Arterial Stiffness in Patients with Bipolar Disorder

Abstract

Objective

Bipolar disorder (BD) is an inflammatory and metabolic disease. The disease and the drugs used to treat it may affect cardiovascular disease (CVD) risk. The aim of this study is to investigate arterial stiffness in patients with BD and compare them with healthy controls.

Methods

Thirty-nine patients with BD type I in remission and 39 healthy control subjects were included in the study. Carotid and femoral artery intima-media thickness (IMT) and arterial thickness parameters were measured by Doppler ultrasonography.

Results

The elastic modulus value of the carotid artery was significantly higher in the patients than in the control group (p = 0.015). Although the IMT of both carotid and femoral artery was thicker in patients than in healthy control subjects, this difference was not statistically significant (p = 0.105; p = 0.391). There was a significant positive correlation between chlorpromazine equivalent dose and femoral elastic modulus value (p = 0.021, r = 0.539). There was a positive correlation between lithium equivalent dose and carotid compliance; a significant negative correlation between lithium equivalent dose and carotid elastic modulus was also determined (both p = 0.007, r = 0.466; p = 0.027, r = −0.391, respect-ively). No predictor was observed between drug dose and arterial stiffness parameters.

Conclusion

Arterial stiffness might be investigated for its potential to reduce CVD risk in patients with BD. Given the established CVD complications in this patient population, further studies are needed to determine whether the results are specific to antipsychotic treatment or BD and to clarify the potential arterial protective effects of mood stabilizers.

INTRODUCTION

Patients diagnosed with bipolar disorder (BD) may have a higher incidence of metabolic disease than the general population because of an exacerbation of an unhealthy lifestyle due to psychiatric symptoms and side effects of psychopharmacologic treatments. Cardiovascular disease (CVD) begins 17 years earlier in these patients [1]. Both type I and type II BD have been found to be associated with premature death, with CVD being the most common cause of death. Risk factors such as obesity, hypertension, diabetes mellitus, and hyperlipidemia, which promote atherosclerosis and occur at high rates, are among the underlying causes of the increased risk of CVD complications in BD [2].

Arterial stiffness is one of the major risk factors for CVD and a strong predictor of CVD morbidity and mortality, especially in hypertension and diabetes mellitus [3]. Measurement of arterial outer wall thickness by high-resolution ultrasonography (USG) is a useful clinical index for early detection of arterial stiffness and a method for noninvasive determination of elastic properties of arterial wall structure [4]. Endothelial dysfunction or intima- media thickness (IMT) is an early sign of atherosclerosis. Arterial stiffness is known as a decrease in the ability of the artery to expand. It is assessed along with distensibility and compliance when evaluating central vascular func-tion. The change in vessel volume with a change in pressure, that is, the distensibility (adaptability) capacity of the vessel, is known as compliance. Distensibility is the change in volume that occurs when the artery is in compliance. Modulus of elasticity is a measure of the deformation of the vessel under pressure. Increased arterial stiffness has negative effects on cardiac output and organ perfusion by impairing the buffering capacity of central arteries, particularly the carotid and femoral arteries [4]. Chronic inflam-mation destroys common targets such as elastin, collagen, and smooth muscle cells in both elastic and muscular arteries, and there is a relationship between the severity of inflammation and arterial stiffness. This information has been studied in chronic inflammatory diseases (such as irritable bowel syndrome, rheumatoid arthritis, ulcerative colitis, and chronic kidney disease) [5-7].

Arterial stiffness is not a relatively unnoticed issue in the field of psychiatry. Schizophrenia, BD, and anorexia nervosa are some of the diseases in which arterial stiffness is investigated by measuring pulse wave velocity (PWV) and using USG [8-10]. In another study, higher arterial stiffness was associated with poorer cognitive function [11]. It is known that increased proinflammatory cytokines and decreased antioxidant regulation in BD lead to neurodegeneration in many key brain regions responsible for mood and cognition [12]. Similarly, the endocrine or immune-inflammatory abnormalities underlying BD and the negative effects of antipsychotic drugs on the CVD system may contribute to the increased CVD risk that we have previously mentioned in BD [13]. In addition, antipsychotic drugs, including second-generation antipsychotics, have been strongly associated with metabolic diseases [14].

