Introduction
It has been well documented that atherosclerosis is the major cause of death in type 2 diabetic patients [
1,
2]. Dyslipidaemia plays an important role in the pathogenesis of atherosclerosis and even to insulin resistance [
3,
4]. Dyslipidaemia is associated with increased lipolysis and the release of higher amounts non-esterified fatty acids (NEFAs) into the bloodstream [
5]. Hyperglycaemia promotes lipolysis and leads to chronic exposure to NEFA [
6]. Several studies have shown that the elevated serum level of NEFA contributes to vascular damage in diabetes [
7,
8].
Plasma NEFAs promote a systemic insulin resistance state and are correlated with inflammation in atheromatous plaques [
5,
7]. The levels of NEFAs seem to be modified by dietary or therapeutic intervention. Glucagon-like peptide-1 (GLP-1) has recently received attention as a novel drug in the antidiabetic field [
9]. The long-acting GLP-1 analogue liraglutide (LRG) exhibits pleiotropic effects on glucolipid metabolism, β-cell insulin secretion and anti-atherosclerosis [
10‐
15]. Nevertheless, the precise mechanism by which liraglutide modulates lipids metabolism remains unclear.
Therefore, the aim of this study is to evaluate whether liraglutide could be more effective at suppressing lipolysis, ameliorating glucose and lipid metabolism, enhancing beta-cell functions, and inhibiting vascular inflammatory markers compared to metformin in recent-onset type 2 diabetes mellitus.
Subjects and methods
Ethical approval
Ethical approval was granted by the Institutional Review Board of Xiamen University Affiliated Zhongshan Hospital according to Helsinki Declaration II. Written informed consent was signed by and obtained from each participant.
Participants
Participants were recruited between October 2015 and December 2016 at Zhongshan Hospital Xiamen University in P. R. China. The inclusion criteria were set to select participants: the patients had an initial diagnosis of type 2 diabetes mellitus who met the World Health Organization criteria for type 2 diabetes and were aged 18–40 years, with a body mass index (BMI) of 25–35 kg/m
2, HbA1c of 6.5–9%, and fasting serum triglycerides of 0.5–5 mmol/L. Main exclusion criteria included the following: type 1 diabetes, recent acute complications including diabetic ketoacidosis and hyperglycaemic hyperosmolar state, impaired liver function, impaired renal function (creatinine clearance < 60 ml/min) [
16], pregnancy and breast bleeding, and other factors that affect glucose and lipid changes.
Study design and procedures
This study is a randomized, parallel, active comparator trial in 60 participants. The treatment order of liraglutide or metformin was determined through random assignment (1:1). All individuals who received liraglutide treatment started injection doses at 0.6 mg/day. After 1 week of injection, the dose was increased to 1.2 mg/day. Participants in another group who received metformin received an oral dose in the range of 1–2 g/day. After an overnight fast, a 75 g oral glucose tolerance test (OGTT) was conducted for each participant. Blood samples were drawn before and 30, 60, 120 min after OGTT, respectively. Participants in the liraglutide treatment group were instructed to administer the Flexpen device. Participants were also informed of medication precautions before metformin treatment. All participants received lifestyle intervention health education from professional nurse. The total planned treatment period was 8 weeks. There were a total of 3 follow-up visits, once per month. The plasma glucose, blood pressure, body weight, waist circumference and hip circumference were measured, and adverse events were monitored by investigators. At the end of the trial, the clinical and laboratory indices were assessed, as previously described.
Clinical measurement and laboratory test
Weight, waist circumference (WC), and hip circumference were measured by professional nurses. Participants removed clothing and shoes prior to measurements. WC was measured midway between the lowest rib and the top of the iliac crest. Hip circumference was measured around the peak of the buttocks. Body mass index (BMI) was calculated as the body weight (BW) in kilograms divided by the square of the patient’s height in metres. Measurements of systolic blood pressure and diastolic blood pressure (SBP and DBP) were performed 2 times with a mercury manometer on the right arm after 5 min in the sitting position. The mean values of measurements were recorded.
