| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Internal Medicine, Dokkyo Medical University Koshigaya Hospital, Koshigaya 343-8555, Japan
Address all correspondence and requests for reprints to: Kohzo Takebayashi, M.D., Department of Internal Medicine, Dokkyo Medical University Koshigaya Hospital, 2-1-50, Minami-Koshigaya, Koshigaya 343-8555, Japan. E-mail: takeb{at}gmail.plala.or.jp.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objectives: The aim was to investigate the association between RBP4 and various markers related to insulin resistance and diabetic complications in type 2 diabetic patients. The effect on RBP4 of the addition of pioglitazone to other diabetic medications was also examined.
Design, Setting, Patients, Intervention, and Main Outcome Measures: RBP4 levels were measured in 101 hospitalized patients with type 2 diabetes and in 22 nonhospitalized control subjects. Endothelial function was evaluated using flow-mediated vasodilatation. In a further 22 nonhospitalized type 2 diabetic patients, pioglitazone (30 mg/d) was administered for 12 wk while other medications for diabetes were continued.
Results: There was a significant elevation of RBP4 levels in diabetic patients compared with healthy subjects. RBP4 showed significant positive correlations with triglyceride, systolic blood pressure, and log urinary albumin excretion, and significant negative correlations with high-density lipoprotein cholesterol and flow-mediated vasodilatation. In stepwise regression analysis, log urinary albumin excretion, triglyceride, and gender showed a significant association with RBP4. RBP4 was significantly elevated in patients with proliferative-diabetic retinopathy compared with nondiabetic retinopathy and simple-diabetic retinopathy patients. The addition of pioglitazone for 12 wk to other diabetic medications the patients were already taking did not affect the serum RBP4 concentration.
Conclusions: The current study shows that RBP4 is associated with variables related to insulin resistance and diabetic complications. The addition of pioglitazone for 12 wk to other diabetic medications the patients were already taking did not affect serum RBP4 levels.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The main purpose of the study was to clarify how RBP4, as a newly identified adipocytokine, is related to the clinical characteristics of patients with type 2 diabetes. This was achieved through investigation of the correlation of circulating RBP4 levels with diabetic nephropathy, diabetic retinopathy, and factors related to insulin resistance: TG; HDL-C; BP, flow-mediated vasodilatation (FMD) reflecting vascular endothelial function (11, 12); 8-iso-prostaglandin F2
(8-iso-PGF2) as a marker of systemic oxidative stress (13, 14); and high-sensitivity C-reactive protein (hsCRP) as a marker of low-grade inflammation. The association between RBP4 and other adipocytokines such as leptin and adiponectin was also explored. Furthermore, although it has been shown that administration of rosiglitazone (a thiazolidinedione) decreases serum RBP4 in mice (3), this relationship has not been studied in humans. Therefore, the effect of pioglitazone, which is also a thiazolidinedione, on serum RBP4 in patients with type 2 diabetes was examined, which was done by adding pioglitazone to other diabetic medications that the patients were already taking, and the effect on RBP4 of the decrease in glucose level caused by insulin therapy was used as a control. In performing the work, we hypothesized that circulating RBP4 is closely associated with variables related to insulin resistance or diabetic complications and that the addition of pioglitazone would decrease the serum RBP4 level.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
From August 2005 to April 2006, 101 type 2 diabetes patients (62 men and 39 women) hospitalized for glycemic control and/or diabetic education were prospectively and consecutively enrolled. As controls, 22 age-matched healthy subjects were also enrolled. The clinical characteristics of the 101 diabetes patients (after diabetes was controlled during admission) and 22 control subjects are shown in Table 1
. None of the patients underwent a change in treatment after enrollment, except for treatment for diabetes.
|
-glucosidase inhibitor (GI); and six, SU and metformin], and two patients had been treated with an -GI and one with metformin as a single-use therapy; these drugs were continued concomitantly with pioglitazone therapy over the observation period. Two patients had been managed using dietary modification alone. Accordingly, 20 of the 22 patients were already taking other diabetic medications. No patients had received a thiazolidinedione, including pioglitazone. In addition, 11 patients who had not received thiazolidinediones and who were changed from SU to insulin therapy were simultaneously prospectively enrolled as controls; these patients were initially hospitalized for about 1–2 wk for induction of insulin therapy. Any patient showing evidence of liver dysfunction, infectious disease, or autoimmune disease was excluded from the study. Methods
All 101 patients in the cross-sectional study were hospitalized for 2–3 wk. Blood, urine, and FMD tests were performed simultaneously in the morning before breakfast after an overnight fast of at least 10 h, and after diabetes was controlled during hospitalization. These analyses were performed between 1 and 2 wk after admission, after the fasting plasma glucose (FPG) level had been lowered to less than 160 mg/dl, the value defined as poor diabetic control by the Japan Diabetes Society (15).