BD is a disease associated with inflammation and metabolic diseases. As with psychiatric diseases, there is insufficient information on arterial stiffness in BD. There is a need for new markers and methods that can be quickly and easily interpreted when evaluating BD patients at risk for CVD. To this end, we aim to assess arterial stiffness in patients with BD independently of the metabolic diseases and conditions we commonly see in these patients using Doppler USG and compare it to healthy control subjects. In this way, we can contribute to predicting CVD risks in these patients.

METHODS

Subjects

Thirty-nine patients diagnosed with BD type I and in remission according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) and 39 healthy control subjects without mental disorder according to DSM- 5 who presented to the Mental Health and Diseases Clinic of Elazığ Fethi Sekin City Hospital were included in the study.

The inclusion criteria for the patients were: Age between 18 and 50 years, BD remission according to DSM- 5, no known metabolic disease, no physical pathology or neurological disease that could affect the distribution of psychiatric symptoms (such as cerebral palsy, epilepsy, cerebrovascular disease), no history of head trauma and cognitive dysfunction, no mental retardation, signing of informed consent, no change in antipsychotic and mood treatment regimens in the past six months, no use of more than one antipsychotic and depot medication (in long- acting intramuscular form), and no additional drug use besides current psychiatric treatment.

The inclusion criteria for the control group were: Age between 18 and 50 years, no history of psychiatric, metabolic, or neurologic diseases, no drug use, signing of informed consent, and no mental retardation. Healthy subjects (whether they needed it or not) had had no psychiatric admissions or treatment histories during their lifetime and had no psychiatric illness at the time of the interview.

In both groups, participants with known alcohol and substance use disorder, hypertension, metabolic syndrome, hyperlipidemia, diabetes, renal insufficiency, history of CVD, concomitant peripheral arterial disease, and drug use for these conditions were not included in the study. Because arterial stiffness may physiologically increase with age [15], participants older than 50 years were not included in the study.

Procedure

The study was conducted between December 2021 and April 2022 at the Mental Health and Diseases Clinic of Elazığ Fethi Sekin City Hospital. After obtaining written informed consent from all participants, sociodemographic questionnaires were completed, routine whole blood and biochemical tests were requested, and waist circumference and systolic and diastolic arterial blood pressure (systolic blood pressure [SBP], diastolic blood pressure [DBP]) were measured. Subsequently, a radiologist measured arterial thickness with a Doppler USG device at the Radiology Outpatient Clinic of Elazığ Fethi Sekin City Hos-pital. The study was approved by the Ethics Committee of Fırat University for Noninterventional Research on Novem-ber 4, 2021, under the number 2021/11-39, and local institutional permissions were obtained.

A complete blood count and general biochemical tests (fasting blood glucose, creatinine, sodium, potassium, chloride, magnesium, calcium, triglycerides, high-density lipoproteins, low-density lipoproteins, total cholesterol, aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase) were requested from the participants, and the pathological values found (fasting blood glucose > 100, triglycerides > 150 mg/dl, high-density lipoprotein < 40 mg/dl in males and < 50 mg/dl in females, low-density lipoprotein > 130 mg/dl, total cholesterol > 220 mg/dl, creatinine > 1.40 mg/dl in males and > 1.30 mg/dl in females) were excluded from the study [16]. The SBP and DBP of all participants were measured with a manual sphygmomanometer in the sitting position, and those who had an SBP > 130 mmHg and a DBP > 80 mmHg in at least three measurements 10 minutes apart were excluded from the study [16]. At the same time, the waist circumference of the participants was measured manually with a tape measure; those with a waist circumference > 102 cm in males and > 88 cm in females and those with a body mass index (BMI) > 30 kg/m² were excluded from the study [16].

