Neuregulin 4 (Nrg4) was proven as a brown fat-enriched secreted factor that can regulate glucose and lipid metabolism. However, the association between circulating Nrg4 levels and diabetes mellitus (DM) in human remains unclear. We conducted a meta-analysis to investigate association of circulating Nrg4 with DM.
Observational studies comparing circulating Nrg4 levels in diabetes patients and health controls were included. Circulating Nrg4, correlation coefficients of clinical indices and circulating Nrg4 were pooled by meta-analysis.
Seven studies were included. The pooled results indicated there were no significant difference in the circulating Nrg4 between diabetes patients and controls (SMD = 0.18, 95%CI = -0.06 to 0.42, P = 0.143). However, diabetes patients had higher circulating Nrg4 than their controls in cross-sectional studies (SMD = 0.55, 95%CI = 0.36 to 0.73, P<0.001). None of the renal function and metabolic syndrome markers were correlated with circulating Nrg4, whereas the HbA1c and BMI were positively correlated (rs = 0.09, 95%CI = 0.03 to 0.16, P = 0.005; rs = 0.20, 95%CI = 0.07 to 0.34, P = 0.003; respectively).
Citation: Wang Y, Huang S, Yu P (2019) Association between circulating neuregulin4 levels and diabetes mellitus: A meta-analysis of observational studies. PLoS ONE 14(12): e0225705. https://doi.org/10.1371/journal.pone.0225705
Editor: Marcello Ciaccio, Institute of Clinical Biochemistry, Clinical Molecular Medicine and Laboratory Medicine, ITALY
Received: July 5, 2019; Accepted: November 11, 2019; Published: December 9, 2019
Copyright: © 2019 Wang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This study was funded by the National Natural Science Foundation of China (no.81600643 to P.Y., no.91746205 to P.Y.) the Tianjin Health Industry Key Research Projects (no.15KG101 to P.Y.), Tianjin Science and Technology Support Project (no. 17JCYBJC27000 to P.Y.), and Key Projects of Tianjin Natural Science Foundation (no.18JCZDJC32900 to P.Y.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
It is well known that adipose tissue contributes to systemic energy balance and metabolic homeostasis. White adipose tissue (WAT) stores fat and interacts with the central nervous system and other peripheral tissues, brown adipose tissue (BAT) serves a role on defense against cold and energy excess [1–3]. Recently, increasing publications have indicated adipose tissue derived factors exert effects on the process of regulating system metabolism and BAT has been evidenced to have an endocrine function in regulating energy metabolism and balance [4–8]. According to rat tissue and brown adipogenesis transcriptome data, a list of brown fat-enriched secreted factors was confirmed . Among these, neuregulin4 (Nrg4), which not only shows universal-expressed characteristic in multiple organs, but also concentrates in the brown adipose tissue (BAT), is central to regulate energy homeostasis and improve diet-induced metabolic disorders, including hepatic steatosis and insulin resistance . Wang  reported that Nrg4 might activate the ErbB3/ErbB4 signaling in mice hepatocytes to coordinate glucose and lipid homeostasis in obesity. Ma  demonstrated that up-regulation of Nrg4 could improve insulin resistance (IR) and prevent weight gain via Nrg4 gene transfer in mice. Wang  showed that Nrg4 overexpression could improve the efficacy of mesenchymal stem cells (ADSCs) in ameliorating IR and other obesity-related metabolic disorders. Rosell and his colleagues suggested Nrg4 could be a crucial factor for promoting the browning process of white fat by sympathetic innervation of white adipose tissue (WAT) depot, thus activating the thermogenic functions, enhancing energy consumption, and reducing adiposity . Taken together, these researches established the therapeutic potential of Nrg4 for treatment of obesity and associated disorders such as diabetes mellitus (DM). But the association between circulating Nrg4 levels and DM is not yet fully understood in humans, moreover, the results of previous studies on this relationship were not inconsistent. For these reasons, the aim of our meta-analysis was to investigate whether circulating Nrg4 levels were associated with DM and clinical indices of diabetes, renal function, metabolic syndrome and obesity in diabetes patients.
