Factors associated with the change in diet in the EPIC Potsdam study

The effect of a change in diet of adults on disease risks has not been investigated in large-scale prospective studies. The European Prospective Investigation into Cancer and Nutrition (EPIC) Potsdam study will analyze this issue. In a First step, the number of persons who changed their diets in the first 2 years of follow-up, characteristics of this group and the most frequent reasons for a change in diet were determined. The EPIC Potsdam study comprises 27548 individuals, 25927 (94 %) of whom returned their Follow-up questionnaire until August 2000. In the follow-up questionnaire it was inquired whether the current diet is different from that of two years ago. Data of 15381 women (age 35-70 years) and 9789 men (age 40-69 years) were included in the analysis. The group of those who answered 'partly' and 'yes' was compared to those who responded with 'no'. Men and women were analysed separately. P values < 0.001 of the chi(2) and the COCHRA-MANTEL-HEANZEL test (adjusted for age, body mass index [BMI], smoking status and educational level) were considered to be statistically significant. 49 % of the women and 42 To of the men reported that they had changed their diet. Men and women who reported a change in diet were those who had more frequently been overweight and obese (BMI 2: 25), who had changed their diets in the past 12 months, who were less content with their health, had been ex-smoker, had lost more than 2 kg body weight per year of follow-up, or in whom diabetes mellitus, hyperlipidemia or heart disease had been diagnosed during the follow-up. Among the overweight subjects, particularly higher educated women reported to have changed their diet during the past two years. Women changed their diet mainly because of body weight problems (63 %), men because of body-weight problems (57 %) ana because of diseases of the metabolic syndrome (50 %). Since the EPIC Potsdam was conducted in a selected population, the extent of changes is not directly applicable to the general population. However, the number of individuals who reported a change in diet is high enough to study its effect on the risk of cardiovascular disease or cancer, particularly if risk factors such as diabetes or hyperlipidemia are present.