Article Text


Coffee and subarachnoid haemorrhage
  1. W T Longstreth, Jr
  1. Department of Neurology, University of Washington, Harborview Medical Center, 325 Ninth Avenue, Seattle, Washington 98104–2420, USA
  1. Correspondence to:
 Dr W T Longstreth;

Statistics from

The link between coffee and subarachnoid haemorrhage is unresolved

You may be making enemies, especially in Seattle, if you conclude from the study of Isaksen et al that coffee is a risk factor for subarachnoid haemorrhage (this issue, pp 185-7).1 Love of java necessitates a critical evaluation. Are we dealing with coincidence, confounding, or causation? In a case-control study, these investigators drew subjects from a population based health survey of inhabitants in the municipality of Tromsø, Norway. At variable times before the bleeding (maximum 186 months) participants had been evaluated. The investigators found that cigarette smoking and high systolic blood pressure increased risk, as have others.2 The trend was for high cholesterol to reduce risk but not significantly, perhaps reflecting the small number of participants (n = 26). Drinking six or more cups of coffee per day yielded an odds ratio of 3.86 even after controlling for these other factors in multivariable analyses.

Is the association a coincidence? Few have examined coffee as a risk factor for stroke,3 let alone subarachnoid haemorrhage, so we cannot look for consistency with previous studies. The investigators did not specify how many candidate risk factors they examined, but the more examined, the greater the risk that a significant finding will emerge by chance alone. Perhaps the association is real but not causal. The search for risk factors is important because, if associations are causal, the possibility exists of understanding better pathophysiology and prevention. Rather than being causal, could the association with coffee have been confounded by other factors related both to coffee consumption and to subarachnoid haemorrhage, such as alcohol consumption? Perhaps the association would have disappeared after controlling for alcohol consumption. Also by examining different doses of coffee, as they did for cigarette smoking, the investigators could have strengthened their argument for a causal association.

If the association were real, how might coffee increase the risk of subarachnoid haemorrhage? Bleeding is typically the culmination of aneurysm formation and rupture. Examining coffee as a risk factor in patients with unruptured intracranial aneurysms or in patients with multiple aneurysms may address questions of formation. How the association varied as a function of the time since the onset of the bleeding may address questions of rupture. Caffeine can cause an increase in blood pressure,4 perhaps putting those who harbour an intracranial aneurysm and who drink six cups or more of caffeinated coffee per day at increased risk for rupture compared with those who drink less or do not drink coffee at all. In the current study we do not know whether the coffee was caffeinated and whether other caffeinated beverages, such as tea and cola drinks, were examined. Questions about rupture would require knowledge about the use of the beverage in the time immediately before the onset of the bleeding, not months before. As is so often the case with such unexpected findings, more studies are needed before we can judge the importance of this intriguing and novel observation, especially about such a common exposure. So for now sip your coffee but with some lingering concern about this unresolved issue.

The link between coffee and subarachnoid haemorrhage is unresolved


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.