SEARCH

SEARCH BY CITATION

Keywords:

  • bipolar disorder;
  • euthymic bipolar patients;
  • hyperinsulinemia;
  • overweight

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Aims:  Insulin resistance, quantified by hyperinsulinemia, has been identified as a preclinical state for metabolic syndrome. Acute medicated bipolar patients are vulnerable to hyperinsulinemia in early remission. We examined the proportion of fasting serum insulin levels in remitted bipolar patients and considered what factors may contribute to hyperinsulinemia.

Methods:  Measurements taken in this study included the fasting plasma levels of insulin, glucose, triglycerides, total cholesterol, high-density-lipoprotein-cholesterol, low-density-lipoprotein-cholesterol, and the body mass index (BMI) among 56 bipolar I manic patients in full remission. We examined serum insulin levels in order to explore the correlation between and within the hyperinsulinemia and non-hyperinsulinemia groups.

Results:  A total of 15 patients (26.8%) were identified as having hyperinsulinemia. Among all factors, only BMI was associated with higher serum insulin level (BMI ≥ 24: odds ratio of 8.57; P < 0.01; 95% confidence interval, 1.65–44.43).

Conclusion:  Hyperinsulinemia was not more prevalent in medicated euthymic bipolar patients compared with the general population. However, increasing bodyweight may make these patients more vulnerable to hyperinsulinemia, irrespective of their mood stabilizers or antipsychotics. Weight management should not be ignored in euthymic bipolar patients to prevent the preclinical state for metabolic syndrome.

BIPOLAR DISORDER IS associated with higher risk for metabolic syndrome,1 which is characterized by a cluster of metabolic risk factors, including obesity, dyslipidemia, hyperglycemia, and hypertension, and is linked to increased risk for type 2 diabetes, cardiovascular disease, and mortality.2,3 The high prevalence of metabolic syndrome in patients with bipolar disorder is suggested to be related to insulin resistance, obesity and/or visceral adiposity,4 and the introduction of novel psychotropic medications.1,5

Insulin resistance, quantified by hyperinsulinemia and considered a preclinical state for metabolic syndrome, plays a key role in the pathophysiology of metabolic syndrome, and has even been postulated to be its underlying cause.6 Our early work7 measured the serum insulin level in 42 physically healthy bipolar I manic (DSM-IV) patients aged ≤45 years with Young Mania Rating Scale (YMRS) scores of ≥26, and then in subsequent remission (YMRS ≤ 12). A total of 14 patients (33.3%) with acute mania and 30 patients (71.4%) in subsequent remission met the Taiwanese criteria for hyperinsulinemia, suggesting that medicated bipolar manic patients are vulnerable to hyperinsulinemia in early remission. Additionally, other factors may contribute to the elevation of insulin levels in bipolar patients, such as the use of beta-blockers for managing side-effects from psychotropic medications instead of the direct effects of mood stabilizers or atypical antipsychotics.7 Because continuation treatment is recommended for bipolar disorder,8 it is crucial to investigate insulin levels after the acute phase and during early remission. However, it remains unclear whether insulin resistance is commonly seen in bipolar patients during full remission.

We hypothesize that insulin levels in those patients may be altered and that those patients with hyperinsulinemia may be characterized by some clinical features. The purpose of this study was to examine the prevalence of fasting serum insulin levels and to explore the correlations in patients with bipolar disorder in full remission.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Subjects

The research protocol was approved by the Institutional Review Board of Taipei Medical University Hospital. All participants signed informed written consent forms before entering the study. Patients meeting the DSM-IV diagnostic criteria for bipolar I disorder, most recent manic, or mixed episode, in full remission, and aged between 16 and 45 years, were invited to participate in this study. The subjects were confirmed and rated by two experienced psychiatrists using the Chinese version of the Structured Clinical Interview for DSM-III-R, patient edition.9 Patients in full remission were defined as having YMRS10 scores <5 and Hamilton Depression Rating Scale11 total scores <7 continuously for at least 2 months. We recruited the patients from 8 to 24 weeks following a manic episode to minimize the long-term effects of medication. The exclusion criteria were the presence of any known surgical conditions and physical illnesses, such as ischemic heart disease, hypertension, type 1 and type 2 diabetes mellitus (characterized by a lack of insulin production and the body's ineffective use of insulin, respectively), organic mental diseases, mental retardation or dementia, any history of alcohol/substance abuse, pregnancy or breasting-feeding in women, and any prior history of lipid-lowering treatments. A total of 56 patients with bipolar I disorder (27 men and 29 women) were recruited and had a mean age of 34.0 ± 8.8 years and mean age at onset of 23.3 ± 7.1 years. The reference range for normal fasting insulin was 2.1 to 30.8 µIU/mL, with hyperinsulinemia being defined by the top quartile cut-off for fasting insulin, specified by gender, in the non-diabetic background Taiwanese population (≥8.7 µIU/mL [52.45 pmol/L] for men and ≥11.3 µIU/mL [67.84 pmol/L] for women).12 Therefore, the two-group comparison (with and without hyperinsulinemia) was performed according to this criterion. A total of 15 patients (26.8%) were identified with hyperinsulinemia.

