- Email: [email protected]
Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus
Accepted Manuscript Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus Pei Wang, He-hong Ma, Xiu-zhen Hou, Li-li Song, Xiao-long Song, Jun-feng Zhang PII: DOI: Reference:
S0168-8227(18)30613-2 https://doi.org/10.1016/j.diabres.2018.05.038 DIAB 7399
To appear in:
Diabetes Research and Clinical Practice
Received Date: Revised Date: Accepted Date:
16 April 2018 11 May 2018 22 May 2018
Please cite this article as: P. Wang, H-h. Ma, X-z. Hou, L-l. Song, X-l. Song, J-f. Zhang, Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus, Diabetes Research and Clinical Practice (2018), doi: https://doi.org/10.1016/j.diabres.2018.05.038
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus Pei Wang1, He-hong Ma1, Xiu-zhen Hou1, Li-li Song1, Xiao-long Song1, Jun-feng Zhang1＃ Obstetrical Department 1, Cangzhou Central Hospital, Cangzhou, Hebei, China, 061000 ＃Correspondence
to: Jun-feng Zhang, Email: [email protected]; Tel: +86-451-82576735
Address: No. 16, Xinhua West Road, Cangzhou 061000, P. R. China Running title: Irisin Levels and Risk of GDM Financial Disclosure The authors did not report any potential conflicts of interest. Each author has indicated that he or she has met the journal’s requirements for authorship.
1
ABSTRACT Background: The aim of this prospective cohort study was to investigate the association of first trimester irisin concentrations and the subsequent development of gestational diabetes mellitus (GDM). Methods: This cohort study was conducted at three maternity centers in China from July 2015 to June 2016. Data for fasting plasma glucose (FPG) and irisin concentrations in the first trimester and one-step GDM screening with 75-g oral glucose tolerance test (OGTT) performed between 24 and 28 weeks of gestation were collected and analyzed. Results: Plasma from women was available for 1150 women, of whom 135 (11.7%) developed GDM. The median value of irisin in those included women was 141.2 (IQR, 99.4–192.9) ng/ml. In multivariate models comparing the first(Q1), second(Q2) and third(Q3) quartiles against the fourth(Q4) quartile of irisin, levels of irisin in Q1 and Q2 were associated with GDM, and increased risk of GDM by 440% (odds ratios[OR]=5.40; 95% confidence intervals[CI]: 2.35-11.40) and 283% (OR: 3.83; 95%CI: 1.63-8.01). A model containing known risk factors plus irisin compared with a model containing known risk factors without irisin showed a greater discriminatory ability to predict GDM, the area under the curve (AUC) increased from 0.776 to 0.809. A significant difference in the AUC between the clinical variables alone and the addition of irisin level was observed (difference, 0.034; P=0.03). Conclusions: Reduced plasma levels of irisin in first trimester was associated with the increased risk of GDM and might be useful in identifying women at risk for GDM for early prevention strategies. Key words: Irisin, gestational diabetes mellitus, association, Chinese
Introduction 2
Gestational diabetes mellitus (GDM) is a first diagnosed diabetes during pregnancy defined as an abnormal glucose tolerance in various degree that may result in negative outcomes for mothers and their fetus and newborn [1]. Over the past years, considering its gradual upward trend in China (the adjusted prevalence of GDM increased by 2.8 times during 1999–2008, from 2.4 to 6.8%) [2], GDM has become a public health priority on the identification of potential predictors of glucose intolerance. Irisin is a small polypeptide hormone that is cleaved from the fibronectin type III domain-containing 5 (FNDC5). The plasma level of irisin is commonly increased by exercise [3]. Irisin is a novel myokine and adipokine that induces an increase in total body energy expenditure, improving insulin sensitivity and glucose tolerance in experimental animals [4]. Accumulated evidence also has demonstrated that irisin contributes to the regulation of glucose and lipid metabolism in skeletal muscle and adipose tissue [4-5]. Another study indicated that FNDC5/irisin ameliorates glucose/lipid metabolic derangements and insulin resistance in obese mice, and enhanced lipolysis via cAMP–PKA–HSL/perilipin pathway [6]. Furthermore, higher irisin levels observed in overweight/obese conditions could be a protective response of organism to early glucose impairments [7]. In healthy subjects aged 40-60 who were first-degree relatives (FDRs) of type 2 diabetes mellitus (T2DM) patients but without previous diagnosis of T2DM, Yang et al. [8] found that serum irisin level was closely related to homeostatic model assessment (HOMA)-β in normal glucose tolerance (NGT), suggesting that irisin may play a crucial role in pancreatic β-cell function. Recent clinical studies in humans have suggested that irisin plays substantial roles in the pathophysiology of obesity [9], insulin resistance [10], type 2 diabetes mellitus [11] and the metabolic syndrome [12]. Although some studies had reported that individuals with type 2 diabetes had lower levels of irisin as compared with nondiabetic controls [13-14], others had reported a positive association between circulating irisin and fasting plasma glucose [15-16] or muscle FNDC5 mRNA expression and 2 h glucose during an oral glucose tolerance test (OGTT) [17]. However, the roles of irisin in GDM were far from being completely understood. Several studies had reported reduced irisin levels in women with GDM [18-21] or in those whom GDM was later developed [22], compared to those who did not develop GDM at all. 3
However, Ebert et al. [23] found no significant different serum irisin levels between patients with GDM and the controls (P=0.616). The differences between those studies require further research to confirm. Thus, the aim of this prospective cohort study was to investigate the association of first trimester irisin concentrations and the subsequent development of GDM. Research design and methods This prospective, multi-center cohort study was conducted at three maternity centers from two cities (Harbin and Beijing) in China from July 2015 to June 2016. The sampling method was purposive sampling. Inclusion and exclusion criteria of the pregnant women have been reported previously [1]. In summary, singleton pregnant women (age≥18 years old) at the first prenatal visit of gestation were recruited. Exclusion criteria are the following: pre-gestational diabetes or diagnosed at the first prenatal visit; pregnancy-induced hypertension; preeclampsia at the first prenatal visit; have a history of childbirth, alcohol abuse; chronic diseases (liver and kidney system disease, autoimmune disease, neurological disorders, active or chronic inflammation); lost blood samples and termination of pregnancy in the follow-up. The study has been approved by the ethics committee of the Cangzhou Central Hospital prior to its initiation. All participants were informed of the study protocol, and written informed consents were obtained before inclusion. Maternal–pregnancy characteristics (maternal age, pre-pregnancy body mass index[BMI], ethnicity, smoking status, mode of conception, gestational weeks at sampling, history of spontaneous abortion, first-degree relative with diabetes and medical history of hypertension, polycystic ovary syndrome or cardiovascular disease[CVD]) were collected and recorded at first prenatal visit. BMI was calculated as the weight in kilograms divided by the height in meters squared. Specific details about physical activity were assessed by a validated questionnaire adapted from the Canadian Community Health Survey(CCHS) [24]. The physical activity index is derived from a variety of physical activities undertaken by respondents to the CCHS. For each leisure time activity an average daily energy expenditure was calculated based on frequency, duration, and intensity of participation in those leisure time activities. The energy expended for all activities is combined to calculate an average daily energy expenditure. Respondents were classified as ACTIVE if average daily energy expenditure was 3 kcal/kg/day, MODERATELY ACTIVE with an energy expenditure between 2.9 and 1.5 kcal/kg/day, and INACTIVE below a 4
value of 1.5 kcal/kg/day. During the first prenatal visit in the hospital, involved subjects were tested for fasting plasma glucose (FPG) and irisin using venous blood sample collected after at least 8 h of fasting. In accordance to the principle of blind method and prior to the understanding of the status of GDM, serum irisin concentrations were measured in duplicate by using the enzyme-linked immunosorbent assay (ELISA) kits (Aviscera Biosciences, Santa Clara, CA), in accordance with the manufacturer's instructions. The sensitivity of the assay was 1.0 ng/ml and the linear range of the standard was 1-500 ng/ ml. The intra-and inter-assay coefficients of variation (CV) were 5.2%-7.5% and 6.3%-8.8%, respectively. Routine blood biomarkers, for instance, triglyceride, cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), C-reactive protein (CRP) and FPG were tested using standard detection methods by BS800M (MINDRAY, Shenzhen, China). Plasma levels of insulin were measured using a chemiluminescence immunoassay method by DPC Immulite 2000 (Diagnostic Products Corporation, CA, USA). Between 24 and 28 weeks during pregnancy, women were informed to return to the hospital in their fasting state for repeat testing. A 75-g Oral Glucose Tolerance Test (OGTT) was performed once at this follow-up. Venous blood samples were collected at 0, 1, and 2 h after a 75-g glucose load. According to the criteria established by Ministry of Health (MOH) of China, diagnosis of GDM can be made when any one of the following values was met or exceeded in the 75-g OGTT: 0h (fasting) ≥5.10mmol/l; 1h≥10.00mmol/l; and 2h≥8.50mmol/l. Statistical Analysis Results were expressed as percentages for categorical variables and as medians (interquartile ranges[IQRs]) for the continuous variables. Univariate data on demographic and clinical features were compared by Mann-Whitney U-test or Chi-Square test as appropriate. In addition, 95% confidence intervals (CI) for incidence of GDM were calculated. Univariate and multivariate regression analyses were performed to investigate the relationship between concentrations of irisin and GDM. Crude models and multivariate models were developed for the adjustment of all associated factors and reported as odds ratios (ORs). For multivariate analysis, variables included the factors significant in univariate regression analysis (maternal age, BMI, pre-existing hypertension and cardiovascular disease [CVD], gestational age 5
at sampling, and blood levels of cholesterol, HDL, insulin, FPG and CRP). For a more detailed exploration of irisin and GDM, a multivariate analysis model was created to estimate adjusted ORs and 95% confidence intervals (CI) of GDM for irisin quartiles (with the highest quartile as the referent). In addition, plasma irisin concentration was also dichotomized according to irisin quartiles, and the grouping of quartiles 2-4 was defined as normal while the grouping of quartile 1 as low. In addition, a propensity score matching (PSM) analysis was performed to reduce the confounding between covariates and irisin levels. Second, we compared different prognostic risk scores from different predictive models by calculating receiver operating characteristic (ROC) analysis. Thereby the area under the receiver operating characteristic curve (AUC) is a summary measure over criteria and cut-point choices. The AUC summary equals the probability that the underlying classifier will score a randomly drawn positive sample higher than a randomly drawn negative sample. To test whether the irisin level improves score performance, we considered the two nested logistic regression models with irisin and known risk factors (maternal age, BMI, gestational age at sampling, smoking, ethnicity, pre-existing hypertension and CVD, history of spontaneous abortion, gestational weeks at admission, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP) as compared with known risk factors only. Furthermore, integrated discrimination improvement (IDI) and net reclassification improvement (NRI) indices were calculated to investigate the utility of adding irisin to recognized risk factors for GDM [25]. All statistical analyses were performed with SPSS for Windows, version 21.0 (SPSS Inc., Chicago, IL, USA), ROCR package (version 1.0-2), Stata release 14 (Stata Corp, College Station, TX), and GraphPad Prism 5.0. Statistical significance was defined as P<0.05. Results Plasma levels of irisin were available for 1150 women, with a median value of 141.2 ng/ml (IQR, 99.4–192.9 ng/ml). The maternal and clinical characteristics were presented in Table 1. As shown in Table 1, low levels of irisin were more likely associated with obesity, pre-existing hypertension, pre-existing CVD, and high physical activity, and higher levels of insulin, FPG, CRP, and triglyceride. In addition, levels of irisin were positively associated with HDL. Furthermore, the maternal and clinical characteristics of the included women with and without GDM were 6
