Oat products and lipid lowering. A meta-analysis

Objectives. —To test the a priori hypothesis that consumption of oats will lower the blood total cholesterol level and to assess modifiers and confounders of this association. Data Sources. —A computerized literature (MEDLINE) search and the Quaker Oats Co identified published and unpublished trials as of March 1991. Raw data were requested for all trials. Study Selection. —Trials were included in summary effect size estimates if they were randomized and controlled, if a formal assessment of diet and body weight changes occurred, and, if raw data were not received, if there was enough information in the published report to perform calculations. Data Synthesis. —Twenty trials were identified. Using the methods of DerSimonian and Laird, a summary effect size for change in blood total cholesterol level of -0.13 mmol/L (-5.9 mg/dL) (95% confidence interval Cl, -0.19 to -0.017 mmol/L -8.4 to -3.3 mg/dL) was calculated for the 10 trials meeting the inclusion criteria. The summary effect size for trials using wheat control groups was -0.11 mmol/L (-4.4 mg/dL) (95% Cl, -0.21 to -0.01 mmol/L -8.3 to -0.38 mg/dL). Calculation of Keys scores demonstrated that substituting carbohydrates for dietary fats and cholesterol did not account for the majority of blood cholesterol reduction. Larger reductions were seen in trials in which subjects had initially higher blood cholesterol levels (≥5.9 mmol/L ≥229 mg/dL), particularly when a dose of 3 g or more of soluble fiber was employed. Conclusion. —This analysis supports the hypothesis that incorporating oat products into the diet causes a modest reduction in blood cholesterol level. (JAMA. 1992;267:3317-3325)