Risk of venous thromboembolism during treatment with antipsychotic agents


Correspondence: Jiří Masopust, MD, PhD, Department of Psychiatry, Faculty of Medicine, Charles University and University Hospital, Sokolská 581, 500 05 Hradec Králové, Czech Republic. Email: masopustj@lfhk.cuni.cz


The evidence to date on the relation between the risk of venous thromboembolic disease (VTE) and antipsychotic agents derives primarily from observational and case history studies. While an increased risk of VTE has been associated with first-generation low-potency antipsychotic agents, particularly clozapine, there appears to be a growing number of reports on the occurrence of this adverse reaction during the use of second-generation antipsychotics, such as risperidone and olanzapine. The highest risk of pathological blood clotting emerges during the first 3 months after initiation of treatment with the product. Potential etiopathogenetic factors leading to VTE during treatment with antipsychotic agents include sedation, obesity, elevation of antiphospholipid antibodies, increased platelet activation and aggregation, hyperhomocysteinemia, and hyperprolactinemia. Diagnoses of schizophrenia and/or bipolar affective disorder, as well as hospitalization or stress with sympathetic activation and elevation of catecholamine levels, have been reported as known prothrombogenic factors. The present article contains the new version of the guideline for the prevention of VTE in psychiatric patients with limited mobility. Further prospective studies are necessary to elucidate the biological mechanisms of the relations between antipsychotic agents and VTE.

PATIENTS WITH SEVERE mental disorders (SMD), such as schizophrenia, bipolar affective disorder, schizoaffective disorder, and severe depressive disorder, have a two- to threefold higher mortality rate than the general population. In 60% of cases, the cause of death is somatic defects, especially cardiovascular disorders. Life expectancy among these patients is 13–30 years shorter. Apart from genetic predispositions, the high occurrence of cardiovascular risk factors is caused by lifestyle issues (e.g. smoking, lack of exercise, improper diet), adverse effects of long-term medication, persistent subsyndromal symptoms, and also by limited availability and quality of somatic care for mentally ill patients.[1-3] In addition to an increased occurrence of obesity, diabetes, hypertension, and dyslipidemia, individuals with SMD, and notably those with schizophrenia, have a greater risk of pathological blood clotting. The risk is even higher during hospitalization, immobilization, and treatment with antipsychotic agents. Diagnosis and treatment of VTE in psychiatric patients is difficult due to sedation and immobilization, as well as symptoms of the disease and patients' non-compliance. The occurrence of VTE has been reported to increase the morbidity and impair the quality of life of psychiatric patients. Moreover, undiagnosed venous thrombosis can have fatal consequences due to the relatively frequent occurrence of pulmonary embolism.

In the past, an association between antipsychotic therapy and VTE was published in a review article by Hägg et al.[4] In the current review, we will describe additional studies from recent years, including the first meta-analysis published on this topic, and we will present an adjusted version of our own guideline for the prevention of VTE in hospitalized psychiatric patients.

Venous Thrombosis and Pulmonary Embolism

Venous thromboembolic disease occurs in 1 out of 1000–2000 adult patient-years. It is a frequent and costly disease, which is associated with high morbidity and mortality rates in the developed countries of Europe and North America. The actual occurrence is likely to be substantially higher because recent studies using systematic lung scans in all patients with acute deep venous thrombosis showed that the occurrence of silent pulmonary embolism was 40–50%.[5] An overview of risk factors for venous thromboembolic disease[6] is provided in Table 1.

Table 1. Risk factors for first venous thromboembolism event
  1. APC, activated protein C resistance.
GeneticAntithrombin deficiency
Protein C deficiency
Protein S deficiency
FV Leiden mutation
FII G20210A mutation
AcquiredIncreasing age
Antiphospholipid antibody syndrome
Inflammatory bowel disease
Systemic lupus erythematosus
Nephrotic syndrome
Overweight and obesity
EnvironmentalInfectious disease
Surgery, trauma, immobilization
Pregnancy and puerperium
Oral contraceptives
Hormone replacement therapy
Antipsychotic use
Air travel (long-haul)
MixedHigh factor VIII
High factor IX
High factor XI
Not well establishedHigh C-reactive protein
Male sex

A hypercoagulation event may occur when at least one of Virchow's triad rules is present (i.e., vessel wall injury, activation of the blood coagulation cascade, venous stasis). Each risk factor is related to at least one point from the said triad. The risk of VTE increases considerably with the combination of individual factors because these factors are not summed up but rather multiplied.

