Introduction

Advanced glycation end-products (AGEs) are a group of compounds formed by non-enzymatic glycation of lipids, proteins, and nucleic acids [1, 2]. AGEs have two endogenous and exogenous sources, while the first is formed during normal metabolism in body and the second is derived from foods or tobacco smoke [3]. AGEs are potential ingredients that lead to oxidative stress and chronic inflammation; they act through binding to their receptors like receptor for advances glycation end products (RAGE) [4]. RAGE is present in numerous differentiated adult cell types, including immune and endothelial, and is a multi-ligand receptor belonging to immunoglobulin superfamily [5,6,7]. AGE binding to its receptor, RAGE, triggers the cell signaling pathways’ activation through p38 and p44/42 MAP kinase, or nuclear factor kappa- B (NF-κB), and leads to production of reactive oxygen species (ROSs) and pro-inflammatory cytokines production [5, 8]. RAGE is expressed on numerous cell types, including vascular cells, adipocytes, podocytes, immune cells, neurons, cardio-myocytes, and lung epithelial cells [9,10,11,12]. RAGE has numerous ligands other than AGEs that can bind to either extracellular V-type immunoglobulin (Ig) domain with numerous special and distinct binding sites or to extracellular C1 and C2-type Ig domains; these further confirm the complexity of RAGE-ligand interactions [13,14,15].

Soluble receptor for advanced glycation end products (sRAGE) are a particular form of RAGE found in plasma and other fluids of the body, such as synovial fluid and cerebrospinal fluids [16]. There are two major forms of sRAGE; most of the circulating sRAGE results from cell surface-cleavage of the full-length receptor by matrix metalloproteinases (MMPs) [17, 18] and the other less prevalent form of sRAGE, known as endogenous secretory or esRAGE, is a product of a splice variant of AGE receptor (AGER) [19]. Numerous evidence are available about the pathogenic role of AGE and RAGE in inducing inflammation, oxidative stress, adipocyte hypertrophy and expansion, as well as ectopic lipid accumulation in different organs [20,21,22]. On the other hand, it is suggested that sRAGE, working as a decoy receptor, can bind to RAGE ligands and prevent membrane RAGE activation and associated detrimental health effects [23, 24]. Several studies have investigated the positive effects of sRAGE in prevention of obesogenic effects of AGE and its receptor RAGE. Dozio E et al. [24] reported lower sRAGE concentrations in women with obesity versus women with normal weight (P < 0.05). In another study by Zaki M et al. [25], similar finding was reported. Some other studies also reported positive associations between BMI and sRAGE concentrations [26,27,28,29]. However, some other studies reported no significant association between BMI and sRAGE [30, 31]. Since the exact association between obesity and sRAGE is not clear, a critical analysis can help to better identify this association. Accordingly, we aimed to quantify and critically review the results of studies reporting the associations between sRAGE with central or general obesity indices in general adult populations.

Methods and materials

To report the results, we used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Sup. Table 1) [32].

Eligibility criteria

In the current systematic review and meta-analysis, inclusion criteria were as follows: (1) observational studies with cross-sectional design (2); studies evaluating the relationship between sRAGE and obesity; (3) studies including measurements of general or central obesity including body mass index (BMI), fat mass, waist circumference (WC), or waist to hip ratio (WHR); (4) studies conducted among apparently healthy adults, with obesity or obesity-related disorders including diabetes, cardiovascular diseases, metabolic syndrome, and aged over 18 years; and (5) studies that provided the mean ± standard deviation (SD) of BMI, fat mass, WC, or WHR of those in the lowest versus highest categories of sRAGE. Interventional trials, case-reports and case-series, animal models and in vitro studies, reviews, congress, seminars, letter to editors and short communications were excluded. The PECO (patients, exposure, control/comparator and outcome) model for selecting the studies is presented in Table 1. This model is one of the most widely used models for formulating clinical questions.

