Effects of Caloric Restriction with Protein Supplementation on Plasma Protein Profiles in Middle-Aged Women with Metabolic Syndrome—A Preliminary Open Study

Background: Clinical studies have demonstrated that higher protein intake based on caloric restriction (CR) alleviates metabolic abnormalities. However, no study has examined the effects of plasma protein profiles on caloric restriction with protein supplementation (CRPS) in metabolic syndrome (MetS). Therefore, using a proteomic perspective, this pilot study investigated whether CRPS ameliorated metabolic abnormalities associated with MetS in middle-aged women. Methods: Plasma samples of middle-aged women with MetS in CR (n = 7) and CRPS (n = 6) groups for a 12-week intervention were obtained and their protein profiles were analysed. Briefly, blood samples from qualified participants were drawn before and after the dietary treatment. Anthropometric, clinical, and biochemical variables were measured and correlated with plasma proteomics. Results: In results, we found that body mass index, total body fat, and fasting blood glucose decreased significantly after the interventions but were not different between the CR and CRPS groups. After liquid chromatography–tandem mass spectrometry analysis, the relative plasma levels of alpha-2-macroglobulin (A2M), C4b-binding protein alpha chain (C4BPA), complement C1r subcomponent-like protein (C1RL), complement component C6 (C6), complement component C8 gamma chain (C8G), and vitamin K-dependent protein S (PROS) were significantly different between the CRPS and CR groups. These proteins are involved in inflammation, the immune system, and coagulation responses. Moreover, blood low-density lipoprotein cholesterol levels were significantly and positively correlated with C6 plasma levels in both groups. Conclusions: These findings suggest that CRPS improves inflammatory responses in middle-aged women with MetS. Specific plasma protein expression (i.e., A2M, C4BPA, C1RL, C6, C8G, and PROS) associated with the complement system was highly correlated with fasting blood glucose (FBG), blood lipids (BLs), and body fat.


Introduction
Metabolic syndrome (MetS) is an escalating global public health challenge. The prevalence of MetS varies among countries and is affected by region, sex, age, and ethnicity [1,2]. According to excluded. Participants who were pregnant, taking any type of lipid-lowering, antihypertension, or hypoglycaemic medication were also excluded.

Study Design
The participants in the cohort study were randomly assigned to one of four groups after a 2-week run-in period, and were then asked to follow a 12-week dietary intervention. The four groups were as follows: CR, CRPS, caloric restriction with fish supplementation, and caloric restriction with protein and fish supplementation. During the 2-week run-in period, the participants were consulted by a dietitian to estimate their regular daily dietary intake and basal metabolic rate ( Figure 1A).
In this pilot study, we focused on only two of the four groups: CR and CRPS. The inclusion criteria were as follows: female aged ≥40 years, relatively high blood biochemical variable improvements (i.e., BMI, WC, fasting blood glucose, and blood lipids) after the 12-week intervention, adequate plasma sample, and provision of written informed consent. After the participant selection process, seven and six subjects' plasma samples were selected in CR and CRPS groups for the proteomic analysis, respectively ( Figure 1B). This additional study was approved by the Ethics Committee of the Joint Institutional Review Board at TMU (N201704088), registered at ClinicalTrials.gov as NCT01768169, and conducted in accordance with the Declaration of Helsinki.
During the 12-week dietary interventions, the CR group was asked to consume calorie-controlled lunches and dinners prepared by the Department of Nutrition at TMUH. The CRPS group was asked to consume calorie-controlled lunches and Herbalife Formula 1 ® (Herbalife, Los Angeles, CA, USA) for dinner. Herbalife Formula 1 ® is a low-calorie nutritional drink (protein powder); the participants were asked to mix 25 g of the protein powder (comprising 11 g of carbohydrates, 0.6 g of fat, and 8 g of protein) with water (81 kcal/serving). Herbalife Formula 1 ® was provided as a protein supplement to increase the participants' daily protein intake. J. Clin. Med. 2019, 8,195 3 of 17 history of cardiovascular events, alcohol or substance abuse, or cardiovascular, hepatic, renal, metabolic, endocrine, psychiatric, cerebrovascular, or peripheral vascular diseases were excluded. Participants who were pregnant, taking any type of lipid-lowering, antihypertension, or hypoglycaemic medication were also excluded.