Occurence of VTE During Treatment with Antipsychotics: Pharmacoepidemiologic Data

The first case histories and retrospective studies of thromboembolic complications in patients treated with chlorpromazine and other antipsychotic agents (AP) appeared in the German literature shortly after the discovery of the antipsychotic activity of chlorpromazine and its introduction into practice during the 1950s.[7, 8] From 1958 to 1961, the risk of venous thrombosis in psychiatric and neurologic patients tripled, compared to the years 1915–1922.[9]

The issue has attracted greater attention in the last decade. Zornberg and Jick[10] documented a significantly increased risk of VTE during treatment with first-generation antipsychotics (FGA) and reported the highest incidence of events during the first months of treatment. The risk was found to be higher in basal antipsychotics than in incisive antipsychotics. Findings from important studies are summarized in Table 2. In this article, we discuss in detail only those studies conducted during the last 2 years.

Table 2. Summary of basic studies evaluating the risk of VTE during treatment with antipsychotics, without group differentiation
StudyYearAge (years)Study designGroupEnd-pointRelative risk (95%CI)
  1. AHR, adjusted hazard ratio; AOR, adjusted odds ratio; AP, antipsychotics; ARR, adjusted risk ratio; FGA, first-generation antipsychotics; H-P, high-potency FGA; L-P, low-potency FGA; OR, odds ratio; PE, pulmonary embolism; SGA, second-generation antipsychotics; VTE, venous thromboembolism; WHO, World Health Organization.
Walker et al. (1997)[11]1991–199310–54Study from the report registry67 072 current and former users of clozapineMortality incl. PERate ratio, current vs past clozapine users: 5.2 (0.7–38.6)
Zornberg and Jick (2000)[10]1990–1998<60Case–control study29 952 AP usersVTEFGA: AOR 7.1 (2.3–22)
L-P: AOR 24.1 (3.3–172.7)
H-P: AOR 3.3 (0.8–13.2)
Hägg et al. (2000)[12]1989–200018–60Case series from the Swedish Adverse Events databaseSpontaneously reported cases of VTEVTEThe risk estimate was 1 case per 2000–6000 patients treated with clozapine
Parkin et al. (2004)[13]1990–199815–59Case–control study75 subjects with fatal PE; 300 healthy controlsPEFGA: AOR 9.7 (2.3–40.9)
L-P: AOR 29.3 (2.8–308.2)
Liperoti et al. (2005)[14]1998–1999≥65Retrospective cohort study19 940 new AP users; 112 078 non-usersVTESGA: AHR 2.01 (1.50–2.70)
FGA: AHR 1.02 (0.67–1.55)
>1 AP: AHR 4.80 (2.29–10.10)
Ray et al. (2002)[15]1994–2000≥65Retrospective cohort study22 514 AP users; 33 033 controlsVTEAP: AHR 1.10 (0.95–1.27)
haloperidol: AHR 1.43 (1.18–1.74)
Hamanaka et al. (2004)[16]1998–2002Retrospective prevalence, control-group study1125 autopsy casesPEAP: AOR 10.49 (3.95–27.85)
Masopust et al. (2007)[17]1996–200418–60Case–control study266 cases; 274 controlsVTEAP: OR 2.8 (1.0–7.6)
Lacut et al. (2007)[18]2000–2004>18Case–control study677 cases; 677 controlsVTEAP: OR 3.5 (2.0–6.2)
FGA: OR 4.1 (2.1–8.2)
SGA: OR 2.7 (0.7–10.0)
Hägg et al. (2008)[19]1975–2004Serious Adverse Events database, WHO3.2 million reported Adverse Events worldwideVTEsignificantly more cases of VTE than expected:
clozapine (n = 375)
olanzapine (n = 91)
sertindole (n = 9)
Jönsson et al. (2008)[20]1992–200518–65Retrospective study of prevalence14 439 casesPEL-P: AOR 2.39 (1.46–3.92)
SGA: AOR 6.91 (3.95–12.10)
Jönsson et al. (2009)[21]1997–2005>18Case–control study5999 cases; 59 990 controlsVTEAP: ARR 1.99 (1.69–2.34)
SGA: ARR 2.47 (1.87–3.28)
L-P: 2.11 (1.51–2.95)
H-P: 1.82 (1.46–2.27)
Kleijer et al. (2010)[22]1998–2008≥60Case–control study111 818 cases; 447 272 controlsVTEAP: OR 0.9 (0.7–1.1)
Parker et al. (2010)[23]1996–200716–100Case–control study25 532 cases; 89 491 controlsVTEAP: AOR 1.32 (1.23–1.42)
SGA: AOR 1.73 (1.37–2.17)
FGA: AOR 1.28 (1.18–1.38)
L-P: AOR 1.99 (1.52–2.62)
H-P: AOR 1.28 (1.18–1.38)
Allenet et al. (2011)[24]2006≥18Retrospective cohort study450 951 AP users; 28 272 820 non-AP-usersPEAP: AOR 1.17 (1.13–1.21)
FGA: AOR 1.19 (1.13–1.25)
SGA: AOR 1.15 (1.09–1.21)
Clozapine: AOR 1.46 (1.05–2.02)

