Interventions for Addressing Anemia Among Children and Adolescents: An Overview of Systematic Reviews

Context: Anemia is a public health problem that can lead to growth, cognitive, and motor impairments. Objective: To collate evidence on interventions for addressing childhood and adolescent anemia. Methods: In this overview of systematic reviews, we included Cochrane as well as non-Cochrane systematic reviews (SRs) irrespective of language and publication status. Two sets of review authors independently screened articles for eligibility and extracted data from relevant SRs. We present data in a tabular format and summarize results based on outcome reported, age of participants, and type of interventions. We also adopt a “measurement for change” approach to assess the utility of measurement for development of interventions in childhood and adolescent anemia. Results: Our search yielded 2,601 records of which 31 SRs were found eligible for inclusion. Results were favorable for fortification and supplementation with clear reduction in the risk of anemia and increase in hemoglobin levels across all age groups. Other interventions reported by the SRs were inconclusive and suggest further research. Conclusions: Current evidence suggests that fortification or supplementation with iron and micronutrients leads to better reduction in the risk of anemia and improvements in hemoglobin levels among children and adolescents. Results of this overview can help decision makers in informing selection of interventions to address childhood and adolescent anemia. Review Registration: PROSPERO CRD42016053687.


INTRODUCTION
Anemia is a condition in which there is either a decrease in the number of red blood corpuscles (RBCs) or a decrease in the total amount of hemoglobin (Hb) (1). The World Health Organization (WHO) defines childhood anemia as Hb concentration below 11 grams per deciliter in children between 6 and 59 months, below 11.5 grams per deciliter in children between 5 and 11 years, and below 12 grams per deciliter in children between 12 and 14 years of age (2).
The highest burden of anemia is in low-and middle-income countries (LMICs). Estimates show that almost 90,000 deaths have occurred due to iron deficiency anemia across all age groups (3). India, with ∼89 million children with anemia is the highest contributor of childhood anemia (4,5). Iron deficiency anemia is the most common cause of anemia on a global level. Other causes of childhood anemia include nutritional deficiencies, such as deficiency of folic acid, iron, or vitamin B 12 ; chronic conditions, such as inflammatory disorders; and parasitic infestations (2). Disorders associated with synthesis of hemoglobin, formation of RBCs, or survival of RBCs are the other, less common causes of anemia in children (2). Malaria, HIV, tuberculosis, and helminthic infestations have been reported to lead to substantial burden of anemia, especially in developing countries (2,6,7). The severity of anemia is higher in children under 8 years old as compared with children over 8 years (8). Regardless of the etiology of anemia, impaired physical growth, motor development, and cognitive development have been observed in anemic children (7,(9)(10)(11)(12).
WHO and the United Nations International Children's Emergency Fund (UNICEF) recommend multipronged public health preventive strategies, such as supplementation of nutrients, fortification of food with nutrients, educational interventions, and prevention and control of parasitic and protozoal infestations (13).
Our comprehensive search strategy reveals many systematic reviews (SRs) that assess the efficacy of different public health interventions for addressing childhood anemia. However, no synthesis of these reviews has been reported in an overview. Hence, the Indian Council of Medical Research (ICMR) task force on childhood and adolescent anemia commissioned this overview in order to collate evidence on the efficacy of various interventions in addressing childhood and adolescent anemia in order to inform programs and a future research agenda. This review could serve as a baseline for clinicians, stakeholders, and decision makers to intervene in childhood and adolescent anemia.

METHODS
A letter of intent was invited from Indian scientists with experience in evidence synthesis. A review team of experienced researchers was constituted to work on the overview of SRs. The protocol was registered in the PROSPERO prospective register of systematic reviews (Registration number: CRD42016053687) after consultation with national experts (14).

Inclusion Criteria
Irrespective of language and publication status, we included SRs, evaluating the effect of 1. Any one intervention, or 2. Combination of interventions, such as iron of ferrous sulfate (FS) or zinc (Zn) or multi micro nutrients (MMN) or vitamin supplementation, or 3. Iron or FS or Zn or MMN or vitamin fortification, or 4. Anti-helminthic treatment 5. Treatment or prevention of malaria, or 6. Treatment of H. pylori, or 7. Water, sanitation, and hygiene (WASH) interventions.
We included SRs designed for addressing anemia in children between 6 months and 19 years of age. We included studies regardless of gender, study setting, and severity of anemia. We excluded non-SRs, SRs targeting populations other than children and adolescents, those done on overtly diseased children or adolescents, those providing data on overlapping age groups, and SRs that did not assess the outcomes of interest of this overview. We assessed outcomes, namely anemia, Hb levels, and adverse events.

