Changes of serum retinol-binding protein 4 associated with improved insulin resistance after laparoscopic sleeve gastrectomy in Chinese obese patients

Background Serum retinol-binding protein 4 (RBP4) plays a critical role in insulin resistance. The mechanism behind the impact of laparoscopic sleeve gastrectomy (LSG) on glucose metabolism is unclear. Hence, we aimed to investigate the triangle relationship between the RBP4, glucose metabolism, and LSG in patients of Chinese ethnicity. Methods The study enrolled eighty-two obese patients. Glucose-lipid metabolic index, uric acid (UA), superoxide dismutase (SOD), free triiodothyronine (FT3), free thyroxin (FT4) and thyrotropin (TSH) were measured. RBP4 levels were detected by enzyme-link immunosorbent assay. 30 obese patients underwent LSG were studied. All these markers were measured again at a time interval of 3 and 6 months after surgery. Results (1) Circulating RBP4 levels were positively associated with body mass index(BMI), blood glucose in 0 min (BG0), BG30, BG120, BG180, fasting inulin(FINS), fasting C peptide(FCP), homeostasis model of assessment for insulin resistance index (HOMA-IR), SOD, TSH and negatively associated with Matsuda index in obesity with a significant difference (P < 0.05). RBP4 levels in the patients with impaired fasting glucose (IFG), insulin resistance or hyperinsulinemia were significantly higher than the patients without IFG, insulin resistance or hyperinsulinemia (P = 0.035, P = 0.001, and P = 0.007). (2) LSG resulted in significantly decreased FBG, FINS, FCP and HOMA-IR at 3, 6 months after surgery (all P < 0.05). The RBP4 levels were significantly decreased after surgery (all P < 0.05) with no gender difference. (3) The change in RBP4 levels was significantly associated with the change in FINS, FCP, HOMA-IR, and HOMA-β at 6 months and the change in TSH at 3 months after surgery in males (all P < 0.05). The change in RBP4 levels were significantly associated with the change in FINS, FCP, HOMA-IR, HOMA-β, and TCH at 3 months after surgery in females (all P < 0.05). Conclusions Overall, our results interpret the significant correlations between RBP4, glucose-lipid metabolism, oxidative stress and thyroid function in obese patients. Further, the LSG brings a decline in RBP4 levels and that may contribute partly to the improved insulin resistance in obese Chinese patients.

as dietary intervention and physical activity are regarded as first-line therapy. However, its efficacy is limited in severe or morbid obesity. The usage of weight-loss drugs are limited due to their side effects. Therefore, bariatric surgery will be a better treatment option for morbidly obese subjects or obesity with comorbidities. Previous studies have proven that bariatric surgery is effective in reducing body weight and improving glucose-lipid metabolism in obese patients [4][5][6][7].
Retinol-binding protein 4 (RBP4) is mainly released by the human liver and adipose tissue [8]. RBP4 is not just a plasma carrier of retinol but also considered as a novel adipokine that plays an important role in insulin resistance [9]. It is indicated that RBP4 concentrations are elevated in insulin-resistant mice and humans with obesity and diabetes and can be normalized by insulinsensitizing drugs [10]. Additionally, RBP4 has a function in the fat deposition by impairing the insulin pathway in mice [11]. A study on 1033 Chinese subjects with various degrees of obesity has shown that Serum RBP4 levels were positively correlated with visceral adipose tissue (P < 0.001) [12]. Several other studies exhibit the interrelation between RBP4, inflammation markers and oxidative stress [8,13].
RBP4 levels were decreased after weight loss, including bariatric surgery [14]. With context to this, a previous study showed that serum RBP4 levels are reduced after gastric banding surgery in 33 morbidly obese patients with body mass index (BMI) of 46 ± 5 kg/m 2 [14], and the decreased RBP4 may contribute to improved insulin resistance after banding surgery with weight loss [14]. Also, a study found that the RBP4 levels start to decrease within 6 months after surgical bypass in 190 morbid obesity [15]. After banding or bypass surgery, RBP4 is decreased by about 16.6% [16], and the change in RBP4 levels are related to reductions of waist circumference (WC), waist-hip ratio (WHR) and visceral-fat diameter (all P < 0.0.5) [16].
To the extent of our knowledge, there is no report about the change in RBP4 levels after laparoscopic sleeve gastrectomy (LSG) in obese patients of Chinese ethnicity. We hypothesis that change in RBP4 may associate with the improvement of glucose metabolism after LSG. Hence, this study was designed to investigate the association between RBP4 and glucose-lipid metabolic markers in patients of Chinese ethnicity and change in RPB4 levels after LSG.