Measurement of FMD of the brachial artery. FMD was measured by a single specialist who had more than 6 months experience with the FMD measurement procedure. After the lumen diameter of the right brachial artery was measured at baseline, the cuff placed above the measurement point was inflated to a pressure 30 mm Hg greater than the patient’s systolic blood pressure (SBP) for 5 min. FMD was calculated using the formula: FMD (%) = [(maximal artery lumen diameter after cuff release – artery lumen diameter at baseline)/artery lumen diameter at baseline] x 100.
To confirm the reproducibility of this test, six young healthy subjects (mean age 30.7 ± 6.4 yr) underwent two consecutive FMD measurements with a 30-min interval, and the change in the FMD value was shown to be less than 5% in each subject (mean change 1.1 ± 1.8%).
Measurement of the intimal-medial complex thickness (not including plaques) (IMT) of the carotid artery. The IMT was measured as described in our previous study (16). To examine the reproducibility of the method, the IMT was measured twice in 11 healthy subjects. The change in the measured IMT was shown to be less than 10% in each subject (mean change 2.4 ± 3.9%). One subject also underwent five consecutive measurements of IMT, and the coefficient of variation (CV) was 4.3%.
Blood test. After collection, the blood was rapidly centrifuged at 1500 rpm for 5 min to separate serum and plasma from clot-containing blood cells. These samples were stored at –70 C until analysis.
Serum RBP4 assay. The tube used for blood collection did not contain any products. Serum was diluted 1000-fold for RBP4 measurements because of the high concentration of RBP4 in human serum. RBP4 was measured using a Retinol Binding Protein-4 (Human) EIA kit (Phoenix Pharmaceuticals, Inc., Burlingame, CA), with each value reported as a mean of duplicate measurements made on the same serum sample. The assay in this kit is linear for purified recombinant RBP4 (not actual human serum samples) from 1.72–28.9 ng/ml, and intraassay and interassay CVs are less than 5% and less than 14%, respectively (17). We did not confirm the linearity of dilution in human samples, and it is unknown if the assay exhibits linearity of dilution in samples from subjects with elevated levels of serum RBP4. Cross-validation of this kit with other methods for measuring serum RBP4, such as quantitative Western blotting, was not performed.
Other measurements.
Plasma glucose, glycosylated hemoglobin (HbA1C), serum lipid concentrations, and serum hsCRP were measured as described in our previous study (16). Measurements of urinary albumin excretion (UAE) (using 24-h urine collected from 1000 to 1000 h the following day) and urinary 8-iso-PGF2 (using morning urine) were also performed as previously described (18). Serum adiponectin was measured using an ELISA kit (Otsuka Pharmaceuticals, Tokyo, Japan), with intraassay and interassay CVs of 4.06% and 4.69%, respectively. Serum leptin was measured using a RIA kit (LINCO Research, Inc., St. Charles, MO), and the intraassay and interassay CVs were less than 8.3% and less than 6.2%, respectively. Vitamin A (retinol) (normal range 65–276 IU/dl) was measured by HPLC. Creatinine clearance (Ccr) was directly measured using urine samples stored for 24 h.
Assessment of diabetic retinopathy. Diabetic retinopathy was assessed by each patient’s ophthalmologist in our hospital according to Davis’ classification as no, simple, and proliferative diabetic retinopathy (NDR, SDR, and PDR, respectively) (19).
Pioglitazone and insulin therapy. These therapies were continued over 12 wk. Blood tests were performed at the initial point and after 12 wk, in each case before breakfast after at least a 10-h period of overnight fasting. All patients in the pioglitazone-treated group received 30 mg/d of pioglitazone, while other medications for diabetes were continued concomitantly, as described in Patients and Methods. For patients who changed from SU to insulin therapy, insulin was initiated at a dose of approximately the body weight (BW) of the patient (kg) x 0.2 U, and then the dose was adjusted based on the degree of glycemic control. No change in administration of any drug occurred for any patient during the 12-wk investigational period.
Ethical considerations. All subjects gave informed consent to inclusion in the study, which was performed according to the guidelines proposed in the Declaration of Helsinki. The study was approved by the Dokkyo Medical Ethics Committee (Koshigaya, Japan).
