Anemia in children aged 6–59 months was significantly associated with maternal anemia status in rural Zimbabwe

Abstract Globally, anemia is a public health problem affecting mostly women of reproductive age (WRA, n = 452) and children aged 6–59 months (n = 452) from low‐ and lower‐middle‐income countries. This cross‐sectional study assessed the prevalence and determinants of anemia in WRA and children aged 6–59 months in rural Zimbabwe. The venous blood sample was measured for hemoglobin utilizing a HemoCue machine. Anthropometric indices were assessed and classified based on World Health Organization standards. Socioeconomic characteristics were assessed. The median (±inter quartile range (IQR)) age of WRA was 29 ± 12 years and that for children was 29 ± 14 months. The prevalence of anemia was 29.6% and 17.9% in children and WRA, respectively, while the median (±IQR) hemoglobin levels were 13.4 ± 1.8 and 11.7 ± 1.5 g/dl among women and children, respectively. Multiple logistic regression analysis was used to assess determinants of anemia. Anemia in children was significantly associated with maternal anemia (odds ratio (OR) = 2.02; 95% CI 1.21–3.37; p = .007) and being a boy (OR = 0.63; 95% CI 0.41–0.95; p = .029), while anemia in WRA was significantly associated with the use of unimproved dug wells as a source of drinking water (OR = 0.36; 95% CI 0.20–0.66; p = .001) and lack of agricultural land ownership (OR = 0.51; 95% CI 0.31–0.85; p = .009). Anemia is a public health problem in the study setting. The positive association between maternal and child anemia reflects the possibility of cross‐generational anemia. Therefore, interventions that focus on improving preconceptual and maternal nutritional status may help to reduce anemia in low‐income settings.


| INTRODUC TI ON
Anemia is a condition in which the number of red blood cells and consequently their oxygen-carrying capacity is insufficient to meet the body's physiologic oxygen needs (WHO & Chan, 2011;WHO, 2020a). Anemia affects approximately one-quarter of the world's population and is concentrated in preschool children (PSC) and women of reproductive age (WRA) (McLean et al., 2009).
However, the situation will be lower in settings with high rates of inflammation and infection (Petry et al., 2016). The current study looked at anemia defined as low hemoglobin (Hb) levels, Hb < 11 g/ dl for children and Hb < 12 g/dl for WRA. Anemia is associated with adverse reproductive outcomes such as preterm delivery, low-birthweight infants, and decreased iron stores for the baby, which leads to impaired health and productivity later in life (Black et al., 2008).
Failure to reduce anemia may result in maternal and neonatal deaths, a combination that accounts for 2.5-3.4 million deaths worldwide (WHO & Worldbank, 2010).
Anemia is an indicator of both poor nutrition and poor health (WHO, 2021). Women of reproductive age and children are deemed the most vulnerable to micronutrient deficiencies due to greater physiological requirements (Ahmed et al., 2017). In 2019, the global anemia prevalence was 29.9% in WRA, equivalent to over half a billion women aged 15-49 years, and 39.8% in children aged 6-59 months, equivalent to 269 million children with anemia. The prevalence of anemia in children under 5 years of age was highest in the African Region (WHO, 2021). According to the 2015 Zimbabwe Demographic Health Survey (ZDHS), more than a third of children aged 6-59 months (37%) were anemic while 27% of WRA had anemia (ZIMSTAT, 2015).
Anemia is often associated with multiple proximal risk factors (iron and vitamin A deficiencies, inflammation, malaria, age, gender, and body mass index [BMI]) and distal risk factors (education status, sanitation and hygiene facilities, and urban or rural residence) (Cane et al., 2022;Mrimi et al., 2022;Wirth et al., 2017). In Zimbabwe, the major causes of anemia in children and WRA are likely to include malaria, helminths, iron, and other nutritional deficiencies and chronic infections (ZIMSTAT, 2015). Zimbabwe does not have any recent national-level statistics on the prevalence and predictors of anemia after the ZDHS 2015 survey. Furthermore, the emergence of the COVID-19 pandemic resulted in restricted mobility and isolation, reduced income, and reduced access to essential health and nutrition services, and these dynamics are likely to have exacerbated malnutrition and food insecurity. Hence, anemia prevalence, as well as predictors, may have changed and thus highlights the need for studies to summarize the evolved scenario in the context of evidence-based health delivery.
The current study was designed to explore demographic, health, and nutritional characteristics associated with anemia among WRA and children aged 6-59 months in selected districts in rural Zimbabwe. The data for the present study were collected as part of the baseline for a micronutrient biomarker survey "Translating

| Primary sampling units (PSU)
The first stage of the sampling frame was made up of 30 enumeration areas (EA) selected per district, each EA had a unique 10-digit geocode that reflected the province, district, ward, and land-use sector in which it was located. EAs were selected by proportional probability sampling (PPS) with inclusion probabilities proportional to the most recently recorded population (ZIMSTAT, 2012

