Eur. J. Psychiat. Vol. 30, N.° 3, (163-171) 2016 Keywords: Mania; Inflammation; Haemodilution. Haematological and biochemical differences between mania and euthymia Köse Çinar Rugüla,* Görgülü Yasemina Sönmez Mehmet Bülenta Kahyaci Kiliç Evnurb a Assistant Professor, MD, Department of Psychiatry, Trakya University Faculty of Medicine, Edirne b MD, Department of Psychiatry, Trakya University Faculty of Medicine, Edirne TURKEY ABSTRACT – Background and Objectives: The effects of transient hypothalamic dys- function on the hypothalamic-pituitary-adrenal and hypothalamic-pituitary-thyroid axes can be shown by haematological and biochemical parameter changes. We hypothesized that manic episodes will be associated with subclinical inflammation, haemodilution and altered thyroid functions compared to euthymic states. Methods: Patients admitted to the psychiatry clinic with manic episodes were identified. Those having comorbidities, except for thyroid dysfunctions, hypertension, hyperlipidemia, and type 2 diabetes mellitus, were excluded. Complete blood counts, total protein, albumin, and thyroid tests were recorded during the admissions (manic episodes) (Maletic, 2014 #24) and one year later (euthymic states) for the same patients. Results: All studied parameters had significant differences between mania and euthymia. During manic episodes, patients had higher peripheral inflammatory indices (neutrophil/lym- phocyte ratio and platelet/lymphocyte ratio), haemodilution (lower haemoglobin, haemat- ocrit, total protein, and albumin), higher thyroxine and lower thyroid-stimulating hormone levels compared to euthymic states. Conclusions: This study supports the hypothesis that compared to euthymic states; manic episodes are associated with low-grade inflammation, haemodilution and thyroid function abnormalities. Monitoring patients’ blood compositions could result in better prognostic evaluations and aid in determining additional systemic treatment options, as well as in generating causal hypothesis to be tested in future studies. Received: 19 January 2016 Revised: 7 April 2016 Accepted: 14 April 2016 164 KÖSE ÇINAR RUGÜL ET AL. Introduction haemodilution in mania and haemoconcen- tration during depressive episodes9. Bipolar disorder (BD) can be counted The hypothalamus produces thyrotropin- among the diseases in which chronic cellular releasing hormone, which stimulates the re- stress plays a role in the aetiology1,2. In- lease of thyroid-stimulating hormone (TSH) flammation contributes to the initiation and from the anterior pituitary gland15. Stress has progression of several diseases that are ag- effects on the hypothalamic-pituitary-thyroid gravated by stress (e.g. dementia, cardiovas- (HPT) axis15,16. HPT axis abnormalities are cular and endocrinologic diseases, and some common in BD17. Thyroid patients present types of cancer)3. These diseases have high with higher rates of BD and hyperthyroidism prevalences and earlier ages of and earlier may increase the risk of developing BD18. ages of onset in patients with BD 4. Stress im- From a neurobiological perspective, patho- pairs the normal regulation of leukocyte func- logical conditions showing similarities and tions by the hypothalamic-pituitary-adrenal differences produce a disease state that we (HPA) axis and causes persistent low-grade call BD19. Current research emphasis is on inflammation2,5. The peripheral pathophysiol- the biological is on the biological markers ogy of BD appears to be related to systemic in- that can be used in the diagnosis and can of- flammatory mechanisms3. Moreover, inflam- fer therapeutic options. Candidate peripheral mation has been proposed to be a causative biomarkers, like inflammatory and oxidative factor for BD progression6. Persistent low- stress markers, neurotrophins and neuro- grade inflammation is more intense during transmitters, seem to change independently mood episodes, especially manic episodes, and from the mood state4,6,19,20. We wanted to less intense in depressive episodes. Even eu- compare mania and euthymia in the same thymia has been associated with detectable pe- patients to see if the blood analyses differ be- ripheral inflammatory activity4. Cytokines, tween the two states. which are produced by immune cells and play important roles in cell signalling, lead to changes in lymphocyte, neutrophil and platelet levels7. There are only a few psychiatric disor- der studies that have examined the neu- Methods trophil/lymphocyte ratio (NLR) and the platelet/lymphocyte ratio (PLR), both of which Subjects are considered inflammatory indices8-12. We reviewed electronic medical records of Between BD episodes, haematological and patients with BD diagnosed between January biochemical parameter differences were de- 2004 and December 2014. We identified and tected, even though the measures remained in confirmed BD type I diagnoses via our inpa- the normal range9. This imbalance was ex- tient and outpatient clinical records. All diag- plained by changes in fluid intake, physical noses were made according to the Diagnostic activity levels or psychotropic drug use. Fluid and Statistical Manual of Mental Disorders, and electrolyte redistribution within body Fourth Edition, Text Revision (DSM-IV-TR). compartments between different episodes Manic episode admissions with Young Mania was another explanation9,13,14. The latter ex- Rating Scale scores greater than 20 were eval- planation has been shown to be more realis- uated. Patients having data from both the tic with a recent study showing a relative manic episode and the euthymic state (ap- INFLAMMATION AND HEMODILUTION IN MANIA 165 proximately 1 year later) were selected. Pa- Statistical Analysis tients having disease episodes within this 1- year period were excluded. From those pa- We used SPSS version 20 (SPSS, Inc., tients with repeated admissions, the final Chicago, IL, USA) for statistical analysis. De- admission was included in our data. scriptive statistics are expressed as mean ± standard deviation (SD) and rate (%). The We excluded patients under 18 years of Kolmogorov-Smirnov test was used to evalu- age. Patients having medical comorbidities, ate variable distributions. Haematological and except for thyroid dysfunctions, hyperten- biochemical parameters were compared be- sion, hyperlipidemia and type 2 diabetes mel- tween manic episodes and euthymic states in litus, were also excluded. We further ex- the same subjects. Student’s t-tests (paired- cluded patients having comorbid organic samples t-tests) were performed on continuous mental disorders or substance use disorders. variables with normal distributions, and This study was approved by the local ethics Wilcoxon signed-ranks tests were used for committee (TÜTF-BAEK 2015/61). continuous variables without normal distribu- tions. Correlations were assessed by Pearson’s or Spearman’s correlation coefficients. The Measures covariates included sex, age in years, cigarette Patients’ psychiatric evaluations and med- smoking, psychotic symptoms during the ical histories were recorded during their ad- manic episode admission, mood stabilizers and missions and in their outpatient clinical ex- antipsychotic treatments. A p value of < 0.05 aminations. Blood samples were collected was considered statistically significant. between 8 to 10 a.m. and analyses were per- formed by the medical faculty’s biochemical laboratories. NLR, PLR, haemoglobin (g/dL), haematocrit (%), total protein (g/dL), albu- Results min (mg/dL), thyroxine (free T4; ng/dL), and TSH (µIU/mL) levels were recorded. While In this retrospective study, we covered the NLR was calculated as the ratio between the absolute neutrophil and the absolute lym- data between 2004 and 2014. Data was gath- phocyte counts, PLR was calculated as the ra- ered from an electronic database and our out- tio between the absolute platelet and the ab- patient and inpatient clinical records. A total solute lymphocyte counts10,21. The total plas ma of 133 patients met the inclusion criteria. protein level was used as a direct measure and Their haematological and biochemical para- haemoglobin, haematocrit and albumin levels meters from their most recent manic episodes were used as indirect measures of haemodi- and from their euthymic states (approxi- lution/haemoconcentration9. mately 1 year after the manic episode) were analysed. The characteristics of the study We retrieved demographic and clinical data population are presented in Table 1. including sex, age in years at mania and eu- thymia, psychotic symptoms during the manic While all measurements were within the episode admission, and cigarette smoking. normal ranges, there were significant differ- We also recorded the use of mood stabilizers ences between mania and euthymia. During and antipsychotic treatments at admission for mania, NLR and PLR were higher than eu- the manic episode and during the euthymic thymia. However, mean haemoglobin, hae - state when blood samples were collected. matocrit, total protein and albumin levels 166 KÖSE ÇINAR RUGÜL ET AL. Table 1 Demographic and clinical characteristics of the study population (n = 133) Characteristics n (%) or mean ± SD Gender, female, n (%) 67 (50.4) Age at mania, mean ± SD 38.3 ± 11.2 Age at euthymia, mean ± SD 39.5 ± 11.7 Psychotic symptom presence (during mania), n (%) 87 (65.4) Treatment during mania, n (%) Lithium 18 (13.5) Valproate 21 (15.7) Atypical antipsychotics 37 (27.8) Typical antipsychotics 4 (3) Treatment during euthymia, n (%) Lithium 65 (48.9) Valproate 80 (60.2) Atypical antipsychotics