A recent case–control study included 25 532 patients with VTE.[23] Four corresponding control patients were assigned to each case. Patients who had been prescribed an antipsychotic in the last 2 years had a 32% higher risk of VTE than patients not treated with antipsychotics. If they were current users of antipsychotic treatment, the risk was 56% higher (odds ratio [OR] 1.56; 95% confidence interval [CI] 1.39–1.75). The risk of thromboembolism during the first 3 months of treatment with an antipsychotic agent was double (OR 1.97; 95%CI 1.66–2.33). By comparing the individual groups of antipsychotics, the authors found a higher risk of VTE when the patients were treated with second-generation antipsychotics (SGA) compared to FGA and with low-potency products compared to high-potency products. Concerning the individual products, the risk was more than double with haloperidol (OR 2.17; 95%CI 1.55–3.02) and quetiapine (OR 2.81; 95%CI 1.75–4.50). Age was a significant risk factor. The overall occurrence of VTE was 4 (3–5) cases per 10 000 patients treated for the period of 1 year (patient-years). In patients above 65 years of age, the risk rose to 10 (7–13). This was confirmed by the results of the study by Liperoti et al.[14] In contrast, the Canadian[15] and recent German studies[22] found no significant relation between the administration of antipsychotics and risk of VTE in elderly people. The dose and pharmaceutical form of the antipsychotic also played a role. When using the maximum dose of the antipsychotic, the probability of VTE was higher than for low doses up to 25% of the recommended daily dose (OR 2.59; 95%CI 1.27–529 vs OR 1.26; 95%CI 1.17–1.36). Administration of the injection form of the drug was associated with a more than threefold higher risk of VTE (OR 3.24; 95%CI 2.04–5.17). The study demonstrated prescription practices in Great Britain. Approximately three-quarters of all patients using antipsychotics were prescribed prochlorperazine for nausea, vomiting, or vertigo. Only a minority of people received antipsychotics for treatment of a psychosis or agitation in dementia.

Allenet et al.[24] analyzed data from more than 28 million patients in 2006. AP were used by 450 951 individuals (1.6%). The prescription of AP was associated with a significantly higher risk of pulmonary embolism, compared to the remaining population (OR 1.17; 95%CI 1.13–1.21). Clozapine appeared to be the most risky. Alternatively, the lowest risk of PE was reported in patients taking quetiapine and aripiprazole. No difference was found between the FGA and SGA.

Sga and Venous Thromboembolic Disease

Concerning individual SGA (except for clozapine), we have only limited data on the relation between their administration and development of VTE in patients younger than 65 years of age. Many case histories and case series have been published, however. Data from the Japanese population[16, 25] suggest a potential association between risperidone and massive pulmonary embolism, which was fatal in the majority of cases. Several cases of the development of VTE shortly after the initiation of treatment with olanzapine were described.[26-32]

Hägg et al.[19] analyzed data from the WHO database containing reports of adverse reactions to antipsychotics. A robust association was found between the occurrence of VTE and the group of SGA. In contrast, the group of low- and high-potency FGA was not associated with a significantly higher occurrence of VTE. Occurrences of VTE were reported more frequently during treatment with clozapine, olanzapine, zuclopenthixol and sertindole than during treatment with other antipsychotics. A case of a patient with recurrent thromboembolic events during treatment with olanzapine and risperidone was recorded. The patient was successfully switched to amisulpride, a product that does not block the 5-HT2A receptors.[26] Furthermore, no case records are available in the literature that would describe the occurrence of VTE in patients treated with ‘metabolically gentler’ ziprasidone and aripiprazole.