Antipsychotic doses taken by patients were converted to chlorpromazine equivalent doses using the equivalent dose method for standardization [17]. Doses of mood stabilizers taken by patients were converted to lithium equivalent doses using the equivalent dose method for standardization [18].

Ultrasonography

A high-resolution Doppler USG device (Philips Affiniti 50 G, L 12-5 Mh linear probe) was used in all patients. Measurements were performed on the right femoral artery in the supine position and on the right carotid artery with the neck hyperextended. IMT of the carotid and femoral arteries was measured 1 cm distal to the bifurcation of the common carotid artery (Fig. 1). Systolic and diastolic arterial diameters were measured using USG modes B and M together (Fig. 2).

Fig. 1.

Intima–media thickness measurement from 1 cm distal to the bifurcation of the common carotid artery.

Fig. 2.

Systolic and diastolic diameter measurement with M mode.

Vessel arterial thickness parameters were calculated using the following formulas described in the literature [4].

Cross-sectional compliance = {π (SD² − DD²)} / (4 ΔP)

Cross-sectional distensibility = (SD² − DD²) / (DD² ΔP)

Diastolic wall stress = {DD / (2 × IMT)} × {(SP + SD) / 2}

Elastic modulus = [3 / {1 + (cross-sectional area of lumen / cross-sectional area of wall)}] / cross-sectional distensibility

Statistical Analysis

Analyzes were performed using the SPSS 22 program (Statistical Package for Social Sciences [SPSS] Inc.). In the study, descriptive data are shown as n and % values for categorical data and mean ± standard deviation for continuous data. Chi-square analysis (Pearson chi-square) was used to compare categorical variables between groups. Conformity of continuous variables to the normal distribution was assessed with the Kolmogorov–Smirnov test. When comparing paired groups, Student’s ttest was applied for the normally distributed variables and Mann–Whitney Utest was applied for the non-normally distributed variables. The Pearson correlation test was used to examine the relationship between continuous variables. Linear regression analysis was performed to determine the predictor for the dependent variables. The Enter method was used in the construction of the model, and those variables that showed a significant relationship in the correlation test were included in the model. The statistical significance level in the analyses was taken as p < 0.05.

RESULTS

A total of 78 participants, 39 patients and 39 healthy control subjects, were included in the study. The mean age of the patient group was 37.8 ± 10.4 years and that of the control group was 35.1 ± 7.3 years, and there was no significant age difference between the groups (p = 0.185). 46.2% of the patient group were female and 53.8% were male; 53.8% of the control group were female and 46.2% were male, and there was no significant difference between groups in terms of sex (p = 0.497) (Table 1).

Table 1.

Comparison of all characteristics of the patient and control groups

Sociodemographic and clinic characteristics	Patient	Control	p valuea
Age (yr)	37.8 ± 10.4	35.1 ± 7.3	0.185b
Sex			0.497
Female	18 (46.2)	21 (53.8)
Male	21 (53.8)	18 (46.2)
Marital status			0.365
Single	21 (53.8)	17 (43.6)
Married	18 (46.2)	22 (56.4)
Educational status			0.090
Middle school and below	20 (51.3)	11 (28.2)
High school	11 (28.2)	19 (48.7)
University	8 (20.5)	9 (23.1)
Residential area			0.096
District	10 (25.6)	17 (43.6)
City	29 (74.4)	22 (56.4)
Economical situation			0.807
Low	12 (30.8)	12 (30.8)
Middle	22 (56.4)	20 (51.3)
High	5 (12.8)	7 (17.9)
Working status			0.173
Working	15 (38.5)	21 (53.8)
Not working	24 (61.5)	18 (46.2)
The use of cigarettes			0.820
Yes	18 (46.2)	17 (43.6)
No	21 (53.8)	22 (56.4)
The use of alcohol/substance			0.431
Yes	5 (12.8)	2 (5.1)
No	34 (87.2)	37 (94.9)
Family history of mental disorder			0.151
Yes	10 (25.6)	5 (12.8)
No	29 (74.4)	34 (87.2)
Self-mutilation (lifetime)			0.104
Yes	8 (20.5)	3 (7.7)
No	31 (79.5)	36 (92.3)
Suicide attempt (lifetime)			0.003
Yes	10 (25.6)	1 (2.6)
No	29 (74.4)	38 (97.4)
Suicide attempt method			0.364
Injury with a cutting tool	7 (70.0)	0 (0)
With drug	1 (10.0)	1 (100.0)
High jump	1 (10.0)	0 (0)
Other	1 (10.0)	0 (0)
Type of drug			-
Mood stabilizer (valproate, lithium, carbamazepine)	20 (51.3)	-
Antipsychotics	8 (20.5)	-
Multiple (antipsychotics + mood stabilizer)	11 (28.2)	-
Chlorpromazine equivalent dose (mg/day)	247.8 ± 120.6	-	-
Lithium equivalent dose (mg/day)	697.4 ± 173.5	-	-

