Nutritional Aspects of Juvenile Idiopathic Arthritis: An A to Z for Dietitians

Juvenile idiopathic arthritis (JIA) represents a chronic, autoimmune, rheumatic musculoskeletal disease with a diagnosis before 16 years of age. Chronic arthritis is a common manifestation in all JIA subtypes. The nature of JIA, in combination to its therapy often results in the development of nutrition-, gastrointestinal (GI)- or metabolic-related issues. The most-common therapy-related nutritional issues involve methotrexate (MTX) and glucocorticosteroids (GCC) adverse events. MTX is a folic acid antagonist, thus supplementation with folic acid in required for improving GI side effects and correcting low serum levels. On the other hand, long-term GCC administration is often associated with hyperglycemia, insulin resistance and growth delay. This relationship is further aggravated when more joints are affected and greater doses of GCC are being administered. Apart from stature, body mass index z-scores are also suboptimal in JIA. Other signs of malnutrition include decreased phase angle and muscle mass, especially among patients with polyarthritis JIA. Evidence also points to the existence of an inverse relationship between disease activity and overweight/obesity. Specific dietary patterns, including the anti-inflammatory diet, might confer improvements in selected JIA outcomes, but the level of available research is yet insufficient to draw safe conclusions. The majority of patients exhibit suboptimal vitamin D status; hence, supplementation is recommended. Collectively, the evidence indicates that, due to the age of onset and the complexity of the disease, along with its pharmacotherapy, children with JIA are prone to the development of several nutritional problems, warranting expert monitoring. Vitamin deficiencies, oral and GI-problems limiting dietary intake, faltering growth, overweight and obesity, physical inactivity, or impaired bone health are among the many nutritional issues in JIA requiring dietitian support.


Introduction
Juvenile idiopathic arthritis (JIA) is an umbrella term, encompassing heterogenous arthritides with an onset before 16 years of age [1]. It consists of the most common chronic, autoimmune, rheumatic and musculoskeletal disease (RMD) diagnosis of unknown etiology encountered during childhood [2]. Recent epidemiological data indicate that approximately two million children are affected by JIA globally [3], with the condition being more prevalent among girls than boys [4].
Based on the most recent classification seven distinct JIA subtypes exist [1], namely (a) systemic-onset JIA (sJIA), characterized by systemic features including a spiking fever   [56] Italy Prospective cohort 4.9 +/-2.8 years N = 58 patients with JRA, who had never received CCS

15/43
In systemic and polyarticular JRA a negative association was noted between disease duration and the ∆ in HAZ, and between the latter and cumulative periods of active disease. The longer the disease duration, the higher the degree of functional joint involvement. In systemic and polyarticular JRA age of puberty is risk factor for stature impairment. Saha [46] Finland Prospective cohort 4 years N = 64 prepubertal children with mild-to-moderate JIA 21/43 Preceding the diagnosis, children were slightly taller than their healthy peers, growing at a faster rate. During the 1 st year of treatment growth velocity decreased, but at further follow-up, it returned to the pretreatment levels. Growth was aggravated among patients with polyarticular JIA.
The cumulative total GCC dose influenced growth. Simon [36,57] France Prospective cohort NR N = 24 patients with JIA on steroid therapy NR A height loss exceeding 2SD during the first years of the disease was noted, correlated with the duration of PRED therapy. After remission and discontinuation of PRED, 70% achieved catch-up growth, but 30% showed a persistent loss of stature. Mean final height correlated with mean stature post-CCS treatment and was different between those attaining catch-up growth, or not. Simon [58] France Prospective cohort 3 years N = 13 patients with JIA on rhGH (0.46 mg/kg/week)

