The nutritional status of children living within institutionalized care: a systematic review

Background There are an estimated 2.7 million children living within institutionalized care worldwide. This review aimed to evaluate currently available data on the nutrition status of children living within institutionalized care. Methods We searched four databases (Pubmed/Medline, CINHAL Plus, Embase and Global Health Database) for relevant articles published from January 1990 to January 2019. Studies that included information on anthropometry or micronutrient status of children living within institutionalized care were eligible for inclusion. The review is registered on PROSPERO: CRD42019117103. Results From 3,602 titles screened, we reviewed 98 full texts, of which 25 papers were eligible. Two (8%) studies reported data from multiple countries, nine (36%) were from Asia, four (16%) from Africa, three (12%) from Eastern Europe, four (16%) from the European Union and one (4%) from each of the remaining regions (Middle East, South America and the Caribbean). Twenty-two (88%) were cross sectional. Ten (40%) of the studies focused on children >5 years, seven (28%) on children <5 years, seven (28%) covered a wide age range and one did not include ages. Low birth weight prevalence ranged from 25–39%. Only five (20%) included information on children with disabilities and reported prevalence from 8–75%. Prevalence of undernutrition varied between ages, sites and countries: stunting ranged from 9–72%; wasting from 0–27%; underweight from 7–79%; low BMI from 5–27%. Overweight/obesity ranged from 10–32% and small head circumference from 17–41%. The prevalence of HIV was from 2–23% and anemia from 3–90%. Skin conditions or infections ranged from 10–31% and parasites from 6–76%. Half the studies with dietary information found inadequate intake or diet diversity. Younger children were typically more malnourished than older children, with a few exceptions. Children living within institutions were more malnourished than community peers, although children living in communities were also often below growth standards. High risk of bias was found. Conclusions This study highlights the limited amount of evidence-based data available on the nutritional status of children in institutions. Of the studies reviewed, children living within institutionalized care were commonly malnourished, with undernutrition affecting young children particularly. Micronutrient deficiencies and obesity were also prevalent. Data quality was often poor: as well as suboptimal reporting of anthropometry, few looked for or described disabilities, despite disability being common in this population and having a large potential impact on nutrition status. Taken together, these findings suggest a need for greater focus on improving nutrition for younger children in institutions, especially those with disabilities. More information is needed about the nutritional status of the millions of children living within institutionalized care to fully address their right and need for healthy development.


INTRODUCTION
Malnutrition impacts millions of children around the world (Black et al., 2013;The World Bank Group, 2019;UNICEF, 2019). In 2018 for children younger than 5 years old, 49 million children were wasted, just under one in four (21.9%) were stunted and 5.9% were overweight (UNICEF, 2019). Almost half of the deaths among children younger than 5 years old have undernutrition as an underlying factor (Black et al., 2013;UNICEF, 2019). In some countries, up to half of adolescents are stunted, as many as 11% are too thin, up to 5% are obese and over 50% are anemic (Black et al., 2013). Being malnourished has many adverse consequences including increased risk and severity of infections, increased risk of disability, and death (Black et al., 2013;Groce et al., 2014;McDonald et al., 2013;Myatt et al., 2018). This can be a part of a cyclical interaction between infections and undernutrition which leads to poor nutritional status, illnesses and impacted growth.The first 1,000 days of a child's life are particularly important because poor nutrition at this stage also predisposes children to long-term impairments such as stunted growth, impaired cognition and poor performance at school and work (Black et al., 2013;UNICEF, 2019).
Some children are at higher risk of malnutrition, such as orphans and children living within institutionalized care (UNICEF, 2019). UNICEF estimates that there are some 140 million orphans worldwide who have lost either one or both of their parents (UNICEF, 2017). Although most orphans live with other family members, some live in institutionalized care or residential care facilities (UNICEF, 2017). Institutionalized care is defined by the United Nations as residential care that is provided in any non-family-based group setting, including all other short-and long-term residential care facilities (United Nations General Assembly, 2009). Many non-orphans live in institutionalized care for a variety of reasons, including social or economic (van IJzendoorn et al., 2011;The Children's Health Care Collaborative Study Group, 1994). These children are also vulnerable (Baron, Baron

