Eur J Clin Invest 2010; 40 (4): 339–343
Background Gestational diabetes (GDM) is an increasing and common complication of pregnancy. The involvement of inflammatory mechanisms in GDM remains unclear. YKL-40 is a novel inflammatory marker that has been recently found to be associated with type 2 diabetes. This is the first study to investigate YKL-40 in GDM.
Material and methods A total of 58 subjects were included, 28 patients with GDM (BMI 33·2 ± 6·1 kg m-², 33 ± 6 years) and 30 healthy pregnant controls (BMI 28·4 ± 5·2 kg m-², 33 ± 4 years; mean ± SD). Standard risk factors for GDM (weight and BMI prior to pregnancy, family history, former GDM, high birthweight offspring) were evaluated. A 2-h 75-g oral glucose tolerance test (oGTT) and measurement of YKL-40 were conducted in gestational week 28 ± 4, as well as 8 weeks after delivery.
Results YKL-40 was not different between GDM and controls, neither during (65·8 ± 44·4 vs. 60·3 ± 30·1 ng mL−1), nor after pregnancy (63·4 ± 30·5 vs. 66·9 ± 32·7 ng mL−1). YKL-40 was correlated with insulin, HOMA and BMI. GDM had higher fasting insulin (14·1 ± 7·4 vs. 8·3 ± 4·3 μU mL−1) and glucose (88 ± 13 – 200 ± 31 – 160 ± 33 vs. 76 ± 10 – 146 ± 37 – 112 ± 28 mg dL−1 for fasting, 1- and 2-h-concentrations in the oGTT, respectively), higher HbA1c (5·3 ± 0·4 vs. 5·0 ± 0·5%;), HOMA (3·1 ± 1·7 vs. 1·6 ± 0·9), and BMI (33·2 ± 6·1 vs. 28·5 ± 5·2 kg m−2) (means ± SD, all P < 0·01).
Conclusions No difference in YKL-40 between GDM and controls suggests similar inflammatory status at the time of measurements. The short duration of metabolic changes during GDM might explain this finding, which is in contrast to results in type 2 diabetes.
Gestational diabetes mellitus (GDM) is a common complication in pregnancy with increasing incidence  and is defined as glucose intolerance with onset or first recognition during pregnancy . While GDM tends to resolve after delivery, it is associated with complications for both infants and mothers at birth, and with a higher risk of developing type 2 diabetes later in life . The association of inflammation with obesity, insulin resistance, and type 2 diabetes has been increasingly recognised in the last decade [4,5]. Women with GDM not only show various features of the metabolic syndrome like obesity , but also low-grade inflammation . Several parameters indicate that inflammation could contribute to insulin resistance during pregnancy and GDM [8–11].
YKL-40 is a glycoprotein secreted by a variety of human cell types, including neutrophils, macrophages and vascular smooth muscle cells. Also called human cartilage glycoprotein-39 (HCgp-39)  or human chitinase-3-like-1 protein (CHI3L1), YKL-40 was named after its N-terminal amino acid sequence and its molecular weight of 40 kDa . YKL-40 has been described in various acute and chronic inflammatory diseases and is elevated in several types of cancer . YKL-40 is also involved in atherosclerosis  and is increased in patients with acute myocardial infarction or stable coronary artery disease, where it correlates with C-reactive protein (CRP) . Recently, YKL-40 has been identified as a marker for type 2 diabetes related to metabolic and inflammatory parameters . Interestingly, YKL-40 is also elevated in type 1 diabetes patients with increasing levels of albuminuria . YKL-40 could therefore represent a link between inflammation, insulin resistance and diabetes, as well as development of microvascular and macrovascular complications.
The role of YKL-40 in GDM remains unclear and has not been investigated previously. We therefore assessed YKL-40 concentrations in women with GDM during and after pregnancy in comparison to healthy pregnant women.
Materials and methods
The study was conducted in accordance with the principles outlined in the Declaration of Helsinki and was approved by the Institutional Review Committee. Prior to the study written informed consent was obtained from all subjects.
A total of 58 female subjects were included in the study, consisting of 28 patients with GDM and 30 healthy pregnant women. Subjects came from a mixed ethnic background and were referred to a specialised clinic for routine GTT. All subjects were evaluated for GDM risk factors (age, weight and BMI prior to pregnancy, family history, former GDM, high birthweight offspring) and underwent measurements of height, weight and blood pressure. GDM was diagnosed according to the American Diabetes Association criteria for GDM . Of those with GDM, 20 women were treated with insulin and eight with diet alone. Baseline parameters of all subjects are given in Table 1.