presented in Sup table I. Among the 1150 included women, GDM was developed in 135 (11.7%). Irisin levels at first visit were significantly lower in women who developed late GDM than those who did not develop it at all [106.5(IQR, 73.8-136.9) ng/ml VS.148.4(107.6-203.5) ng/ml], as shown in Figure 1. The GDM distribution across the irisin quartiles ranged between 22.2% (first quartile) and 2.4% (fourth quartile), figure 2. In univariate logistic regression analysis, we calculated the odds ratio (OR) of irisin level to predict GDM as compared with other risk factors. With an unadjusted OR of 0.983 (95% CI, 0.977–0.987; P<0.001), irisin had a strong association with GDM. In multivariate models comparing the first(Q1), second(Q2) and third(Q3) quartiles against the fourth(Q4) quartile of irisin (Table 2), levels of irisin in Q1, Q2 and Q3 were associated with GDM, and increased risk of GDM by 440% (OR=5.40; 95%CI: 2.35-11.40), 283% (OR: 3.83; 95%CI: 1.63-8.01) and 60% (OR: 1.60; 95%CI: 0.62-4.18). The independent association of irisin with GDM was confirmed using the likelihood ratio test (P=0.001). In a multivariate model using the Q1 of irisin vs. Q2-4 together with the clinical variables, the marker displayed prognostic information and increased risk of GDM by 89% (OR for Q1, 1.89 [95% CI, 1.16-2.84]). In the entire cohort, there was a certain amount of bias between the two groups. The women with GDM have higher levels of insulin. Therefore, a propensity score was calculated for each woman (N=135) based on above identified variables. Interestingly, irisin still had an association with GDM (OR=0.988, 95%CI: 0.982-0.995; P=0.009) and pseudo R2 showed that we explained 28.9 percent of variation. Area under ROC curve is calculated as 0.75. Using ROC curves, irisin levels with 155ng/ml at first visit predicted the development of GDM with the highest sensitivity and specificity [67.6% and 70.7%, respectively; area under the curve (AUC)=0.723; 95% CI, 0.683–0.763]. With an AUC of 0.723, irisin showed a significantly greater discriminatory ability to predict GDM as compared with CRP (AUC, 0.618; 95% CI, 0.559–0.672; P<0.001), FPG (AUC, 0.685; 95% CI, 0.631–0.729; P=0.017) and insulin (AUC, 0.669; 95% CI, 0.616–0.705; P=0.003). In addition, a model containing known risk factors plus irisin compared with a model containing known risk factors without irisin showed a greater discriminatory ability to predict 7
GDM, the area under the curve (AUC) increased from 0.776 to 0.809 (95% confidence interval [CI], 0.763-0.854). A significant difference in the AUC between the clinical variables alone and the addition of irisin level was observed (difference, 0.034 [95% CI, 0.029-0.041]; P=0.03), table 3. The NRI statistic showed that the addition of irisin to established risk factors significantly increased the correct reclassification of GDM women and non-GDM women(P=0.004). The IDI statistic also found that the irisin level significantly increased the discrimination between GDM women and non-GDM women (P=0.04). Discussion Irisin is an exercise-regulated myokine inducing browning of white adipose tissue and has gained interest as a potential new strategy to combat obesity and its associated disorders, such as type 2 diabetes mellitus [11]. To the best of our knowledge, the present study is the first report to investigate the predictive value of irisin for GDM in a substantial cohort of Chinese women from three maternity centers. Maternal irisin levels at first visit were significantly lower in women who developed late GDM compared with those who did not. Data confirms evidence linking lower circulating irisin concentrations with signs of GDM and adds significant predictive information to the established risk factors. Furthermore, the addition of irisin to established risk factors improved the NRI and the IDI. Several studies found a positive association of irisin levels with BMI, insulin resistance (IR), and metabolic syndrome (MetS) [16, 26], whereas others reported negative correlations [13, 18-19, 27]. A significant association between circulating irisin and a favorable lipid profile in the general population was found [28], while another study found that circulating irisin levels were positively associated with parameters related to an increased cardiometabolic risk [29]. Previous study showed that plasma irisin levels appear to be associated with important metabolic factors (including BMI and fasting blood glucose) in non-diabetic subjects but not in individuals with type 2 diabetes [30]. Furthermore, some previous studies also reported that irisin-encoding gene (FNDC5) variant could change blood pressure in men with type 2 diabetes [31], while irisin improved endothelial function in type 2 diabetes [32] and in a mouse model of obesity [33]. Future study regarding interaction between irisin and other factors might be helpful to allow better understanding of irisin. 8
In this study, a negative correlation between irisin and BMI, insulin, FPG and lipid metabolism had been found. We considered irisin as a protective biomarker for GDM. In addition, irisin still was a useful predictor for GDM after being adjusted for above confounding factors. Consistent with our findings, Erol et al. [22] also suggested that irisin might be a useful biomarker in early pregnancy to predict the development of GDM. We considered that irisin could be seen as a predictive or clinical marker. Furthermore, irisin showed a significantly greater discriminatory ability to predict GDM as compared with other risk factors, for instance, CRP, FPG and insulin. Thus, clinicians would collect blood samples and test irisin levels in the first trimester for all women. Investigations on irisin in circulation from women have drawn the attention from many previous studies. In this study, we found that median irisin levels were significantly lower in women who developed late GDM compared with those who did not [106.5(IQR, 73.8-136.9) VS.148.4(10.7.6-203.5); P<0.001]. The level of irisin in this study was lower than in previous studies (213.4ng/ml ~ 1703.3ng/ml [18-23]). These controversial results might mostly be due to the small number of sample sizes from previous studies (range from 20 to 130) [18-23], which could
affect
the
power
of
statistical
analysis
conducted.
Furthermore,
different measurement ELISA kits may also cause differences. Lastly, blood levels of irisin were tested at different gestation age. In this study, we tested at first prenatal visit, which were similar with Erol et al. [22]. However, different time-point, including at the time of birth [19], before and within 24 h after delivery [23] and at 26-30th weeks of gestation [18] were assessed in other studies. Those variations make it difficult to compare different studies. This observational study design does not allow drawing a causal conclusion. We could not assess whether irisin is the cause or result of GDM. However, there are possible mechanisms in which irisin might be associated with GDM. First, we found that low levels of irisin were more likely with higher levels of insulin. Thus, irisin might be a biomarkers of insulin resistance. However, in the multivariable models, irisin still associated with GDM after adjusted for insulin resistance. Therefore, irisin is a specific unique predictor of GDM rather than just another measure of insulin resistance. Second, irisin is a novel myokine and adipokine that improves insulin sensitivity and glucose tolerance. Previous studies speculated the underlying mechanisms linking 9
irisin with GDM through glucose utilization and lipid metabolism [4, 6-8]. Third, another possible explanation is an interplay with other myokines, adipocytokines, or classical cytokines [34]. Furthermore, adipokines [leptin, adiponectin, resistin, chemerin, and visfatin] and myokines [interleukin (IL)-8, IL-15, irisin, and myonectin] play a crucial role in the communication between adipose tissue, skeletal muscle, and other organs [35]. Skeletal muscle-derived factors “myokines” and adipose tissue-derived factors “adipokines” exert their effects mainly in the organs of adipose tissue and/or skeletal muscle to form the muscle-adipose axis that regulates body composition [36]. The relationship between adipokine [37], leptin [38], retinol-binding protein 4[39], resistin and IL-6 [40] and GDM had been proposed in previous studies. Furthermore, one study provides evidence of an association between irisin and homocysteine, which may be due to nicotinamide metabolism [41]. Homocysteine level as a significant risk factor for development of diabetes in women with previous GDM [42]. Lastly, endothelial dysfunction has been documented in patients with diabetes and in individuals with insulin resistance or at high risk for developing type 2 diabetes. Lower levels of irisin are independently associated with endothelial dysfunction [43], and irisin alleviates endothelial dysfunction in type 2 diabetes partially via reducing oxidative/nitrative stresses through inhibiting signaling pathways implicating PKC-β/NADPH oxidase and NF-κB/iNOS [32]. The following limitations of our study must be taken into account. First, lifestyle advice/referral to diet would be changed the levels of irisin in those pregnant women. However, we did not collect data on lifestyle advice/referral to diet. Thus, we could not determine the association of those factors with blood levels of irisin and GDM. Second, the observational study design does not allow drawing secure conclusions for a causal relationship. In addition, Hedderson et al. [44] suggested that high rates of gestational weight gain, especially early in pregnancy, may increase a woman’s risk of GDM. However, in this study, we only assess the pre-pregnancy BMI, and gestational weight gain. Thus, we the relationship between irisin, gestational weight gain and GDM cannot be assessed. Third, irisin is usually measured by ELISA, but the quantification varies greatly between the kits. These differences probably come from the variety in the irisin epitopes being targeted for measurement by the manufacturing companies. Thus, the results should be confirmed in similar studies using an Irisin-Elisa kit targeting other epitopes. In addition, it has 10
also been suggested that some other adipokines, including adipokine [37], leptin [38] and retinol-binding protein 4[39] were associated with GDM. However, in this study, we did not test those adipokines. Therefore, we cannot determine the relationship between irisin, GDM and those factors. Finally, most participants of the present study were Chinese Han (89.0%) and whether these observations can also be extended to other ethnic groups with different body composition remains to be determined. Interestingly, in this study, Manchu, Hui, Mongolian and Korean accounted for 3.1%, 2.5%, 1.7% and 1%. However, there is no significant difference in the incidence of GDM among different ethnicities. In addition, some women with alcohol abuse and chronic diseases were excluded. Those women were also susceptible to GDM. Thus, the issue of selection bias must be considered. Conclusion Research results suggested that the reduced plasma levels of irisin in first trimester independently predicted GDM and significantly improved reclassification and discrimination. Irisin shows a potential as a novel biomarker for GDM prediction in Chinese women. Further studies, which may open the way to the proposal of new therapeutic options, are proposed to confirm this association.
