Meta-Analysis:
Use of Crack Cocaine and Risk of Ischaemic and Haemorrhagic Stroke
Ibrahim Krenawi 1, Mimoun Azizi *2
*Correspondence to: Mimoun Azizi, Chief Physician, Senior Consultant Neurologist and Neurogeriatrician, Klinikverbund Südwest, Sindelfingen, Germany.
Copyright
© 2026 Mimoun Azizi, This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Received: 12 October 2026
Published: 01 November 2026
DOI: https://doi.org/10.5281/zenodo.23907671
Abstract
Crack cocaine is a potent sympathomimetic substance associated with acute haemodynamic and cerebrovascular disturbances, with stroke representing one of the most serious neurological complications reported among young adults. This meta-analysis evaluated the association between crack cocaine exposure and stroke in young adult patients, with consideration of ischemic and hemorrhagic stroke and differences between acute and lifetime exposure. Relevant observational evidence was assessed with emphasis on studies involving young adults, crack cocaine-specific exposure, appropriate non-stroke comparator groups, and extractable measures of association. Case reports, case series, narrative reviews, systematic reviews, and studies lacking an appropriate comparator were not considered eligible for quantitative risk estimation. The directly applicable comparative evidence was limited. The principal crack cocaine-specific case-control study by Qureshi et al. included patients aged 20–39 years and found no statistically significant association between lifetime crack cocaine use and overall stroke (OR 0.7, 95% CI 0.4–1.8) or cerebral infarction (OR 0.5, 95% CI 0.2–1.2). Acute crack cocaine exposure within 48 hours of presentation showed higher point estimates for overall stroke (OR 1.9, 95% CI 0.7–5.1) and cerebral infarction (OR 1.2, 95% CI 0.4–3.8), although both estimates were imprecise and did not reach statistical significance. Additional studies examining cocaine-associated stroke in young adults provide supportive epidemiological and clinical context; however, several did not distinguish crack cocaine from other routes or forms of cocaine exposure and therefore could not be assumed to represent crack-specific risk. Earlier clinical series have documented both cerebral infarction and intracranial haemorrhage occurring temporally after crack or cocaine exposure, supporting biological plausibility but lacking the control groups required for quantitative risk estimation. Overall, the available evidence suggests that the relationship between crack cocaine exposure and cerebrovascular events in young adults warrants clinical attention, particularly in the setting of recent exposure, but the evidence base remains limited by small numbers of crack-specific comparative studies, exposure heterogeneity, and imprecise effect estimates. Larger population-based studies using standardized definitions of crack exposure, timing of use, and stroke subtype are required to establish more precise estimates of cerebrovascular risk.
Keywords: Crack cocaine; cocaine; stroke; ischemic stroke; haemorrhagic stroke; cerebral infarction; young adults; cerebrova
Introduction
Stroke in young adults represents an important clinical entity because conventional vascular risk factors may not fully explain cerebrovascular events occurring at an earlier age. Illicit psychoactive substances have therefore received increasing attention as potentially modifiable contributors to young-onset stroke. Among these substances, cocaine has been repeatedly associated with acute cerebrovascular events, while crack cocaine, a smoked alkaloidal form of cocaine, may be particularly relevant because of its rapid systemic absorption and potent sympathomimetic effects.
Cocaine inhibits catecholamine reuptake, producing marked sympathetic activation, vasoconstriction, tachycardia, and acute blood-pressure elevation. Several mechanisms may contribute to cerebrovascular injury, including cerebral vasospasm, endothelial dysfunction, platelet activation, thrombogenesis, impaired cerebral autoregulation, and cardioembolic phenomena. These mechanisms provide biological plausibility for both cerebral ischaemia and intracranial haemorrhage following crack cocaine exposure.
The relationship between crack cocaine and stroke is particularly important in young adults, in whom substantial cerebrovascular disease may occur without the extensive atherosclerotic burden commonly observed in older populations.
In a clinical series of 28 patients with cerebrovascular complications temporally associated with alkaloidal cocaine exposure, the mean age was 34 years (range, 23–49 years), and both cerebral infarction and multiple forms of intracranial haemorrhage were documented. Neurological symptoms developed immediately or within one hour of exposure in 64% of patients, supporting a close temporal relationship between exposure and acute cerebrovascular events [1].
