JCEM ONLINE Brief Report Endocrine Research Distinct Changes in Serum Fibroblast Growth Factor 21 Levels in Different Subtypes of Diabetes Yang Xiao, Aimin Xu, Lawrence S. C. Law, Cheng Chen, Hui Li, Xia Li, Lin Yang, Shiping Liu, Zhiguang Zhou, and Karen S. L. Lam Diabetes Center (Y.X., H.L., X.L., L.Y., S.L., Z.Z.), Institute of Metabolism and Endocrinology, Second Xiangya Hospital, Key Laboratory of Diabetes Immunology, Ministry of Education, Central South University, Changsha, Hunan 410081, China; Departments of Medicine (A.X., L.S.C.L., C.C., K.S.L.L.) and Pharmacology and Pharmacy (A.X.) and Research Center of Heart, Brain, Hormone, and Healthy Aging (Y.X., A.X., K.S.L.L.), University of Hong Kong, Hong Kong Aims: Fibroblast growth factor (FGF) 21 is an endocrine factor with multiple beneficial effects on glucose and lipid metabolism in animals. This study aimed to investigate the association of serum FGF21 levels with type 1 diabetes, latent autoimmune diabetes in adults (LADA) and type 2 diabetes. Methods: Serum FGF21 levels were determined by ELISA in patients with type 1 diabetes (n 76), LADA (n 68), type 2 diabetes (n 77), and their age- and sex-matched controls. The association of serum FGF21 with markers of autoimmunity was studied. Results: In type 1 diabetic patients, serum FGF21 levels were significantly lower than controls [108.3 (61.5 180.1) vs. 196.0 (103.7 330.9) pg/ml, P 0.001]. In LADA patients, serum FGF21 levels were significantly lower than controls after adjustment for body mass index [210.9 (121.4 441.6) vs. 268.3 (159.5 443.6) pg/ml, P 0.003]. By contrast, serum FGF21 levels in type 2 diabetic patients were significantly higher than controls [381.2 (244.7 531.3) vs. 301.4 (173.9 444.2) pg/ml, P 0.006]. FGF21 levels increased progressively from type 1 diabetes, LADA, to type 2 diabetes (P 0.001 for global trend). Furthermore, FGF21 levels correlated inversely with titers of glutamic acid decarboxylase and insulinoma-associated protein 2 autoantibodies in type 1 diabetic and LADA patients. Conclusions: Serum FGF21 level is increased in type 2 diabetes but decreased in type 1 diabetes and LADA. In autoimmune diabetes, the reduction in circulating FGF21 is closely associated with markers of pancreatic -cell autoimmunity. (J Clin Endocrinol Metab 97: E54 E58, 2012) Fibroblast growth factor 21 (FGF21) is a member of the endocrine FGF subfamily and a major metabolic regulator (1). Despite the multiple salutatory effects of FGF21 on glucose and lipid metabolism in animals, circulating FGF21 levels are elevated in obesity (2) and patients with type 2 diabetes mellitus (3, 4), dyslipidemia (2, 5), impaired glucose tolerance (3), nonalcoholic fatty liver disease (6), and coronary heart disease (7). It has been proposed that the elevated level of FGF21 (hyper-fgf21-nemia) in obesity-related disorders is attributed to FGF21 resistance, a phenomenon reminiscent of hyperinsulinemia and insulin resistance (8). Although several previous studies have reported the elevated levels of FGF21 in type 2 diabetic patients (2 4), the ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright 2012 by The Endocrine Society doi: 10.1210/jc.2011-1930 Received July 2, 2011. Accepted September 22, 2011. First Published Online October 19, 2011 relationship of FGF21 with autoimmune diabetes remains unclear. In this study, we investigated the changes in serum FGF21 levels in patients with type 1 diabetes, latent autoimmune diabetes in adults (LADA), and type 2 diabetes, in comparison with their respective age- and sex-matched controls. In addition, we analyzed the association of serum FGF21 levels with markers of autoimmunity in our patients. Participants and Methods Participants Three groups of Chinese diabetic patients, within 1 yr of their disease diagnosis, and their respective age and sex-matched con- Abbreviations: BMI, Body mass index; FGF21, fibroblast growth factor 21; GADA, glutamic acid decarboxylase autoantibody; IA2A, insulinoma-associated protein 2 autoantibody; LADA, latent autoimmune diabetes in adults. E54 jcem.endojournals.org J Clin Endocrinol Metab, January 2012, 97(1):E54 E58
