Abstract
This study assesses homeostatic model assessment of insulin resistance and stress hormone levels among drug-naive patients with first-episode schizophrenia.
Before the introduction of antipsychotics, links between schizophrenia and abnormal glucose metabolism levels were found in the late 1800s as an increased prevalence of diabetes in families with a history of “insanity.”1 Furthermore, it is known that some patients with psychosis require higher dosages than other patients when applying insulin therapy, suggestive of insulin resistance.2 A recent meta-analysis by Pillinger et al1 assessed insulin resistance and found an elevated homeostatic model assessment of insulin resistance (HOMA-IR) among drug-naive patients with first-episode schizophrenia (n = 560) compared with controls (n = 450).1 They highlighted hormonal stress axis activation and lifestyle factors as potential confounders.1 Stress hormones, such as cortisol and catecholamines, are catabolic and functional antagonists of insulin. Antipsychotic-naive individuals with first-episode psychosis exhibit higher baseline cortisol levels and a blunted cortisol awakening response compared with controls.1,3 To test whether insulin resistance in schizophrenia can be discerned from stress-related and medication effects, we assessed HOMA-IR and stress hormone levels among drug-naive patients with first-episode schizophrenia and matched controls.
Methods
Morning urine and blood samples (taken at 8 am after the patients fasted) were obtained from 24 acute, drug-naive patients with first-episode schizophrenia and 24 controls. All participants were white and matched for sex, age, body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared), and waist-hip ratio (Table 1). All procedures were approved by the local institutional review board and written informed consent was obtained. Methodological details, inclusion/exclusion criteria, a psychopathology assessment (Positive and Negative Syndrome Scale [PANSS]), and the measurement of serum cortisol and urinary metanephrine/normetanephrine levels are available in Jordan et al.4 Insulin and glucose levels were determined as described in Steiner et al.5 Homeostatic model assessment of insulin resistance, a measure of basal insulin resistance, was calculated using the formula (insulin [mU/L] × glucose [mmol/L])/22.5. Group differences were detected applying Mann-Whitney U tests, and associations of HOMA-IR with other measures were determined using the Spearman rank test. An analysis of covariance using an aligned rank transformation of the data (http://www.r-project.org) was performed to control for the diagnostic group effects regarding HOMA-IR for the covariates BMI, waist-hip ratio, cortisol levels, metanephrine levels, normetanephrine levels, and smoking.
Table 1. Demographic Data, Clinical Scores, and Laboratory Parameters Related to Insulin Resistance and Hormonal Stress Axis Activation (Cortisol and Catecholamine Metabolites)a.
| Median (Quartile 1-Quartile 3) | Mann-Whitney U Test | P Value | ||
|---|---|---|---|---|
| Patients With Schizophrenia (n = 24) |
Controls (n = 24) |
|||
| Sex, No. male/female | 13/11 | 14/10 | >.99b | |
| Age, y | 31.0 (25.0-37.5) | 31.5 (25.3-37.5) | 286.0 | .98 |
| PANSS P scorec | 17.0 (12.5-19.2) | NA | ||
| PANSS N scorec | 9.5 (4.5-19.8) | NA | ||
| PANSS G scorec | 24.5 (22.8-34.2) | NA | ||
| PANSS sum scorec | 55.0 (44.0-68.2) | NA | ||
| BMI, kg/m2 | 22.2 (21.0-27.4) | 24.7 (21.6-27.6) | 239.0 | .32 |
| Waist-hip ratio, cm/cm | 0.87 (0.82-0.91) | 0.85 (0.81-0.91) | 295.5 | .88 |
| HOMA-IR, mU*mmol/L | 0.72 (0.38-2.28) | 0.36 (0.26-0.84) | 382.0 | .02d |
| Insulin, μIU/mL (to convert to picomoles per liter, multiply by 6.945) | 3.07 (1.62-7.43) | 1.79 (1.27-3.22) | 374.0 | .04d |
| Glucose, mg/dL (to convert to millimoles per liter, multiply by 0.0555) | 89 (86-108) | 83 (79-90) | 409.0 | .01d |
| Cortisol, µg/dL (to convert to nanomoles per liter, multiply by 27.588) | 26.65 (17.78-40.30) | 18.60 (15.30-24.50) | 374.0 | .04d |
| Metanephrine, µg/g creatinine | 71.5 (42.9-127.1) | 42.9 (31.7-75.0) | 395.5 | .03d |
| Normetanephrine, µg/g creatinine | 147.7 (89.1-217.7) | 98.7 (50.6-158.0) | 373.0 | .08 |
| Smoking, cigarettes/d | 10.0 (0.25-17.00) | 0 (0-3) | 444.0 | .001d |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); G, general psychopathology scale; HOMA-IR, Homeostasis model assessment of insulin resistance; N, negative scale; P, positive scale; NA, not applicable; PANSS, Positive and Negative Syndrome scale.
Statistical tests were 2-tailed, P < .05 was considered significant, and a P value of less than .10 was considered a statistical trend.
Fisher exact test.
Corrected scores (subtraction of minimum scores representing “no symptoms” from the PANSS scores).
Significant P values.