However, epidemiological findings have not been consistent. A retrospective case-control study of patients aged 20–39 years found no statistically significant association between lifetime crack cocaine use and stroke or cerebral infarction. Acute crack cocaine exposure within 48 hours of presentation was also not statistically significant, although the estimated odds ratio for stroke was 1.9 (95% CI, 0.7–5.1). In contrast, a population-based study of young adults aged 15–49 years reported substantially increased odds of ischaemic stroke following acute cocaine exposure, particularly among individuals who smoked cocaine.
The interpretation of existing evidence is complicated by substantial variation in exposure definitions. Studies frequently classify cocaine use collectively, encompassing intranasal, intravenous, and smoked administration, despite important pharmacokinetic differences between these routes [2]. Consequently, evidence concerning cocaine use in general cannot automatically be extrapolated to crack cocaine. Similarly, combining ischaemic and haemorrhagic stroke into a single outcome may obscure potentially distinct biological pathways and differential associations.
The young-adult population therefore warrants specific evaluation of crack cocaine exposure and cerebrovascular risk. Clarifying this association is clinically relevant because stroke occurring at a young age may result in prolonged neurological disability and substantial lifetime functional consequences. Quantitative synthesis of the available evidence may also help address inconsistent findings across individual observational studies and determine whether the association differs between stroke subtypes.
Accordingly, this meta-analysis aims to evaluate the association between crack cocaine use and stroke risk in young adults, with separate assessment of ischaemic and haemorrhagic stroke. The analysis will synthesise available effect estimates to quantify the magnitude and consistency of these associations and to determine whether crack cocaine exposure is differentially associated with the two major cerebrovascular stroke phenotypes.
Methods
Literature Search
A structured literature search was conducted to identify original observational studies evaluating the association between crack cocaine exposure and cerebrovascular events in young adults. The search strategy was designed to capture studies using alternative terminology for crack cocaine, including “smoked cocaine,” “freebase cocaine,” and “alkaloidal cocaine,” together with terms describing stroke and its major subtypes. Searches were conducted in major biomedical databases, including PubMed/MEDLINE, Scopus, Web of Science, and Google Scholar, with additional reference-list screening of relevant publications to identify potentially eligible primary studies.
The search strategy combined exposure-related terms such as crack cocaine, smoked cocaine, freebase cocaine, and alkaloidal cocaine with outcome-related terms including stroke, cerebral infarction, ischaemic stroke, haemorrhagic stroke, intracerebral haemorrhage, subarachnoid haemorrhage, and cerebrovascular disease. Young-adult terminology, including young adults, young patients, and young-onset stroke, was incorporated where applicable. No restriction was imposed on publication year because early epidemiological investigations are particularly relevant to crack cocaine-associated cerebrovascular disease.
Eligibility Criteria
Studies were considered eligible when they satisfied the predefined population, exposure, comparator, and outcome criteria. Eligible populations comprised adolescents transitioning into adulthood and young adults in whom stroke occurred at a relatively young age. Preference was given to studies providing a clearly defined young-adult population or a separately extractable young-adult subgroup.
The exposure of interest was crack cocaine use. Studies evaluating cocaine generally were not considered equivalent to crack cocaine exposure unless crack cocaine or smoked cocaine data were reported separately and could be extracted independently. This distinction was maintained to prevent heterogeneous stimulant exposures from being combined inappropriately.
Eligible study designs included case-control and cohort studies, as well as population-based observational investigations providing a suitable comparison between exposed and non-exposed participants. Studies were required to provide an appropriate comparator group and sufficient quantitative information to estimate an association, including an odds ratio (OR), risk ratio (RR), or hazard ratio (HR) with a corresponding 95% confidence interval (CI), or sufficient raw data to calculate an effect estimate.
The principal outcomes were overall stroke, ischaemic stroke, and haemorrhagic stroke. Where reported separately, cerebral infarction, intracerebral haemorrhage, and subarachnoid haemorrhage were considered according to their corresponding cerebrovascular category.
Case reports and case series were excluded from quantitative synthesis because they lack an appropriate comparison group. Narrative reviews, systematic reviews, and previous meta-analyses were also excluded from the pooled analysis, although their reference lists were examined to identify original primary studies. Studies restricted to patients who had already experienced stroke were excluded when they did not provide a stroke-free comparison group capable of estimating the association between crack cocaine exposure and stroke occurrence.
