| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Endocrinology, University College London Hospitals (J.E.O., G.S.C.), London W1T 3AA, United Kingdom; and Adipokines and Metabolism Research Group, Center for Clinical Pharmacology, Department of Medicine, University College London (M.J.H.A., V.M.-A.), London WC1E 6JJ, United Kingdom
Address all correspondence and requests for reprints to: Dr. Gerard S. Conway, Department of Endocrinology, Middlesex Hospital, Mortimer Street, London W1T 3AA, United Kingdom. E-mail: g.conway{at}ucl.ac.uk.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Obesity is common in TS (4, 5) and is associated with elevated serum triglyceride and cholesterol concentrations (4, 6), which may contribute to the increased cardiovascular risk. The obesity in TS appears to be predominantly central, with an increased waist/hip ratio and a reduced fat-free mass (5).
The associations between obesity and type 2 diabetes mellitus (7, 8) as well as other features of the insulin resistance/metabolic syndrome, such as hypertension, hypertriglyceridemia, and low high-density lipoprotein (HDL) concentrations, are well established in the general population (9, 10). Because central obesity, hypertension, dyslipidemia, and impaired glucose homeostasis are common in TS (4, 5), we set out to determine whether these are indeed features of metabolic/insulin resistance syndrome in TS.
The aim of this study was to investigate the relationships between markers of the metabolic syndrome and obesity, with particular reference to the adipokines IL-6 and leptin in women with TS.
| Subjects and Methods |
|---|
|
|
|---|
Clinical history and case notes were reviewed for previous medical history. History of estrogen replacement, previous GH and oxandrolone administration, smoking, exercise, and current medications were recorded.
Subjects attended after a 12-h overnight fast. Antihypertensive medication was withheld for at least 24 h. Clinical parameters, such as height, weight, and waist and hip measurements, were recorded. Recumbent blood pressure was recorded at the right brachial artery using an automated sphygmomanometer (Dinamap) after subjects had rested for at least 15 min. Five readings were taken, and mean systolic and diastolic blood pressures were calculated. A large cuff was used for obese individuals.
A blood sample was taken for analysis of lipids, glucose, insulin, leptin, high sensitivity C-reactive protein (CRP), IL-6 concentration, and karyotype. The samples were stored on ice before either being assayed in the laboratory or frozen at 80 C and were analyzed later. The homeostasis model insulin resistance index (HOMA-R) was calculated as the product of fasting insulin and glucose concentrations divided by 22.5.
CRP was determined with an enzyme immunoassay using rabbit antihuman antibodies (X0293) from Dako Diagnostics (Ely, UK), validated against the United Kingdom Reference Preparation, with an assay range of 0.150.48 mg/liter and inter- and intraassay coefficients of variation (CVs) less than 10% as previously described (11). IL-6 was measured by high sensitivity two-site ELISA (R&D Systems, Inc., Minneapolis, MN). The inter- and intraassay CVs were less than 10%. The minimum detectable dose of IL-6 was 0.09 pg/ml. This method has previously been described (12). Leptin was measured in serum using a well-validated in-house RIA, with an antihuman rabbit monoclonal antibody and 125I-labeled leptin. The assay was sensitive to 0.1 µg/liter, with inter- and intraassay CVs less than 10% as previously described (13). Each assay was performed in three batches, each consisting of a random selection of TS, POF, and normal control samples. All assays included a low, medium, and high internal control from which inter- and intraassay CVs were calculated.
Statistical analysis
Statistical analysis was performed using SPSS version 11.0 for Windows (SPSS, Inc., Chicago, IL). For the cross-sectional analyses, continuous variables were compared by ANOVA, using log transformation where appropriate, and Scheffés post hoc analysis. Categorical variables were compared by
2 test. Estrogen exposure in TS and POF groups was compared by Mann-Whitney U test. Associations between variables were assessed using Pearsons and Spearmans correlation coefficients for parametric and nonparametric variables, respectively. Linear regression analysis (entry method) was used to assess the strength and independence of associations between variables. Interaction analysis was performed by univariate ANOVA to assess the effect of TS status in regression equations (Fig. 1
).