Values are presented as mean ± standard deviation or number (%).

-, not available.

^aChi-square analysis, ^bStudent’s ttest was performed.

Body mass index (p = 0.008), SBP (p < 0.001), DBP (p = 0.007) and carotid elastic modulus (p = 0.015) were significantly higher in the patient group than in the control group (Table 2).

Table 2.

Comparison of BMI, blood pressure, and arterial pressure of the patient and control groups

Clinical parameters	Patient	Control	p value
BMI (kg/m2)	24.7 ± 2.6	23.1 ± 2.6	0.008a,*
SBP (mmHg)	123.6 ± 6.9	109.8 ± 12.1	<0.001b,*
DBP (mmHg)	74.1 ± 6.3	70.4 ± 6.1	0.007b,*
Carotid IMT (mm)	0.54 ± 0.13	0.50 ± 0.08	0.105a
Carotid compliance (mm2/mmHg)	0.15 ± 0.06	0.16 ± 0.08	0.465b
Carotid distendibility (%)	0.005 ± 0.002	0.006 ± 0.003	0.086b
Carotid diastolic wall stress	381.4 ± 68.6	375.4 ± 87.6	0.671a
Carotid elastic modulus (mmHg)	188.0 ± 94.0	148.0 ± 98.3	0.015b,*
Femoral IMT (mm)	0.47 ± 0.10	0.45 ± 0.08	0.391b
Femoral compliance (mm2/mmHg)	0.16± 0.07	0.17 ± 0.09	0.384a
Femoral distenbility (%)	0.004 ± 0.002	0.004 ± 0.002	0.648b
Femoral diastolic wall stress	487.4 ± 130.5	466.9 ± 107.1	0.450a
Femoral elastic modulus (mmHg)	194.6 ± 128.7	169.5 ± 94.6	0.398b

Values are presented as mean ± standard deviation.

BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; IMT, intima–media thickness.

^aStudent’s ttest, ^bMann–Whitney Utest was performed, *p < 0.05.

A positive and significant correlation was found between chlorpromazine equivalent dose and BMI, SBP, and femoral elastic modulus values. There was a positive correlation between lithium equivalent dose and carotid compliance; a significant negative correlation was determined between lithium equivalent dose and carotid elastic modulus (Table 3).

Table 3.