Statistical methods. All data are presented as mean ± SD. Comparison of two time points for an individual was performed using a paired t test. Comparison of RBP4 levels between the two groups was performed using an unpaired t test, and comparison among three groups was performed using a Bonferroni test. A P value < 0.05 was considered to indicate statistical significance in all analysis.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
. The correlations between RBP4 and TG, FMD, and log UAE are also shown in Fig. 1, A–C, respectively. RBP4 did not show a gender-related difference or a difference between smokers and nonsmokers, and there were no significant differences in RBP4 levels between patients who had and had not received antihypertensive drugs, statins, or insulin. In addition, no significant difference in RBP4 levels was observed between patients who received insulin and those who received SU (n = 45) (22.6 ± 7.8 vs. 25.2 ± 9.8 µg/ml, respectively; P = 0.1474). Regarding retinopathy, there was a significant elevation of RBP4 in patients with PDR (n = 28) compared with those with NDR (n = 58) or SDR (n = 15): 29.3 ± 9.6 vs. 21.4 ± 6.1 (P < 0.0001); and 22.7 ± 11.6 µg/ml (P = 0.0128) (Fig. 1D). When the RBP4 levels were adjusted for UAE, TG, and FMD using a general linear model, multiple analysis of covariance showed that a tendency for a significant difference between RBP4 levels remained in patients with PDR and NDR (26.7 ± 7.9 vs. 22.5 ± 6.8 µg/ml; P = 0.0970). In contrast, the difference in RBP4 levels between patients with PDR and NDR was significant by analysis of covariance when adjusted only for TG or FMD (P < 0.0001, respectively), and a tendency for a significant difference was obtained with adjustment for UAE only (P = 0.1067). Furthermore, the significant difference in RBP4 levels between patients with PDR and NDR or SDR also remained after adjustment for HbA1C (P < 0.0001, P = 0.0400, respectively, by analysis of covariance). We also tried to stratify the patients into two groups based on the degree of UAE. The limiting value giving a significant difference in RBP4 levels among retinopathy groups was estimated by performing principal component analysis based on the mean ± 2 SD; the significant difference between PDR (n = 8) and NDR (n = 45) or SDR (n = 12) patients persisted in patients with UAE less than 30 mg/gCr (P = 0.0398, P = 0.0170, respectively). The mean value of vitamin A (retinol) in the 101 patients was 157.3 ± 50.2 IU/dl, and only one patient showed a vitamin A value (56 IU/dl) less than the normal reference range. There was a significant positive correlation between RBP4 and vitamin A levels (R = 0.8931; P < 0.0001).
|
|
. Changes in RBP4 and adiponectin are also shown in Fig. 2, A and B. In the 11 diabetic patients who changed to insulin therapy, significant decreases in FPG and HbA1C occurred (206.8 ± 42.9 vs. 128.9 ± 60.1 mg/dl, P = 0.0087; and 10.8 ± 1.8 vs. 8.0 ± 1.8%, P = 0.0059), and a significant elevation of HDL-C was obtained (49.9 ± 20.1 vs. 60.0 ± 17.2 mg/dl; P = 0.0176). However, there was no significant change in RBP4 (19.4 ± 7.0 vs. 20.1 ± 6.8 µg/ml; P = 0.6746) or in BW, SBP, diastolic blood pressure (DBP), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and adiponectin in these patients; insulin levels and the homeostasis assessment model ratio (HOMA-R) were not investigated.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
We also found a strong positive correlation between RBP4 and TG, and a negative correlation between RBP4 and HDL-C. TG and HDL-C are closely associated with insulin resistance (5), and these results support the findings of Graham et al. (6), although the patient setting was different. Regarding the mechanism underlying the association between RBP4 and TG, a direct effect of RBP4 on activities such as progression of lipogenesis in liver may be plausible. In addition, production of TG in the liver and release of very low-density lipoprotein into blood generally increase in an insulin-resistant state, and this may also partially explain the close association between RBP4 and TG. We also found a significant correlation between RBP4 and SBP, although this association was not retained in multiple regression analysis. Because BP is associated with insulin resistance (5), this result appears to be reasonable and is also consistent with Graham et al. (6).
Next, we noted a significant negative correlation between circulating RBP4 and FMD; this association was also retained in multiple regression analysis, in which FMD showed a tendency to associate with RBP4 as a dependent variable. This result suggests that RBP4 may be associated with vascular endothelial dysfunction as a diabetic complication. The mechanism underlying this association may involve RBP4 inhibition of insulin action not only in skeletal muscle cells, but also in vascular endothelial cells; if this occurs, RBP4 may affect endothelial function directly through inhibition of insulin-mediated pathways for nitric oxide (NO) production in endothelial cells because in the insulin-resistant state, NO-dependent vasodilation is impaired by decreased NO production (21).