| Secondary sampling units (SSU)
At the EA level, all eligible households (defined as a unit of members eating food prepared from the same pot with at least one WRA and one child 6-59 months of age) were listed. Household lists provided the frame for the second stage, where systematic random sampling without replacement approach was used to select participating households. The location of each household within the EAs was determined using a global positioning system receiver (GPS) and verified through matched shape files ( Figure 1). Only 10 households were selected in each of the sampled EAs using the formula: where: • Phij is the selection probability of household number j in EA number hi • nhi is the size of the household sample from EA number hi • Nhi is the listed number of households in EA number hi

| Tertiary sampling units (TSU)
In households where more than one set of eligible individuals were identified, one mother-child pair was randomly selected using the Kish grid method (Jabkowski, 2017). The degrees of freedom for variance components of a variable for which one measurement was made per household were as follows: 2 between-district, 87 between-EA within-district, and 510 between-household withindistrict. Based on this approach, the study selected 1200 participants (600 WRA and 600 children aged 6-59 months). Recruitment was done at the household level, with participants then directed to the nearest health facility for data collection.

| Inclusion and exclusion criteria
Women aged 15-49 years and children aged 6-59 months were eligible for the study. The exclusion criteria consisted of the following: chronically ill; outside the target age groups; diagnosed with chronic disease or long-term illness requiring treatment; with acute or chronic infection; on regular prescribed medicine; regularly using dietary supplements including iodine, zinc, or selenium supplements; children with a physical disability that would affect either height or weight measurement, and if the mother-child pair were not resident in the study. The eligibility screening criteria were adapted from a similar survey conducted in Malawi (Phiri et al., 2019). Illnesses were either self-reported and/or assessed by health practitioners.

| Socioeconomic characteristics
Data were collected using face-to-face interviews at selected health facilities, observing national COVID-19 guidelines.

| Anthropometry
Weight, recumbent length, and height were measured according to WHO standard protocols (WHO, 2008 Maternal short stature was defined as a height below 145 cm (Sumarmi, 2016). In children aged 6-59 months old, wasting was defined as WHZ below −2 standard deviations (SD), stunting as HAZ below −2SD, underweight as WAZ below −2SD, and overweight as WHZ above +2SD (WHO, 2006). Low birth weight (LBW) was defined as a birth weight below 2.5 kg regardless of gestational age (WHO, 2011b).

| Hemoglobin levels and anemia
Hemoglobin (Hb) testing is the primary method of anemia diagnosis (WHO, 2007). Hemoglobin (g/dl) concentration was measured in the field on a venous blood sample using a portable Hb801 HemoCue machine (HemoCue AB). The last drop of blood from venepuncture was drawn into a microcuvette and measured. Anemia was defined as Hb < 11 g/dl for children and Hb < 12 g/dl for WRA. Mild anemia corresponds to a Hb concentration of 10.0-10.9 g/dl for children under age 5 and 10.0-11.9 g/dl for nonpregnant WRA; moderate anemia corresponds to a level of 7.0-9.9 g/dl and severe anemia corresponds to a level less than 7.0 g/dl cases of anemia (WHO & Chan, 2011).

| Data collection and management
Two teams of enumerators conducted data collection for approximately 12 days per district. Each team consisted of eight members and a team leader. All participants were assigned a unique numeric ID that was used in data capture forms and subsequent analyses to maintain anonymity. Questionnaire, anthropometry, and specimen data were collected via passcode-protected tablets using KoboToolbox (Android v2022.1.2). Data were transferred daily to the central data processing server. The team leaders and data manager reconciled samples and data forms daily to check for participant totals and completeness of questionnaires.

| Statistical analysis
Statistical analysis of data was performed using SPSS for Windows

| RE SULTS
Hemoglobin measurements were matched for 452 mother-child pairs from Shamva and Mutasa districts from which data are presented.

| Socioeconomic characteristics
The median (±IQR) age for children was 29 ± 14 months with a proportional boy-to-girl ratio (

| Hemoglobin levels and anemia
The overall prevalence of anemia in WRA was 17.9% with a me- TA B L E 1 (Continued) a median (±IQR) Hb level of 11.7 ± 1.5 g/dl. There were no severe cases of anemia in both WRA and children, only moderate and mild cases were observed (Figure 2).