11
Abbreviations list GDM=Gestational diabetes mellitus BMI= Body mass index MS=Metabolic syndrome FPG=Fasting plasma glucose HDL=High-density lipoprotein LDL=Low-density lipoprotein CRP=C-reactive protein OGTT=Oral Glucose Tolerance Test IQR= Interquartile ranges CI=Confidence intervals OR=Odds ratios ROC=Receiver operating characteristic curves AUC=Area under the curve IDI=Integrated discrimination improvement NRI=Net reclassification improvement NGT=Normal glucose tolerance
12
Declarations Consent for publication Consent to publish had be obtained from all included women. Availability of data and material Please contact author for data requests Ethics approval and consent to participate Prior to the initiation of the study, it has been approved by the ethics committee of the Cangzhou Central Hospital. All participants or their relatives were informed of the study protocol, and written informed consents were obtained before inclusion. Acknowledgements The authors thank the nurses, physicians, and patients who participated in our study, and the staff of the central laboratories of the hospitals. We also thanks Dr. Xiao-Tong Shi (China Rehabilitation Research Center) and Dr. Lei Yu (Harbin Medical University) who help us collect blood samples and clinical information. Competing interests None Funding None Guarantor’s statement Dr. Zhang JF is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Authors' contributions Zhang JF had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Wang P, Ma HL, Hou XZ, Song LL, Song XL, Zhang JF Acquisition of data: Wang P, Ma HL, Hou XZ 13
Analysis and interpretation of data: Song LL, Song XL, Zhang JF Drafting of the manuscript: Wang P, Ma HL, Hou XZ, Song LL Critical revision of the manuscript for important intellectual content: Song XL, Zhang JF Administrative, technical, or material support: Song XL, Zhang JF Study supervision: Zhang JF
14
References 1. Tu WJ, Guo M, Shi XD, et al. The Association of First Trimester Serum Fatty Acid–Binding Protein 4 and Subsequent Gestational Diabetes Mellitus. Obstetrics & Gynecology; 2017, 130(5): 1011-1016. 2. Zhang F, Dong L, Zhang CP et al. Increasing prevalence of gestational diabetes mellitus in Chinese women from 1999 to 2008. Diabet Med 2011; 28: 652–657. 3. Huh J Y, Mougios V, Kabasakalis A, et al. Exercise-induced irisin secretion is independent of age or fitness level and increased irisin may directly modulate muscle metabolism through AMPK activation. The Journal of Clinical Endocrinology & Metabolism 2014; 99(11): E2154-E2161. 4. Xin C, Liu J, Zhang J, et al. Irisin improves fatty acid oxidation and glucose utilization in type 2 diabetes by regulating the AMPK signaling pathway. International journal of obesity, 2016, 40(3): 443-451. 5. Lee H J, Lee J O, Kim N, et al. Irisin, a novel myokine, regulates glucose uptake in skeletal muscle cells via AMPK. Molecular Endocrinology, 2015, 29(6): 873-881. 6. Xiong X Q, Chen D, Sun H J, et al. FNDC5 overexpression and irisin ameliorate glucose/lipid metabolic derangements and enhance lipolysis in obesity. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease, 2015, 1852(9): 1867-1875. 7. Huerta A E, Prieto-Hontoria P L, Fernandez-Galilea M, et al. Circulating irisin and glucose 15
metabolism in overweight/obese women: effects of α-lipoic acid and eicosapentaenoic acid. Journal of physiology and biochemistry, 2015, 71(3): 547-558. 8. Yang M, Chen P, Jin H, et al. Circulating levels of irisin in middle-aged first-degree relatives of type 2 diabetes mellitus—correlation with pancreatic β-cell function. Diabetology & metabolic syndrome, 2014, 6(1): 133. 9. Perakakis N, Triantafyllou G A, Fernández-Real J M, et al. Physiology and role of irisin in glucose homeostasis. Nature Reviews Endocrinology 2017; 13(6): 324-337. 10. Qiu S, Cai X, Yin H, et al. Association between circulating irisin and insulin resistance in non-diabetic adults: a meta-analysis. Metabolism, 2016, 65(6): 825-834. 11. Shoukry A, Shalaby S M, El‐Arabi Bdeer S, et al. Circulating serum irisin levels in obesity and type 2 diabetes mellitus. IUBMB life 2016, 68(7): 544-556. 12. Park K H, Kang J H, Lee H J, et al. Irisin and its relation to metabolic syndrome among Korean adolescents. Obesity Reviews, 2016, 17: 128. 13. Moreno-Navarrete JM, Ortega F, Serrano M, Guerra E, Pardo G, Tinahones F, et al. Irisin is expressed and produced by human muscle and adipose tissue in association with obesity and insulin resistance. J Clin Endocrinol Metab 2013; 98: E769–E778. 14. Choi YK, Kim MK, Bae KH, Seo HA, Jeong JY, Lee WK, et al. Serum irisin levels in new-onset type 2 diabetes. Diabetes Res Clin Pract 2013;100:96–101. 15. Huh JY, Panagiotou G, Mougios V, Brinkoetter M, Vamvini MT, Schneider BE, et al. FNDC5 and irisin in humans: I. Predictors of circulating concentrations in serum and plasma and II. mRNA expression and circulating concentrations in response to weight loss and exercise. Metab Clin Exp 2012; 61:1725–1738. 16. Park KH, Zaichenko L, Brinkoetter M, Thakkar B, Sahin-Efe A, Joung KE, et al. Circulating irisin in relation to insulin resistance and the metabolic syndrome. J Clin Endocrinol Metab 2013; 98:4899–4907 17. Staiger H, Bo¨hm A, Scheler M et al. Common genetic variation in the human FNDC5 locus, encoding the novel musclederived ‘browning’ factor irisin, determines insulin sensitivity. PLoS One 2013; 8: e61903. 16
18. Kuzmicki M, Telejko B, Lipinska D, et al. Serum irisin concentration in women with gestational diabetes. Gynecological Endocrinology, 2014, 30(9): 636-639. 19. Yuksel M A, Oncul M, Tuten A, et al. Maternal serum and fetal cord blood irisin levels in gestational diabetes mellitus. Diabetes research and clinical practice, 2014, 104(1): 171-175. 20. Zhao L, Li J, Li Z L, et al. Circulating irisin is lower in gestational diabetes mellitus[J]. Endocrine journal, 2015, 62(10): 921-926. 21. Du X, Jiang W. Lower circulating irisin level in patients with diabetes mellitus: a systematic review and meta-analysis. Hormone and Metabolic Research, 2016, 48(10): 644-652. 22. Erol O, Erkal N, Ellidağ H Y, et al. Irisin as an early marker for predicting gestational diabetes mellitus: a prospective study. The Journal of Maternal-Fetal & Neonatal Medicine, 2016, 29(22): 3590-3595. 23. Ebert T, Stepan H, Schrey S, et al. Serum levels of irisin in gestational diabetes mellitus during pregnancy and after delivery. Cytokine, 2014, 65(2): 153-158. 24. Anonymous. Canadian Community Health Survey, Cycle 2.2, Nutrition (2004): A Guide to Accessing and Interpreting the Data [Health Canada, 2006]. 25. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008; 27:157–172. 26. Jang H B, Kim H J, Kang J H, et al. Association of circulating irisin levels with metabolic and metabolite profiles of Korean adolescents. Metabolism, 2017; 73:100-108. 27. Hwang YC, Jeon WS, Park CY, Youn BS. The ratio of skeletal muscle mass to visceral fat area is a main determinant linking circulating irisin to metabolic phenotype. Cardiovasc Diabetol 2016; 15: 9. 28. Oelmann S, Nauck M, Völzke H, et al. Circulating irisin concentrations are associated with a favourable lipid profile in the general population. PloS one, 2016, 11(4): e0154319. 29. Moreno M, Moreno-Navarrete J M, Serrano M, et al. Circulating irisin levels are positively associated with metabolic risk factors in sedentary subjects. PLoS One, 2015, 10(4): e0124100. 30. Liu J J, Wong M D S, Toy W C, et al. Lower circulating irisin is associated with type 2 diabetes mellitus. Journal of Diabetes and its Complications, 2013, 27(4): 365-369. 17
31. Brondani LA, Boelter G, Assmann TS, Leitao CB, Canani LH, Crispim D. Irisin-encoding gene (FNDC5) variant is associated with changes in blood pressure and lipid profile in type 2 diabetic women but not in men. Metabolism 2015; 64:952–957. 32. Zhu D, Wang H, Zhang J, Zhang X, Xin C, Zhang F, et al. Irisin improves endothelial function in type 2 diabetes through reducing oxidative/nitrative stresses. J Mol Cell Cardiol 2015; 87:138–147. 33. Han F, Zhang S, Hou N, Wang D, Sun X. Irisin improves endothelial function in obese mice through the AMPK-eNOS pathway. Am J Physiol Heart Circ Physiol 2015;309:H1501–H1508. 34. Lee M J, Lee S A, Nam B Y, et al. Irisin, a novel myokine is an independent predictor for sarcopenia and carotid atherosclerosis in dialysis patients. Atherosclerosis, 2015, 242(2): 476-482. 35. Chung H S, Choi K M. Adipokines and Myokines: A Pivotal Role in Metabolic and Cardiovascular
Disorders[J].
