Skip to main content

Antipsychotic drug use and risk of stroke and myocardial infarction: a systematic review and meta-analysis



The effect of antipsychotic (AP) drugs on risk of stroke and myocardial infarction (MI) remains unclear due to methodological limitations of, and inconsistencies across, existing studies. We aimed to systematically review studies reporting on the associations between AP drug use and stroke or MI risk, and to investigate whether associations differed among different sub-populations.


We searched Medline, EMBASE, PsychINFO and Cochrane Library (from inception to May 28, 2017) for observational studies reporting on AP drug use and MI or stroke occurrence. We performed random-effects meta-analyses for each outcome, performing sub-groups analyses by study population – specifically general population (i.e. those not restricted to patients with a particular indication for AP drug use), people with dementia only and psychiatric illness only. Where feasible we performed subgroup analyses by AP drug class.


From 7008 articles, we included 29 relevant observational studies, 19 on stroke and 10 on MI. Results of cohort studies that included a general population indicated a more than two-fold increased risk of stroke, albeit with substantial heterogeneity (pooled HR 2.31, 95% CI 1.13, 4.74, I2 = 83.2%). However, the risk among patients with dementia was much lower, with no heterogeneity (pooled HR 1.16, 95% CI 1.00, 1.33, I2 = 0%) and there was no clear association among studies of psychiatric populations (pooled HR 1.44, 95% CI 0.90, 2.30; substantial heterogeneity [I2 = 78.8])). Associations generally persisted when stratifying by AP class, but few studies reported on first generation AP drugs. We found no association between AP drug use and MI risk (pooled HR for cohort studies: 1.29, 95% CI 0.88, 1.90 and case-control studies: 1.07, 95% CI 0.94, 1.23), but substantial methodological and statistical heterogeneity among a relatively small number of studies limits firm conclusions.


AP drug use may be associated with an increased risk of stroke, but there is no clear evidence that this risk is further elevated in patients with dementia. Further studies are need to clarify the effect of AP drug use on MI and stroke risk in different sub-populations and should control for confounding by indication and stratify by AP drug class.

Peer Review reports


Major mental disorders represent a growing and, until relatively recently, under-recognized global public health burden. Schizophrenia, bipolar disorder, major depression and anxiety feature among the top 20 causes of years lived with disability (YLD) [1] and are associated with marked premature mortality [2,3,4]. Much of this excess mortality is due to a higher burden of cardiovascular and cerebrovascular disease (largely ischaemic heart disease and stroke) compared to the general population [5,6,7]. Several factors related to mental illness, including low socioeconomic status, lifestyle, physical comorbidities, genetic predisposition and healthcare access [7,8,9] could contribute to increased cardio- and cerebrovascular disease incidence in this vulnerable group. Prescription medication, including antipsychotic (AP) drug use, has also been raised as a possible cardiovascular and cerebrovascular disease risk factor, potentially operating through effects on body weight, metabolic factors and thrombosis [10].

Despite their potential to increase risk of circulatory disease, AP drugs are also being increasingly used off-label, for the treatment of dementia, anxiety, insomnia and post-traumatic stress [11], with little understanding of the long-term side-effects, including risks of major cardiovascular events such as myocardial infarction and stroke. Their increased use in patients with dementia is particularly concerning, with around 20% of patients with dementia in nursing homes in the USA and UK treated with APs [12, 13]. Use of these drugs in this population has been linked to increased risk of stroke as well as other adverse outcomes, including increased mortality [14] and thus regulatory bodies have discouraged the use of AP drug use in people with dementia [15]. However, the evidence for increased stroke risk is based largely on analyses of serious adverse events in randomized controlled trials of AP drugs in patients with dementia [10, 16, 17]. Confirmation of the appropriateness of these attached warnings to AP drug use is important, given that pharmacological alternatives for the treatment of behavioral and psychological symptoms of dementia are quite limited [18].

A number of observational studies have reported on the association between AP use and stroke risk, but until very recently these had not been systematically reviewed. Since commencing the present review, one systematic review and meta-analysis has been published which concluded that first generation, but not second-generation AP drug use was associated with an increased risk of cerebrovascular disease. The authors also reported that among those with dementia in particular, use of any AP was associated with a low risk of cerebrovascular disease [19]. However, this study identified almost half as many studies as in the present review, thus omitting relevant additional studies on this topic. Earlier reviews focused on summarizing the effects of AP drug use specifically in the elderly or those with dementia [10, 20], the most recent of which suggests that AP drug use may be associated with an increased risk of stroke in this population [20]. Existing systematic reviews on the association between AP drug use and risk of myocardial infarction (MI) are conflicting. The first review concluded that the evidence for an association is inconclusive [21], whereas a more recent review concluded that AP drug use is associated with an increased risk of MI [22]. Interestingly, a subsequent additional review (which was published after we completed the screening phase of the present review) included the same studies as this previous review, but drew different, more cautious, conclusions [23]. There is therefore a lack of clarity on the associations between AP drug use and the occurrence of major cardio- and cerebrovascular events. To address this, we performed a systematic review and meta-analysis of studies reporting on the association between antipsychotic drug use and risk of stroke and MI.


Search strategy

We searched Embase, MEDLINE, PsychINFO (via OVID) and the Cochrane Library from their origin to May 28, 2017 using a comprehensive search strategy comprising medical subject heading terms and free text words for the exposure (AP drugs) and outcomes (stroke and MI) of interest (Additional file 1). We restricted our search to English language articles only and perused reference lists from previous reviews and relevant included studies to identify any additional studies. The first reviewer (SZ) and one of 2 second reviewers (NK and AK) independently screened all titles and abstracts, and, for potentially relevant studies, screened full text articles to determine eligibility for inclusion. We included conferences abstracts that were indexed in the search engines.