Subjects
An observational study of 34 men and 48 women subjects with obesity were enrolled from the Endocrine and Metabolism Department of Shanghai Tenth People's Hospital. The obesity was defined by BMI over 30 kg/ m 2 . 10 healthy persons with normal BMI (average BMI 20.22 ± 1.15 kg/m 2 ) were included as controls. Among obese patients, 30 subjects (males:female = 10:20) underwent LSG. Inclusion criteria: BMI ranged from 31.16 to 52 kg/m 2 and age ranged from 18 to 60 years old which meets the recommended cutoff for bariatric surgery in Asians. Exclusion criteria: (1) secondary cause of obesity: hypothalamic obesity, Cushing syndrome, and hypophysis dysfunction, etc., (2) pregnancy or lactation, (3) taking antipsychotic medication, (4) presence of severe complications of obesity and diabetes, (5) Gastrointestinal diseases that contraindications for laparoscopic surgery, such as intra-abdominal infection, adhesions, etc., (6) Severe organic and systemic diseases intolerant of surgery (e.g. malignant tumor, severe heart, liver and kidney dysfunction, mental illness, autoimmune disease, acute or chronic inflammation, etc.). These subjects were examined before the LSG and followed up at 3, 6 months after surgery. For all patients in this cohort, this was the primary bariatric procedure and none have required an additional bariatric operation. The operation was performed by a professional surgeon under standard procedures, and under general anesthesia induced with a tracheal cannula. After CO 2 pneumoperitoneum was established, a laparoscope was used to explore the abdominal cavity. Four other trocars were inserted through incisions alongside the bilateral rectus abdominis muscles, the right side of the midline, and below the costal margin. A LigaSure Atlas (Covidien, Inc.) was used to take down the inner side of the hemal arch from the greater curvature to the fundus of the stomach entirely. A stent was planted in the lesser curvature, and an Echelon (Ethicon Endo-Surgery, Inc.) was used to cut and close the stomach from the pylorus ring to the fundus of the stomach several times to make the stomach like a tube. A Stratafix (Ethicon Endo-Surgery, Inc.) was used to make a continuous suture in the greater curvature of the stomach. Then, the incision was checked to avoid bleeding, and part of the stomach from navel incision was taken out. Negative pressure drainage was applied in the splenic recess and led to the right incision. Informed consent was signed and obtained by all individual participants included in this study. This study was approved by the Shanghai Tenth People's ethical committee.

Anthropometric measurements
Anthropometric data including body weight and height were measured without shoes and with light clothing by professional staff. We calculated the BMI by weight in kilograms divided by height in meters squared. WC and hip circumference (HC) were also measured. WC Page 3 of 10 Wang et al. Diabetol Metab Syndr (2020) 12:7 was measured at the midway between the lower rib margin and the iliac crest. HC was measured at the top point around the buttocks. The WHR was calculated as HC divided by WC.

Laboratory tests
Venous blood samples were adopted from all the subjects at least fasting for 8 h at baseline and 3, 6 months after LSG in the subjects under LSG. The serum was centrifuged immediately and stored at − 80 °C. Lipid metabolic markers [total cholesterol(TCH), triglyceride (TG), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), free fatty acid (FFA)], uric acid (UA) and superoxide dismutase (SOD) were measured by Roche Cobas c701 fully automatic biochemical analyzer. Free triiodothyronine (FT3), free thyroxin (FT4) and thyroid-stimulating hormone (TSH) were measured by ADVIA Centaur XP Immunoassay System. A 75 g-oral glucose tolerance test (OGTT) was performed to estimate glucose metabolism and insulin clearance according to the methods demonstrated by World Health Organization [17]. The glucose, insulin and C-peptide levels at 0, 30, 60, 120 and 180 min were measured. Glucose was measured by by Roche Cobas c701 fully automatic biochemical analyzer, insulin and C-peptide were measured by Roche Cobas e 601 analyzer. The HOMA of insulin resistance (HOMA-IR) was calculated to assess insulin resistance by the following formula: fasting serum insulin concentration (uIU/mL) × fasting blood glucose concentration (mmol/L)/22.5 [18]. HOMAβ was calculated to evaluate beta-cell function: HOMAβ = 20 × fasting insulin/(fasting glucose − 3.5) [18]. Insulin sensitivity was evaluated by the Matsuda index. Matsuda index was calculated as the following formula: Matsuda [19]. Serum retinol-binding protein 4 was detected by enzymelink immunosorbent assay (R&D Systems, Catalog Number DRB400).

