Paroxysmal atrial fibrillation occurs often in cryptogenic ischaemic stroke. Final results from the SURPRISE study.

BACKGROUND AND PURPOSE: Atrial fibrillation (AF) increases the risk of stroke fourfold and is associated with a poor clinical outcome. Despite work-up in compliance with guidelines, up to one-third of patients have cryptogenic stroke (CS). The prevalence of asymptomatic paroxysmal atrial fibrillation (PAF) in CS remains unknown. The SURPRISE project aimed at determining this rate using long-term cardiac monitoring. METHODS: Patients with CS after protocolled work-up including electrocardiography (ECG) and telemetry were included after informed consent. An implantable loop recorder (ILR) was implanted subcutaneously. PAF was defined by events of atrial arrhythmia >2 min with a correlating one-lead ECG confirming the diagnosis. RESULTS: Eighty-five patients were monitored for a mean of 569 days (SD ±310). PAF was documented in 18 patients (20.7%) during the study period and detected by ILR in 14 patients (16.1%). In three patients PAF was detected by other methods before or after monitoring and was undiscovered due to device sensitivity in one case. The first event of PAF was documented at a mean of 109 days (SD ±48) after stroke onset. PAF was asymptomatic in all cases and occurred in episodes lasting predominantly between 1 and 4 h. Four recurrent strokes were observed, three in patients with PAF; all three patients were on oral anticoagulation (OAC). CONCLUSIONS: One in five patients with CS had PAF, which occurred at low burden and long after stroke. Future studies should determine the role of implantable cardiac monitors after stroke and determine the potential therapeutic benefit of OAC treatment of patients with PAF.

Stress and stereotypic behaviour in mink (Mustela vison): a focus on adrenocortical activity.

We examined whether female mink with low (LS) and high (HS) occurrence of stereotypic behaviour differ in their adrenocortical activity in baseline conditions or in response to immobilisation (Experiment 1), handling, adrenocorticotropic hormone (ACTH) challenge (Experiment 2) and excretion of circulating cortisol (Experiment 3). Faeces are the predominating excretory route of cortisol (83%), with peak concentrations after 4.2 h (urine: 3.4 h). Faecal cortisol metabolites (FCMs) reflected changes in relation to handling/ACTH challenge. In Experiment 1 (n = 162), HS mink had approximately 54% higher baseline level of FCM than LS mink (P < 0.001), with markedly elevated FCM on the days after an immobilisation stressor. In Experiment 2 (n = 48), LS and HS mink did not differ in adrenocortical activity after an ACTH challenge. However, HS mink reacted more in response to handling, evident in the FCM level 4-20 h after the handling (P = 0.001). In Experiment 3 (n = 16), the excretion of infused (3)H-cortisol did not differ between LS and HS mink. Stereotypic behaviour is concurrent with higher baseline concentrations of FCM, which cannot be explained by a greater adrenocortical reactivity or a different excretion of the circulating cortisol. Instead, we conclude that mink with a high level of stereotypic behaviour have a greater perception of stress, or increased sensitivity to stressors at the pituitary level.

Stereotypic behaviour in farm mink (Neovison vison) can be reduced by selection.

In this article we present the first estimation of genetic variation of stereotypic behaviour (SB). Stereotypic behaviour is defined as an unvarying behaviour without any specific goal or function repeated at least five times. All types of SB were included in the analyses. Altogether 1484 adult mink females of the brown colour type were assessed for behaviour traits: SB, active or inactive behaviour, staying in nest box. Genetic correlations were based on estimates of additive genetic (co)variances obtained from a trivariate linear animal model fitted to behaviour traits, body weight and litter size. The SB has an intermediate genetic variation (h(2) approximately 0.3) and divergent selection for SB confirmed that the frequency of SB can be altered by selection. The results confirmed the hypotheses of negative genetic correlation between SB and body weight and negative genetic correlation between body weight and litter size. The hypotheses of positive correlation between SB and active behaviour and SB and litter size were not confirmed. Consequences of selection for reduced SB can be changes in other behaviour traits, body weight and litter size, depending on the genetic correlation between the traits.

Synthesis and pharmacological evaluation of dimeric muscarinic acetylcholine receptor agonists.