Inclusion and exclusion criteria

We sought studies that included a general population (i.e. those not restricted to patients with a particular indication for AP drug use) or people with specific mental disorders often treated with AP drugs (e.g. schizophrenia, bipolar disorder, major depression or dementia), and compared AP drug use versus no AP drug use in relation to risk of stroke or MI. We anticipated that confounding by indication [24] could be a limitation of many studies and so we did include studies which attempted to address this by comparing people taking APs versus those on other medications for mental illness (but not taking APs). Confounding by indication could occur because the indication for AP drug use, such as having a serious mental illness, is itself associated with increased risk of stroke or MI, thus potentially leading to a spurious association between AP drug use and stroke or MI risk. We excluded studies: that compared AP drug use with active comparators; and that reported only on stroke or MI mortality, since we were interested in the association between AP drug use and risk of stroke/MI occurrence and not death following stroke/MI. Where two studies overlapped in terms of study population we selected the larger of the two studies.

Data extraction

Pairs of reviewers (SZ and CK or SZ and NK) independently extracted data on: study design; sample size; country; patient population; baseline characteristics of each comparator arm; definition of exposure and outcome; ascertainment of exposure and outcome; length of and loss to follow up; number of events; main results including crude and adjusted effect estimates with accompanying 95% confidence intervals; method of statistical analysis; and adjustments for confounders. A fourth investigator (CAJ) reviewed any disagreements. We assessed study quality and risk of bias using the SIGN checklist [25].


We used Stata version 14.0 to conduct meta-analyses on the association between AP drug use and each of stroke and MI occurrence. We pooled studies according to study design and type of effect estimate, grouping together: cohort and nested case-control studies reporting hazard ratios (HRs); cohort/nested case-control studies reporting odds ratios (ORs); and standard case-control studies reporting ORs. For stroke, we stratified these meta-analyses by type of study population, creating sub-groups of studies that included: patients diagnosed with psychiatric disorders; patients diagnosed with dementia; and a general population with unspecified indications for AP use. Whilst we summarized findings from self-controlled case series and case-crossover studies, we did not include these studies in our meta-analyses, given the differences in study design and analytical approach. Due to a fewer number of studies, almost all of which were conducted using general population data, we were unable to perform sub-group analysis by type of study population when pooling together studies reporting on MI. We pooled effect estimates using the random effects method, assessing heterogeneity using the I2 and Chi2 statistics. In accordance with Cochrane Collaboration guidance, we interpreted heterogeneity based on I2 values as follows: 0–40% - may not be important; 30–60% - may be moderate; 50–90% - may be substantial; 75–100% may be considerable. We interpreted I2 values in the context of the magnitude and direction of effects as well as the chi2 p-value for heterogeneity. In order to improve comparability between studies, where results on the association between AP drug use and cerebrovascular disease risk were reported for different exposure periods we included the longest exposure period, often defined as the ‘ever exposed’ period, or as close to is as possible. We narratively summarized findings from self-controlled studies, since it is methodologically inappropriate to pool this type of study design with cohort or case-control studies. In addition, methodological differences (including the reporting of different types of effect estimates) precluded separate pooling of these studies. Where outcome risks by type of AP drug were reported (i.e. first generation antipsychotic [FGA] or second generation antipsychotic [SGA]), we included these separately in the analyses, thus obtaining sub-group summary effect estimates for type of AP as well as overall summary estimates. If studies reported on multiple APs separately, we included the largest group of AP users in our meta-analysis.

We reported the findings of this systematic review in accordance with the PRISMA and MOOSE guidelines [26, 27].


Our search strategy identified 8163 articles. After de-duplication we screened 7008 titles and abstracts, 146 of which were potentially relevant. Following full-text review of these articles, we included 19 articles reporting on cerebrovascular disease [28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46] and 10 reporting on MI [31, 47,48,49,50,51,52,53,54,55] in the review (Fig. 1).

Fig. 1

Flow diagram of literature search and study selection

Quality assessment of included studies

All but three studies [38, 53, 54] used routine administrative health datasets to ascertain AP prescribing and stroke/MI ascertainment, thus eliminating reporting bias and minimizing loss to follow-up (Additional file 1: TablesS1 and S2). The remaining three studies relied on self-reported AP drug use and/or outcome occurrence. Twelve studies reporting on stroke and 7 reporting on MI reported baseline characteristics of comparison groups. All studies adjusted for age and sex and the majority also accounted for cardiovascular comorbidities. Among studies that included a general population (e.g. a cohort identified from primary care/medical insurance records), indication for AP drug use was generally not accounted for. One study compared stroke and MI outcomes in people taking SGAs versus a comparison group taking anti-depressants but not SGAs, in order to partially control for confounding factors associated with initiation of SGAs (related to mental illness) [31]. Socioeconomic status and lifestyle factors, which may be associated with the disorder for which AP drug medication was prescribed, were rarely adjusted for (Additional file 1: Tables S3 and S4). We identified one conference abstract by Wang et al. which was assessed to be of a reasonable quality overall. However, we couldn’t assess some elements due to the lack of available information [55].

Antipsychotic drug use and stroke risk

Eight cohort studies [28, 29, 31, 36, 38, 40, 41, 43], 6 case-control studies (5 nested case-control) [30, 33,34,35, 37, 39], two self-controlled case series studies [32, 42], two case-crossover studies [44, 46] and one case-case-time control study [45] reported on the association between AP drug use and stroke risk. Five studies included a population of dementia patients [28, 29, 37, 39, 40], three included patients with psychiatric disorders [30, 34, 36] and the remaining 11 studies investigated AP drug use in a general population (Table 1). All but one [38] of the included studies used large routine medical databases to identify patients, determine AP drug use and identify outcome. AP drug exposure was defined as new AP drug use [28, 29, 42, 43] or ‘ever’ AP drug use. Only 3 studies specified a minimum exposure period, which ranged from 3 days to 4 weeks [31, 33, 45].

Table 1 Characteristics of included studies, stratified by outcome

The mean baseline age of study participants varied by the study populations included (Table 1). The mean age in studies that included people with dementia only ranged from 78 to 81 years, whilst studies of patients with psychiatric conditions had a younger range (43 to 57 years). The age range in studies that included a general population was much broader (45 to 85 years), reflecting different age restrictions and data sources used. Studies were heterogeneous in terms of follow-up duration, ranging from 1 week [42] to 12 years [37]. Among studies that included a general population, the indication for AP drug use was not reported.