J Clin Endocrinol Metab, January 2012, 97(1):E54 E58 jcem.endojournals.org E55 trols were recruited. Diabetes was diagnosed according to American Diabetes Association criteria (9). Newly diagnosed type 1 diabetic patients (n 76) were recruited at the Diabetes Center of the Second Xiangya Hospital, Central South University. Type 2 diabetic patients (n 77) were enrolled from an ongoing clinical project (inclusion criteria shown in supplementary data). LADA patients (n 68) were selected from the LADA China Studies using the following inclusion criteria (10): 1) age at onset more than 30 yr; 2) glutamic acid decarboxylase autoantibody (GADA) positive; and 3) no ketosis within the first 6 months after diagnosis. Age- and, sex-matched controls for the type 1 diabetic group (n 77) were recruited from children in the community who came for health screening at the Children Health Center of the Second Xiangya Hospital, Central South University. Controls for LADA (n 69) and type 2 diabetes mellitus (n 75) were recruited from the China National Diabetes and Metabolic Disorders Study (inclusion criteria for controls shown in Supplementary Data, published on The Endocrine Society s Journals Online web site at http://jcem.endojournals.org). The study protocol was approved by the Institutional Review Board of Second Xiangya Hospital, Central South University. All study subjects provided written informed consent. Clinical and biochemical assessments After overnight fasting, a venous blood specimen was collected in the morning (around 0800 h) for analysis of various biochemical parameters (as shown in Supplementary Data). Serum FGF21 was measured using an ELISA kit established in our laboratory (Antibody and Immunoassay Services, the University of Hong Kong) (11). The titers of GADA and insulinoma-associated protein 2 autoantibody (IA2A) were determined using an in-house radioligand assay as previously described (12, 13). The cutoff titers of positivity for GADA and IA2A were 0.05 and 0.02, which were determined according to the 99th percentile of 188 and 171 healthy controls, respectively (14). The sensitivity and the specificity were 72 and 98% for GADA assay and were 66 and 99% for the IA2A assay, respectively. Statistical analysis All statistical analyses were performed with the Statistical Package for Social Science version 16.0 (SPSS, Inc., Chicago, IL). Data were expressed as mean SD or median with interquartile range as appropriate. Data that were not normally distributed, as determined using the Kolmogorov-Smirnov test, were logarithmically transformed before analysis. Correlations between FGF21 and biochemical variables were analyzed with Pearson correlation or partial correlation as appropriate. Data comparisons between the groups were performed using 2 tests for categorical variables and an unpaired t test or a univariate general linear model as indicated for continuous variables. Comparisons among groups were performed using one-way ANOVA or the univariate general linear model. P 0.05 was considered statistically significant. Results Serum FGF21 levels in different types of diabetes Characteristics of patients with type 1 diabetes mellitus (n 76), LADA (n 68), and type 2 diabetes mellitus (n 77), and their respective age- and sex-matched controls are described in Table 1. Among 76 type 1 diabetic patients, 60 were on insulin treatment. The disease duration was 16 (7 306]) (median and interquartile range) days. Compared with healthy controls, patients with type 1 diabetes mellitus had higher fasting glucose and lower body mass index (BMI), blood pressure, and fasting C-peptide levels. They also had significantly lower serum FGF21 levels compared