Two dimeric analogs of the muscarinic acetylcholine receptor (mAChR) agonist phenylpropargyloxy-1,2,5-thiadiazole-quinuclidine (NNC 11-1314) were synthesized and pharmacologically evaluated. In radioligand binding assays on Chinese hamster ovary (CHO) cell membranes expressing the individual human M(1) to M(5) mAChR subtypes, both dimers [(3S)-1,4-bis-(3-[(3-azabicyclo[2.2.2]octanyl)-1,2,5-thiadiazol-4-yloxy]-1-propyn-1-yl)benzene,2-L-(+)-tartrate (NNC 11-1607) and (3S)-1,3-bis-(3-[(3-azabicyclo[2.2.2]octanyl)-1,2,5-thiadiazol-4-yloxy]-1-propyn-1-yl)benzene,2-L-(+)-tartrate (NNC 11-1585)] exhibited higher binding affinities than the monomeric NNC 11-1314. Only NNC 11-1585, however, displayed significant selectivity for the M(1) and M(2) mAChRs relative to the other subtypes. Although binding studies in rat brain homogenates supported the selectivity profile of NNC 11-1585 observed in the CHO membranes, rat heart membrane experiments revealed complex binding behavior for all three agonists that most likely reflected differences in species and host cell environment between the heart and CHO cells. Subsequent functional assays with phosphatidylinositol hydrolysis revealed that all three novel ligands were partial agonists relative to the full agonist oxotremorine-M at the CHO M(1), M(3), and M(5) mAChRs, with NNC 11-1607 displaying the highest functional selectivity. In the CHO M(2) and M(4) mAChR cells, agonist-mediated effects on forskolin-stimulated cAMP accumulation were characterized by bell-shaped concentration-response curves, with the exceptions of NNC 11-1607, which had no discernible effects at the M(2) mAChR, and NNC 11-1585, which could only inhibit cAMP accumulation at the M(4) mAChR. Thus, we identified NNC 11-1607 as a novel functionally selective M(1)/M(4) mAChR agonist. Our data suggest that dimerization of mAChR agonists is a viable approach in designing more potent and functionally selective agonists, as well as in providing novel tools with which to probe the nature of agonism at these receptors.

The muscarinic receptor agonist BuTAC, a novel potential antipsychotic, does not impair learning and memory in mouse passive avoidance.

(5R,6R)-6-(3-butylthio-1,2,5-thiadiazol-4-yl)-1-azabicyclo[3.2.1]octane) (BuTAC) is a novel, selective muscarinic receptor ligand with partial agonist mode of action at muscarinic M2 and M4 and antagonist mode of action at M1, M3 and M5 receptor subtypes in cloned cell lines. BuTAC exhibits functional dopamine receptor antagonism despite its lack of affinity for dopamine receptors, and parasympathomimetic effects in mice are produced only at doses well beyond the doses exhibiting the antipsychotic-like effects. In the present study we investigated the effects of BuTAC and the antipsychotic compounds clozapine, sertindole and olanzapine using one trial passive avoidance with mice as a model of learning and memory. Pharmacologically relevant doses of BuTAC and reference antipsychotics were identified, based on inhibition of apomorphine-induced climbing in mice as an assay measuring antidopaminergic potency. When ratios between the minimum effective dose (MED) for impairment of retention in passive avoidance and the MED for inhibition of apomorphine-induced climbing were calculated, BuTAC displayed a high ratio of >10, compared with clozapine (0.3), sertindole (3) and olanzapine (3). These data suggest that BuTAC is a potential novel antipsychotic which may have favourable effects on aspects of learning and memory.

Muscarinic receptor agonists decrease cocaine self-administration rates in drug-naive mice.

(5R,6R)-6-(3-Propylthio-1,2,5-thiadiazol-4-yl)-1-azabicyclo[ 3.2.1]octane (PTAC) is a selective muscarinic receptor ligand. The compound exhibits high affinity for central muscarinic receptors with partial agonist mode of action at muscarinic M(2) and M(4) and antagonist mode of action at muscarinic M(1), M(3) and M(5) receptor subtypes. The compound was earlier reported to exhibit functional dopamine receptor antagonism in rodents despite its lack of affinity for dopamine receptors. In the present study, we report that PTAC, as well as the muscarinic receptor agonists pilocarpine and oxotremorine, dose-dependently decreased rates of intravenous self-administration (fixed ratio 1) of the indirect dopamine receptor agonist cocaine in drug naive mice. Similar decreases in cocaine self-administration rates were obtained with the dopamine receptor antagonists olanzapine, clozapine, risperidone, fluphenazine and haloperidol. These findings suggest that compounds with partial muscarinic receptor agonist mode of action may be used in the medical treatment of cocaine abuse.