We included all but one [41] of the cohort and case-control studies in our meta-analysis. We separately pooled 13 study populations from 8 cohort studies (including one nested case-control study reporting HRs) [28,29,30,31, 36, 38, 40, 43] and 9 study populations from 5 case-control studies [33,34,35, 37, 39]. Overall, among studies including a general population, use of any AP drug was associated with a 2.3-fold increased risk of stroke (pooled HR 2.31, 95% CI 1.14, to 4.74), but with substantial heterogeneity between studies (I2 = 83.2%; p < 0.001; Fig. 2a). This appeared to be due to larger effect estimates in the study which examined new (as opposed to ‘ever’) AP drug use [43]. Findings were more mixed for studies including only psychiatric patients, where use of any AP drug was associated with a statistically non-significant increased risk of stroke (pooled HR 1.44, 95% CI 0.90 to 2.30), with substantial heterogeneity between studies (Fig. 2a). Similar findings were observed when we pooled case-control studies, although effect estimates were weaker and not significant (Fig. 2d). When we pooled studies that included a general population or patients with psychiatric conditions and stratified by AP drug use, SGA drug use was significantly associated with a 71% increased risk of stroke (pooled HR 1.71, 95% CI 1.16 to 2.53; Fig. 2b), again with substantial heterogeneity between studies (I2 = 74.9%, p = 0.003). Only two studies reported on FGA drug use, with contradictory findings (Fig. 2b).

Fig. 2

Forest plots for the association between antipsychotics and stroke, by study population/antipsychotic class subgroups NB In plots a-c, n/N represents number of cerebrovascular events/total number in comparison group for those using antipsychotic medication and those not using antipsychotic medication. Chen (a) = haemorrhagic stroke; Chen (b) = ischaemic stroke. AP = antipsychotic; CI = confidence interval

Findings between studies of patients with dementia were more consistent. Use of any AP drug was marginally significantly associated with a 16% increase risk of stroke among cohort studies (pooled HR 1.16, 95% CI 1.00 to 1.33), with no heterogeneity between studies (Fig. 2a). Findings were consistent across studies, irrespective of whether AP drug use was defined as ‘new’ or ‘ever’. When dementia studies were stratified according to AP type, neither FGA nor SGA drug use were statistically significantly associated with stroke risk (pooled HR 1.02, 95% CI 0.65 to 1.60 and pooled HR 1.18, 95% CI 0.83 to 1.68, respectively; Fig. 2c), however this is based on just two studies in each sub-group. In keeping with findings from cohort studies, findings from case-control studies showed no clear association between AP drug use and stroke risk in studies of patients with dementia (pooled OR 0.94, 95% CI 0.66 to 1.33), but there was substantial heterogeneity between studies (I2 = 94.3%, p < 001; Fig. 2d).

An additional five self-controlled studies (among which three different variations of the self-controlled approach were adopted) [32, 42, 44,45,46] compared pre- and post-periods of AP drug use for each patient. All five studies found AP drug use was associated with an increased risk of stroke, although follow-up was much shorter than in other study designs, ranging from 1 week to 1 year. Three found a significant association between use of any AP drug and stroke [32, 45, 46]. Shin found a significant association between SGA drug use and increased stroke risk (OR 3.90, 95% CI 3.30 to 4.60) [44] while Pratt reported a significant association between FGA drug use (but not SGA drug use) and increased stroke risk during the week following AP drug initiation (IRR 2.25, 95% CI 1.32 to 3.83) [42]. These findings are also consistent with those from Sacchetti, which found a substantial increased risk of stroke within one-month post-AP drug use initiation [56]. This study wasn’t included in our review since the study population overlapped with that of another study already included [43].

AP drug use and MI risk

Five cohort studies [31, 48, 49, 52, 54], three case control studies [47, 51, 53], one case crossover study [50] and one case-case time control study [55] reported on AP drug use and MI risk (Table 1). Seven studies included a general population and thus did not specify the indication for AP drug use [31, 47, 49, 51, 53,54,55], one included patients with schizophrenia [48], one included patients with psychiatric illness or dementia disorders [50] and one included patients with dementia [52]. In five studies the included population had a mean age of more than 65 years, four included a population aged < 50 years, and one study did not report on age of participants. The follow-up period ranged from 90 days [49] to 13 years [54].

Study findings among five cohort studies were inconsistent, with substantial heterogeneity between studies reporting HRs (I2 = 62.6%; p = 0.05) and those reporting ORs (I2 = 65.7; p = 0.09). There were too few studies to allow sub-group analysis by type of study population and a limited sub-group analysis by AP drug type (Fig. 3).

Fig. 3

Forest plots for the association between APs and MI, by study design and AP class. *Number of MI events in the group prescribed antipsychotic drugs (exposed) and those not prescribed antipsychotic drugs (non-exposed). †Number of people prescribed antipsychotic drugs among cases (those with MI) and controls (those without MI). AP = antipsychotic; HR = hazard ratio; MI = myocardial infarction; NR = not reported

Similarly, there was considerable heterogeneity between the findings from case-control studies (I2 = 78.3%; p = 0.003). Given this statistical heterogeneity, along with methodological heterogeneity in terms of study population and type of AP drug included, summary estimates should be interpreted with caution. Based on existing studies, there was no clear evidence that AP drug use was associated with increased risk of MI (Fig. 3).

In contrast to the findings from cohort and case-control studies, findings from the two self-controlled studies were consistent in showing a significant association between AP drug use and increased risk of MI (50,55). As with the self-controlled studies that examined stroke outcomes, these studies differ from cohort and case-control studies in terms of including a much shorter follow-up period (30–120 days; Table 1).


We found that AP drug use is associated with an increased risk of stroke, but that this association is much weaker in studies that included only patients with dementia than in studies which included the general population or patients with psychiatric disorders. Among study populations that were not restricted to patients with dementia, an increased stroke risk was observed for SGAs only, with less available data on FGAs. However, with the exception of the subgroup of dementia-focused studies, in all analyses there was substantial statistical heterogeneity between studies, even after stratifying by study population and type of AP drug.