Results
Homeostatic model assessment of insulin resistance increased 2-fold among patients (median [IQR], 0.72 [0.38-2.28]) compared with controls (median [IQR], 0.36 [0.26-0.84]) (P = .02) (Table 1). Serum cortisol levels increased approximately 1.4-fold (median [IQR], 26.65 [17.78-40.30]; P = .04) and urinary metanephrine (median [IQR], 71.5 [42.9-127.1]; P = .03) and normetanephrine (median [IQR], 147.7 [89.1-217.7]; P = .08) levels increased approximately 1.7-fold and 1.5-fold, respectively, although only the latter trended toward significance (P < .10). Homeostatic model assessment of insulin resistance was correlated with BMI (r = 0.463; P = .001), but no significant correlations were found between HOMA-IR and levels of cortisol (r = 0.097; P = .52), metanephrine (r = 0.123; P = .41), normetanephrine (r = 0.161; P = .28) or smoking (r = 0.094; P = .53) (Table 2). Homeostatic model assessment of insulin resistance was not related to PANSS-P (r = 0.086; P = .69), PANSS-N (r = 0.099; P = .65), PANSS-G (r = −0.183; P = .39), or PANSS sum (r = 0; P = .99). Additional testing showed that the schizophrenia-related HOMA-IR increase was present among nonsmokers (patients: n = 8, controls: n = 18; P = .02). The diagnostic group difference regarding HOMA-IR was confirmed by an analysis of covariance using an aligned rank transformation (F = 10.172, P = .003). Only the covariate BMI had significant influence in this model (patients: median [IQR], 22.2 [21.0-27.4]; controls: median [IQR], 24.7 [21.6-27.6]; F = 9.001, P = .005).
Table 2. Correlation of Homeostasis Model Assessment of Insulin Resistance With Demographic Data, Clinical Scores, and Laboratory Parameters.
| HOMA-IR [mU*mmol/L] | ||
|---|---|---|
| r | P Value | |
| Age, y | 0.059 | .69 |
| PANSS P score | 0.086 | .69 |
| PANSS N score | 0.099 | .65 |
| PANSS G score | −0.183 | .39 |
| PANSS sum score | 0.000 | .99 |
| BMI, kg/m2 | 0.463 | .001a |
| Waist to hip ratio, cm/cm | 0.173 | .25 |
| Cortisol, µg/dL (to convert to nanomoles per liter, multiply by 27.588) | 0.097 | .52 |
| Metanephrine, µg/g creatinine | 0.123 | .41 |
| Normetanephrine, µg/g creatinine | 0.161 | .28 |
| Smoking, cigarettes/d | 0.094 | .53 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); G, general psychopathology scale; HOMA-IR, Homeostasis model assessment of insulin resistance; N, negative scale; P, positive scale; PANSS, Positive and Negative Syndrome Scale.
Significant P value.
Discussion
We found that although serum and urinary stress hormone levels were increased in acute psychosis; this was not related to HOMA-IR. Notably, glucose metabolism is regulated redundantly and may be impaired by stress or biological insults in many ways (eg, central sympathetic control through multiple neurotransmitters). The patient and control groups were well-matched for sex, age, BMI, and waist-hip ratio to rule out the possibility that being overweight and visceral adiposity could be factors that explained increased HOMA-IR scores among patients. While the lifestyle factors “exercise,” “sedentary behavior,” and “diet” were not examined in this study, HOMA-IR was not related to smoking.
These results support the recently raised notion of impaired insulin/glucose homeostasis in drug-naive patients with first-episode schizophrenia, small sample and effect sizes notwithstanding.1,6 Reduced insulin sensitivity may arise independently from pharmacotherapy, hormonal stress axis activation, or obesity.6 Glucose metabolism levels should be assessed in larger samples and other psychiatric disorders to examine specificity for schizophrenia.
References
- 1.Pillinger T, Beck K, Gobjila C, Donocik JG, Jauhar S, Howes OD. Impaired glucose homeostasis in first-episode schizophrenia: a systematic review and meta-analysis. JAMA Psychiatry. 2017;74(3):261-269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sakel M. Insulin therapy in the future of psychiatry. Can Med Assoc J. 1938;39(2):178-179. [PMC free article] [PubMed] [Google Scholar]
- 3.Berger M, Kraeuter AK, Romanik D, Malouf P, Amminger GP, Sarnyai Z. Cortisol awakening response in patients with psychosis: Systematic review and meta-analysis. Neurosci Biobehav Rev. 2016;68:157-166. [DOI] [PubMed] [Google Scholar]
- 4.Jordan W, Dobrowolny H, Bahn S, et al. . Oxidative stress in drug-naive first episode patients with schizophrenia and major depression: effects of disease acuity and potential confounders [published online December 2, 2016]. Eur Arch Psychiatry Clin Neurosci. doi: 10.1007/s00406-016-0749-7 [DOI] [PubMed] [Google Scholar]
- 5.Steiner J, Bernstein HG, Schiltz K, et al. . Decrease of serum S100B during an oral glucose tolerance test correlates inversely with the insulin response. Psychoneuroendocrinology. 2014;39:33-38. [DOI] [PubMed] [Google Scholar]
- 6.Kirkpatrick B, Miller BJ, Garcia-Rizo C, Fernandez-Egea E, Bernardo M. Is abnormal glucose tolerance in antipsychotic-naive patients with nonaffective psychosis confounded by poor health habits? Schizophr Bull. 2012;38(2):280-284. [DOI] [PMC free article] [PubMed] [Google Scholar]