A strong positive association between RBP4 and log UAE was also observed. In the current study, we did not measure urinary RBP4 for comparison with serum RBP4, but there are reports demonstrating that the urinary RBP4 levels (but not the serum RBP4 level) reflect tubular injury or dysfunction (22, 23). Furthermore, urinary RBP4 and known urinary protein markers for glomerular dysfunction increase in diabetic nephropathy; marked excretion of urinary RBP4 is especially observed in patients with macroalbuminuria (24). Therefore, we speculate that the close association between serum RBP4 and log UAE is due to hyperproduction of RBP4, rather than impaired excretion of RBP4, because it has been proposed that increased production of RBP4 in adipose tissue may contribute to elevated RBP4 levels in insulin-resistant states (3), and because our patients did not have advanced renal failure (anuric or oliguric end-stage). The close association between serum RBP4 and log UAE observed in the current study suggests that serum RBP4 could be a marker for diabetic nephropathy.
Interestingly, we also found a significant elevation of RBP4 in patients with PDR compared with those with NDR or SDR; however, the difference in RBP4 levels was not statistically significant when the RBP4 level was adjusted for UAE, even between PDR and NDR patients, but a tendency for a significant difference persisted. A significant difference in RBP4 levels among retinopathy groups was also found in patients with UAE less than 30 mg/gCr. Although there is no apparent evidence that increased retinol transport mediated by RBP4 is toxic to the retina, this finding suggests that RBP4 is closely associated with diabetic retinopathy, which is one of the diabetic microangiopathies.
Together, the association of RBP4 levels with decreased FMD, increased UAE, and retinopathy suggests that the serum RBP4 level is predictive of and/or contributory to vascular complication in type 2 diabetes. However, additional studies are needed to examine the validity of this hypothesis.
We also investigated the association between RBP4 and oxidative stress, as evaluated by urinary 8-iso-PGF2. Oxidative stress in adipose tissue is reported to influence adipocytokines such as TNF-, plasminogen activator inhibitor (PAI)-1, and adiponectin (25). However, contrary to our expectation, the RBP4 level was not correlated with urinary 8-iso-PGF2. Therefore, our results suggest that, unlike other adipocytokines, the level of RBP4 is not influenced by oxidative stress, at least not to an extent that is clinically significant. In addition, RBP4 was not associated with leptin or adiponectin as other adipocytokines, showing that RBP4 is independent of these adipocytokines. Furthermore, RBP4 was not associated with hsCRP as a marker of inflammation, or with the IMT, which reflects systemic atherosclerosis (26, 27, 28), suggesting that RBP4 is not a surrogate marker for inflammation or atherosclerosis.
We failed to find an association between RBP4 and FPG or HbA1C. Yang et al. (3) reported no difference in glucose levels between rbp4 transgenic mice and wild-type mice, despite elevated RBP4 and insulin levels. This result suggests that RBP4 is not always associated with glucose levels because compensatory elevation of insulin can occur. In addition, we measured RBP4 after diabetes was controlled by treatment after admission as we simultaneously measured FMD, which is influenced by hyperglycemia (29). The lowered glucose levels under these conditions may have weakened the real association between RBP4 and glucose levels because RBP4 probably does not change as acutely as glucose levels (30); however, because we did not investigate serum RBP4 concentrations before intervention, we cannot prove directly whether the acute glucose-lowering therapy influenced the RBP4 levels. Furthermore, we recognize that the interval after intervention was insufficient to perform cross-sectional analyses of other variables, including glucose concentrations. Furthermore, no correlation of RBP4 with fasting insulin levels or HOMA-R was observed in hospitalized patients who did not receive insulin therapy; these results differ from those of Graham et al. (6). However, unlike our study, the patients in the study by Graham et al. (6) were not hospitalized, and the blood was measured in a steady state of glucose homeostasis. This is the primary reason why we failed to find associations between RBP4 and insulin or HOMA-R, and between RBP4 and glucose. We emphasize that this is a major limitation of this study. We also found no correlation between RBP4 and insulin or HOMA-R at the initial time point (steady state of glucose homeostasis) in pioglitazone-treated patients without apparent obesity; however, it is important to note that 20 of the 22 pioglitazone-treated subjects were already being treated with drugs that influence glucose homeostasis (metformin, SU, and -GI) at the time of initial measurement, unlike the patients in the study by Graham et al. (6). The preceding treatment with these drugs would influence the association between RBP4 and insulin or HOMA-R; this is also a major limitation in interpreting our data.