with controls [108.3 (61.5 180.1) vs. 196.0 (103.7 330.9) pg/ml, P 0.001, Table 1]. After adjustment for BMI, serum FGF21 was still significantly lower in type 1 diabetic patients (P 0.004, Supplemental Fig. TABLE 1. controls Clinical and metabolic parameters of different groups of diabetic patients vs. age- and sex-matched T1DM LADA T2DM Parameters Cases Controls Cases Controls Cases Controls n 76 77 68 69 77 75 Age (yr) 11.5 4.8 11.5 6.5 51.4 10.9 51.8 5.3 55.3 8.9 54.9 8.3 Sex (men/women) 28/48 36/41 42/26 35/34 42/35 41/34 BMI (kg/m 2 ) 16.6 3.2 a 17.7 3.8 23.6 3.5 b 21.0 1.8 24.9 2.8 b 21.0 2.2 Waist circumference (cm) Men 62.3 9.9 61.9 12.5 85.7 10.1 b 72.6 7.2 87.9 7.2 b 73.9 7.2 Women 61.4 10.1 56.4 11.6 83.3 11.0 b 71.8 7.4 83.5 9.4 b 71.4 7.4 Systolic blood pressure (mm Hg) 98 13.5 b 110 12.1 124 21.0 b 112 11.1 123 18.6 b 114 10.9 Diastolic blood pressure (mm Hg) 63 10.3 a 67 8.3 79 11.9 b 72 8.1 76 11.5 c 71 7.3 Fasting glucose (mmol/liter) 11.5 7.0 b 4.5 0.6 9.2 4.5 b 4.2 0.5 8.1 3.1 b 4.3 0.5 Fasting C-peptide (pmol/liter) d 70.1 (34.5 175.8) b 271.4 (217.6 321.6) 158.3 (120.0 387.6) c 230.8 (157.1 311.4) 421.6 (320.7 593.8) b 297.3 (155.0 378.1) LDL cholesterol (mmol/liter) 2.3 0.9 2.2 0.6 3.0 1.1 b 2.1 0.5 3.1 1.0 b 2.2 0.5 HDL cholesterol (mmol/liter) 1.5 0.4 b 1.2 0.2 1.5 0.2 1.3 0.5 1.4 0.4 1.4 0.2 Triglycerides (mmol/liter) d 0.8 (0.5 1.1) 0.8 (0.6 1.1) 1.6 (1.0 2.8) b 0.8 (0.5 1.0) 1.5 (1.2 2.3) b 0.8 (0.6 1.0) FGF21 (pg/ml) d 108.3 (61.5 180.1) b 196.0 (103.7 330.9) 210.9 (121.4 441.6) 268.3 (159.5 443.6) 381.2 (244.7 531.3) c 301.4 (173.9 444.2) T1DM, Type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; HDL, high-density lipoprotein; LDL, low-density lipoprotein. a Compared with control, P 0.05. b Compared with control, P 0.01. c Compared with control, P 0.001. d Natural logarithm transformed before analysis.
E56 Xiao et al. FGF21 and Autoimmune Diabetes J Clin Endocrinol Metab, January 2012, 97(1):E54 E58 1A). In subjects with type 1 diabetes, hemoglobin A1c was 11.43 3.87%, indicating poor metabolic control. In addition, total serum protein levels were significantly lower in subjects with type 1 diabetes (60.74 5.65 vs. 67.17 4.72 g/liter in controls, P 0.001). Nevertheless, serum FGF21 remained significantly decreased in type 1 diabetic patients after adjusting for total protein levels (P 0.01). Among 68 LADA patients, 16 were on insulin treatment, 27 were on oral antidiabetic drugs, and six were on insulin and oral antidiabetic drugs. The disease duration was 48 (10 237) days. Compared with controls, patients with LADA had higher fasting glucose, BMI, waist circumference, blood pressure, triglycerides, and low-density lipoprotein cholesterol but lower fasting C-peptide. There was a moderate but not significant decrease in serum FGF21 in LADA patients compared with controls [210.9 (121.4 441.6) vs. 268.3 (159.5 443.6) pg/ml, P 0.274, Table 1]. After adjustment for BMI, serum FGF21 was significantly lower in LADA patients (P 0.003, Supplemental Fig. 1A). Among 77 type 2 diabetic patients, 46 were on oral antidiabetic drugs, three were on insulin treatment, and three were on insulin and oral antidiabetic drugs. The disease duration was 64 (1.5 161.5) days. Compared with controls, patients with type 2 diabetes mellitus had higher fasting glucose, BMI, waist circumference, blood pressure, triglycerides, low-density lipoprotein cholesterol, and C-peptide. Serum FGF21 levels were significantly increased in type 2 diabetic patients compared with controls [381.2 (244.7 531.3) vs. 301.4 (173.9 444.2) pg/ml, P 0.006, Table 1], and such a significance remained, even after adjustment for BMI (P 0.045, Supplemental Fig. 1A). FGF21 levels were increased from type 1 diabetes mellitus, LADA to type 2 diabetes mellitus (Supplemental Fig. 1B, P 0.001 for global trend). After adjustment for age and BMI, the increasing trend remained significant (P 0.039). In 