PDF
Citation
Print
INTRODUCTION
Type 2 diabetes mellitus (T2DM) has become the most common chronic disease worldwide; the number of cases and the significance of T2DM both continue to rise as economic development and urbanization lead to changing lifestyles characterized by reduced physical activity and increased obesity [1]. Previous research has shown that T2DM increases the risk of many diseases, such as cardiovascular diseases, some cancers, and eye and kidney diseases [2]. In particular, lipid metabolism disorders are a common but severe complication of T2DM that could further lead to the development of atherosclerosis [3]. Extensive evidence has demonstrated that inflammation could play a vital role in the pathogenesis of T2DM [4,5]. Serum amyloid A (SAA) has been identified as a protein involved in acute-phase inflammation in humans and many mammals, and it has been extensively studied [6,7].
SAA regularly acts as an allele cluster on chromosome 11 in humans and is mainly synthesized in white adipose tissues and hepatocytes [7], and is an essential component of apolipoprotein and high-density lipoprotein (HDL) particles [8]. Increasing evidence indicates that high SAA levels in vitro have a number of adverse effects, including mediating the binding of HDL to differentiated macrophages and endothelial cells [9] and impairing the capacity of HDL to promote cholesterol efflux from macrophages, which could potentially promote dyslipidemia and even atherosclerosis [10]. Free SAA in vitro also has been shown to induce expression of extracellular matrix-degrading metalloproteinases and promote chemotaxis and adhesion of both monocytes and T lymphocytes [11,12].
Studies have documented that high levels of SAA are independently associated with a series of chronic diseases, including obesity, rheumatoid arthritis, coronary heart disease, and cancer [7,13-15]. However, consistent findings regarding the relationship between SAA levels and T2DM have not been published. The aim of this study was to synthesize a large number of reports about SAA levels and T2DM to further elucidate their relationship.
METHODS
The present analysis was performed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and was registered on the PROSPERO registry for systematic reviews (registration number: CRD 42020203008).
Data sources and searches
We searched the following databases from inception up to August 2021 for studies reporting on relationships between SAA and T2DM: PubMed, Cochrane Library, Embase, Web of Science, and MEDLINE. The keywords consisted of “serum amyloid A” OR “SAA” OR “amyloid protein AA” OR “amyloid A protein-related serum component” OR “amyloid protein AA precursor” OR “serum A related protein” AND “type 2 diabetes” OR “type 2 diabetes mellitus” OR “T2DM” OR “Non-insulin dependent diabetes mellitus.” All search results were screened by title and abstract, and potentially relevant full text articles were retrieved and assessed. Reference lists were also searched for additional relevant publications.
Inclusion and exclusion criteria
Studies were selected for the meta-analysis according to whether they met the following criteria: (1) the study investigated the potential association between SAA levels and T2DM; (2) complete data were available on SAA concentrations in both patients and controls or the correlations of SAA with some factors; (3) the study was a cross-sectional or case-control study; and (4) the language of publication was English. If an article had been published more than one time, we only used the latest edition. Articles were excluded based on the following criteria: (1) review or commentary articles; (2) repeated publications or similar studies; (3) animal experiments rather than human research; and (4) studies with incomplete and/or unavailable data.
Quality assessment
The quality of studies was separately evaluated by two independent authors (T.L. and M.L.). The Newcastle-Ottawa Quality Assessment Scale was used to evaluate the quality of included observational studies [16]. Only those studies for which the majority of the questions were deemed satisfactory (i.e., with a score of 6 or higher) were categorized as being of high methodological quality.
Data extraction and data conversion
Two investigators (T.L. and J.Z.) filtered the abstracts and collected data independently. When divergences appeared, we strived to reach an agreement after panel discussion. The following characteristics of the selected articles are shown in Table 1: (1) publication details, including the first author and the publication year; (2) study design; (3) characteristics of the study population, such as sample size, age, sex, continent (country); (4) SAA detection method; (5) the SAA levels in cases and controls, or the correlations of SAA levels to body mass index (BMI), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), fasting plasma glucose (FPG), hemoglobin A1c (HbA1c), homeostasis model assessment for insulin resistance (HOMA-IR), C-reactive protein (CRP), and interleukin-6 (IL-6) in patients with T2DM.
Since SAA levels sometimes showed a skewed distribution, some studies reported the mean and standard deviation of SAA levels in the original data, while others reported the median and quartiles. Therefore, it was necessary to convert the original data into usable means and standard deviations prior to analysis. The interquartile ranges of SAA levels were converted to the mean and standard deviation in accordance with methods reported in the literature [17-20].
Statistical methods
All statistical analyses were carried out using Stata version 12.0 (StataCorp LLC, College Station, TX, USA). Standardized mean differences (SMDs) were used to explore SAA differences in SAA levels between T2DM and healthy control groups. Heterogeneity between studies was assessed with the I2 statistic (0%<I2 <50%, no heterogeneity; I2 >50%, large heterogeneity) [21]. If heterogeneity was significant (I2 >50%, P<0.05), a randomeffect model was adopted to pool the SMD values; otherwise, a fixed-effect model was used. Sensitivity analyses were performed to assess the stability of our results. In order to identify the sources of heterogeneity, subgroup and meta-regression analyses were performed by the sex, mean age, and BMI of participants; study design; SAA detection method; and the continents that participants were from.