Serum RBP4 levels were relatively low, even in diabetic patients, compared with those reported by Graham et al. (6). In the Japanese population, the normal range of RBP4 is 24–79 µg/ml (30), suggesting that serum RBP4 levels in the current study were relatively low. Vitamin A deficiency could explain this result, but only one patient showed such a condition. Another possible explanation is that the assay used in this study underestimated the RBP4 level, compared with methods such as quantitative Western blotting, and a further study is needed to investigate variations among RBP4 measurement methods. In addition, the preceding treatment for diabetes in the current study may also explain the relatively low serum RBP4 levels. Furthermore, the diabetic patients were enrolled during hospitalization for glycemic control or diabetic education. Because RBP4 is known to be a negative acute-phase reactant, hospitalization may decrease the serum RBP4 levels because it might be stressful for the patients. This might partially account for the low RBP4 in this study population.
We studied the effects on RBP4 of the addition of pioglitazone (a thiazolidinedione) to other diabetic medications that patients with type 2 diabetes were already taking. However, contrary to our expectation, the addition of pioglitazone (30 mg/d) for 12 wk did not affect the serum RBP4 concentration. Regarding the discrepancy between our results and those of Yang et al. (3), we emphasize that our study design did not involve treatment of drug-naive subjects but, rather, used the addition of pioglitazone to an established medical regimen for the majority of the subjects; unlike our study, no medications were performed before the administration of pioglitazone in the study by Yang et al. (3). Because our study does not control for potential effects of the preexisting medical regimen on RBP4, this may have influenced the results. In addition, the species difference may account for the discrepancy, or the effects of the thiazolidinediones (pioglitazone vs. rosiglitazone) may be different, and an investigation of the effect of rosiglitazone on RBP4 in humans would be of value.
In conclusion, our results suggest that RBP4 is associated with certain markers related to insulin resistance or diabetic complications, but it remains uncertain whether RBP4 has a direct influence on these markers.
|
Footnotes |
|---|
First Published Online April 17, 2007
Abbreviations: BMI, Body mass index; BP, blood pressure; BW, body weight; Ccr, creatinine clearance; CV, coefficient of variation; DBP, diastolic blood pressure; FMD, flow-mediated vasodilatation; FPG, fasting plasma glucose; GI, glucosidase inhibitor; HbA1C, glycosylated hemoglobin; HDL-C, high-density lipoprotein cholesterol; HOMA-R, homeostasis assessment model ratio; hsCRP, high-sensitivity C-reactive protein; IMT, intimal-medial complex thickness (not including plaques); 8-iso-PGF2, 8-iso-prostaglandin F2; LDL-C, low-density lipoprotein cholesterol; NDR, no diabetic retinopathy; NO, nitric oxide; PAI, plasminogen activator inhibitor; PDR, proliferative diabetic retinopathy; RBP, retinol binding protein; SBP, systolic blood pressure; SDR, simple diabetic retinopathy; SU, sulfonylurea; TC, total cholesterol; TG, triglyceride; UAE, urinary albumin excretion.
Received June 13, 2006.
Accepted April 5, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Lim, S. H. Choi, I.-K. Jeong, J. H. Kim, M. K. Moon, K. S. Park, H. K. Lee, Y.-B. Kim, and H. C. Jang Insulin-Sensitizing Effects of Exercise on Adiponectin and Retinol-Binding Protein-4 Concentrations in Young and Middle-Aged Women J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2263 - 2268. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Mody, T. E. Graham, Y. Tsuji, Q. Yang, and B. B. Kahn Decreased clearance of serum retinol-binding protein and elevated levels of transthyretin in insulin-resistant ob/ob mice Am J Physiol Endocrinol Metab, April 1, 2008; 294(4): E785 - E793. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Qi, Z. Yu, X. Ye, F. Zhao, P. Huang, F. B. Hu, O. H. Franco, J. Wang, H. Li, Y. Liu, et al. Elevated Retinol-Binding Protein 4 Levels Are Associated with Metabolic Syndrome in Chinese People J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4827 - 4834. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ziegelmeier, A. Bachmann, J. Seeger, U. Lossner, J. Kratzsch, M. Bluher, M. Stumvoll, and M. Fasshauer Serum Levels of Adipokine Retinol-Binding Protein-4 in Relation to Renal Function Diabetes Care, October 1, 2007; 30(10): 2588 - 2592. [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 |
