Silencing of acid-labile subunit (ALS) improved glucose metabolism in animal models. The aim of this study is to evaluate the effects of rosiglitazone (RSG) on ALS levels in individuals with type 2 diabetes. A randomized, double-blind, placebo-controlled trial was conducted. Subjects with type 2 diabetes mellitus were randomly distributed to an RSG-treated
Acid-labile subunit (ALS) is a 63.3 kDa glycoprotein that is encoded by the
Peroxisome proliferator-activated receptor (PPAR
Recent studies have shown that PPAR
This protocol was approved by the Human Research Committee of the National Taiwan University College of Medicine, National Taiwan University Hospital, and Taiwan Department of Health and is registered in the Clinical Trials Protocol Registration System (
A double-blind, placebo-controlled, parallel-group comparative study was conducted between 1999 and 2000 to evaluate the effects of RSG (BRL 49653C) and concurrent sulfonylurea therapy [
Flowchart of the study design.
Body weight, height, blood pressure, and heart rate were measured. Body mass index (BMI) was calculated as weight (kg)/height (m2). As the WHO suggested population-specific cut-off points for BMI to identify those with increased risk for type 2 diabetes and cardiovascular disease [
The primary endpoint of the study was the treatment-induced change in HbA1c from baseline to week 24. A sample size of 52 patients (26 for each treatment group) was determined after considering a 20% dropout rate and a 90% power to detect a difference of 1.1 in HbA1c between treatment groups (if the standard deviation of the response is 1.1, based on an
Descriptive data are presented as means ± standard deviations or as percentages for categorical variables. Student’s
Of the 61 enrolled subjects, 30 were randomly assigned to the RSG treatment group and 31 were assigned to the placebo group. The mean patient age was
Clinical characteristics of the study subjects at baseline.
Placebo group | RSG group |
| |
---|---|---|---|
|
31 | 30 | |
Age |
|
|
0.668 |
Female (%) | 58.1% | 56.7% | 1.000 |
Body weight (kg) |
|
|
0.943 |
Body length (m) |
|
|
0.932 |
BMI (kg/m2) |
|
|
0.868 |
Systolic blood pressure, mmHg |
|
|
0.037 |
Diastolic blood pressure, mmHg |
|
|
0.053 |
HbA1c, (%) |
|
|
0.646 |
Fasting plasma glucose, mmol/L |
|
|
0.801 |
Fasting plasma insulin, pmol/L |
|
|
0.391 |
Total cholesterol, mmol/L |
|
|
0.055 |
Total triglyceride, mmol/L |
|
|
0.583 |
LDLc, mmol/L |
|
|
0.134 |
HDLc, mmol/L |
|
|
0.450 |
HOMA2-%S |
|
|
0.383 |
HOMA2-%B |
|
|
0.631 |
HOMA-IR |
|
|
0.368 |
HOMA- |
|
|
0.502 |
ALS, mU/mL |
|
|
0.854 |
Each value represents the mean ± standard deviation. The
Women had higher ALS levels than men (
Relationships between acid-labile subunit (ALS) levels and metabolic parameters. The correlations of ALS levels with low-density lipoprotein cholesterol (LDLc) concentrations (a) and homeostatic model assessment version 2 insulin sensitivity (HOMA2-%S) (b) were significant at baseline.
Five subjects (2 in the RSG-treated group and 3 in the placebo group) were lost to followup for personal or nonmedical reasons and/or had missing data of ALS levels at the 24-week endpoint.
Using separate mixed models, we found that the RSG-treated group experienced progressive improvements in HbA1c levels, fasting plasma glucose levels, and the HOMA2-%B compared with the placebo group over the 24-week study period. The RSG-treated group had decreased HOMA-insulin resistance (HOMA-IR) as compared with the placebo group at 12 weeks (
Changes in metabolic parameters after 24 weeks of treatment.
12 weeks | 24 weeks |
| |
---|---|---|---|
|
|||
Placebo group | 29 | 28 | |
RSG group | 28 | 28 | |
Body weight (kg) | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
BMI (kg/m2) | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
Systolic blood pressure, mmHg | |||
Placebo group |
|
|
0.083 |
RSG group |
|
|
|
Diastolic blood pressure, mmHg | |||
Placebo group |
|
|
0.901 |
RSG group |
|
|
|
HbA1c (%) | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
Fasting plasma glucose, mmol/L | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
Fasting plasma insulin, pmol/L | |||
Placebo group |
|
|
0.505 |
RSG group |
|
|
|
Total cholesterol, mmol/L | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
Total triglyceride, mmol/L | |||
Placebo group |
|
|
0.822 |
RSG group |
|
|
|
LDLc, mmol/L | |||
Placebo group |
|
|
<0.001 |
RSG group |
|
|
|
HDLc, mmol/L | |||
Placebo group |
|
|
0.566 |
RSG group |
|
|
|
HOMA2-%S | |||
Placebo group |
|
|
0.533 |
RSG group |
|
|
|
HOMA2-%B | |||
Placebo group |
|
|
0.011 |
RSG group |
|
|
|
HOMA-IR | |||
Placebo group |
|
|
0.197 |
RSG group |
|
|
|
HOMA- |
|||
Placebo group |
|
|
0.211 |
RSG group |
|
|
|
Acid-labile subunit, mU/mL | |||
Placebo group |
|
|
0.627 |
RSG group |
|
|
Each value represents the mean ± standard deviation (
Owing to the fact that growth hormone (GH) secretion is blunted in obese individuals [
Changes of serum ALS levels by treatment in subjects categorized as nonobese (BMI < 24 kg/m2) or obese (BMI ≥ 24 kg/m2).