The first meta-analysis of studies evaluating the relation between antipsychotic treatment and the occurrence of VTE[33] was published recently. The authors included seven case–control studies from the PubMed database. The risk of development of VTE during antipsychotic treatment was more than double (OR 2.39; 95%CI 1.71–3.35), as compared to the controls. When classifying the products into groups, the most risky were low-potency antipsychotics (OR 2.91; 95%CI 1.81–4.71), followed by SGA (OR 2.20; 95%CI 1.22–3.96) and traditional neuroleptics (OR 1.72; 95%CI 1.31–2.24). The safest in terms of thromboembolic episodes were high-potency products (OR 1.58; 95%CI 1.50–1.67).

All of the studies described above are limited due to their designs, which were retrospective analyses of case–controls. The main limitations are the great heterogeneity of the involved disorders, the doses and duration of antipsychotic treatment, and the cause of VTE. Nevertheless, the results are relatively consistent. It appears that the risk of VTE is the highest during the first 3 months after the initiation of antipsychotic treatment. The overall risk from the use of antipsychotics is more than double, compared to their non-use. The highest risk group is low-potency products.

Clozapine and Venous Thromboembolic Disease

Clozapine is an antipsychotic for which there is repeated pharmacoepidemiologic evidence of its relation to VTE. Based on analysis, Hägg[12] estimated the incidence of VTE to be 1 in 2000–6000 individuals treated with clozapine for 1 year. Based on the adverse reaction reports, the Food and Drug Administration found the overall risk of death in association with pulmonary embolism during the therapy with clozapine to be 1 in 3450 patients per year.[34] The Food and Drug Administration recorded 99 relevant reports on the incidence of VTE in relation to treatment with clozapine in the years 1990–1999.[35] Walker et al.[11] in their major mortality study, confirmed the anti-suicidal effects of clozapine but also found a significantly higher number of fatal VTE in patients using clozapine. In addition, the literature describes 22 case histories of VTE during treatment with clozapine.[36] The mean age of these patients was 38 years, which was low for the occurrence of thromboembolism. Death of the patient occurred in 44% of cases. Data from Great Britain show that the risk of development of fatal agranulocytosis during treatment with clozapine is lower than the risk of fatal pulmonary embolism (1:10 000 vs 1:4500).[37] The possible mechanisms leading to VTE during the administration of clozapine include sedation and obesity, which reduce movement activities and lead to venous stasis. Obesity, moreover, serves as an independent risk factor for VTE. Serotonin-induced platelet aggregation due to the high affinity of clozapine for the 5-HT2A receptors can also contribute to the development of VTE. Effects of clozapine on the increase of platelet adhesion and aggregation in vitro[38] have been reported. Another factor to consider is the clozapine-induced creation of antiphospholipid antibodies (APA) with prothrombogenic effects[4] or the elevation of hsCRP (hsCRP is the C-reactive protein as a marker of inflammation assessed by a highly sensitive method) during the initiation of treatment, with negative impacts, including endothelial damage.[39] Given the high risk, monitoring for the signs of VTE is recommended, especially during the first 3 months of clozapine administration.[37] It should be noted that usually patients treated with clozapine are carefully monitored due to its potential severe adverse effects (agranulocytosis, myocarditis and cardiomyopathy, epileptic paroxysms), and so complications can be reported more frequently. In this respect, findings from the Finnish cohort study[40] showed that treatment with clozapine was associated with the lowest mortality (including that from cardiovascular causes) of all antipsychotic agents.

Mechanisms Involved in the Etiopathogenesis of VTE During Treatment with Antipsychotics

Mechanical injury of blood vessel walls is usually not present in psychiatric patients except for shortly after surgery or injury. The only exception is injury or contusion of the extremities, which can occur due to physical restraints. One of the risk factors is compression resulting in venous occlusion in patients fastened to the bed. Increased activation of blood coagulation was demonstrated by measuring fibrinolysis parameters during compression of the arm with the cuff of a blood pressure monitor. Inflammation of the endothelium and its dysfunction are also associated with thrombogenesis and atherogenesis. Functioning of the vascular endothelium is damaged by various factors, such as high blood pressure, hyperglycemia, low-density lipoprotein cholesterol, hyperhomocysteinemia or smoking. Often these factors are already present in young patients with psychosis and may further worsen during the treatment.

