부산대학교 도서관

Background/objectives: We aimed to examine dietary patterns and their longitudinal associations with socio-demographic and lifestyle factors in older adults.Subjects/methods: A cohort of 1098 participants aged 50-80 years were followed for 5 years. Dietary intake was assessed at baseline, 2.6 and 5 years using a validated food frequency questionnaire. Dietary patterns were identified at baseline using exploratory factor analysis and pattern scores for each calculated using the weighted sum score method. Associations of dietary pattern scores with participants' characteristics were assessed using linear mixed-effects models.Results: The three dietary patterns identified and the food groups of which they were predominantly composed were as follows: a healthy dietary pattern (vegetables, fruits, nuts, and whole grains); a western dietary pattern (pizza, hamburgers, chips, and potatoes); and a meat and vegetable dietary pattern (red meat, fish, poultry, vegetables, potatoes, and legumes). Being a man, unemployed, a current smoker, less educated, and residing in a socially disadvantaged area were associated with lower healthy dietary pattern scores, but these differences lessened over time, except in current smokers (p < 0.03 for interactions with time). Being a man was associated with higher, but being a current smoker with lower western dietary pattern scores (β = 8.0, 95% CI: 5.3,10.7 and - 6.7: - 10.1,- 3.3, respectively). For the meat and vegetable dietary pattern, being a man and a current smoker were associated with lower scores (β = - 24.9, 95% CI: - 44.9,- 4.9 and - 66.8: - 98.3,- 35.3, respectively), while being unemployed was associated with higher scores but this difference lessened over time (p = 0.018 for interaction with time).Conclusions: In older adults, men, smokers, and those experiencing social disadvantage could be target groups for interventions to improve diets. [ABSTRACT FROM AUTHOR]