Anti-Inflammatory and Pro-Inflammatory Adipokine Profiles in Children on Vegetarian and Omnivorous Diets

Adipose tissue is a highly active endocrine organ that secrets many pro-inflammatory as well anti-inflammatory adipokines. The aim of the study was to assess serum adipokine profile in prepubertal vegetarian and omnivorous children. Sixty-two children on a vegetarian diet and fifty-five children on an omnivorous diet, aged 5 to 10 years, were studied. Dietary assessment was performed using a nutritional software program. Body composition was measured by dual-energy X-ray absorptiometry. Serum concentrations of adipokines: leptin, soluble leptin receptor (sOB-R), adiponectin (total and high molecular weight), resistin, visfatin, vaspin, and omentin were determined by immunoenzymatic assays. Both studied groups of children were comparable in terms of age, weight, height, body mass index, and body composition. Vegetarians had a lower (p = 0.017) leptin/sOB-R ratio and lower serum concentrations of resistin (p = 0.051), compared with omnivores. Average levels of other adipokines did not differ between both groups of children. However, we observed significantly higher ratios of anti-inflammatory to pro-inflammatory adipokines: adiponectin/leptin 0.70 (0.37–0.93) vs 0.39 (0.28–0.74), p = 0.005, and omentin/leptin 0.40 (0.23–0.83) vs. 0.33 (0.15–0.48), p = 0.011 in vegetarians compared with omnivores. A well-planned vegetarian diet might beneficially affect the adipokine profile and inflammatory status expressed by the ratios of anti-inflammatory to pro-inflammatory adipokines in prepubertal children.

We examined 62 prepubertal children (54.8% male, age range 5-10 years) following a lacto-ovo-vegetarian diet, which included dairy and egg products. The inclusion criteria regarding vegetarians were: being on a lacto-ovo-vegetarian diet from birth, in the prepubertal period, generally healthy (without development and nutrition disorders, normal-weight with a body mass index (BMI) z-score between −1 + 1). The exclusion criteria were: not in the prepubertal period, history of low birth weight, gastrointestinal diseases accompanied by malabsorption, history of acute and chronic infection, and drug consumption, except of standard vitamin D supplementation. In this study, we recruited the maximum possible number of prepubertal vegetarian children who consumed a lacto-ovo-vegetarian diet from birth.
The control group included 55 healthy children (49.1% male, age range 5-10 years) following a traditional omnivorous diet, which consisted of consuming meat, poultry, and fish.
Pubertal stage was assessed according to the Tanner's criteria. The studied children (vegetarians and omnivores) attained the World Health Organisation (WHO)recommendation regarding physical activity and accumulated about 60-90 min/day of MVPA (moderate-to-vigorous physical activity) and approximately 30 min per day of VPA (vigorous physical activity). VPA included activities after school twice a week for 1 or 2 h. The level of physical activity was similar in both studied groups of children.

Dietary Assessment
Dietary assessment in the studied children had been described in detail previously [40]. The parents of studied children (advised by a nutritionist) were asked to prepare a food diary for their children. Three consecutive days (two weekdays and one weekend day) were selected, and data were analyzed by a nutritional software program Dieta5®(National Food and Nutrition Institute, Warsaw, Poland). The average daily energy intake, dietary protein, fat, carbohydrates, and fiber intakes was assessed in the studied children, and compared with the current recommendations for Polish children [41].

Anthropometric Parameters
Physical examination included height and weight measurements, and were performed in all children. Body mass index (BMI) was calculated as the body weight (kg) divided by height squared (m 2 ). Fat mass, lean mass, and fat free mass were measured by dual-energy X-ray absorptiometry (DXA) using Lunar Prodigy (General Electric Healthcare, Madison, WI, USA). Fat mass index (FMI) and lean mass index (LMI) were calculated as the fat mass or lean mass (kg) divided by the height squared (m 2 ), respectively.