There was no clear evidence of an association between AP drug use and risk of MI. However, substantial statistical and methodological heterogeneity between among a relatively small number of studies on MI risk makes it difficult to draw firm conclusions.

We are aware of just one other systematic review and meta-analysis on AP drug use and stroke risk, published during the conduct of our review [19]. This review included fewer studies than we did, but this is partly due to our use of a broader set of inclusion criteria. Whilst the authors of the previous review did perform sub-group analyses, these were not identical to those that we performed, which makes comparability of findings difficult. Consistent with the findings of our review, the authors did report a weaker association between AP drug use and stroke risk among the dementia population compared to studies not restricted to this sub-population. We found that AP drug use was associated with a 16% increased odds of stroke in those with dementia, which was similar to the 17% (OR 1.17, 95% CI 1.08 to 1.26) reported by Hsu and colleagues [19]. Previous narrative reviews drew similar conclusions, but focused on summarizing evidence on AP use and stroke risk in the elderly or those with dementia [10, 20]. Interestingly, Sacchetti concluded that the excess cerebrovascular disease risk does not appear to be confined to patients with dementia, but applies to elderly patients in general [20]. In contrast to our findings, Hsu et al. concluded that the association with stroke risk is stronger for FGAs than SGAs [19]. Conversely, although relatively few studies reported findings by AP drug type, we found more convincing evidence of an association between SGAs and stroke risk, with more limited data available for FGAs. However, there are a number of key differences between the two reviews, which may account for these differing conclusions. In addition to including a greater number of studies, we also stratified by study design in all analyses. In contrast, Hsu et al. appeared to have pooled studies irrespective of study population, type of study and type of effect estimate. Interestingly, a recent meta-analyses found no difference in risk of cerebrovascular events in those prescribed FGAs versus SGAs when pooling data from 5 studies that directly compared the effect of FGAs versus SGAs on cerebrovascular disease risk in people with dementia [57].

As with stroke, studies on MI were heterogeneous in terms of study design, AP drug type, study population and measure of effect. Although we found no clear evidence that AP drug use is associated with increased MI risk, the inconsistency in findings across studies, in the presence of substantial statistical heterogeneity, makes it difficult to draw firm conclusions. This concurs with the findings of a previous review by Brauer [21], despite our review having included an additional 5 studies. Our conclusion contradicts that of a more recent review, which concluded that there is an association between AP drug use and MI risk, albeit with substantial heterogeneity between pooled effect estimates [22]. We identified two additional studies not included in this review [31, 49]. Whilst the authors of this review included sub-group analyses, they still pooled results from methodologically different studies or different study sub-populations and often pooled different effect estimates, which affects reliability and interpretability of pooled results. This may account for the contradiction in findings between the present review and this previous review. Our conclusions do however concur with those of Huang et al., who performed a similar review concurrent to our own review [23]. In line with our findings, the authors conclude that there is insufficient evidence to draw firm conclusions about the association between APs and MI risk and highlight the need for further prospective studies in this area. Given that the conclusions drawn by Yu et al. and Huang et al. are based on the same included studies, but are inconsistent, our review makes an important further addition to the literature, providing support for the conclusions and recommendations by Huang et al. In our review we did not synthesize data on differing time periods of AP drug exposure and risk of stroke and MI. Effect estimates for different AP drug exposure periods were reported in relatively few studies, which were extremely methodologically heterogeneous, limiting formal synthesis. However, within individual studies, there is some evidence that the risk of stroke and MI might be greater in the very acute period of AP drug use, with a lower risk in the longer term. This is consistent with findings from other reviews [20, 22, 23]. These findings largely stem from studies that included a general, older population and so it is unclear whether such a pattern is also present amongst people with major psychiatric disorders for which AP drugs are indicated. Given that relatively few studies have investigated this across different sub-populations, further careful investigation in future studies is needed.

There is biological plausibility for a causal association between AP drug use and increased risk of vascular disease. AP drug use has long been associated with weight gain, which is possibly due to the AP drug affinity for the histamine-1 receptor which, when blocked, modulates feeding behavior and stimulates weight gain [58, 59] and the effect of AP drugs on hypothalamic regulation [60, 61]. AP drug use is also associated with risk of metabolic syndrome, perhaps through the increased affinity for specific serotonin, muscarinic and histamine receptors, among others [62]. APs may also increase risk of thrombosis. A recent meta-analysis [63] found that AP drug use is associated with a 50% increased odds of venous thromboembolisms, albeit with substantial heterogeneity between studies, with possible underlying mechanisms including increased platelet aggregation, AP drug-induced sedation leading to venostasis and clot formation [64]. APs are also thought to be associated with insulin resistance and weight gain, which in turn are risk factors for cardio- and cerebro-vascular disease [65].

Our review benefits from a number of strengths: we identified studies using a detailed and comprehensive search strategy; we performed meta-analyses using a carefully considered approach which took due account of important methodological differences between studies, including study design and type of effect estimate (whereas previous meta-analyses have been less careful in their meta-analytical approach); and we assessed individual study methodological quality, paying particular attention to the potential for confounding in many studies, which previous reviews have rarely highlighted.

Our review has limitations, mainly due to shortcomings of individual studies and the challenges of meaningfully synthesizing such heterogeneous studies. Whilst selection and information bias were minimized through the use of routinely collected health datasets in almost all studies, confounding by indication is a major limitation, since having a major mental disorders is itself associated with increased risk of stroke and MI [5]. The excess risk of cardiovascular disease in people with major mental illness is thought to be multifactorial and to include poor lifestyle behaviours (such as smoking, alcohol misuse, physical inactivity and obesity), increased risk of diabetes, shared genetic factors and direct physiological effects of the mental illness [66,67,68]. Whilst APs might themselves increase the risk of cardiovascular disease, particularly through weight gain, it is interesting that metabolic dysregulation for example, has been observed in AP naïve people with major mental illness [69]. In the studies identified in our review confounders such as lifestyle factors and socioeconomic status were rarely adequately adjusted for. The definition of stroke varied considerably across studies, with some also including non-stroke cerebrovascular disease ICD codes and others adopting a very narrow definition of stroke. Finally, studies were also heterogeneous in their definition of AP drug use and duration of follow-up varied widely, from just weeks to 13 years.