Current
medicinal
chemistry,
2017;
Doi:10.2174/0929867325666171205144627 36. Li F, Li Y, Duan Y, et al. Myokines and adipokines: Involvement in the crosstalk between skeletal muscle and adipose tissue[J]. Cytokine & growth factor reviews, 2017, 33: 73-82. 37. Williams M A, Qiu C, Muy-Rivera M, et al. Plasma adiponectin concentrations in early pregnancy and subsequent risk of gestational diabetes mellitus[J]. The Journal of Clinical Endocrinology & Metabolism, 2004, 89(5): 2306-2311. 38. Qiu C, Williams M A, Vadachkoria S, et al. Increased maternal plasma leptin in early pregnancy and risk of gestational diabetes mellitus[J]. Obstetrics & Gynecology, 2004, 103(3): 519-525. 39. Krzyzanowska K, Zemany L, Krugluger W, et al. Serum concentrations of retinol-binding protein 4 in women with and without gestational diabetes[J]. Diabetologia, 2008, 51(7): 1115-1122. 40. Kuzmicki M, Telejko B, Szamatowicz J, et al. High resistin and interleukin-6 levels are associated with gestational diabetes mellitus[J]. Gynecological endocrinology, 2009, 25(4): 258-263. 41. Alis R, Sanchis-Gomar F, Pareja-Galeano H, et al. Association between irisin and homocysteine in euglycemic and diabetic subjects. Clinical biochemistry 2014; 47(18): 333-335. 18
42. Cho N H, Lim S, Jang H C, et al. Elevated homocysteine as a risk factor for the development of diabetes in women with a previous history of gestational diabetes mellitus[J]. Diabetes care, 2005, 28(11): 2750-2755. 43. Hou N, Han F, Sun X. The relationship between circulating irisin levels and endothelial function in lean and obese subjects. Clinical endocrinology, 2015, 83(3): 339-343. 44. Hedderson M M, Gunderson E P, Ferrara A. Gestational weight gain and risk of gestational diabetes mellitus[J]. Obstetrics and gynecology, 2010, 115(3): 597-604.
Table 1 the maternal and clinical characteristics of the included women at the time of irisin assessment P￡
Pξadjusted
27(25-32)
0.409
0.772
28.5(27.0-30.2)
24.3(23.1-27.5)
<0.001
0.002
73(25.4)
79(9.2)
<0.001
<0.005
Ethnicity-Han Chinese, N (%)
256(89.2)
768(89.0)
0.923
0.954
Pre-existing hypertension, N (%)
32(11.1)
50(5.8)
0.002
0.018
Pre-existing CVD, N (%)
27(9.4)
31(3.6)
<0.001
0.009
Family history of diabetes-Yes, N (%)
55(19.2)
136(15.8)
0.179
0.415
History of spontaneous abortion, N (%)
25(8.7)
70(8.1)
0.749
0.853
Smoker-Yes, N (%)
33(11.5)
106(12.3)
0.724
0.873
Gestational age at sampling (median, range)
5 (2-8)
6 (4-11)
0.026
0.093
Physical activity, kcal/kg day, median(IQR)
1.1(0.6-1.7)
1.7(1.0-2.5)
<0.001
0.002
Insulin, mIU/l
17.3(14.2-20.3)
12.3(9.0-15.1)
<0.001
0.001
Total cholesterol, mmol/l
4.90(4.22-5.09)
4.75(4.10-4.93)
0.041
0.206
Low irisin(Q1)
High irisin(Q2-4)
N
287
863
Maternal age, years, median(IQR)
27(26-33)
Pre-pregnancy BMI, kg/m2, median(IQR) Obesity, N (%)
Plasma samples, median(IQR)
19
HDL cholesterol, mmol/l
1.41(1.22-1.75)
1.81(1.39-2.07)
0.007
0.015
Triglyceride, mmol/l
1.75(1.27-1.99)
1.13(0.97-1.42)
<0.001
0.009
FPG, mg/dl
100.3(92.3-11.6)
82.6(77.3-91.1)
<0.001
<0.001
CRP, mg/l
5.0(3.8-7.5)
4.0(2.9-5.7)
<0.001
0.006
<0.001
<0.001
Irisin, ng/ml 77.4(65.8-88.0) 163.3(132.6-217.0) Data are median (interquartile range) or n (%) unless otherwise specified ￡P ξ
value tested by chi-square or Mann–Whitney test
logistic regression adjusted for maternal age, BMI, pre-existing CVD, pre-existing hypertension, gestational age
at blood collected, family history of diabetes, and blood levels of cholesterol, HDL, insulin, FPG and CRP. GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose.
Table 2 Odds ratios for GDM according to irisin quartiles at first visit Irisin quartiles ǂ Q1, N=287
GDM, N (%) 64(22.3)
Unadjusted OR (95% CI) ξ 11.61(5.61-25.84)
Adjusted OR (95%CI) *ξ 5.40(2.35-11.40)
Q2, N=288
45(15.6)
7.49(3.31-16.91)
3.83(1.63-8.01)
Q3, N=288
19(6.6)
2.83(1.17-6.85)
1.60(0.62-4.18)
Q4, N=287
7(2.4)
References
References
Q1 VS. Q2-4
－
3.22(2.23-4.67)
1.89(1.16-2.84)
ǂ
Plasma levels of irisin in quartile 1 (<99.4ng/ml), quartile 2 (99.4–141.2ng/ml), quartile 3 (141.3–192.9ng/ml),
and quartile 4 (>192.9ng/ml) *
Adjusted for maternal age, BMI, pre-existing CVD, pre-existing hypertension, gestational age at sampling, and
blood levels of cholesterol, HDL, insulin, FPG and CRP. ξ
p value for the trend <0.001
GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose; OR, odds ratios; CI, confidence intervals
20
Table 3 Plasma irisin concentrations at first visit prediction of GDM with AUROC GDM
AUROC Irisin
First visit
‡
Risk factors
0.723(0.683–0.763) ǂ
ǂ‡
0.776 (0.721-0.812)
Risk factors with Irisin‡
Incremental area (P)†
NRI (P)
IDI(P)
0.809(0.763-0.854)
0.034(0.03)
0.089(0.004)
0.031(0.04)
Established risk factors including: maternal age, BMI, gestational age at sampling, smoking, ethnicity,
pre-existing hypertension and CVD, history of spontaneous abortion, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP †
Comparison of AUROCs: established risk factors without irisin levels vs. established risk factors with irisin
levels. ‡
The results present with 95% confidence intervals
GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose; AUROC: Area under the Receiver Operating Characteristic
21
Figure legends Figure 1 Plasma levels of irisin in first trimester in women with GDM and without GDM. Mann–Whitney U Test. All data are medians and interquartile ranges (IQR). GDM, gestational diabetes mellitus.
Figure 2 The unadjusted incidence for GDM according to the baseline irisin quartiles. Plasma levels of irisin in Quartile 1 (<99.4ng/ml), Quartile 2 (99.4–141.2ng/ml), Quartile 3 (141.3–192.9ng/ml), and Quartile 4 (>192.9ng/ml). GDM= Gestational diabetes mellitus
22
Figure 3 Receiver operator characteristic curve demonstrating sensitivity as a function of 1-specificity for predicting the GDM based on the logistic model incorporating all known risk factors model /irisin and the relative contribution of each model alone (initial cohort). The known risk factors model included maternal age, BMI, gestational age at sampling, smoking, ethnicity, pre-existing hypertension and CVD, history of spontaneous abortion, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP.
23
Highlights 1. Irisin is a novel myokine and adipokine which improves insulin sensitivity and glucose tolerance 2. Irisin were significantly lower in women with GDM later developed than those without 3. Irisin shows potential as a novel biomarker for GDM prediction in Chinese women 4. The finding may open the way to the proposal of new therapeutic options
24
Sup Table I The maternal and clinical characteristics of the included women with and without GDM P￡
Pξadjusted
25(22-27)
0.031
0.106
28.1(26.5-29.7)
24.5(23.4-27.8)
<0.001
0.001
31(23.0)
121(11.9)
0.001
0.005
Ethnicity-Han Chinese, n (%)
119(88.1)
905(89.2)
0.704
0.825
Pre-existing hypertension, n (%)
14(10.4)
68(6.7)
0.119
0.335
Pre-existing CVD, n (%)
13(9.6)
45(4.4)
0.010
0.106
Family history of diabetes-Yes, n (%)
46(34.1)
145(14.3)
<0.001
0.008
History of spontaneous abortion, n (%)
12(8.9)
83(8.2)
0.778
0.866
Smoker-Yes, n (%)
18(13.3)
121(11.9)
0.636
0.778
Gestational weeks at blood collected,
6(4-8)
6(4-9)
0.365
0.569
1.4(0.8-2.0)
1.3(0.7-1.9)
0.406
0.615
Insulin, mIU/L
16.2(13.8-19.7)
12.5(9.2-15.6)
<0.001
0.003
Total cholesterol, mmol/L
4.98(4.33-5.19)
4.70(4.05-4.873)
0.033
0.194
HDL cholesterol, mmol/L
1.50(1.31-1.90)
1.89(1.52-2.18)
0.010
0.022
Triglyceride, mmol/L
1.32(1.13-1.60)
1.23(1.06-1.49)
0.086
0.332
FPG, mmol/L
5.35(5.02-5.88)
4.63(4.31-5.09)
<0.001
0.003
CRP, mg/l
4.8(3.6-7.0)
4.1(3.0-5.9)
0.001
0.022
IRISIN, ng/ml
106.5(73.8-136.9)
148.4(107.6-203.5)
<0.001
<0.001
Maternal age, years(IQR)
Women with GDM
Women without GDM
N
135
1015
Maternal age, years(IQR)
29(25-34)
BMI at admission, kg/m2(IQR) Obesity, n (%)
weeks(IQR) Physical activity, kcal/kg day(IQR) Blood samples at admission, median(IQR)
Data are median (interquartile range) or n (%). ￡P ξ
value tested by chi-square or Mann–Whitney test
logistic regression adjusted for maternal age, BMI, pre-existing CVD, family history of diabetes, and blood levels
of cholesterol, HDL, insulin, FPG and CRP. GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose.