METHODS
The aim of our review was to better understand the current nutritional status of children in care by looking at anthropometric and nutritional status indicators in relation to age, disability, geography, gender and related factors, with an ultimate goal of improving policy and practice to better meet the needs of this unique and vulnerable population.
We analyzed existing published peer-reviewed literature on the nutrition status of children in institutional care by examining anthropometric data, micronutrient status and other factors including disability status, gender and age. PRISMA guidelines were followed throughout the review process and a PROSPERO registration was completed prior to the start of the study (PROSPERO 2019: CRD42019117103, https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=117103) (Moher et al., 2009;National Institute for Health Research, 2019).
The review primarily evaluated observational and intervention studies. Inclusion criteria included material published between January 1990 and January 2019 in English and contained research related to orphanages/institutionalized care, children, nutrition, anthropometric data or micronutrient status. We selected these dates because the Convention on the Rights of the Child went into effect in 1990, and since then, there have been significant changes in institutional care and changes in the understanding of the needs of children in institution-based care (IBC) (Frank et al., 1996; United Nations Human Rights Office of the High Commissioner).
In order to be included in this review, the studies must have addressed a population of children younger than 18 years old (with the exception of one study which included children as old as 20 years but was retained for informational value), been peer reviewed and included at least one measurement of nutrition status through standardized tools, such as WHO Growth Standards or WHO Growth References and definitions (World Health Organization, 2019a;World Health Organization, 2019b). Anthropometric indicators of interest included: weight for age, length/height for age, weight for length/height, head circumference for age and mid-upper arm circumference for age. Micronutrient status, clinical signs/symptoms and dietary information were also included when available. Emily DeLacey, the principal investigator, and Dr. Marko Kerac determined and used the search strategy. Four electronic databases were searched through OVID from December 2018 through January 2019: Pubmed/Medline, CINHAL Plus, Embase and Global Health Database. For details of our search strategy, see Appendix S1. Initial article screening was based on title and abstract, following which full texts were assessed for eligibility against our pre-specified inclusion/exclusion criteria. Discussions with the research team resolved any questions of eligibility with Dr. Cally Tann deciding any discords. A data extraction table was used to summarize key information from the final selection of articles into tables and columns organized by related themes and areas.
Nutritional status was determined according to reported anthropometry, whether reported by z-scores (standard deviations from a reference population) or percentiles. Micronutrient status and intake were also reported on and included prevalence of anemia or micronutrient deficiencies. Other key data areas included disability status, birth weight, sex, age, dietary intake and any reported disease, illness or infection which could impact nutrition status. Heterogeneity in the type of interventions prevented our ability to conduct a meta-analysis of the study, so a narrative synthesis was used.

RESULTS
We found a total of 3,973 papers. After 371 duplicates were removed, the remaining 3,602, were screened by title and abstract. All but 98 articles were excluded during this phase. Of the 98 identified as potentially eligible, we were unable to locate seven, 53 had insufficient anthropometry or used non-standard measurements, 10 did not have appropriate population or study type and three were excluded because the anthropometric data existed in another study. Twenty-five studies met our inclusion criteria (Fig. 1). Most studies (22, 88%) were cross sectional ( Table 1). The most commonly researched region was Asia with nine studies (36%), followed by Africa with four studies (16%). Three (12%) were from Eastern Europe, four (16%) from the European Union and one (4%) from each of the remaining regions (Middle East, South America and the Caribbean). Kenya and India were the most commonly researched countries and were each included in four studies (Table 1). Ten (40%) focused on children older than 5 years, seven (28%) on children younger than 5 years, seven (28%) covered a wide age range and one did not include ages. Twelve (48%) included control or comparison groups of children who were community children (CC) or orphaned, separated or abandoned children living in family-based care (FBC), or children living on the streets (CLS). Control groups were typically orphaned children living in family-based care (FBC) or community children (CC) with no history of institutional care and the groups were selected from different settings including from local schools, communities, clinics or hospitals, lists, house-to-house census or other child-related programs (Braitstein et al., 2013;Johnson et al., 2010;Whetten et al., 2014). Eight (32%) studies mentioned or analyzed gender differences (Tables 1 and 2). A history of low birth weight (LBW) were also common (25% to 39%).

