Background

With the global economy thriving, the Western diet and the sedentary living habits have been disseminated, and physical labor has been largely reduced. As a result, the prevalence of Mets has increased drastically in recent years [1]. The 2010–2012 Chinese National Nutrition and Health Survey suggested that the general prevalence of Mets has reached 24.2%, including 24.6% for men and 23.8% for women [2]. Mets is highly concerned as it doubles the risk of cardiovascular disease, while the all-cause mortality rate for Mets patients increases by 1.5 times [3]. Controlling the incidence of Mets and reducing complications is world-wide urgent. However, current Mets guidelines recommend treatments mainly based on lifestyle interventions, including smoking cessation, Mediterranean diet, 30–60 min of physical exercise per day, and a minimum of 5% weight-loss goal for obese patients [4]. Though the molecular-targeted drug therapy has been implemented for a variety of diseases, there is no specific drug for Mets treatment yet [4]. New options are urgently required for Mets, especially in the field of gene-targeted therapy.

Mets is a heterogenic and multifactorial diagnosis. The whole-gene linkage analysis is failed to identify loci that correspond to functional genes [5]. The Genome-Wide Association Study (GWAS) has inherent limitations on the minimum frequency of SNP, and the total effect of identified loci only explains a small proportion of Mets prevalence [5]. Under these circumstances, the SNP study is still the most important method for Mets gene research. It has been found that over 870 SNPs are associated with obesity [6], 477 SNPs with lipid metabolism [7], more than 200 SNPs with blood pressure [8], and around 250 SNPs with glucose tolerance [9]. However, most of these gene loci were related to single Mets components, among which the lipid metabolism-associated SNPs showed the strongest relevance to Mets [10]. Furthermore, current researches have been inadequate in gene-environmental interaction study. We expect to find a locus that participates in multiple components of Mets, and to carry out research on the interaction of gene polymorphism and environments.

As universally acknowledged, lipid metabolism, especially triglyceride (TG) metabolism plays a central role in Mets pathogenesis [11]. TG elevation is one of the components of Mets and a risk factor for abdominal obesity [12], and the synthesis of TG is associated with glucose metabolism through the tricarboxylic acid cycle [13]. Consequently, the gene locus featured by TG regulation is expected to become the target of Mets gene-directed therapy. Lipoprotein lipase (LPL) regulates TG hydrolysis in circulation and in adipose tissue. Angiopoietin-like protein 4 (ANGPTL4) is characterized by a reversible inhibitor of LPL [14]. It has been reported that ANGPTL4 is not only involved in the regulation of blood lipids [14], but also blood pressure [15, 16], glucose tolerance [17,18,Mets interventions

Participants with increased body mass index (BMI), hypertension, diabetes, dyslipidemia, and other conditions related to Mets and cardiovascular risks were suggested to have weight management and lifestyle interventions, which include reducing dietary sugar and fat, having regular and moderate physical exercise, tobacco and alcohol cessation, visiting physician regularly. Prescription drugs including antihypertensives, hypoglycemics, and statins were also recommended as needed. The out-come of weight management was measured by changes in body weight (ΔWeight), and participants in the follow-up survey were subdivided into weight loss group (ΔWeight < 0) and weight gain (ΔWeight ≥ 0) group accordingly.