25
S0168-8227(18)30613-2 https://doi.org/10.1016/j.diabres.2018.05.038 DIAB 7399
To appear in:
Diabetes Research and Clinical Practice
Received Date: Revised Date: Accepted Date:
16 April 2018 11 May 2018 22 May 2018
Please cite this article as: P. Wang, H-h. Ma, X-z. Hou, L-l. Song, X-l. Song, J-f. Zhang, Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus, Diabetes Research and Clinical Practice (2018), doi: https://doi.org/10.1016/j.diabres.2018.05.038
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reduced plasma level of irisin in first trimester as a risk factor for the development of gestational diabetes mellitus Pei Wang1, He-hong Ma1, Xiu-zhen Hou1, Li-li Song1, Xiao-long Song1, Jun-feng Zhang1＃ Obstetrical Department 1, Cangzhou Central Hospital, Cangzhou, Hebei, China, 061000 ＃Correspondence
to: Jun-feng Zhang, Email: [email protected]; Tel: +86-451-82576735
Address: No. 16, Xinhua West Road, Cangzhou 061000, P. R. China Running title: Irisin Levels and Risk of GDM Financial Disclosure The authors did not report any potential conflicts of interest. Each author has indicated that he or she has met the journal’s requirements for authorship.
1
ABSTRACT Background: The aim of this prospective cohort study was to investigate the association of first trimester irisin concentrations and the subsequent development of gestational diabetes mellitus (GDM). Methods: This cohort study was conducted at three maternity centers in China from July 2015 to June 2016. Data for fasting plasma glucose (FPG) and irisin concentrations in the first trimester and one-step GDM screening with 75-g oral glucose tolerance test (OGTT) performed between 24 and 28 weeks of gestation were collected and analyzed. Results: Plasma from women was available for 1150 women, of whom 135 (11.7%) developed GDM. The median value of irisin in those included women was 141.2 (IQR, 99.4–192.9) ng/ml. In multivariate models comparing the first(Q1), second(Q2) and third(Q3) quartiles against the fourth(Q4) quartile of irisin, levels of irisin in Q1 and Q2 were associated with GDM, and increased risk of GDM by 440% (odds ratios[OR]=5.40; 95% confidence intervals[CI]: 2.35-11.40) and 283% (OR: 3.83; 95%CI: 1.63-8.01). A model containing known risk factors plus irisin compared with a model containing known risk factors without irisin showed a greater discriminatory ability to predict GDM, the area under the curve (AUC) increased from 0.776 to 0.809. A significant difference in the AUC between the clinical variables alone and the addition of irisin level was observed (difference, 0.034; P=0.03). Conclusions: Reduced plasma levels of irisin in first trimester was associated with the increased risk of GDM and might be useful in identifying women at risk for GDM for early prevention strategies. Key words: Irisin, gestational diabetes mellitus, association, Chinese
Introduction 2
Gestational diabetes mellitus (GDM) is a first diagnosed diabetes during pregnancy defined as an abnormal glucose tolerance in various degree that may result in negative outcomes for mothers and their fetus and newborn [1]. Over the past years, considering its gradual upward trend in China (the adjusted prevalence of GDM increased by 2.8 times during 1999–2008, from 2.4 to 6.8%) [2], GDM has become a public health priority on the identification of potential predictors of glucose intolerance. Irisin is a small polypeptide hormone that is cleaved from the fibronectin type III domain-containing 5 (FNDC5). The plasma level of irisin is commonly increased by exercise [3]. Irisin is a novel myokine and adipokine that induces an increase in total body energy expenditure, improving insulin sensitivity and glucose tolerance in experimental animals [4]. Accumulated evidence also has demonstrated that irisin contributes to the regulation of glucose and lipid metabolism in skeletal muscle and adipose tissue [4-5]. Another study indicated that FNDC5/irisin ameliorates glucose/lipid metabolic derangements and insulin resistance in obese mice, and enhanced lipolysis via cAMP–PKA–HSL/perilipin pathway [6]. Furthermore, higher irisin levels observed in overweight/obese conditions could be a protective response of organism to early glucose impairments [7]. In healthy subjects aged 40-60 who were first-degree relatives (FDRs) of type 2 diabetes mellitus (T2DM) patients but without previous diagnosis of T2DM, Yang et al. [8] found that serum irisin level was closely related to homeostatic model assessment (HOMA)-β in normal glucose tolerance (NGT), suggesting that irisin may play a crucial role in pancreatic β-cell function. Recent clinical studies in humans have suggested that irisin plays substantial roles in the pathophysiology of obesity [9], insulin resistance [10], type 2 diabetes mellitus [11] and the metabolic syndrome [12]. Although some studies had reported that individuals with type 2 diabetes had lower levels of irisin as compared with nondiabetic controls [13-14], others had reported a positive association between circulating irisin and fasting plasma glucose [15-16] or muscle FNDC5 mRNA expression and 2 h glucose during an oral glucose tolerance test (OGTT) [17]. However, the roles of irisin in GDM were far from being completely understood. Several studies had reported reduced irisin levels in women with GDM [18-21] or in those whom GDM was later developed [22], compared to those who did not develop GDM at all. 3
However, Ebert et al. [23] found no significant different serum irisin levels between patients with GDM and the controls (P=0.616). The differences between those studies require further research to confirm. Thus, the aim of this prospective cohort study was to investigate the association of first trimester irisin concentrations and the subsequent development of GDM. Research design and methods This prospective, multi-center cohort study was conducted at three maternity centers from two cities (Harbin and Beijing) in China from July 2015 to June 2016. The sampling method was purposive sampling. Inclusion and exclusion criteria of the pregnant women have been reported previously [1]. In summary, singleton pregnant women (age≥18 years old) at the first prenatal visit of gestation were recruited. Exclusion criteria are the following: pre-gestational diabetes or diagnosed at the first prenatal visit; pregnancy-induced hypertension; preeclampsia at the first prenatal visit; have a history of childbirth, alcohol abuse; chronic diseases (liver and kidney system disease, autoimmune disease, neurological disorders, active or chronic inflammation); lost blood samples and termination of pregnancy in the follow-up. The study has been approved by the ethics committee of the Cangzhou Central Hospital prior to its initiation. All participants were informed of the study protocol, and written informed consents were obtained before inclusion. Maternal–pregnancy characteristics (maternal age, pre-pregnancy body mass index[BMI], ethnicity, smoking status, mode of conception, gestational weeks at sampling, history of spontaneous abortion, first-degree relative with diabetes and medical history of hypertension, polycystic ovary syndrome or cardiovascular disease[CVD]) were collected and recorded at first prenatal visit. BMI was calculated as the weight in kilograms divided by the height in meters squared. Specific details about physical activity were assessed by a validated questionnaire adapted from the Canadian Community Health Survey(CCHS) [24]. The physical activity index is derived from a variety of physical activities undertaken by respondents to the CCHS. For each leisure time activity an average daily energy expenditure was calculated based on frequency, duration, and intensity of participation in those leisure time activities. The energy expended for all activities is combined to calculate an average daily energy expenditure. Respondents were classified as ACTIVE if average daily energy expenditure was 3 kcal/kg/day, MODERATELY ACTIVE with an energy expenditure between 2.9 and 1.5 kcal/kg/day, and INACTIVE below a 4
value of 1.5 kcal/kg/day. During the first prenatal visit in the hospital, involved subjects were tested for fasting plasma glucose (FPG) and irisin using venous blood sample collected after at least 8 h of fasting. In accordance to the principle of blind method and prior to the understanding of the status of GDM, serum irisin concentrations were measured in duplicate by using the enzyme-linked immunosorbent assay (ELISA) kits (Aviscera Biosciences, Santa Clara, CA), in accordance with the manufacturer's instructions. The sensitivity of the assay was 1.0 ng/ml and the linear range of the standard was 1-500 ng/ ml. The intra-and inter-assay coefficients of variation (CV) were 5.2%-7.5% and 6.3%-8.8%, respectively. Routine blood biomarkers, for instance, triglyceride, cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), C-reactive protein (CRP) and FPG were tested using standard detection methods by BS800M (MINDRAY, Shenzhen, China). Plasma levels of insulin were measured using a chemiluminescence immunoassay method by DPC Immulite 2000 (Diagnostic Products Corporation, CA, USA). Between 24 and 28 weeks during pregnancy, women were informed to return to the hospital in their fasting state for repeat testing. A 75-g Oral Glucose Tolerance Test (OGTT) was performed once at this follow-up. Venous blood samples were collected at 0, 1, and 2 h after a 75-g glucose load. According to the criteria established by Ministry of Health (MOH) of China, diagnosis of GDM can be made when any one of the following values was met or exceeded in the 75-g OGTT: 0h (fasting) ≥5.10mmol/l; 1h≥10.00mmol/l; and 2h≥8.50mmol/l. Statistical Analysis Results were expressed as percentages for categorical variables and as medians (interquartile ranges[IQRs]) for the continuous variables. Univariate data on demographic and clinical features were compared by Mann-Whitney U-test or Chi-Square test as appropriate. In addition, 95% confidence intervals (CI) for incidence of GDM were calculated. Univariate and multivariate regression analyses were performed to investigate the relationship between concentrations of irisin and GDM. Crude models and multivariate models were developed for the adjustment of all associated factors and reported as odds ratios (ORs). For multivariate analysis, variables included the factors significant in univariate regression analysis (maternal age, BMI, pre-existing hypertension and cardiovascular disease [CVD], gestational age 5
at sampling, and blood levels of cholesterol, HDL, insulin, FPG and CRP). For a more detailed exploration of irisin and GDM, a multivariate analysis model was created to estimate adjusted ORs and 95% confidence intervals (CI) of GDM for irisin quartiles (with the highest quartile as the referent). In addition, plasma irisin concentration was also dichotomized according to irisin quartiles, and the grouping of quartiles 2-4 was defined as normal while the grouping of quartile 1 as low. In addition, a propensity score matching (PSM) analysis was performed to reduce the confounding between covariates and irisin levels. Second, we compared different prognostic risk scores from different predictive models by calculating receiver operating characteristic (ROC) analysis. Thereby the area under the receiver operating characteristic curve (AUC) is a summary measure over criteria and cut-point choices. The AUC summary equals the probability that the underlying classifier will score a randomly drawn positive sample higher than a randomly drawn negative sample. To test whether the irisin level improves score performance, we considered the two nested logistic regression models with irisin and known risk factors (maternal age, BMI, gestational age at sampling, smoking, ethnicity, pre-existing hypertension and CVD, history of spontaneous abortion, gestational weeks at admission, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP) as compared with known risk factors only. Furthermore, integrated discrimination improvement (IDI) and net reclassification improvement (NRI) indices were calculated to investigate the utility of adding irisin to recognized risk factors for GDM [25]. All statistical analyses were performed with SPSS for Windows, version 21.0 (SPSS Inc., Chicago, IL, USA), ROCR package (version 1.0-2), Stata release 14 (Stata Corp, College Station, TX), and GraphPad Prism 5.0. Statistical significance was defined as P<0.05. Results Plasma levels of irisin were available for 1150 women, with a median value of 141.2 ng/ml (IQR, 99.4–192.9 ng/ml). The maternal and clinical characteristics were presented in Table 1. As shown in Table 1, low levels of irisin were more likely associated with obesity, pre-existing hypertension, pre-existing CVD, and high physical activity, and higher levels of insulin, FPG, CRP, and triglyceride. In addition, levels of irisin were positively associated with HDL. Furthermore, the maternal and clinical characteristics of the included women with and without GDM were 6