Children with disabilities
Of the 25 studies reviewed, 12 (48%) did not state whether they included children with disabilities (Tables 1 and 2). Eight (32%) of the studies stated that children with disabilities were excluded, leaving only five (20%) mentioning children with disabilities in their reporting, but either excluded them from analysis or did not state whether or not they were excluded. Only one study included any anthropometric measurements for children with disabilities (Lewindon et al., 1997). The St. Petersburg-USA Orphanage Research Team found that 21% of children had disabilities (The St Petersburg-USA Orphanage Research Team, 2005). Miller and colleagues found 16% of institution-based children (IBC) had significant disabilities and 75% had developmental delays (Miller et al., 2006).

Anthropometrics
Undernutrition, micronutrient deficiencies and overweight/obesity were reported in varying ways. Prevalence of undernutrition differed markedly: stunting (low length/height for age) from 9 to 72%; wasting (low weight for length/height) from 0 to 27%; underweight (low weight for age) from 7 to 79%; low BMI (body mass index) ranged from 5 to 27% (Table 2). Ten to 32% of children were overweight or obese. Panpanich et al. found children younger than 5 years old to be more stunted, wasted and underweight than older children and below WHO growth standards (Panpanich et al., 1999). The prevalence of small head circumference ranged from 17 to 41%.

Micronutrients, clinical signs/symptoms and infections
Clinical signs or symptoms were reported in 48% (12) of the studies (Table 3). Five (20%) mentioned HIV but two of these were conducted in institutions for children with HIV (Kapavarapu et al., 2012;Myint et al., 2012). Excluding the facilities for children with HIV, HIV prevalence was from 2 to 23%. One study found a higher prevalence of morbidity among children in IBC than CC (p < 0.05) (Mwaniki, Makokha & Muttunga, 2014). The prevalence of parasites ranged from 6 and 76%, with Lesho and colleagues finding 10% of children in IBC having three or more parasites (Lesho et al., 2002). Skin infections, varicella zoster, tuberculosis, impetigo, dental issues, ear/nose/throat problems, respiratory infections, diarrhea and other conditions or illnesses were frequently reported among IBC (Table 3). Skin conditions or infections ranged between 10 and 31%, and Kapavarapu and colleagues found 75% of children had an infection within the first three months of admission to a site (Kapavarapu et al., 2012). Seven (28%) reported on micronutrient status or intake and the prevalence of anemia ranged from 3 to 90%. Hearst and colleagues found over a third of children had low vitamin D . Other micronutrient deficiencies discussed included iodine, zinc, albumin, as well as vitamins A and B (Table 3). Edema, conjunctival pallor, xerophthalmia and goiters were found more in children in IBC than those living in FBC (Aboud et al., 1991).

Dietary diversity, intake and food security
Eight (32%) studies discussed dietary diversity, intake or food security (Table 3). Mwaniki and colleagues found that diet diversity was lower in children living in IBC than for CC (p < 0.05). Diets were reported to have a high reliance on starches and legumes (Mwaniki, Makokha & Muttunga, 2014). Of the studies that assessed dietary intake, 50% found adequate intake. Dietary adequacy varied; from children in IBC at 3.9 times higher risk of consuming inadequate calories to having 362% higher intake than estimated average requirements for some nutrients. The one study which reported on food security found that children in IBC had higher food security when compared to children in FBC, 42% vs. 2% (Braitstein et al., 2013).