Xanomeline, an M(1)/M(4) preferring muscarinic cholinergic receptor agonist, produces antipsychotic-like activity in rats and mice.

Xanomeline is an M(1)/M(4) preferring muscarinic receptor agonist which decreased psychotic behaviors in patients with Alzheimer's disease, suggesting that xanomeline might be useful in the treatment of psychotic symptoms in patients with schizophrenia. The purpose of the present studies was, therefore, to compare the pharmacologic profile of xanomeline with that of known antipsychotic drugs. Electrophysiologically, xanomeline, after both acute and chronic administration in rats, inhibited A10 but not A9 dopamine cells in a manner which was blocked by the muscarinic receptor antagonist scopolamine. Behaviorally, xanomeline, like haloperidol, clozapine and olanzapine, blocked dopamine agonist-induced turning in unilateral 6-hydroxydopamine-lesioned rats, as well as apomorphine-induced climbing in mice. However, unlike the dopamine antagonist antipsychotic haloperidol, xanomeline did not produce catalepsy in rats. Moreover, xanomeline, like haloperidol, clozapine and olanzapine, inhibited conditioned avoidance responding in rats, an effect which also was blocked by scopolamine. The present results thus demonstrate that xanomeline has a pharmacologic profile which is similar to that of the atypical antipsychotics clozapine and olanzapine, thus indicating that xanomeline has the potential to be a novel approach in the treatment of psychotic symptoms in patients with schizophrenia.

Muscarinic receptor agonists, like dopamine receptor antagonist antipsychotics, inhibit conditioned avoidance response in rats.

The purpose of our studies was to determine the effects of muscarinic receptor agonists on conditioned avoidance responding in the rat. Rats were trained to avoid or escape an electric shock delivered to the feet in a discrete trial procedure. The muscarinic receptor agonists pilocarpine and [2-ethyl-8-methyl-2,8-diazaspiro(4. 5)decane-1,3-dione] hydrochloride (RS86) and the cholinesterase inhibitor physostigmine all decreased the percentage of avoidance responses at doses that produced less than approximately 30% response failures. Similar results were obtained with the antipsychotic drugs haloperidol, trifluoperazine, chlorpromazine, and clozapine. However, the benzodiazepine anxiolytic diazepam did not decrease avoidance responding up to doses that produced ataxia. On the other hand, oxotremorine and arecoline decreased avoidance responding only by producing response failures, whereas aceclidine produced intermediate changes. The muscarinic receptor antagonists scopolamine, trihexyphenidyl, and benztropine were without effect when administered alone but antagonized the decreases in avoidance responding produced by pilocarpine and RS86. Scopolamine had little effect on the decreases in avoidance responding produced by haloperidol. The newer muscarinic receptor partial agonists or agonist/antagonists [R-(Z)-(+)-alpha-(methoxyimino)-1-azabicyclo[2.2. 2]octane-3-acetonitrile] hydrochloride, talsaclidine, milameline, and xanomeline also produced dose-related decreases in avoidance responding. Our results demonstrate that muscarinic receptor agonists can decrease avoidance responding in a manner similar to dopamine-receptor antipsychotic drugs, suggesting that muscarinic receptor agonists may provide an alternative approach to the treatment of psychosis.

1-(1,2,5-Thiadiazol-4-yl)-4-azatricyclo[2.2.1.0(2,6)]heptanes as new potent muscarinic M1 agonists: structure-activity relationship for 3-aryl-2-propyn-1-yloxy and 3-aryl-2-propyn-1-ylthio derivatives.

Two new series of 1-(1,2,5-thiadiazol-4-yl)-4-azatricyclo[2.2.1.0(2, 6)]heptanes were synthesized and evaluated for their in vitro activity in cell lines transfected with either the human M1 or M2 receptor. 3-Phenyl-2-propyn-1-yloxy and -1-ylthio analogues substituted with halogen in the meta position showed high functional potency, efficacy, and selectivity toward the M1 receptor subtype. A quite unique functional M1 receptor selectivity was observed for compounds 8b, 8d, 8f, 9b, 9d, and 9f. Bioavailability studies in rats indicated an oral bioavailability of about 20-30%, with the N-oxide as the only detected metabolite.

Potential role of muscarinic receptors in schizophrenia.

