And the sentence now becomes: Childhood anemia is defined as hemoglobin concentration in the blood below 11g/dl.

3.2. Childhood anemia is defined by age group and gender. I disagree with your definition. Anemia in children 6-59 months specifically is defined by WHO as <11g/dL hgb. Maybe that is what you meant? I would suggest stating in your methods that you defined anemia as hemoglobin <11g/dL based on xxxx definition from the WHO. For

� We defined anemia in 6-59 month age children as hemoglobin <11 g/dL according to WHO criteria.

3.3. I would suggest starting with sentence on prevalence of anemia in children worldwide versus in Ethiopia, why it is important/what the complications are, then why hemoglobin is used as a measure of anemia, then the WHO definition of anemia in children 6-59 months of age, and then discuss more details on Ethiopian anemia classification and problems specific to anemia.

� We carefully reviewed the introduction, and we rewrite it based on your suggestion in the revised document.

3.4. Specifically state the aims. The aims of this study are to 1), 2), 3)….

� As per the comment, the aims of this study are :

1. To identify individual level factors associated with anemia among children 6-59 months of age in Ethiopia.

2. To identify community level factors associated with anemia among children 6-59 months of age in Ethiopia

4. Methods

4.1. - Where are your inclusion and exclusion criteria? What is your study population?

� Based on your comment, we included the inclusion and exclusion criteria as follows. for general information, the EDHS kid record file is source of our sample, and this data contains under-five children (children 0-59 months of age).

� Inclusion criteria: children 6-59 months of age who live in the selected enumeration areas (community).

� Exclusion criteria: children 6-59 months of age who have no hemoglobin test result. So, according to this inclusion and exclusion criteria, we have had a total of 10,641 under-five children regardless of their hemoglobin test result. Then from these children, we exclude 1137 children less than 6 months of age (10641-1137=9504 children 6-59 months age regardless of their hemoglobin test result). Then from these children we exclude again 1714 children 6-59 months age who have no hemoglobin test result. Finally, we consider 7790 children 6-59 months age as eligible for our study.

� The study population is children 6-59 months of age who were living in selected enumeration areas.

4.2. I would remove all formulas from your methods unless they are novel or controversial methods of analysis.

� As mentioned in the method section of the document, the analysis method we applied is multilevel binary logistic regression analysis. These multilevel models are different from the ordinary single level logistic regression in such a way that they can handle hierarchical nature of data like that of the Demographic and Health Survey (DHS) data. Recently, these models are becoming popular models. However, in our country Ethiopia, these models are still rarely used. That is why we preferred to include their formula in our document.

Data Source

4.3. I would again spell out EDHS the first time in this section, but that is up to you.

� EDHS is abbreviation for the Ethiopia Demographic and Health Survey and mostly it is the name given for the respective Ethiopia Demographic Health Surveys conducted in Ethiopia.

4.4. Which MOH? Specify Ethiopian Ministry of Health.

� We correct it and the sentence becomes “The 2016 EDHS is the fourth survey which is implemented by the Central Statistical Agency (CSA) in collaboration with the Ethiopian Ministry of Health under the technical assistance of International Classification of Functioning, Disability, and Health (ICF) thorough the DHS Program”.

4.5. Yes, we need to know that you have permission to access the EDHS database, but I don’t think we need to know how you obtained permission.

� Our intention to write how we obtained permission to access the EDHS data was just to convince readers of this paper that we are legal. But based on your comment we removed the sentences how we obtained permission.

4.6. “through” not “thorough”

� We correct it.

4.7. Study Design and Sample Size. The cross sectional survey you describe is the original method of the EDHS data collection? Or are we talking about your study?? If this is your study, then very briefly describe the methods of the EDHS in the first paragraph.

� The cross sectional survey we described in our document is the original method of the EDHS data collection. We did nothing on the study design and sampling design just we used the original method of the EDHS data collection methods.

4.8. We need your inclusion and exclusion criteria in order to understand why it is important that 9504 children were 6-59 months of age.