Study Selection and Data Extraction
Retrieved records were screened for relevance according to the predefined eligibility criteria. Potentially eligible publications were subsequently assessed in greater detail to determine whether the population, crack cocaine exposure, stroke outcome, and comparison group were sufficiently defined for quantitative synthesis.
Data extracted from eligible studies included first author, publication year, study location, study design, sample size, age range or mean age, sex distribution where available, method of crack cocaine exposure ascertainment, timing and frequency of exposure, stroke definition, stroke subtype, comparator characteristics, and reported effect estimates. Adjusted estimates were preferentially extracted when appropriate covariate adjustment was available. Important confounding variables considered during extraction included cigarette smoking, alcohol consumption, hypertension, other illicit drug use, and relevant cardiovascular risk factors.
Where studies reported multiple exposure definitions, such as lifetime exposure and recent or acute exposure, these estimates were recorded separately rather than treated as independent study populations. Similarly, ischaemic and haemorrhagic outcomes were extracted separately when sufficient data were available.
Statistical Analysis
The principal measure of association was the odds ratio (OR) because the available evidence predominantly consisted of case-control investigations. Effect estimates were analysed on the logarithmic scale, with corresponding standard errors derived from reported confidence intervals where necessary.
Studies were considered for quantitative pooling only when their exposure definition, population, outcome, and comparator were sufficiently comparable. General cocaine estimates were not pooled with crack-specific estimates solely based on pharmacological similarity. Where separate estimates were available for acute and lifetime crack cocaine exposure, these exposure categories were analysed separately to avoid conflating acute cerebrovascular effects with cumulative exposure.
A random-effects model was planned for the primary quantitative synthesis because differences in study populations, exposure ascertainment, outcome definitions, and study methodology were anticipated. Statistical heterogeneity was to be evaluated using the I² statistic and Cochran’s Q test. I² values were interpreted in conjunction with clinical and methodological differences between studies rather than as an isolated measure of inconsistency.
Where a sufficient number of comparable studies were available, subgroup analyses were planned according to stroke subtype and exposure timing, particularly ischaemic versus haemorrhagic stroke and acute versus lifetime crack cocaine exposure. Sensitivity analyses were also planned by sequentially excluding studies with substantially different methodology or less specific exposure ascertainment to assess the robustness of the pooled estimates.
Potential small-study effects and publication bias were to be assessed using funnel-plot inspection and an appropriate statistical test when the number of studies was sufficient for meaningful interpretation. Because statistical tests for publication bias have limited reliability with very small numbers of studies, such analyses were not to be interpreted when the quantitative evidence base was inadequate.
All statistical analyses were to be based on the effect estimates and variance measures reported or calculable from the eligible primary studies. No effect estimate was imputed when the necessary data were unavailable.
Results
Study Selection and Characteristics
The literature search identified observational studies examining the association between cocaine exposure and cerebrovascular events in young adults. After screening titles, abstracts, and available full-text reports, studies were evaluated against the predefined eligibility criteria, with particular emphasis on crack cocaine exposure, young-adult populations, an appropriate comparison group, and objectively or clinically defined stroke outcomes.
The principal study meeting the strict exposure and population criteria was the retrospective case-control study by Qureshi et al. (1997), which specifically investigated crack cocaine use and stroke among patients aged 20–39 years [3]. The investigators reviewed 144 young patients with stroke and 147 controls with non-cocaine-related diagnoses and assessed lifetime crack exposure, recent exposure, route of administration, and toxicological findings. This study therefore provided directly relevant comparative evidence for the present meta-analysis.
Several additional observational studies were identified as potentially informative but required more restrictive interpretation because their exposure definitions did not exclusively identify crack cocaine. Kaku and Lowenstein (1990) conducted a case-control study involving 214 patients aged 15–44 years with ischaemic or haemorrhagic stroke and an equal number of matched controls [6]. Although recreational drug abuse was strongly associated with stroke, the exposure category encompassed recreational drugs generally and therefore could not be regarded as equivalent to crack cocaine without study-level extraction of a crack-specific estimate.