The role of muscarinic receptors in schizophrenia was investigated using the muscarinic agonist PTAC. PTAC was highly selective for muscarinic receptors, was a partial agonist at muscarinic M2/M4 receptors and an antagonist at M1, M3 and M5 receptors. PTAC was highly active in animal models predictive of antipsychotic behavior including inhibition of conditioned avoidance responding in rats and blockade of apomorphine-induced climbing behavior in mice. d-Amphetamine-induced Fos expression in rat nucleus accumbens was inhibited by PTAC, thus directly demonstrating the ability of PTAC to modulate DA activity. In electrophysiological studies in rats, PTAC acutely inhibited the firing of A10 DA cells and after chronic administration decreased the number of spontaneously firing DA cells in the A10 brain area. However, PTAC did not appreciably alter the firing of A9 DA cells. Thus, PTAC appears to have novel antipsychotic-like activity and these data suggest that muscarinic compounds such as PTAC may represent a new class of antipsychotic agents.

Muscarinic agonists exhibit functional dopamine antagonism in unilaterally 6-OHDA lesioned rats.

(5R,6R) 6-(3-propylthio-1,2,5-thiadiazol-4-yl)-1-azabicyclo[3.2.1]oc tane (PTAC) is a selective muscarinic ligand with high affinity for central muscarinic receptors, agonist mode of action at the muscarinic M2 and M4 receptor subtypes and substantially less or no affinity for central dopamine receptors. In the present study PTAC, as well as the muscarinic agonists oxotremorine, RS86 and pilocarpine, inhibited dopamine D1 and D2 receptor agonist induced contralateral rotation in unilaterally 6-OHDA lesioned rats. The dose of SKF 38393 used to induce contralateral rotation also caused an intense Fos protein immunoreactivity in the rat dorsolateral striatum on the lesioned site which was inhibited by PTAC indicating that the inhibition of rotation by PTAC was not due to non-specific peripheral side effects.

Muscarinic agonists with antipsychotic-like activity: structure-activity relationships of 1,2,5-thiadiazole analogues with functional dopamine antagonist activity.

Muscarinic agonists were tested in two models indicative of clinical antipsychotic activity: conditioned avoidance responding (CAR) in rats and inhibition of apomorphine-induced climbing in mice. The standard muscarinic agonists oxotremorine and pilocarpine were both active in these tests but showed little separation between efficacy and cholinergic side effects. Structure-activity relationships of the alkylthio-1,2,5-thiadiazole azacyclic type muscarinic partial agonists are shown, revealing the exo-6-(3-propyl/butylthio-1,2, 5-thiadiazol-4-yl)-1-azabicyclo[3.2.1]octane analogues (4a,b and 9a, b) to be the most potent antipsychotic agents with large separation between efficacy and cholinergic side effects. The lack of enantiomeric selectivity suggests the pharmacophoric elements are in the mirror plane of the compounds. A model explaining the potency differences of closely related compounds is offered. The data suggest that muscarinic agonists act as functional dopamine antagonists and that they could become a novel treatment of psychotic patients.

Xanomeline compared to other muscarinic agents on stimulation of phosphoinositide hydrolysis in vivo and other cholinomimetic effects.

Activation of muscarinic m1 receptors which are coupled to the phosphoinositide (PI) second messenger transduction system is the initial objective of cholinergic replacement therapy in Alzheimer's disease. Thus, we evaluated the ability of the selective muscarinic receptor agonist (SMRA) xanomeline to stimulate in vivo phosphoinositide (PI) hydrolysis and compared it to a number of direct acting muscarinic agonists, two cholinesterase inhibitors and a putative m1 agonist/muscarinic m2 antagonist. Using a radiometric technique, it was determined that administration of xanomeline robustly stimulated in vivo PI hydrolysis and the effect was blocked by muscarinic antagonists, demonstrating mediation by muscarinic receptors. The non-selective muscarinic agonists pilocarpine, oxotremorine, RS-86, S-aceclidine, but not the less active isomer R-aceclidine, also effectively stimulated PI hydrolysis in mice. Amongst the putative m1 agonists, thiopilocarpine, hexylthio-TZTP as well as xanomeline effectively stimulated PI hydrolysis, but milameline, WAL 2014, SKB 202026 and PD 142505 did not significantly alter PI hydrolysis. Furthermore, WAL 2014 and SKB 202026 inhibited agonist-induced PI stimulation, suggesting that they act as antagonists at PI-coupled receptors in vivo. The cholinesterase inhibitors, tacrine and physostigmine, and the mixed muscarinic m1 agonist/m2 antagonist LU25-109 did not activate in vivo PI hydrolysis. Xanomeline, hexylthio-TZTP and thiopilocarpine were relatively free of cholinergic side effects, whereas milameline, WAL 2014 and SKB 202026 produced non-selective effects. Therefore, these data demonstrate that xanomeline selectively activates in vivo PI hydrolysis, consistent with activation of biochemical processes involved in memory and cognition and xanomeline's beneficial clinical effects on cognition in Alzheimers patients.