Table 1 The PECO criteria used for the systematic review
Full size table

Information sources, search strategy and selection process

Our systematic search in electronic databases, including Scopus, PubMed and Embase up to 26 October, 2023 yielded a total of 21,612 articles. The search was limited only to English language articles. No missing document was found through hand-searching the reference lists of the papers. The search strategy was created with a combination of the MeSH (Medical Subject Headings) (Sup. Table 2). Two independent reviewers evalutaed the articles for meeting the inclusion criteria. Any problem was resolved by a third reviewer in case of inconsistencies.

Data collection and extraction process

Four independent researchers performed data extraction, of the following information, including the name of first author, journal, country, publication year, demographic information of participants (e.g., age, gender distribution, and percent of male participants), baseline BMI and WC, study design, sample size, adjusted covariates, study setting, sRAGE measurement tools, and main findings.

Risk of bias assessment

The methodological quality of included studies for risk of bias assessment were assessed using the Agency for Healthcare Research and Quality (AHRQ) checklist [33, 34].

Synthesis methods

Data analysis was performed by STATA version 16 (STATA Corp, College Station, TX, USA). P-values less than 0.05 were considered as statistically significant. The studies that reported the comparison of BMI or WC [mean (SD)] in those with highest versus lowest sRAGE categories were evaluated. Therefore, the mean and SD of variable was used to calculate weighted mean difference (WMD) with 95% confidence interval (CI). When the median and range were reported and mean and SD were not available, the median values were considered as the best estimate of mean and the SD was calculated as the Hozo et al method as below: \({SD}^{2}\approx \left(\frac{1}{12}\right(\frac{{\left(a-2m+b\right)}^{2}}{4}+{\left(b-a\right)}^{2}\)) where “SD” is standard deviation, “a” and “b” are upper and lower limits of range, and “m” is the median value [35, 36]. For missing SDs, the method of Walter and Yao was used [37]. Cochran’s Q and I2 tests were used for heterogeneity measurements; for statistically significant heterogeneities (e.g., either P value for Q statistics of less than 0.1 or I2 greater than 50%), random effects model we used [38,39,40]. To identify the source of heterogeneity, subgrouping and meta-regression approaches were performed. Begg’s Funnel plots, Begg’s correlation coefficient and Egger’s asymmetry tests were used for assessment of publication bias.

Results

Study selection

Our search results yielded a total of 21,612 articles that were imported into the EndNote software. In the first phase, 12,704 articles were removed due to duplication. Next, 8,192 articles were removed after screening the title and abstract due to not meeting inclusion criteria, other designs and age groups, and being seminars, congresses, and review articles. Finally, a total of 716 articles remained for full-text evaluation by two independent reserachers. Consequently, 13 manuscripts were included in the final meta-synthesis. The study’s selection flowchart is presented in Fig. 1.

Fig. 1
figure 1

Study flowchart

Full size image

Study characteristics

In the meta-analysis of the comparison of BMI between highest versus lowest sRAGE categories, six individual studies with a total number of 1,865 participants were included. General characteristics of these studies are shown in Table 2. Two out of eight studies were performed among obese, diabetic patients [31, 41] and the six other studies [25,26,27,28, 30, 42] were performed among apparently healthy population. The age range of study participants was 18–83 years old and BMI was 18–42 kg/m2. In the study by Momma H (2014) [26], BMI was significantly lower in the highest versus the lowest category of sRAGE among general Japanese adult population (22.7 versus 24.5 kg/m2; P < 0.001). Similarly, in another study by the same author [27], BMI was significantly lower in the highest versus the lowest category of esRAGE (21.9 versus 23.3 kg/m2; P < 0.001). Other studies by Moriya S et al. in general healthy population of Japan [28], by Moy KA et al. in Finland [30], and by Zaki M et al. in Egypt [25] found similar results. Similarly, in the meta-analysis of the comparison of WC between highest versus lowest sRAGE categories, four individual studies with 1,876 participants from the USA [42], Japan [26, 27], and Egypt [25], WC was significantly lower in the highest versus the lowest category of sRAGE in all the studies. The mean sRAGE concentration in serum was compared between individuals with obesity versus individuals without obesity in three individual studies with 165 participants [31, 41]; the individuals with obesity had lower mean sRAGE compared to individuals without obesity.

