Nutritional perspectives on sickle cell disease in Africa: a systematic review

Background Sickle cell disease (SCD) is an inherited blood disorder that predominantly affects individuals in sub-Saharan Africa. However, research that elucidates links between SCD pathophysiology and nutritional status in African patients is lacking. This systematic review aimed to assess the landscape of studies in sub-Saharan Africa that focused on nutritional aspects of SCD, and highlights gaps in knowledge that could inform priority-setting for future research. Methods The study was conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Inclusion criteria comprised original, peer-reviewed research published between January 1995 and November 2020 involving individuals in Africa with any phenotypic variant of SCD and at least one nutritional status outcome. Nutritional status outcomes were defined as those that assessed dietary intakes, growth/anthropometry, or nutritional biomarkers. Databases used were Ovid Embase, Medline, Biosis and Web of Science. Results The search returned 526 articles, of which 76 were included in the final analyses. Most investigations (67%) were conducted in Nigeria. Studies were categorized into one of three main categories: descriptive studies of anthropometric characteristics (49%), descriptive studies of macro- or micronutrient status (41%), and interventional studies (11%). Findings consistently included growth impairment, especially among children and adolescents from sub-Saharan Africa. Studies assessing macro- and micronutrients generally had small sample sizes and were exploratory in nature. Only four randomized trials were identified, which measured the impact of lime juice, long-chain fatty acids supplementation, ready-to-use supplementary food (RUSF), and oral arginine on health outcomes. Conclusions The findings reveal a moderate number of descriptive studies, most with small sample sizes, that focused on various aspects of nutrition and SCD in African patients. There was a stark dearth of interventional studies that could be used to inform evidence-based changes in clinical practice. Findings from the investigations were generally consistent with data from other regional settings, describing a significant risk of growth faltering and malnutrition among individuals with SCD. There is an unmet need for clinical research to better understand the potential benefits of nutrition-related interventions for patients with SCD in sub-Saharan Africa to promote optimal growth and improve health outcomes.


Background
Sickle cell disease (SCD) is the most common inherited blood disease worldwide, with the vast majority of cases occurring in sub-Saharan Africa [1]. The condition derives from a point mutation of the β-globin gene found on the short arm of chromosome 11 through which the hydrophilic amino acid glutamic acid is substituted with the hydrophobic amino acid valine at the sixth position [2,3]. The result is a change in the structure and dynamics of hemoglobin such that certain conditions including deoxygenation and acidosis predispose to hemoglobin polymerization. When this occurs, erythrocytes assume a misshapen and rigid form that promotes pathological processes leading to intravascular inflammation and occlusion of small blood vessels [4]. Since these processes can take place anywhere in the body, the disease is highly complex and characterized by multiple potential life-threatening complications that include acute splenic sequestration, aplastic crises, acute chest syndrome, infection, heart failure, and stroke [5,6]. The disease's clinical hallmarks include acute painful crises and severe anemia [1]. In sub-Saharan Africa, it has been estimated that up to 90% of children born with SCD die before 5 years of age [7].
In high resource countries, mortality from SCD has decreased dramatically over the past five decades. The improvements in outcomes have been attributed in part to newborn screening and comprehensive care programs designed to prevent disease complications to the extent possible and to treat complications of disease when they occur [8]. Early detection of disease enables clinicians and families to institute measures to proactively prevent complications and facilitate timely treatment when needed. For example, the risk of fatal infection has been shown to be reduced through vaccination and administration of prophylactic antibiotics [9]. Furthermore, treatments with blood transfusion and hydroxyurea therapy has led to superior outcomes in the long term [10,11]. Unfortunately, the availability of vaccines, medicines, and other interventions is highly variable in sub-Saharan Africa. Increasing access to proven preventative and treatment modalities is therefore an urgent priority [12]. At the same time, there is a need to identify new ways of maximizing the well-being of individuals with SCD in Africa and it is in this context that nutritional interventions could possibly play an important role.
There is evidence that the pathophysiology of SCD has substantial nutritional implications including higher energy and nutrients requirements, nutrient deficiencies, and growth abnormalities [13][14][15]. It is theorized that a main driver of disease complications is higher rates of metabolic expenditure in individuals with SCD resulting from increased hematopoiesis, increased cardiac output, chronic inflammation, and related processes [16,17].
Since nutrition interventions could be a mechanism for addressing increased energy expenditure, attention to nutritional care is increasingly seen to be an important aspect of supportive management for patients with SCD [18,19], especially in resource poor settings. However, evidence-based nutritional guidelines for patients with SCD in Africa are lacking and the extent of nutritionfocused research involving individuals in Africa with SCD is unclear. We undertook this systematic review to evaluate the existing literature focused on nutritional aspects of SCD in sub-Saharan Africa. Specifically, we sought to assess the number and nature of relevant studies, review their findings, and identify gaps in knowledge that could inform priority-setting for future research.