Study Design
The participants in the cohort study were randomly assigned to one of four groups after a 2-week run-in period, and were then asked to follow a 12-week dietary intervention. The four groups were as follows: CR, CRPS, caloric restriction with fish supplementation, and caloric restriction with protein and fish supplementation. During the 2-week run-in period, the participants were consulted by a dietitian to estimate their regular daily dietary intake and basal metabolic rate ( Figure 1A).
In this pilot study, we focused on only two of the four groups: CR and CRPS. The inclusion criteria were as follows: female aged ≥40 years, relatively high blood biochemical variable improvements (i.e., BMI, WC, fasting blood glucose, and blood lipids) after the 12-week intervention, adequate plasma sample, and provision of written informed consent. After the participant selection process, seven and six subjects' plasma samples were selected in CR and CRPS groups for the proteomic analysis, respectively ( Figure 1B). This additional study was approved by the Ethics Committee of the Joint Institutional Review Board at TMU (N201704088), registered at ClinicalTrials.gov as NCT01768169, and conducted in accordance with the Declaration of Helsinki.
During the 12-week dietary interventions, the CR group was asked to consume caloriecontrolled lunches and dinners prepared by the Department of Nutrition at TMUH. The CRPS group was asked to consume calorie-controlled lunches and Herbalife Formula 1 ® (Herbalife, Los Angeles, CA, USA) for dinner. Herbalife Formula 1 ® is a low-calorie nutritional drink (protein powder); the participants were asked to mix 25 g of the protein powder (comprising 11 g of carbohydrates, 0.6 g of fat, and 8 g of protein) with water (81 kcal/serving). Herbalife Formula 1 ® was provided as a protein supplement to increase the participants' daily protein intake.
(A) (B) We focused on only two groups in the current study: CR and CRPS. The participants' plasma was digested into peptides and then analysed using LC-MS/MS. BMI: body mass index, WC: waist circumference, FBG: fasting blood glucose, BLs: blood lipids, MetS: Metabolic syndrome.

Anthropometry and Analysis of Clinical and Biochemical Variables
The participants' height, body weight, WC, and body composition were measured at baseline (week 0) and after intervention (week 12). An oral glucose tolerance test and blood drawing were also conducted at both time points. The blood samples were then stored at −80 °C until analysis. Briefly, serum albumin, TG, and TC were measured using an automated analyser (Ortho Clinical Diagnostics VITROS 950, Johnson & Johnson, New Brunswick, NJ, USA). Serum LDL-C and HDL-C were analysed using an automated analyser (Toshiba TBA-c16000, Toshiba, Tokyo, Japan). FBG was determined using an automated analyser (VITOR 5, 1FS, Ortho Clinical Diagnostics, Johnson & Johnson) with Vitros Chemistry Products GLU slides. Serum insulin was analysed using a radioimmunoassay kit (DIAsource, Lovain-La-Nueve, Belgium). HbA1c was determined using an HLC-723 GHb G7 analyser (Tosoh, Tokyo, Japan). Serum C-reactive protein (CRP) concentrations were analysed using an automated analyser (Toshiba TBA-c16000). IL6 was measured using a Qunatikine high-sensitivity commercial enzyme-linked immunosorbent assay kit (R&D Systems, Minneapolis, MN, USA). In this study, the qualified and subject-matched plasma samples were used to analyse the basic biochemical characteristics and protein profile (or proteome) (Tables 1 and 2). (B) We focused on only two groups in the current study: CR and CRPS. The participants' plasma was digested into peptides and then analysed using LC-MS/MS. BMI: body mass index, WC: waist circumference, FBG: fasting blood glucose, BLs: blood lipids, MetS: Metabolic syndrome.