| Predictors of anemia in children and WRA
Pearson Chi-square test indicated that the male sex (p = .040), unimproved dug wells as a source of drinking water (p = .006), on-premise toilet facility (p = .004) (Table 1), and maternal anemia (p = .002) (Table 3) were positively associated with anemia in children, whereas in WRA, unimproved dug wells (p < .001) as a source of drinking water, on-premise toilet facility (p = .060), and lack of agricultural land ownership (p = .007) ( Table 4) were positively associated with anemia.
Multiple regression analysis showed significant associations of anemia in children with having an anemic mother (p = .007) and being a male child (p = .029), while in WRA, anemia was TA B L E 2 Nutritional status and morbidities in children aged 6-59 months in rural Zimbabwe by anemia status.
significantly associated with using unimproved dug wells as a source of drinking water (p = .001) and lack of agricultural land ownership (p = .009) ( Table 5). Children whose mothers were anemic were 2.02 times more likely to be anemic compared to those whose mothers were nonanemic (OR = 2.02; 95% CI 1.21-3.37; p = .007) and the boy child was 0.63 times more likely to be anemic than the girl child (OR = 0.63; 95% CI 0.41-0.95; p = .029). Women from households with unimproved dug wells as sources of drinking water were 0.36 times more likely to be anemic compared to those drinking water from improved wells (OR = 0.36; 95% CI 0.20-0.66; p = .001), and women who did not own agricultural land were 0.51 times more likely to be anemic than those owning agricultural land (OR = 0.51; 95% CI 0.31-0.85; p = .009).

| DISCUSS ION
The study sought to determine the prevalence of anemia in WRA and children aged 6-59 months as well as potential predictors.
According to the WHO classification, our results show that anemia is of moderate public health significance in children (20%-39%) and mild among WRA (5%-19%) (WHO, 2012) in the selected districts.
Thus, we speculate that anemia is potentially a public health issue Zimbabwe. Micronutrient powder supplementation is known to significantly reduce the prevalence of anemia and iron deficiency anemia, and improve Hb levels (Mahfuz et al., 2016;Salam et al., 2013).
In addition, this can also be due to a higher exclusive breastfeeding rate (68.3%) noted in the current study compared to 2015 when it was reported to be 48% (ZIMSTAT, 2015) since breastmilk is an important source of micronutrients (Allen et al., 2006). Higher rates of mild anemia were indicated compared to moderate anemia, and there were no cases of severe anemia in both women and children.
This indicates reduced severity of anemia within the selected district populations. Median Hb concentrations were comparable to other studies conducted in the region; Togo and sub-Saharan African countries (Nambiema et al., 2019;Worku et al., 2022).

| Predictors of anemia in children 6-59 months old (model 1)
Our results show that gender was significantly associated with anemia risk in children. Boys were more likely to be anemic than girls, which was consistent with previous studies in Ethiopia (Melku et al., 2018), Togo (Nambiema et al., 2019), and in Boricha Woreda, Southern Ethiopia (Yoseph & Beyene, 2020). The ZDHS 2015 indicated a small variation by gender (38% of boys were anemic compared with 36% of girls) while other studies in rural Tanzania have found the opposite (Mrimi et al., 2022). The higher prevalence in boys could be related to the higher growth rate in boys resulting in a greater need for iron by the body, which might not be met by the diet (Silla et al., 2013). This could be explained by the growth, development, and maturation stages after birth, including the postnatal spurt followed by the transition from infancy to childhood (Hermanussen, 2016). The growth spurt coupled with termination of lactation by about age 2 years poses a high nutrient demand which may be difficult to meet particularly among boys when compared to girls at age 6-59 months (Khara et al., 2018). Therefore, there is a need to emphasize differences in diet/nutrient requirements based on sex within the generic "one size fits all" recommendations currently being employed in infant and young child feeding (IYCF) education and counseling sessions in health centers.
In the current study, the occurrence of anemia in the mothers was a significant positive predictor of anemia in children. Compared to nonanemic mothers, anemic mothers were significantly more likely to have anemic children. These findings were consistent with results from a multilevel analysis conducted for Southern Africa covering Malawi, Mozambique, and Namibia (Ntenda et al., 2018). This can be attributed to low levels of essential minerals such as iron, zinc, and folate, as well as vitamins A and B12, in the breast milk of the anemic mother, which could also affect the Hb level of the breastfeeding child (Wang et al., 2015). Furthermore, it can potentially be a result of the depletion of iron stores due to successive pregnancies and lactation resulting in iron deficiency anemia (Cane et al., 2022;Cardoso et al., 2012). Additionally, it might be due to that mothers and their children share similar socioecological environments. Thus, their dietary patterns, disease profiles or exposures, and quality of life may be similar (Ntenda et al., 2018), suggestive of intergenerational transmission of poor maternal socioeconomic, health, and F I G U R E 2 Types of anemia in children aged 6-59 months and WRA (overall Anemia: Hb < 11 g/dl for children and Hb < 12 g/dl for WRA; mild anemia; Hb of 10.0-10.9 g/dl for children and 10.0-11.9 g/dl WRA; moderate anemia; Hb of 7.0-9.9 g/dl; and severe anemia Hb <7.0 g/dl).