Procedure

After recruiting the subjects, we evaluated the available clinical data and all other available information, including a review of the subjects' medical records and family member contacts, so that accurate diagnoses of all patients could be obtained. After an overnight fasting from 24:00 h on the preceding night, heparinized blood was obtained via a venous puncture and plasma was collected and frozen at −80°C until use. The intra- and inter-assay coefficient of variation (CV) were <12% for insulin. We then grouped all the subjects into two groups according to the insulin level.

The well-recognized clinical features of the illness were collected for analysis, with the continuous variables adopted for this study including: current age, age at onset, years of illness, daily dosage of antipsychotics (expressed as a chlorpromazine equivalent) and mood stabilizers. The categorical variables included gender, smoking habits, coexisting psychotic features, use of atypical antipsychotics or mood stabilizers, and a body mass index (BMI) ≥ 24 kg/m2. We used the Taiwanese standard (BMI ≥ 24 kg/m2) for overweight proposed by the Department of Health, Executive Yuan in 2002,13 which was modified from the Asia–Pacific standard as a cut-off point.14

Statistical analyses

The data were analyzed using SPSS 12.0 (SPSS, Chicago, IL, USA). We used Levin's test for evaluating variable distribution normality. The two-group comparison (with/without hyperinsulinemia) was conducted using the Student's t-tests and Fisher's exact test. The Mann–Whitney U test was performed for non-normal variables in each group. A two-tailed P-value of <0.05 was considered to be significant for all of the tests in this study. Furthermore, a multivariate analysis was performed on the grounds of statistical significance (P < 0.05) in independent variables shown in the univariate analysis.

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Patients with hyperinsulinemia had significantly higher mean value BMI than those without hyperinsulinemia (Table 1). There was no significant difference between the other variables except for insulin and BMI.

Table 1.  Comparison between the continuous variables of euthymic bipolar patients with and without hyperinsulinemia
Continuous variablesPatients with hyperinsulinemia (n = 15)Patients without hyperinsulinemia (n = 41)tP-value
Mean (SD)Mean (SD)
  • *

    P < 0.05.

  • BMI, body mass index; CPZ, chlorpromazine; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

Age, years37.8 (9.3)32.6 (8.3)−2.0070.050
Age at onset, years23.0 (6.7)23.4 (7.4)0.1940.847
Lithium dosage, mg/day620.0 (551.9)452.7 (511.9)−1.0490.299
Valproate dosage, mg/day400.0 (603.6)278.9 (469.6)−0.7790.440
Dosage of typical antipsychotics, CPZ equivalent/day92.9 (198.1)5.5 (22.4)−1.7050.110
Dosage of atypical antipsychotics, CPZ equivalent/day31.1 (82.9)35.1 (89.7)0.1520.880
Dosage of total antipsychotics, CPZ equivalent/day123.9 (199.9)38.6 (95.7)−1.5880.131
BMI kg/m227.38 (4.12)24.03 (3.31)−2.9660.005*
Fasting glucose, mmol/L6.73 (2.39)4.99 (0.31)−1.7830.134
Triglycerides, mmol/L1.48 (0.57)1.39 (1.33)−0.1990.844
Total cholesterol, mmol/L4.67 (0.92)4.35 (0.96)−0.9580.344
HDL cholesterol, mmol/L1.32 (0.25)1.49 (0.51)1.4990.144
LDL cholesterol, mmol/L2.93 (0.89)2.53 (0.93)−1.1850.243
Insulin, pmol/L254.89 (188.21)22.22 (13.89)−4.771<0.001*

The categorical variables of sampled patients with and without hyperinsulinemia and by clinical characteristics are presented in Table 2. Patients with hyperinsulinemia were more likely to be overweight. Furthermore, a relatively higher risk of 8.57 times for hyperinsulinemia was demonstrated in overweight patients (BMI ≥ 24: odds ratio, 8.57; 95% confidence interval, 1.65–44.43).