Anthropometry and Analysis of Clinical and Biochemical Variables
The participants' height, body weight, WC, and body composition were measured at baseline (week 0) and after intervention (week 12). An oral glucose tolerance test and blood drawing were also conducted at both time points. The blood samples were then stored at −80 • C until analysis. Briefly, serum albumin, TG, and TC were measured using an automated analyser (Ortho Clinical Diagnostics VITROS 950, Johnson & Johnson, New Brunswick, NJ, USA). Serum LDL-C and HDL-C were analysed using an automated analyser (Toshiba TBA-c16000, Toshiba, Tokyo, Japan). FBG was determined using an automated analyser (VITOR 5, 1FS, Ortho Clinical Diagnostics, Johnson & Johnson) with Vitros Chemistry Products GLU slides. Serum insulin was analysed using a radioimmunoassay kit (DIAsource, Lovain-La-Nueve, Belgium). HbA1c was determined using an HLC-723 GHb G7 analyser (Tosoh, Tokyo, Japan). Serum C-reactive protein (CRP) concentrations were analysed using an automated analyser (Toshiba TBA-c16000). IL6 was measured using a Qunatikine high-sensitivity commercial enzyme-linked immunosorbent assay kit (R&D Systems, Minneapolis, MN, USA). In this study, the qualified and subject-matched plasma samples were used to analyse the basic biochemical characteristics and protein profile (or proteome) (Tables 1 and 2).  During the study period, the participants attended ten weekly sessions of a nutrition course that provided information on healthy diets, exercise, dietary habits, and dietary behaviour modification. Participant compliance was assessed as attendance in the weekly sessions.

Liquid Chromatography-Tandem Mass Spectrometry
Five microliter qualified plasma samples were spiked with protein standard β-lactoglobulin (LACB) and were processed with dithiothreitol (DTT) and iodoacetamide (IAA) for reduction and alkylation, respectively. Processed plasma samples were digested with SMART Digest Kit (SMART Digest Trypsin Kit, P/N60109-101, Thermo Fisher, Bedford, MA, USA), reduced, desalted (Millipore®Ziptips Micro-C 18 , P/NZ720003, Sigma, St. Louis, MO, USA), and purified by SOLAµ™ SPE Plates (Thermo Fisher, Bedford, MA, USA). Finally, the processed samples were dissolved in 0.1% formic acid for LC-MS/MS analysis. After sample preparation, digested peptides were loaded into a LTQ-Orbitrap Elite mass spectrometer with a nanoelectrospray ionisation source (Thermo Electron, Waltham, MA, USA) connected to a nanoACQUITY UPLC system (Waters, Milford, MA, USA). Peptide samples were first loaded with a single injection model into the nanoACQUITY UPLC system, and then peptides were captured and desalted on a C 18 trap column (2 cm × 180 µm, Symmetry C 18 , Waters, Milford, MA, USA). The peptide samples were separated using a BEH130 C 18 column (25 cm × 75 µm, Waters, Milford, MA, USA) with a 0-95% segmented gradient of 3-40% B for 168 min, 40-95% B for 2 min, and 95% B for 10 min at a flow rate of 0.5 µL/min. Mobile phases were prepared as solution A (0.1% formic acid in water) and solution B (0.1% formic acid in acetonitrile). The eluted peptides were ionised with a spray voltage of 2.33 kV and introduced into the LTQ-Orbitrap Elite mass spectrometer. The mass spectrometer was conducted in the positive ion mode and on the basis of a data-dependent acquisition method (isolation width: 1.5 Da). Mass spectrum data of the peptides were obtained using a full mass spectrometer survey scan (m/z range of 350-1600) of 30,000. According to the data-dependent acquisition method, the first 15 most intensively charged peptide ions were scanned. High-energy collisional dissociation of the selected precursor peptide ions was stimulated with helium. The MS data were deposited as mzML to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) with identifier PXD012213 (Project DOI: 10.6019/PXD012213). An enzyme-linked immunosorbent assay (ELISA) was conducted to confirm the liquid chromatography-tandem mass spectrometry (LC-MS/MS) proteomic results.