Statistics

The χ2 goodness-of-fit test is used to test the Hardy–Weinberg equilibrium at rs1044250 locus. Continuous variables are presented in \(\overline{{\text{x}}} \pm {\text{S}}\) and compared by independent-samples t test or one-way analysis of variances. Categorical variables are presented in proportions and compared by χ2 test, χ2 is calibrated by Bonferroni when the minimum sample size is lower than 5. Counting variables are presented in Median [Quartile1, Quartile3] (M [Q1, Q3]) and analyzed by Nonparametric test. Multi-factor logistic regression is used to analyze the risk factors for Mets and its components. The differentials in laboratory indicators after 5-year follow-up are calculated by subtracting the cross-section value from the follow-up value, differentials are recorded in "Δ". Multiple stepwise linear regression is used to analyze the effects of gene polymorphism on the number of Mets components and on changes of systolic blood pressure (SBP), diastolic blood pressure (DBP) and high-density lipid-c (HDL-c). Ordinal logistic regression is used to elucidate the impact of gene polymorphism on the number changes of Mets components. Crossover analysis is applied to analyze the interaction between two independent variables on Mets. All statistical analysis is operated using SPSS 26.0 (Chicago, Illinois SPSS). A two-tailed p value of less than 0.05 is considered to be statistically significant.

Results

Basic characteristics of the control and the Mets group

Participants in the control group (n = 1029) and the Mets group (n = 1018) are matched by gender and age, differences of the weight, WC, BMI, SBP, DBP, TG, total cholesterol (TC), low-density lipid-c (LDL-c), HDL-c, fasting plasma glucose (FPG) levels between the Control group and the Mets group are statistically significant (p < 0.001, Table 1). Frequencies of the C allele and the T allele on rs1044250 locus account for 92.1% and 7.9% respectively in the study population. Distribution of CC, CT, TT genotypes in the control group (χ2 = 0.02; p = 0.99), the Mets group (χ2 = 1.48; p = 0.48), and the study population (χ2 = 1.58; p = 0.45) are followed with Hardy–Weinberg genetic balance.

Table 1 Baseline characteristics of Control group and Mets group
Full size table

Risks of the SNP rs1044250 on Mets and the components of Mets

The laboratory indicators among patients with CC, CT and TT genotypes at ANGPTL4 gene rs1044250 locus are compared respectively, the result indicates that the baseline data among genotypes in the control group do not show difference (p > 0.05, Table 2), while in the Mets group, HDL-c is significantly higher in patients with TT genotype than patients with CC or CT genotypes (p < 0.05, Table 2).

Table 2 Baseline characteristics of various rs1044250 genotypes in the Control group and the Mets group
Full size table

The frequencies of rs1044250 genotypes are significantly different between the control group and the Mets group (χ2 = 25.556; p < 0.001, Table 3), and the frequency of T alleles in the control group is significantly lower than that in the Mets group (χ2 = 25.991; p < 0.001, Table 3). The study population is subdivided into groups respectively according to the presence or absence of each of the five components of Mets. The frequencies of CC, CT, TT genotypes as well as the frequencies of C allele and T allele are significantly different between the normal WC group and the increased WC group, statistical significances were also found between the normal TG group and the elevated TG group, the normal blood pressure group and the elevated blood pressure group, the normal FPG group and the elevated FPG group (p < 0.05, Table 3).

Table 3 Distribution of SNP rs1044250 genotypes and alleles in Mets and Mets components
Full size table

Multi-factor stepwise logistic regression analysis shows the SNP rs1044250 is an independent risk factor for metabolic syndrome (OR 1.875 [95% CI 1.363–2.580]; p < 0.001, Table 4). Accordingly, the five components of Mets are studied respectively, the SNP rs1044250 is an independent risk factor for increased WC (OR 1.618 [95% CI 1.119–2.340]; p = 0.011, Table 4) and increased blood pressure (OR 1.323 [95% CI 1.002–1.747]; p = 0.048, Table 4). In addition, patients with various rs1044250 genotypes are significantly different in the numbers of Mets components (M [Q1, Q3]: CC 2 [0, 3], CT 3 [0, 3], TT 3 [3, 3]; p = 0.001). Multivariate linear regression analysis confirms that the SNP rs1044250 (β = 0.044; p = 0.007), age (β = 0.068; p < 0.001), weight (β = 0.221; p < 0.001), BMI (β = 0.387; p < 0.001) and LDL-c (β = 0.200; p < 0.001) are independent risk factors for the increased number of Mets components carried by patients.