Materials and methods
We searched the PubMed, Embase, Web of Science databases from establishment to July 2, 2019 using the terms ‘‘diabetes” or ‘‘diabetes mellitus” or ‘‘DM” and ‘‘neuregulin 4” or ‘‘Nrg4” or ‘‘adipokine”. We selected only English-language, and the reference lists of identified publications were hand-examined, and the grey literature or ongoing relevant studies were searched on http://www.opengrey.eu/ and http://www.ntis.gov/.
Cross-sectional studies and case control studies comparing circulating Nrg4 levels in diabetes patients and health controls were included. The primary outcome was the association between circulating Nrg4 levels and DM. The second outcome was correlation coefficients between circulating Nrg4 levels and the measured clinical indices in subjects with DM. The clinical indices include DM markers (fasting plasma glucose (FPG); HOMA-IR, hemoglobin A1c (HbA1c)), renal function markers (estimated glomerular filtration rate (eGFR), creatinine), metabolic syndrome markers (systolic blood pressure (SBP), diastolic blood pressure (DBP), low-density lipoprotein (LDL-C), high-density lipoprotein (HDL-C), triglyceride (TG), total cholesterol (TC)), obesity markers (body mass index (BMI), waist circumference (WC)). Studies were excluded for the following reasons: publication of abstracts, duplicate data, overlapping population, unavailable data, literature reviews, animal experiment, and case report.
The information extracted from each original study by two investigators (Y.W. and S.H.), the name of first author, publication year, country, the number of subjects, mean age, circulating Nrg4 levels and correlation coefficients between Nrg4 and the measured clinical indices dealt by spreadsheets of Microsoft Excel. If there were opposite views between the two investigators, the discrepancy would be resolved through discussion. Moreover, one of the investigators would email the authors of the enrolled articles to get more information if data were absent.
Statistical analysis and quality assessment
Drafted on the basis of a preset protocol registered with PROSPERO 2018 (CRD42018103571), the present meta-analysis was conducted in line with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement . Stata 14.0 (Stata Corp, College Station, TX, USA) was used to carry out the meta-analysis. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated on pooled effects for dichotomous variables, while the standard mean difference (SMD) with 95% confidence intervals (CIs) were calculated for continuous variables. The I2 test was used to examine heterogeneity among the studies, I2 values of 25–50%, 50–75% and 75% are considered indicative of low, moderate and high heterogeneity, respectively . If I2<50%, a fixed-effect model was conducted; otherwise, radom-effect model was used. Sensitivity and subgroup analyses were attempted to identify potential sources of the between-study heterogeneity if necessary. Subgroup analyses were conducted by region (Asia and non-Asia), age (matched or higher in DM), and study type (cross-sectional and case-control). Begg’s and Egger’s test contributed to publication bias. P <0.05 was regarded as statistically significant.
Besides, as for the association between circulating Nrg4 levels and clinical indices in diabetes patients, Spearman correlation coefficients were selected for present meta-analysis, because they are unaffected by monotonic transformations. Pearson correlation coefficients could transform into Spearman correlation coefficients by equation: rs = 6/πsin-1 rp/2, where rp represents the Pearson correlation coefficient and rs represents the Spearman correlation coefficient . Aiming to adjust the sampling distribution of the Spearman correlation coefficients, the Fisher transformation shown in equation: z = 1/2*[ln (1+rs)-ln (1-rs)] was performed to transform correlation coefficient to an almost normally-distributed variable fisher’s z with SE (SE = 1/√(n-3)), where n is the sample size of diabetes patients. After these conversion, forest plots could be established to show the summary fisher’s z (Zs) and its 95% confidence intervals (CIs) . Subsequently, the Zs was converted to the summary correlation coefficient (rs) by the transformation shown in equation: rs = e2Zs-1/e2Zs+1.
Newcastle-Ottawa Quality Assessment Scale (NOS) was used to assess the quality of case-control studies , while, Agency for Healthcare Research and Quality (AHRQ) was appropriate for assessment of cross-sectional studies . Disagreements were resolved by argumentum.