Protein Identification
The acquired proteomic raw data files were then applied to search against a UniProt human protein database (containing 162,989 protein sequences; released on April 2017; http://www.uniprot. org/) by using PEAKS Studio 7.5 (Bioinformatics Solutions, Waterloo, Ontario, Canada). The settings in PEAKS Studio 7.5 combined with UniProt for searching the protein database were as follows: enzyme set as trypsin with a maximum of two missed cleavage sites; precursor and fragment mass tolerance of 20 ppm and 0.8 Da, respectively; and false discovery rate <1%, obtained through search against a decoy database in all protein and peptide characteristics. A protein was identified when at least one unique peptide was matched. Protein quantification was based on label-free quantitative analysis. Furthermore, spectrum counts were normalised with the total identified spectra per biological sample and the proteins. The proteins (containing at least two matched peptides or one unique peptide) with statistically higher or lower peptide counts in the participants (nonparametric Quade's test was conducted in SAS version 9.4, Cary, NC, USA) were considered as different expressions. All mass data of this study have been documented as raw files and peak lists in ProteomeXchange. The selected proteins were based on the biochemical characteristics improvements (included: blood pressure-, coagulation-, complement system-, glucose metabolism-, inflammatory, lean body mass-and lipid metabolism-associated proteins) and missing values in nanoLC-MS/MS based proteomics dataset ( Figure 2).
(nonparametric Quade's test was conducted in SAS version 9.4, Cary, NC, USA) were considered as different expressions. All mass data of this study have been documented as raw files and peak lists in ProteomeXchange. The selected proteins were based on the biochemical characteristics improvements (included: blood pressure-, coagulation-, complement system-, glucose metabolism-, inflammatory, lean body mass-and lipid metabolism-associated proteins) and missing values in nanoLC-MS/MS based proteomics dataset ( Figure 2).

Statistical Analysis
Differences between the postintervention clinical and biochemical characteristics of the treatment groups were compared using the Mann-Whitney U test. A comparison between the baseline and postintervention clinical and biochemical measurements between the groups was conducted using the Wilcoxon signed-rank test. Using the LC-MS/MS-derived proteomics data, the nonparametric Quade's test was adopted to compare the different postintervention protein expressions between the treatment groups with baseline measurements as covariates. Moreover, a Spearman's rank correlation coefficient was calculated to evaluate the relationship between the specific plasma protein expressions and clinical variables. All statistical analyses were performed using SAS version 9.4. Data are presented as the median (75th percentile values in parentheses), and p < 0.05 was considered statistically significant.

Anthropometric and Clinical Characteristics
Among those in the CR (n = 44) and CRPS (n = 45) groups (Figures 1 and 2), there were seven and six middle-aged female participants, respectively, enrolled in this pilot study. Their plasma samples were processed, and their protein profiles were analysed using LC-MS/MS. According to the database from the cohort study, the groups showed no significant differences in age: CR = 61.97 years and CRPS = 55.84 years, p = 0.52. Body weight, BMI, android fat, gynoid fat, total body fat (TBF), and FBG decreased significantly in the groups throughout the 12-week interventions, but no significant differences were observed between the groups ( Table 1). The basic characteristics of subjects were shown in Supplementary Materials.