Participants were instructed to refrain from excessive physical exercise and to eat normally 3 days before the test, with no less than 200 grams of carbohydrates daily. After fasting overnight (8–10 h), a 2-h OGTT test was performed on each subject. Fasting plasma glucose (FPG), 30 min PG, 60 min PG, and 120 min PG were measured by the hexokinase method. Subjects’ renal and liver functions, plasma lipids and lipoprotein concentrations including triglycerides (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) were assayed using standard methods (Roche cobas8000 automatic biochemical analyser). HbA1c was analysed by HPLC (Bio-Rad, Inc., Hercules, CA, USA). Plasma insulin levels were measured using the electrochemiluminescence immunoassay (ECLI). NEFA levels were measured in the Xiamen Diabetes Hospital using Abbott c16000 automatic biochemical analyser. Soluble vascular cell adhesion molecule-1 (sVCAM-1) and soluble intercellular adhesion molecule-1 (sICAM-1) were measured in the plasma of patients using quantikine ELISA kits (R&D systems, Inc., Minneapolis, USA). Plasminogen activator inhibitor 1 (PAI-1) concentrations were assessed using ELISA assays (Multi sciences, Inc., Hangzhou, PRC). The intra-assay and inter-assay coefficients of variation (CV) of the ELISA kits mentioned above were all less than 10%. The early phase insulin secretion index was calculated as (ΔI30/ΔG30) = ([insulin at 30 min] − [insulin at 0 min])/([glucose at 30 min] − [glucose at 0 min]) [
17]. The area under the curve (AUC) for plasma insulin during the OGTT was calculated using the trapezoidal rule [
18]. Deltas (Δ) are presented as the difference before and after treatment, which were available for the variables ΔNEFA, ΔsVCAM-1, ΔAUCins, and ΔI30/ΔG30. Deltas were also considered to express the difference between 120 min values and fasting values in ΔTC and ΔLDL-C.
Statistical analysis
SPSS version 21 (SPSS software, IBM Inc., USA) and GraphPad Prism version 5.0 (GraphPad software, Inc., La Jolla, CA, USA) were utilized for statistical analysis and the construction of graphs. The normal distribution data are presented as the mean ± standard deviation. Comparisons of basic characteristics between groups were made using an unpaired t test. Comparisons of baseline and post-treatment data in the same group were made using a paired t test. The skewed distribution variables are expressed as the median (interquartile rang), and the Mann–Whitney U test or Wilcoxon signed rank test was performed. The Mann–Whitney U test was used to determine the differences from baseline with a final measurement at 8 weeks for 120 min NEFA, insulin secretion capacity including fasting insulin, 30 min insulin, 60 min insulin, 120 min insulin, ΔI30/ΔG30, and AUCins between the liraglutide and metformin group. At the same time, the Mann–Whitney U test was carried out to determine the differences between 120 min values and fasting values on TC and LDL-C in the metformin group. Spearman rank correlations were performed to determine the relationship of ΔNEFA with ΔsVCAM-1, ΔAUCins and ΔI30/ΔG30 after 8 weeks of treatment. The Chi squared (χ
2
) test was used for categorical variables. Two-tailed significance was set at P < 0.05.
Discussion
Our results indicate that prolonged liraglutide treatment for 8 weeks effectively inhibited 120 min non-esterified free fatty acids (NEFAs). We utilized a 75 g oral glucose tolerance test (OGTT) to evaluate the changes of NEFA, which should avoid the effects of a high-fat diet on postprandial lipid metabolism. As observed in this study, NEFA levels gradually decreased after glucose uptake, and 120 min NEFAs are dramatically lower than fasting NEFAs. Furthermore, our data show that liraglutide exhibits greater stimulation on the insulin secretion capacity, including early phase insulin secretion, fast plasma insulin, 60 min plasma insulin, 120 min plasma insulin and insulin area under the curve (AUCins) relative to those at baseline. Previous research has shown that insulin could suppress chylomicron synthesis from human jejunal explants [
19]. Additionally, circulating FFAs derived from chylomicron are reduced in the hyperinsulinaemia state, as insulin prevents lipolysis in the regulation of hepatic and intestinal lipoprotein production [
20,
21].
Our data demonstrated that after the 8-week treatment period, the reduction in 120 min NEFA (ΔNEFA) was greater increased, and the increase in AUCins (ΔAUCins) was also enhanced in liraglutide treatment relative to metformin treatment. Consequently, in this study, liraglutide therapy exhibited beneficial effects in ameliorating the β-cell secretion capacity and suppressing free fatty acid production. However, the precise mechanisms by which glucagon-like peptide-1 affects lipids metabolism are less certain. It has been hypothesized that enterocyte GLP-1 receptor signalling is essential for postprandial lipoprotein synthesis and secretion, reduces intestinal lipid production and absorption, and prevents the postprandial increases in triacylglycerol, cholesterol and apo-B48 levels [
22]. Moreover, GLP-1 is related to the direct modulation of fatty acid binding protein 2 (FABP2), which is required for the formation of apo-B48 containing chylomicrons [
23]. In addition, even through enhanced insulin secretion, GLP-1 induced prolonged reductions in NEFA concentrations after meals, suppressed postprandial rises in ApoCIII [
24]. ApoCIII is a small glycoprotein synthesized in the liver and intestine that resides predominantly on the surface of ApoB-containing lipoproteins and HDL. ApoCIII inhibits lipoprotein lipase activity and interferes with the receptor-mediated uptake of TG-rich lipoproteins, therefore delaying the clearance of TG-rich lipoproteins [
25,
26]. On the other hand, a long-term increase in NEFA levels has been associated with ROS accumulation, which causes mitochondrial stress and leads to β-cell apoptosis [
27,
28]. In a catch-up growth rat model by re-feeding with high-fat diet, liraglutide prevented the increase of plasma NEFA, increased insulin secretion, increased islet pancreatic duodenal homeobox-1 (Pdx-1) and B cell lymphoma-2 (Bcl-2) expression, and reduced procaspase-3 transcription and Caspase-3 p11 subunit protein expression, which suggested that liraglutide treatment could antagonize β-cell apoptosis caused by elevated NEFA [
11].