DISCUSSION
The nutritional status of children living in institutions has the potential to adversely impact their health and well-being, yet out of 3,602 papers from four major databases, only 25 peer-reviewed papers presented evidence based findings on the children's nutrition status ( Fig. 1). All 25 reviewed studies indicated that many of the children in institutionalized care faced some form of malnutrition. The available data suggests that children living within institutionalized care are commonly malnourished: affected by undernutrition, overweight and micronutrient deficiencies. With few exceptions, mostly of older children, children living within institutionalized care were significantly below standards for growth, diet and micronutrient status and were often below comparison groups of their community peers. Nutrition status varied between care centers and between the ages of children, with younger children at a higher risk of being malnourished. There may be a number of reasons why this is the case, such as younger children have a harder time feeding themselves, especially if disabilities are present, and young, poorly nourished children are at risk of not surviving to become older children in institutional settings (McDonald et al., 2013;Myatt et al., 2018;The Children's Health Care Collaborative Study Group, 1994). Diet inadequacy, micronutrient deficiencies and illnesses or infections were also found to be prevalent in children of all ages. To our knowledge, this is the first systematic review of the nutrition status of children living within institutionalized care. It is important because 2.7 million children worldwide   ---''Irrespective of the ART status, a decrease in underweight, stunting and wasting was seen at the end of 36 months. There was an observed higher rate of zscore increase among children not yet on ART compared to that of those who were on ART was probably attributable to the fact that children on ART had a more advanced forms of disease along with comorbidities which resulted in slower rate of improvement in growth than children with a milder form of disease and who did not need to be treated with ART. All received age and gender appropriate nutrition along with additional nutrition supplements such as iron when required. These results suggest that dietary support (both macronutrients and micronutrients) may have a role in improving nutritional outcomes in HIV-infected individuals, thereby improving quality of life and perhaps indirectly reducing disease-related mortality.''        Mwaniki, Makokha & Muttunga (2014) Using a 24hr diet recall and Nutri Survey program, diets were assessed. A total of 63 and 37 food items were consumed by the CC and IBC respectively. Only 7.2% of IBC consumed more than three food groups compared to 45.2% of CC. 92.9% of IBC and 54.8% of CC consumed less than four food groups (p < 0.05). CC had significantly (p < 0.05) higher diversity of foods served than IBC. Energy intake: The total mean energy intake among CC was 1,890 Kcal per day and was significantly higher (p < 0.05) than that of IBC. The intake of energy by IBC who took lunch was 1,547 Kcal compared to the energy intake of CC who also took the three meals of the day (p < 0.05). The mean energy intake of IBC who did not take lunch was less than half of that of CC. IBC who attended school away from the orphanage had two meals (mainly breakfast and supper) in a day during school days and three meals during the weekend and did not meet their daily needs compared to CC who always had three meals. IBC had 3.9 times higher risk of consuming inadequate calories compared to CC. Orphanages tend toward exclusive reliance on starches and legumes. Food in orphanages mainly depended on donations.  HIV rates: IBC 23% (3/13) Asia Hearst et al. (2014) -The nutritional status, based on blood biomarkers, revealed that 37.1% of the children were anemic, 21.4% had low albumin, 38.1% had low vitamin D, 5.5% were iodinedeficient and 2% had low serum zinc.    Food intake was obtained by 24 h food-weighing method for seven days. The average food intake were calculated by using the Institute of Nutrition and Food Science. Total food intake was about double the intake of similar children in the 1995-96 nutrition survey. Mean energy (2,270 kcal), protein (65 grams), carbohydrate (335 grams) and fat intake (73 grams). Carbohydrates, protein and fat provide 59%, 12% and 29% of total calories respectively. Protein intake was 65 grams, about 50% higher than the requirement and the 1995-96 nutrition survey of the urban location of the same group. Energy intake was found 20% higher than requirement and about 42% higher compared to 1995-96 nutrition survey. Average intake of IBC was higher than the national intake and the nutritional status of IBC was also found to be better than the national average by any nutritional criteria. Studies consider this to be potentially attributed to better health and care system prevailing in the orphanage apart provision of highcalorie and protein-rich food and that the nutritional status IBC, who are nutritionally disadvantageous, can be improved through organized feeding and better hygienic conditions.   (2015) Diets were chemically analyzed using the Kjeldahl method and Soxhlet method and compared to Polish Estimated Average Requirements. Results indicate that daily diets meet about 80% of recommended intake of energy, fat and carbohydrates. The intake of protein with daily diets exceeded EAR value and ranged from 115 to 362% (average 214.2%). It has been also found that the intake of basic nutrients was varied, coefficient variation (CV) ranged from 22.2% to 27.1%. Boys, compared to girls, spent almost twice as much time on physical activity.   live in IBC and there are a multitude of factors and reasons why they may be affected by different types of malnutrition. The extent and direction of this has not been well studied nor is it currently being effectively monitored or assessed (Petrowski, Cappa & Gross, 2017). Representing an inherently high-risk population, there are many reasons why we would expect undernutrition to be common: this was indeed observed in our review. Conversely, there are some reasons why IBC may offer opportunities for good nutrition and access to services, such as better food security, more reliable funding sources and access to specialized therapy or treatment. These ideal factors may not be possible or available for families affected by different economic circumstances living in the same communities (Braitstein et al., 2013;Panpanich et al., 1999;Whetten et al., 2014). Our review, which used a comprehensive search strategy, also notably highlights a lack of well reported and standardized evidence. Only 19 countries were represented in our findings, despite Petrowski and colleagues finding 140 countries with data on children in institutions and this limited our ability to determine trends or region-specific patterns and risk factors (Petrowski, Cappa & Gross, 2017).