NR
Median growth velocity increased from 2.1 to 6.0 cm/year in the 1 st year of rhGH therapy and remained above baseline in the 2 nd year. HAZ did not change, but growth response varied across patients. NR rhGH therapy increased IGF-1 and IGFBP3 levels above normal. All patients showed an increase in growth velocity.
According to Saha et al. [46], a year preceding the JIA diagnosis, children are generally taller than their healthy peers, growing at a faster rate. Post-diagnosis, a decline in growth is noted during the first year, with children with polyarticular disease being the most affected [46]. Similar results were also reported by the Childhood Arthritis Prospective Study (CAPS) cohort [55], where children with JIA were followed for three years after their initial presentation to the rheumatology clinic. Although height at three years was within the population norm for most patients, as a cohort, youngsters with JIA demonstrated a growth reduction in height over the first three years of disease diagnosis [55]. Interestingly, patients with the lowest HAZ at presentation were the most likely to show improvement at three years [55].
Miyamae [35] suggested that in sJIA HAZ is inversely correlated with the duration of the disease. Apart from disease duration however, the age of puberty, and the higher degree of functional joint involvement have been shown to act synergistically in augmenting growth impairment in CCS-naive patients with systemic and polyarticular JIA [56].
Nonetheless, research is unanimous on the deleterious effects of CCS use on the stature of children with JIA [35][36][37]. In an early study [57], a significant loss of height was observed (exceeding two standard deviations [SD) during the first years post-JIA diagnosis. This height loss correlated well with the duration of prednisone (PRED) treatment. After remission was induced and PRED therapy was discontinued, the majority (70%) of patients achieved catch-up growth but some (30%) retained a persistent loss of stature. The mean final height of participants was correlated with height at the end of steroid treatment and was different between the children who achieved catch-up growth and those who did not [57]. In parallel, at prepubertal stages [37] HAZ has been shown to differ between patients and controls. Polyarticular JIA and greater cumulative GCC doses have been shown to synergistically contribute to short stature [37]. It appears that, long post-GCC cessation, children with polyarticular/sJIA still remain susceptible to low height and delayed puberty [37]. Apart from disrupting growth through the GH/IGF1 axis [34], prolonged GCC treatment also alters the secretion and pulsatility of GH, through the somatostatin inhibitor tone and by reducing the expression of GH receptors through the Janus kinase (JAK)-2/signal transducers and activators of the transcription (STAT)-5 pathway, tamping down IGF-1 mRNA levels within the liver cells [34,62]. In a comparative study of patients on different CCS treatment durations, Wang et al. [61] showed that continuous systemic CCS use for less than a year does not affect attained adult height. On the other hand, prolonged CCS treatment exceeding a year in duration may lead to irreversible growth impairment [61]. Among boys in particular, CCS dose and age of CCS initiation are associated with delayed puberty [48].
With the timely diagnosis of JIA available today, and the use of biologics, the number of children with short stature is expected to fall. In sJIA in particular, biologics targeting the interleukin (IL)-1β or IL-6 pathway are increasingly used [52]. Biologics appear to restore growth retardation by inhibiting inflammation and limiting CCS daily doses [63]. Uettwiller [60] postulated that growth delay is observed in children with JIA long before biologics are initiated. In fact, although biologic treatment normalizes growth velocity, growth delay does not appear to be fully corrected [60]. Furthermore, when more than one biological agents are prescribed, this is associated with poorer growth [59,60]. Subsequently, biologic therapy is often insufficient in restoring normal growth velocity, indicating that rhGH might be required [59].
In a recent proof-of-concept study, Belgian scientists [52] suggested that stature and adipose tissue stores in sJIA can be restored with a hormonal combination therapy including (i) a GnRH analog (triptorelin) to postpone the onset of puberty until a minimum height is attained (or until prepubertal growth spurt is exhausted), and (ii) rhGH at doses of 50 µg/kg/day to promote height gain, once inflammation is controlled and high GCC doses are not required. Of note, rhGH is the cornerstone of growth delay treatment in patients with JIA who have received GCC [64]. Bechtold et al. [50] revealed an increase by 1.5 in the total pubertal growth of rhGH-treated patients in comparison to patients not receiving rhGH, leading to an improved final height. Thus, for the maximization of final height, rhGH treatment must be initiated early on, in order to reduce the height deficit observed at the onset of puberty.
David and associates [51] followed patients with JIA on rhGH treatment to adulthood and showed that median adult height still remained below target height. The authors concluded that long-term rhGH treatment increases growth spurts in JIA; however, it cannot fully restore the genetic growth potential of these children [51]. When the growth velocity of children on rhGH therapy was recorded [29] rhGH was efficient in increasing IGF-1 and IGFBP3 plasma concentrations to values above normal thresholds, improving growth velocity. However, one year after the cessation of rhGH therapy, growth velocity fell to pretreatment values and HAZ was lower than prior to the initiation of rhGH treatment. Nonetheless, rhGH appears to reduce the deficit in stature that occurs during the active phase of JIA, producing an adult height that is closer to the genetically determined height [65]. Overall, on rhGH treatment, catch-up growth markedly varies depending on the severity of inflammation and the CCS doses administered during treatment [64]. According to a systematic review [25] most studies evaluating the effect of rhGH on children with JIA reported positive effects on growth; however, great variability in the treatment response is observed. Suffice it to say, the early initiation of rhGH therapy has been shown to be more effective in maintaining normal growth [64], and the use of combined therapies in parallel to rhGH appears promising in further accelerating growth velocity. It is still unclear, however, whether greater rhGH doses at remission, with patients on biologic therapy post-CCS, can confer a prepubertal growth acceleration, comparable to that observed among GH-deficient children [47].
TNF-α antagonist therapy can also ameliorate growth velocity in JIA [64]. However, the use of etanercept (ETN), a recombinant human soluble TNF-α receptor fusion protein, has been shown to have limited efficacy [64]. In the British Society for Paediatric and Adolescent Rheumatology Etanercept Cohort Study [53], 191 children with JIA treated with ETN were followed for two years. All participants exhibited an improvement in HAZ from baseline; however, a lower baseline HAZ and the use of no oral CCS at baseline were associated with improved HAZ at two years [53]. When the effects of tocilizumab (TCZ), an anti-interleulin-6 (IL-6) receptor monoclonal antibody were examined [35] TCZ was efficient in improving HAZ velocity one year post-baseline. Reduction in CCS use was associated with further improvement in the observed HAZ [35].
Collectively, the literature indicates the important problem of correcting growth delay in children with JIA. Furthermore, the heterogeneity in disease manifestations, therapeutic approaches and medication responses make it difficult to suggest safe, horizontal regimes to exploit the genetic growth potential reserve in JIA.