presented in Sup table I. Among the 1150 included women, GDM was developed in 135 (11.7%). Irisin levels at first visit were significantly lower in women who developed late GDM than those who did not develop it at all [106.5(IQR, 73.8-136.9) ng/ml VS.148.4(107.6-203.5) ng/ml], as shown in Figure 1. The GDM distribution across the irisin quartiles ranged between 22.2% (first quartile) and 2.4% (fourth quartile), figure 2. In univariate logistic regression analysis, we calculated the odds ratio (OR) of irisin level to predict GDM as compared with other risk factors. With an unadjusted OR of 0.983 (95% CI, 0.977–0.987; P<0.001), irisin had a strong association with GDM. In multivariate models comparing the first(Q1), second(Q2) and third(Q3) quartiles against the fourth(Q4) quartile of irisin (Table 2), levels of irisin in Q1, Q2 and Q3 were associated with GDM, and increased risk of GDM by 440% (OR=5.40; 95%CI: 2.35-11.40), 283% (OR: 3.83; 95%CI: 1.63-8.01) and 60% (OR: 1.60; 95%CI: 0.62-4.18). The independent association of irisin with GDM was confirmed using the likelihood ratio test (P=0.001). In a multivariate model using the Q1 of irisin vs. Q2-4 together with the clinical variables, the marker displayed prognostic information and increased risk of GDM by 89% (OR for Q1, 1.89 [95% CI, 1.16-2.84]). In the entire cohort, there was a certain amount of bias between the two groups. The women with GDM have higher levels of insulin. Therefore, a propensity score was calculated for each woman (N=135) based on above identified variables. Interestingly, irisin still had an association with GDM (OR=0.988, 95%CI: 0.982-0.995; P=0.009) and pseudo R2 showed that we explained 28.9 percent of variation. Area under ROC curve is calculated as 0.75. Using ROC curves, irisin levels with 155ng/ml at first visit predicted the development of GDM with the highest sensitivity and specificity [67.6% and 70.7%, respectively; area under the curve (AUC)=0.723; 95% CI, 0.683–0.763]. With an AUC of 0.723, irisin showed a significantly greater discriminatory ability to predict GDM as compared with CRP (AUC, 0.618; 95% CI, 0.559–0.672; P<0.001), FPG (AUC, 0.685; 95% CI, 0.631–0.729; P=0.017) and insulin (AUC, 0.669; 95% CI, 0.616–0.705; P=0.003). In addition, a model containing known risk factors plus irisin compared with a model containing known risk factors without irisin showed a greater discriminatory ability to predict 7
GDM, the area under the curve (AUC) increased from 0.776 to 0.809 (95% confidence interval [CI], 0.763-0.854). A significant difference in the AUC between the clinical variables alone and the addition of irisin level was observed (difference, 0.034 [95% CI, 0.029-0.041]; P=0.03), table 3. The NRI statistic showed that the addition of irisin to established risk factors significantly increased the correct reclassification of GDM women and non-GDM women(P=0.004). The IDI statistic also found that the irisin level significantly increased the discrimination between GDM women and non-GDM women (P=0.04). Discussion Irisin is an exercise-regulated myokine inducing browning of white adipose tissue and has gained interest as a potential new strategy to combat obesity and its associated disorders, such as type 2 diabetes mellitus [11]. To the best of our knowledge, the present study is the first report to investigate the predictive value of irisin for GDM in a substantial cohort of Chinese women from three maternity centers. Maternal irisin levels at first visit were significantly lower in women who developed late GDM compared with those who did not. Data confirms evidence linking lower circulating irisin concentrations with signs of GDM and adds significant predictive information to the established risk factors. Furthermore, the addition of irisin to established risk factors improved the NRI and the IDI. Several studies found a positive association of irisin levels with BMI, insulin resistance (IR), and metabolic syndrome (MetS) [16, 26], whereas others reported negative correlations [13, 18-19, 27]. A significant association between circulating irisin and a favorable lipid profile in the general population was found [28], while another study found that circulating irisin levels were positively associated with parameters related to an increased cardiometabolic risk [29]. Previous study showed that plasma irisin levels appear to be associated with important metabolic factors (including BMI and fasting blood glucose) in non-diabetic subjects but not in individuals with type 2 diabetes [30]. Furthermore, some previous studies also reported that irisin-encoding gene (FNDC5) variant could change blood pressure in men with type 2 diabetes [31], while irisin improved endothelial function in type 2 diabetes [32] and in a mouse model of obesity [33]. Future study regarding interaction between irisin and other factors might be helpful to allow better understanding of irisin. 8
In this study, a negative correlation between irisin and BMI, insulin, FPG and lipid metabolism had been found. We considered irisin as a protective biomarker for GDM. In addition, irisin still was a useful predictor for GDM after being adjusted for above confounding factors. Consistent with our findings, Erol et al. [22] also suggested that irisin might be a useful biomarker in early pregnancy to predict the development of GDM. We considered that irisin could be seen as a predictive or clinical marker. Furthermore, irisin showed a significantly greater discriminatory ability to predict GDM as compared with other risk factors, for instance, CRP, FPG and insulin. Thus, clinicians would collect blood samples and test irisin levels in the first trimester for all women. Investigations on irisin in circulation from women have drawn the attention from many previous studies. In this study, we found that median irisin levels were significantly lower in women who developed late GDM compared with those who did not [106.5(IQR, 73.8-136.9) VS.148.4(10.7.6-203.5); P<0.001]. The level of irisin in this study was lower than in previous studies (213.4ng/ml ~ 1703.3ng/ml [18-23]). These controversial results might mostly be due to the small number of sample sizes from previous studies (range from 20 to 130) [18-23], which could
affect
the
power
of
statistical
analysis
conducted.
Furthermore,
different measurement ELISA kits may also cause differences. Lastly, blood levels of irisin were tested at different gestation age. In this study, we tested at first prenatal visit, which were similar with Erol et al. [22]. However, different time-point, including at the time of birth [19], before and within 24 h after delivery [23] and at 26-30th weeks of gestation [18] were assessed in other studies. Those variations make it difficult to compare different studies. This observational study design does not allow drawing a causal conclusion. We could not assess whether irisin is the cause or result of GDM. However, there are possible mechanisms in which irisin might be associated with GDM. First, we found that low levels of irisin were more likely with higher levels of insulin. Thus, irisin might be a biomarkers of insulin resistance. However, in the multivariable models, irisin still associated with GDM after adjusted for insulin resistance. Therefore, irisin is a specific unique predictor of GDM rather than just another measure of insulin resistance. Second, irisin is a novel myokine and adipokine that improves insulin sensitivity and glucose tolerance. Previous studies speculated the underlying mechanisms linking 9
irisin with GDM through glucose utilization and lipid metabolism [4, 6-8]. Third, another possible explanation is an interplay with other myokines, adipocytokines, or classical cytokines [34]. Furthermore, adipokines [leptin, adiponectin, resistin, chemerin, and visfatin] and myokines [interleukin (IL)-8, IL-15, irisin, and myonectin] play a crucial role in the communication between adipose tissue, skeletal muscle, and other organs [35]. Skeletal muscle-derived factors “myokines” and adipose tissue-derived factors “adipokines” exert their effects mainly in the organs of adipose tissue and/or skeletal muscle to form the muscle-adipose axis that regulates body composition [36]. The relationship between adipokine [37], leptin [38], retinol-binding protein 4[39], resistin and IL-6 [40] and GDM had been proposed in previous studies. Furthermore, one study provides evidence of an association between irisin and homocysteine, which may be due to nicotinamide metabolism [41]. Homocysteine level as a significant risk factor for development of diabetes in women with previous GDM [42]. Lastly, endothelial dysfunction has been documented in patients with diabetes and in individuals with insulin resistance or at high risk for developing type 2 diabetes. Lower levels of irisin are independently associated with endothelial dysfunction [43], and irisin alleviates endothelial dysfunction in type 2 diabetes partially via reducing oxidative/nitrative stresses through inhibiting signaling pathways implicating PKC-β/NADPH oxidase and NF-κB/iNOS [32]. The following limitations of our study must be taken into account. First, lifestyle advice/referral to diet would be changed the levels of irisin in those pregnant women. However, we did not collect data on lifestyle advice/referral to diet. Thus, we could not determine the association of those factors with blood levels of irisin and GDM. Second, the observational study design does not allow drawing secure conclusions for a causal relationship. In addition, Hedderson et al. [44] suggested that high rates of gestational weight gain, especially early in pregnancy, may increase a woman’s risk of GDM. However, in this study, we only assess the pre-pregnancy BMI, and gestational weight gain. Thus, we the relationship between irisin, gestational weight gain and GDM cannot be assessed. Third, irisin is usually measured by ELISA, but the quantification varies greatly between the kits. These differences probably come from the variety in the irisin epitopes being targeted for measurement by the manufacturing companies. Thus, the results should be confirmed in similar studies using an Irisin-Elisa kit targeting other epitopes. In addition, it has 10
also been suggested that some other adipokines, including adipokine [37], leptin [38] and retinol-binding protein 4[39] were associated with GDM. However, in this study, we did not test those adipokines. Therefore, we cannot determine the relationship between irisin, GDM and those factors. Finally, most participants of the present study were Chinese Han (89.0%) and whether these observations can also be extended to other ethnic groups with different body composition remains to be determined. Interestingly, in this study, Manchu, Hui, Mongolian and Korean accounted for 3.1%, 2.5%, 1.7% and 1%. However, there is no significant difference in the incidence of GDM among different ethnicities. In addition, some women with alcohol abuse and chronic diseases were excluded. Those women were also susceptible to GDM. Thus, the issue of selection bias must be considered. Conclusion Research results suggested that the reduced plasma levels of irisin in first trimester independently predicted GDM and significantly improved reclassification and discrimination. Irisin shows a potential as a novel biomarker for GDM prediction in Chinese women. Further studies, which may open the way to the proposal of new therapeutic options, are proposed to confirm this association.