Selection of Reviews
We undertook screening of SRs using Covidence SR software (15). Two overview authors (PM and RH) independently screened all SRs (retrieved in the search process) on the basis of title and abstract. Two overview authors (RH and NK) then independently screened the full texts of identified reviews for inclusion. A third reviewer (BU) resolved disagreement among primary reviewers through discussion. We prepared a PRISMA flow diagram to map the inclusion and exclusion of SRs (16).

Data Extraction
We developed and pilot-tested our data extraction form for its suitability and usability. Three review authors (AS, PM, and NK) independently extracted data regarding characteristics of the included participants, interventions, comparisons, outcomes, and methodological quality of the included reviews. We contacted reviewers to seek missing information if any.

Data Synthesis
We followed standard procedures given by the Cochrane Handbook for Systematic Reviews of Interventions (19). We obtained data from included SRs and presented them in a tabular and graphical format. In case of duplicate publications, we considered them as a single report. We summarized results based on outcome reported, age of participants, and type of interventions as risk ratio (RR), mean difference (MD), standardized mean difference (SMD), or weighted mean difference (WMD) with 95% confidence intervals (95% CI). As an overview, we mainly focused on summarizing and thematically categorizing the intervention and its effectiveness and not taking a more interpretative approach during synthesis.

Search Results
Our search yielded 2,601 records. After removal of duplicates, 2,204 records remained, and they were screened for title and abstract. At this stage, we excluded 2,043 articles, and a total of 161 full-text articles were screened for eligibility. We excluded 130 articles and included 31 SRs that met our inclusion criteria (Figure 1).

Description of Included Reviews
We included 31 SRs in this overview, the salient features of which are depicted in Table 1.
We included nine Cochrane reviews (20,21,28,31,38,41,46,49,50) and 22 non-Cochrane SRs (22-27, 29, 30, 32-37, 39, 40, 42-45, 47, 48). Except for Cembranel et al. (23), all other reviews are in the English language. Updating the version of existing SRs was reported only by Cochrane reviews. The review by Neuberger et al. (21) is the updated version of previous publications (51,52). Taylor-Robinson et al. (46) is updated from a previous five Cochrane reviews (53)(54)(55). One Cochrane review by De-Regil et al. (38) has been reprinted in a Cochrane review journal (56), and hence, to avoid duplication, we considered both reviews as a single report (38) Included reviews differed in the age groups of participants: 11 reviews included participants <5 years of age (23, 25-27, 30, 33, 35, 36, 38, 40, 41). Low et al. (20), Das et al. (34), and Huang et al. (48) included participants of age groups beyond the inclusion range in our overview, but they reported separate data for children as a subgroup analysis and, hence, were included. Aaron et al. (32) included a trial on pregnant women that was not included in the meta-analysis of anemia. Another review (35) considered included women but did not identify studies on women and, hence, was included in this overview. Except for Low et al. (24), all included reviews included participants of both genders. All SRs except McDonagh et al. (42) included studies from LMICs and developing countries. Sun et al. (45) included specifically the Chinese population. One SR included studies each on children from malaria endemic areas (21), areas endemic for intestinal helminths (46), and areas endemic for anemia (50).

Effects of Interventions
Anemia A total of 18 SRs addressed the effect of different interventions on anemia in children and adolescents (20-25, 28, 30-36, 40, 41, 45, 50) (Table 2 and Figure 2). We did not find any review that assessed the effect of treatment of deworming, H. pylori infection, or WASH intervention on anemia status.

Adverse Events
A total of six reviews addressed the adverse effects of different interventions for preventing, treating, and controlling anemia (21,22,24,25,28,31) (Table 4). We did not find any review that assessed the adverse effects of antimalarial intervention, deworming, H. pylori treatment, or WASH interventions.
In children under 18 years old residing in malaria endemic areas, there is a significantly higher risk of diarrheal episodes per patient-month by 15% with iron supplements (RR = 1. 15