Dyslipidaemia has been reported to affect vascular endothelial function by the inflammatory pathway [
29]. Raised non-esterified fatty acids impair insulin’s effect on glucose uptake in skeletal muscle and could thus have detrimental effects on the vascular endothelium, which leads to premature cardiovascular disease [
30]. As reported previously, the adhesion molecules usually act as biomarkers, which mediate interactions between leukocytes and the vascular endothelium, shed from cell surfaces and become soluble isoforms in the blood, namely, sVCAM-1 and sICAM-1 [
31]. Our study has shown that the levels of the vascular inflammatory marker sVCAM-1 are dramatically suppressed after liraglutide treatment compared to metformin treatment. However, we did not identify appreciable differences in the levels of sICAM-1 and PAI-1 after liraglutide therapy.
Considerable evidence has demonstrated that sVCAM-1 plays an important role in the pathophysiological mechanisms of atherosclerosis [
32]. During a 10-year follow-up, a longitudinal study revealed that elevated circulating sVCAM-1 concentrations in females with a history of gestational diabetes mellitus (GDM) not only represented the earliest high risk stage for developing type 2 diabetes, but also reflected the early stage of the pathway for the manifestation of future cardiometabolic disorders [
33]. It is widely known that atherosclerosis leads to macroangiopathy and is responsible for most deaths in patients with diabetes [
34]. Liraglutide was shown to significantly reduce the risk of major cardiovascular events in the large prospective LEADER trial [
12,
35‐
37]. However, the precise mechanisms behind the antiatherogenic effect of liraglutide are not entirely clear, although several hypotheses have been proposed. First, it has been shown that in cultured human aortic endothelial cells (HAECs), liraglutide reduces the protein expression of VCAM-1 in response to TNFα and LPS stimulation and increases intracellular calcium and cAMP levels, leading to the phosphorylation of AMPK, which is an evolutionarily conserved fuel and stress-sensing enzyme that can be activated by calmodulin dependent protein kinase kinase-b (CAMKKb), and then downregulates monocyte adhesion, which presumably inhibits atherogenesis [
38,
39]. Second, in a cellular model of human umbilical vein endothelial cells (HUVECs) obtained from umbilical cords of women affected by GDM, liraglutide exposure significantly mitigated TNFα induced endothelial monocyte adhesion as well as VCAM-1 and ICAM-1 expression, reduced the phosphorylation of the MAPK42/44 signalling pathway, inhibited NF-kB p65 nuclear translocation, and decreased peroxynitrite levels and endothelial microvesicle (EMV) release [
40]. Third, after 8-week metformin treatment, TC and LDL-C were remarkably lower compared to baseline, but no significant differences were found between 120 min values and fasting values in ΔTC and ΔLDL-C. Moreover, no significant differences were found in plasma TC and LDL-C after liraglutide treatment. We presumed that liraglutide exert antiatherogenic mechanism, which might be not related to a decrease of LDL-C levels but might be related to an enhancement of HDL anti-inflammatory capacity [
41].
We carried out a Spearman rank correlation assay and the data showed that the reduction of 120 min NEFA (ΔNEFA) was positively correlated with the decrease of sVCAM-1 (ΔsVCAM-1) after 8-week liraglutide treatment. Our results suggest that liraglutide has beneficial effects on suppressing plasma NEFA as well as sVCAM-1 levels, which is likely to indicate that liraglutide could protect the endothelia by inhibiting monocyte cell adhesion and sVCAM-1 activation and reducing lipocyte oxidative stress and free fatty acid production, revealing an antiatherogenic effects [
42,
43].
The strengths of this study include the randomized, parallel, active comparator design and the similarities between the liraglutide and metformin groups at baseline. To the best of our knowledge, this is the first study to verify that liraglutide treatment is more effective than metformin treatment for recent-onset type 2 diabetes mellitus in reducing NEFA after glucose intake and suppressing sVCAM-1 levels at the same time. However, this study has certain limitations. First, it has a non-blinded design, lacks a non-treatment control group, and has a small sample size for the study. Second, further studies are needed to reveal the relevant signalling pathways by which liraglutide might influence glucose and lipid metabolism, regulate vascular endothelial function, and antagonize atherosclerosis.
Authors’ contributions.
XC conceived and designed the research. XC, LW, WZ, YT, CC and CQ performed the experiments. XC and WZ analysed the data. XC drafted the manuscript. All authors contributed to the writing of the manuscript. All authors read and approved the final manuscript.