Overall, 165 studies were identified in our initial search, of which 12 underwent full-text review (Fig 1). Finally, 7 studies [20–26] involving 2257 patients were identified and enrolled in the meta-analysis (Fig 1). Detailed information for another 5 excluded articles is accessible in S1 Table. The baseline demographics of these enrolled studies were shown in Table 1. Results of quality assessment of all included studies were relatively satisfying and shown in S2 Table.
DM and circulating Nrg4 levels
There were seven studies [20–26] regarding the association between circulating Nrg4 levels and DM, the P value of heterogeneity between these studies was significant, so we used the random-effect model indicating that there was no significant statistical difference between diabetes patients and their normal controls (SMD = 0.18, 95% CI = -0.06 to 0.42, P = 0.143) (Fig 2). The results of subgroup analyses by region (Asian[20–23,26] or non-Asian[24–25]), age (matched[21,24–26] or higher [20,22–23] in DM) study type (case-control [20–24,26] or cross-sectional) are summarized in Table 2. A subgroup of cross-sectional study demonstrated that diabetes patients had higher circulating Nrg4 levels than their normal controls (SMD = 0.55, 95% CI = 0.36 to 0.73, P<0.001). Moreover, subgroup analysis according to region and age demonstrated that no significant difference of circulating Nrg4 levels between two groups was recorded. The sensitivity analysis presented a robust result that was not influenced by individual studies. Neither the Begg’s (P = 0.548) test nor the Egger’s (P = 0.316) test showed publication bias.
Diabetes markers and circulating Nrg4 levels
Six studies [20–21,23–26] presented data on the correlation coefficients of circulating Nrg4 levels and fasting plasma glucose (FPG) in diabetes patients, and five studies [20,23–26] reported HOMA-IR. Neither index was significantly correlated with circulating Nrg4 levels (rs = 0.12, 95% CI = -0.12 to 0.36, P = 0.330 and rs = 0.02, 95% CI = -0.15 to 0.19, P = 0.806, respectively) (Fig 3A and 3B). No publication bias was noted in these parameters. The sensitivity analyses of studies with significant heterogeneity didn’t show results sensitive to any individual studies, and the subgroup analyses did not discover the sources of the heterogeneity. Hemoglobin A1c (HbA1c) were measured in four studies [21,23–25]. The overall correlation coefficient was not statistically significant (rs = 0.03, 95% CI = -0.11 to 0.17, P = 0.685). Subgroup analysis had not found the sources of the significant heterogeneity. Based on sensitivity analysis, the most influential study by Kurek Eken M  was excluded, this resulted in a summary coefficient coefficient of 0.09 (95% CI = 0.03 to 0.16, P = 0.005) (Fig 3C), neither the Begg’s (P = 1.000) test nor the Egger’s (P = 0.708) test showed publication bias.
Renal function markers and Nrg4
The renal function markers estimated glomerular filtration rate (eGFR) and creatinine were each discussed in two of the included studies. It was found that eGFR and creatinine were no significantly correlated with circulating Nrg4 (rs = -0.14, 95% CI = -0.30 to 0.03, P = 0.099 and rs = -0.04, 95% CI = -0.21 to 0.13, P = 0.624; respectively) (Fig 4A and 4B). No publication bias was noted in these parameters.
Metabolic syndrome markers and circulating Nrg4 levels
Blood pressure was reported in four studies [20–21,24–25]. Our meta-analysis showed that systolic blood pressure and diastolic blood pressure were not correlated with circulating Nrg4 levels (rs = -0.01, 95% CI = -0.07 to 0.05, P = 0.735 and rs = 0.05, 95% CI = -0.07 to 0.05, P = 0.740, respectively) (Fig 5A and 5B). Also, the low-density lipoprotein cholesterol (LDL-C) (rs = 0.01, 95% CI = -0.05 to 0.07, P = 0.463) (Fig 5C), high-density lipoprotein cholesterol (HDL-C) (rs = -0.01, 95% CI = -0.13 to 0.11, P = 0.882) (Fig 5D), triglycerides (TG) (rs = 0.10, 95% CI = -0.07 to 0.28, P = 0.238) (Fig 5E) and total cholesterol (TC) (rs = -0.00, 95% CI = -0.06 to 0.06, P = 0.900) (Fig 5F) were not significantly correlated with circulating Nrg4 levels. The sensitivity analyses of studies with significant heterogeneity did not show results sensitive to any individual studies, and the subgroup analyses did not discover the sources of the heterogeneity. No publication bias was noted in these parameters.