Eligibility criteria
We sought to include all studies involving original research that focused on the nutritional status of individuals with SCD in an African population. Studies involving nutritional status were defined as those that investigated topics of dietary intake, measurements of growth or anthropometry, or nutrition-related biomarkers. Studies that did not differentiate the cause of the anemia were excluded, as were studies that only included nutrition interventions as part of comprehensive care programmes (i.e., studies that did not clearly report nutrition-related outcomes). Studies involving both children and adults were included. The focus of this analysis was on studies involving individuals with various forms of SCD including HbSS, HbSC, and rarer genetic variants of disease; studies were excluded that only involved individuals with sickle cell trait. Case reports and review articles were also excluded.

Informational sources and search strategy
The databases employed for this search were Ovid Embase, Medline, Biosis, and Web of Science. The date range was January 1st, 1995, through November 30th, 2020, such that the reference list covered a period of approximately 25 years. We performed a Boolean search using specific Boolean operators and the following search terms: "Sickle cell disease" or "sickle cell anemia" or "hemoglobinopathy" AND Africa or specific African countries (all African countries were individually listed) AND various nutrition-related terms (i.e., nutrition, growth, macronutrient, micronutrient, vitamin, mineral, anthropometric, height, length, weight, head circumference, mid-upper arm circumference, MUAC, dietary intake, recommended dietary allowance, RDA, nutritional status) along with associated terms (both indexed and non-indexed) for nutrition, diet and growth, and specific vitamins and minerals. We also allowed for inclusion of articles that were identified through review of the bibliographies of papers that underwent full-text review. Investigations written in any language were included provided sufficient translation into English could be assured.

Data management and selection process
Titles and abstracts were each screened by two independent reviewers. Any title or abstract that appeared to meet inclusion criteria or for which there was uncertainty prompted a full text review. Reviewing of full text articles were assigned to individual investigators. If eligibility of a full text article was unclear, it was resolved by discussing it with at least two other reviewers on the research team who were not earlier assigned the full text article using the inclusion criteria. To maximize consistency among reviewers, each reviewer initially reviewed 10 articles and the review team together discussed the initial dataset that had been extracted to ensure accuracy and completeness. The review process then proceeded according to the process described.

Data synthesis
Study data were extracted into standardized forms using Microsoft Excel (Microsoft, Redmond, USA) where they were organized for analyses. Depending on the nature of the articles that met inclusion criteria, we extracted information relating to geography, subject age group, sample size, comparison groups, biomarkers, and other relevant variables. Since the main purpose of this investigation was to broadly understand the types of nutritionrelated studies that have been conducted involving individuals with SCD in Africa, we chose not to systematically judge the quality of evidence or risks of bias within individual studies. Rather, we discussed specific merits and limitations of individual studies where appropriate in the context of major themes that would emerge in the analyses. We planned for a quantitative categorization of the types of articles (e.g., descriptive versus interventional studies) and a narrative synthesis of data in table and text format to summarize and assess the results.