Effects of the M1 agonist xanomeline on processing of human beta-amyloid precursor protein (FAD, Swedish mutant) transfected into Chinese hamster ovary-m1 cells.

Complementary DNA (cDNA) encoding human beta-amyloid precursor protein familial Alzheimer's disease (FAD) Swedish mutant (beta APPSM) form was cloned into a mammalian expression vector (PK255) containing the CMV promoter. The vector was transfected into Chinese hamster ovary cells containing human muscarinic m1 receptors (CHO-m1), and clonal cells stably expressing beta APPSM were isolated. The effects of m1-receptor activation by the selective m1 agonist xanomeline and the non-selective muscarinic agonist carbachol on processing of beta APPSM to release soluble APP (APPs) and beta-amyloid peptide (A beta) were compared. Xanomeline stimulated APP release with a potency 1000-fold greater than that observed for carbachol. Concentrations of carbachol and xanomeline producing maximal effects on APPs release reduced the secretion of A beta by 28 and 46%, respectively. These results extend previous studies with xanomeline and suggest that cholinergic replacement therapy for Alzheimer's disease may reduce amyloid deposition.

1,2,5-Thiadiazole analogues of aceclidine as potent m1 muscarinic agonists.

The acetyl group of the muscarinic agonist aceclidine 4 was replaced by various 1,2,5-thiadiazoles to provide a new series of potent m1 muscarinic agonists 17 and 18. Optimal m1 muscarinic agonist potency was achieved when the 1,2,5-thiadiazole substituent was either a butyloxy, 17d, or butylthio, 18d, group. Although 1,2,5-oxadiazole 37 and pyrazine 39 are iso-pi-electronic with 1,2,5-thiadiazole 17d, both analogues were substantially less active than 17d. Compounds with high muscarinic affinity and/or m1 muscarinic agonist efficacy were also obtained when the 3-oxyquinuclidine moiety of 17d or 18c was replaced by ethanolamines, hydroxypyrrolidines, hydroxyazetidine, hydroxyisotropanes, or hydroxyazanorbornanes. The structure-activity data support the participation of the oxygen or sulfur atom in the substituent on the 1,2,5-thiadiazole in the activation of the m1 receptor. Several of these new 1,2,5-thiadiazoles have m1 agonist efficacy, potency, and selectivity comparable to those of xanomeline 2 in the muscarinic tests investigated.

Identification of side chains on 1,2,5-thiadiazole-azacycles optimal for muscarinic m1 receptor activation.

Series of analogs to the functional m1 selective agonist, xanomeline (hexyloxy-TZTP), were evaluated for their in vitro m1 efficacy in cell lines transfected with the human m1 receptor. Systematic variation of the side chain and the azacyclic ring led to the discovery of potent muscarinic agonists with robust m1 efficacy, all having the phenylpropargyloxy/thio as the side chain. The most selective compound was the phenylpropargylthio-[3.2.1] endo analog 28, which is a potent and efficacious m1 agonist with no m2 activity.

Tissue plasminogen activator is elevated in women with ischemic stroke.

A recent study suggests that a high plasma level of tissue plasminogen activator (t-PA antigen) is a risk factor for stroke in men. Whether t-PA antigen is a risk factor for stroke in women is unknown. We measured plasma levels of t-PA antigen in 302 nonselected patients with acute ischemic stroke and in 138 healthy control subjects. In a subgroup of the patients, plasma t-PA antigen was remeasured 6 months after the stroke. Women with acute ischemic stroke (n=171) had median plasma t-PA antigen that was 39% higher than the healthy female control subjects (n=86): 10.3 (8.0 to 13.7) versus 7.4 (6.1 to 9.1) ng/mL (median [interquartile range]), P=.0001. At the reexamination of the patients after 6 months, plasma t-PA antigen was unchanged in the female patients. This suggests that the difference in plasma t-PA antigen between the female patients and the healthy control subjects did not result from an acute phase reaction. In a multivariate regression analysis, high t-PA antigen was an independent risk factor for stroke, and high plasma level of t-PA antigen was associated with severe stroke in women. The current data suggest that plasma t-PA antigen is elevated in women with ischemic stroke.