Further research is needed to address the gaps identified in this review. For ethical reasons, RCTs in this area are less feasible than observational studies, since, for some conditions, there are no suitable potential alternatives to AP drug treatment. However, where potential alternative treatments do exist, RCTs may be appropriate. In the dementia population for example, future studies could examine how the risk of stroke and MI amongst AP drug users compares to alternative treatment for behavioural and psychotic symptoms. For example, findings from a cohort study conducted by Finkel may suggest that AP drug use has a potentially lower risk of stroke compared to the use of benzodiazepines in dementia patients [70]. Where RCTs are not feasible, robust cohort and case-control studies should be performed. Such studies should be sufficiently large in order to be adequately powered to detect significant associations with high precision. Pooling of individual patient data across multiple settings would be one approach to enhance study power. Routinely collected national health data which contains information on mental health diagnosis and treatment, previous and subsequent cardiovascular events and confounding factors would be the ideal data source for future studies. This relies on the existence of robust mental health registries linked to other health data or linkage of primary and secondary care data as well as prescription data, which exists in relatively few settings. However, such studies would eliminate bias, facilitate control of confounders and establish temporality between AP drug use and cardiovascular outcome, particularly in case-control studies. Future studies should also distinguish between FGA and SGA users, given their differing pharmacological properties and potentially different risk of stroke and MI. Confounding by indication should be minimized and risk estimates should be determined based on indication for AP drug use. Studies should also assess the risk over multiple time periods in order to clarify whether and how the association may change with time. A minimum duration of AP use should also be clearly established (such as multiple prescriptions) especially if large registry databases are used to ascertain exposure.

In the absence of conclusive evidence that AP drug use increases risk of stroke or MI, we recommend that clinicians should carefully weigh the potential benefits and risks of AP drug use on an individual basis, and evaluate cardiovascular risk prior to AP drug initiation and during treatment. We urge clinical caution in the initiation of off-label AP drug use and encourage the exploration of alternative treatment routes.


In conclusion, there is some evidence that AP drug use is associated with an increased risk of stroke, but no clear evidence that it is associated with an increased risk of MI. The risk of stroke was less apparent in studies specifically reporting on a population with dementia. However, these conclusions are drawn with the caveat that there was substantial methodological and statistical heterogeneity between studies and we highly recommend further methodologically robust studies which control for confounding factors, account for confounding by indication and examine associations by AP drug type and exposure period. Whilst further research is undertaken, psychiatrists and other physicians prescribing antipsychotic medications to people with or without a major mental disorder should be vigilant in monitoring and improving their cardiovascular risk profile.

Availability of data and materials

Not applicable.





Confidence interval


First-generation antipsychotic


Hazard ratio


Incident rate ratio


Myocardial infarction


Odds ratio


Second-generation antipsychotic


  1. 1.

    Global Burden of Disease Study 2013 Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: a systematic analysis for the global burden of disease study 2013. Lancet. 2015;386(9995):743–800.

    Article  PubMed Central  Google Scholar 

  2. 2.

    Hennekens CH, Hennekens AR, Hollar D, Casey DE. Schizophrenia and increased risks of cardiovascular disease. Am Heart J. 2005;150(6):1115–21.

    CAS  Article  PubMed  Google Scholar 

  3. 3.

    Laursen TM, Munk-Olsen T, Vestergaard M. Life expectancy and cardiovascular mortality in persons with schizophrenia. Curr Opin Psychiatry. 2012;25(2):83–8.

    Article  PubMed  Google Scholar 

  4. 4.

    Osborn DPJ, Levy G, Nazareth I, Petersen I, Islam A, King MB. Relative risk of cardiovascular and cancer mortality in people with severe mental illness from the United Kingdom’s general practice research database. Arch Gen Psychiatry. 2007;64(2):242–9.

    Article  PubMed  Google Scholar 

  5. 5.

    Correll CU, Solmi M, Veronese N, Bortolato B, Rosson S, Santonastaso P, Thapa-Chhetri N, Fornaro M, Gallicchio D, Collantoni E, Pigato G, Favaro A, Monaco F, Kohler C, Vancampfort D, Ward PB, Gaughran F, Carvalho AF, Stubbs B. Prevalence, incidence and mortality from cardiovascular disease in patients with pooled and specific severe mental illness: a large-scale meta-analysis of 3,211,768 patients and 113,383,368 controls. World Psychiatry. 2017;16(2):163–80.

    Article  PubMed  PubMed Central  Google Scholar 

  6. 6.

    De Hert M, Cohen D, Bobes J, Cetkovich-Bakmas M, Leucht S, Ndetei DM, Newcomer JW, Uwake R, Asai I, Moller HJ, Gauman S, Detraux J, Correl C. Physical illness in patients with severe mental disorders. II. Barriers to care, monitoring and treatment guidelines, plus recommendations at the system and individual level. World Psychiatry. 2011;10(2):138–51.

    Article  Google Scholar 

  7. 7.

    Liu NH, Daumit GL, Dua T, Aquila R, Charlson F, Cuijpers P, Druss B, Dudek K, Freeman M, Fujii C, Gaebel W, Hegerl U, Levav I, Laursen TM, Ma H, Maj M, Elena Medina-Mora M, Nordentoft M, Prabhakaran D, Pratt K, Prince M, Rangaswamy T, Shiers D, Susser E, Thornicroft G, Wahlbeck K, Wassie AF, Whiteford H, Saxena S. Excess mortality in persons with severe mental disorders: a multilevel intervention framework and priorities for clinical practice, policy and research agendas. World Psychiatry. 2017;16(1):30–40.

    Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    De Hert M, Dekker JM, Wood D, Kahl KG, Holt RIG, Möller HJ. Cardiovascular disease and diabetes in people with severe mental illness position statement from the European psychiatric association (EPA), supported by the European Association for the Study of diabetes (EASD) and the European Society of Cardiology (ESC). Eur Psychiatry. 2009;24(6):412–24.

    Article  PubMed  Google Scholar 

  9. 9.