PRISMA
The study was conducted and reported according to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines [20,21].

Search results
In total, 526 unique titles and abstracts were identified through the literature search. Of those, 347 did not meet inclusion criteria and 179 full-text articles were assessed. Seventy-six studies were deemed eligible and included in the final analyses ( Fig. 1).

Results overview
Most investigations (67%) were conducted in a single country (Nigeria). Each study was placed into one of three main categories according to the primary nature of the investigation: (a) descriptive study of anthropometric characteristics (37 articles; 49%); (b) descriptive study of macro-or micronutrient status (31 articles; 41%); and (c) interventional studies (8 articles; 11%). The studies are summarized in Tables 1, 2, 3 and described in greater detail below.

Descriptive studies of anthropometric characteristics
Nearly one-half of all studies identified focused on anthropometric characteristics. The studies typically assessed height, weight, and body mass index (BMI).     • Significantly increased urea in non-malaria infected individuals with SCD compared with non-malaria infected controls • Significantly reduced levels of total antioxidants in non-malaria infected individuals with SCD compared with non-malaria infected controls • No differences in magnesium, copper, chromium, cadmium, and selenium in nonmalaria infected individuals with and without SCD • No differences in levels of serum albumin   • 54% of children treated with arginine were discharged compared to 24% in placebo arm by day 5 • Arginine treatment appeared to ameliorate some measures of pain • No significant differences in adverse events but arginine arm trended more towards vomiting compared to placebo Arginine had previously been shown to have benefits in individuals with SCD in studies performed in the Unite

States
Other measurements included head circumference, arm span, and various body composition parameters. The majority (25/37; 68%) of studies were conducted in Nigeria. Other studies involved populations in Democratic Republic of Congo (DRC) [49,50,57], Ghana [33,53], Tanzania [32], Egypt [23], Cameroon [54], and Algeria [37]. In addition, two multi-country studies involved patients in Cameroon, Ivory Coast, Gabon, Mali, and Senegal [43,55]. The majority of reports focused on children and adolescents; only one study exclusively involved adults [24]. Most studies evaluated approximately 50-200 patients and a similar number of age-and gender-matched healthy controls; three large studies enrolled over 1000 SCD patients each [32,43,55]. Several studies used WHO growth standards for comparison rather than a non-sickle cell disease control group. Male and female subjects were generally equally represented in the study populations.
The three largest studies found significant growth defects compared to healthy controls. A multi-country study of more than 3500 SCD patients (aged 10-24 years) with nearly 1000 controls in Cameroon, Ivory Coast, Gabon, Mali, and Senegal was designed to evaluate determinants of vascular complications [43]. Anthropometric analyses showed that SCD patients were significantly shorter and had lower BMI than controls; weight was not reported. A caveat of the study was that the control group was significantly older (median age 24 vs 16 for patients) and more likely to be female (60% vs 54% for patients) than the SCD group. A follow-up study involving the same subject population revealed significantly higher rates of growth failure, defined as a height and/or weight and/or BMI below the 5th percentile using WHO growth reference [55]. Another large study followed a cohort of approximately 1000 SCD patients aged 6 months to 48 years over 5 years and found SCD to be significantly associated with stunting, underweight, and wasting, with the most pronounced effects associated with adolescent age and male gender [32]. Adult men were seven times more likely than adult women to be underweight and were significantly more likely to be stunted and wasted. Females demonstrated improved catch-up growth compared with males following growth deficits that were identified during adolescence.
The studies involving smaller sample sizes showed greater variation in the results, but some trends emerged. Several reports confirmed the finding that males were more likely to show growth defects than females [22,27,29,45]. Multiple studies also noted that growth deficits became more pronounced with age. For example, a study that involved young patients aged 6-35 months showed no association with wasting, stunting, or underweight status [40]. A study of 233 children aged 2-17 years with SCD in Lagos, Nigeria found that the factor most significantly associated with both wasting and stunting was older age [45]. Additional studies similarly detected one or more growth deficits in adolescents but not in younger children [25,28,34,39,41,46]. In several Nigerian studies, SCD patients were found to be underweight or to have low BMI, but showed no difference in height compared to controls [22,25,35,42,44,52]; while other Nigerian studies showed differences in both weight and height [24,26,27,29,44,51]. These variable findings may have resulted from the fact that the studies were not powered to detect significant differences in height specifically. Reports from DRC, Egypt, and Ghana found that children with SCD had a higher prevalence of stunting compared to controls, but did not always show differences in wasting or BMI [23,33,50]. SCD was also associated with delayed puberty [23,31]. Three studies from Nigeria reported the presence of overweight and obesity among patients with SCD although in lower proportions (an average of less than 3% of the sample population) [36,42,45].