Table 2 Characteristics of the studies included in the systematic review
Full size table

Risk of bias in the included studies

The results of risk of bias assessment is provided in Table 3 by AHRQ checklist [33]. As shown in this Table, four out of eight studies had moderate quality and the remaining four studies had high quality. The lowest quality score was 5 and the highest quality score was 9.

Table 3 Agency for Healthcare Research and Quality (AHRQ) checklist to assess quality of the cross-sectional studies
Full size table

Results of synthesis

The results of meta-analysis (Fig. 2) showed that being at the highest category of sRAGE with median values of 934.92 pg/ml of sRAGE, was associated with lower BMI among apparently healthy adults (WMD: -1.927; CI: -2.868, -0.986; P < 0.001) compared with those at the lowest category of sRAGE concentration with median values of 481.88 pg/ml. Similarly, those at the highest sRAGE category with the median values of 1302.3 pg/ml sRAGE had about 6 cm lower WC compared with those at the lowest category of sRAGE concentration with median values of 500.525 pg/ml (WMD: -5.602; CI: -8.820, -2.383; P < 0.001; Fig. 3). Comparing the sRAGE concentrations in individuals with obesity versus individuals without obesity, individuals with obesity had significantly lower circulating sRAGE concentrations (WMD: -135.105; CI: -256.491, -13.72; P = 0.029; Fig. 4). Subgrouping results of the comparison of BMI between the highest versus the lowest category of sRAGE are presented in Table 4. In studies performed in Japan that had baseline BMI of lower than 23 kg/m2, initial heterogeneity reduced from 86.4 to 0. Therefore, country and baseline BMI might be the sources of heterogeneity. The results of subgrouping for the comparison of WC between sRAGE categories are shown in Table 5. Accordingly, country might be a heterogeneity source. However, interpretation of results is challenging due to the low number of studies in each subgroup. The results of meta-regression (Tables 6 and 7) also confirmed these findings. The results of funnel plots (Sup. Figures 1, 2) and Begg’s adjusted rank correlation and Egger’s regression asymmetry tests showed no publication bias (for BMI; Egger P-value: 0.224; Begg’s P-value: 0.851 and for WC: Egger P-value: 0.497; Begg’s P-value: 0.297).

Fig. 2
figure 2

Weighted mean difference (WMD) with 95% confidence interval (CI) of body mass index (BMI) in the highest versus the lowest soluble receptor for advanced glycation end products (sRAGE) categories. I2 represents the degree of heterogeneity

Full size image
Fig. 3
figure 3

Weighted mean difference (WMD) with 95% confidence interval (CI) of waist circumference (WC) in the highest versus the lowest soluble receptor for advanced glycation end products (sRAGE) categories. I2 represents the degree of heterogeneity

Full size image
Fig. 4
figure 4

Weighted mean difference (WMD) with 95% confidence interval (CI) of circulating soluble receptor for advanced glycation end products (sRAGE) concentrations in individuals with obesity versus individuals without obesity. I2 represents the degree of heterogeneity

Full size image
Table 4 Subgroup analysis for the comparison of BMI between the highest versus the lowest category of sRAGE
Full size table
Table 5 Subgroup analysis for the comparison of WC between the highest versus the lowest category of sRAGE
Full size table
Table 6 Meta regression analysis for finding the possible sources of heterogeneity for the association between sRAGEs and BMI
Full size table
Table 7 Meta regression analysis for finding the possible sources of heterogeneity for the association between sRAGEs and WC
Full size table