Table 4 Logistic regression analysis of Mets and Mets components
Full size table

Baseline characteristic changes of the study population

Compared with the control group, the weight, WC, BMI, SBP, DBP, TG, TC, LDL-c and FPG of the Mets group are significantly reduced in 5 years (Table 5). In the Mets group, HDL-c level is significantly elevated in patients with TT genotype than that in CC genotype or CT genotype (p < 0.05, Table 5). Linear regression analysis shows that the SNP rs1044250 (β = − 0.065; p = 0.017), Sex (β =  − 0.069; p = 0.012) and TG (β = − 0. 611; p < 0.001) are independent risk factors for HDL-c reduction in patients with Mets in the 5-year follow-up survey.

Table 5 Baseline characteristics changes in 5-year follow-up survey of various rs1044250 genotypes in the Control group and the Mets group
Full size table

Changes in the number of Mets components

The study population is subdivided into Weight loss group and Weight gain group, the distribution of rs1044250 genotypes (fisher = 1.162; p = 0.532) and T alleles (χ2 = 0.878; p = 0.349) in these groups do not show a significant difference. Among patients with CC, CT and TT genotypes at rs1044250 locus, changes of the number of Mets components do not show any difference (M [Q1, Q3]: CC 0 [− 1, 1], CT 0 [− 1, 1], TT 0 [− 1, 1]; p = 0.529, Table 6). However, subgroup analysis based on weight management indicates that the number changes of Mets components between CC genotype and CT or TT genotype in the weight gain group are significantly different (M [Q1, Q3]: CC 1 (0, 1), CT + TT 0 [− 1, 1]; p = 0.021, Table 6), while there is no difference showed in the weight loss group (M [Q1, Q3]: CC 0 [− 1, 1], CT + TT − 0 [− 1, 1], p = 0.732, Table 6). Indeed, under an ordinal regression model, the SNP rs1044250 (β = − 0.703; p = 0.024), TG (β = − 0.337; p < 0.001) and FPG (β = − 0.242; p = 0.003) are independent protective factors for the number of Mets components when the body weight is increased, which suggests that people with CC genotype are more likely to catch-up with the number of Mets component in the 5-year survey.

Table 6 The number changes of Mets components carried by various subgroups
Full size table

Interaction of SNP rs1044250 and weight management

In the weight loss group, the HDL-c level is significantly decreased in patients with TT genotype at rs1044250 locus compared with CC (p = 0.002, Table 7) and CT (p = 0.008, Table 7) genotype. In the weight gain group, the SBP (p = 0.002, Table 7) and DBP (Table 7; p = 0.004) of CC genotype are significantly higher than that of CT genotype. Accordingly, the interaction of SNP rs1044250 and weight management on SBP (F = 3.291; p = 0.038, Table 7), DBP (F = 3.026; p = 0.049, Table 7) and HDL-c (F = 6.269; p = 0.002, Table 7) are statistically significant.

Table 7 Factorial analysis SNP rs1044250 and weight management on clinical characteristic changes
Full size table

The independent variable ΔWeight*rs1044250 is included in the multi-factor stepwise linear regression equations, of which the dependent variable is ΔSBP and ΔDBP respectively. Thereafter, the interaction of SNP rs1044250 and weight management is an independent risk factor for elevated SBP (β = 0.075; p < 0.001, Table 8) and elevated DBP (β = 0.097; p < 0.001, Table 8). The independent variable ΔWeight*rs1044250 is excluded from the linear regression equations of ΔHDL-c (β = 0.004, p = 0.851, Table 8).