    Ringen PA, Engh JA, Birkenaes AB, Dieset I, Andreassen OA. Increased mortality in schizophrenia due to cardiovascular disease - a non-systematic review of epidemiology, possible causes and interventions. Front Psychiatry. 2014;5:137.

    Article  PubMed  PubMed Central  Google Scholar 

  10. 10.

    Herrmann N, Lanctôt KL. Do atypical antipsychotics cause stroke? CNS Drugs. 2005;19(2):91–103.

    Article  PubMed  Google Scholar 

  11. 11.

    Alexander GC, Gallagher SA, Mascola A, Moloney RM, Stafford RS. Increasing off-label use of antipsychotic medications in the United States, 1995-2008. Pharmacoepidemiol Drug Saf. 2011;20(2):177–84.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Cioltan H, Alshehri S, Howe C, Lee J, Fain M, Eng H, Schachter K, Mohler J. Variation in use of antipsychotic medications in nursing homes in the United States: a systematic review. BMC Geriatr. 2017;17(1):32.

    Article  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Barnes TRE, Banerjee S, Collins N, Treloar A, McIntyre SM, Paton C. Antipsychotics in dementia: prevalence and quality of antipsychotic drug prescribing in UK mental health services. Br J Psychiatry. 2012;201(3):221–6.

    Article  PubMed  Google Scholar 

  14. 14.

    Schneider S, Dagerman S, Insel P. Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of randomized placebo-controlled trials. JAMA. 2005;294(15):1934–43.

    CAS  Article  PubMed  Google Scholar 

  15. 15.

    Kuehn B. FDA warns antipsychotic drugs may be risky for elderly. JAMA. 2005;293(20):2462.

    CAS  Article  Google Scholar 

  16. 16.

    Wooltorton E. Risperidone (Risperdal): increased rate of cerebrovascular events in dementia trials. CMAJ. 2002;167(11):1269–70.

    PubMed  PubMed Central  Google Scholar 

  17. 17.

    Wooltorton E. Olanzapine (Zyprexa): increased incidence of cerebrovascular events in dementia trials. CMAJ. 2004;170(9):1395.

    Article  PubMed  PubMed Central  Google Scholar 

  18. 18.

    Corbett A, Burns A, Ballard C. Don’t use antipsychotics routinely to treat agitation and aggression in people with dementia. BMJ. 2014;349:g6420.

    CAS  Article  PubMed  Google Scholar 

  19. 19.

    Hsu W-T, Esmaily-Fard A, Lai C-C, Zala D, Lee S-H, Chang S-S, Lee C-C. Antipsychotics and the risk of cerebrovascular accident: a systematic review and meta-analysis of observational studies. J Am Med Dir Assoc. 2017;18(8):692–9.

    Article  PubMed  Google Scholar 

  20. 20.

    Sacchetti E, Turrina C, Valsecchi P. Cerebrovascular accidents in elderly people treated with antipsychotic drugs: a systematic review. Drug Saf. 2010;33(4):273–88.

    CAS  Article  PubMed  Google Scholar 

  21. 21.

    Brauer R, Douglas I, Smeeth L. The association between antipsychotic agents and the risk of myocardial infarction: a systematic review. Br J Clin Pharmacol. 2011;72(6):871–8.

    Article  PubMed  PubMed Central  Google Scholar 

  22. 22.

    Yu Z -h, Jiang H-Y, Shao L, -y ZY, -y SH, Ruan B. Use of antipsychotics and risk of myocardial infarction: a systematic review and meta-analysis. Br J Clin Pharmacol. 2016;82(3):624–32.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  23. 23.

    Huang KL, Fang CJ, Hsu CC, Wu SI, Juang JJM, Stewart R. Myocardial infarction risk and antipsychotics use revisited: a meta-analysis of 10 observational studies. J Psychopharmacol. 2017;31(12):1544–55.

    Article  PubMed  Google Scholar 

  24. 24.

    Joseph KS, Mehrabadi A, Lisonkova S. Confounding by indication and related concepts. Curr Epidemiol Rep. 2014;1(1):1–8.

    Article  Google Scholar 

  25. 25.

    Sign. Critical appraisal: Notes and checklists [Internet]. Sign. 2014. Available from:

  26. 26.

    Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA, Thacker SB. Meta-analysis of observational studies in epidemiology. JAMA. 2000;283(15):2008–12.

    CAS  Article  PubMed  Google Scholar 

  27. 27.

    Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 2009;6(7):e1000097.

    Article  PubMed  PubMed Central  Google Scholar 

  28. 28.

    Barnett MJ, Wehring H, Perry PJ. Comparison of risk of cerebrovascular events in an elderly VA population with dementia between antipsychotic and nonantipsychotic users. J Clin Psychopharmacol. 2007;27(6):595–601.

    CAS  Article  PubMed  Google Scholar 

  29. 29.

    Chan MC, Chong CSY, Wu AYK, Wong KC, Dunn ELW, Tang OW, Chan WF. Antipsychotics and risk of cerebrovascular events in treatment of behavioural and psychological symptoms of dementia in Hong Kong: a hospital-based, retrospective, cohort study. Int J Geriatr Psychiatry. 2010;25(4):362–70.

    Article  PubMed  Google Scholar 

  30. 30.

    Chen Y, Guo JJ, Li H, Wulsin L, Patel NC. Risk of cerebrovascular events associated with antidepressant use in patients with depression: a population-based, nested case-control study. Ann Pharmacother. 2008;42(2):177–84.

    Article  PubMed  Google Scholar 

  31. 31.

    Correll CU, Joffe BI, Rosen LM, Sullivan TB, Joffe RT. Cardiovascular and cerebrovascular risk factors and events associated with second-generation antipsychotic compared to antidepressant use in a non-elderly adult sample: results from a claims-based inception cohort study. World Psychiatry. 2015;14(1):56–63.

    Article  PubMed  PubMed Central  Google Scholar 

  32. 32.

    Douglas IJ, Smeeth L. Exposure to antipsychotics and risk of stroke: self controlled case series study. BMJ. 2008;337:a1227.

    Article  PubMed  PubMed Central  Google Scholar 

  33. 33.