Table 2.  Comparison between categorical variables of euthymic bipolar patients with and without hyperinsulinemia
Categorical variablesPatients with hyperinsulinemia (n = 15)Patients without hyperinsulinemia (n = 41)P-value
n%n%
  • *

    P < 0.05 by Fisher's exact test.

Male533.32253.70.566
Without coexisting psychotic features533.31946.30.058
Smoking426.7512.20.264
Body mass index ≥241280.01434.10.009*
Use of atypical antipsychotics213.3922.00.438
Use of lithium853.31741.50.547
Use of valproate426.71331.70.754
Use of propranolol320.0717.11.000

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

To the best of our knowledge, this is the first study to examine insulin levels among medicated bipolar patients in full remission. We found that hyperinsulinemia was present in about a quarter of medicated euthymic bipolar patients, and bodyweight may be more tightly associated with hyperinsulinemia than other clinical characteristics among those patients, irrespective of the type or dosage of mood stabilizers or antipsychotics. Patients with BMI over 24 had 8.57 times higher risk for hyperinsulinemia than those with BMI under 24. However, with regard to fasting sugar and lipid profiles, there was no difference between the hyperinsulinemia and non-hyperinsulinemia groups. These findings suggest that change of bodyweight and insulin levels prior to other metabolic factors is consistent with the notion that obesity and insulin resistance, which are reflected by hyperinsulinemia, are the preclinical state for metabolic syndrome.6

Hyperinsulinemia is defined by the top quartile cut-off for fasting insulin in the general population.12 In this study, we found that about a quarter of patients had hyperinsulinemia, which is similar to the prevalence in the general population. Therefore, unlike our earlier report (in which 33.3% of cases of hyperinsulinemia had acute manic episodes and there was a subsequent remarkable increase in the proportion [71.4%] in early remission7), in the current study, 26.8% of the fully remitted bipolar patients had hyperinsulinemia, which is comparable to that of the general population. Taken together, there seemed to be a tendency among bipolar patients for insulin to gradually increase during the initial stage of a manic/mixed episode, peak soon after early remission, and finally return to a lower level in full remission. Hence, hyperinsulinemia may not be a persistent physiological state throughout the whole course of bipolar disorder, but only a transient physiological phenomenon for bipolar disorder. Although the exact role of insulin in the pathophysiology of bipolar disorder remains obscure, it merits further investigation because the different proportion of hyperinsulinemia was found in a different affective state.

Our findings fail to demonstrate the impact of medication on hyperinsulinemia in euthymic bipolar patients. However, bodyweight gains among bipolar patients have been thought to be associated with the illness itself or with medications.15

The major limitations in this study were the relatively restricted sample size, a lack of age-, gender- and BMI-matched healthy controls, and the cross-sectional study design. Repeated measurements may demonstrate the alteration of insulin levels in the long run. Also, obesity, analyzed in terms of BMI, instead of waist or waist-to-hip ratio, may restrict the generalization of the current study, because some previous studies have revealed that abdominal obesity and visceral fat are significantly associated with insulin resistance.16 Furthermore, some factors probably confounded the results of the study, such as diet and exercise, which were not taken into consideration for patients from outpatient clinics where these factors could not be easily controlled. However, the present study suggests that, compared with the general population, there is no increase of hyperinsulinemia among medicated euthymic bipolar patients, which supports the theory that hyperinsulinemia is only a transient physiological state for bipolar disorder. Moreover, in a euthymic state, overweight medicated bipolar patients are prone to develop hyperinsulinemia, irrespective of the mood stabilizers or antipsychotics used. Because hyperinsulinemia is a preclinical state for metabolic syndrome, a weight-reduction or weight-gain-prevention strategy should be emphasized among euthymic bipolar patients in order to reduce the risk of insulin resistance and ensuing metabolic disarrangement.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

The authors wish to thank Miss Pin-Fe Lai and Miss Nai-Ying Liu for their assistance in data collection.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES