Migraine with aura, as well as high total cholesterol, associated with increased stroke in women

Shelley Wood

April 28, 2004

San Francisco, CA - Women who suffer from migraines with aura are much more likely to have strokes than women who get migraines without auras or who get nonmigraine headaches, a new analysis from the Women's Health Study (WHS) suggests. In a separate study by the same researchers, cholesterol levels also predicted stroke risk, in a finding that did not appear to be mediated by the presence or absence of hypertension.

Dr Tobias Kurth (Brigham and Women's Hospital, Boston, MA) and colleagues presented the results of both studies at the American Academy of Neurology 56th Annual Meeting.

Previous research linking stroke risk to headache type has been inconclusive and contradictory, Kurth commented. "There've been a couple of retrospective case-controlled studies that suggest an association between particular migraines with aura and ischemic stroke; however, no prospective data have previously been available. Other studies have suggested that migraine overall, and headache overall, are associated with various stroke types."

In their migraine study, Kurth et al reviewed the incidence of self-reported migraines, aura symptoms, and headaches in the 39754 female health professionals participating in the WHS. Over an average of nine years of follow-up, 385 strokes occurred, of which 309 were ischemic, 72 hemorrhagic, and four undefined.

One less headache

Compared with women who did not get migraines, women who reported migraine without aura had no increased risk of stroke, whereas women who reported migraine with aura had a statistically significant increased risk of stroke.

Stroke risk in women with migraine plus aura


Adjusted hazard ratio

95% CI

Total stroke

1.53 1.01-2.30

Ischemic stroke

1.70 1.10-2.64

Total stroke, age <55

1.74 1.01-3.00

Ischemic stroke, age <55

2.25 1.30-3.90

Risk of stroke was no higher in women with nonmigraine headaches than in women with no headaches whatsoever, they report.

"There has been a lot of speculation among experts at this meeting as to why this association might be there," Kurth commented to heartwire . "It looks like migraine with aura might be one factor among other factors that increases the risk of ischemic stroke, and why exactly that is, is a question for further research. I can't give you an exact reason, but there is speculation on the vascular components of the migraine, although this alone might not explain the whole phenomenon."

Lipid levels and stroke

In their cholesterol analysis, Kurth et al examined lipid levels in relation to stroke risk in the 27939 US women participating in the WHS who had provided blood samples at baseline. As the investigators point out, while the link between lipid levels and coronary artery disease is well established, the link between lipids and stroke risk is less clear.

Over nine years of follow-up, 199 ischemic strokes occurred in the WHS subset. Increased ischemic stroke risk was linked with higher total cholesterol and higher triglyceride levels. Higher LDL cholesterol appeared to have no strong association with ischemic stroke, with wide confidence intervals, whereas high HDL was associated with decreased ischemic stroke risk. "This might be explained by other factors such as smoking, body mass index, exercise, and other confounding factors," the researchers said.

Risk of stroke based on lipid levels

Lipid level

Adjusted hazard ratio

95% CI

Total cholesterol >240 mg/dL (compared with <200 mg/dL)

1.47 1.02-2.13

LDL >160 mg/dL (compared with <100 mg/dL)

1.30 0.82-2.06

TG >200 mg/dL (compared with <150 mg/dL)

1.98 1.41-2.77

HDL >60 mg/dL (compared with <40 mg/dL)

0.66 0.42-1.03


When the results were adjusted for the presence or absence of hypertension, the association between all lipid subfractions and ischemic stroke was "only slightly attenuated."

Kurth emphasized that since arterial stiffening caused by high lipids can raise blood pressure, it is important to try to extricate hypertensive effects from the lipid effects. "In our analysis, it seems likely that some of the effect can certainly be explained by the presence of hypertension, but not all of the effect of total cholesterol and triglycerides is explained by hypertension."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.