Using the pooled statistical values, the Spearman correlation coefficient (r) was utilized to assess the correlations of SAA levels to BMI, TC, TG, HDL-C, LDL-C, FPG, HbA1c, HOMA-IR, CRP, and IL-6 levels in patients with T2DM. The Egger and Begg tests were used to detect potential publication bias. A P value less than 0.05 was set as the threshold for statistical significance.
RESULTS
Characteristics of included studies
A total of 1,407 articles were retrieved from PubMed, Cochrane Library, Embase, Web of Science, and MEDLINE, of which 782 articles were filtered out because they were editorials, letters, reviews, commentaries, interviews, or studies of non-human subjects. After carefully examining each article’s title, abstract, and full text, 587 studies were excluded due to being duplicates, having an irrelevant topic or incomplete data, or dealing with an undefined disease. Eventually, 21 articles with 1,780 patients and 2,070 healthy controls were included in this meta-analysis (Fig. 1). Of these, 18 studies contained data on SAA levels (mean± standard deviation) in cases and controls, while five studies reported the correlations of SAA levels with BMI, TC, TG, LDL-C, HDL-C, FPG, HbA1c, HOMA-IR, CRP, and IL-6 levels in T2DM.
The basic features of the articles included in this meta-analysis are displayed in Table 1. The studies were published between 1994 and 2017. Sixteen studies indicated that they included both men and women, one study contained only men, two studies included only women, and two studies did not indicate participants’ sex. Five studies were conducted in China (Asia), one in Germany (Europe), two in the USA (North America), one in Sweden (Europe), two each in Ireland (Europe), Brazil (South America), and Japan (Asia), one each in Netherlands (Europe), Finland (Europe), Italy (Europe), Turkey (Asia), and Spain (Europe), and one in multiple countries, including Australia, New Zealand, and Finland (Oceania and Europe). Eleven cross-sectional and 10 case-control studies were included in the current meta-analysis. SAA levels were detected using enzyme-linked immunosorbent assays in 16 studies and immunonephelometry in four studies, while one study did not provide this information. Nine of the studies included in the meta-analysis were considered to be good quality [22-30], while 12 studies were categorized as poor quality [31-42].
Meta-analyses of SAA and T2DM
A meta-analysis was conducted of differences in SAA levels between T2DM patients and healthy controls. The average SAA levels in T2DM patients were obviously higher than those in healthy individuals (SMD, 0.68; 95% confidence interval [CI], 0.39 to 0.98), although a forest plot visually conveyed significant heterogeneity (I2=94.4%, P<0.001) (Fig. 2). In order to identify the source of heterogeneity, subgroup analyses were conducted according to the sex, mean age, and BMI of participants; the study design, the SAA detection method, and the continents that participants were from (Table 2). The SAA levels of T2DM patient groups who had mean ages of 50 to 59 years or over 60 years were remarkably higher than those of healthy controls (SMD, 0.69; 95% CI, 0.24 to 1.15; and SMD, 0.50; 95% CI, 0.13 to 0.87, respectively). Markedly higher SAA levels in T2DM patients than in normal controls were found among participants from Europe (SMD, 0.33; 95% CI, 0.09 to 0.56), Asia (SMD, 1.04; 95% CI, 0.53 to 1.54), and South America (SMD, 0.37; 95% CI, 0.07 to 0.66). To further clarify the source of heterogeneity, meta-regression analysis was also performed. The results of meta-regression indicated that the sex, mean age and BMI of participants; study type; SAA detection method; and continents where participants were from could not explain the origin of heterogeneity (Table 3).
Sensitivity analyses also showed that removing studies individually did not substantially change the associations (Fig. 3). Therefore, the overall magnitude of differences in SAA levels between T2DM patients and healthy controls was stable.
Meta-analysis of correlations between SAA and several factors in T2DM
Next, a meta-analysis of the correlations between SAA levels and cardiometabolic risk factors, glucose metabolic homeostasis indexes, and inflammation factors was conducted to further explore the associations of SAA levels with T2DM (Fig. 4). The result revealed that, in T2DM patients, SAA levels were positively linked with the levels of BMI (r=0.34; 95% CI, 0.03 to 0.66) and TG (r=0.12; 95% CI, 0.01 to 0.24), but negatively associated with HDL-C (r=–0.23; 95% CI, –0.44 to –0.03) levels. Furthermore, the pooled analysis suggested that, in T2DM patients, SAA levels were positively associated with FPG (r=0.26; 95% CI, 0.07 to 0.45), HbA1c (r=0.24; 95% CI, 0.16 to 0.33), HOMA-IR (r=0.22; 95% CI, 0.10 to 0.34), CRP (r=0.77; 95% CI, 0.62 to 0.91), and IL-6 (r=0.42; 95% CI, 0.31 to 0.54) levels.