� Inclusion criteria: children 6-59 months of age who live in the selected enumeration areas (community).

� Exclusion criteria: children 6-59 months of age who have no hemoglobin test result. So, according to this inclusion and exclusion criteria, we have had a total of 10,641 under-five children regardless of their hemoglobin test result. Then from these children, we exclude 1137 children less than 6 months of age (10641-1137=9504 children 6-59 months age regardless of their hemoglobin test result). Then from these children we exclude again 1714 children 6-59 months age who have no hemoglobin test result. Finally, we consider 7790 children 6-59 months age as eligible for our study”.

4.9. Revise grammar for the sentence “the 2016 EDHS was used…”

� We corrected it as: The 2016 EDHS had used a stratified two-stage cluster sampling design.

4.10. Study Variables (Maybe a better title is Definitions?)

� We amend it as: Definitions of study variables

4.11. What do you mean the variables were selected from the literature?

� During the conception of our study, we reviewed literatures just to know what was done before regarding anemia among children 6-59 months age and what is left to solve the problem. So, during this we observed and identified variables included in the literatures done before. Bay the way, we include also new variables for the first time in our study.

4.12. I would not go to the extent of telling us if your variables are dichotomous.

� Yes you are right. But we don’t think that all readers have the same knowledge in understanding the outcome variable (anemia) whether it is binary variable. That is why we indicated as it is dichotomous.

4.13. Perhaps simplify this paragraph by first stating that “We assessed the impact of individual and community-level variables on our primary outcome of anemia. We defined anemia as … (see below). Individual level variables were …. Community-level variables were …..”

� We reviewed this paragraph and based on your comment, we rewrite it as:

� We assessed the impact of individual and community-level variables on anemia among children 6-59 months of age. We defined anemia in 6-59 month children as hemoglobin (hg) <11 g/dL according to WHO criteria. Individual level variables were: sex, age, birth order, birth weight, religion, number of under-five children, childhood wasting, underweight, stunting, symptoms of acute respiratory infection, child fever and diarrhea, maternal anemia and age, parents’ educational and employment status, wealth index, source of drinking water, and type of toilet facility, whereas community-level variables were: Region, community-poverty, community-women education and community- women unemployment”

4.14. “We defined anemia in 6-59 month children as <11 g/dL according to WHO criteria (reference).” Please see earlier comments. Remember that unless you specify, I don’t even know that you are discussing children 6-59 months only.

� We corrected it as per your comment as” We defined anemia in 6-59 month age children as hemoglobin <11 g/dL according to WHO criteria”.

4.15. Definitions are important. Consider using full sentences such as, “Wealth Index is a composite measure…” instead of : .

� As per the comments you provided, we rewrite it as “Wealth index: is a composite measure of a household’s cumulative living standard. It was calculated based on household ownership of selected assets such as televisions and bicycles, cars; materials used for the housing construction; source of drinking water; and type of sanitation facilities. It was then generated using principal components analysis and the individual households were placed on a continuous scale of relative wealth. In the EDHS all mothers and children were assigned a standardized wealth index score. It was measured as a composite variable made up of five quintiles as poorest, poorer, middle, richer and richest.

4.16. Consider combining anthropometrics into a paragraph “Anthropometrics: Stunting was defined as height or length for age, HFA) <-2, wasting as ….

� We corrected it as “Anthropometrics: Stunting was defined as height or length for age, (HFA) <-2SD (standard deviation), wasting as weight-for-height (WFH) <-2 SD and underweight as weight-for-age (WFA) <-2 SD.

4.17. What about children without a mother? Are they included? I wonder if some of the “mothers” are other relatives who are the child’s guardian.

� Yes. We include children without a mother that is children with their guardian or caretakers. So the whole document is corrected by considering this comment.

4.18. What do you mean by “poor” or “poorest” etc.? Are these economic quintiles? Please define objectively this subjective term. This makes it difficult for me to interpret Figure 2.