Table 8 Linear regressions analysis of ΔSBP, ΔDBP and ΔHDL-c
Full size table

A crossover analysis is conducted to study the interaction of ANGPTL4 gene SNP rs1044250 and weight management on ΔSBP, ΔDBP, ΔHDL-c under linear regression model. In the linear regression model with ΔSBP as the dependent variable, the independent variables SNP rs1044250 (β = 0.193; p < 0.001, Table 9) and rs1044250*Weight Management (β = − 0.093; p = 0.013, Table 9) are included in the regression equation, and the interaction between SNP rs1044250 and weight management on ΔSBP is negative (synergy index = 0.558); in the linear regression model with ΔDBP as the dependent variable, the SNP rs1044250 (β = 0.241; p < 0.001, Table 9) and rs1044250* Weight Management (β = − 0.078; p = 0.035, Table 9) are included in the regression equation, and the synergistic effect between SNP rs1044250 and weight management are negative (synergy index = 0.696).

Table 9 Crossover analysis of SNP rs1044250 and WM on ΔSBP, ΔDBP and ΔHDL-c
Full size table

Discussion

This is the first study that comprehensively identifies the ANGPTL4 gene SNP rs1044250 as an independent risk factor for Mets by increasing the WC and blood pressure. The number of Mets components in CC genotype individuals increases when body weight raised. Consistently the SNP rs1044250 and weight management are negatively correlated on the interaction with blood pressure.

In 2007, a GWAS study found that FTO polymorphism was associated with weight gain and increased BMI [34]. It was the first GWAS study which ushered in the era of gene SNP research on Mets. Since then, the effects of gene polymorphisms have been related to obesity [6], lipid metabolism [7], glucose tolerance and hypertension [8, 9]. Because of the heterogeneity and multifactorial nature of Mets, current SNP studies have been limited on single Mets components, the effects of multifunctional gene loci on Mets are seldom reported. Under this circumstance, various studies have suggested that ANGPTL4 is involved in the regulation of multiple components of Mets, including lipid metabolism [14], obesity [24], blood pressure and glucose tolerance [15,16,17]. Regarding the multifunctional feature, the major role of ANGPTL4 is regulating the TG content in circulation and maintaining the balance of adipopexis in WAT [35]. The N-terminal oligomerized ANGPTL4 in circulation inhibits LPL activity, therefore, increases the circulating TG level [2], while the SNP rs1044250 mainly alters the activity of ANGPTL4 C-terminus [36]. Previous study has reported that the SNP rs1044250 only accounts for 0.8% of patients with decreased serum TG, and the TG reducing effect of SNP rs1044250 will not be significant after the impact of ANGPTL4 N-terminus related SNP rs110843064 (E40K) is excluded, whereas E40K is positively related to an increased overall coronary heart disease risk [2, 37]. Accordingly, our result suggests that the rs1044250 polymorphism does not have independent interaction with circulating TG, but Mets.

This study proves that the SNP rs1044250 is an independent risk factor for increased WC. It has been reported previously that overexpression of purified ANGPTL4 C-terminus in mouse accelerates the decomposition of WAT lipid, suggesting a lipolytic activity of ANGPTL4 C-terminal domain in fat cells independent from LPL [38]. Therefore, the SNP rs1044250 is likely to induce an increase in WC by lowering the level of WAT lipolysis. In addition, the WC and waist-to-hip ratio is increased in adipocyte ANGPTL4 knockout mice [22]. Another research has found that the ANGPTL4 knockout mice fed with high-fat diet show granuloma lesions in the intestine and WAT, as well as lymphangitis and mesenteric lymphadenitis [39, 40], which suggests that ANGPTL4 is essential to the lymphatic drainage of lipids from WAT to the liver. A study has also reported that ANGPTL4 reduces appetite by inhibiting hypothalamic adenosine monophosphate-activated protein kinase (AMPK) activity, accordingly, the ANGPTL4 knockout mice show increased appetite after fasting [41]. Regarding features of ANGPTL4 discussed above, we speculate that SNP rs1044250 causes abdominal fat accumulation and increased WC by promoting the WAT lipid recruitment, hindering lipolysis, destroying the integrity of WAT-hepatic lymphoid tissue as well as increasing appetite.