Undernutrition
Apart from faltering growth and short stature, underweight is also prevalent in children with JIA. Knops [42] suggested that children with JIA exhibited lowed fat-free mass (FFM) compared to controls, except from the oligo-and polyarticular patients, who demonstrated FFM values within the normal range. In general, a lower body mass index (BMI) z-score (BMIz) is observed compared to healthy children [37], and this has been associated with greater disease activity [66]. Research indicates that between 8.3-30% of children with JIA suffer from malnutrition [67][68][69], with dietary intake factors, disease subtype, more affected joints with restricted range of motion (ROM) and younger patient age consisting of effectors of lower BMIz [67].
Most studies, however, suggest that not all JIA subtypes are associated with the same malnutrition risk and/or severity. Young patients with polyarticular disease demonstrate more signs of malnutrition compared to the rest of the subtypes [54,70]. Więch and associates [70] assessed malnutrition through phase angle score, a proxy for nutritional status assessment [71], and a reliable prognostic measure of disease [72]. Among children with JIA, phase angle, percentage of body cell mass and muscle mass were lower compared to that of healthy children [70]. When a subgroup analysis was performed, youngsters with polyarthritis exhibited lower phase angle and muscle mass compared to healthy children; however, no significant differences were observed between children with oligoarthritis and controls, indicating that the pathophysiology of polyarthritis is associated with a greater malnutrition risk. In a similar study, lower total body water was observed in more nutritionally deprived patients with JIA, indicating lower lean mass [54].
In a UK-based case-control study [54], 18.1% of children with JIA demonstrated body weight below the third percentile. Mid-upper arm circumference (MUAC)-an proxy for estimating body fat stores-was below the fifth percentile in 36.4% of the sample. Again, children with polyarticular disease showed significantly more signs of malnutrition than patients with pauciarticular disease, including lower stature, body weight, MUAC, body fat (as a % of body weight) and total body water [54].
Undernutrition has a variety of short-and long-term developmental, behavioral and physiologic effects on children, including increased susceptibility to central fat accumulation, reduced fat oxidation and resting energy expenditure (REE) levels, IR, hypertension and dyslipidemia in adulthood, and a reduced capacity for manual work [73,74]. Thus, although specific screening tools for malnutrition in JIA do not exist to date, frequent weighing of children and the use of appetite enhancers to improve nutritional intake are warranted. Furthermore, the available evidence indicates the lack of JIA-specific dietary intervention trials targeting undernutrition, using enhanced protein formulas, high-energy diets, dietary supplements or combination medical nutrition therapy (MNT).

Overweight and Obesity
Apart from undernutrition though, the other end of malnutrition spectrum, namely, overweight and obesity, is also apparent in children with JIA, in particular those with low disease activity [75][76][77] (Table 2). Young patients with JIA are more frequently overweight/obese than healthy controls, exhibiting greater serum leptin concentrations even after adjustment for fat mass [77]. An increased risk for the development of metabolic syndrome (MetS) has also been noted among youngsters with a JIA diagnosis [78]. The pathophysiological link between obesity, adipokines and cytokines released from the adipose tissue and JIA is not yet clearly understood [79].  Obesity was found in 18% of the patients, whereas 12% were overweight. There was no association between obesity and JADAS-27, physician's assessment of disease activity, parent's assessment of child's well-being, ESR, active joint count, or CRP. Rego [84] Pediatric Rheumatology Department, Hospital Regional Universitario de Málaga, Spain  Obesity and body weight accumulation is greatly dependent on the environment and the genetics. The genetic link of obesity is apparent in a recent study, where 80% of the overweight children and adolescents with RMDs had parents who were also overweight [87]. In Brazil, approximately 1/5 (21%) of the youngsters with JIA were classified as overweight/obese [68]. When on remission children with JIA exhibit higher measures of central and peripheral adiposity, a greater prevalence of being overweight, including obesity (30%), and larger biceps skinfolds compared to their healthy peers [75]. Disease activity or JIA subtype does not appear to affect body composition, energy intake (EI), or other nutritional biomarkers. Despite the fact that disease activity does not appear to be related to adiposity [66,81,83], it still contributes to the degree of functional impairment and the response to treatment [79]. Similar findings were also reported in Brazil [76], with age, disease subtype, number of affected joints, months since diagnosis and use of CCS or MTX not affecting adiposity. Nonetheless, within the sample, girls with JIA exhibited greater BMIz, body fat (% body weight), trunk fat and fat mass index (FMI) compared to healthy controls [76]. Furthermore, patients with JIA demonstrated lower muscle strength and total bone mineral density (BMD) compared to the control group, although no differences were noted regarding cardiorespiratory fitness (CRF) and body composition.
Only one study [84] reported a positive association between adiposity and disease activity. According to the results, overweight/obesity in JIA correlated with the duration of biological therapy and PA levels, while the FMI of patients was associated with age, the 27-joint Juvenile Arthritis Disease Activity score (JADAS-27) and PA levels [84].
It has been suggested that obesity may also negatively influence the course of the disease, as well as the response to treatment [81]. Collectively, these studies suggest that the changes observed in the body composition of children with JIA increase the risk for developing hypertension, DM and CVD [76]. According to Held [79], the development of early atherosclerosis and CVD in children with obesity consists of an important complication. Furthermore, given that children with JIA often demonstrate IR, higher triglycerides (TG) concentrations, greater systolic blood pressure and early atherosclerosis, as evidenced by ultrasonographic evidence [88], a comprehensive treatment approach targeting overweight and obesity is important for improving disease prognosis and reducing future CVD-related complications.