LC-MS/MS Proteomic Analysis
A proteomic analysis conducted using LC-MS/MS (Thermo Scientific LTQ-Orbitrap Elite, Thermo Fisher, Bedford, MA, USA) revealed that postintervention alpha-2-macroglobulin (A2M) protein expression in the CRPS group was significantly higher than in the CR group. Furthermore, C4BPA, complement C1r subcomponent-like protein (C1RL), C6, C8G, and PROS expression was significantly lower in the CRPS group than in the CR group after intervention (Table 2)

Correlation of Specific Protein Candidates with Biochemical Variables
A Spearman's rank correlation coefficient was calculated for the specific protein candidates and showed significant differences between the groups with the selected anthropometric and biochemical variables. A significantly positive correlation was observed between C6 and blood LDL-C levels in the CR group (rs = 0.93, p <0.01) ( Figure 4A). Furthermore, C1RL plasma levels were negatively correlated with the android-to-gynoid ratio (A/G%) (rs = −0.90, p = 0.01) and TBF (rs = −0.84, p = 0.04) in the CRPS group. C1RL expression was also negatively correlated with HbA1c levels in the CRPS group (rs = 0.81, p = 0.05). C6 levels were positively correlated with TC (rs = 0.94, p <0.01) and LDL-C (rs = 0.99, p <0.01) levels, whereas there was a significantly positive correlation between C8 and FBG levels in the CRPS group (rs = 0.99, p <0.01) ( Figure 4B).

Correlation of Specific Protein Candidates with Biochemical Variables
A Spearman's rank correlation coefficient was calculated for the specific protein candidates and showed significant differences between the groups with the selected anthropometric and biochemical variables. A significantly positive correlation was observed between C6 and blood LDL-C levels in the CR group (rs = 0.93, p < 0.01) ( Figure 4A). Furthermore, C1RL plasma levels were negatively correlated with the android-to-gynoid ratio (A/G%) (rs = −0.90, p = 0.01) and TBF (rs = −0.84, p = 0.04) in the CRPS group. C1RL expression was also negatively correlated with HbA1c levels in the CRPS group (rs = 0.81, p = 0.05). C6 levels were positively correlated with TC (rs = 0.94, p < 0.01) and LDL-C (rs = 0.99, p < 0.01) levels, whereas there was a significantly positive correlation between C8 and FBG levels in the CRPS group (rs = 0.99, p < 0.01) ( Figure 4B).    Figure 5 displays a volcano plot of the plasma selected proteomic ratio measurements within each group throughout the trial. The proteins depicted as red dots represent the plasma protein levels that increased significantly (p < 0.05) with a 1.25-fold change (25% change threshold), whereas the proteins depicted as blue dots represent the plasma protein levels that decreased significantly to 25% changes of the original value. Apolipoprotein A2 (APOA2), complement component 3 (C3), A2M, fibrinogen alpha chain, alpha-1-microglobulin/bikunin precursor (AMBP), and fibronectin (FINC) levels decreased significantly following the 12-week intervention in the CR group. Moreover, apolipoprotein B-100 (APOB-100), apolipoprotein C2 (APOC2), and C6 levels increased significantly throughout the intervention in the CR group ( Figure 5). Moreover, FINC levels decreased significantly following the 12-week intervention in the CRPS group ( Figure 6). The detailed proteins characteristics were shown in Supplementary Materials.  Figure 5 displays a volcano plot of the plasma selected proteomic ratio measurements within each group throughout the trial. The proteins depicted as red dots represent the plasma protein levels that increased significantly (p <0.05) with a 1.25-fold change (25% change threshold), whereas the proteins depicted as blue dots represent the plasma protein levels that decreased significantly to 25% changes of the original value. Apolipoprotein A2 (APOA2), complement component 3 (C3), A2M, fibrinogen alpha chain, alpha-1-microglobulin/bikunin precursor (AMBP), and fibronectin (FINC) levels decreased significantly following the 12-week intervention in the CR group. Moreover, apolipoprotein B-100 (APOB-100), apolipoprotein C2 (APOC2), and C6 levels increased significantly throughout the intervention in the CR group ( Figure 5). Moreover, FINC levels decreased significantly following the 12-week intervention in the CRPS group ( Figure 6). The detailed proteins characteristics were shown in Supplementary Materials.