Obesity markers and circulating Nrg4 levels
There were four included studies [20–21, 23–24] discussing about body mass index (BMI). The overall correlation coefficient was not significantly correlated with circulating Nrg4 levels (rs = 0.11, 95% CI = -0.10 to 0.33, P = 0.301). Subgroup analysis did not find the sources of the significant heterogeneity. Based on sensitivity analysis, the most influential study by Cai C  was omitted, this resulted in a summary coefficient of 0.20 (95% CI = 0.07 to 0.34, P = 0.003) (Fig 6A). There was no publication bias found with Begg’s (P = 1.000) and Egger’s (P = 0.830) tests. Waist circumference (WC) was measured in two studies [20–21], and the overall correlation coefficient was also not correlated with circulating Nrg4 (rs = 0.02, 95% CI = -0.26 to 0.30, P = 0.883) (Fig 6B). There was no publication bias found with Begg’s (P = 1.000) test.
It is well known that BAT could enhance energy expenditure through adjusting whole-body energy balance . Meanwhile, activation of BAT is related to reduce adiposity, plasma lipid and glucose . Besides, several recent studies indicated that BAT could secret endocrine factors working on other tissues to regulate systemic metabolism. One intriguing finding is Nrg4. Evidence showed that Nrg4 deficiency in mice and human obesity adipose tissue could aggravate insulin resistance and diet-induced lipid metabolic disorder . But the correlation between circulating Nrg4 levels and DM in human has not been well characterized.
Our study is the first meta-analysis to investigate the association of Nrg4 with DM and clinical indices, indicating that circulating Nrg4 levels were not significantly associated with DM, but subgroup analysis according to study type demonstrated that circulating Nrg4 levels were higher in diabetes patients than their normal controls in the cross-sectional study group. Furthermore, neither renal nor metabolic syndrome markers was related to Nrg4, and HbA1c and BMI were positively associated with circulating Nrg4.
In our meta-analysis, circulating Nrg4 levels are not associated with the risk of DM. Contrastingly, Chen  demonstrated circulating Nrg4 levels was elevated in diabetes patients, similar results were concluded by Kang . In contrast, some researchers reached the opposite conclusion [25, 26]. Therefore, meta-analysis concerning this relationship was carried out which showed no difference. We speculated the source of discrepancy might be from age, region, and study type. Therefore, subgroup analyses of them were performed. Our result of the subgroup of cross-sectional study raised the possibility that circulating Nrg4 levels might be a role in the development of DM, which was similar to Chen . However, this result conflicted with those of a previous study reporting that Nrg4 stimulates insulin secretion and controls de novo liver lipid synthesis under metabolic stress in mice , and this study also revealed increased Nrg4 expression could improve insulin resistance and glucose intolerance, thus suggesting the prospect of treating obesity-related comorbidities such as diabetes. We harbored the idea that this phenomenon might be attributed to several reasons. Firstly, our result might embody a compensatory response to decreased expression of Nrg4 in adipose tissue or a kind of impairment in the Nrg4/ErbB signaling pathway by receptor resistance. That is to say, there may be “Nrg4 resistance” just like “insulin resistance” . Secondly, although mainly released by adipose tissue, Nrg4 also could be secreted by other tissues such as liver, pancreas, stomach and muscle [29, 30]. Because the Nrg4 expression patterns might be evidently different among those tissues, the expression of Nrg4 in circulation under the condition of metabolic disorders could be distinct from that in adipose tissues. Thirdly, cold condition can affect the Nrg4 expression , leading to the different results under different temperatures. Unfortunately, such an analysis on this topic could not be performed owing to limited data. Finally, given that Nrg4 measurement might influence the results from different investigators, so the standardized Nrg4 assays are necessary.