� Poor or poorest are economic quintiles. In the 2016 Ethiopia Demographic and Health Survey, these are quintiles of wealth index (economic quintiles) calculated based on household ownership of selected assets such as televisions and bicycles, cars; materials used for the housing construction; source of drinking water; and type of sanitation facilities. It was then generated using principal components analysis and the individual households were placed on a continuous scale of relative wealth. In the EDHS all mothers (caretakers) and children were assigned a standardized wealth index score. It was measured as a composite variable made up of five quintiles as poorest, poorer, middle, richer and richest

Methods of data analysis

4.19. I would cut this section down to two paragraphs. The explanation of your model also needs to be condensed to one or two paragraphs. You may consider sharing your variable selection as an Appendix or supplement.

� We aimed to explain more about the models to our reader, since it is unusual model especially in our country. But we annexed Table 1 as “annex 1” at the end of the main document.

5. Results

5.1. How many children were included?

� We included 7790 children 6 -59 months of age who have hemoglobin test result

5.2. Please first state how many children you included and basic demographics (age, gender). Then state prevalence of anemia. Then go on to other variables.

� We accepted the comment and we rewrite it again as” Seven thousand seven hundred ninety (7790) children 6 -59 months of age were included in this study. Above half (52%) of the children were male, and 34.2%, 33.3% and 32.2% where in the age category of 6-23, 24-41 and 42-59 months of age, respectively with mean ± SD (standard deviation) of 32 +15 months. The prevalence of anemia was 57.6% with a median hemoglobin concentration of 10.7 (IQR: 9.6-11.6)”.

5.3. The sentence tells me that children were living with their partners. Surely this is not what is meant. Consider using “caretakers” to clarify when you discuss characteristics of the caretaker or mother.

� We accepted the comment, and we rearranged the sentence as” Almost all of the respondents mothers (caretakers) (95%) were living with their respective partners and most of them were Muslims (40%) followed by Orthodox Christians (34%).

5.4. I would be more interested in seeing anemia sorted by economic status than a pie chart only of economic status. I think seeing child anemia sorted by maternal anemia (what is the definition of this, by the way?) would also provide useful information, as maternal anemia is a known risk factor for childhood anemia especially for the youngest.

� All the significant variables have sorted by anemia and this is indicated in table 4.

� Maternal anemia is defined as hemoglobin less than or equal to 11 g/dL (hg<=11 g/dL) according to WHO criteria. In the EDHS, hemoglobin levels were adjusted for pregnancy because during pregnancy the increase in maternal blood volume and the iron needs of the fetus decreases the blood Hb level.

5.5. Table 2. I think this should be Table 1. I would not present the current Table 1. Good to have this table (2), but some of this information can be presented as supplemental data. Please indicate the units, for example child age (months). Please consider labeling this as something like “Table x. Demographic characteristics of included children age 6-59 months selected from the 2016 EDHS.”

� We annexed table 1, and we rename Table 2 as Table 1. We accept all the other comments and we amend our document accordingly.

5.6. Why does birth order 4-5 matter more than >=6?

� According to our result presented in Table 4, birth order >=6 matter more than birth order 4-5. The odds of anemia were 1.26 (AOR=1.26; 95% CI: 1.00, 1.61) times higher for children with birth order six and above than first-order children (Table 4).

5.7. Community Level characteristics table: Please note this is mislabeled as Table 1. Please also see above suggestion for the title.

� We corrected it accordingly.

Individual and community-level factors associated with anemia

5.8. Consider presenting the odds ratios in Table 1 next to the demographics. This would condense a lot of information into less space. Alternatively, cut down your demographics to the essential for Table 1 (currently labeled 2) and then keep Table 4 as is. Then present some of your data in graphs as well.

� Table 1 (currently labeled 2) has presented simple frequency of all included variables. In addition, this table did not show any distribution of anemia by these factors, whereas Table 4 in the second column to right shows the anemia distribution by the significant factors. Then next to this, the Adjusted OR is written with their p-values. Therefore, these two tables are independent.