    Franchi C, Sequi M, Tettamanti M, Bonometti F, Nobili A, Fortino I, Bortolotti A, Merlino L, Pasina L, Djade CD, Marengoni A. Antipsychotics prescription and cerebrovascular events in Italian older persons. J Clin Psychopharmacol. 2013;33(4):542–5.

    CAS  Article  PubMed  Google Scholar 

  34. 34.

    Hsieh P-H, Hsiao F-Y, Gau SS-F, Gau C-S. Use of antipsychotics and risk of cerebrovascular events in schizophrenic patients: a nested case-control study. J Clin Psychopharmacol. 2013;33(3):299–305.

    CAS  Article  PubMed  Google Scholar 

  35. 35.

    Kleijer B, van Marum R, Egberts A, Jansen P, Knol W, Heerdink E. Risk of cerebrovascular events in elderly users of antipsychotics. J Psychopharmacol. 2009;23(8):909–14.

    CAS  Article  PubMed  Google Scholar 

  36. 36.

    Lan CC, Liu CC, Lin CH, Lan TY, McInnis MG, Chan CH, Lan TH. A reduced risk of stroke with lithium exposure in bipolar disorder: a population-based retrospective cohort study. Bipolar Disord. 2015;17(7):705–14.

    CAS  Article  PubMed  Google Scholar 

  37. 37.

    Laredo L, Vargas E, Blasco AJ, Aguilar MD, Moreno A, Portolés A. Risk of cerebrovascular accident associated with use of antipsychotics: population-based case-control study. J Am Geriatr Soc. 2011;59(7):1182–7.

    Article  PubMed  Google Scholar 

  38. 38.

    Liebetrau M, Steen B, Skoog I. Depression as a risk factor for the incidence of first-ever stroke in 85-year-olds. Stroke. 2008;39(7):1960–5.

    Article  PubMed  Google Scholar 

  39. 39.

    Liperoti R, Gambassi G, Lapane KL, Chiang C, Pedone C, Mor V, Bernabei R. Cerebrovascular events among elderly nursing home patients treated with conventional or atypical antipsychotics. J Clin Psychiatry. 2005;66(9):1090–6.

    Article  PubMed  Google Scholar 

  40. 40.

    Liu M-E, Tsai S-J, Chang W-C, Hsu C-H, Lu T, Hung K-S, Chiu W-T, Chang W-P. Population-based 5-year follow-up study in Taiwan of dementia and risk of stroke. PLoS One. 2013;8(4):e61771.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. 41.

    Percudani M, Barbui C, Fortino I, Tansella M, Petrovich L. Second-generation antipsychotics and risk of cerebrovascular accidents in the elderly. J Clin Psychopharmacol. 2005;25(5):468–70.

    Article  PubMed  Google Scholar 

  42. 42.

    Pratt NL, Roughead EE, Ramsay E, Salter A, Ryan P. Risk of hospitalization for stroke associated with antipsychotic use in the elderly a self-controlled case series. Drugs Aging. 2010;27(11):885–93.

    Article  PubMed  Google Scholar 

  43. 43.

    Sacchetti E, Trifirò G, Caputi A, Turrina C, Spina E, Cricelli C, Brignoli O, Sessa E, Mazzaglia G. Risk of stroke with typical and atypical anti-psychotics: a retrospective cohort study including unexposed subjects. J Psychopharmacol. 2008;22(1):39–46.

    Article  PubMed  Google Scholar 

  44. 44.

    Shin J-Y, Choi N-K, Jung S-Y, Lee J, Kwon JS, Park B-J. Risk of ischemic stroke with the use of risperidone, quetiapine and olanzapine in elderly patients: a population-based, case-crossover study. J Psychopharmacol. 2013;27(7):638–44.

    CAS  Article  PubMed  Google Scholar 

  45. 45.

    Wang S, Linkletter C, Dore D, Mor V, Buka S, Maclure M. Age, antipsychotics, and the risk of ischemic stroke in the veterans health administration. Stroke. 2012;43(1):28–31.

    CAS  Article  PubMed  Google Scholar 

  46. 46.

    Wu CS, Wang SC, Gau SSF, Tsai HJ, Cheng YC. Association of stroke with the receptor-binding profiles of antipsychotics - a case-crossover study. Biol Psychiatry. 2013;73(5):414–21.

    CAS  Article  PubMed  Google Scholar 

  47. 47.

    Brauer R, Smeeth L, Anaya-Izquierdo K, Timmis A, Denaxas SC, Farrington CP, Whitaker H, Hemingway H, Douglas I. Antipsychotic drugs and risks of myocardial infarction: a self-controlled case series study. Eur Heart J. 2015;36(16):984–92.

    CAS  Article  PubMed  Google Scholar 

  48. 48.

    Enger C, Weatherby L, Reynolds RF, Glasser DB, Walker AM. Serious cardiovascular events and mortality among patients with schizophrenia. J Nerv Ment Dis. 2004;192(1):19–27.

    Article  PubMed  Google Scholar 

  49. 49.

    Hwang YJ, Dixon SN, Reiss JP, Wald R, Parikh CR, Gandhi S, Shariff SZ, Pannu N, Nash DM, Rehman F, Garg AX. Atypical antipsychotic drugs and the risk for acute kidney injury and other adverse outcomes in older adults: a population-based cohort study. Ann Intern Med. 2014;161(4):242–8.

    Article  PubMed  Google Scholar 

  50. 50.

    Lin ST, Chen CC, Tsang HY, Lee CS, Yang P, Da CK, Li DJ, Wang CJ, Hsieh YC, Yang WC. Association between antipsychotic use and risk of acute myocardial infarction: a nationwide case-crossover study. Circulation. 2014;130(3):235–43.

    CAS  Article  PubMed  Google Scholar 

  51. 51.

    Nakagawa S, Pedersen L, Olsen ML, Mortensen PB, Sørensen HT, Johnsen SP. Antipsychotics and risk of first-time hospitalization for myocardial infarction: a population-based case–control study. J Intern Med. 2006;260(5):451–8.

    CAS  Article  PubMed  Google Scholar 

  52. 52.

