In the present study, we examined concentrations of circulating FGF19, FGF21 and Klotho proteins among the normal weight, obese and overweight children and adolescents and their relationships with metabolic parameters. The results complement existing reports with thus far lacking paediatric-specific data.
We noted increased FGF21 concentrations in children and adolescents with MS compared to other subjects. Moreover, FGF21 levels correlated with clinical (adiposity, arterial hypertension) as well as biochemical (TG, HDL-C) features of MS. Despite its role in metabolism regulation, the reports on its usefulness as a biomarker for obesity and abnormalities associated with MS are conflicting [8, 24, 25, 26]. FGF21, produced during fasting mainly in the liver, promotes gluconeogenesis, lipolysis, ketogenesis, ameliorates glucose uptake and improves insulin sensitivity [5]. It was previously shown that systemic administration of FGF21 has therapeutic benefits against obesity-related medical complications in obese animals [27, 28, 29]. Despite the beneficial effects of FGF21, increased endogenous FGF21 level has been observed in adults with obesity that may result from the tissue resistance to FGF21. This paradoxical phenomenon led to the hypothesis that central obesity, strangely associated with MS, is a state of FGF21 resistance resulting from decreased of FGF coreceptor (betaKlotho) expression in white adipose tissue. This hypothesis seems to corroborate our results as well other authors previous findings [30, 31]. However, as Reinehr et al. we did not confirm the relationship between FGF21 concentrations and insulin resistance occurrence [32]. The results may be explained by new data showing that elevated FGF21 levels in obesity serve as a defence mechanism to protect against systemic IR by upregulating adiponectin in subcutaneous but not visceral fat with the following anti-inflammatory action resulting from local M2 macrophage polarization [33].
To our knowledge, this is the first study in children and adolescents to show that FGF19 beside pubertal stage and BMI Z-score, is an independent predictor for IR. The studies on animal model indicated that administration of recombinant FGF19 to obese mice led to reduction of body mass, decreased blood glucose level as well as increased insulin sensitivity by central FGF19 action [34]. However, the data focused on relationship between FGF19 levels and metabolic parameters including IR are conflicting [6, 7, 35]. Despite the fact that FGF19 actions lead to the similar metabolic processes like insulin, the differences between two hormones are still not well understood [5]. What is certain, FGF19 unlike insulin, is released from the small intestine, not pancreas, and it reaches its peak serum level at 3 h after a meal, not at 1 h [5]. Furthermore, FGF19 acts using FGFR1/betaKlotho or FGFR4/betaKlotho pathways. The activation of FGFR1/betaKlotho leads to regulation of glucose metabolism, including suppression of gluconeogenesis, stimulation of glucose catabolism and reduction of TG synthesis. On the other hand, FGFR4/betaKlotho receptor activation is connected with reduction of bile acid levels and alteration in bile acid pool composition, which potentially may promote increased TG levels [5, 6, 36]. It has been speculated that insulin and FGF19 may have an inverse effect on each other [6]. Consequently, the insulin-resistant state leading to increased levels of circulating insulin may provoke observed decreased FGF19 levels.
In our study we noted that children and adolescents affected by obesity show higher serum Klotho concentrations than those with normal body weight. This contrasts with other studies on the matter. Amitani et al. showed markedly lower plasma Klotho levels in patients with obesity and anorexia nervosa than in the control group, which suggests that Klotho may reflect normal nutritional status [14]. On the other hand, Wojcicki et al., in the group of healthy Latino neonates, found no association between either weight, length at birth or obesity in early childhood and cord blood Klotho levels [16]. However, in the literature the data focused on the relationship between obesity and Klotho levels in children and adolescents are insufficient. Our results may be supported by the fact that Klotho is one of the regulators of adipogenesis. It was previously revealed that Klotho increases adipocyte differentiation in vitro. Mice without Klotho have less detectable adipose tissue content than wild-type animals. Moreover, mice that lack the Klotho gene are resistant to obesity induced by a high-fat diet [37].
Interestingly, we noted different Klotho levels between the groups of patients with IR and normal insulin sensitivity. This may be explained by the fact that Klotho takes part in enzymatic modification of N-glycans in insulin and IGF-1 receptors and thus inhibits the intracellular insulin/IGF-1 signalling pathway. It blocks insulin-stimulated glucose uptake that contributes to IR development [12, 38]. What is important, probably inhibition of IGF-1 signalling cascade is associated with increased resistance to the oxidative stress and lead to the extension of life, which is one of the major functions of Klotho [39].
There are potential limitations of our study. Firstly, we studied only circulating hormones without their local (i.e. liver, adipose tissue) expression, which was out of scope for this study. Secondly, we relied on BMI z-score to recognize overweight and obesity without body content assessed by Dual-energy X-ray absorptiometry (DXA), which was not available during this study. Estimated BF% could not be translated into sex- and age-independent Z-scores or percentiles due to the lack of modern paediatric charts for Polish population. Moreover, serum Klotho concentrations may depend on vitamin D and calcium-phosphate homeostasis which we did not examine in the studied subjects. However, the above-mentioned dependence was observed especially in patients with chronic kidney diseases who were excluded from this study. Our multivariate model for HOMA-IR explained only a small fraction of between-patients variability. This demonstrates that there are likely other factors which might be associated with IR in a stronger and more direct way. Moreover, the subjects were enrolled to the study in the hospital and this might be a potential limitation. On the other hand, the recruitment in inpatient condition assured the similar exposure to potential confounding factors including diet, physical activity, ambient temperature. Finally, the study was carried out in a hospital single centred with sample derived from local population, which prohibits generalisation of results onto Polish or European children.