TA B L E 4
Sociodemographic and nutritional status of WRA in rural Zimbabwe anemia status.  (Christian, 2015), the approach should be considered in low socioeconomic settings like Zimbabwe to ensure positive impacts on anemia prevalence.

| Predictors of anemia in WRA (model 2)
Our results showed that anemia risk in women was positively associated with lack of agricultural land ownership, with those owning land being at a lower risk of being anemic compared to those without land ownership. This can be explained by improved dietary diversity and food security through diversified crop production (Kristjanson et al., 2010). Lack of access to land may predispose populations to food insecurity which consequently increases susceptibility to anemia (FAO, 2000;WFP/FAO, 2014). This is, however, contrary to a similar study in India, which found that women who had their own agricultural land suffered more from anemia compared to those who did not own any agricultural land. This was explained by time tradeoffs between women's participation in agriculture and food preparation in the household which negatively impacted their nutritional status (Jana et al., 2022). The recommendation is for government to devote itself to food security programs that promote land ownership. However, literature is limited on the linkages between agricultural land ownership and anemia. Further research could help clarify the linkage so as to inform livelihood programs.
Results from the current study indicated that unimproved dug wells as a source of drinking water were a positive predictor of anemia in WRA, with those relying on unimproved dug wells being at a greater risk of anemia than those using improved wells. This result was consistent with a study undertaken in Nepal (Gautam et al., 2019), India, and Bangladesh (Jana et al., 2022) where women in households using dug wells as sources of drinking water were significantly associated with an increased risk of developing anemia.
In the case of Bangladesh, it was partly attributable to fluoride and arsenic in groundwater (Del Bello, 2020;Jana et al., 2022). In some contexts, it may be due to intestinal worm infections (Chelkeba et al., 2022;WHO, 2022) which cause gut blood losses, malabsorption of nutrients, loss of appetite, and anemia due to loss of iron and impaired protein metabolism (Pullan et al., 2014). However, in other contexts it might be due to higher incidences of diarrheal diseases and upper respiratory tract infections (Marakwet & SNV, 2018).
Consistent findings were observed in Peru (Mougenot et al., 2020) and Nepal (Coffey & Geruso, 2015) where safe drinking water from improved dug wells was associated with reduced anemia prevalence.
While it is expected that access to improved water and sanitation facilities lowers anemia prevalence, in the current study, on-premise toilet facilities increased the prevalence of anemia. Similar mixeddirection findings were observed in previous studies in Zimbabwe (SHINE, 2015) and other surveys (Yu et al., 2019) which showed that access to improved water and sanitation does not consistently predict lower anemia prevalence. Despite the heterogeneity, hygiene interventions including hand washing with soap and improved water and sanitation can reduce the disease risk associated with anemia (Borghi et al., 2002;Ejemot-Nwadiaro et al., 2015;Ejemot-Nwadiaro et al., 2008;WHO, 2022). Therefore, promoting policies and practices that strengthen access to improved WASH should still be considered for reducing anemia prevalence. Additionally, the promotion of regular water consumption as a source of minerals is warranted since iron in groundwater provides a good source of absorbable iron for humans and can contribute to optimal iron and Hb status among populations (Choudhury et al., 2022).

| Limitations of study
The cross-sectional nature of the present study limits the ability to make inferences on causation. The current study is exploratory and a substudy from a larger trial with preplanned outcomes. Tests of statistical significance should be interpreted bearing this in mind, as well as the possibility of false positives (Type 1 error) due to multiple comparisons. Adjustment for multiple comparisons was not conducted since this was an exploratory study and we wanted to avoid false negatives (Type 2 error). Additional dedicated studies are needed to confirm the results. The study did not differentiate the types of anemia, and sample size would be a limiting factor for greater refinement of several related research questions. Nevertheless, the present study contributes to the body of knowledge that shows the associations between maternal and child anemia, and environmental and nutritional characteristics in vulnerable populations.

| CON CLUS IONS
The current study showed that the prevalence of anemia was moderate, mainly affecting male children; with maternal anemia being a positive predictor of child anemia. Unimproved dug wells as a source of drinking water and lack of ownership of agricultural land were positive predictors of anemia in WRA. Interventions that focus on improving preconceptual and maternal nutritional status, MNP supplementation, access to improved water and sanitation, and agricultural land ownership may be important strategies to reduce anemia in vulnerable populations from low-and lower middle-income countries.

ACK N OWLED G M ENTS
The authors declare no conflict of interest and thank all the parents and caregivers of the infants for participating in the study. Gratitude

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy/ethical restrictions.

E TH I C A L A PPROVA L
The funding bodies had no influence on the study design, data collection, analysis or interpretation of the data, writing of the manuscript, or the decision to submit the manuscript for publi-