Descriptive studies of macro-or micronutrient status
The second most common group of studies identified involved assessment of biomarkers from serum samples for macronutrients or micronutrients. Most studies were conducted in Nigeria (21/31, 68%); other studies took place in DRC [78], Tanzania [32,77,79], Egypt [62], Kenya [40], Ghana [82], Uganda [69], and Malawi [64]. Nutritional parameters measured included proteins/ amino acids, fatty acids, vitamins, and minerals. Most studies included less than 100 individuals with SCD. Both children and adults were studied, with male and female subjects generally equally represented.
Serum protein levels were investigated in one small study (13 children with SCD and 17 healthy controls) in Nigeria in which no significant differences were reported in the concentrations of total protein or albumin between SCD patients and controls [22]. However, serum prealbumin levels were significantly lower for the population of patients with SCD, which was hypothesized to result from poor nutrition or existing disease-related inflammation. The serum concentrations of all amino acids except alanine, glutamic acid, and proline were significantly reduced in SCD patients. A small study involving 23 participants in Tanzania measured the steady state nutrition status of SCD patients who later died (n = 11) compared with those who were alive at the end of the study period. Those who suffered mortality had a significantly lower BMI, plasma taurine levels and arginine bioavailability before succumbing [59].
Proportions of fatty acid and the state of metabolism were evaluated in four related studies of young SCD patients in Nigeria [26,29,60,61]. These reports found perturbed pathways of fatty acid elongation and desaturation in children with SCD. Specifically, arachidonic acid, eicosapentanoic acid (EPA), and decosahexanoic acid (DHA) were significantly reduced, whereas saturated (palmitic acid) and monounsaturated (oleic acid) were significantly elevated in patients compared to controls. Another study in a population of 26 SCD patients aged 11-43 in Enugu, Nigeria, confirmed the finding that EPA and DHA fatty acids are reduced in SCD patients [63]. The authors of these studies hypothesized that reduced polyunsaturated fatty acids in the phospholipids of the cell membrane of SCD patients could lead to their being more rigid, thereby contributing to disease symptoms. A study of 30 children with SCD in Egypt found that patients also had significantly lower cholesterol, triglycerides, and LDL (but not HDL) in blood plasma compared to healthy controls [62].
Acknowledging that interpretation of plasma concentrations of vitamins and minerals can be problematic in patients with ongoing inflammation, analysis of serum vitamin levels in SCD patients generally indicated lower concentrations of vitamin A [63,65] vitamin C [65,66] and vitamin E [62][63][64][65]. One study of 14 SCD patients in Kenya found no association of HbSS phenotype and low concentrations vitamin A [40]. Three reports of a related study population in Ilese, Nigeria, investigated vitamin D status of young SCD patients [48,67,68]. When compared to healthy controls, mean 25-hydroxyvitamin D levels were significantly lower in SCD patients and suboptimal vitamin D levels were seen in greater than 10% of patients. However, no SCD patients with severe vitamin D deficiency (defined as < 20 ng/ml) were observed. A limitation of the latter two studies was the lack of a healthy comparator group; each used vitamin D deficiency cut-off values for a healthy population in other published studies as reference.