DISCUSSION
We evaluated 31 SRs that assessed the effects of different interventions for addressing anemia in children and adolescents. The highest number of SRs addressed food fortification and iron supplementation. None of the reviews were from India or included participants from India. Overall, the quality of the evidence reported by the SRs varied between moderate to low (Supplementary File 2).
In infants, iron fortification showed better results as compared with MMN fortification (30,34). In children between 4 months and 2 years of age; home fortification, when used as a preventive measure, showed significant reduction in the risk of anemia followed by iron supplementation (25,38,40). However, when used as a treatment measure, home fortification insignificantly increased the risk of anemia (40). This difference in the effect of home fortification when used as a preventive measure and as a treatment measure might be because the prevention trials were conducted on unselected populations, and the participants in the control group did not receive any treatment. On the contrary, the therapeutic trials were conducted on anemic subjects, and the participants in the control received medicinal iron drops (40).
In children under 5 years old and in school-age children, fortification of food with iron and MMN showed better improvement in Hb (moderate QoE) (34,36). In apparently healthy children under 12 years old, intermittent supplementation of iron alone or in combination with other nutrients significantly increased Hb concentrations, and this positive response did not differ with frequency or duration of supplementation or age of children (28). The risk of anemia is higher with intermittent iron/FA/MMN supplementation as compared with daily iron/FA/MMN supplementation (28). A school-based deworming program resulted in a decrease in the prevalence of anemia in children (47). Programs that combined interventions, such as deworming with iron and nutrition supplementation, led to better effects as compared with single interventions (47). Daily oral iron supplementation with or without other vitamins (20,24) and MMN-fortified beverages (32) showed maximum reduction in the risk of anemia in children between 3.5 and 18 years. In malaria endemic areas, iron supplementation with antimalarial treatment reduced the risk of anemia (21). There is low-quality evidence that community deworming has little or no effect (46).
In summary, results were favorable for iron with MMN fortification until school age and iron with MMN supplementation in older children with a clear reduction in the risk of anemia and increase in Hb among the intervention groups of individual SRs. However, other interventions reported by SRs were inconclusive and all SRs suggested further research on the same.
The chances of developing anemia were less in children who received daily supplements as compared with those who received intermittent supplements. However, studies suggest that, in places and settings in which daily supplementation is not practical, intermittent supplementation can be used as a public health intervention to address childhood anemia. We did not find any review that assessed the effect of treatment of deworming, H. pylori infection, or WASH intervention on anemia status.
Limitations of the review: This overview examined the available evidence with respect to interventions addressing childhood and adolescent anemia without any restrictions on the type of interventions or language of publication. We did not restrict the search to Cochrane reviews, and inclusion of non-Cochrane reviews makes it more comprehensive. However,  wide physiological variations during different stages of childhood could make the uniform application of interventions difficult. We acknowledge that not all SRs included in this overview were up to date. We found limited reviews on deworming, antimalarial interventions, H. pylori intervention, and WASH intervention, and hence, there is a need for primary studies in these areas. Some SRs did not consider socio-economic or nutritional status at baseline or did not report this clearly, which could have had an influence on some findings. In some reviews, there were too few data to reach a firm conclusion. Strengths of the review: We estimate the potential bias in this overview as low. We followed the Cochrane handbook for methodology (19). The search was as comprehensive as possible. Screening of studies, data extraction, and assessing the methodological quality of reviews were done in duplicate. The authors of this review are from diverse disciplines (e.g., public health, biostatistics, nutrition, physiology, and maternal and child health), and this internal diversity may be an asset of this overview.
After an extensive literature search, we did not come across any overview of SRs that has addressed this area. This overview can help healthcare providers, consumers, public health specialists, and decision makers by providing a "bird's-eye view" of the efficacy of different public health interventions across reviews and age groups.

Implications of All the Available Evidence
This overview identified interventions that had favorable effect sizes (which public health experts might contemplate using) as well as interventions that were mostly ineffective (which public health experts might want to restrict). More research is needed to assess the effect of deworming, antimalarial intervention, H. pylori treatment, and WASH intervention on anemia. Also, better recording of adverse events in primary studies is needed.

CONCLUSION
Results were favorable for iron and MMN fortification and supplementation with a clear reduction in the risk of anemia and increase in Hb levels across all age groups. Other interventions reported by the SRs were inconclusive and suggest further research. MMN improves outcomes better than single micronutrients. Also, better recording of adverse effects of different intervention is needed.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
AS envisioned the idea of the review. RV developed and ran the search strategies. AS and PM developed the protocol. PM and RH screened title and abstract. RH and NK screened full texts. BU resolved disagreement among primary reviewers (for screening) through discussion. AS, PM, and MK extracted data. AS and PM assessed the methodological quality of included SRs. NN, AG, and SQ provided technical advice. MK and PM drafted the manuscript with input from AS, BU, NN, AG, and SQ. All authors have contributed significantly to this overview, read the manuscript, and participated in writing and revision.