Publication biases
The Begg’s correlation test and the Egger’s regression test indicated no evidence of significant publication bias in these studies (all P>0.05) (Table 4).
DISCUSSION
The findings of the present meta-analysis indicated that high SAA levels showed statistically significant associations with the presence of T2DM, which is consistent with the studies of Ebtehaj et al. [22], Yang et al. [23], and Griffiths et al. [24]. More importantly, our study found that SAA levels were positively correlated with BMI, TG, FPG, HbA1c, HOMA-IR, CRP, and IL-6 levels in T2DM patients, suggesting that measuring SAA may be beneficial for assessing impaired lipid metabolism homeostasis and the inflammatory response in patients with T2DM. These findings enhance our knowledge of the usefulness of SAA as a marker of inflammation and may also provide evidence of a mechanistic link between inflammation and lipid metabolism dysregulation in T2DM patients.
SAA has been found to be expressed at high levels in vivo under conditions of decreased insulin sensitivity [43], which can activate the neutrophil nuclear factor κB, inhibitory factor, kinase β signaling pathway in the liver, skeletal muscle, and adipose tissue, inducing liver and systemic local insulin resistance [44,45]. Furthermore, long-term insulin resistance increases glycosylation end products that stimulate macrophages and monocytes to release more inflammatory factors, inducing a chronic low-grade inflammatory state in the body [44,45] and accelerating the progression of T2DM and its complications. In addition, animal evidence has demonstrated that the administration of recombinant A-SAA (the acute-phase isoform) in the adipocytes of mice resulted in the downregulation of genes that were critical for insulin sensitivity in the treated cells [46].
We observed that the average SAA levels in T2DM patients were notably higher than those in healthy controls. However, subgroup analyses showed that this difference was associated with the mean age of participants and the continent that they were from, with particularly strong relationships in studies with a mean age of more than 50 years and those from Europe, Asia, or South America. A previous study showed that age was tightly correlated with obesity and T2DM [47,48], and that SAA was associated with obesity [13]. Moreover, obesity was reported to increase the risk of T2DM in numerous studies [49,50]. However, previous studies placed less emphasis on differences in SAA levels between T2DM cases and normal controls in different continents. We assumed that environmental, genetic, and lifestyle-related factors may be responsible for these differences. However, a meta-regression model indicated that the source of heterogeneity could not be explained by the sex, mean age, and BMI of participants; study type; SAA detection method; or the continents that participants were from.
The meta-analysis of correlations found that SAA levels were positively linked with cardiometabolic risk factors (BMI and TG) in T2DM patients, which is in line with previous studies [13,51-53]. Elevated FPG is the most common indicator of T2DM, and HbA1c and insulin resistance are the two key indicators in the pathogenesis of T2DM. This meta-analysis found that SAA levels were positively associated with FPG, HbA1c, and insulin resistance, indicating a link between SAA levels and the occurrence and development of T2DM. These data further support the association of SAA levels with T2DM. SAA is an acute-phase response protein and can serve as a sensitive marker of the acute inflammatory response [15]. Its acute-phase isoform (A-SAA) is up-regulated up to 1,000-fold in response to inflammatory stimuli such as trauma, infection, injury, and stress [54], thereby stimulating the production of CRP, IL-6, and other inflammatory mediators [13], as exhibited in our study.
This study has some strengths. First, the meta-analysis was conducted based on a comprehensive search of the literature and addressed concerns about professionally determined SAA levels and T2DM. Second, the sensitivity analysis revealed that excluding any study did not affect the pooled effect of differences in SAA levels between T2DM patients and healthy controls, suggesting that our results are robust. Third, considering some potential confounding factors, the meta-analysis was stratified by age, sex, BMI, and other variables, which could increase the stability and accuracy of the results. Fourth, the findings of a systematic review and meta-analysis using secondary data are valuable and can provide new directions for future studies. Direct measurements of SAA will be necessary to demonstrate whether it may be of pathological significance in T2DM.
Nonetheless, this article is also subject to some unavoidable limitations. Foremost, correlation does not mean causation in terms of T2DM. Second, the large amount of heterogeneity was a weakness of the present study, although it could be explained by various subgroup analyses. Therefore, caution in interpreting these results is necessary for research focusing on these aspects. Third, a small number of participants may experience worsened symptoms of T2DM, and these outliers are not well represented in sample means. Fourth, participants’ age, sex, and BMI were inconsistent in the reported articles. These factors could not be adjusted to calculate the pooled effect due to the limited data. Finally, the specific mechanism underlying the relationship between SAA levels and T2DM must be validated through animal experiments in the future.
In summary, high SAA levels were found to show statistically significant associations with the presence of T2DM. Furthermore, these findings demonstrate that SAA may have an impact on the development of T2DM by elevating the levels of BMI, TG, FPG, HbA1c, HOMA-IR, CRP, and IL-6, or by decreasing HDL-C levels. However, firm confirmation of the role played by SAA in T2DM will require direct testing in both transgenic and SAA-deficient animal models.
XML Download