Interventional studies
A very small number of clinical studies involving nutritional interventions in SCD patients in African countries were identified. There were four randomized trials [88][89][90][91]. The first was a study of 125 SCD patients in Nigeria (involving children aged 1 year and above) that tested the effect of lime juice on SCD parameters. All patients in the treatment group (n = 65) as well as controls (n = 60) were given folic acid, vitamin B complex, and proguanil, with the treatment group also given twice-daily oral lime juice with weight-based dosing ranging from 5 to 15 mL. Each child was assessed monthly for 6 months. The group receiving lime juice was reported to have significantly fewer pain episodes, febrile illnesses, and hospital admission rates. There was no change in transfusion rate, organomegaly, or jaundice. The positive effect was postulated to result from vitamin C, amino acids (in particular, phenylalanine), and flavonoids contained in the juice, but no direct evidence for this was provided [88].
A second study of 128 SCD patients aged 2-24 years in Sudan investigated the impact of 1 year of treatment with oral omega-3 capsules containing EPA and DHA fatty acids (using age-and weight-dependent dosing) compared to placebo. The hypothesis was that omega-3 fatty acids could reduce red blood cell aggregation, adherence, and inflammation that occur during sickle cell disease-mediated vaso-occlusive crises. The treatment group had significantly fewer clinical vaso-occlusive events, as well as reduced rates of severe anemia and need for blood transfusions. The study was not powered to detect changes in other outcomes such as stroke, sequestration crisis, or vascular necrosis [89].
A third trial utilized a double-blind, random order design. Ready-to-use-supplementary food was studied in 119 children with SCD in Tanzania [90]. Two different formulations of the supplements were compared: a commercially available (Nutriset, France) ready-to-use-supplementary food (RUSF) fortified with vitamins and minerals according to recommended daily allowances and an "enhanced" version of the same RUSF (providing 500 kcal/day) that was additionally fortified with arginine and citrulline. Arginine is the substrate for endothelial nitric oxide synthase, a natural vasodilator, and has been implicated in pathophysiology of SCD complications. In the cross-over study design, children received each treatment for 4 months, with 4-month washout periods following the intervention. Ready-to-use-supplementary food led to small weight gains, an increased arginine bioavailability ratio, and improved measures of endothelial function compared to baseline; addition of arginine and citrulline to the supplement did not provide additional benefits [90].
The final randomized study identified involved the regular administration of oral arginine therapy to 35 hospitalized patients with SCD in Nigeria and compared the effects with 33 control subjects. Plasma arginine levels increased by 125% in the arginine arm compared with 29% in the control arm [91]. Arginine treatment was associated with quicker discharge and reduced pain events. The rate of adverse events was non-significant between the two treatment arms, however there was a trend towards increased vomiting in the patients treated with arginine. A previous study outside of Africa also found positive clinical effects associated with the use of arginine [92].
One of the non-randomized interventional studies identified investigated the use of vitamin D supplementation. A small treatment arm was nested in a Nigerian study comparing blood levels of vitamin D and proinflammatory cytokines [48]. The hypothesis was that low vitamin D levels might lead to a pro-inflammatory environment that exacerbates SCD symptoms. Twelve children with SCD who were determined to have low vitamin D levels were given 3 months of oral vitamin D supplementation (2000 U). At the end of treatment, mean serum 25-hydroxyvitamin D levels were significantly increased compared to baseline, levels of several proinflammatory cytokines were significantly decreased, and the levels of anti-inflammatory cytokine IL-11 were significantly increased.