Venous stasis is an important factor, especially in hospitalized patients. The lifestyle of patients in long-term treatment with antipsychotic agents is characterized by a lack of physical activity, improper health regimen and increased bodyweight. Hospitalization itself, and the various levels of immobilization related to it, is a risk factor for VTE. Antipsychotics and often co-administered benzodiazepines cause sedation and overall reduction of physical activity, which contributes to venous stasis. The highest levels of sedation are associated with treatment with low-potency FGA and clozapine.[41] Immobilization due to physical restraint also plays a significant role.[42] An alarming fact described in case histories is the presence of disease in young individuals after a relatively short period of immobilization. The main causative factor in these cases is immobility associated with venous stasis. Injuries to the extremities during physical restraint or compression of an extremity with fasteners may also have prothrombogenic effects. In addition to VTE, asphyxia or cardiac disorders have also been reported as causes of death during physical immobilization.[43] Aspiration or blunt thoracic injury may occur. Catecholamines released as a result of stress during physical immobilization have negative effects. Other risk factors include immobilization-related conditions, such as catatonia, stupor and neuroleptic malignant syndrome.[42, 44]

A third factor in thrombogenesis is abnormality in the coagulation pathway (activation of blood coagulation). Our knowledge in this area has improved greatly during the last years. The diagnosis of schizophrenia and/or bipolar affective disorder is associated with an increased risk of venous thromboembolism. In a Danish study based on a large registry of psychiatric patients (almost 26 000 patients), a higher incidence of pulmonary embolism was reported in patients with bipolar affective disorder, especially in male patients. In addition, an increased incidence of pulmonary embolism was found in women diagnosed with schizophrenia.[45] The recent observational study and autopsy study suggest a relatively high incidence of pulmonary embolism among patients with schizophrenia and other psychotic disorders. However, these works contain little or no information on current treatment with antipsychotics. The acute phase of a psychotic disease is characterized by an elevated release of adrenalin, which stimulates platelet aggregation.[43, 46] In our study, we found significantly increased activation of the markers of thrombogenesis (i.e., sP-selectin, d-dimers, FVIII) in patients having the first attack of psychosis before initiation of antipsychotic therapy compared to healthy volunteers.[47] Increased levels in the markers of thrombogenesis continued after 1 year of therapy.

Some psychopharmaceuticals are associated with increased levels of APA. These are antibodies against phospholipid-dependent plasma proteins bound to a negatively charged, most often phospholipid surface. This is a very non-homogeneous group of antibodies, which interfere (with varying levels of significance and in various ways) with the blood coagulation process, usually resulting in thrombophilia. These pharmaceuticals include FGA, in particular, represented by chlorpromazine[48, 49] and clozapine.[12, 50] Of note, increased antibody titers are present primarily in patients with schizophrenia. Firer et al.[51] found increased levels of class immunoglobulin (Ig)G anticardiolipin antibodies (ACLA) in a group of patients with schizophrenia and their relatives, as compared to a control group of healthy volunteers. ACLA IgM was significantly more present in the patients. In 75% of families, at least one family member had positive ACLA, and the same findings were usually also reported in the patient's relatives. A high prevalence of autoantibodies (e.g. APA, antinuclear antibodies or rheumatoid factor) was demonstrated in 70% of individuals from a group of 184 men treated with neuroleptic drugs.[48] Chengappa et al.[52] and Schwartz et al.[49] demonstrated a higher prevalence of APA (approximately 30%) in patients with schizophrenia. Antipsychotic therapy further increased the already elevated APA titer. This increases the risk of thromboembolic complications. Shena et al.[53] demonstrated a relation between clozapine serum levels and increased APA titers in patients with schizophrenia. Therapy with risperidone in schizophrenic patients was also associated with increased APA levels, compared to the normal population.[54] Halacheva et al.[55] reported an increased titer of anticardiolipin antibodies in schizophrenic patients before and after treatment with antipsychotics but without any relation to psychotic symptoms and treatment administration. A disorder of membrane phospholipids in the erythrocytes and platelets was reported earlier in patients with schizophrenia. Similar abnormalities were found in the frontal lobes of previously untreated patients having the first episode of schizophrenia. The role of autoimmune mechanisms with the production of autoantibodies has been considered in the cause and pathophysiology of schizophrenic disease.[56] Antiphospholipid antibodies (and in particular ACLA) may serve as a marker of autoimmune reactivity or may directly interfere with phospholipid metabolism in patients with schizophrenia. In the event of APA, a coincidence of two factors is expected. The first factor is the presence of APA in the blood, while the other factor is a local thrombogenic trigger. Patients with psychosis and increased titers of APA (approximately one-third of patients) appear to be at a higher risk of death from cardiovascular and respiratory causes during the administration of antipsychotics or physical restraint.[57]