Correlations between Chlorpromazine and Lithium equivalent doses, diagnosis time, and other variables

Diagnosis time and other variables	Chlorpromazine equivalent dose (mg/day)			Lithium equivalent dose (mg/day)
	r	p value		r	p value
Age (yr)	0.116	0.647		−0.257	0.155
Psychiatric diagnosis time (yr)	0.021	0.934		−0.193	0.289
Duration of psychotropic drug use	−0.005	0.984		−0.137	0.456
BMI (kg/m2)	0.704	0.001		0.053	0.771
SBP (mmHg)	0.470*	0.049*		−0.239	0.188
DBP (mmHg)	0.063	0.805		0.013	0.942
Carotid IMT (mm)	0.342	0.165		−0.310	0.084
Carotid compliance	−0.248	0.321		0.466*	0.007*
Carotid distendibility	−0.233	0.352		0.333	0.063
Carotid diastolic wall stress	−0.330	0.182		0.165	0.368
Carotid elastic modulus	0.335	0.174		−0.391*	0.027*
Femoral IMT (mm)	−0.198	0.432		−0.109	0.551
Femoral compliance	−0.285	0.252		0.108	0.555
Femoral distendibility	−0.304	0.220		0.150	0.412
Femoral diastolic wall stress	−0.281	0.259		−0.023	0.899
Femoral elastic modulus	0.539*	0.021*		−0.195	0.286

BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; IMT, intima–media thickness.

*Correlated ones.

Multiple linear regression analysis showed that BMI, SBP, and femoral elastic modulus did not predict chlorpromazine equivalent dose. Carotid artery compliance and carotid elastic modulus values also did not predict lithium equivalent dose. And BMI did not predict lithium equivalent dose (Table 4).

Table 4.

Linear regression analysis of factors associated with drug equivalent doses

Clinical parameters	Linear regression analysis of factors
	β	SE	Standard β	t	p value
Chlorpromazine equivalent dose (mg/day)a
BMI (kg/m2)	20.611	11.560	0.448	−2.136	0.051
SBP (mmHg)	4.456	3.956	0.278	1.783	0.096
Femoral elastic modulus	0.147	0.201	0.149	1.126	0.279
Lithium equivalent dose (mg/day)b
Carotid compliance	431.992	600.512	0.153	0.719	0.478
Carotid elastic modulus	−0.676	0.401	−0.359	−1.684	0.103
BMI	1.356	10.797	0.021	0.126	0.901

BMI, body mass index; SBP, systolic blood pressure; SE, standard error.

^aR² = 0.470; F = 4.140; p = 0.027. ^bR² = 0.222; F = 4.143; p = 0.026.

DISCUSSION

Carotid elastic modulus, one of the parameters of arterial stiffness, was significantly higher in the patients than in the control group. Although both carotid IMT and femoral IMT were higher in patients, there was no statistically significant difference between them and the control group. A recent study examined the parameters of our study in patients with schizophrenia [8]. Based on the assumption that patients with schizophrenia have a high risk of dying from CVD, this study compared arterial stiffness in these patients with healthy controls and found significantly higher IMT in patients with schizophrenia but no significant difference in arterial stiffness compared with the control group. The researchers explained these results by the fact that antipsychotic treatment may have a protective effect on arterial stiffness [9]. This is possible both because of the drugs’ lowering of systolic arterial pressure and because of their possible anti-inflammatory effects. The anti-inflammatory properties of first- and second-generation antipsychotic drugs are supported by the results of studies [19,20]. A meta-analysis showed that antipsychotic treatment selectively reduced inflammation in psychosis [21]. On the other hand, it has been reported that the changes induced by antipsychotic treatment may lead to a further increase in proinflammatory cytokine levels in the long term, along with the metabolic risks that accompany them [22]. In other words, these drugs may also have a proinflammatory effect because of their metabolic side effects. Severe inflammation may increase arterial stiffness by causing pathological degenerative changes in the arterial wall and increasing CVD risk [5]. Apart from antipsychotics, about half of our patients were taking mood- stabilizer drugs. The effects of mood-stabilizer drugs on inflammation in BD are inconsistent [23]. Although there are some conflicting results, studies suggesting that mood- stabilizer drugs (especially lithium) reduce the production of proinflammatory mediators predominate [24]. Because the individual mechanisms of action of the various mood- stabilizer agents differ widely, one can also assume different effects on the cardiovascular system. Furthermore, the complex relationship between antipsychotic and mood- stabilizer treatments and inflammation limits the unam-biguous interpretation of our data on inflammatory burden based on the nature of BD.