Low Physical Activity (PA) Levels
Due to the nature of JIA, children with a diagnosis are, in general, less physically active than their healthy peers [89]. With the environment being an important contributor to PA levels, 56% of the less active children with JIA seem to have parents with similar PA levels [87]. In Canada [89], children with JIA reported participating in less daily moderate-to-vigorous-physical activity (MVPA) than healthy controls, with peak oxygen consumption (VO 2 max ) gradually decreasing with ascending age. Furthermore, patients with JIA demonstrate lower muscle strength as an epiphenomenon of the reduced PA levels [90]. According to Rochette [91], lipid oxidation rates are lower in JIA and respiratory exchange ratios beyond 50% of VO 2 max are higher compared to healthy peers, both indicative of metabolic disturbance during exercise, even at times when the disease is inactive.
Furthermore, the observed lower levels of PA are associated with increased functional disability, with patients avoiding positions commonly associated with joint inflammation, intra-articular vascularity and pain [92]. As a result, they tend to apply greater initial hip flexion and lower knee and hip extensions during terminal stance [92].
Low levels of PA can contribute to disease and disability by increasing body weight accumulation and the number of comorbidities associated with overweight (DM, hypertension, etc.), catapulting CVD risk [93,94]. In a study using data from the German National Registry [86], low (or lack of) participation in sports was associated with greater chances for overweight and obesity among children with JIA. Furthermore, low PA is associated with sarcopenia, dynamopenia, low CRF and osteoporosis, as well as with a variety of mental health issues, including anxiety and depression [93,94]. As per Bohr [95], the inactive lifestyle observed in young patients with JIA appears to be a pivotal contributor to the development of subclinical atherosclerosis; thus, the promotion of an active lifestyle during childhood and adolescence is an eminent necessity.

Gastrointestinal Manifestations
Microbial dysbiosis and gut permeability have been shown to contribute to the inflammation and the immunological imbalance in a synergistic manner [96]. With JIA manifestations often being extended at the GI system, common relevant symptoms include eosinophilic infiltrates [97], gut dysbiosis and small intestinal bacterial overgrowth (SIBO) [96], GI inflammation (mostly affecting the colon (80%)) [97] and autoantibodies indicative of inflamed GI mucosa [98]. Patients with JIA frequently report suffering from chronic abdominal pain [97,99], constipation and diarrhea [100], all of which have been associated with MTX and SSZ use [101]. In parallel, sucrose excretion is increased compared to that of healthy controls, suggesting the existence of gastric and intestinal mucosal lesions (leaky gut) [99]. The observed damage of the intestinal lining is further aggravated as the disease progresses [102]. Among the JIA subtypes ERA is associated with more severe and diverse GI symptoms compared to the rest [98].
Fecal calprotectin (fCal) levels, a marker used in inflammatory bowel diseases (IBD) screening, has been reported to exceed the norm levels in approximately 1/3 of children with JIA [100,103], although extensive non-steroidal anti-inflammatory drugs (NSAIDs) has also been suggested to alter gut inflammation and increase fCal levels [104], although not all studies appear to agree on this issue [105]. Among distinct disease subtypes, ERA is the type associated with greater fCal concentrations [105]. Rare cases of patients with sJIA who were subsequently diagnosed with IBD have also been reported [106], mainly among those with a relevant family history.
A variety of in vivo and in vitro studies indicate that ONS with probiotics can exert immunomodulatory effects through multiple pathways, by (i) regulating intestinal immune function and tampering down inflammation, (ii) inhibiting the entrance of proinflammatory gut cytokines to the joints, (iii) preventing the progression of intestinal permeability (leaky gut) and the translocation of gut bacteria, and by (iv) inhibiting the production of autoantibodies inside the inflamed gut [107,108]. According to a recent meta-analysis probiotic supplementation improves JIA-related symptoms [107]; however, routine probiotic supplementation is not recommended by the scientific societies.