Children with disabilities and children with low birth weight
A key observation is that few studies mentioned children with disabilities and only one included anthropometric analysis (Tables 1 and 2) (Lewindon et al., 1997). Children with disabilities are disproportionately present in institutionalized care settings. (Baron, Baron & Spencer, 2001;The Children's Health Care Collaborative Study Group, 1994;The St Petersburg-USA Orphanage Research Team, 2005). They are already at increased risk when they enter care centers because disabilities can increase the likelihood of being malnourished due to feeding challenges, malabsorption and/or intake needs. In addition, children with disabilities face the risk of their disabilities worsening in environments that do not meet their individual needs (Groce et al., 2014;Kroupina et al., 2014). Children with some types of disabilities may have higher caloric needs or require specialized diets or additional supports at mealtimes (Groce et al., 2014;Johnson & Gunnar, 2011;Johnson et al., 2010;Kroupina et al., 2014;The Children's Health Care Collaborative Study Group, 1994;The St. Petersburg-USA Orphanage Research Team, 2008).
Children in care or those who stay in care the longest may have more disabilities, more underlying diseases or more complex backgrounds-including a history of low birth weight (LBW), and therefore may require more focused care (The Children's Health Care Collaborative Study Group, 1994;The St Petersburg-USA Orphanage Research Team, 2005). Even when provided with adequate diet and medical care, these groups may be more dependent on caregivers for feeding, or need specialized approaches to feeding such as supportive seating and positioning, adaptive skill development and an extended time to eat (Johnson & Gunnar, 2011). When children enter into care they are often in poor health and those who stay the longest, such as some children with disabilities, are frequently in worse condition compared to children who are healthy at admission (Groce et al., 2014;The St Petersburg-USA Orphanage Research Team, 2005). These issues are important to highlight because becoming malnourished while living in an institution can also increase the risk of children developing a disability (Groce et al., 2014).
High prevalences of low birth weight infants were common within institutions; although child history, records or tracking were often limited (Johnson et al., 2010;Kroupina et al., 2014;The Children's Health Care Collaborative Study Group, 1994;The St Petersburg-USA Orphanage Research Team, 2005). Health status at birth was found to be a significant determinant of development. Growth trajectories and pre-and perinatal circumstances influence children's development in care: nutrition needs vary depending on individual growth rates and the presence of preexisting nutrition deficiencies (Johnson et al., 2010;Kroupina et al., 2014;Martins et al., 2013). Gunnar (2011) andJohnson et al. (2010) found that during early rapid-growth phases, the effects of even modest nutritional deficits can become magnified. Age, age at admission and length of stay were other key factors identified that were associated with nutritional status (Chowdhury et al., 2017;Kroupina et al., 2014;Martins et al., 2013;Panpanich et al., 1999).

