Abstract
BACKGROUND: Food intake patterns provide a summary of dietary intake. Few studies have examined trends in food intake patterns over time in large, population-based studies. We examined food intake patterns and related sociodemographic and individual characteristics in the large Northern Sweden Diet Database during the two time windows 2000-2007 and 2008-2016. METHODS: In total, 100 507 participants (51% women) who had filled in a 64-item food frequency questionnaire and provided background and sociodemographic data between 2000 and 2016 were included. Food intake patterns were evaluated for women and men separately for the two time windows 2000-2007 and 2008-2016, respectively. Latent class analysis was used to identify distinct, latent clusters based on 40 food groups. RESULTS: Among both women and men, a greater proportion of participants were classified into food intake patterns characterized by high-fat spread and high-fat dairy during 2008-2016 compared to 2000-2007. In the earlier time window, these high-fat clusters were related to lower educational level and smoking. Simultaneously, the proportion of women and men classified into a cluster characterized by high intake of fruit, vegetables, and fibre decreased from the earlier to the later time window. CONCLUSION: From a public health perspective, the increase in clusters with a high conditional mean for high-fat spread and high-fat dairy and decrease in clusters with a high conditional mean for fruit and vegetables, during the time period 2008-2016 compared to 2000-2007, is worrisome as it indicates a shift away from the recommended food habits. Subgroups of women and men with less healthy dietary patterns in the time window 2008-2016 with lower education, lower age, higher body mass index, lower levels of physical activity and more smoking were identified and future interventions may be targeted towards these groups.