Stroke: triggers for brain bleeds probed

Friday May 6 2011

“Coffee, vigorous exercise and nose-blowing may trigger a stroke,” reported The Guardian . It said that a study has identified eight everyday activities that often precede a type of haemorrhagic stroke caused by a bleed in the brain.

This was a case-crossover study that looked at 250 people who had experienced a particular type of bleed in the membranes covering the brain, called a subarachnoid haemorrhage (SAH). This is a rupture of a ballooned blood vessel (an aneurysm).

The study examined peoples’ exposure to 30 different factors in the hours leading up to the SAH that could have potentially caused the rupture. These exposures were then compared to the person’s typical exposure over the previous year. Eight of the 30 assessed factors were found to be associated, including anger, sexual activity, exercise, straining at the toilet and being startled. All of these exposures would be expected to cause a brief increase in blood pressure, so it is quite plausible that these could trigger the rupture of an aneurysm within the skull, if one is present.

This was a good quality study, furthering our understanding of what can potentially trigger these types of strokes. However, its design has several limitations, and the results cannot be generalised to any other type of stroke, including haemorrhagic stroke where the bleed has occurred within the brain itself. It must be noted that only a small proportion of the population have an aneurysm (reported as 2% in the study) and even fewer of these will actually rupture. The general population are not all at risk from these common activities, as might be thought from reading the news headlines.

Where did the story come from?

The study was carried out by researchers from the Utrecht Stroke Center in the Netherlands. The study was funded by the Julius Center for Health Sciences and Primary Care and the department of neurology of the University Medical Center Utrecht. The study was published in the peer-reviewed medical journal Stroke , a Journal of the American Heart Association.

As discussed below, the news headlines are overly simplistic and do not clearly convey the small proportion of the population for whom these findings would be applicable.

What kind of research was this?

This was a case-crossover study that aimed to investigate the activities that can trigger the rupture of an intracranial aneurysm (a ballooned weak region of a blood vessel in the skull). These ruptures can cause a haemorrhagic stroke, where the weakened blood vessel bursts and the subsequent build up of blood leads to brain damage. In this study the researchers were interested in a type of haemorrhagic stroke called a subarachnoid haemorrhage. This is a bleed in the membranes surrounding the brain rather than in the brain itself. The aim was to get a better idea of the triggers that may lead to the rupture of an intracranial subarachnoid aneurysm.

A case-crossover study is a type of study similar to a case control, but where the person who has had the stroke (the case) acts as their own control. In this study the researchers looked at what the person was doing in the time immediately prior to their stroke to try and identify a trigger event. They then compared this with what the same person was doing at another time when they did not experience the event (the control period).

In case-crossover studies, the researchers often pick several control periods (for example, looking at several weeks before the event) to try and get an idea of the person’s normal habits. Essentially, the aim of the case-crossover study is to examine what happened to this person before this event (in this case, a haemorrhagic stroke) that is not normal for them? What could have triggered their stroke? These study designs have strengths but they also have numerous limitations.

What did the research involve?

The researchers recruited people who had been admitted to the Utrecht Stroke Center and had suffered a subarachnoid haemorrhage (SAH) as a result of a ruptured aneurysm. The brain and spinal cord are covered with protective layers of membranes – the dura, arachnoid and pia mater. The dura mater is the lining closest to the skull and the pia mater is the lining adhering directly to the brain. An SAH means that the bleed occurs between the arachnoid and pia layers – it is a bleed within the skull but outside of the brain, and is a type of haemorrhagic stroke (its main symptom is a sudden, very severe headache). The other type of haemorrhagic stroke is caused by an intracerebral haemorrhage – a bleed within the brain.

Eligible people had arrived at the clinic with an abrupt severe headache or loss of consciousness and had their SAH confirmed by CT scan. The researchers interviewed either the person themselves, if they were well enough, or a family member or friend if the person was seriously ill or had died from the bleed (though the researchers say that few proxies were willing to participate in the study in such circumstances).

Over a three-year period the researchers assessed 250 people who had had an SAH caused by a ruptured aneurysm. The participants’ average age was 55 years (around middle age is known to be the average for an SAH). They, or their family member or friend, completed a structured questionnaire assessing exposure to 30 potential triggers in the "hazard period" (the time before the stroke occurred, varying from two to 24 hours depending on the exposure being assessed). The respondent also supplied how often these exposures had occurred throughout the previous year when they had not resulted in an SAH.

