Prevalence of Metabolic Syndrome in Brazilian Adults: A Systematic Review and Meta- Analysis of Published Studies

A cluster of interconnected cardiometabolic risk factors characterizes metabolic Syndrome (MetS). The prevalence of MetS is increasing worldwide, but there is not a meta-analysis of this prevalence in the Brazilian population. We aimed to determine the prevalence of metabolic syndrome among adult general population in Brazil through a meta ‐ analysis study. Original research studies were searched at PubMed, Scopus, Web of Science, and SciELO databases, from 2011 to 2021. We used the Joanna Briggs Institute tool to assess the quality of included studies. The random effect model was used to estimate the pooled prevalence of MetS. Subgroup and meta-regression analysis were conducted for explored heterogeneity and used the Funnel Plot to assess publication bias. The study was performed based on the criteria of Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). and many appropriate

Syndrome X or MetS, and''prevalence,'' and ''Brazil.'' No language restrictions were imposed. A manual review of the reference lists in each identi ed study was also conducted.

Study selection
The search was performed independently by three authors (LTSV, LSBS, and VASJ). This reviewers independently identi ed potentially eligible articles by performing an initial screen of titles and abstracts. All potentially relevant titles and abstracts were selected for full text examination. Any discrepancies among the reviewers were resolved through consensus. Then, the following inclusion criteria were applied: (1) original type studies (e.g., cohort study, cross sectional studies); (2) studies that were conducted among 18 years of age or older and reportedly healthy individuals of both sexes; (3) There were no restrictions geographic region (urban, rural) and (4) to de ne MetS, studies that used any de ned criteria to determine the prevalence of MetS.
The exclusion criteria for our study were as follows: (1) the reviews and letters to the editors, (2) studies that used animal models or in vitro, (3) studies performed outside of Brazil, (4) the study population comprising individuals who were reported to have other health complications, (5) studies with incomplete information (6) or in a speci c population.

Data extraction
The three investigators extracted the data independently. The following information collected from each study was: rst author's name, year of publication, gender, age range, city and region of study and area in which the study was carried out, population, study design, criteria for diagnosis of metabolic syndrome, and the prevalence of metabolic of syndrome and its components.

Quality of studies
Study quality was assessed independently and blindly by three reviewers using the Joanna Briggs Institute tool for cross-sectional studies (JBI Critical Appraisal Checklist for Analytical Cross Sectional Studies) (21). This tool consists of a checklist of nine items, which determine the adequacy of the inclusion criteria; sample description; were study participants recruited in an appropriate way, was the sample size adequate, were the study subjects and setting described in detail, su cient coverage of the identi ed sample, standardization of diagnostic criteria, reliability and validity of the results, use of adequate statistical analysis, and response rate adequate. The answer options were yes, no, unclear and not applicable. The divergences in the analysis were resolved by consensus.

Statistical analysis
The meta-analysis was performed using R software (R Foundation for Statistical Computing, Vienna, Austria, URL http://www.R-project.org, 2020). The prevalence of MetS reported in the selected studies among healthy Brasilian adult populations was analyzed based on different diagnostic criteria used. In each study, we extracted the total number of participants and the number of individuals with the outcome. If one of these data was not provided by the article, we obtained this value through the prevalence of metabolic syndrome.
We used random effect models to calculate pooled prevalence and 95% con dence intervals. Inter-study heterogeneity was explored quantitatively using Cochran's Q and I 2 tests (22). In this regard, an I 2 of 50 % and 75 % indicated substantial and considerable heterogeneity, respectively. We used the xed effect for I 2 < 50% (low heterogeneity). We explored sources of heterogeneity by comparing MetS prevalence across subgroups de ned by several study-level characteristics and meta-regression analyses. We assessed the presence of publication bias graphically using the funnel plot.

Results
The ow of the literature search is shown in Figure 1. An initial search of the electronic databases identi ed 1598 records. Overall, 1560 records were excluded that did not meet the inclusion criteria. Therefore, 38 studies were assessed for eligibility through full-text reading. Of these, 12 studies were excluded for consisting of speci c population. Finally, 26 studies were selected for systematic review and meta-analysis.

Subgroup analysis
Subgroup analysis based on criteria used to de ne metabolic syndrome

Meta-regression analyses
To assess the sources of heterogeneity, we performed a meta-regression. In these analyses, age and year of implementation variables were not signi cantly associated with heterogeneity (p = 0.73, p = 0.62, respectively).

Analysis of quality of studies
The quality of the studies was assessed according to the set of criteria based on JBI guidance and are summarized in Table 2. A set of nine criteria was used to assess the quality of the studies. The sample frame was appropriate to address a target population in almost all articles with one exception (32). Fourteen study participants were sampled appropriately (6, 7, 24, 25, 28-31, 33, 34, 36, 40, 44, 45). The sample size was adequate in 19 studies (6, 7, 23, 24, 26-29, 31-33, 37, 38, 41-46). Study subjects and setting was described in detail in all articles. The data analysis was conducted with su cient coverage of the identi ed sample in 77% studies (6, 7, 23, 24, 26, 27, 29, 31-33, 35, 37, 39-46). Valid methods were used of identify of the condition in almost all articles with one exception (46). The condition was measured in a standard and reliable way for all participants and there was an appropriate statistical analysis in all the studies. The response rate was adequate and, if not, the low response rate was adequately managed in almost all articles with two exceptions (31,44).

