[Skip to Navigation]
Sign In
December 24, 1997

Inverse Association of Dietary Fat With Development of Ischemic Stroke in Men

Author Affiliations

From the Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care (Dr Gillman), Boston University School of Public Health (Drs Cupples and Millen), and Evans Department of Medicine (Dr Ellison) and Department of Neurology (Dr Wolf), Boston University School of Medicine, Boston, Mass.

JAMA. 1997;278(24):2145-2150. doi:10.1001/jama.1997.03550240035030

Context.  —A few ecological and cohort studies in Asian populations suggest an inverse association of the intake of both fat and saturated fat with risk of stroke. However, data among western populations are scant.

Objective.  —To examine the association of stroke incidence with intake of fat and type of fat among middle-aged US men during 20 years of follow-up.

Design and Setting.  —The Framingham Heart Study, a population-based cohort study.

Participants.  —A total of 832 men, aged 45 through 65 years, who were free of cardiovascular disease at baseline (1966-1969).

Measurements and Data Analysis.  —The diet of each subject was assessed at baseline by a single 24-hour dietary recall, from which intakes of energy and macronutrients were estimated. In Kaplan-Meier analyses, we calculated age-adjusted cumulative incidence rates of stroke. Using Cox regression, we estimated stroke incidence relative risks during 20 years of follow-up.

Main Outcome Measure.  —Incidence of ischemic stroke, which occurred in 61 subjects during the follow-up period.

Results.  —Mean intakes were 10 975 kJ for energy; 114 g (39% of energy) for total fat; 44 g (15%) for saturated fat; 46 g (16%) for monounsaturated fat; and 16 g (5%) for polyunsaturated fat. Risk of ischemic stroke declined across the increasing quintile of total fat (log-rank trend P=.008), saturatedfat(P=.002), and monounsaturated fat (P=.008) but not polyunsaturated fat (P=.33). The age- and energy-adjusted relative risk for each increment of 3% of energy from total fat was 0.85 (95% confidence interval [CI], 0.78-0.94); for an increment of 1% from saturated fat, 0.91 (95% CI, 0.85-0.98); and for 1% from monounsaturated fat, 0.89 (95% CI, 0.83-0.96). Adjustment for cigarette smoking, glucose intolerance, body mass index, blood pressure, blood cholesterol level, physical activity, and intake of vegetables and fruits and alcohol did not materially change the results. Too few cases of hemorrhagic stroke (n=14) occurred to draw inferences.

Conclusion.  —Intakes of fat, saturated fat, and monounsaturated fat were associated with reduced risk of ischemic stroke in men.

Add or change institution