Biochemical Measurements
Blood samples were obtained after an overnight fast (between 8:00-10:00 a.m.), and serum samples were frozen at −20 • C until analysis (no longer than after two months). Concentrations of serum adipokines were determined by a commercial enzyme-linked immunosorbent assay (ELISA) kit, according to the manufacturer's instructions. Serum leptin and soluble leptin receptor (sOB-R) concentrations were determined using ELISA kits from DRG Instruments GmbH (Marburg, Germany), with intra-assay variation coefficients (CV) ranging between 4.2-7.3%, and inter-assay CV ranging between 3.7-9.1% for leptin, and intra-assay CV 7.10-7.23% and inter-assay CV 6.21-9.81% for sOB-R. Serum total and HMW adiponectin levels were measured using a kit from ALPCO (Salem, MA, USA), with intra-assay CV 5.3-5.4% for total adiponectin and 3.3-5.0% for HMW adiponectin, and inter-assay CV 5.0% for total adiponectin and 5.7% for HMW adiponectin, respectively. The resistin level was determined using a kit from DRG Instruments GmbH (Marburg, Germany), with intra-assay CV 5.2-6.6% and inter-assay CV 7.0-8.1%. The concentrations of vaspin and visfatin were determined using a Vaspin (human) ELISA kit and Nampt (Visfatin/pre-B cell colony-enhancing factor (PBEF) (human) kits from Adipogen Life Science (Liestal, Switzerland). The intra-assay and inter-assays were 1.31-3.85% and 3.27-9.06% for vaspin, and 2.31-9.11% and 4.66-7.24% for visfatin, respectively. Serum omentin was measured using a kit from SunRed Biotechnology (Shanghai, China), with a within-assay variability of less than 10% and a between-assay variability of less than 12%.

Statistical Analysis
All statistical analyses were carried out using IBM-SPSS software, version 23.0 (SPSS INC., Chicago, IL, USA). The Kolmogorov-Smirnov test was used to test the normality of the distribution of the variables. For compare categorical data the chi-squared test was used. Baseline characteristics were presented as mean ± standard deviation (SD) for the variables with normal distribution, or as median values and interquartile ranges (IQRs) for the variables that were non-normally distributed. The ratios of anti-to pro-inflammatory cytokines: adiponectin/leptin, adiponectin/resistin, omentin/leptin, omentin/resistin, vaspin/leptin, and vaspin/resistin were calculated. The differences between groups regarding anthropometrical, nutritional, and biochemical parameters were evaluated using the Student's t-test for normally distributed variables, and the Mann-Whitney U test for non-parametric variables. A univariate correlation analysis was performed using the Pearson or Spearman test, as appropriate. A p value of <0.05 was considered to be statistically significant.