A 2020 study assessed arterial stiffness using the noninvasive applanation tonometry method in patients with BD depressive episodes who were taking medication and found that the augmentation index (percentage of central pulse pressure), another indicator of arterial stiffness, was not significantly high. In interpreting their results, the researchers emphasized that the higher BMI, higher risk of hyperlipidemia, and higher smoking rates in BD patients complicate the reality and that the contribution of depression per se to inflammation in BD may be masked [25]. To open a parenthesis on the inflammatory burden in BD, previous biochemical and genetic studies have shown that impaired oxidative phosphorylation and mitochondrial dysfunction are essential features of BD. The chronic low course of phosphocreatine, a cellular reservoir for adenosine triphosphate (ATP) synthesis, independent of mood episodes, the impaired tricarboxylic acid cycle and the identification of some mitochondrial DNA polymorphisms are indicators [26-28]. This mitochondrial dysfunction observed in BD may be due to a disturbance in autophagy [29]. Supported by molecular and cellular data, the perspective that accepts mood disorders as neurodegenera-tive disorders draws attention to abnormal autophagy in BD. There is ample evidence that inflammation affects autophagy in the brain and that autophagy has a neuroprotective function [30]. In BD, it is observed that perturbations of intracellular signaling cascades, such as protein kinase B/mammalian target of rapamycin (AKT/mTOR), are involved in the pathophysiology of psychiatric disorders and autophagy [31]. This signaling pathway is critical for the regulation of cellular metabolism and growth and directly affects neuroplasticity. Impaired AKT and mTOR activity has been observed in BD, and genetic variants of AKT1 have been associated with BD. Decreased mTOR activity and downregulated autophagy have been associated with the development of depressive episodes in BD [28,31,32]. Studies linking autophagy and BD show that lithium increases autophagy and positively upregulates mitochondrial function [33,34].

In all patients taking second-generation antipsychotics (atypical antipsychotics), BMI, SBP, and femoral elastic modulus increased significantly when the equivalent dose of chlorpromazine was increased, but no predictive effect was found. When the equivalent dose of lithium increased in patients, carotid compliance increased significantly, and carotid elastic modulus decreased significantly, but no predictive effect was found. We know that an artery with a low elastic modulus value is less stiff and deforms more easily, whereas an artery with high compliance has a good compliance degree. In this regard, mood stabilizers seem to be associated with a protective function for the carotid arm. PWV was measured in BD and schizophrenia patients taking second-generation antipsychotics were compared with healthy controls, and it was found that blood pressure, an indicator of arterial stiffness, was significantly higher and PWV values were significantly faster [10]. The researchers attributed their findings to the use of antipsychotics and/or the disease itself. On the other hand, Fiedorowicz et al. [35] suggested that taking first-generation antipsychotics was associated with arterial stiffness, but taking second-generation antipsychotics was not. In another study, adults with schizophrenia taking atypical antipsychotics for an average of 2 years or adults with BD were found to have higher aortic PWV compared with control subjects. The researchers reported that several factors, such as BMI, influence CVD risk, but mentioned that antipsychotic therapy might contribute independently of CVD risk [36]. In our study, the high positive correlation between chlorpromazine equivalent dose and BMI but the lack of predictability of regression might suggest that there is a mediating role between these two factors. In a study of mentally retarded children (6−18 years), all of whom were taking atypical antipsychotics, children treated with atypical antipsychotics were found to have higher BMI, PWV, and SBP than children not taking atypical antipsychotics, as in our study. They also emphasized that high BMI is one of the metabolic complications in children treated with atypical antipsychotics. The same study reported that the use of atypical antipsychotics was not associated with IMT and arterial stiffness. Again, it associated the increase in PWV and SBP with early vascular structural changes [37]. Although we tried to minimize behavioral and metabolic factors (smoking, physical inactivity, hypertension, diabetes mellitus) in our study, CVD risk is subject to complicated influences. The parameters we examined are only part of a complex combination of in-flammatory, pathophysiological, and immunological mech-anisms that influence CVD risk in the BD population. As mentioned previously, the use of atypical antipsychotics and mood stabilizers may mediate some of these effects.