5.9. Consider instead of simply stating all the OR, giving some context of importance. For example, “the strongest odds of anemia occurred with x and y (data).” Or “anemia was most strongly associated with ….”

� We accepted the comment. we rewrite it as “From the individual-level factors, anemia was most strongly associated with child age, wealth index, maternal anemia and child stunting, whereas from the community-level, the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region (Table 4).

5.10. Table 4: I would remove crude OR and present the adjusted OR with p value in the next column. Also, see above to cut down on demographics table as much of the same information is repeated here. Please explain what you adjusted for. Tell me more about the results of your multivariate model? Which variables made it into your model? This is not clear.

� We removed crude OR and we present the adjusted OR with p value in the next column.

� The demographic information written in Table 4 is not the same as the demographic information written in Table 1. Table 1 (previously labeled as Table 2) has presented simple frequency of all included variables. In addition, this table did not show any distribution of anemia by these factors, whereas Table 4 in the second column to right shows the anemia distribution by the significant factors. Then next to this, the Adjusted OR is written with their p-values.

5.11. Please explain what you adjusted for. Tell me more about the results of your multivariate model?

� First bivariate analysis was performed to see the effect of each predictor variable on the outcome variable using a significance level of p<0.25 independently. Accordingly, anemia among children aged 6-59 months was associated with number of <5 children in the household, parents educational level and employment status, child age, religion of mother, birth order, maternal age, type of toilet facility , source of drinking water, wealth index, child stunting, wasting and underweight, fever, diarrhea, child deworming, symptoms acute respiratory infection, maternal anemia status, women community-education , place of residence, community-poverty , region and community- women unemployment status. Then all this variables were included simultaneously in to the final model to estimate adjusted odds ratios with 95% Confidence Interval (CI) at a significance level of p<0.05.

� In short, in the multivariable analysis the following variables were adjusted and controlled.

1. number of <5 children in the household

2. child age

3. religion of mother

4. birth order

5. maternal and husband employment status

6. maternal age, type of toilet facility

7. source of drinking water

8. wealth index

9. child stunting, wasting, underweight, fever, diarrhea

10. child deworming, acute respiratory infection

11. maternal anemia status

12. community- women education

13. place of residence, community-poverty

14. region

15. community- women unemployment

� After adjusting for all this variables, we found that anemia was most strongly associated with child age, wealth index, maternal anemia, child stunting, underweight, child fever, birth order, community-poverty and region (Somali, Harari, Dire Dawa, Afar, Oromia, Addis Ababa, Amhara and Benishangul). The result in

� Table 4 shows only the significant variables found at the multivariable analysis.

6. Discussion

6.1. “This study aimed to ….” (not was aimed to)

� We correct it as” This study aimed to identify individual and community-level factors associated with anemia among children aged 6-59 months”.

6.2.“This finding indicated that” is too specific to one finding when you are discussing an entire study. Consider wording like: “We found that anemia was most strongly associated with … “ or “We found that … were the strongest predictors of anemia followed by…”

� We accepted the comment and we rearranged as” We found that anemia among children aged 6-59 was most strongly associated with individual-level factors such as child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order, whereas from the community-level the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa”.

6.3. Consider making a figure using a flowchart of the strongest predictors of anemia (from your model) with the adjusted OR or other information along the arrows that lead to anemia. This would be an excellent visual representation of your most important conclusion.

� We corrected based on the, comment that order the strongest predictor in text, you provided to us. But the comments consider making a figure using a flowchart of the strongest predictors of anemia is not clear, sorry.

6.4. What about confounders or collinearity? How confident are you that those factors did not influence your model?

� We have checked confounders or collinearity to our model and no confounders or collinearity was detected.

7. Conclusion

7.1. I would emphasize here the strongest of all predictors first.

� We found that anemia among children aged 6-59 was most strongly associated with individual-level factors such as child age, wealth index, maternal anemia and child stunting followed by child underweight, child fever and birth order, whereas from the community-level the strongest odds of anemia occurred among children from Somali, Harari, Dire Dawa and Afar region followed by Oromia and Addis Ababa.