Statistical analysis
SPSS20.0 statistical software was used to analyze the data. Descriptive statistics were expressed as mean ± standard deviation (X ± SD). Count data were expressed as number (n). An independent sample-test was used to compare the data between two independent groups. Statistical significance before and after surgery was evaluated using the paired two-tailed t-test. Nonparametric tests were used for non-normal distributed data. Pearson's and Spearman's correlation coefficient was used for correlations between RBP4 levels and other markers. The difference was considered significant at a P value of less than 0.05.

Improved metabolism and decreased serum RBP4 levels after surgery
Weight, BMI, WC, HC and, TG were significantly decreased in males and females at 3, 6 months after surgery (all P < 0.05). TG at 3 and 6 months, UA at 6 months, SOD at 3 months after surgery were significantly decreased and HDL-C at 3, 6 months after surgery were significantly increased in females (all P < 0.05). SOD and HDL-C were significantly increased at 6 months after surgery in males (all P < 0.05). All the results were showed in Table 2. LSG led to significantly decreased fasting glucose, FINS, and FCP at 3, 6 months after surgery (all P < 0.05) as shown in Fig. 3. HOMA-IR at 3, 6 months in both genders, HOMA-β at 3 months in females and at 6 months in males were significantly decreased after surgery (all P < 0.05) as presented in Fig. 4. Additionally, RBP4 levels were significantly decreased in all subjects after surgery with a significant difference (all P = 0.002). RBP4 levels were significantly decreased from 43.50 ± 12.11 to 31.6 ± 4.74 mg/L at 3 months, and to 28.83 ± 13.74 mg/L at 6 months after surgery in males (P = 0.020 and P = 0.017). RBP4 levels were significantly decreased from 43.07 ± 15.48 to 34.79 ± 11.76 mg/L at 3 months, and to 32.00 ± 11.94 mg/L at 6 months after surgery in females (P = 0.037 and P = 0.027). It need be mentioned that RBP4 levels were slightly decreased from 3 months to 6 months after surgery without statistic difference in females, males and all subjects with obesity (all P > 0.05). All these results were showed in

Change in RBP4 levels associated with a change in metabolic markers after surgery
In males, the change in serum RBP4 levels were significantly associated with a change in FINS, FCP, HOMA-IR and HOMA-β at 6 months and were associated with a change in TSH at 3 months after surgery (all P < 0.05).
In females, the change in serum RBP4 levels were significantly related to change in FINS, FCP, HOMA-, HOMA-β and TCH at 3 months after surgery (all P < 0.05). The change in serum RBP4 levels were not significantly associated with the change in lipid profile while were significantly associated with FCP at 6 months in all obesity (all P > 0.05). All results are presented in Table 3.