Results
Both groups of children were similar in terms of age, body weight, height, BMI, and lean and fat mass ( Table 1). The mean values of LMI were similar in both studied groups of children, but FMI was significantly (p = 0.044) lower in the vegetarian than in the omnivorous children. Analyzing the children's diets, we found that the average daily energy intake in both groups of children was within recommendations, whereas the proportions of macronutrient intakes in vegetarians and omnivores were different. Compared with omnivores, vegetarians had a similar percentage of energy from fat (p = 0.286), a significantly lower percentage of energy from protein (p < 0.001), and a higher percentage of energy from carbohydrates (p = 0.002). As expected, in the vegetarian diet, the daily intake of plant protein was significantly (p = 0.024) higher, and the intake of animal protein was significantly (p < 0.001) lower than in the omnivorous diet. Additionally, the dietary intake of fiber was higher (p = 0.011) in vegetarians compared with omnivores. Among adipokine levels, we observed a significantly (p = 0.017) lower ratio of leptin to soluble leptin receptors, and the resistin concentration was on the border of statistical significance (p = 0.051) in vegetarians compared with omnivores ( Table 2). Serum concentrations of other adipokines, such as adiponectin, visfatin, omentin, and vaspin, did not differ significantly between both groups of children; however, there was a slight increase in the median value of circulating visfatin and omentin. When analyzing the correlations, we found a significant positive association between resistin and visfatin concentrations (r = 0.485, p < 0.001), and between resistin and the HMW/adiponectin ratio (r = 0.327, p = 0.009) in the group of vegetarian children, and between vaspin level and the leptin/sOB-R ratio (r = 0.287, p = 0.045) in the omnivorous group. We did not observe associations between other serum adipokine levels in any of the studied groups of children (data not shown). Regarding significant correlations between adipokines and anthropometric parameters, we noted that in vegetarians, the ratio of leptin/sOB-R was positively correlated with body weight (r = 0.371, p = 0.003), height (r = 0.400, p = 0.001), fat mass (r = 0.475, p = 0.000), fat/lean mass ratio (r = 0.346, p = 0.006), and FMI (r = 0.379, p = 0.002), and it was negatively correlated with LMI (r = −0.291, p = 0.022) ( Table 3). Additionally, the HMW/total adiponectin ratio was negatively correlated with BMI (r = −0.379, p = 0.002) and FMI (r = −0.263, p = 0.039). Also, resistin was negatively correlated with fat mass (r = −0.305, p = 0.016), fat/lean mass (r = −0.303, p = 0.017), and FMI (r = −0.321, p = 0.011). In contrast, we did not observe correlations between adipokine concentrations and anthropometric data in the omnivorous group, except for a relation between the leptin/sOB-R ratio and weight (r = 0.292, p = 0.031), and between vaspin and LMI (r = 0.366, p = 0.016).
In the studied groups of children, we observed differences in the ratios of anti-inflammatory to pro-inflammatory adipokines (Table 4).