34 type 1 diabetic patients, 28 LADA patients, and 76 type 2 diabetic patients who had available data on serum creatinine (type 1: 48.4 20.6; LADA: 83.5 18.5; type 2: 68.4 19.8 mol/liter). FGF21 levels still increased progressively from type 1 diabetes, LADA to type 2 diabetes (P 0.024 for global trend) after adjustment for age, BMI, and estimated glomerular filtration rate. Relationship of serum FGF21 with markers of autoimmunity In type 1 diabetic and LADA patients, FGF21 level correlated inversely with GADA and IA2A autoantibody titer (r 0.215, P 0.011; r 0.429, P 0.001, respectively, Fig. 1). FIG. 1. Correlations between serum FGF21 and titers of GADA (n 138) (A) or IA2A (n 101) (B) in type 1 diabetic and LADA patients. Ln, Natural logarithm. Discussion Although the function of FGF21 as a metabolic regulator has been extensively characterized in animals using both gain-of-functional and loss-of-functional approaches, its relevance to metabolic disorders in human remains poorly characterized. The present study provides the first evidence showing distinct changes in serum FGF21 levels in different subtypes of diabetes. Serum FGF21 levels are significantly increased in type 2 diabetic patients, modestly reduced in LADA patients, but markedly decreased in type 1 diabetic patients. Furthermore, our study is the first to demonstrate, in patients with type 1 diabetes or LADA, the link between serum FGF21 levels and circulating autoantibodies directed against the pancreatic -cell. The elevation in circulating FGF21 levels in type 2 diabetic patients may be attributed to a compensatory response, secondary to FGF21 resistance, which has been previously reported (3, 11). The decreased circulating FGF21 levels in type 1 diabetes and LADA may reflect the impaired -cell function in these two subtypes of autoimmune diabetes. This con-
J Clin Endocrinol Metab, January 2012, 97(1):E54 E58 jcem.endojournals.org E57 clusion was supported by our findings that circulating FGF21 was associated positively with fasting C-peptide (Supplemental Fig. 2) and negatively with titers of autoimmune antibodies related to -cell destruction. Notably, animal studies have demonstrated that FGF21 is highly expressed in the pancreas (15) and may act in an autocrine manner to protect the mass and function of pancreatic -cells (16). Rat Islets and INS-1E cells treated with FGF21 are partially protected from glucolipotoxicity and cytokine-induced apoptosis. In islets isolated from diabetic mice, FGF21 treatment increases islet insulin content and potentiates glucose-induced insulin secretion (16). In addition, FGF21 is expressed in pancreatic acinar cells, and its expression is markedly elevated during ceruleininduced pancreatitis (17). Both gain-of-function and lossof-function studies in mice showed that cerulein-induced pancreatitis-induced elevation of FGF21 expression in acinar cells represented a protective response to prevent pancreatic acini from overt damage. Taken together, these findings suggest that decreased circulating levels of FGF21 in autoimmune diabetes could be attributed in part to the reduced FGF21 production in pancreas, which in turn further aggravates islet -cell destruction. Nevertheless, whether the pancreatic islet is a potential source of circulating FGF21 and how islet-derived FGF21 is regulated remain to be investigated. Conflicting data have been reported with respect to the effect of insulin on FGF21 expression in humans (18 20). Two studies demonstrated that insulin stimulated FGF21 expression during hyperinsulinemic clamp (18, 19), whereas another study observed a significant increase in FGF21 during hypoinsulinemia (20). In our study, type 1 diabetic patients and some of the LADA patients had already received insulin replacement therapy, and fasting insulin level in these patients was similar to that of type 2 diabetic patients, suggesting that decreased circulating