Dyslipidemia
Barsalou [109] postulated that in JIA, the enhanced atherogenic profile of patients is promoted by the dual exposure to both traditional and non-traditional CVD risk factors. In this population, the inadequate intake of energy, paired with the physical inactivity, overweight, chronic inflammation, autoantibodies and medication use (GCC in particular), tend to aggravate dyslipidemia [69]. Children on ETN therapy exhibit elevated total cholesterol (TC), low-density lipoprotein cholesterol (LDLc) and high-density lipoprotein cholesterol (HDLc) levels, whereas, on the other hand, TG and atherogenic index (TC/HDLc) are decreased, evolving to a less-atherogenic profile post-anti-TNF-α treatment [110]. Similar improvements were also shown following MTX treatment alone, or in combination with immunobiological agents [111]. Disease duration does not appear to be related to the observed dyslipidemia [112].
In Brazil, 83.3% of the patients with JIA demonstrated dyslipidemia based on the TC, HDLc, LDLc, TG, non-HDLc, apolipoprotein A 1 (ApoA 1 ), and apolipoprotein B (ApoB) concentrations [68]. Among JIA subtypes, a more atherogenic profile is manifested in sJIA [113,114]. The observed dyslipidemia appears to be aggravated in bouts of elevated disease activity [113]. During the active stage of JIA [115], children with the systemic and RF-positive polyarticular subtype present a reduction in HDLc and a concomitant increase in TG levels. These changes in lipid profile show improvement during disease remission. Inevitably however, dyslipidemia is also associated with increased cardiometabolic risk, as indicated by the elevated uric acid (UA) concentrations [116] observed among patients with JIA.
Nonetheless, aside from the observed disturbances regarding HDL particle distribution, non-lipid transporting activities and cholesterol efflux are also altered [117]. As a result, patients with JIA exhibit reduced arylesterase activity and endothelial cell migration compared to healthy controls. In parallel, macrophages exposure to serum from patients with JIA revealed a smaller increase in the mRNA expression of ATP binding cassette A1 and a greater increase in the expression of both ATP binding cassette G-1 and SR-B1, compared with healthy controls [117].
Interestingly, results from the Lehigh Valley Health Network (LVHN) [118] showed that, despite that fact that most scientific societies advocate for the frequent (every three months) lipid screenings of patients with RMDs [119], approximately one third of the patients on JAK and IL-6 inhibitors fulfill the indications for the initiation of lipid lowering therapies, including lifestyle modifications, but are left untreated.

Bone Health
Systemic autoimmune disorders such as JIA can affect the skeletal system, reducing BMD and increasing the risk of fragility fractures during childhood [120]. In parallel, pharmacotherapy, and in particular CCS use has been shown to affect bone health and BMD deposition. As a result, lower total BMD [90,[121][122][123] and increased bone resorption as measured by greater carboxyterminal telopeptide (CTX) of type I collagen levels are observed in patients with JIA [77]. These marked deficits in BMD coincide with reduced bone strength and are associated with disease severity and duration [124]. Greater CCS use, TNF-α blocking agents and sJIA are associated with extended bone deficits [123,125]. On the other hand, rhGH treatment seems to improve bone characteristics, including bone mineral content (BMC) in children with JIA and short stature [126]. Table 3 presents the results of randomized controlled trials (RCTs) delivering dietary interventions in children with JIA, with the aim of improving bone health.

Outcomes Results
Warady [132] USA N = 10 children with RMD and OP (n = 6 with JRA), on CCS (1) Ca ONS (2) vitamin D ONS 6 months OC, BMD Spinal BMD improved with supplementation but OC levels remained low throughout the study.
Abbreviations In a Calcium (Ca) ONS RCT, Carrasco et al. [127] randomized children with JIA to Ca and vitamin D supplementation, or vitamin D alone, for a total of two years. The results revealed that, among the arm receiving combined Ca and vitamin D ONS, serum levels of 1,25-dihydroxyvitamin D (25(OH)D), osteocalcin, parathyroid hormone (PTH), and urine Phosphorus (P) were lower, whereas no incidences of hypercalciuria were noted on spot testing based on the urinary Ca-to-creatinine ratio. The authors concluded that ONS with Ca met the physiologic needs of children, causing an increased Ca urine loss. In a similar trial, Lovell et al. [129] showed that 24 months of Ca supplementation significantly improved BMD in children with JIA compared to placebo; however, participant's sex, initial BMD, Tanner stage, treatment adherence and body composition were factors explaining the higher BMD in the Ca-arm. In a cross-over manner, Hillman [128] suggested that the percentage of true Ca absorption in JIA is, in fact, in the low range and not affected by any treatment with ONS (placebo, vitamin D 3 , Ca, or combination of vitamin D 3 + Ca). Thus, although vitamin D 3 at 2000 IU/day may increase serum 25(OH)D and Ca levels, it does not confer any improvements in bone mass accretion. In parallel, supplementation with Ca (1000 mg/day) was similarly unsuccessful in improving bone mass.
Tang [131] evaluated the administration of a high oral cholecalciferol dose (2000 IU/day) for 24 weeks, compared to no intervention in children with JIA. At the end of the trial, no differences were noted in the BMD or the JADAS-27 score between the two arms, indicating a lack of effect of vitamin D in improving bone health. Nonetheless, given that dual energy x-ray absorptiometry (DXA) can only partly explain bone quality, it has been suggested that bone abnormalities in JIA might be better assessed using complementary imaging techniques, including peripheral quantitative computed tomography (pQCT) or quantitative ultrasonography (QUS) [125,133].
The available research shows that improving bone health in JIA through supplementation is not always successful. Although serum levels of specific biomarkers might be improved post-supplementation, the level of improvement regarding BMC and BMD is questionable, as a possible adverse event of medication and the low levels of PA.

Energy Requirements and Dietary Intake
According to Knops et al. [42], although crude resting energy expenditure (REE) appears to be indifferent when comparing patients with JIA to healthy controls, REE correction for body mass and FFM showed a 18% higher REE/kg of body weight among young patients with sJIA and 8% greater REE/kg of body weight for oligo-and polyarticular patients. As a result of these elevated REE levels, patients often report feeling overwhelmingly and unremittingly fatigued [101,134,135], complaining for having inadequate energy irrespective of the PA levels [136].
In an early study [43], using direct calorimetry, decreased energy expenditure (EE) was recorded in comparison to the "ideal" values. Nonetheless, greater EE was observed among children with long-time disease and longer follow-up. Furthermore, greater values of IL-6 and TNF were associated with reduced EE [43]. Table 4 details all the studies conducted to date, assessing the dietary intake of pediatric patients with JIA. According to the literature more than half (54%) of the children fail to reach energy intake levels to meet their daily requirements [43]. Furthermore, proteinenergy malnutrition (PEM) has been previously identified in 36% of the young patients with JIA, irrespective of the subtype [137].

Studies Assessing Dietary Intake in Children with JIA
Of note, low dietary intake in JIA can also be the result of mouth-related problems. Dental problems and temporomandibular joint disease can affect the child's ability to eat, as may functional difficulties resulting from arthritis affecting the upper limb [138]. Mouth sores are also frequently encountered, especially in children on MTX treatment [24].

Energy and Macronutrient Intake
In a sample of Brazilian children with JIA, Caetano et al. [69] observed an excess in the intake of energy, protein, and lipids was by 12.5, 75, and 31.3%, respectively, paired with a low Iron (Fe), Zinc (Zn), and vitamin A dietary intake. On the other hand, 41.7, 8.3, and 31.3% of the patients reported consuming less energy, protein, and lipids, than required. The food groups identified as being under-consumed involved milk and derivatives (78%), meat (32%), vegetables (98%) and fruits (84%). As a result, Fe intake was suboptimal in 29.2% of the sample, Ca in 62.5%, Zn and vitamin A in 87.5% and vitamin B6 in 64.6%. Nearly half (42.1%) of the children consumed a juice with added sugar on a daily basis.

Suboptimal Levels of Immunonutrients
In an early study, Johansson and associates [146] showed that girls with JIA exhibited lower plasma Selenium (Se) levels compared to their healthy peers. Amancio et al. [139] suggested that, in children with JIA, serum Copper (Cu) concentrations are related to the number of inflamed joints, and greater compared to that of controls. Similar findings were also reported by Haugen et al. [143], where the concentrations of Zn were lower among children with polyarticular JIA compared to controls, and Cu levels were elevated. Se and Zn are important cofactors of antioxidant enzymes, regulating a plethora of inflammatory and immune responses and important micronutrients, in rheumatic diseases in particular [17]. Aside from participating in the antioxidant defense, they also regulate the innate and adaptive immune responses [17,147,148]. With regard to Se in particular, we are yet unsure if the decreased Se concentrations consist of the residue of malnutrition associated with chronic disease or are due to other factors [17,149].

The Curious Case of Ferritin
With regard to Fe levels, a plethora of studies report low Fe and hemoglobin (Hgb) levels among children and adolescents with JIA [143,[150][151][152][153][154][155], despite the dietary Fe intake. In parallel, sideroblasts are also reduced in number, pointing to the fact that, in these patients, Fe does not appear to be sufficiently transferred to the erythroid series, limiting the ability for it to be used by erythroblasts, resulting in an absolute iron deficiency anemia (IDA) [156]. Subsequently, Fe is mainly stored in the form of ferritin, although not in an accessible form, hampering metabolism in JIA [156]. Serum ferritin is a known inflammatory marker, although we are not yet sure if it participates in the development of inflammation, or it simply reflects an elevated inflammatory profile [157]. It has been suggested that serum ferritin arises from damaged cells, as a proxy of cellular damage [157]. As a result, the exis-tence of macrophage activation syndrome (MAS) skyrockets the ferritin values [158,159]. However, many patients with active disease (mainly sJIA) without a MAS diagnosis often demonstrate extreme serum ferritin levels, indicating that a distinct inflammatory pathway may exist that is unique to this JIA subtype [154,159,160]. In sJIA in particular, extremely high ferritin levels may be exhibited, even exceeding 1000 ng/mL [161,162], and may assist in establishing the diagnosis of sJIA [159]. Nonetheless, recently, the ferritin/erythrocyte sedimentation rate (ESR) ratio has been suggested for the identification of MAS in sJIA [163].
Anti-TNF therapy has been shown to improve Hgb and the mean corpuscular volume of red blood cells (MCV) without the use of Fe ONS, with a parallel improvement in the concentrations of serum Fe and percent transferrin saturation [156].

Fatty Acid Intake and Status
During the active disease phase children with JIA [141] exhibit low levels of arachidonic acid and docosahexaenoic acid (DHA), a finding that suggests the possible role of polyunsaturated fatty acids (PUFA) in the pathogenesis of JIA, while pointing out the possible need for supplementation. On the other hand, fat intake of children with JIA is also suboptimal, with a recent study [68] showing that 75.8% and 79% of the youngsters with JIA, respectively, presented a SFA and trans-fat intake greatly exceeding the recommendations.

Considerations for the Assessment of Energy Requirements in JIA
With the exception of one study [69], it appears that children with JIA consume suboptimal amounts of energy and several micronutrients. This could be due to diseasespecific characteristics and medication adverse effects, and/or to the lack of specific energy equations to assess the actual needs of patients with JIA, leading to an overestimation of the energy requirements. Thus, it is likely that the use of stress factors for the estimation of energy requirements might not be warranted for children with JIA.
In sJIA in particular, where, apart from inflammation, fever is presented frequently during active disease phases, the estimation of REE becomes extremely difficult. Active disease status is associated with reduced body weight, indicating a hypercatabolic state [66]. Overall, in children and adolescents with JIA, the energy expenditure is affected by the fat-free mass of patients, the underlying inflammation (if any), the growth requirements of the children, the exercise-induced activity and the thermogenic effect of food. Fluctuations in disease activity and inflammation induce concomitant alterations in the REE, and thus, when REE cannot be measured, different equations must be used for children in remission compared to those with exacerbated JIA.

Dietary Interventions for Children with JIA
In a pilot uncontrolled study, Berntson et al. [164] evaluated the effect of an antiinflammatory diet in children and adolescents with JIA. After four weeks of intervention, the sensation of pain, morning stiffness and inflammatory proteins concentrations were reduced, while, in parallel, physical function, fecal butyrate levels and arthritis were improved in the majority of participants. Previously, Berntson had tested the effect of exclusive enteral nutrition (EEN) in a patient with polyarticular JIA for two periods of almost seven weeks each, several months apart [165]. EEN had a remarkable anti-inflammatory effect that was sustained for several months after the treatment periods.
The importance of MNT as a supplementary treatment in JIA becomes apparent in a recent study, where one third of young patients with JIA reported trialing some "special diet" throughout the course of the disease, and more than half of the parents perceived symptom improvement post-intervention [166]. Overall, research on specific dietary interventions targeting JIA is limited, indicating the need for the conduction of more studies evaluating the effect of different diets on JIA outcomes.

Interventions Targeting Insufficient Vitamin D Levels
Vitamin D is an important modulator of both the innate and adaptive immune systems [167,168]. Several lines of evidence reveal suboptimal vitamin D status in children with JIA [169][170][171][172][173][174][175][176][177][178][179][180][181][182][183][184][185][186][187][188] (Table 5). Kondratyeva and associates [182] suggested that among patients with JIA, vitamin D status does not appear to be not affected by the presence of vitamin D receptor (VDR) genetic variants. Similar findings were also suggested in an Italian cohort [177], where, additionally, both the TT genotype and the T allele were more frequently observed among patients. Furthermore, VDR polymorphisms have also been shown to affect lipid metabolism in JIA, with the VDR FokI FF genotype in particular, demonstrating protective effects [189].  Some researchers reported the existence of a negative association between vitamin D status and disease activity, or the number of affected joints [173][174][175]181,183,185,187], whereas others suggest that low levels of vitamin D are independent of disease status and particularities [169,172,176,178,180,182,186,188,191]. Recent global geographical mapping of the available research revealed the existence of a north-south geographical gradient in the association between low vitamin D status and increased arthritis activity [170]. A Canadian study [187] was the only one reporting greater 25(OH)D levels among children with JIA compared to the same-age general population, a finding attributed to the more frequent use of dietary supplements containing vitamin D. In parallel, birth seasonality has also been suggested to influence vitamin D status, with children with JIA and vitamin D deficiency/insufficiency being more often born during the fall and winter months [187].
Nonetheless, the observed low levels may partly explain why children with JIA often fail to correct bone-related parameters, despite the use of more effective current drugs [184]. Low levels of vitamin D have also been associated with increased CVD risk, as evidenced by greater carotid intima media thickness (cIMT) and lower flow mediated dilatation of the brachial artery (FMD) of children with JIA [190]. Furthermore, a positive correlation has been shown to exist between 25(OH)D levels and FMD, and an inverse one between 25(OH)D concentrations and cIMT. This may explain, at least in part, why despite more effective current drugs, young patients with JIA do not achieve a bone-normal condition over time. Patients with a more severe form of the disease might require higher supplementation of vitamin D to maintain normal 25(OH)D serum levels.
MTX therapy has also been associated with vitamin D insufficiency in JIA [192] and other chronic diseases [193]; however, no specific biologic explanation has been suggested for this phenomenon. It is possible that suboptimal vitamin D levels are participating in the onset and progression of JIA, but it is also highly likely that children with chronic diseases in general are less active outdoors, limiting vitamin D synthesis. Nonetheless, vitamin D ONS has been suggested as an adjuvant in JIA therapy, aiming to correct for the observed low concentrations and the increased CVD risk associated with JIA [167,168,183,184,194]. Supplementation with higher doses of vitamin D (2000 IU/day) is also recommended as an effective generic adjuvant treatment in active JIA [131]. According to Kondratyeva and associates [182], the prophylactic ONS doses of 500-1000 IU/day and 1500-2000 IU/day do not seem to meet the augmented requirements of children with JIA, as the majority seem to attain low levels, despite ONS use, calling for an update of the evidence behind the ONS recommendations.