Discussion

In the current meta-analysis, higher sRAGE concentrations was associated with lower BMI among 1,865 apparently healthy individuals. Also, lower WC was accompanied with higher sRAGE concentrations among 1,876 adults. This is the first meta-analysis evaluating the association between sRAGE and obesity indices, and reporting obesity-reducing effects of sRAGE among adults. There are several underlying mechanisms suggested for obesity prevention by sRAGE, such as protection against obesity-induced lipid accumulation by preventing RAGE hyper-expression [43]. Also, it is suggested that higher AGE flux in individuals with obesity reduces sRAGE levels; it is well-known that sRAGE has other ligands besides AGEs that could decrease it. So, a lower sRAGE among individuals with obesity is a reflection of greater binding of AGEs to its ligands [44, 45]. Since numerous studies revealed lower circulating AGEs concentrations in individuals with obesity, one interesting finding is that, possibly, CML concentrations are not a good biomarker of its status in individuals with obesity, and further studies are needed to confirm whether it is the case for different AGE compounds [46,47,48,49]. This might be attributed to the fact that the most preferable AGEs measured in the studies is CML, and it is suggested that circulating CML concentrations is in inverse association with body fat storage, because it deposits in fat tissue and by fat mass expansion (like in individuals with obesity) its deposition in fat mass increases and lowers its circulating amount [48, 50, 51]. As a result, circulating sRAGE values would be a better marker of the AGE-RAGE interaction in the body and its lower concentrations is a reflection of detrimental effects of this interaction. Impaired adipocyte function is also suggested as possible mechanism of increased central obesity in lower sRAGE concentrations [26, 52]. The possible underlying mechanisms of the effects of AGE-RAGE interaction in developing obesity and the protective role of sRAGE is presented in Fig. 5.

Fig. 5
figure 5

The accumulation of AGEs in adipocytes may explain the inverse association between AGE levels and BMI; high-AGE diets are associated with greater fat consumption and an increased risk of abdominal obesity [55]. RAGE activation also raises circulating inflammatory cytokines while decreasing anti-inflammatory adipokines like adiponectin. IL-6 and leptin are recognized as inflammatory mediators in the presence of obesity [56]. Moreover, adipocyte hypertrophy is associated with enhanced RAGE expression [57]. However, the involvement of sRAGE in the prevention of age-related muscle mass loss must be addressed [58]. As a decoy receptor for AGEs and other inflammatory ligands, sRAGE attributed to the reduced ligand-mediated disruption, preventing AGEs from binding to the cell-bound full-length receptor RAGE [58] and inhibiting the obesogenic effects of AGE-RAGE interaction by diminishing AGES accumulation visceral fat, regulating leptin and Adiponectin release [56], reducing Insulin resistance [58]. AGEs = advanced glycation end products; RAGE = receptor for advanced glycation end products; sRAGE = soluble receptor for advanced glycation end products; IL-1 = interleukin-1; IL-6 = interleukin-6; TNF-α = Tumor Necrosis Factor alpha; BMI = Body mass index; WC = Waist Circumference

Full size image

In our study, all the included studies had a moderate or high quality, and no study had poor quality. In the subgrouping, geographical area and baseline BMI were the possible sources of heterogeneity. However, the current meta-analysis has some limitations. First, causal inference was not possible due to cross-sectional design of the included studies. Second, due to the low number of studies in each subgroup, making a reliable conclusion is not possible. Third, as the adipose tissue differs between men and women [53, 54], we needed separate data in this regard; however, there was no separate data to evaluate the association of sRAGE levels with BMI or WC by gender.

As a conclusion, in the current systematic review and meta-analysis, for the first time, we quantified and critically reviewed the studies that evaluated the association between general or central obesity indices with circulating sRAGE levels. There was a negative association between BMI, WC, and circulating sRAGE concentrations among adults. Further studies are warranted to confirm our results.