The researchers compared the participants’ exposure to the triggers during the hazard periods with the usual rate of frequency, calculating the risk of having an SAH after each potential trigger.

What were the basic results?

Of the 30 triggers assessed, the researchers identified eight that were associated with an increased relative risk of SAH:

  • Coffee consumption: 70% increased risk (relative risk [RR] 1.7, 95% confidence interval [CI], 1.2 to 2.4)
  • Cola consumption: over three-fold increase (RR 3.4, 95% CI 1.5 to 7.9)
  • Anger: over six-fold increase (RR 6.3, 95% CI 1.6 to 25)
  • Being startled: over 23-fold increase (RR 23.3, 95% CI, 4.2 to 128)
  • Straining for defecation: over seven-fold increase (RR, 7.3, 95% CI, 2.9 to 19)
  • Sexual intercourse: over 11-fold increase (RR 11.2, 95% CI, 5.3 to 24)
  • Nose blowing: over two-fold increase (RR 2.4, 95% CI, 1.3 to 4.5)
  • Vigorous physical exercise: over two-fold increase (RR 2.4, 95% CI, 1.4 to 4.2)

How did the researchers interpret the results?

The researchers conclude that they have identified eight trigger factors for aneurysmal rupture, all of which are possible common causes as they can cause a sudden and short increase in blood pressure. They say that some of these triggers are modifiable and further studies should assess whether lessening people’s exposure to these factors could be of benefit to those known to have an intracranial aneurysm.


This is a well-conducted, good quality study, but there are several points that need to be considered when interpreting its findings to ensure that they are put in the correct context. The news headlines may give people the wrong impression that they should avoid drinking coffee, sex and blowing their nose to reduce their risk of having a stroke, and this is not the case.

This study examined people who had a subarachnoid haemorrhage. This is caused by a ruptured aneurysm in the membranes between the skull and the brain. The risk factors for developing an aneurysm are not fully established (though genetic vulnerability and high blood pressure are possible causes), and people who have them usually do not know of their presence. SAH is quite rare and accounts for a small proportion of all strokes. As the study highlights, only about 2% of the population have an intracranial aneurysm and only a few of these actually rupture. As such, though some triggers could potentially be avoided to try and reduce the risk of rupture, this would only be of relevance to people who are known to have an aneurysm within the skull. The vast majority of the population would not be at increased risk from carrying out any of these activities because they do not have an intracranial aneurysm.

The study tested the risk associations for 30 potential triggers, each requiring its own statistical tests. Carrying out a large number of statistical tests always increases the possibility of chance findings. Of the eight that had a positive association, many had extremely wide confidence intervals (4.2 to 128 for being startled), which greatly reduces the confidence that these associations are reliable. The true risk association may be quite different from that calculated.

Though it is has benefits, the case-crossover design also has several limitations, many of which the researchers point out themselves.

  • One of the strengths of this design is that it does not require controls for comparison as the cases act as their own controls. As such, many of the other potential confounders (for example, genetic and medical factors) that can differ between people are removed. A case-crossover is also the only study design that can be used to ask why a particular event happened at this point in time in this person rather than the day before or the week before, for example. They are a good study design for examining the effect of brief, transient exposures in an individual that differ from their normal habits.
  • The design’s potential limitations include recall bias. The person (or their friend or family member) knows they have suffered a SAH. They may therefore be searching for reasons why this happened and may recall exposures differently in an attempt to try and find an answer to what could have triggered it. The possibility for recall bias increases with the length of time after the event, and for 40% of the cases in this study, the respondents completed questionnaires over six weeks after the stroke had occurred.
  • Another limitation is that the participants may not be representative of those who have suffered the most severe SAH. This is largely because family members or friends of people who died or were seriously ill after the event understandably often did not wish to participate in the study. Therefore, this study may not represent the general population of people who have an SAH, and may only represent people who survive from their SAH and make a good recovery.
  • In a case-crossover study, the researchers have to choose what they are going to consider to be an appropriate "hazard period" before the event, and what they are going to consider the "control period". This all leads to potential inaccuracies.

This study is of value for understanding the potential triggers of a subarachnoid haemorrhage in the small number of people who are at risk due to the presence of an aneurysm. All of these triggers are quite plausible ones, being things that cause a sudden and short increase in blood pressure and so could be expected to rupture the aneurysm.

Analysis by Bazian
Edited by NHS Choices