    Pariente A, Fourrier-Réglat A, Ducruet T, Farrington P, Béland S, Dartigues J, Moore N, Moride Y. Antipsychotic use and myocardial infarction in older patients with treated dementia. Arch Intern Med. 2012;172(8):648–53.

    Article  PubMed  Google Scholar 

  53. 53.

    Penttinen J, Valonen P. Use of psychotropic drugs and risk of myocardial infarction: a case-control study in Finnish farmers. Int J Epidemiol. 1996;25(4):760–2.

    CAS  Article  PubMed  Google Scholar 

  54. 54.

    Pratt LA, Ford DE, Crum RM, Armenian HK, Gallo JJ, Eaton WW. Depression, psychotropic medication, and risk of myocardial infarction. Prospective data from the Baltimore ECA follow-up. Circulation. 1996;94(12):3123–9.

    CAS  Article  PubMed  Google Scholar 

  55. 55.

    Wang S, Mor V, Maclure M, Dore D, Linkletter C, Buka S. Treatment with antipsychotics and risk of myocardial infarction: a case –case time control study. Am J Epidemiol. 2011;173:S130.

    Google Scholar 

  56. 56.

    Sacchetti E, Turrina C, Cesana B, Mazzaglia G. Timing of stroke in elderly people exposed to typical and atypical antipsychotics: a replication cohort study after the paper of Kleijer, et al. J Psychopharmacol. 2010;24(7):1131–2.

    CAS  Article  PubMed  Google Scholar 

  57. 57.

    Rao A, Suliman A, Story G, Vuik S, Aylin P, Darzi A. Meta-analysis of population-based studies comparing risk of cerebrovascular accident associated with first- and second-generation antipsychotic prescribing in dementia. Int J Methods Psychiatr Res. 2016;25(4):289–98.

    Article  Google Scholar 

  58. 58.

    Haddad PM, Sharma SG. Adverse effects of atypical antipsychotics : differential risk and clinical implications. CNS Drugs. 2007;21(11):911–36.

    CAS  Article  Google Scholar 

  59. 59.

    Kroeze WK, Hufeisen SJ, Popadak BA, Renock SM, Steinberg S, Ernsberger P, Jayathilake K, Meltzer HY, Roth BL. H1-histamine receptor affinity predicts short-term weight gain for typical and atypical antipsychotic drugs. Neuropsychopharmacology. 2003;28(3):519–26.

    CAS  Article  PubMed  Google Scholar 

  60. 60.

    Reynolds GP, Kirk SL. Metabolic side effects of antipsychotic drug treatment – pharmacological mechanisms. Pharmacol Ther. 2010;125(1):169–79.

    CAS  Article  PubMed  Google Scholar 

  61. 61.

    Kim SF, Huang AS, Snowman AM, Teuscher C, Snyder SH. Antipsychotic drug-induced weight gain mediated by histamine H1 receptor-linked activation of hypothalamic AMP-kinase. Proc Natl Acad Sci. 2007;104(9):3456–9.

    CAS  Article  PubMed  Google Scholar 

  62. 62.

    Riordan J, Antonini P, Murphy F. Atypical antipsychotics and metabolic syndrome in patients with schizophrenia: risk factors, monitoring, and healthcare implications. Am Heal drug benefits. 2011;4(5):292–302.

    Google Scholar 

  63. 63.

    Barbui C, Conti V, Cipriani A. Antipsychotic drug exposure and risk of venous thromboembolism: a systematic review and meta-analysis of observational studies. Drug Saf. 2014;37(2):79–90.

    CAS  Article  PubMed  Google Scholar 

  64. 64.

    Hagg S, Jonsson AK, Spigset O. Risk of venous thromboembolism due to antipsychotic drug therapy. Expert Opin Drug Saf. 2009;8(5):537–47.

    CAS  Article  PubMed  Google Scholar 

  65. 65.

    Burghardt KJ, Seyoum B, Mallisho A, Burghardt PR, Kowluru RA, Yi Z. Atypical antipsychotics , insulin resistance and weight; a meta-analysis of healthy volunteer studies. Prog Neuro-Psychopharmacol Biol Psychiatry. 2018;83:55–63.

    CAS  Article  Google Scholar 

  66. 66.

    Holt RIG, Mitchell AJ. Diabetes mellitus and severe mental illness: mechanisms and clinical implications. Nat Rev Endocrinol. 2015;11(2):79.

    Article  PubMed  Google Scholar 

  67. 67.

    De Hert M, Correll CU, Bobes J, Cetkovich-Bakmas M, Cohen DAN, Asai I, et al. Physical illness in patients with severe mental disorders. I. Prevalence, impact of medications and disparities in health care. World Psychiatry. 2011;10(1):52–77.

    Article  Google Scholar 

  68. 68.

    Amare AT, Schubert KO, Klingler-Hoffmann M, Cohen-Woods S, Baune BT. The genetic overlap between mood disorders and cardiometabolic diseases: a systematic review of genome wide and candidate gene studies. Transl Psychiatry. 2017;7(1):e1007.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  69. 69.

    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):261–9.

    Article  PubMed  PubMed Central  Google Scholar 

  70. 70.

    Finkel S, Kozma C, Long S, Greenspan A, Mahmoud R, Baser O, Engelhart L. Risperidone treatment in elderly patients with dementia: relative risk of cerebrovascular events versus other antipsychotics. Int Psychogeriatrics. 2005;17(4):617–29.

    Article  Google Scholar 

Download references


Not applicable.


This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author information




SZ and CAJ designed the study. SZ, CHK and NK performed the electronic search, screened articles and extracted relevant data. All authors were involved in the interpretation of the review findings. SZ and CAJ constructed the draft manuscript. All authors edited and agreed the final manuscript for submission.

Corresponding author

Correspondence to Caroline A. Jackson.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests


Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Additional file

Additional file 1:

Supplemental Material. (DOCX 61 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Zivkovic, S., Koh, C.H., Kaza, N. et al. Antipsychotic drug use and risk of stroke and myocardial infarction: a systematic review and meta-analysis. BMC Psychiatry 19, 189 (2019).

Download citation


  • Antipsychotics
  • Myocardial infarction
  • Stroke
  • Dementia
  • Meta-analyses