|Age (years)||33 ± 6||33 ± 4||0·988|
|Gestational week||28 ± 4||26 ± 4||0·076|
|Height (cm)||161 ± 6||163 ± 6||0·258|
|Weight (kg)||86 ± 17||75 ± 15||0·210|
|BMI (kg m−2)||33·2 ± 6·1||28·4 ± 5·2||0·006*|
|HbA1c (%)||5·3 ± 0·4||5·0 ± 0·5||0·010*|
|SBP (mmHg)||117 ± 13||114 ± 13||0·418|
|DBP (mmHg)||71 ± 8||75 ± 10||0·242|
Oral glucose tolerance test (oGTT)
Investigations were carried out at week 28 ± 4 of pregnancy and repeated 8 weeks after birth in both groups. Test procedures were performed according to American Diabetes Association recommendations . In brief, after overnight fasting, an antecubital intravenous cannula was placed and blood samples for glucose, insulin and inflammatory parameters were obtained, subjects then received a 75 g standardised glucose drink (Glucose 75 g Trinklösung, Mayrhofer Pharmazeutika GmbH&CoKG, Leonding, Austria). Venous blood samples for glucose and insulin were repeated after 1 and 2 h.
Glucose and insulin concentrations were determined immediately after sampling, using routine laboratory methods. Insulin resistance was estimated by homeostasis model assessment of insulin resistance (HOMA-IR), calculated using the formula: = fasting insulin (μU mL−1) × fasting glucose (mg dL−1)/405. Glycosylated haemoglobin (HbA1c) was measured by high performance liquid chromatography (Diamat, Bio-Rad, Hercules, CA, USA). Remaining samples were centrifuged and stored at -80 °C until batch analysis. YKL-40 was measured by ELISA (Metra®YKL-40, Quidel Corporation, San Diego, CA, USA), with means of duplicate measurements used for analysis. YKL-40 results > 3× SD of mean values from all subjects were excluded from analysis.
All data were tested for normal distribution and compared using Student’s paired or unpaired t-test, as appropriate. Data with skewed distribution were log-transformed prior to analysis. Pearson’s product moment correlation was used for calculation of associations between variables, followed by multivariate linear regression analysis. Statistical calculations were carried out using the SPSS software package (Release 14·0, SPSS Inc, Chicago, IL, USA). Values are expressed as mean values and SD, unless indicated otherwise; P < 0·05 was considered significant.
Both study groups were well comparable in most parameters at inclusion. As expected, GDM patients had a higher BMI and HbA1c than healthy controls (Table 1), as well as a higher weight and BMI prior to pregnancy, and a higher rate of macrosomia in previous pregnancies (data not shown). There was a tendency for positive family history of diabetes in GDM (P = 0·069), but no difference in the number of previous pregnancies between groups.
Oral glucose tolerance test
In the diagnostic oGTT at gestational week 28 ± 4, women with GDM had higher glucose values at all three sample points, and fasting insulin was significantly higher in women with GDM, with no difference between groups in the 1- and 2-h insulin concentrations (Table 2). HOMA-IR was significantly different between groups.
|YKL-40 (ng mL−1)||65·8 ± 44·4||60·3 ± 30·1||63·4 ± 30·5||66·9 ± 32·7|
|HOMA-IR||3·1 ± 1·7*||1·6 ± 0·9||4·2 ± 6·2||1·9 ± 1·8|
|fasting glucose (mg dL−1)||88 ± 13*||76 ± 10||92 ± 14*||83 ± 9|
|1-h-glucose (mg dL−1)||200 ± 31*||146 ± 37||159 ± 38*||114 ± 28|
|2-h-glucose (mg dL−1)||160 ± 33*||112 ± 28||117 ± 35*||94 ± 26|
|fasting insulin (μU mL−1)||14·1 ± 7·4*||8·3 ± 4·3||17·3 ± 26·0||9·3 ± 9·4|
|1-h-insulin (μU mL−1)||101·2 ± 69·9||97·2 ± 54·7||83·5 ± 76·3||53·2 ± 37·9|
|2-h-insulin (μU mL−1)||97·4 ± 47·1||89·5 ± 67·7||49·1 ± 29·8||40·0 ± 54·7|
Eight weeks after birth, differences in glucose concentrations in the oGTT (and HbA1c) remained significant between groups. However, HOMA-IR was no longer significantly different between groups. As expected, 1- and 2-h-glucose and -insulin concentrations were lower in both groups after birth when compared to values during pregnancy.
YKL-40 was not different between groups (Table 2), neither during (P = 0·584), nor after pregnancy (P = 0·681) (Fig. 1). YKL-40 did not differ between pregnancy and 8 weeks after delivery, neither within the GDM group (P = 0·607), nor in the controls (P = 0·368). We observed no difference in ΔYKL-40 between groups (P = 0·165).