FGF21 in our patients with autoimmune diabetes was not due to low serum insulin levels. However, we cannot exclude the possibility that endogenous insulin has greater effects on FGF21 levels than exogenously administered insulin because circulating FGF21 levels correlated with fasting C-peptide (Supplemental Fig. 2), a marker of endogenous insulin secretion. Another notable observation of the present study is that the magnitude of decrease in circulating FGF21 in LADA patients is much smaller than in type 1 diabetic patients. This phenomenon may be due to the fact that LADA is only a mild form of autoimmune diabetes with a slow progression to pancreatic islet destruction and the levels of C-peptide are not as low as those in type 1 diabetic patients (Table 1). In addition, LADA patients share some common clinical features with type 2 diabetes, such as components of the metabolic syndrome and insulin resistance, which may cause a compensatory up-regulation of FGF21 in the liver. In summary, the findings from this study demonstrate that circulating FGF21 levels may be explored as a potential biomarker to discriminate different subtypes of diabetes. In patients with autoimmune diabetes, serum FGF21 level is associated positively with circulating C-peptide levels but negatively with the markers of -cell directed autoimmunity. The precise mechanism that leads to decreased FGF21 production in autoimmune diabetes remains to be further investigated. Acknowledgments Authors Y.X. and A.X. interpreted the data and drafted and revised the article. L.S.C.L. and C.C. analyzed and interpreted the data and revised the article. H.L., X.L., L.Y., and S.P.L. collected and interpreted the data. Z.G.Z. and K.S.L.L. supervised the studies, contributed to the discussion, and revised/edited the article. All of the authors approved the final version of the article. Address all correspondence and requests for reprints to: Karen S. L. Lam, Department of Medicine and Research Center of Heart, Brain, Hormone, and Healthy Aging, the University of Hong Kong, Hong Kong. E-mail: ksllam@hku.hk; or Zhiguang Zhou, Diabetes Center, Institute of Metabolism and Endocrinology, Second Xiangya Hospital, Key Laboratory of Diabetes Immunology, Ministry of Education, Central South University, Changsha, Hunan, China. E-mail: zhouzg@hotmail.com. This work was supported by Collaborative Research Fund (Grant HKU3/09C) from the Research Grant Council of Hong Kong (to K.S.L.L.), Seeding Fund for Basic Research and a matching fund for National 973 projects from the University of Hong Kong (to A.X.), Key Laboratory of Diabetes International Cooperation and Exchange of the National Natural Science Foundation of China (Grant No. 30831160518), the National Key Technology Research and Development Program for the 11th 5-yr plan (Grant 2006BA102B08), and the European Foundation for the Study of Diabetes (Grant EFSD/CDS/Lilly-2010) (to Z.Z.). Disclosure Summary: The authors declare that there is no duality of interest associated with this manuscript. References 1. Itoh N, Ornitz DM 2011 Fibroblast growth factors: from molecular evolution to roles in development, metabolism and disease. J Biochem 149:121 130 2. Zhang X, Yeung DC, Karpisek M, Stejskal D, Zhou ZG, Liu F, Wong RL, Chow WS, Tso AW, Lam KS, Xu A 2008 Serum FGF21 levels are increased in obesity and are independently associated with the metabolic syndrome in humans. Diabetes 57:1246 1253 3. Chavez AO, Molina-Carrion M, Abdul-Ghani MA, Folli F, Defronzo RA, Tripathy D 2009 Circulating fibroblast growth factor-21 is elevated in impaired glucose tolerance and type 2 diabetes