Treatment with antipsychotics, especially phenothiazines, results in an increased platelet aggregation.[58] Serotonin is involved in the primary hemostasis. An increased platelet aggregation was reported during treatment with phenothiazines and was found to be serotonin-induced. The great affinity of clozapine and risperidone to 5-HT2A receptors may be a cause of the 5-HT2A-induced increased aggregation of platelets.[25] Depression, which is often present in patients with schizophrenia, is associated with changes in serotonin metabolism not only in the neurons but also peripherally in the platelets. Elevated adrenalin secretion during the acute psychotic condition may serve as another factor leading to blood coagulation.[43] Clozapine was found to have an effect on platelet adhesion.[38] The influence of antipsychotics on platelets was confirmed by the results of a recent experimental study.[59] In another in vitro study, however, no direct effect of risperidone on platelet activity or plasma coagulation was demonstrated.[60] Contradictory findings were reported in a Polish study:[61] clozapine and olanzapine reduced ADP-induced platelet aggregation in vitro. Carrizo et al.[62] found no differences between markers of coagulation and inflammatory markers in patients with schizophrenia treated with antipsychotics and in their first-degree relatives.

Hyperprolactinemia is another factor to be considered in addition to the effects of antiphospholipid antibodies and platelet activation. Increased prolactin levels are reported during treatment with traditional neuroleptic agents, while among the SGA, risperidone and amisulpride increase prolactinoma. Transient elevation may be associated with olanzapine, zotepine and ziprasidone. Quetiapine and aripiprazole do not increase prolactinaemia.[63] Based on clinical observations, Wallaschofski et al. formulated a hypothesis that increased prolactin levels are involved in the platelet activation process. In their study, they observed pregnant women, patients with a pituitary gland tumor (with either normal or increased prolactin levels) and healthy controls. They also tested platelet activation and aggregation in vitro. Based on the results of this study, the authors concluded that prolactin may be a physiological co-factor for the balance of coagulation during pregnancy and may explain a higher risk of venous thromboembolic complications in the period around birth.[64] Platelet stimulation was found to be statistically significantly higher in patients with hyperprolactinemia induced by antipsychotics than in controls with normal prolactin levels. Prolactin levels significantly correlated with the ADP-stimulated expression of P-selectin, a marker of platelet activation.[65]

Metabolic symptoms caused by antipsychotics, such as bodyweight increase, hyperleptinemia, hyperglycemia and dyslipidemia, are known to be risk factors for VTE and to interfere with the coagulation and fibrinolysis processes.[66, 67] Obesity and other cardiovascular risk factors have been reported to occur in patients with schizophrenia more frequently than in the general population.[1] Obesity (body mass index ≥ 30) is associated with a twofold increase in the occurrence of venous thrombosis. Obese individuals have higher levels of factors VIII and IX.[68] The thrombophilic condition also reduces fibrinolytic activity and capacity. Abdominal obesity is associated with the effects of hormonally active mediators and leads to a pro-coagulation and pro-inflammatory condition with a high risk of thrombosis.[69] Increasing bodyweight and metabolic abnormalities are most frequently reported during the administration of clozapine and olanzapine. Use of other SGA, such as quetiapine, risperidone and zotepine, may also be associated with bodyweight increases of varying significance.[1] A recent work describes significantly increased hsCRP levels shortly after the initiation of therapy with clozapine, which, however, showed no difference from healthy controls, in patients using clozapine for more than 1 year.[39]

Some works have reported hyperhomocysteinemia in patients with schizophrenia.[70-72] Elevated homocysteine levels were probably related more to the disease itself in combination with poor dietary habits, such as smoking cigarettes and excess consumption of black coffee and tea, than with the treatment with antipsychotics.[73] A mild hyperhomocysteinemia with levels not exceeding 18.5 μmol/L showed a 2–2.5-fold increase in the risk of venous thrombosis, while levels above 20 μmol/L showed a 34-fold increased risk.[74]

Although the literary data on the relation between antipsychotics and pathological blood clotting is not fully consistent, antipsychotic medication is evidently one of the factors inducing coagulation and hemostasis.