YKL-40 during pregnancy was correlated with fasting (r = 0·352, P = 0·011), 1- and 2-h insulin concentrations in the oGTT (r = 0·461, P = 0·001 and r = 0·321, P = 0·02, respectively), as well as with HOMA (r = 0·335, P = 0·016) and BMI (r = 0·458, P = 0·001). These parameters were entered in multivariate linear regression analysis, and 1-h-insulin remained as strongest determinant for YKL-40 concentrations (β = 0·398, P = 0·038).
In this study, we did not find any difference in YKL-40 concentrations between women with GDM and healthy pregnant controls, neither during pregnancy nor after delivery. However, YKL-40 was associated with metabolic parameters in the oGTT during pregnancy.
Our findings are in contrast to several reports in the literature linking insulin resistance in pregnancy and GDM to inflammatory processes. Elevated CRP in the first trimester is associated with subsequent GDM , but CRP appears to be determined by maternal obesity rather than GDM status . Tumour necrosis factor-α (TNFα), on the contrary, was correlated with insulin resistance in late gestation irrespective of fat mass . While we did not measure these parameters, earlier results from our own group concerning other inflammatory parameters, monocyte chemoattractant protein-1 (MCP-1) and soluble CD40 ligand (sCD40L) , were confirmed in this study with no difference between groups (data not shown). It is important to note that we previously found increased MCP-1 in GDM vs. healthy pregnant controls only later in gestation (week 33 ± 5), while blood sampling was performed at week 28 ± 4 in this study. Sub-clinical inflammation might progress during pregnancy, as do insulin resistance and GDM. Therefore, it is possible that results for YKL-40 could have been different at a later point in pregnancy as well. Interestingly, Madazli et al.  analysed YKL-40 in normal and pre-eclamptic pregnancies after gestational week 30 and reported higher concentrations than those we found in our study, but also no difference between groups. In contrast, a recent study by Seol et al.  found significantly elevated YKL-40 in pre-eclampsia before delivery, but with lower concentrations than Madazli et al., and with less than half the concentrations in healthy pregnant Asian women than in our study. We did not measure YKL-40 in age-matched healthy nonpregnant women, and both lower [17,22] and higher  YKL-40 concentrations have been reported for healthy controls in the literature. Considering that the cited reports investigated older populations, and that YKL-40 increases with age , it remains unclear whether YKL-40 might be elevated in pregnancy irrespective of GDM status. Our data indicate a similar inflammatory status at the beginning of the third trimester in both GDM and healthy pregnancies.
YKL-40 did not change 8 weeks after delivery, neither in GDM nor in controls and was not different between groups. In GDM patients, increased TNFα and CRP have been reported 3 months after delivery  and insulin resistance and elevated TNFα can persist even longer . Again, it remains unclear whether sampling at a later time point might have yielded different results, especially as metabolic changes were still apparent in our subjects after GDM (Table 2).
In correlation analysis, YKL-40 was correlated with insulin concentrations in the oGTT as well as with HOMA-IR and BMI during pregnancy. Increased levels of YKL-40 have been reported in type 2 diabetes, but these were not correlated with BMI , which is in contrast to our findings. However, an association of YKL-40 with insulin and HOMA-IR has been described in the same study, which is confirmed by our results. Nevertheless, the association of YKL-40 with these metabolic parameters did not result in increased concentrations of YKL-40 in GDM. However, YKL-40 concentrations in pregnancy were nearly as high as those reported in type 2 diabetes . Several studies have demonstrated a connection of YKL-40 to microvascular and macrovascular disease. In type 1 diabetic patients, YKL-40 increases with levels of albuminuria . YKL-40 is elevated in patients with coronary artery disease  and after acute myocardial infarction [16,26]. Metabolic changes in GDM tend to resolve after delivery but an increased risk for type 2 diabetes remains. While the time of overt GDM might be too short for development of vascular complications and a significant elevation of YKL-40, it could be possible that a continued increase in YKL-40 after GDM might contribute to a later emergence of type 2 diabetes.
There are several limitations to our study. All women participating in the current study were pre-selected by referral to our specialised clinic for oGTT. Women included in this study were from a mixed ethnic background, which might have influenced our results. Furthermore, GDM severity differed between subjects, and not all required insulin therapy at time of measurements.
In conclusion, our findings indicate a similar inflammatory status in gestational week 28 ± 4 in both GDM and healthy pregnancies, and inflammatory parameters did not change 8 weeks after delivery. Further studies are needed to clarify the role of inflammation in GDM and possible consequences for related complications later in life.
Department of Medicine I, Rudolfstiftung Hospital, Vienna, Austria (G. Schaller, J. M. Brix, G. Placher-Sorko, F. Höllerl, G. Schernthaner); Department of Medicine II, Medical University of Vienna, Vienna, Austria (G.-H. Schernthaner).