E58 Xiao et al. FGF21 and Autoimmune Diabetes J Clin Endocrinol Metab, January 2012, 97(1):E54 E58 and correlates with muscle and hepatic insulin resistance. Diabetes Care 32:1542 1546 4. Mraz M, Bartlova M, Lacinova Z, Michalsky D, Kasalicky M, Haluzikova D, Matoulek M, Dostalova I, Humenanska V, Haluzik M 2009 Serum concentrations and tissue expression of a novel endocrine regulator fibroblast growth factor-21 in patients with type 2 diabetes and obesity. Clin Endocrinol (Oxf) 71:369 375 5. Li H, Bao Y, Xu A, Pan X, Lu J, Wu H, Lu H, Xiang K, Jia W 2009 Serum fibroblast growth factor 21 is associated with adverse lipid profiles and gamma-glutamyltransferase but not insulin sensitivity in Chinese subjects. J Clin Endocrinol Metab 94:2151 2156 6. Li H, Fang Q, Gao F, Fan J, Zhou J, Wang X, Zhang H, Pan X, Bao Y, Xiang K, Xu A, Jia W 2010 Fibroblast growth factor 21 levels are increased in nonalcoholic fatty liver disease patients and are correlated with hepatic triglyceride. J Hepatol 53:934 940 7. Lin Z, Wu Z, Yin X, Liu Y, Yan X, Lin S, Xiao J, Wang X, Feng W, Li X 2010 Serum levels of FGF-21 are increased in coronary heart disease patients and are independently associated with adverse lipid profile. PLoS One 5:e15534 8. Fisher FM, Chui PC, Antonellis PJ, Bina HA, Kharitonenkov A, Flier JS, Maratos-Flier E 2010 Obesity is a fibroblast growth factor 21 (FGF21)-resistant state. Diabetes 59:2781 2789 9. 2007 Diagnosis and classification of diabetes mellitus. Diabetes Care 30(Suppl 1):S42 S47 10. Fourlanos S, Dotta F, Greenbaum CJ, Palmer JP, Rolandsson O, Colman PG, Harrison LC 2005 Latent autoimmune diabetes in adults (LADA) should be less latent. Diabetologia 48:2206 2212 11. Chen C, Cheung BM, Tso AW, Wang Y, Law LS, Ong KL, Wat NM, Xu A, Lam KS 2011 High Plasma Level of Fibroblast Growth Factor 21 Is an Independent Predictor of Type 2 Diabetes: A 5.4-year population-based prospective study in Chinese subjects. Diabetes Care 34:2113 2115 12. Petersen JS, Hejnaes KR, Moody A, Karlsen AE, Marshall MO, Høier-Madsen M, Boel E, Michelsen BK, Dyrberg T 1994 Detection of GAD65 antibodies in diabetes and other autoimmune diseases using a simple radioligand assay. Diabetes 43:459 467 13. Savola K, Bonifacio E, Sabbah E, Kulmala P, Vähäsalo P, Karjalainen J, Tuomilehto-Wolf E, Meriläinen J, Akerblom HK, Knip M 1998 IA-2 antibodies a sensitive marker of IDDM with clinical onset in childhood and adolescence. Childhood Diabetes in Finland Study Group. Diabetologia 41:424 429 14. Yang L, Luo S, Huang G, Peng J, Li X, Yan X, Lin J, Wenzlau JM, Davidson HW, Hutton JC, Zhou Z 2010 The diagnostic value of zinc transporter 8 autoantibody (ZnT8A) for type 1 diabetes in Chinese. Diabetes Metab Res Rev 26:579 584 15. Fon Tacer K, Bookout AL, Ding X, Kurosu H, John GB, Wang L, Goetz R, Mohammadi M, Kuro-o M, Mangelsdorf DJ, Kliewer SA 2010 Research resource: comprehensive expression atlas of the fibroblast growth factor system in adult mouse. Mol Endocrinol 24: 2050 2064 16. Wente W, Efanov AM, Brenner M, Kharitonenkov A, Köster A, Sandusky GE, Sewing S, Treinies I, Zitzer H, Gromada J 2006 Fibroblast growth factor-21 improves pancreatic -cell function and survival by activation of extracellular signal-regulated kinase 1/2 and Akt signaling pathways. Diabetes 55:2470 2478 17. Johnson CL, Weston JY, Chadi SA, Fazio EN, Huff MW, Kharitonenkov A, Köester A, Pin CL 2009 Fibroblast growth factor 21 reduces the severity of cerulein-induced pancreatitis in mice. Gastroenterology 137:1795 1804 18. Mai K, Andres J, Biedasek K, Weicht J, Bobbert T, Sabath M, Meinus S, Reinecke F, Möhlig M, Weickert MO, Clemenz M, Pfeiffer AF, Kintscher U, Spuler S, Spranger J 2009 Free fatty acids link metabolism and regulation of the insulin-sensitizing fibroblast growth factor-21. Diabetes 58:1532 1538 19. Hojman P, Pedersen M, Nielsen AR, Krogh-Madsen R, Yfanti C, Akerstrom T, Nielsen S, Pedersen BK 2009 Fibroblast growth factor-21 is induced in human skeletal muscles by hyperinsulinemia. Diabetes 58:2797 2801 20. Mai K, Schwarz F, Bobbert T, Andres J, Assmann A, Pfeiffer AF, Spranger J 2011 Relation between fibroblast growth factor-21, adiposity, metabolism, and weight reduction. Metabolism 60:306 311