Gender and malnutrition
Gender is also important to consider because programs and policies should be evidencebased and equitable, offering support to those most in need (Theobald et al., 2017). However, our review found that only nine of the studies compared genders. Of these, two found that girls were more malnourished, three found boys were more malnourished than girls and another four found both groups had similarly high prevalence of malnutrition or no significant difference in nutritional status by gender. We thus have mixed and inconclusive evidence of malnutrition or risk of malnutrition being linked to gender of children in institutional care (Table 2). This may be a very context specific issue where social as well as biological factors play a role.

Anthropometrics
Frequently the prevalence of low birth weight, stunting, wasting, underweight, anemia, and overweight was higher in IBC compared to the global prevalence for children younger than 5 years old (The World Bank Group, 2019). Paralleling global trends, the triple burden of malnutrition (undernutrition, micronutrient deficiencies and overweight/ obesity) also needs to be examined in IBC (Black et al., 2013;UNICEF, 2019). Although only a few studies reported on overweight, when it was reported, the prevalence was high, especially for adolescents. Future studies should report on overweight as well as underweight and micronutrient deficiencies. A positive feature of the studies reviewed was that many had peer groups for comparison; this is helpful because many children in the surrounding community may also deviate from WHO growth standards and it is helpful to see the nutritional status of children in IBC in local as well as global context. Multiple studies found that children in IBC were more undernourished than community children (CC) or children living in family-based care (FBC) ( Table 2). Six studies indicated that peers within the community were more likely to be malnourished than children living within IBC, although this varied a bit by age. This could be in part due to children in care receiving adequate nutrition, routine meals and health screenings, especially for children who have HIV, and/or it could reflect the challenges faced by families in those communities (Braitstein et al., 2013;Panpanich et al., 1999;Sarma et al., 1991;Whetten et al., 2014;Whetten et al., 2009;Zahid & Karim, 2013).

Clinical signs/symptoms, micronutrient status and infections
HIV prevalence was higher than global percentages for the few sites that reported it (The World Bank Group, 2019). HIV can be a significant risk factor for becoming malnourished and is also a contributing factor to children ending up in care (Kotler, 1989;Leyenaar, 2005). Another clear gap was that less than a third of the studies reported on micronutrient status and less than half reported on clinical signs/ symptoms or infections (Table 3). Micronutrient deficiencies were common with a prevalence of anemia higher than the global average in the majority of studies (The World Bank Group, 2019). The prevalence of micronutrient deficiencies in children in IBC is likely linked to their increased risk of sickness or morbidities (Black et al., 2013). Hearst et al. (2014) concluded that the growth-related indicators coincide with the high prevalence of low albumin, indicating generalized chronic undernutrition, and suggested macronutrient deficiencies could be due to inadequate diets, infections and/or inflammation, or impaired nutrient absorption or utilization secondary to the psychosocial stress of living in an institution.

Dietary diversity, intake and food security
Only eight (32%) studies included information on dietary intake and, of those, half found intake or diet diversity to be inadequate. Dietary diversity was reported to be low for children in IBC, especially in terms of fruits, vegetables and protein. Limited funding and reliance on donations for food were frequently mentioned issues, and resulted in diets high in starches and legumes (Mwaniki, Makokha & Muttunga, 2014). Dietary adequacy varied; in some IBC sites children received an adequate amount or more than recommended dietary allowances and in others they received below the recommendations. Interestingly, the one study which reported on food security found that children in a Kenyan orphanage had higher food security when compared to children in FBC (Braitstein et al., 2013). However, it is impossible to generalize from this one study to say anything more broadly about food security.

