Dietary behaviors associated with total fat and saturated fat intake.

OBJECTIVE: To estimate percentages of US adults who have adopted behaviors promoted by dietary guidance about how to reduce fat intake, and to assess relationships between these behaviors and intake of energy from total and saturated fat.

DESIGN: Relationships were examined between intake of total and saturated fat from two 24-hour recalls in the US Department of Agriculture's 1994-1996 Continuing Survey of Food Intakes by Individuals and responses to 19 fat-related behavior questions on the follow-up Diet and Health Knowledge Survey (DHKS).

SUBJECTS: Data are from a national sample of 5,649 individuals 20 years of age and older.

STATISTICAL ANALYSES PERFORMED: Multiple regression models are used to identify dietary behaviors, demographic factors, and personal characteristics that are determinants of fat intake.

RESULTS: In this study, the percentage of US adults who consistently followed the low-fat behaviors ranged from 8% to 70%. The most highly adopted behaviors (45% or more of adults) included trimming fat from meat, removing skin from chicken, and eating chips infrequently. The least highly adopted behaviors (15% or less of adults) included eating baked or boiled potatoes without added fat, avoiding butter or margarine on breads, eating low-fat instead of regular cheeses, and having fruit for dessert when dessert is eaten. Together, the 19 fat-related behavior questions on the DHKS formed a statistically significant predictor for total fat and saturated fat intake, expressed as a percent of energy (P<.0001). Key behaviors in terms of their predicted effect on lowering both total and saturated fat intake were never adding fat to baked or boiled potatoes, not eating red meats, eating less than 3 eggs per week, and never eating chicken fried. Predicted effects of these key behaviors in terms of lowering fat intake as a percentage of energy were > or = 1.5 percentage points for total fat and > or = 0.5 percentage point for saturated fat.

CONCLUSIONS: Results have applications for designing brief fat assessment instruments and for identifying key nutrition education messages that promote important fat-lowering behaviors.