Possible biological mechanisms of action of antipsychotics in relation to venous thrombosis and pulmonary embolism are shown in Figure 1. Table 3 contains the prothrombogenic factors of selected antipsychotics.

Figure 1.

Possible mechanisms involved in the etiopathogenesis of venous thromboembolic disease (VTE) during the treatment with antipsychotics.

Table 3. Possible post-thrombogenic mechanisms during treatment with selected antipsychotics
  1. APA, antiphospholipid antibodies.
PhenothiazinesSedation, hypotension, APA, 5-HT2 receptor blockade, bodyweight increase, hyperprolactinemia
ClozapineSedation, hypotension, APA, 5-HT2 receptor blockade, bodyweight increase, metabolic symptoms
Risperidone5-HT2 receptor blockade, hyperprolactinemia, APA, bodyweight increase (?)
Olanzapine5-HT2 receptor blockade, bodyweight increase, metabolic symptoms, APA, sedation (?)

Risk of thromboembolism in hospitalized psychiatric patients

According to the Leiden Thrombophilia Study, a prolonged hospitalization increases the risk of VTE more than 11 times.[75] Patients with SMD are hospitalized frequently and repeatedly on a long-term basis and sometimes permanently. This fact itself means a predisposition for a higher risk of VTE. The risk factors for VTE that play a role during hospitalization at psychiatry clinics are summarized in Table 4. Cases of death of immobilized young patients treated with antipsychotics and without other known risks of VTE have been reported.[42, 43, 76]

Table 4. Risk factors of venous thromboembolic diseases in psychiatric patients
Long-term hospitalization
Immobilization due to physical restraint
Catatonic conditions, stupor, neuroleptic malignant syndrome
Overall limitation of mobility
Pharmacological sedation (benzodiazepines)
Administration of antipsychotics
Diagnosis of schizophrenia and bipolar affective disorder

Based on the literature regarding non-surgical and surgical patients with respect to specificities in mental disorders, we have created the first and still the last guideline for the prevention of VTE in psychiatric patients with limited mobility.[77] The scoring system (Table 5) contains general important risk factors for VTE and markers involved in the pathogenesis of venous thrombosis in hospitalized psychiatric patients.[77] After verifying the prevention guideline in practice in an acute admission setting together with prospective monitoring of immobilized patients, we have further adjusted and simplified the prevention guideline (only two grades of risk instead of three were implemented, the age limit was removed) (Table 6). According to our experience, the adjusted guideline ‘as is’ can be used easily. Despite this, prevention of VTE in accordance with the guideline was used in only 55% of cases during the prospective monitoring.[78] If the contraindications (a high risk of bleeding or an active bleeding) to low-molecular-weight heparin are absent, prevention of VTE is safe and effective.[79]

Table 5. Score of risk factors for venous thromboembolism in hospitalized psychiatric patients with reduced mobility
Risk factor: score 2 pointsRisk factor: score 1 point
  1. aTrunk varicose veins in the areas drained by the greater and lesser saphenous veins, swellings and hyperpigmentation.
  2. bDehydration is defined as a clinically significant loss of body fluids (severe hydration is a loss of >10%), for example due to diarrhea, hyperthermia, vomiting, burns, gastroenteritis, malnutrition with electrolyte imbalance, severe hyperglycemia in patients with diabetes, and so on.
  3. cFactor V Leiden mutation, prothrombin mutation 20210 G/A, antithrombin deficiency, protein C deficiency or protein S deficiency, antiphospholipid syndrome, hyperhomocysteinemia, high factor VIII levels, dysfibrinogenemia.
  4. BMI, body mass index; DVT, deep venous thrombosis; HCC, hormonal contraception; HRT, hormone replacement therapy; PE, pulmonary embolism; VTE, venous thromboembolic disease.
Personal history of VTE (DVT/PE)Immobilization (extremity paralysis, physical restraint ≥ 8 h, catatonia)
Malignancy (active/treated)Estrogen therapy (HCC, HRT)
Age ≥ 75 yearsObesity (BMI ≥ 30)
Acute infection (severe infection/sepsis) or acute respiratory tract disease (including exacerbation of a chronic disorder)Age 60–74 years
Varicose veins / venous insufficiencya
Thrombophilia (laboratory)c
Treatment with antipsychotics
Table 6. Guideline for thromboprophylaxis – adjusted version. Recommended prevention of venous thromboembolic disease according to the level of risk
Low risk ≤ 3 pointsModerate to high risk ≥ 4 points
  1. LMWH, low-molecular-weight heparin.
  • Regular lower extremity exercise
  • Sufficient hydration
  • Compressive antithrombotic stockings
No pharmacoprophylaxisLMWH 1 × daily s.c. (enoxaparin 40 mg, nadroparin 0.4 mL, dalteparin 5000 IU)