Interventions Delivering Supplementation/Dietary Intake with Fatty Acids
Active and short-lasting JIA is associated with lower levels of selected fatty acids, including arachidonic acid (AA) and docosahexaenoic acid (DHA). Furthermore, serum α-linolenic acid (ALA) levels are more elevated in poly-JIA than in oligo-JIA. What is more interesting, though, is that serum n-6 and n-3 levels appear to be negatively related to the active joint count, ESR and CRP concentrations, and positively correlated to the platelet count [141].
In patients with rheumatoid arthritis, meta-analyses suggest that the use of n-3 fatty acid supplements may confer clinical benefits [195]. In the case of JIA, uncontrolled clinical trials indicate that ONS with n-3 FA can reduce the daily requirements of NSAIDs, improve the clinical manifestations associated with the disease and lower the degree of inflammation [196][197][198] (Table 6). Furthermore, ONS with n-3 PUFA also induces improvements in several immune parameters, in particular Immunoglobulin A (IgA), Immunoglobulin M (IgM), CD8, CD4/CD8, CD22, IL-1, and IL-4 concentrations, stabilizes lymphocyte cell membranes and activates the organoid level of intracellular regeneration, as indicated by the circulating immune complexes (CIC) [198]. The only RCT performed on pediatric patients with JIA randomized them either to ibuprofen treatment alone, or in combination to a diet with a high omega-3 PUFA content [199]. After five months of intervention, the ibuprofen doses were reduced in the arm receiving ibuprofen in parallel to the high-n-3 diet, although the result was not different compared to that of controls [199]. Table 6. Clinical trials delivering dietary interventions for improving bone health in children with JIA.

First Author
Origin Population Intervention(s) and Comparator(s)

days
Inflammation markers, ROM, pain, CRP and ESR levels Treatment with Kre-Celazine improved and, in some cases, resolved symptoms among youngsters who had not achieved satisfactory relief from the use of NSAIDs. Vargová [199] Slovakia N = 23 children with JCA (1) ibuprofen treatment (2) ibuprofen + diet with an increased n-3 PUFA content

months Ibuprofen dose reduction
In the ibuprofen + diet arm, in the course of treatment the original ibuprofen dose declined by 17.3%, while in the control arm there was a decline equal to 6.5%.

Conclusions
Due to the age of onset, the complexity of the disease and the associated medication, children with JIA are prone to the development of several nutritional issues, warranting expert monitoring. Dietary factors are important regulators of the physiological development of children and adolescents, and in those with a JIA diagnosis their impact on growth, body weight status, bone health, inflammation, and disease activity is augmented (Figure 1). In parallel, most studies indicate a suboptimal dietary intake among children with JIA, and the existence of several micronutrient deficiencies, all of which could be corrected with the appropriate dietary education conducted by experts. For several nutrients, it appears that the recommended allowances fail to meet the requirements of children with JIA. Thus, a multi-disciplinary effort including dietitians and rheumatologists should re-evaluate these recommendations and set JIA-specific targets with regards to the energy, protein and vitamin D intake of young patients in particular.
( Figure 1). In parallel, most studies indicate a suboptimal dietary intake among children with JIA, and the existence of several micronutrient deficiencies, all of which could be corrected with the appropriate dietary education conducted by experts. For several nutrients, it appears that the recommended allowances fail to meet the requirements of children with JIA. Thus, a multi-disciplinary effort including dietitians and rheumatologists should re-evaluate these recommendations and set JIA-specific targets with regards to the energy, protein and vitamin D intake of young patients in particular. Collectively, it appears that although MNT can aid in improving JIA prognosis, the number of studies delivering dietary interventions remains low. More recently, the importance of nutrition in JIA has been acknowledged, with a couple of reviews and opinion papers being published [200,201], although the present work consists of the most Collectively, it appears that although MNT can aid in improving JIA prognosis, the number of studies delivering dietary interventions remains low. More recently, the importance of nutrition in JIA has been acknowledged, with a couple of reviews and opinion papers being published [200,201], although the present work consists of the most comprehensive review conducted to date. In parallel, patients and health providers would greatly benefit from the development of specific nutrition and lifestyle guidelines for the management of JIA. The development of a food pyramid, as recently conducted for rheumatoid arthritis [202], but with the use of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system [203], might be of great use for those involved in the management of JIA.
As the link between nutrition and JIA prognosis is bidirectional and strong, dietitians should be educated on JIA and work closely with rheumatologists and other health professionals in an effort to improve the health of patients in a holistic manner, achieve and sustain maximum growth velocity. Furthermore, rheumatologists would greatly benefit from the inclusion of a specialist dietitian in their team, in an effort to identify nutritional issues in a timely manner and offer the best available MNT in each case.