Nonobese subjects (BMI < 24 kg/m2) | Obese subjects (BMI ≥ 24 kg/m2) | |||
---|---|---|---|---|
Placebo | RSG | Placebo | RSG | |
|
10 | 9 | 21 | 21 |
Baseline ALS levels, mU/mL |
|
|
|
|
24-week ALS levels, mU/mL |
|
|
|
|
Difference |
|
|
|
|
The correlations between HOMA2-%S and serum LDLc with ALS concentrations, analyzed with 4 linear regression models. Baseline serum ALS concentrations and change of ALS at 24 weeks were the dependent variables.
Baseline ALS | Change of ALS at 24 weeks | |||||||
---|---|---|---|---|---|---|---|---|
HOMA2-%S |
|
LDLc |
|
Change in |
|
Change in |
| |
Unadjusted |
|
0.025 |
|
0.025 |
|
0.156 |
|
0.930 |
Model 1: |
|
0.156 |
|
0.010 |
|
0.154 |
|
0.718 |
Model 2: |
|
0.158 |
|
0.009 |
|
0.060 |
|
0.976 |
Model 3: |
|
0.196 |
|
0.048 | ||||
Model 4: |
|
0.012 |
|
0.889 |
Baseline levels of ALS and HOMA2-%S were log-transformed. ALS: acid-labile subunit; HbA1c: hemoglobin A1c; BMI: body mass index; HOMA2-%S: homeostatic model assessment version 2 insulin sensitivity; LDLc: low-density lipoprotein cholesterol.
Change in total cholesterol and acid-labile subunit (ALS) in the placebo (▲) and the rosiglitazone group (○).
To the best of our knowledge, this is the first study to examine the correlation of ALS levels with metabolic phenotypes and the effect of the insulin-sensitizer RSG on ALS levels in subjects with type 2 diabetes. We found that, at baseline, ALS levels were highly correlated with age, height, fasting plasma insulin levels, HOMA2-%S, and serum LDLc concentrations. After 24 weeks, we observed a significant decrease in ALS levels in nonobese subjects with type 2 diabetes treated with RSG, as compared with the placebo group. The effect of treatment on ALS levels was not observed in obese individuals, indicating a heterogeneous response to RSG therapy according to BMI of the subjects.
Serum ALS levels were lower in men than women with type 2 diabetes. In a previous meta-analysis, it had been shown that testosterone levels were lower in men and higher in women with type 2 diabetes as compared to the healthy controls [
Our results are consistent with previous observations in animals and provide the first demonstration that an increase in serum ALS levels is associated with insulin resistance in patients with type 2 diabetes. Our study showed that subjects with lower serum ALS had less insulin resistance. When considering the lower ALS levels in older people [
RSG was demonstrated to improve insulin resistance in large clinical trials [
On the basis of the correlations between ALS, insulin sensitivity, and LDLc at baseline, we tested if RSG decreased ALS levels and if a change in ALS is beneficial. An inhibitory effect of RSG treatment on ALS levels was noted only in the subgroup of nonobese individuals with type 2 diabetes (Table
ALS was negatively correlated with LDLc levels at baseline (Table
Despite the strengths of this study, such as the standardized methods used to collect the information and proper blood sample storage, there are certain limitations that need to be considered. First, the sample sizes were relatively small, which might severely reduce the statistical power for subgroup analyses according to BMI status. In the nonobese group, the total sample size was 19. A power analysis showed a 1-
Our findings revealed a BMI-dependent effect of RSG treatment on ALS levels in individuals with type 2 diabetes. In nonobese diabetic subjects, serum ALS levels decreased after RSG treatment. Insulin sensitivity and LDLc were associated with ALS levels at baseline. The change in plasma ALS concentration predicted the change in total cholesterol concentration and correlated with the change in insulin sensitivity. Further studies will be required to clarify the effects of RSG on insulin sensitivity via the GH-IGF1-ALS axis and the mechanism of ALS influencing cholesterol metabolism.
The authors declare that there is no conflict of interests regarding the publication of this paper.
Lee-Ming Chuang, Ying-Chuen Lai, Ta-Jen Wu, and Chi-Yuan Jeng conceived and designed the experiments; Lee-Ming Chuang and Ying-Chuen Lai performed the experiments; Lee-Ming Chuang, Ta-Jen Wu, and Chi-Yuan Jeng contributed reagents/materials/analysis tools; Ying-Chuen Lai, Lee-Ming Chuang, and Hung-Yuan Li analyzed the data; Ying-Chuen Lai wrote the paper; Ying-Chuen Lai, Lee-Ming Chuang, Hung-Yuan Li, Ta-Jen Wu, and Chi-Yuan Jeng discussed the results; and Lee-Ming Chuang, Ying-Chuen Lai, Hung-Yuan Li, and Ta-Jen Wu commented on the paper.
The authors thank Ms. Kuan-Ching Lee and Ms. Jao-Ping Wang for their technical assistance. The study was funded by a Grant from the Department of Education (89-B-FA01-1-4) of Taiwan and by the Diabetes Research Foundation of the National Taiwan University Hospital, Taipei, Taiwan. The funding agencies had no role in study design, data collection and analysis, decision to publish, or preparation of the paper.