Discussion
As a metabolic risk factor in obesity, RBP4 has been reported to be associated with insulin resistance and adipose accumulation. RBP4 levels are elevated in obesity and are positively associated with BMI and WHR [22][23][24][25][26][27]. In our study, RBP4 levels were also significantly higher in obese patients than healthy controls with normal BMI and RBP4 levels were also significantly associated with BMI in obesity. Additionally, RBP4 levels may decrease after weight loss. A study proved that RBP4 levels were reduced 6 months after gastric banding in morbidly obese subjects [14]. Decreased RBP4 is associated with changes in BMI, HOMA-IR and TCH (all P < 0.05) [14]. Another study compared the SG to roux-en-y  gastric bypass(GBP) in subjects with diabetes, the results showed that RBP4 was significantly decreased as early as 3 days after GBP [28]. By taking these facts into consideration, in this study, we have investigated the relationship between RBP4 and the other metabolic markers and the change of its levels after LSG. We demonstrated a marked decrease in serum RBP4 levels at 3 and 6 months after LSG. To our knowledge, this study is the first study to assess the change in RBP4 serum levels after LSG.
RBP4 was mostly secreted by the human liver and adipose tissue as an adipokine [8]. RBP4 contributes to the pathogenesis of type 2 diabetes. A study by Kahn BB has first documented the importance of RBP4 in insulin resistance [10]. Adipose-specific GLUT4 knockout [adipose-Glut4 (−/−)] mice show insulin resistance is related to increased RBP4 expression [10]. Additionally, the overexpression of RBP4 or injection of recombinant RBP4 in wild-type mice leads to the development of insulin resistance [10]. In this study, RBP4 levels were significantly and positively associated with BG0, BG30, BG120, BG180, INS0, CP0, HOMA-IR and HGB in females with obesity, and also significantly and positively associated with HOMA-IR and negatively with Matsuda index in males with obesity. Additionally, RBP4 levels of the patients with IFG, insulin resistance or hyperinsulinemia were significantly higher than the patients without IFG, insulin resistance or hyperinsulinemia. These results are consistent with previous studies that show RBP4 is significantly related to glucose metabolism in the human [10].
In addition to the influence of RBP4 on glucose metabolism, it also has a specific role in lipid metabolism. A previous study proved that elevated serum RBP4 is associated with increased serum TG levels and decreased HDL-C levels [27]. In this similar pattern, our study results show RBP4 levels were significantly and positively associated with TCH and TG in females with obesity but not in males. Additionally, RBP4 levels of patients with hypertriglyceridemia or hypercholesterolemia were significantly higher than patients with normal TG or TCH in females. Therefore, we can conclude that RBP4 interference with lipid metabolism is gender-dependent.
As a novel adipokine secreted by adipocyte, RBP4 is significantly higher in obese subjects than adults with normal weight [29,30]. RBP4 levels are also associated with oxidative stress and inflammatory markers  [13,31]. There exists a positive association of RBP4 levels with oxidative stress markers such as SOD in males and females with obesity in this study. An interesting study demonstrates that plasma RBP4 levels are significantly associated with UA in the Chinese population [32].
In conclusion, RBP4 levels were found significant with serum UA levels in females with obesity. TSH levels present a positive association with BMI and are increased in obese patients when compared to the control group with normal weight [33]. Obese patients with the complication of mild thyroid hormone deficiency exert more serious metabolic disturbance [34]. However, few studies involved studying the association of RBP4 levels with thyroid function in obesity. A previous study showed that RBP4 levels were positively associated with TSH in postmenopausal women [35]. RBP4 levels are higher in patients with clinical hypothyroidism than controls [36]. In this study, RBP4 levels were significantly and positively associated with TSH in obesity and the change in serum RBP4 levels were associated with change in TSH in 3rd month after LSG in males. Hence, we suggest that RBP4 may play a role in thyroid dysfunction in obesity.
Metabolic surgery is more and more popular worldwide for reducing body weight and improving metabolism. Insulin resistance is significantly improved earlier than weight loss after surgery [37]. In our study, body weight was significantly reduced and the metabolism was significantly improved at 3, 6 months after LSG. Glucose metabolism improved especially accompanied by decreased fasting glucose, FINS, FCP, HOMA-IR and HOMA-β after surgery. However, the mechanism by which bariatric surgery improves glucose metabolism is unclear. Change of hormone secretion in the gastrointestinal tract may be one of those mechanisms [38,39]. RBP4 plays an important role in insulin resistance, and the improved glucose metabolism after LSG may be related to RBP4. A previous study illustrated the changes in RBP4 levels after gastric banding in morbidly obese subjects [14]. Another study also confirmed the change in RBP4 levels after GBP as early as 3 days [28]. While, they did not found change in RBP4 levels at 3 days after SG [28]. Therefore, we investigated the change in RBP4 levels after LSG with follow up time of 6 months. The results of our study showed that RBP4 levels were significantly decreased at 3, 6 months after surgery in males and females. Therefore, there exists change in RBP4 levels with long term after SG. RBP4 levels were closely associated with insulin resistance. Additionally, we found that the change in serum RBP4 levels were associated with change in FINS, FCP, HOMA-IR and HOMA-β at 6th month after LSG in males, and the change in serum RBP4 levels were related to change in FINS, FCP, HOMA-IR, and HOMA-β at 3rd month after LSG in females. Therefore, we assume that decreased RBP4 levels may contribute to improved glucose metabolism after LSG. There also some limitations to our study. The sample size is relatively smaller, and the follow-up time is relatively short. We will further expand the sample size and extend the follow-up time to observe the change in RBP4 for the long-term.

Conclusion
Overall, we relate the association between RBP4 with glucose metabolism, oxidative stress and thyroid function in patients of Chinese ethnicity. After the LSG surgery, the patients have shown improved glucose metabolism and decreased RBP4 levels in the 3rd and 6th months. Our study also suggests, decreased RBP4 levels may partly account for the improved glucose metabolism in obese patients after LSG. Future in-depth analyses on this area could project LSG as the most reliable treatment method for obesity and RBP4 as a trustable biomarker.