作者
James H. O’Keefe,Nathan L. Tintle,William S. Harris,Evan L. O’Keefe,Aleix Sala‐Vila,John Attia,Manohar L. Garg,Alexis Hure,Christian Sørensen Bork,Erik Berg Schmidt,Stine Krogh Venø,Kuo‐Liong Chien,Yun-yu Chen,Sarah Egert,Tobias Feldreich,Johan Ärnlöv,Lars Lind,Nita G. Forouhi,Johanna M. Geleijnse,Kamalita Pertiwi,Fumiaki Imamura,Vanessa D. de Mello,Matti Uusitupa,Jaakko Tuomilehto,Markku Laakso,Maria Lankinen,Danielle Laurin,Pierre‐Hugues Carmichael,Joan Lindsay,Karin Leander,Federica Laguzzi,Brenton R. Swenson,W. T. Longstreth,Jo Ann E. Manson,Samia Mora,Nancy R. Cook,Matti Marklund,Debora Melo van Lent,Rachel A. Murphy,Vilmundur Gudnason,Toshiharu Ninomiya,Yoichiro Hirakawa,Fangfei Qian,Qi Sun,Frank B. Hu,Andres V Ardisson Korat,Ulf Risérus,Iolanda Lázaro,Cécilia Samieri,Mélanie Le Goff,Catherine Helmer,Marinka Steur,Trudy Voortman,M. Kamran Ikram,Toshiko Tanaka,Jayanta Das,Luigi Ferrucci,Stefania Bandinelli,Michael Y. Tsai,Weihua Guan,Parveen K. Garg,W.M.M. Verschuren,Jolanda M. A. Boer,Anneke Biokstra,Jyrki K. Virtanen,Michael Wagner,Jason Westra,Luc Albuisson,Kazumasa Yamagishi,David S. Siscovick,Rozenn N. Lemaître,Dariush Mozaffarian
摘要
BACKGROUND: The effect of marine omega-3 PUFAs on risk of stroke remains unclear. METHODS: We investigated the associations between circulating and tissue omega-3 PUFA levels and incident stroke (total, ischemic, and hemorrhagic) in 29 international prospective cohorts. Each site conducted a de novo individual-level analysis using a prespecified analytical protocol with defined exposures, covariates, analytical methods, and outcomes; the harmonized data from the studies were then centrally pooled. Multivariable-adjusted HRs and 95% CIs across omega-3 PUFA quintiles were computed for each stroke outcome. RESULTS: Among 183 291 study participants, there were 10 561 total strokes, 8220 ischemic strokes, and 1142 hemorrhagic strokes recorded over a median of 14.3 years follow-up. For eicosapentaenoic acid, comparing quintile 5 (Q5, highest) with quintile 1 (Q1, lowest), total stroke incidence was 17% lower (HR, 0.83 [CI, 0.76–0.91]; P <0.0001), and ischemic stroke was 18% lower (HR, 0.82 [CI, 0.74–0.91]; P <0.0001). For docosahexaenoic acid, comparing Q5 with Q1, there was a 12% lower incidence of total stroke (HR, 0.88 [CI, 0.81–0.96]; P =0.0001) and a 14% lower incidence of ischemic stroke (HR, 0.86 [CI, 0.78–0.95]; P =0.0001). Neither eicosapentaenoic acid nor docosahexaenoic acid was associated with a risk for hemorrhagic stroke. These associations were not modified by either baseline history of AF or prevalent CVD. CONCLUSIONS: Higher omega-3 PUFA levels are associated with lower risks of total and ischemic stroke but have no association with hemorrhagic stroke.