Effects of the Dietary Intervention on the Clinical Variables
After comparing the anthropometric and biochemical variables between the groups, we found no significant differences in baseline measurements. According to the literature review, we found that the short dietary interventions (12 weeks), loose target for the CR group (1500 kcal/per day), and relatively low protein supplementation dose provided (17% kcal/per day) in this study might explain the reasons for the minor effects of CRPS. A study by Flechtner-Mors et al. showed that a relatively high protein intake (35% kcal/day) combined with 1300 kcal/per day contributed to a significant body weight loss throughout a 12-week intervention [19]. Furthermore, Josse et al. suggested that a relatively high protein intake (28% kcal/day) combined with 1500 kcal/per day contributed to significant body weight and body fat mass losses and a significant increase in lean body mass after a 16-week dietary intervention [10]. Further study on protein supplementation should be considered to enable the use of a relatively high protein dose for a longer intervention period.

Effect of Protein Supplementation on Plasma Protein Profiles
The results showed that eight plasma proteins (A2M, C4BPA, C1RL, C6, C8G, and PROS) were significantly different between the groups. Those plasma protein candidates have been associated with inflammation (A2M and C4BPA), the complement system (C1RL, C6, and C8), and coagulation responses (PROS).
A2M is regarded as a powerful protease inhibitor that is involved in proteolysis [20]. Studies have also shown that A2M is an acute phase protein (APP) that increases slightly during inflammatory responses (e.g., surgical trauma, myocardial infarction, and severe burns) in humans [21][22][23]; however, A2M levels are significantly decreased in people who are overweight or obese. Moreover, a significantly negative correlation was found between A2M levels and BMI [24]. A study reported that

Effects of the Dietary Intervention on the Clinical Variables
After comparing the anthropometric and biochemical variables between the groups, we found no significant differences in baseline measurements. According to the literature review, we found that the short dietary interventions (12 weeks), loose target for the CR group (1500 kcal/per day), and relatively low protein supplementation dose provided (17% kcal/per day) in this study might explain the reasons for the minor effects of CRPS. A study by Flechtner-Mors et al. showed that a relatively high protein intake (35% kcal/day) combined with 1300 kcal/per day contributed to a significant body weight loss throughout a 12-week intervention [19]. Furthermore, Josse et al. suggested that a relatively high protein intake (28% kcal/day) combined with 1500 kcal/per day contributed to significant body weight and body fat mass losses and a significant increase in lean body mass after a 16-week dietary intervention [10]. Further study on protein supplementation should be considered to enable the use of a relatively high protein dose for a longer intervention period.

