Dietary Pattern Trajectories over Time and Diabetes among Chinese Adults Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 22, 2019
  • Batis Ruvalcaba, Carolina
    • Affiliation: Gillings School of Global Public Health, Department of Nutrition
  • Dietary patterns, instead of single nutrients or foods, are a useful approach to study diet and diet-disease associations. However, most studies examine dietary patterns only at one point in time. The purpose of this dissertation was to identify the longitudinal changes or stability of dietary patterns and their association with Diabetes in the China Health and Nutrition Survey from 1991 to 2009 (7 waves of diet data). Aim 1: we derived two dietary patterns using factor analysis in each wave: a traditional southern pattern (rice, vegetables, meat, poultry and fish) and a modern high-wheat pattern (wheat products, nuts, fruits, eggs, milk and instant noodles/frozen dumpling). The structure of these patterns remained stable over time, but the tracking was lower and the adherence increased over time for the modern high-wheat. Aim 2: among 4,316 adults not previously diagnosed with diabetes the adjusted Odds Ratio for diabetes prevalence in 2009, comparing the highest versus the lowest dietary pattern score quartile in 2006, was 1.25 (0.78, 2.01) for the modern high-wheat pattern, 0.79 (0.51, 1.21) for the traditional southern pattern and 2.36 (1.55, 3.58) for a pattern derived with Reduced Rank Regression (with HbA1c, HOMA-IR and glucose as response variables). This pattern combined items of the modern high-wheat pattern (wheat products and soy milk) with items opposite to the traditional southern (low rice, poultry and fish). Aim 3: A score for the third dietary pattern was estimated for each subject at each wave and with Latent Class Trajectory Analysis subjects with similar trajectories of their dietary pattern's score over time were grouped in 5 classes. Among two classes with similar scores in 2006, the one with lower scores from 1991-2004, had significantly lower HbA1c [-1.64 (-3.17, -0.11)], and non-significantly lower odds of diabetes. All together our findings suggest that the popularity of a modern high-wheat pattern was increasing and that part of this pattern, when combined with low intake of rice, poultry, fish and legumes, was associated with diabetes. In addition, even if the diets were similar recently, the long-term trajectories of this dietary pattern were also associated.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Popkin, Barry
  • Doctor of Philosophy
Graduation year
  • 2013

This work has no parents.