With our project, entitled ANTRE (ANtipsychotics, ThRombosis, Embolism), we are trying to manage the issue of venous thromboembolism in patients treated with antipsychotics. In an epidemiological analysis, we found a higher exposure to antipsychotic treatment in patients with VTE compared to patients with arterial hypertension (OR 2.76; 95%CI 1.01–7.55).[17] In a case series of patients with VTE while on olanzapine treatment, we focused on identification of the present risk factors. We proposed possible mechanisms involved in the etiopathogenesis of VTE during the administration of olanzapine and/or other antipsychotics.[30] We designed a needed guideline for the prevention of VTE in hospitalized psychiatric patients.[77] We verified the guideline in practice using prospective monitoring, and we made adjustments to make its use easier and simple.[78] This long-term prospective study enables us to monitor and evaluate the etiopathogenetic mechanisms involved in pathological blood coagulation in patients suffering from schizophrenia. We compare hemostatic, coagulatory, immunological and biochemical parameters before the initiation of antipsychotics and after several months of treatment in patients having a first attack of psychosis, as well as in a group of healthy volunteers. The results of the pilot study (i.e. significantly higher d-dimer and sP-selectin levels in the patient group) support the idea of a pro-coagulation condition in the acute disease phase and possible effects of antipsychotic agents on thrombogenesis.[47]


Observational studies show an increased risk for the development of thromboembolic complication in patients treated with antipsychotics. Despite differing methods and many limitations, these studies provide relatively consistent results. The occurrence of VTE during treatment with antipsychotics is not frequently reported in clinical practice, but it can have fatal consequences. The highest risk is present during the first 3 months of treatment with antipsychotics. The highest-risk products are clozapine and low-potency FGA. Less evidence is available for olanzapine, risperidone and high-potency FGA. Concerning the pathogenesis of VTE, it is difficult to differentiate between the role of mental disease and the effects of antipsychotics. The risk significantly increases with immobilization of the patient, most frequently when physical restraints are used. Despite many possible explanations, the precise biological mechanisms for the development of VTE during treatment with antipsychotics remain unclear. Given the current level of data and knowledge, no specific prevention of VTE is recommended in non-hospitalized patients treated with antipsychotics. It is advisable to monitor the physical health of patients with SMD treated with antipsychotics on a regular basis. When selecting the product, physicians should screen for risk factors of VTE (e.g. obesity, dyslipidemia, venous insufficiency, patient's history of VTE, and so on) and in the event of multiple risk factors, they should consider metabolically gentle antipsychotics without marked sedation. High-risk patients and their relatives should be instructed about the possible adverse effects of the treatment, including signs and symptoms of venous thrombosis and pulmonary embolism, and the necessity of seeking medical attention immediately in the event of their occurrence. Diagnosis and treatment of VTE is within the competence of an internist. A psychiatrist should not hesitate to consult with specialists or refer the patient for a specialist examination given the potentially fatal outcome of VTE. In hospitalized patients with reduced mobility, prevention of VTE should be undertaken according to the guidelines. ‘Low-threshold’ use of low-molecular-weight heparin is recommended in patients with multiple risk factors. Further studies will be needed to elucidate the relations between antipsychotics and venous thromboembolism, focusing on quantification of risk carried by the respective products in this heterogeneous group of treatments.


This study was supported by the research program PRVOUK P37/03. The authors declare no conflict of interest.