Effect of Protein Supplementation on Plasma Protein Profiles
The results showed that eight plasma proteins (A2M, C4BPA, C1RL, C6, C8G, and PROS) were significantly different between the groups. Those plasma protein candidates have been associated with inflammation (A2M and C4BPA), the complement system (C1RL, C6, and C8), and coagulation responses (PROS).
A2M is regarded as a powerful protease inhibitor that is involved in proteolysis [20]. Studies have also shown that A2M is an acute phase protein (APP) that increases slightly during inflammatory responses (e.g., surgical trauma, myocardial infarction, and severe burns) in humans [21][22][23]; however, A2M levels are significantly decreased in people who are overweight or obese. Moreover, a significantly negative correlation was found between A2M levels and BMI [24]. A study reported that the expression of A2M is likely to increase under a relatively low consumption of calories and animal-derived proteins [21]. In the current study, A2M plasma levels were significantly higher in the CRPS group than in the CR group after intervention. However, there were no significant differences in IL-6 or CRP levels between the groups throughout the interventions. Thus, we infer that the increased A2M plasma levels in the CRPS group were influenced by the plant-derived protein supplementation.
C4BP has been shown to be involved in not only the complement system but also inflammation. Moreover, it is regarded as a positive APP. Human C4BP is mainly composed of an α chain (C4BPA) and a β chain (C4BPB), and there are three isoforms (α 7 β 1 , α 6 β 1 , and α 7 β 0 ) of C4BP in humans [25]. A study suggested that C4BP α 7 β 0 isoform levels increase significantly in inflammation [26]. Clinical research also showed a significant increase in C4BPA expression in inflammatory responses [27]. We observed a significant decrease in C4BPA protein expression in the CRPS group, which shows the potential benefit of CRPS in inflammatory responses in people with obesity.
C1RL has been associated with the classical pathway and has a catalytic effect on proteolysis of pro-C1s, stimulating the activation of the pathway [28]. Furthermore, animal experiments have demonstrated a significant increase in mRNA C1RL levels under inflammation conditions [29]. C6 and C8G are thought to be part of the membrane attack complex (MAC), a structure typically formed from the activation of a host's complement system [30]. The MAC is involved in cell lysis, and recent studies have demonstrated that it is also an inflammatory trigger [31,32]. Moreover, studies have suggested that the MAC stimulates the release of proinflammatory substrates following the activation of endothelial cells and platelets [33]. In the current study, we found that C1RL, C6, and C8G plasma levels were significantly lower in the CRPS group compared with the CR group, which suggests that CRPS reduces inflammatory responses in people with obesity.
Studies have reported that PROS is an anticoagulant protein and is mainly involved in catalysing the inactivation process of factor V and factor VIII through activated protein C (APC) [33,34]. A study showed that PROS expression was inhibited by NF-κB in inflammation [35]. However, in the current study, postintervention PROS plasma levels were significantly lower in the CRPS group compared with the CR group. Moreover, IL-6 and CRP levels, which have been associated with inflammatory responses, were not significantly different between the groups. According to the absolute amount of PROS derived from ELISA, the mean plasma level of PROS in the CRPS group was 20.34 µg/mL, which was still within the normal range (20-25 µg/mL) [36].

Spearman's Correlation of Selected Plasma Protein Candidates with Biochemical Variables
Although we observed several significant differences between the plasma protein candidates and the selected biochemical variables, C6 plasma levels were significantly correlated with LDL-C levels in both groups. A study showed that LDL-C plasma levels decreased through an LDL apheresis (LDL-A) operation in patients with familial hypercholesterolaemia. Moreover, the LDL-A operation was shown to prevent cardiovascular events [37]. In another proteomic study, Yuasa et al. suggested that C6 is involved in the cause and pathophysiology of atherosclerosis [38]. To the best of our knowledge, the current study is the first to report a highly significant difference between C6 and LDL-C levels. Further studies with a larger sample size are required in the future.

Effects of CRPS on Plasma Protein Profiles
FINC plasma levels decreased significantly in both the CR and CRPS groups after the 12-week dietary interventions. A significantly positive correlation was found between FINC plasma levels and the degree of obesity in the participants [39]. Furthermore, the adipose tissue gene expression of FINC was significantly downregulated in those who successfully maintained their weight after the CR intervention and the following weight maintenance phase [40]. Studies have suggested that increased FINC plasma levels are related to DM, cardiovascular disease, kidney disease, and pre-eclampsia [41][42][43][44][45][46]; therefore, the significant decrease in FINC plasma levels throughout the interventions in both groups suggests that CRPS is beneficial in metabolic abnormalities.

Limitations
This study has several limitations. First, although this study was designed as a pilot study, a small sample size led to inadequate statistical power. Second, an inadequate dietary intervention might have influenced the effect of the protein supplementation. Large double-blinded randomised studies on patients with MetS are required to confirm our findings.

Conclusions
The findings suggest that CRPS improves inflammatory responses in middle-aged women with MetS. Furthermore, specific plasma protein expression associated with the complement system was highly correlated with FBG, blood lipids (BLs), and body fat. The precise effect of CRPS and the relevance of specific plasma proteins as potential predictors of biochemical variables should be investigated in large clinical trials.  Acknowledgments: The authors thank the participants for their cooperation, and the staff of Taipei Medical University Hospital and Wan Fan Hospital in Taiwan for their assistance throughout this study.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.