Mediterranean diet may lower risk for stroke in women

NASHVILLE, Tenn. — Women who adhered to the Mediterranean diet had a lower incidence of total stroke and ischemic stroke, according to new data from the California Teachers Study.

Researchers reported no association between Mediterranean diet adherence and incidence of hemorrhagic stroke in women, Ayesha Z. Sherzai, MD, from Columbia University Medical Center, said during a presentation at the International Stroke Conference.

Sherzai and colleagues analyzed 104,268 women from the California Teachers Study aged 26 to 99 years with no history of stroke at baseline. The mean age of the women was 53 years and 87% were white.

Participants were given a score of 0 to 9 based on adherence to Mediterranean diet. A score of 9 reflected the strongest adherence to the diet. The women were then stratified into quintiles based on adherence: score of 0 to 2 (16.1%), 3 (18.2%), 4 (21.4%), 5 (20.1%) or 6 to 9 (24.27%).

Since 1995, researchers have documented 3,165 strokes (2,270 ischemic, 895 hemorrhagic) in the California Teachers Study population, Sherzai said.

In an unadjusted model, compared with the lowest quintile, women in all other quintiles of Mediterranean diet adherence had a reduced risk for stroke. Risk decreased the higher the quintile of adherence (P for trend < .0001), Sherzai said.

In an adjusted model, the trend of decreased risk for overall stroke with greater adherence to the Mediterranean diet remained (P for trend = .009). However, compared with the lowest quintile, the reduced risk was significant only in the fourth (HR = 0.86) and the fifth quintiles of adherence (HR = 0.83), she said.

For ischemic stroke, compared with the lowest quintile, risk was reduced for the other quintiles in an unadjusted model (P for trend < .0001), and was significantly reduced for the third quintile (HR = 0.84), fourth quintile (HR = 0.85) and fifth quintile (HR = 0.82; P for trend = .01) in an adjusted model, she said.

For hemorrhagic stroke, compared with the lowest quintile, risk was not significantly reduced for any other quintile in an unadjusted (P for trend = .01) or adjusted (P for trend = .06) model, Sherzai said.

“We were not really surprised, because this has been seen in the past as well,” Sherzai said. “Of course, we need better studies to address the issue of generalizability,” as all participants were of the same sex, lived in the same state and worked in the same profession.

“But given the burden of stroke and the fact that less than 1% of the U.S. adult population has an ideal diet, our finding addresses the importance of addressing diet as an important modifiable risk factor,” she said. – by Erik Swain

Reference:

Sherzai AZ, et al. Plenary Session II: LB5. Presented at: International Stroke Conference; Feb. 11-13, 2015; Nashville, Tenn.

Disclosure: Sherzai reports no relevant financial disclosures.