11
Abbreviations list GDM=Gestational diabetes mellitus BMI= Body mass index MS=Metabolic syndrome FPG=Fasting plasma glucose HDL=High-density lipoprotein LDL=Low-density lipoprotein CRP=C-reactive protein OGTT=Oral Glucose Tolerance Test IQR= Interquartile ranges CI=Confidence intervals OR=Odds ratios ROC=Receiver operating characteristic curves AUC=Area under the curve IDI=Integrated discrimination improvement NRI=Net reclassification improvement NGT=Normal glucose tolerance
12
Declarations Consent for publication Consent to publish had be obtained from all included women. Availability of data and material Please contact author for data requests Ethics approval and consent to participate Prior to the initiation of the study, it has been approved by the ethics committee of the Cangzhou Central Hospital. All participants or their relatives were informed of the study protocol, and written informed consents were obtained before inclusion. Acknowledgements The authors thank the nurses, physicians, and patients who participated in our study, and the staff of the central laboratories of the hospitals. We also thanks Dr. Xiao-Tong Shi (China Rehabilitation Research Center) and Dr. Lei Yu (Harbin Medical University) who help us collect blood samples and clinical information. Competing interests None Funding None Guarantor’s statement Dr. Zhang JF is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Authors' contributions Zhang JF had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Wang P, Ma HL, Hou XZ, Song LL, Song XL, Zhang JF Acquisition of data: Wang P, Ma HL, Hou XZ 13
Analysis and interpretation of data: Song LL, Song XL, Zhang JF Drafting of the manuscript: Wang P, Ma HL, Hou XZ, Song LL Critical revision of the manuscript for important intellectual content: Song XL, Zhang JF Administrative, technical, or material support: Song XL, Zhang JF Study supervision: Zhang JF
14
References 1. Tu WJ, Guo M, Shi XD, et al. The Association of First Trimester Serum Fatty Acid–Binding Protein 4 and Subsequent Gestational Diabetes Mellitus. Obstetrics & Gynecology; 2017, 130(5): 1011-1016. 2. Zhang F, Dong L, Zhang CP et al. Increasing prevalence of gestational diabetes mellitus in Chinese women from 1999 to 2008. Diabet Med 2011; 28: 652–657. 3. Huh J Y, Mougios V, Kabasakalis A, et al. Exercise-induced irisin secretion is independent of age or fitness level and increased irisin may directly modulate muscle metabolism through AMPK activation. The Journal of Clinical Endocrinology & Metabolism 2014; 99(11): E2154-E2161. 4. Xin C, Liu J, Zhang J, et al. Irisin improves fatty acid oxidation and glucose utilization in type 2 diabetes by regulating the AMPK signaling pathway. International journal of obesity, 2016, 40(3): 443-451. 5. Lee H J, Lee J O, Kim N, et al. Irisin, a novel myokine, regulates glucose uptake in skeletal muscle cells via AMPK. Molecular Endocrinology, 2015, 29(6): 873-881. 6. Xiong X Q, Chen D, Sun H J, et al. FNDC5 overexpression and irisin ameliorate glucose/lipid metabolic derangements and enhance lipolysis in obesity. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease, 2015, 1852(9): 1867-1875. 7. Huerta A E, Prieto-Hontoria P L, Fernandez-Galilea M, et al. Circulating irisin and glucose 15
metabolism in overweight/obese women: effects of α-lipoic acid and eicosapentaenoic acid. Journal of physiology and biochemistry, 2015, 71(3): 547-558. 8. Yang M, Chen P, Jin H, et al. Circulating levels of irisin in middle-aged first-degree relatives of type 2 diabetes mellitus—correlation with pancreatic β-cell function. Diabetology & metabolic syndrome, 2014, 6(1): 133. 9. Perakakis N, Triantafyllou G A, Fernández-Real J M, et al. Physiology and role of irisin in glucose homeostasis. Nature Reviews Endocrinology 2017; 13(6): 324-337. 10. Qiu S, Cai X, Yin H, et al. Association between circulating irisin and insulin resistance in non-diabetic adults: a meta-analysis. Metabolism, 2016, 65(6): 825-834. 11. Shoukry A, Shalaby S M, El‐Arabi Bdeer S, et al. Circulating serum irisin levels in obesity and type 2 diabetes mellitus. IUBMB life 2016, 68(7): 544-556. 12. Park K H, Kang J H, Lee H J, et al. Irisin and its relation to metabolic syndrome among Korean adolescents. Obesity Reviews, 2016, 17: 128. 13. Moreno-Navarrete JM, Ortega F, Serrano M, Guerra E, Pardo G, Tinahones F, et al. Irisin is expressed and produced by human muscle and adipose tissue in association with obesity and insulin resistance. J Clin Endocrinol Metab 2013; 98: E769–E778. 14. Choi YK, Kim MK, Bae KH, Seo HA, Jeong JY, Lee WK, et al. Serum irisin levels in new-onset type 2 diabetes. Diabetes Res Clin Pract 2013;100:96–101. 15. Huh JY, Panagiotou G, Mougios V, Brinkoetter M, Vamvini MT, Schneider BE, et al. FNDC5 and irisin in humans: I. Predictors of circulating concentrations in serum and plasma and II. mRNA expression and circulating concentrations in response to weight loss and exercise. Metab Clin Exp 2012; 61:1725–1738. 16. Park KH, Zaichenko L, Brinkoetter M, Thakkar B, Sahin-Efe A, Joung KE, et al. Circulating irisin in relation to insulin resistance and the metabolic syndrome. J Clin Endocrinol Metab 2013; 98:4899–4907 17. Staiger H, Bo¨hm A, Scheler M et al. Common genetic variation in the human FNDC5 locus, encoding the novel musclederived ‘browning’ factor irisin, determines insulin sensitivity. PLoS One 2013; 8: e61903. 16
18. Kuzmicki M, Telejko B, Lipinska D, et al. Serum irisin concentration in women with gestational diabetes. Gynecological Endocrinology, 2014, 30(9): 636-639. 19. Yuksel M A, Oncul M, Tuten A, et al. Maternal serum and fetal cord blood irisin levels in gestational diabetes mellitus. Diabetes research and clinical practice, 2014, 104(1): 171-175. 20. Zhao L, Li J, Li Z L, et al. Circulating irisin is lower in gestational diabetes mellitus[J]. Endocrine journal, 2015, 62(10): 921-926. 21. Du X, Jiang W. Lower circulating irisin level in patients with diabetes mellitus: a systematic review and meta-analysis. Hormone and Metabolic Research, 2016, 48(10): 644-652. 22. Erol O, Erkal N, Ellidağ H Y, et al. Irisin as an early marker for predicting gestational diabetes mellitus: a prospective study. The Journal of Maternal-Fetal & Neonatal Medicine, 2016, 29(22): 3590-3595. 23. Ebert T, Stepan H, Schrey S, et al. Serum levels of irisin in gestational diabetes mellitus during pregnancy and after delivery. Cytokine, 2014, 65(2): 153-158. 24. Anonymous. Canadian Community Health Survey, Cycle 2.2, Nutrition (2004): A Guide to Accessing and Interpreting the Data [Health Canada, 2006]. 25. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008; 27:157–172. 26. Jang H B, Kim H J, Kang J H, et al. Association of circulating irisin levels with metabolic and metabolite profiles of Korean adolescents. Metabolism, 2017; 73:100-108. 27. Hwang YC, Jeon WS, Park CY, Youn BS. The ratio of skeletal muscle mass to visceral fat area is a main determinant linking circulating irisin to metabolic phenotype. Cardiovasc Diabetol 2016; 15: 9. 28. Oelmann S, Nauck M, Völzke H, et al. Circulating irisin concentrations are associated with a favourable lipid profile in the general population. PloS one, 2016, 11(4): e0154319. 29. Moreno M, Moreno-Navarrete J M, Serrano M, et al. Circulating irisin levels are positively associated with metabolic risk factors in sedentary subjects. PLoS One, 2015, 10(4): e0124100. 30. Liu J J, Wong M D S, Toy W C, et al. Lower circulating irisin is associated with type 2 diabetes mellitus. Journal of Diabetes and its Complications, 2013, 27(4): 365-369. 17
31. Brondani LA, Boelter G, Assmann TS, Leitao CB, Canani LH, Crispim D. Irisin-encoding gene (FNDC5) variant is associated with changes in blood pressure and lipid profile in type 2 diabetic women but not in men. Metabolism 2015; 64:952–957. 32. Zhu D, Wang H, Zhang J, Zhang X, Xin C, Zhang F, et al. Irisin improves endothelial function in type 2 diabetes through reducing oxidative/nitrative stresses. J Mol Cell Cardiol 2015; 87:138–147. 33. Han F, Zhang S, Hou N, Wang D, Sun X. Irisin improves endothelial function in obese mice through the AMPK-eNOS pathway. Am J Physiol Heart Circ Physiol 2015;309:H1501–H1508. 34. Lee M J, Lee S A, Nam B Y, et al. Irisin, a novel myokine is an independent predictor for sarcopenia and carotid atherosclerosis in dialysis patients. Atherosclerosis, 2015, 242(2): 476-482. 35. Chung H S, Choi K M. Adipokines and Myokines: A Pivotal Role in Metabolic and Cardiovascular
Disorders[J].