Discussion
To our knowledge this is the first review of nutritionrelated studies involving individuals living with SCD in sub-Saharan Africa. While a moderate number of studies were identified, most were descriptive in nature and small in terms of numbers of subjects. Approximately two-thirds of studies took place in a single country (Nigeria). In addition, there were very few interventional trials designed to measure the impact of an isolated nutritional intervention and only four randomized studies. The findings of this review suggest an outstanding need for nutrition-focused research relating to the care of individuals with SCD in Africa, with a particular emphasis on research with practical implications for clinical management in order to improve patient outcomes.
The findings of studies identified through this review are generally consistent with nutrition-related investigations involving SCD patients in other parts of the world. More than 50 years ago, poor growth was first reported in patients with SCD, and that observation has since been repeated in multiple studies involving SCD populations in Jamaica, Brazil, India, and North America [13,15,[93][94][95][96][97][98][99]. Many of these studies specifically note that the growth faltering occurred in patients that were receiving recommended daily protein and calorie intakes. The pathophysiology of growth problems in SCD patients has come into sharper focus in recent decades. A leading view is that the increased rate of red cell turnover, a primary feature of SCD patients, underlies a hypermetabolic state. The biochemical and physiological factors that contribute to hypermetabolism include increased protein turnover, increased myocardial activity, and production of proinflammatory cytokines [100][101][102][103][104]. The supposition is that the energy and nutrient requirements normally recommended are not adequate in patients with SCD given their increased energy expenditures and other unusual metabolic demands, which compete directly with energy needs required to sustain adequate growth.
Evidence derived from robust interventional studies is important to support recommendations for specific nutrition-related practices for patients with sickle cell disease. Only four randomized trials were identified. The studies were small, each involving less than 150 individuals with SCD. Positive clinical benefits were found with the use of lime juice, long-chained fatty acid supplementation, RUSF, and oral arginine; ideally these findings would be confirmed in larger follow-up investigations. It is worth noting the paucity of robust interventional trials designed to test the effect of macronutrient supplementation in individuals with SCD despite the evidence, as described above, that nutrition deficits in this population are likely to be caused at least in part from the increased energy demands that result from altered metabolism.
Guidelines for clinical management of patients with SCD published by internationally recognized organizations do not provide special guidance for nutritional care [105,106]. Given that the risk of poor growth in SCD patients is increasingly reported, and the fact that there is plausible pathophysiologic drivers of nutritional disturbances in SCD patients, there appears to be a substantial gap in research in this area to inform much needed evidence-based recommendations.
Limitations of this systematic review include the fact that nearly half of studies identified were largely anthropometry-based descriptive studies. Few studies involving nutrition interventions in sub-Saharan Africa were identified, only several had robust methodologies, and none have been validated in repeated studies. In addition, the studies involving analyses of vitamin and mineral levels in SCD patients in sub-Saharan Africa overall involved small numbers of patients and generally were unable to link findings with meaningful clinical correlations in ways that might influence nutritional care practices. Another limitation is that most investigations identified took place in the single country of Nigeria (at the same time, acknowledging that Nigeria is home to the largest population of SCD patients globally).

Conclusion
Despite the reality that most SCD patients globally live in sub-Saharan Africa, and the fact that nutritional disturbances in SCD patients are increasingly well described, there has been limited research focused on ways that nutritional care might help to improve clinical outcomes in this patient population. A systematic review of the literature revealed studies that consistently reported stunted growth and malnutrition in African SCD patients during childhood and adolescence, but failed to identify robust, validated studies that could be used to inform clinical management. Our study suggests an outstanding need to determine if and how supportive nutritional care can reduce disease severity and improve health outcomes for individuals with SCD in sub-Saharan Africa. As such, priority research in this area in the future may include systematic assessment of the drivers of nutritional status in SCD patients, studies that directly advance the understanding of macro-and micronutrient deficiencies associated with clinically significant physiologic effects, and investigations that evaluate the impact of nutritional interventions to inform evidence-based nutrition guidance.