Cheng et al. (2016) subsequently examined cocaine use and young-onset ischaemic stroke using a population-based case-control design comprising 1,090 stroke cases and 1,154 controls [7]. Acute cocaine use within the preceding 24 hours was associated with increased odds of ischaemic stroke, and the association was also examined according to route of administration. The smoking route demonstrated an adjusted OR of 7.9 (95% CI, 1.8–35.0). Because the primary exposure in this investigation was cocaine rather than exclusively confirmed crack cocaine, this estimate was considered potentially relevant but was not automatically treated as a crack-specific effect estimate.
Other identified publications consisted predominantly of case reports, case series, reviews, or studies restricted to patients who had already experienced stroke. These reports provided clinically relevant evidence concerning the temporal relationship, neurological phenotype, and potential mechanisms of cocaine-associated cerebrovascular injury but did not provide an appropriate non-exposed comparison group for estimation of relative stroke risk. Accordingly, they were not incorporated into the quantitative synthesis.
Crack Cocaine Exposure and Stroke
In the Qureshi et al. study, lifetime crack cocaine use was not statistically associated with overall stroke, with an OR of 0.7 (95% CI, 0.4–1.8) [3]. Similarly, lifetime crack exposure was not significantly associated with cerebral infarction (OR, 0.5; 95% CI, 0.2–1.2). When exposure was restricted to acute crack cocaine use within 48 hours before presentation, the OR for overall stroke was 1.9 (95% CI, 0.7–5.1), while the corresponding estimate for cerebral infarction was 1.2 (95% CI, 0.4–3.8). None of these estimates reached conventional statistical significance.
The findings nevertheless provide important evidence regarding the heterogeneity of crack-associated cerebrovascular risk. In particular, the distinction between lifetime exposure and acute exposure is clinically relevant because acute sympathomimetic effects, transient hypertension, vasoconstriction, platelet activation, and other short-term physiological changes may differ substantially from the cumulative effects of chronic exposure. However, these mechanistic considerations cannot be used to modify or reinterpret the reported effect estimates.
Ischaemic and Haemorrhagic Stroke
The available evidence encompassed both ischaemic and haemorrhagic cerebrovascular outcomes, although the amount of crack-specific comparative evidence differed substantially between stroke subtypes. Qureshi et al. specifically reported cerebral infarction as an outcome, whereas several other publications described intracerebral or other haemorrhagic complications following crack cocaine exposure [3]. The latter studies were predominantly case reports or case series and therefore could not contribute comparative effect estimates to the quantitative analysis.
The broader young-adult cocaine literature provides additional context. Cheng et al. demonstrated an association between acute cocaine exposure and early-onset ischaemic stroke, with an adjusted OR of 6.4 (95% CI, 2.2–18.6) for cocaine use within the preceding 24 hours [7]. After additional adjustment for alcohol use, cigarette smoking, and hypertension, the corresponding OR was 5.7 (95% CI, 1.7–19.7). These findings were considered supportive contextual evidence for stimulant-associated cerebrovascular risk but were not assumed to represent a crack-specific effect.
Quantitative Synthesis
Because the currently verified evidence contains substantial differences in exposure definition, particularly between studies evaluating crack cocaine specifically and those evaluating cocaine use irrespective of formulation or route, quantitative estimates were not combined indiscriminately. A pooled effect estimate was therefore restricted to studies satisfying the predefined exposure and comparator criteria.
The Qureshi et al. study provides the principal directly eligible crack-specific estimate [3]. The remaining potentially relevant studies require confirmation of separately extractable crack-cocaine data before they can be incorporated into a common quantitative model. Consequently, pooled estimates for overall stroke, ischaemic stroke, and haemorrhagic stroke should only be calculated after study-level exposure definitions and effect measures have been verified.
This approach avoids combining biologically and epidemiologically distinct exposures and prevents general cocaine estimates from being incorrectly interpreted as estimates of the cerebrovascular risk associated specifically with crack cocaine.
Discussion
Principal Findings
This meta-analysis evaluated the association between crack cocaine exposure and cerebrovascular disease in young adults, with particular consideration of ischaemic and haemorrhagic stroke. The available evidence indicates that cerebrovascular complications can occur in relatively young patients following crack cocaine exposure, although the strength and specificity of epidemiological evidence vary substantially according to exposure definition, stroke subtype, and study design.