Current
medicinal
chemistry,
2017;
Doi:10.2174/0929867325666171205144627 36. Li F, Li Y, Duan Y, et al. Myokines and adipokines: Involvement in the crosstalk between skeletal muscle and adipose tissue[J]. Cytokine & growth factor reviews, 2017, 33: 73-82. 37. Williams M A, Qiu C, Muy-Rivera M, et al. Plasma adiponectin concentrations in early pregnancy and subsequent risk of gestational diabetes mellitus[J]. The Journal of Clinical Endocrinology & Metabolism, 2004, 89(5): 2306-2311. 38. Qiu C, Williams M A, Vadachkoria S, et al. Increased maternal plasma leptin in early pregnancy and risk of gestational diabetes mellitus[J]. Obstetrics & Gynecology, 2004, 103(3): 519-525. 39. Krzyzanowska K, Zemany L, Krugluger W, et al. Serum concentrations of retinol-binding protein 4 in women with and without gestational diabetes[J]. Diabetologia, 2008, 51(7): 1115-1122. 40. Kuzmicki M, Telejko B, Szamatowicz J, et al. High resistin and interleukin-6 levels are associated with gestational diabetes mellitus[J]. Gynecological endocrinology, 2009, 25(4): 258-263. 41. Alis R, Sanchis-Gomar F, Pareja-Galeano H, et al. Association between irisin and homocysteine in euglycemic and diabetic subjects. Clinical biochemistry 2014; 47(18): 333-335. 18
42. Cho N H, Lim S, Jang H C, et al. Elevated homocysteine as a risk factor for the development of diabetes in women with a previous history of gestational diabetes mellitus[J]. Diabetes care, 2005, 28(11): 2750-2755. 43. Hou N, Han F, Sun X. The relationship between circulating irisin levels and endothelial function in lean and obese subjects. Clinical endocrinology, 2015, 83(3): 339-343. 44. Hedderson M M, Gunderson E P, Ferrara A. Gestational weight gain and risk of gestational diabetes mellitus[J]. Obstetrics and gynecology, 2010, 115(3): 597-604.
Table 1 the maternal and clinical characteristics of the included women at the time of irisin assessment P￡
Pξadjusted
27(25-32)
0.409
0.772
28.5(27.0-30.2)
24.3(23.1-27.5)
<0.001
0.002
73(25.4)
79(9.2)
<0.001
<0.005
Ethnicity-Han Chinese, N (%)
256(89.2)
768(89.0)
0.923
0.954
Pre-existing hypertension, N (%)
32(11.1)
50(5.8)
0.002
0.018
Pre-existing CVD, N (%)
27(9.4)
31(3.6)
<0.001
0.009
Family history of diabetes-Yes, N (%)
55(19.2)
136(15.8)
0.179
0.415
History of spontaneous abortion, N (%)
25(8.7)
70(8.1)
0.749
0.853
Smoker-Yes, N (%)
33(11.5)
106(12.3)
0.724
0.873
Gestational age at sampling (median, range)
5 (2-8)
6 (4-11)
0.026
0.093
Physical activity, kcal/kg day, median(IQR)
1.1(0.6-1.7)
1.7(1.0-2.5)
<0.001
0.002
Insulin, mIU/l
17.3(14.2-20.3)
12.3(9.0-15.1)
<0.001
0.001
Total cholesterol, mmol/l
4.90(4.22-5.09)
4.75(4.10-4.93)
0.041
0.206
Low irisin(Q1)
High irisin(Q2-4)
N
287
863
Maternal age, years, median(IQR)
27(26-33)
Pre-pregnancy BMI, kg/m2, median(IQR) Obesity, N (%)
Plasma samples, median(IQR)
19
HDL cholesterol, mmol/l
1.41(1.22-1.75)
1.81(1.39-2.07)
0.007
0.015
Triglyceride, mmol/l
1.75(1.27-1.99)
1.13(0.97-1.42)
<0.001
0.009
FPG, mg/dl
100.3(92.3-11.6)
82.6(77.3-91.1)
<0.001
<0.001
CRP, mg/l
5.0(3.8-7.5)
4.0(2.9-5.7)
<0.001
0.006
<0.001
<0.001
Irisin, ng/ml 77.4(65.8-88.0) 163.3(132.6-217.0) Data are median (interquartile range) or n (%) unless otherwise specified ￡P ξ
value tested by chi-square or Mann–Whitney test
logistic regression adjusted for maternal age, BMI, pre-existing CVD, pre-existing hypertension, gestational age
at blood collected, family history of diabetes, and blood levels of cholesterol, HDL, insulin, FPG and CRP. GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose.
Table 2 Odds ratios for GDM according to irisin quartiles at first visit Irisin quartiles ǂ Q1, N=287
GDM, N (%) 64(22.3)
Unadjusted OR (95% CI) ξ 11.61(5.61-25.84)
Adjusted OR (95%CI) *ξ 5.40(2.35-11.40)
Q2, N=288
45(15.6)
7.49(3.31-16.91)
3.83(1.63-8.01)
Q3, N=288
19(6.6)
2.83(1.17-6.85)
1.60(0.62-4.18)
Q4, N=287
7(2.4)
References
References
Q1 VS. Q2-4
－
3.22(2.23-4.67)
1.89(1.16-2.84)
ǂ
Plasma levels of irisin in quartile 1 (<99.4ng/ml), quartile 2 (99.4–141.2ng/ml), quartile 3 (141.3–192.9ng/ml),
and quartile 4 (>192.9ng/ml) *
Adjusted for maternal age, BMI, pre-existing CVD, pre-existing hypertension, gestational age at sampling, and
blood levels of cholesterol, HDL, insulin, FPG and CRP. ξ
p value for the trend <0.001
GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose; OR, odds ratios; CI, confidence intervals
20
Table 3 Plasma irisin concentrations at first visit prediction of GDM with AUROC GDM
AUROC Irisin
First visit
‡
Risk factors
0.723(0.683–0.763) ǂ
ǂ‡
0.776 (0.721-0.812)
Risk factors with Irisin‡
Incremental area (P)†
NRI (P)
IDI(P)
0.809(0.763-0.854)
0.034(0.03)
0.089(0.004)
0.031(0.04)
Established risk factors including: maternal age, BMI, gestational age at sampling, smoking, ethnicity,
pre-existing hypertension and CVD, history of spontaneous abortion, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP †
Comparison of AUROCs: established risk factors without irisin levels vs. established risk factors with irisin
levels. ‡
The results present with 95% confidence intervals
GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose; AUROC: Area under the Receiver Operating Characteristic
21
Figure legends Figure 1 Plasma levels of irisin in first trimester in women with GDM and without GDM. Mann–Whitney U Test. All data are medians and interquartile ranges (IQR). GDM, gestational diabetes mellitus.
Figure 2 The unadjusted incidence for GDM according to the baseline irisin quartiles. Plasma levels of irisin in Quartile 1 (<99.4ng/ml), Quartile 2 (99.4–141.2ng/ml), Quartile 3 (141.3–192.9ng/ml), and Quartile 4 (>192.9ng/ml). GDM= Gestational diabetes mellitus
22
Figure 3 Receiver operator characteristic curve demonstrating sensitivity as a function of 1-specificity for predicting the GDM based on the logistic model incorporating all known risk factors model /irisin and the relative contribution of each model alone (initial cohort). The known risk factors model included maternal age, BMI, gestational age at sampling, smoking, ethnicity, pre-existing hypertension and CVD, history of spontaneous abortion, family history of diabetes, physical activity, and blood levels of cholesterol, HDL, triglyceride, insulin, FPG and CRP.
23
Highlights 1. Irisin is a novel myokine and adipokine which improves insulin sensitivity and glucose tolerance 2. Irisin were significantly lower in women with GDM later developed than those without 3. Irisin shows potential as a novel biomarker for GDM prediction in Chinese women 4. The finding may open the way to the proposal of new therapeutic options
24
Sup Table I The maternal and clinical characteristics of the included women with and without GDM P￡
Pξadjusted
25(22-27)
0.031
0.106
28.1(26.5-29.7)
24.5(23.4-27.8)
<0.001
0.001
31(23.0)
121(11.9)
0.001
0.005
Ethnicity-Han Chinese, n (%)
119(88.1)
905(89.2)
0.704
0.825
Pre-existing hypertension, n (%)
14(10.4)
68(6.7)
0.119
0.335
Pre-existing CVD, n (%)
13(9.6)
45(4.4)
0.010
0.106
Family history of diabetes-Yes, n (%)
46(34.1)
145(14.3)
<0.001
0.008
History of spontaneous abortion, n (%)
12(8.9)
83(8.2)
0.778
0.866
Smoker-Yes, n (%)
18(13.3)
121(11.9)
0.636
0.778
Gestational weeks at blood collected,
6(4-8)
6(4-9)
0.365
0.569
1.4(0.8-2.0)
1.3(0.7-1.9)
0.406
0.615
Insulin, mIU/L
16.2(13.8-19.7)
12.5(9.2-15.6)
<0.001
0.003
Total cholesterol, mmol/L
4.98(4.33-5.19)
4.70(4.05-4.873)
0.033
0.194
HDL cholesterol, mmol/L
1.50(1.31-1.90)
1.89(1.52-2.18)
0.010
0.022
Triglyceride, mmol/L
1.32(1.13-1.60)
1.23(1.06-1.49)
0.086
0.332
FPG, mmol/L
5.35(5.02-5.88)
4.63(4.31-5.09)
<0.001
0.003
CRP, mg/l
4.8(3.6-7.0)
4.1(3.0-5.9)
0.001
0.022
IRISIN, ng/ml
106.5(73.8-136.9)
148.4(107.6-203.5)
<0.001
<0.001
Maternal age, years(IQR)
Women with GDM
Women without GDM
N
135
1015
Maternal age, years(IQR)
29(25-34)
BMI at admission, kg/m2(IQR) Obesity, n (%)
weeks(IQR) Physical activity, kcal/kg day(IQR) Blood samples at admission, median(IQR)
Data are median (interquartile range) or n (%). ￡P ξ
value tested by chi-square or Mann–Whitney test
logistic regression adjusted for maternal age, BMI, pre-existing CVD, family history of diabetes, and blood levels
of cholesterol, HDL, insulin, FPG and CRP. GDM, gestational diabetes mellitus; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; FPG, Fasting plasma glucose.
25