Discussion
We have conducted this review including studies performed in the last decade to obtain a comprehensive estimate of burden of MetS in Brazilian adult population. In total, we analysed data from 26 studies that involved 84,522 participants. We have also captured the gender distribution, habitat differences, geographical region, criteria used to de ne metabolic syndrome, age of study participants and year of the study implementation estimates to nd any signi cant difference in the estimates of MetS.
Our meta-analysis revealed that the pooled estimate of MetS prevalence among subjects in Brazil was 33%. This estimate was higher than the prevalence of 29.6% observed in Brazil in 2013 and approached the worldwide prevalence of 20-25 % (3,14). The prevalence was also higher than that found in Malaysia Environmental factors related to lifestyle, such as physical inactivity, unbalanced food and stress and are closely linked with higher prevalence of obesity and especially for the accumulation of adipose tissue in the abdominal region, tissue directly involved in the genesis of insulin resistance, which is a possible connection with MetS. The decrease in insulin action in tissues, such as adipose tissue, leads to an increase in the in ammatory process, which induces this resistance. As a consequence, the accumulation of visceral adipose tissue in the body generating a high-risk cardiometabolic condition (56). In addition, insulin promotes renal sodium reabsorption and, in hyperinsulinemic conditions, an exacerbation of this action is expected. In fact, comparing individuals with and without MetS, it was observed that patients with the syndrome had signi cantly greater proximal sodium reabsorption, which can cause hypertension (57).
Study quality assessment shows that in many studies participants were not sampled in an appropriate way and the sample size was inadequate, which is a concern. Furthermore, some studies did not present su cient coverage of the identi ed sample for data analysis. These criteria for evaluating the quality of studies demonstrate that some studies may have publication bias, which corroborates with evidente asymmetry on the funnel plot.
We observed considerable heterogeneity among the included studies to estimate the prevalence of MetS in the Brazilian adult population. Prevalence of metabolic syndrome was the same in males and females, remaining with high heterogeneity. The wide variation in the prevalence of MetS among populations in Brazil can be attributed to heterogeneity among the included studies. The country, in addition to being continental in size, has great epidemiology, demographic and socio-economic variability and multicultural characteristics, which makes the population very diverse, making it di cult to generalize the ndings of this study in Brazil.
The subgroup analysis based on habitat, geographical region, criteria, age and year of study implementation was conducted in order to try to overcome this limitation. However, heterogeneity remained even after theses subgroup analysis. Hence, we tried to explain the between-study variability using metaregression and found the potential sources of heterogeneity. However, meta-regression analyses did not indicate enough factors to explain the observed heterogeneity. We suggest that other factors such as lifestyle, alcohol and tobacco consumption, stress, diet and physical inactivity may in uence MetS heterogeneity. Furthermore, the small number of studies in some regions of Brazil did not allow for a more robust analysis of the prevalence in these areas.
Other studies that assessed the prevalence of MetS in different countries also observed high heterogeneity among their data. Meta-analyses performed with data from the general population of Bangladesh (19), Iran (53), China (58), Middle East (59) and Mexico (18) showed heterogeneity greater than 90%. The study carried out in Bangladesh identi ed that the main source of heterogeneity was the geographical area of the population. In the study conducted in China, the age of participants was associated with lack of homogeneity. In Mexico, the diagnostic criteria used were signi cantly associated with the heterogeneity.
However, as in our work, the studies carried out in Iran and the Middle East, after performing analyzes by subgroups such as habitat, genus and diagnostic criteria, it was not possible to identify the source of this heterogeneity.
Like other studies, this our systematic review and meta-analysis study has some limitations, like there is no uniformity of metabolic syndrome de nitions, age groups, waist circumference and hyperglycemia cut-offs, and study settings in the studies included in the present review, resulting in limitations in comparability. Furthermore, we could not estimate the role of important risk factors on MetS such as physical activity and diet, since the studies included had not measured the effects of these factors. This review, we conduct some subgroup analyzes with limited data, such as MetS prevalence based on age of participants, because many included studies did not present this information.
The major strength of the study is that we have tried to provide the rst review with metanalisys on burden of MetS among adult population in Brazil. In addition, the strength is the comprehensiveness of the process, which included a search of four different databases, well-de ned inclusion/exclusion criteria, and extensive use of reference lists.

Conclusion
This systematic review and meta-analysis evaluated the scienti c literature on the prevalence of metabolic syndrome in Brazil. Our review indicates a high prevalence of MetS in the healthy Brazilian adult population, when compared to numerous countries and with a world estimate. Furthermore, the high prevalence remained when we subdivided the data according to different criteria, such as diagnostic, gender, age and geographic area of subjects studied,

Declarations
Ethics approval and consent to participate Not Applicable.

Consent for publication
Not Applicable.
Availability of data and materials The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests  Forest plot of prevalence of metabolic syndrome in Brazilian population. * prevalence according to the JIS criteria, ** prevalence according to the IDF criteria, *** prevalence according to the modi ed NCEP_ATPIII criteria and **** prevalence according to the NCEP-ATPIII criteria.

Figure 4
Forest plot of prevalence of metabolic syndrome in adult females in Brazilian population.

Figure 9
Forest plot of prevalence of metabolic syndrome according year of study implementation in Brazilian population.