The ANGPTL4 gene SNP rs1044250 is an independent risk factor for elevated blood pressure. It has been reported that the C-terminal domain of ANGPTL4 protein inhibits vascular epithelial growth factor (VEGF) and basic fibroblast growth factor (bFGF)-mediated angiogenesis [15], and SNP rs1044250 is capable of altering the activity of ANGPTL4 C-terminus, therefore, the SNP rs1044250 may lead to a decreased angiogenesis and dysfunctional endothelial repairment. ANGPTL4 gene knockout mice are prone to have coronary arteritis and mesenteric vasculitis when fed with a high-fat diet [39, 42, 43]. The dysfunction of endothelial repairment can increase vascular resistance and lead to an increased blood pressure in both direct and indirect manners. Consistently, circulating ANGPTL4 protein levels are significantly up-regulated in patients with hypertension [16, 24]. Based on the risk of the SNP rs1044250 on WC and blood pressure, the ANGPTL4 gene polymorphism is considered a risk factor for Mets. A step forward, we attempt to cast light on the interaction between ANGPTL4 polymorphisms and lifestyle interventions, and to find a feasible way for rs1044250-targeted Mets therapy.

At present, the management of Mets is mainly based on lifestyle interventions. The body weight reflects a time superposition effect of lifestyle. Therefore, weight management has become a recommended indicator of lifestyle interventions to Mets. In 2017, the international panel recommended a minimum 5% weight loss target for obese patients [4]. Our result shows that patients with rs1044250 CC genotype are more likely to have an increased number of Mets components as body weight raised. Elevated blood pressure is considered the major cause of increased number of Mets components. Accordingly, the synergistic effect of weight management and SNP rs1044250 on blood pressure is negative, in other words, the superimposed effect of these two independent variables is less than the sum of their effects alone. Studies have shown that muscle-derived ANGPTL4 levels are increased during exercise or fasting [31, 44], and that WAT lipolysis is positively correlated to the circulating ANGPTL4 levels [22]. We speculate the lack of exercise leads to a decrease in myogenic ANGPTL4, which in turn results in the accumulation of WAT and weight gain. Therefore, the transient decrease of ANGPTL4 level in circulation during weight gain is negatively correlated to the risk of SNP rs1044250 on blood pressure. Under this circumstance, the blood pressure of rs1044250 wild-type patients shows a catch-up effect, which may relate to a fasting involved metabolism disorder.

Participants with Mets were suggested to have lifestyle interventions and take prescription drugs. As a result, the Mets group showed improvement in various physical and laboratory indicators in the follow-up survey. Participants with TT genotype have elevated HDL-c, which was significantly decreased among the weight loss group in the follow-up survey. ANGPTL4 is characterized by its lipolytic effect and HDL-c is negatively correlated to TG [26]. However, previous research has reported that elevated HDL-c is associated with ANGPTL4 overexpression [22], which is considered discrepant to the role of ANGPTL4. In this research, we have demonstrated a catch-up of HDL-c among participants with CC and CT genotype, in another word, weight management overtakes the negative correlation of r1044250 to HDL-c, unfortunately, the cross-over assay was failed to provide more evidence.

This study elaborated on the interaction between the ANGPTL4 gene SNP rs1044250 and weight management on Mets. Future studies need to clarify the effect of SNP rs1044250 on the structure of ANGPTL4 C-terminus, and to detect the interaction between SNP rs1044250 and neovascularization. The effect of SNP rs1044250 on circulating lipid metabolism under fasting and exercise conditions also requires experimental verification.

Conclusion

We have found that the ANGPTL4 gene SNP rs1044250 increases the incidence of Mets in the Shandong Han population by increasing blood pressure and WC. The number of Mets components in patients with CC genotype at rs1044250 locus shows a catch-up effect when the body weight is increased, while weight-loss could significantly inhibit the increase of SBP and DBP caused by rs1044250 polymorphism.