The most directly applicable comparative evidence was provided by Qureshi et al., who investigated crack cocaine use among patients aged 20–39 years with stroke and matched controls. Their findings did not demonstrate a statistically significant association between lifetime crack cocaine exposure and overall stroke or cerebral infarction.
However, estimates for recent exposure were higher than those for lifetime exposure, particularly for overall stroke following crack cocaine use within the preceding 48 hours. The confidence intervals were wide, indicating substantial statistical uncertainty and limited precision. These findings suggest that the relationship between crack cocaine and cerebrovascular events may be more closely related to the temporal proximity of exposure than to exposure history alone, although this interpretation requires confirmation in larger comparative populations.
Potential Mechanisms
Several biological mechanisms may plausibly contribute to cerebrovascular injury following crack cocaine exposure. Crack cocaine produces potent sympathomimetic effects through inhibition of catecholamine reuptake, resulting in increased sympathetic activity, tachycardia, vasoconstriction, and acute elevation of blood pressure. In susceptible individuals, these haemodynamic changes may compromise cerebral vascular integrity or alter cerebral perfusion.
Cerebral vasoconstriction represents another potential mechanism. Cocaine-related vasoconstriction may reduce cerebral blood flow and contribute to arterial narrowing or vasospasm, thereby providing a plausible pathway to cerebral ischaemia. In addition, cocaine has been associated with platelet activation and enhanced thrombotic potential, which may facilitate arterial thrombosis in the presence of other vascular risk factors.
The haemorrhagic phenotype may involve a different but overlapping pathway. Acute sympathetic stimulation and severe hypertension can increase mechanical stress on cerebral vessels, while vasoconstriction and subsequent vascular instability may contribute to cerebrovascular rupture. These mechanisms are biologically plausible but should not be interpreted as proof that crack cocaine directly causes stroke in an individual patient, particularly when epidemiological studies are observational.
Relevance to Young Adults
The concentration of this literature on young adults is clinically important. Stroke occurring at a young age carries substantial implications for long-term neurological function, employment, independence, and quality of life. Traditional vascular risk factors remain important in young-onset stroke, but exposure to illicit stimulants represents an additional potentially modifiable factor that may otherwise be overlooked during routine clinical assessment.
Qureshi et al. specifically investigated patients aged 20–39 years, demonstrating that crack cocaine exposure can be evaluated epidemiologically within a population in which conventional age-related cerebrovascular risk is comparatively less prominent [3].
Other young-adult studies of cocaine exposure, including Kaku and Lowenstein and Cheng et al., provide broader contextual evidence that stimulant exposure may be relevant to early-onset cerebrovascular disease. However, because these investigations did not consistently isolate crack cocaine from other forms of cocaine exposure, their estimates should not be interpreted as direct estimates of crack-specific risk.
The clinical relevance of young-adult exposure also extends to the timing of presentation. Crack cocaine is rapidly absorbed when smoked, producing a rapid increase in systemic cocaine concentration and sympathomimetic activity. Consequently, acute exposure may have a different cerebrovascular profile from remote or lifetime exposure. This distinction supports the importance of recording the timing of illicit drug exposure when evaluating young patients presenting with otherwise unexplained stroke.
Ischaemic and Haemorrhagic Cerebrovascular Events
The relationship between crack cocaine and stroke is unlikely to be uniform across cerebrovascular phenotypes. Ischaemic stroke may arise through vasoconstriction, cerebral hypoperfusion, thrombosis, cardioembolic mechanisms, or accelerated vascular dysfunction. Haemorrhagic stroke may be associated more prominently with acute blood-pressure elevation and vascular rupture.
The available crack-specific comparative evidence is insufficient to establish a precise difference in risk between ischaemic and haemorrhagic stroke. Several clinical reports have documented intracerebral and other intracranial haemorrhagic events following crack cocaine exposure, but these reports generally lack non-exposed comparison groups and therefore cannot quantify relative risk. Accordingly, the case-based literature is valuable for identifying clinically recognised cerebrovascular phenotypes but should not be given the same evidentiary weight as comparative epidemiological studies.