Discussion
Adipokines secreted by adipose tissue may act both locally and peripherally, regulating a large number of physiological processes. The dysregulation of adipokine synthesis can alter homeostasis and lead to pathological conditions [8]. The majority of research has focused on obesity and its impact on the adipokine profile, although a few papers consider the adipokine pattern in an undernutrition state, such as anorexia nervosa [42,43]. Analyzing adipokine status, Mantovani et al. [44] suggested that leptin, resistin, or adiponectin might be early markers of changing from lean to obesity, even before the occurrence of metabolic alterations. Generally, the researchers concluded that changes in nutritional status, as well in body weight and fat content, may influence the hormonal activity of adipose tissue, with special attention being paid to the balance between adipokines with pro-inflammatory and anti-inflammatory activities.
It is widely known that consuming a properly composed diet has been associated with better health status, due to the protective effect against various chronic diseases [37,45]. Evidence regarding the role of diet quality on glucose and lipid metabolism and inflammatory markers has been established in multiple observational studies [45][46][47][48][49]. The authors reported that a plant-based pattern and the Mediterranean diet (characterized by a high consumption of vegetables, fruits, legumes, olive oil, and low consumption of red meats) are associated with favorable health status and less inflammation [25,27]. Additionally, they observed that this kind of diet was associated with higher levels of adiponectin, lower levels of leptin, and unchanged levels of resistin and vaspin.
There is a paucity of data on the effects of the vegetarian diet on serum adipokine status [36,39]. We studied prepubertal children and found a different adipokine profile with a significantly lower leptin/soluble leptin receptor ratio, and resistin concentrations in vegetarians compared with omnivores. The changes in other adipokine levels were not statistically significant between the groups, but we noticed slightly higher serum omentin and visfatin concentrations in the vegetarian group. It is worth noting that we determined visfatin as nicotinamide phosphoribosyltransferase (NAMPT)-an enzyme form that plays a role in insulin secretion in pancreatic beta-cells, and in the delayed neutrophil apoptosis of sepsis [50,51]. However, visfatin is an adipokine that exerts pleomorphic effects, and it is not excluded that other forms of visfatin might have different properties and associations with anthropometric and nutritional parameters [52].
Significant positive correlations between the leptin/sOB-R ratio and anthropometric parameters (body weight, height, fat mass), between resistin and fat, and negative correlations between the HMW/total adiponectin and fat observed in our vegetarians (but not in omnivores) suggest exact association between adipokines and body composition in this group of children. We think that conditions with lower leptin and resistin levels, coexisting with lower fat mass and FMI in vegetarians, are favorable for health.
The interplay between adipokines of opposite effects, may determine their better metabolic profile regarding the inflammatory state. This is why we presented the ratios of anti-to pro-inflammatory adipokines reflecting an inflammatory balance that is better than the measurement of individual adiponectins. It is known that adiponectin and omentin have anti-inflammatory effects that contribute to their protective role against metabolic stress. The possible mechanism is that adiponectin plays an anti-inflammatory role through the blockade of NF-kappa B signaling, and it further inhibits the production of TNF-alfa. An adiponectin deficiency is an independent risk factor for endothelial dysfunction and cardiovascular complications. Also, omentin is negatively correlated with chronic inflammation, endothelial dysfunction, atherosclerosis, and calcification [53]. This adipokine is involved in lipid metabolism (through inhibiting oxLDL) and proper synthesis of nitric oxide through the inhibition of TNF-alpha-induced cyclooxygenase-2 (COX-2) expression, with further activation of the endothelial nitric oxide synthase (eNOS)/NO pathway [54]. Moreover, omentin is able to promote the anti-inflammatory M2 phenotype of macrophages during the differentiation of monocytes into macrophages, thus reducing the production of pro-inflammatory factors [16,19]. In this context, the increased ratios of adiponectin/leptin and omentin/leptin observed in our vegetarians may be related to a better metabolic profile in children who consume a well-planned vegetarian diet. It may protect against various chronic diseases in later life.
Other ratios regarding resistin and vaspin were without significant differences between the studied groups. Resistin plays an important role in inflammatory responses since the main sources of resistin secretion are macrophages [10]. Unexpectedly, in our study, serum resistin levels were positively correlated with the HMW/adiponectin ratio, and negatively with fat mass in vegetarians. No relations between resistin concentration and anthropometric parameters or other adipokine levels were observed in omnivores. The exact role of vaspin in the body is still a matter of debate. Some authors believe that vaspin has anti-inflammatory properties. We did not observe any associations of vaspin with body composition indicators in vegetarians; however, in omnivorous children, we found positive correlations between vaspin and LMI, as well between vaspin and the leptin/sOB-R ratio. Further studies are needed to clarify this issue.
In the analysis of correlations between adipokine ratios and dietary nutrient intakes, we observed that the percentage of energy from protein was significantly negatively related with the vaspin/resistin ratio, plant protein intake was positively related with the omentin/leptin ratio (p = 0.052), and dietary fiber with the adiponectin/resistin ratio in vegetarians. In the omnivores group, the percentage of energy from protein was negatively correlated with the adiponectin/leptin ratio and the omentin/leptin ratio, and positively with the vaspin/resistin ratio. The data regarding resistin, visfatin, and vaspin seems to be inconclusive and need further investigation.