Limitations
We focused on nutritional status of children living in care but note that many other issues (e.g., development, cognition, puberty, catch-up growth, care practices, length of stay, age at admission, cause of institutionalization, illnesses, health of children who have been adopted or cultural practices) affect the demographics, health and well-being of children who are in institutions. It could be that all children coming into care are at risk due to the adverse events and trauma of being abandoned or orphaned (Baron, Baron & Spencer, 2001;Martins et al., 2013;The Children's Health Care Collaborative Study Group, 1994;The St. Petersburg-USA Orphanage Research Team, 2008). These wider factors were beyond the scope of this study (as well as infrequently reported in sufficient details in papers). Given biological links between poor nutrition and sub-optimal child development, evaluating these topics in more depth is critical in future work.
Although we found some research, there was limited recent information on this population of children. This may be because of practical or ethical considerations or it may reflect the desire to move away from institution-based care to family-based living situations for children (Kelley et al., 2016). This review also only analyzed data from research published in English from January 1990 to January 2019. The studies were of differing designs and types. The review did not find enough studies to be able to examine differences between IBC, FBC, CC and CLS (children living on the streets). Other weaknesses included the common use of non-standard reporting methods or lack of clarity around measurement methods, such as how studies assessed micronutrient status or clinical signs and symptoms or determined disability status. Many of the studies were examining other subjects and nutritional/anthropometric information was only supplementary. Furthermore, growth measurements may have been affected by measurement or other errors (e.g., incorrect birthdate estimates leading to incorrect z-score calculations for age-related indices). Additionally, children with some types of disabilities may be shorter or lighter not because of inadequacy of dietary intake but because of their specific underlying conditions (e.g., disabilities such as Down syndrome and many others are associated with non-standard growth and development). It is also possible that there is under-diagnosis or misdiagnosis of medical conditions, chronic diseases or disabilities in these settings, which can also impair the growth and development of children (Byass, Kahn & Ivarsson, 2011). Another consideration is the potential for healthy survivor bias and sampling bias: some of the most vulnerable children may have died prior to measurement; younger children and healthier children may more quickly leave institutions with the remaining older residents more likely to have deficiencies (van IJzendoorn et al., 2011;The St Petersburg-USA Orphanage Research Team, 2005).
Risk of bias was apparent in most of the studies. We had originally considered using a formal risk of bias tool to differentiate study quality but did not do so because it became apparent that all of the studies had a high risk of bias and could not be representative of all the institutions in the countries. Another concern was that many used convenience sampling. It is also plausible that the sites included in the research were better-off facilities, which welcomed researchers, who were looking to share positive results and good performance. These are unlikely to be representative of all sites; we speculate that the overall situation is likely worse at many facilities with higher prevalence of malnutrition indicators. There is also wide variation between different institutional care facilities (van IJzendoorn et al., 2011;Petrowski, Cappa & Gross, 2017;Whetten et al., 2014).

CONCLUSIONS
A key finding from this study was the limited amount of quality evidence-based data available on the nutritional status of children in institutions. Equally as important, our review found that where data was available, children living in institutionalized care were consistently at high risk of malnutrition, commonly experiencing undernutrition, overweight and/or micronutrient deficiencies. The implications for caregivers, clinicians, institutional administration and policy makers is that work is needed to ensure all children's basic rights to nutrition are met. Children living within care are at risk and require special attention. This is especially true for children with disabilities and low birth weight infants.
Although institutionalized care is not the ideal setting for children to grow up in, living within care continues to be a reality for many children. This study is in agreement with other papers and reports that support optimizing current institutional environments when alternative placements for orphaned or abandoned children are not available. These children have a right to good nutrition, both to maintain their health now and to allow them to grow into healthy adults. Interventions will need to be multifaceted to address all of the root causes of malnutrition faced by children living in care. The need for much more evidence as well as a commitment to monitoring and evaluation of nutritional status in all institutions, should be acknowledged and children supported through improved nutrition programming as part of broader policy and child rights initiatives.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
The authors received no funding for this work.