|
| Results |
|---|
|
|
|---|
Table 1
compares clinical characteristics among the three groups of subjects. In addition, women with TS smoked less (84.5, 56.7, and 51.6% of nonsmokers for TS patients, normal subjects, and 46,XX POF patients, respectively; P < 0.001), were more sedentary (60.0, 43.3, and 48.4% took no exercise among TS patients, normal subjects, and 46,XX POF patients, respectively; P = 0.002). The prevalence of karyotype 45,X was 59.6% in women with TS, with the remainder having TS karyotypes isochromosome X, partial X deletion, ring X, Y-chromosome fragment, mosaic 45,X/46,XX, and complex forms. Thirty-five percent of TS women had received previous GH therapy, and 26.5% had previously been treated with oxandrolone. TS women started estrogen therapy at a younger age than women with POF [median (range), 14 (534) vs. 20 (1238) yr; P < 0.001], but the overall years of estrogen deficiency were similar [3 (022) vs. 5 (016) yr; P = 0.164]. In the TS group, the percentages of women taking oral synthetic, oral natural, and patch estrogens were 20, 69, and 11%, respectively, with 13, 80, and 7%, respectively, in the 46,XX POF group (by
2 test, P = 0.517). Twenty-two (18.8%) women with TS, but none of the normal or 46,XX POF controls, were taking antihypertensive medication.
|
Women with TS had greater body mass index (BMI), waist/hip ratio, absolute waist circumference, and blood pressure than women in the other two groups (Table 1
). Triglyceride concentrations in the TS group were similar to those in the 46,XX POF group and greater than those in normal controls. The group differences in triglycerides were no longer significant when adjusted for waist circumference (P = 0.071), but remained so when adjusted for BMI (P = 0.030).
Fasting glucose and insulin concentrations and calculated HOMA-R score were lower in women with TS. CRP and IL-6 concentrations were greater, and leptin concentrations were lower in TS women than in both other groups (Fig. 1
). On interaction analysis, there was no difference in the relationship between waist and insulin, leptin, IL-6, or CRP in TS and control women (Fig. 1
). TS status was, however, associated with insulin, HOMA-R, CRP, and leptin concentrations independently of either waist circumference or BMI (P < 0.001 for all). The group differences in IL-6 concentrations were not significant after adjustment for these physical markers of obesity.
Analyses within the TS cohort
Anthropometric measures of obesity (weight, waist circumference, and BMI) were strongly intercorrelated (r
0.800; P < 0.001 for all) and were associated with biochemical markers of obesity. In comparison, the waist/hip ratio was weakly associated with other markers of obesity (correlation with BMI: r = 0.243; P = 0.010; weight: r = 0.228; P = 0.016). Waist circumference was associated with height (r = 0.263; P = 0.005); BMI (r = 0.855; P < 0.001); and total, low density lipoprotein, and HDL cholesterol (r = 0.215, P = 0.024; r = 0.229, P = 0.016; and r = 0.262, P = 0.006, respectively). There were also associations between waist circumference and triglycerides (r = 0.410; P < 0.001), fasting insulin (r = 0.494; P < 0.001), and HOMA-R (r = 0.491; P < 0.001), but not fasting glucose (r = 0.154; P = 0.110). Serum leptin, CRP, and IL-6 were all strongly correlated with waist circumference (r = 0.556, 0.402, and 0.370, respectively; P < 0.001 for all; Fig. 1
). In addition, leptin was strongly correlated, and CRP and IL-6 were weakly correlated with insulin concentrations (r = 0.267, P = 0.004; r = 182, P = 0.055; and r = 0.185, P = 0.051, respectively). The associations between waist circumference and all three adipokines/proinflammatory markers were independent of the association with insulin.
There was no significant association between measures of obesity and karyotype or previous history of GH and oxandrolone therapy. There was no significant difference in any metabolic parameter between groups of women with either TS or POF receiving different types of estrogen replacement.
| Discussion |
|---|
|
|
|---|
Our findings confirm increased central adiposity in women with TS (4, 5). We found that the waist/hip ratio, which has been used as a marker of obesity and cardiovascular risk in the general population (14, 15, 16), appears to be a poor index of obesity in TS. This may be because hip development is less pronounced, and even slim women, therefore, have a relatively high waist/hip ratio. Waist circumference and BMI appear to be better markers of obesity and associated cardiovascular risk in TS, consistent with data from other populations (17) and current guidelines for risk assessment (10).
With regard to metabolic profiles, we show similarly elevated triglyceride concentrations in TS and POF women compared with normal controls and suggest that this may be an effect of estrogen therapy, as previously demonstrated in postmenopausal women (18). The similarity of HDL concentrations in the three groups may also be attributable to estrogen therapy.
This study found women with TS to have lower fasting glucose and insulin concentrations than normal controls and women with POF. A reduction of fasting insulin and glucose concentrations paradoxically associated with a deterioration in glucose tolerance has been described in TS (5) and postmenopausal women (19) treated with estrogen. This phenomenon has been attributed to glucagon antagonistic and glucocorticoid stimulatory effects of estrogen (20). In addition, a lower incidence of diabetes mellitus has been described in postmenopausal women receiving estrogen replacement compared with nonusers (21). Because both the TS and POF groups in our study were treated with estrogen, however, it is unlikely that the lower glucose and insulin concentrations in the TS women are attributable to estrogen therapy.