Systolic blood pressure and DBP and BMI levels [4], which are claimed to influence arterial stiffness, did not significantly predict arterial stiffness in our study, but BMI, SBP, and DBP levels in patients were significantly higher than those in healthy controls. Depending on the increase in arterial stiffness, SBP can be expected to increase, DBP can be expected to decrease, and elastic modulus can be expected to change in the direction of increase. Arterial stiffness may be both a consequence and a cause of increased blood pressure, so there is a bidirectional relationship between arterial stiffness and high SBP [4]. In addition, we should consider the use of antipsychotic and mood-stabilizer drugs (sometimes both) when evaluating this finding. Although the in vivo effects of lithium are controversial, it has been shown to slightly increase endothelium-dependent vasodilation [38], whereas valproic acid increases nitric oxide production in an in vitro model [39]. Antipsychotic drugs act directly on the arteries via blockade of adrenergic and cholinergic receptors and indirectly via central autonomic regulation. This mechanism is responsible for the majority of cardiovascular side effects of antipsychotics [40].

In a study investigating the relationship between vascular function and mood in BD, severe manic symptoms were associated with increased DBP [41]. In addition, brachial artery dilation was measured in patients with a greater longitudinal burden of manic symptoms, and endothelial function was found to be weaker [35]. However, there are no prospective studies that establish a temporal relationship between mood episodes and endothelial dysfunction. Our study did not examine the number of mood episodes in euthymic patients. Again, the persistence and duration of mood episodes have been associated with impaired vascular function in BD [35]. Our study also found no correlation between disease duration and arterial stiffness parameters in patients.

The strength of our study is that we tried to exclude premorbid risk factors affecting arterial stiffness as much as possible. Our study excluded factors that predispose to atherosclerosis, such as a history of CVD, hypertension, hypercholesterolemia, metabolic syndrome, and diabetes mellitus. Because arterial stiffness increases with age [15], care was taken to select patients as young as possible.

This study has several limitations. Although there was no abuse in the sample group, the use of cigarettes and alcohol was not reset. Although we believe that evaluation of these parameters would provide clearer results in drug-free BD patients, this does not seem possible. More valuable data may be obtained by increasing the sample size. On the other hand, although it is not clear which receptor is responsible for the positive vascular effects of estrogen, it has been observed that loss of endothelial ERα, an estrogen receptor, decreases arterial stiffness, whereas loss of another estrogen receptor, Erβ, reduces arterial dilatory capacity [42,43]. The inclusion of mixed genders in our sample is a limitation because the ER signal effective in arterial stiffness may differ with respect to gender. On the other hand, although BMI did not predict lithium and chlorpromazine equivalent dose, weight could affect drug doses.

In conclusion, arterial stiffness was higher in patients than in healthy controls because of the elastic modulus of the carotid artery. Although both carotid and femoral IMT were thicker in patients than in healthy control subjects, this difference was not statistically significant.

As the equivalent dose of lithium was increased in patients, carotid compliance significantly increased and carotid elastic modulus significantly decreased. With increasing equivalent dose of chlorpromazine in patients, BMI, SBP, and femoral elastic modulus increased signifi-cantly. Our study provides data for further studies to determine whether these effects are specific to antipsychotic treatment or to BD and to clarify the possible arterial protective effects of mood stabilizers. Arterial stiffness should be investigated for its potential to reduce CVD risk in BD.

The results of our study can only be generalized to our sample. Given the established CVD complications in this patient population, further studies with repeated measurements of arterial stiffness parameters are needed to identify and confirm atherosclerosis.

Funding Statement

Funding None.