Predictors of self-initiated, healthful dietary change

Objective To examine demographic and psychosocial factors that predict heal thful dietary change. Design A cohort study, examining how factors assessed at baseline predicted change in fat-related dietary habits and fruit and vegetable intakes 2 yea rs later. Participants Participants were recruited in 1995 and 1996 by random-digit d ialing (response rate 0.63), and followed-up in 1997 and 1998 (follow-up ra te 0.82). The final sample included 336 men and 502 women. Main outcome measures Fruit and vegetable intake and fat-related dietary pa tterns, measured by telephone-administered surveys. Statistical analyses chi (2) tests and linear regression were used to test associations of baseline characteristics with dietary change. Results Fat intake (energy from fat) decreased by approximately 2 percentag e points and fruits and vegetables intake increased by 0.17 servings per da y (both P < .001). Changes were significantly larger among women and person s who were well educated. Persons in the maintenance stage of change and pe rsons who believed there was a strong relationship between diet and cancer made the largest dietary changes. Use of food labels was strongly associate d with fat reduction, but not with increases in fruits and vegetables. Applications These results suggest that food labels are useful for helping people reduce fat intake, that interventions should target persons at all s tages of dietary change, and that new efforts are needed to reach men and p ersons who are less well educated.