The high prevalence of impaired glucose tolerance in women with TS (1, 5) appears to be secondary to an insulin secretory defect (22) or a glucose storage defect (23). As recently as 35 yr ago, it was suggested that the pathogenesis of diabetes in TS may be distinct from that of type 2 diabetes mellitus (24), and the absence of fasting hyperinsulinemia, despite excess central obesity, is consistent with this (25). The role of the metabolic syndrome, therefore, becomes questionable in TS.
We report elevated CRP and IL-6 and reduced leptin concentrations in TS patients. These parameters were measured as indicators of body fat (26, 27, 28, 29) and inflammatory markers of cardiovascular risk (30, 31). CRP and IL-6 are predominantly associated with visceral fat (32, 33, 34), although the relationship is complex. Oral estrogen therapy itself is known to increase CRP concentrations via an effect on the liver (35), but CRP concentrations have been shown to predict cardiovascular risk independently of estrogen therapy (36). The effect of oral estrogen on IL-6 concentrations is inconsistent (35, 37, 38, 39).
Given the significantly higher CRP concentrations in TS women compared with both other groups, it is unlikely that they can be attributed to the effect of estrogen, because the POF women were also treated with estrogen. IL-6 concentrations were elevated in TS women compared with normal controls, but not compared with POF women, so an estrogen effect is difficult to determine and may contribute. Notably, both CRP and IL-6 were strongly correlated with waist circumference. Thus, their elevated concentrations in TS women are likely to be attributable to the increased central adiposity, although other factors may also contribute, because they are nonspecific inflammatory markers.
The concentration of leptin, which originates predominantly from sc fat (26, 40), but is generally elevated in subjects of increased adiposity, was lower in women with TS than in the other two groups. A previous study did not demonstrate, but inferred, hypoleptinemia in TS women compared with normal controls because of the greater percentage of body fat (41). It is possible that low leptin levels in TS result from reduced sc adiposity. Another possible explanation is that the reduced leptin concentrations are related to low fasting insulin concentrations, because leptin production is regulated by chronic, rather than acute, insulin concentrations (42). The correlation between leptin and insulin in TS women in this study was not independent of the effect of waist size, however. The metabolic syndrome is typically, but not unequivocally, associated with increased leptin concentrations (43, 44). A unifying hypothesis to explain the low insulin and leptin concentrations in the context of greater obesity would be that the insulin secretory defect is the primary abnormality, causing low leptin concentrations, which then result in inhibited satiety mechanisms (45, 46).
In conclusion, women with TS have various physical and biochemical features suggestive of the metabolic/insulin resistance syndrome, but the evidence that they are all interrelated in a true metabolic syndrome is questionable. An excess prevalence of central obesity in TS is evident. Hypertriglyceridemia in TS may be partly related to estrogen therapy, although obesity probably contributes. The abnormalities of glucose homeostasis do not follow the classical pattern associated with the metabolic syndrome and may result from a unique metabolic defect. The elevated CRP and IL-6 concentrations are consistent with visceral adiposity, but the hyperleptinemia often associated with the metabolic syndrome was not evident. Although these markers are undoubtedly related to obesity in TS, other factors, such as estrogen therapy, low fasting insulin, and perhaps other sources of proinflammatory cytokines, appear to contribute. The apparent features of the metabolic syndrome in TS are not fully explained by the obesity, the history of estrogen deficiency, or indeed current estrogen therapy, and it may be that one or more TS-specific metabolic defects explain this apparent paradox.
| Footnotes |
|---|
First Published Online February 22, 2005
Abbreviations: BMI, Body mass index; CRP, C-reactive protein; CV, coefficient of variation; HDL, high-density lipoprotein; HOMA-R, homeostasis model insulin resistance index; IHD, ischemic heart disease; POF, premature ovarian failure; TS, Turner syndrome.
Received October 12, 2004.
Accepted February 10, 2005.
| References |
|---|
|
|
|---|
, in vivo. J Clin Endocrinol Metab 82:41964200This article has been cited by other articles:
![]() |
C. H. Gravholt, B. E. Hjerrild, L. Mosekilde, T. K. Hansen, L. M. Rasmussen, J. Frystyk, A. Flyvbjerg, and J. S. Christiansen Body composition is distinctly altered in Turner syndrome: relations to glucose metabolism, circulating adipokines, and endothelial adhesion molecules. Eur. J. Endocrinol., October 1, 2006; 155(4): 583 - 592. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |