Nonsikelelo Mathe1, Pedro T Pisa2, Jeffrey A Johnson1, Steven T Johnson3. 1. Alliance for Canadian Health Outcomes Research in Diabetes, School of Public Health, 2-040 Li Ka Shing Centre for Health Research Innovation, University of Alberta, Edmonton, Alberta, Canada. 2. Wits Reproductive Health & HIV Institute, University of Witwatersrand, Johannesburg, South Africa. 3. Alliance for Canadian Health Outcomes Research in Diabetes, School of Public Health, 2-040 Li Ka Shing Centre for Health Research Innovation, University of Alberta, Edmonton, Alberta, Canada; Centre for Nursing and Health Studies, Faculty of Health Disciplines, Athabasca University, Alberta, Canada. Electronic address: sjohnson@athabascau.ca.
Abstract
OBJECTIVES: Examining the diets of people living with type 2 diabetes may improve understanding of how diet affects disease progression. We derived dietary patterns in adults living with type 2 diabetes and explored associations among patterns, sociodemographic variables and cardiometabolic risk factors. METHODS: Dietary patterns were derived from food frequency questionnaires (FFQs) in 196 adults with type 2 diabetes using principal components analysis (PCA). Multilinear regression models were fitted for the differing dietary pattern scores so as to estimate the marginal contribution of each variable explaining variations in diet. Differences in clinical variables across dietary patterns, adjusting for sex, smoking and total energy intake, were assessed. RESULTS: Three principal components (PCs), or patterns, were identified, explaining 56.5% of the total variance in diet: (PC1) fried foods, cakes and ice cream; (PC2) fish and vegetables; and (PC3) pasta, potatoes and breads. Female sex, current smoker and total energy were significant associated with patterns. Total energy accounted for the greatest amount of variance in each pattern (11.2% for fried foods, cakes and ice cream, 3.89% for fish and vegetables and 9.21% for pasta, potatoes and breads). After adjustment for sex, smoking and total energy, the pasta, potatoes and breads pattern was inversely associated with systolic blood pressure and low-density lipoprotein-cholesterol. CONCLUSIONS: Of the 3 distinct diet patterns characterized, the carbohydrate-based pattern was most closely associated with cardiometabolic risk factors. To better understand and improve self-management by people living with type 2 diabetes through dietary modifications, further improvements in measuring and assessing diet using comparable instruments and comparisons with apparently healthy populations is required.
OBJECTIVES: Examining the diets of people living with type 2 diabetes may improve understanding of how diet affects disease progression. We derived dietary patterns in adults living with type 2 diabetes and explored associations among patterns, sociodemographic variables and cardiometabolic risk factors. METHODS: Dietary patterns were derived from food frequency questionnaires (FFQs) in 196 adults with type 2 diabetes using principal components analysis (PCA). Multilinear regression models were fitted for the differing dietary pattern scores so as to estimate the marginal contribution of each variable explaining variations in diet. Differences in clinical variables across dietary patterns, adjusting for sex, smoking and total energy intake, were assessed. RESULTS: Three principal components (PCs), or patterns, were identified, explaining 56.5% of the total variance in diet: (PC1) fried foods, cakes and ice cream; (PC2) fish and vegetables; and (PC3) pasta, potatoes and breads. Female sex, current smoker and total energy were significant associated with patterns. Total energy accounted for the greatest amount of variance in each pattern (11.2% for fried foods, cakes and ice cream, 3.89% for fish and vegetables and 9.21% for pasta, potatoes and breads). After adjustment for sex, smoking and total energy, the pasta, potatoes and breads pattern was inversely associated with systolic blood pressure and low-density lipoprotein-cholesterol. CONCLUSIONS: Of the 3 distinct diet patterns characterized, the carbohydrate-based pattern was most closely associated with cardiometabolic risk factors. To better understand and improve self-management by people living with type 2 diabetes through dietary modifications, further improvements in measuring and assessing diet using comparable instruments and comparisons with apparently healthy populations is required.