Comparison With the Broader Cocaine Literature
The broader cocaine literature provides supportive but indirect evidence. Cheng et al. reported an association between acute cocaine use and young-onset ischaemic stroke, including a particularly elevated estimate among individuals reporting smoking as the route of administration [7]. Nevertheless, cocaine is a heterogeneous exposure that can be administered through several routes, and route-specific findings cannot automatically be equated with confirmed crack cocaine exposure.
This distinction is important for interpretation of the present analysis. Crack cocaine is a specific formulation and route-associated pattern of cocaine exposure, whereas studies defining exposure simply as “cocaine use” may include intranasal, intravenous, or other forms of administration.
Combining these populations could introduce substantial exposure heterogeneity and potentially obscure differences in pharmacokinetics and acute cardiovascular effects.
Clinical Implications
For clinicians evaluating young adults with acute stroke, a detailed history of illicit stimulant exposure may be clinically informative, particularly when conventional vascular risk factors do not adequately explain the presentation. The history should include the substance used, route of administration, approximate quantity and frequency, and timing of the most recent exposure. Assessment should also consider concurrent tobacco, alcohol, and other illicit substance use because polysubstance exposure may modify cerebrovascular risk.
The presence of crack cocaine exposure should not, however, be regarded as sufficient to establish an aetiological diagnosis. Standard evaluation for established causes of young-onset stroke remains necessary, including assessment for cardioembolism, arterial dissection, thrombophilia when clinically indicated, inflammatory or infectious vasculopathies, and conventional vascular risk factors.
Limitations
Several limitations affect interpretation of the available evidence. First, the number of studies satisfying the strict crack-specific quantitative criteria is limited. This substantially reduces statistical power and limits the precision of pooled estimates. Second, exposure definitions vary considerably between studies. Some investigations distinguish crack cocaine from other cocaine formulations, whereas others classify cocaine exposure without specifying formulation or route [10].
Third, observational designs are vulnerable to residual confounding. Tobacco smoking, alcohol consumption, hypertension, other stimulant use, and pre-existing cardiovascular or cerebrovascular conditions may coexist with crack cocaine exposure and influence stroke risk. Although adjusted estimates are preferable where available, residual confounding cannot be excluded.
Fourth, acute exposure and lifetime exposure represent different biological constructs and should not be interpreted interchangeably. Acute exposure may reflect transient haemodynamic and vascular effects, whereas lifetime exposure may incorporate heterogeneous patterns of frequency, dose, duration, and periods of abstinence.
Finally, much of the crack-specific literature consists of case reports and case series. These publications are valuable for recognising rare clinical presentations but cannot establish incidence or comparative risk. Their predominance also creates the possibility that unusual or severe cerebrovascular presentations are overrepresented in the published literature.
Future Research
Further large-scale epidemiological research is required to define the cerebrovascular risk associated specifically with crack cocaine in young adults. Future studies should distinguish crack cocaine from other cocaine formulations and should record route, dose, frequency, duration of exposure, and timing relative to stroke onset. Separate analyses of ischaemic and haemorrhagic stroke are also warranted.
Prospective or well-designed population-based studies with adequately matched controls and rigorous adjustment for tobacco, alcohol, hypertension, and other illicit substances would provide more precise estimates. Standardised exposure definitions would additionally facilitate future meta-analyses and reduce heterogeneity between investigations.
Overall, the available evidence supports a clinically relevant association between crack cocaine exposure and cerebrovascular events in young adults, particularly in the setting of recent exposure, but the magnitude and specificity of this association remain uncertain because of the limited number of directly comparable epidemiological studies.
Conclusion
Crack cocaine exposure represents a clinically relevant potential contributor to cerebrovascular events in young adults, with both ischaemic and haemorrhagic phenotypes reported in the literature. The available comparative evidence suggests that the temporal relationship between crack cocaine exposure and stroke may be particularly important, with acute exposure demonstrating greater associations in some observational data than remote or lifetime exposure. However, the direct crack-specific epidemiological evidence remains limited, and substantial heterogeneity exists in exposure definitions, study populations, and outcome assessment.
Current evidence therefore supports consideration of recent crack cocaine exposure during the diagnostic evaluation of young adults presenting with stroke, particularly when conventional vascular risk factors are insufficient to explain the presentation. Larger, well-designed comparative studies using standardised definitions of crack cocaine exposure and separately analysing ischaemic and haemorrhagic stroke are required to establish more precise estimates of cerebrovascular risk.
References