7.2. I think going from economic quartile as a risk factor to suggesting food subsidies is a big stretch. Perhaps say something like “our model predicts that … are most strongly associated with anemia in Ethiopian children 6-59 months with additional contribution from …. This suggests that interventions targeted to the improvement of …. may contribute to a reduction in childhood anemia and its devastating complications.”

� Our model predicts that child age, maternal anemia and child stunting followed by child underweight, child fever and birth order and region are most strongly associated with anemia in Ethiopian children 6-59 months with additional contribution from wealth index. This suggests that interventions targeted to the improvement of Economic subsidy may contribute to a reduction in childhood anemia and its devastating complications.

Reviewer #2

Dear reviewer, thank you for your thorough review of our manuscript and for encouraging us.

7 Oct 2020

PONE-D-20-09803R1

INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA AMONG CHILDREN 6 - 59 MONTHS OF AGE IN ETHIOPIA: A FURTHER ANALYSIS OF 2016 ETHIOPIA DEMOGRAPHIC AND HEALTH SURVEY

PLOS ONE

Dear Dr. Gebremeskel,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please consider the comments of reviewer 3 carefully, especially regarding the statistical analysis and subsequent interpretation.

Please submit your revised manuscript by Nov 21 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Astrid M. Kamperman

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #3: No

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Overall, much improved flow and language. Excellent abstract.

Statistics are still overly described, but that is your choice if you have the room for it.

Throughout:

The term “lower poverty status” is confusing. Are these communities less poor? I would say “less poverty” or “more poverty” instead of “higher poverty status” and vv. Another option is “lower wealth index”.

Check on small details like punctuation, missing brackets in tables around CI, spelling.

Re the use of “caretakers,” if mothers are mothers and there were no non-mother caretakers (ex. grandmothers), then there is no need to continue to say both words: mothers (caretakers).

I would use paragraph form and not bullets or check marks for inclusion/exclusion criteria, but that is an editorial choice.

Figure 2: the word “Characteristics” is misspelled

Tables:

I am of the opinion that tables should stand alone. If I look at these tables now without the rest of the article, I am lost and need more definitions and clarity.

For each table, give definitions in the captions. What do you mean by “low” education for example? How is frequency weighted? Give a description as a footnote/caption to the table.

What do you mean by community-women education or community-women unemployment? Are these grouped per region? If it is instead a community-level characteristic, then you have already labelled that in the title.

Discussion

First sentence of first paragraph has . and , in the middle of the sentence.

Great work.

Reviewer #3: The paper and analysis initially struck me as to offer interesting insight on individual and group level effects. On second look there are some omissions (e.g. in reporting on missing data) and some irregularities with what was reported about model building (checking for interaction effects). By looking at the data myself, I quickly find a mismatch between what the authors reported in their model building approach and what I find. I attach the R script I used and the descriptive plot I produced.

Aside from that, some revisions are required in formatting and presenting the results. Additionally I have left comments on some relatively small clarifications need to be added addressing choices the authors made when constructing and reporting on their model.

Please check the attached PDF file to access my notes on the manuscript.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: Yes: Milan Zarchev

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: interaction INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA.png

Click here for additional data file.

Submitted filename: script INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA.R

Click here for additional data file.

Submitted filename: INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA.pdf

Click here for additional data file.

13 Oct 2020

Responses to reviewers

Reviewer #1

Dear reviewer, thank you for your thorough review of our manuscript. Your comments are constructive and very good lessons for us. We have tried to put our responses to your constructive comments and questions. The responses are put immediately after the questions, suggestions or comments.

Reviewer #1: Overall, much improved flow and language. Excellent abstract.

Statistics are still overly described, but that is your choice if you have the room for it.

Throughout:

The term “lower poverty status” is confusing. Are these communities less poor? I would say “less poverty” or “more poverty” instead of “higher poverty status” and vv. Another option is “lower wealth index”.