An Insertion/Deletion polymorphism in the promoter region of the plasminogen activator inhibitor-1 gene is associated with plasma levels but not with stroke risk in the elderly.

The purpose of the present study was to examine the effects of an insertion/deletion (ins/del) polymorphism in the promoter region of the plasminogen activator inhibitor-1 (PAI-1) gene on plasma PAI-1 antigen and activity levels and on stroke risk in the elderly. The ins/del genotype and PAI-1 antigen and activity plasma levels were determined in 177 patients with ischemic stroke (mean age, 75 years) and 93 healthy elderly subjects (mean age, 74 years). There was no difference in the frequencies of the ins and del alleles between stroke patients and healthy elderly subjects. The del/del genotype was associated with the highest plasma PAI-1 antigen levels in the healthy subjects: those with the ins/ins genotype had 36% lower plasma PAI-1 antigen levels than those with the del/del genotype (effect of genotype, P=0.3). In contrast, the ins/del genotype was not associated with plasma PAI-1 antigen and activity levels in 89 patients who had a stroke less than 10 days before blood sampling. However, an association of ins/del genotype with plasma PAI-1 activity levels could be demonstrated in 88 other patients more than 5 months after the stroke. This may suggest that PAI-1 metabolism is temporarily perturbed after a stroke. The present data suggest that an ins/del polymorphism in the PAI-1 promoter region affects plasma PAI-1 levels but has little or no effect on stroke risk in the elderly.

Decreased serum testosterone in men with acute ischemic stroke.

Serum levels of total and free testosterone and 17 beta-estradiol were determined in 144 men with acute ischemic stroke and 47 healthy male control subjects. Blood samples from patients were drawn a mean of 3 days after stroke onset and also 6 months after admission in a subgroup of 45 patients. Initial stroke severity was assessed on the Scandinavian Stroke Scale and infarct size by computed tomographic scan. Mean total serum testosterone was 13.8 +/- 0.5 nmol/L in stroke patients and 16.5 +/- 0.7 nmol/L in control subjects (P = .002); the respective values for free serum testosterone were 40.8 +/- 1.3 and 51.0 +/- 2.2 pmol/L (P = .0001). Both total and free testosterone were significantly inversely associated with stroke severity and 6-month mortality, and total testosterone was significantly inversely associated with infarct size. The differences in total and free testosterone levels between patients and control subjects could not be explained by 10 putative risk factors for stroke, including age, blood pressure, diabetes, ischemic heart disease, smoking, and atrial fibrillation. Total and free testosterone levels tended to normalize 6 months after the stroke. There was no difference between patients and control subjects in serum 17 beta-estradiol levels. These results support the idea that testosterone affects the pathogenesis of ischemic stroke in men.

Body temperature in acute stroke: relation to stroke severity, infarct size, mortality, and outcome.

BACKGROUND: In laboratory animals, cerebral ischaemia is worsened by hyperthermia and improved by hypothermia. Whether these observations apply to human beings with stroke is unknown. We therefore examined the relation between body temperature on admission with acute stroke and various indices of stroke severity and outcome. METHODS: In a prospective and consecutive study 390 stroke patients were admitted to hospital within 6 h after stroke (median 2.4 h). We determined body temperature on admission, initial stroke severity, infarct size, mortality, and outcome in survivors. Stroke severity was measured on admission, weekly, and at discharge on the Scandinavian Stroke Scale (SSS). Infarct size was determined by computed tomography. Multiple logistic and linear regression outcome analyses included relevant confounders and potential predictors such as age, gender, stroke severity on admission, body temperature, infections, leucocytosis, diabetes, hypertension, atrial fibrillation, ischaemic heart disease, smoking previous stroke, and comorbidity. FINDINGS: Mortality was lower and outcome better in patients with mild hypothermia on admission; both were worse in patients with hyperthermia. Body temperature was independently related to initial stroke severity (p < 0.009), infarct size (p < 0.0001), mortality (p < 0.02), and outcome in survivors (SSS at discharge) (p < 0.003). For each 1 degrees C increase in body temperature the relative risk of poor outcome (death or SSS score on discharge < 30 points) rose by 2.2 (95% CI 1.4-3.5) (p < 0.002). INTERPRETATION: We have shown that, in acute human stroke, an association exists between body temperature and initial stroke severity, infarct size, mortality, and outcome. Only intervention trials of hypothermic treatment can prove whether this relation is causal.