The indices of the renal function and metabolic syndrome were not correlated with circulating Nrg4 levels in diabetes patients. Whereas the meta-analysis suggested that circulating Nrg4 levels were positively correlated with one of the glucose parameters (HbA1c), this finding was similar to some researchers [21, 23]. They concluded that the theory of “Nrg4 resistance” contributes to this correlation. Besides, although Nrg4 preserves glucose homeostasis in mice hepatocytes by activation of ErBb3/ErBb4 signaling, the universal distribution characteristic of ErBb3/ErBb4 signaling may lead to a totally different biological role of Nrg4 in other sites of human body [9, 32, 33]. On the contrary, Wang and Zhang [9, 26] argued that circulating Nrg4 levels were negatively correlated with glucose parameters. They explained that Nrg4 could stimulate insulin release from islet cells and improve insulin resistance under metabolic stress in mice [9, 34]. Taken together, the exact mechanism of Nrg4 in regulating metabolism has not been illuminated and requires further study.
Additionally, a positive correlation between adiposity proxies and circulating Nrg4 levels was confirmed in the meta-analysis. Intriguingly, Wang’s study encapsulated a negative correlation between obesity and adipose Nrg4 mRNA , and Cai  found that circulating Nrg4 was negatively associated with indices of obesity. Whereas Dai  showed that circulating Nrg4 was not correlated with obesity indices, similar results were found by Zhang . Furthermore, Chen and Kang’s works were confirmed by our study [21, 23]. Expression of Nrg4 in each unit of adipose tissue, together with the total fat mass, forms a delicate balance and therefore gives an explanation to the discrepancy. In other words, although expression of Nrg4 in one unit adipose tissue decreases in obesity , the total fat mass should rise in obese subjects, which would increase the overall expression of Nrg4 through the whole body . Thus the eventual expressing direction of Nrg4 for the whole body in obesity would be determined by the interaction of these two opposite notions, and each of the three different results might be reasonable.
In our meta-analysis, although we observed higher levels of circulating Nrg4 in diabetes patients in the cross-sectional studies, only one case-control study investigated this correlation , which suggested no significant association between circulating Nrg4 levels and DM. Since insufficient data on other study types, it was still inconclusive whether there was significant relationship of circulating Nrg4 levels and DM. Therefore, large prospective studies are required to confirm whether circulating Nrg4 levels are associated with the risk of DM in the future. Furthermore, standardized Nrg4 binding assays and analysis of Nrg4 expression pattern in multiple organs are necessary.
Our meta-analysis had some strengths. First of all, we evaluated the association of circulating Nrg4 levels with the risk of DM and glucose-related/non-glucose-related parameters in diabetes patients in the round, which added to the evidences on the role of circulating Nrg4 in human metabolism. Second, our meta-analysis could avoid some potential bias, because that results of stratification according to age and region still demonstrated no significant association between circulating Nrg4 levels and DM. Several limitations cannot be avoided in our meta-analysis. Firstly, the limited studies on this correlation could influence the strength of our meta-analysis conclusions. Secondly, the most of studies were based on cross-sectional data with a limited sample size, and given its cross-sectional design, it could not confirm the causal association between circulating Nrg4 levels and DM. Thirdly, the enrolled articles were all observational studies. The possibilities of selection and confounding bias could not be avoided. Indeed, diabetes patients could differ from their health controls in terms of economic power, access to healthcare services, attention to their own health and quality of our life. Finally, the included studies were lack of more useful markers of the metabolic syndrome, such as Visceral Adiposity Index (VAI) and Lipid Accumulation Product Index (LAP), which are more realistic than HOMA, BMI or WC. Therefore, more powerful original research with new and realistic markers of metabolic syndrome are warranted to elucidate the exact role of circulating Nrg4 in the development of DM in larger samples and in other ethnic groups.
In summary, although our meta-analysis suggested circulating Nrg4 levels didn’t play a role in the process of DM, subgroup of the cross-sectional study revealed that circulating Nrg4 is higher in diabetes patients. As for clinical indices in diabetes patients, the glucose and adiposity parameters were positively associated with circulating Nrg4. The exact role of the circulating Nrg4 levels in the development of DM still needs to be further identified.
S1 Table. Reasons why 5 studies were excluded after full-text review.