� We correct it based on your comment in the manuscript.

Check on small details like punctuation, missing brackets in tables around CI, spelling.

Re the use of “caretakers,” if mothers are mothers and there were no non-mother caretakers (ex. grandmothers), then there is no need to continue to say both words: mothers (caretakers).

� We correct it based on your comment in the manuscript.

I would use paragraph form and not bullets or check marks for inclusion/exclusion criteria, but that is an editorial choice.

� We write it in paragraph form as” The inclusion criteria were children 6-59 months of age who live in the selected enumeration areas (community). And Exclusion criteria were children 6-59 months of age who have no hemoglobin test result.”

Figure 2: the word “Characteristics” is misspelled

� We correct it.

Tables:

I am of the opinion that tables should stand alone. If I look at these tables now without the rest of the article, I am lost and need more definitions and clarity.

For each table, give definitions in the captions. What do you mean by “low” education for example? How is frequency weighted? Give a description as a footnote/caption to the table.

What do you mean by community-women education or community-women unemployment? Are these grouped per region? If it is instead a community-level characteristic, then you have already labelled that in the title.

� We add footnote in to our manuscript based on your comment

Discussion

First sentence of first paragraph has . and , in the middle of the sentence.

� We correct it.

Great work.

Reviewer #3

Dear reviewer, thank you for your thorough review of our manuscript and for encouraging us.

Reviewer #3: The paper and analysis initially struck me as to offer interesting insight on individual and group level effects. On second look there are some omissions (e.g. in reporting on missing data) and some irregularities with what was reported about model building (checking for interaction effects). By looking at the data myself, I quickly find a mismatch between what the authors reported in their model building approach and what I find. I attach the R script I used and the descriptive plot I produced.

� The R script cannot open. So we cannot see it, sorry.

1. Reporting on missing data

� From the total of 10,641 under-five years’ old children, 9504 were children 6-59 months of age. Data on hemoglobin test result from the survey were available for 7790 children. As a result, 1714 children aged 6-59 month were excluded from the study due to missing data of hemoglobin test result. In addition, the variables of dietary intake and child feeding practices were not included due to missing value. These variables were missing for nearly half of observations. This is due to the reason that our study sample was age 6-59 months that the variables of dietary intake and child feeding practices were not a concern for children less than 6 months old. Other possible child-related explanatory variables such as parasitic infection and chronic illness were not included in the analysis because these variables are not in the EDHS data.

2. Regarding Interaction

� Interaction between variables was checked for those variables found significant at the final model. As a result, there were significant interactions (p<0.05) between these variable. However, as we examined the interaction effect by fitting regression models that contained interaction terms yields no significant (p>0.05) interaction effect. “We have attached evidences as supplementary material”

Aside from that, some revisions are required in formatting and presenting the results. Additionally I have left comments on some relatively small clarifications need to be added addressing choices the authors made when constructing and reporting on their model.

� We corrected it based on the comments you provided on the manuscript.

20 Oct 2020

INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA AMONG CHILDREN 6 - 59 MONTHS OF AGE IN ETHIOPIA: A FURTHER ANALYSIS OF 2016 ETHIOPIA DEMOGRAPHIC AND HEALTH SURVEY

PONE-D-20-09803R2

Dear Dr. Gebremeskel,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Astrid M. Kamperman

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: All comments were either incorporated or addressed adequately. I have no further major recommendations for this paper.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: Yes: Milan Zarchev

4 Nov 2020

PONE-D-20-09803R2

INDIVIDUAL AND COMMUNITY LEVEL FACTORS ASSOCIATED WITH ANEMIA AMONG CHILDREN 6 - 59 MONTHS OF AGE IN ETHIOPIA: A FURTHER ANALYSIS OF 2016 ETHIOPIA DEMOGRAPHIC AND HEALTH SURVEY

Dear Dr. Gebremeskel:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Astrid M. Kamperman

Academic Editor

PLOS ONE