There are only a few studies on the associations between a plant-based diet, adipokine secretion, and inflammation states. Kahleova et al. [35] reported that the vegetarian diet is more effective in increasing insulin sensitivity, reducing visceral fat volume, and improving serum adipokine levels than a conventional diet in patients with type 2 diabetes. The authors observed a decrease in leptin (by 35%) and resistin levels (19%), and an increase (by 19%) in total and HMW adiponectin levels in the studied patients after a 3-month dietary intervention. Also, McCarty et al. [55] suggested that, owing to their amino acid composition, plant proteins preferentially stimulate glucagon secretion, and as a consequence, they have favorable effects on adipocyte function, as increased adiponectin secretion, which directly exerts anti-inflammatory effects. In an experimental study on rats, Chen et al. [56], found that vegan diets have shown hypolipidemic effects and an increased expression of adiponectin and its receptor Adipo R1, which is involved in the reduction of gluconeogenesis via the AMP-activated receptor signaling pathway. It is difficult to explain whether changed levels of adipokines are directly associated with diet. Some authors suggested that polyphenols, vitamins, and fiber derived from plant-based foods are the main bioactive compounds that are associated with the anti-inflammatory and antioxidant properties [57][58][59]. It is not excluded that a lower dietary intake of total protein and animal protein, but a higher intake of plant protein and fiber observed in vegetarians may have a direct beneficial effect on body composition, fat tissue secretion, lipid profile, and glycemic control. One of the mechanisms that are likely to be responsible for improved glycemic control is increased insulin sensitivity in response to plant-based diets that partially replace of meats with soy products. Another potential mechanism is improved gastrointenstinal hormone response, especially of incretins [23]. Generally, it is known that too much protein (especially animal protein) in the diet is a predisposing factor for overweight/obesity in children. Also, high fiber consumption has been linked to reduced blood lipids, plaque formation, and cardiovascular risk.
The interpretation of serum adipokine concentrations with regard to their clinical use in pediatrics is hampered by a relative lack of reference values for children and adolescents. To our knowledge, our study is the first to determine and analyze the adipokine profile in children on a vegetarian diet. This represents a first attempt to study the interaction of diet and adipokines in the modulation of important metabolic processes. We think that future studies will provide information about implications of adipokine profile on public health research, and on public health practice. The beneficial adipokine profile in vegetarian children demonstrated in our study may indicate the benefits that are associated with the use of this diet in the context of the prevention of diseases underscored by inflammatory mechanisms. Some authors postulated strong association between the adipokine ratio (high leptin/adiponectin ratio) with a higher risk of type 2 diabetes, cardiovascular disease, or insulin sensitivity [60,61]. We suggest that ratios of other pro-inflammatory/anti-inflammatory adipokines may be useful in the prevention of dietary-related diseases as well as other inflammatory disorders as diagnostic tools in the future.
The results of our research are also an important contribution to the understanding of the pathomechanism of the impact of diet components on metabolic processes. Our previous study suggests that the leptin/adiponectin ratio may be a reliable parameter to assess the risk of obesity-related disorders in prepubertal children with simple obesity carrying different genotypes of the-11377C>G polymorphism in adiponectin gene [62]. The association between polimorphisms in other adipokine genes and pro-inflammatory/anti-inflammatory adipokine ratio may also reflect the relations between genetic variants, serum adipokine levels, and the inflammatory state. However, gaining more knowledge about these mechanisms requires future research on both children and adults, using different nutrition models.
Our study has several possible limitations. First, we analyzed relatively small samples, which reduced the power of our results. However, our data come from two groups of children who were similar in terms of age, weight, height, BMI, and body composition parameters. Second, we measured fat mass and percentage of fat mass, but did not examine the volume of subcutaneous and visceral fat, which could have provided a better tool for adipokine assessment than basic anthropometric parameters. Third, the obtained results were based on single measurements of the adipokines, and therefore do not reflect the long-term exposure of these hormones. However, we provided unique simultaneous evaluation of several adipokines with anti-and pro-inflammatory properties in children consuming vegetarian and omnivorous diets. Some of them such as visfatin, vaspin, and omentin, were examined in vegetarians for the first time. Fourth, we did not measure inflammatory markers, such as CRP, IL-6, or TNF-alpha, but the studied children are generally healthy without acute or chronic infections. Finally, the cross-sectional nature of our study does not permit for causality statements. However, we believe that our results may create basic knowledge for further studies evaluating the possible role of diet on the adipokine profile, inflammation, and metabolic abnormalities.

Conclusions
The results of our study suggest that the vegetarian diet affects the adipokine profile, and it may have a protective effect on inflammatory status. Higher ratios of anti-to pro-inflammatory adipokines, expressed as the adipectin/leptin and the omentin/leptin ratios, may suggest a better metabolic panel of adipokines in children consuming the vegetarian diet. Further studies with larger samples are necessary to elucidate the specific mechanisms linking dietary patterns and adipokine levels with metabolic and inflammatory markers.