- 1. Kadowaki T, Yamauchi T, Kubota N, Hara K, Ueki K, Tobe K. Adiponectin and adiponectin receptors in insulin resistance, diabetes, and the metabolic syndrome. J Clin Invest. 2006; 116: 1784–92. pmid:16823476
- 2. Trujillo ME, Scherer PE. Adipose tissue-derived factors: impact onhealth and disease. Endocr Rev. 2006; 27: 762–78. pmid:17056740
- 3. Waki H, Tontonoz P. Endocrine functions of adipose tissue. Annu Rev Pathol. 2007; 2: 31–56 pmid:18039092
- 4. Parimisetty A, Dorsemans AC, Awada R, Ravanan P, Diotel N, Lefebvre d’Hellencourt C. Secret talk between adipose tissue and central nervous system via secreted factors-an emerging frontier in the neurodegenerative research. J. Neuroinflammation. 2016;13(1): 67. pmid:27012931
- 5. Cannon B, Nedergaard J. Brown adipose tissue: Function and physiological significance. Physiol. Rev. 2004; 84(1):277–359. pmid:14715917
- 6. Wang GX, Zhao XY, Lin JD. The brown fat secretome: Metabolic functions beyond thermogenesis. Trends. Endocrin. Met. 2015; 26(5):231–37.
- 7. Villarroya F, Giralt M. The beneficial effects of brown fat transplantation: Further evidence of an endocrine role of brown adipose tissue. Endocrinology. 2015;156(7):2368–70. pmid:26091427
- 8. Villarroya F, Cereijo R, Villarroya J, Giralt M. Brown adipose tissue as a secretory organ. Nat. Rev. Endocrinol. 2017; 13(1): 26–35. pmid:27616452
- 9. Wang GX, Zhao XY, Meng ZX, Kern M, Dietrich A, Chen Z, et al. The brown fat-enriched secreted factor Nrg4 preserves metabolic homeostasis through attenuation of hepatic lipogenesis. Nature medicine. 2014; 20(12):1436–43. pmid:25401691
- 10. Pfeifer A. NRG4: an endocrine link between brown adipose tissue and liver. Cell Metab. 2015; 21:13–4. pmid:25565202
- 11. Ma Y, Gao M, Liu D. Preventing high fat diet-induced obesity and improving insulin sensitivity through neuregulin 4 gene transfer. Sci. Rep. 2016;6:26242. pmid:27184920
- 12. Wang W, Zhang Y, Yang C, Wang Y, Shen J, Shi M, et al. Feature Article: Transplantation of neuregulin 4-overexpressing adipose-derived mesenchymal stem cells ameliorates insulin resistance by attenuating hepatic steatosis. Exp Biol Med (Maywood). 2019; 244(7):565–578.
- 13. Rosell M, Kaforou M, Frontini A, Okolo A, Chan YW, Nikolopoulou E et al. Brown and white adipose tissues: intrinsic differences in gene expression and response to cold exposure in mice. Am J Physiol Endocrinol Metab. 2014; 306(8): 945–64.
- 14. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000; 283(15): 2008–12. pmid:10789670
- 15. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat. Med. 2002; 21:1539–58. pmid:12111919
- 16. Rupinski MT, Dunlap WP. Approximating pearson product-moment correlations from Kennall’s tau and Spearman’s rho. Educ. Psychol. Measur. 1996; 56: 419–29.
- 17. Wilson DB, Lipsey MW. The role of method in treatment effectiveness research: evidence from meta-analysis. Psychol Methods. 2001; 6(4): 413–29. pmid:11778681
- 18. Wells G, Shea B, O’Connell J, Peterson J, Welch V, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analysis. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. [updated 2011].
- 19. Rostom A, Dubé C, Cranney A, Saloojee N, Sy R, Garritty C, et al. Celiac Disease. Rockville (MD): Agency for Healthcare Research and Quality (US); 2004. (Evidence Reports/Technology Assessments, No.104.) Appendix D. Quality Assessment Forms. http://www.ncbi.nlm.nih.gov/books/NBK35156/.
- 20. Cai C, Lin M, Xu Y, Li X, Yang S, Zhang H. Association of circulating neuregulin 4 with metabolic syndrome in obese adults: a cross-sectional study. BMC Med. 2016;14(1):165. pmid:27772531
- 21. Chen LL, Peng MM, Zhang JY, Hu X, Min J, Huang QL, et al. Elevated circulating neuregulin4 level in patients with diabetes. Diabetes Metab. Res. Rev. 2016; 33: 2870.
- 22. Jiang J, Lin M, Xu Y, Shao J, Li X, Zhang H, et al. Circulating neuregulin 4 levels are inversely associated with subclinical cardiovascular disease in obese adults. Sci. Rep. 2016; 6: 36710. pmid:27819316
- 23. Kang YE, Kim JM, Choung S, Joung KH, Lee JH, Kim HJ, et al. Comparison of circulating Neuregulin 4 (Nrg4) levels in adults with newly diagnosed type 2 diabetes mellitus and controls without diabetes. Diabetes Res. Clin. Pract. 2016;117:1–3. pmid:27329015
- 24. Kurek Eken M, Yayla Abide C, Sahin Ersoy G, Altun Ensari T, Pekin O, Cevik O. Clinical significance of neuregulin 4 (NRG4) in gestational diabetes mellitus. Gynecol Endocrinol. 2018; 34(7): 605–8. pmid:29282998
- 25. Kralisch S, Hoffmann A, Kratzsch J, Blüher M, Blüher M, Stumvoll M, Fasshauer M, et al. The brown-fat-secreted adipokine neuregulin 4 is decreased in gestational diabetes mellitus. Diabetes Metab. 2018; 44(2):150–54. pmid:28709749
- 26. Zhang L, Fu Y, Zhou N, Cheng X, Chen C. Circulating neuregulin 4 concentrations in patients with newly diagnosed type 2 diabetes: a cross-sectional study. Endocrine. 2017; 57(3): 535–38. pmid:28523627
- 27. Oelkrug R, Polymeropoulos ET, Jastroch M. Brown adipose tissue: physiological function and evolutionary significance. J Comp Physiol B. 2015;185(6): 587–606. pmid:25966796
- 28. Bartelt A, Bruns OT, Reimer R, Hohenberg H, Ittrich H, Peldschus K, et al. Brown adipose tissue activity controls triglyceride clearance. Nat Med. 2011;17(2): 200–5. pmid:21258337
- 29. Harari D, Tzahar E, Romano J, Shelly M, Pierce JH, Andrews GC, et al. Neuregulin-4: a novel growth factor that acts through the ErbB-4 receptor tyrosine kinase. Oncogene. 1999;18(17):2681–9. pmid:10348342
- 30. Hayes NV, Blackburn E, Boyle MM, Russell GA, Frost TM, Morgan BJ, et al. Expression of neuregulin 4 splice variants in normal human tissues and prostate cancer and their effects on cell motility. Endocr Relat Cancer. 2011;18(1):39–49 pmid:20959441
- 31. Rosell M, Kaforou M, Frontini A, Okolo A, Chan YW, Nikolopoulou E, et al. Brown and white adipose tissues: intrinsic differences in gene expression and response to cold exposure in mice. Am. J. Physiol. Endocrinol. Metab. 2014; 306(8): 945–64.
- 32. Hayes NV, Newsam RJ, Baines AJ, Gullick WJ. Characterization of the cell membrane-associated products of the Neuregulin 4 gene. Oncogene. 2008; 27:715–20. pmid:17684490
- 33. Schneider MR, Wolf E. The epidermal growth factor receptor ligands at a glance. J Cell Physiol. 2009; 218:460–6. pmid:19006176
- 34. South JC, Blackburn E, Brown IR, Gullick WJ. The neuregulin system of ligands and their receptors in rat islets of langerhans. Endocrinology. 2013;154:2385–92. pmid:23610133
- 35. Dai YN, Zhu JZ, Fang ZY, Zhao DJ, Wan XY, Zhu HT, et al. A case-control study: association between circulating neuregulin 4 level and non-alcoholic fatty liver disease. Metab. 2015;64(12):1667–73.