Identification of dietary patterns and relationship with weight and health outcomes in older adults
Open Access
- Author:
- Hsiao, Pao
- Graduate Program:
- Nutrition
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 04, 2012
- Committee Members:
- Gordon Lee Jensen, Dissertation Advisor/Co-Advisor
Terryl Johnson Hartman, Dissertation Advisor/Co-Advisor
Sharon M Nickols Richardson, Committee Member
Donna Coffman, Committee Member
Martin John Sliwinski, Committee Member - Keywords:
- dietary patterns
older adults
obesity
diet
aging - Abstract:
- The proportion of older adults (≥ 65 years old) is dramatically increasing. The first of the Baby Boomer generation (adults born between 1946 and 1964) turned 65 years old in 2011. The prevalence of obesity among this age group is higher than ever. According to data from the National Health and Nutrition Examination Survey (NHANES) 2007-2008, 34% of women and 37% of men aged ≥ 60 years old are obese. Obesity is clearly related to a number of adverse health outcomes, including cardiovascular disease (CVD), type 2 diabetes mellitus, hypertension, and metabolic syndrome (MetSyn). However, the role that diet plays related to these outcomes in an aging population is less clear. The prevalence of nutrition-related chronic disease outcomes is higher among low socioeconomic and rural populations. These at-risk populations are not readily sampled in national nutritional surveys. Studies characterizing populations of rural, older adults are needed to better understand the role of dietary patterns and health disparities. In particular, limited data are available for the old-older person aged 80 years and above. Data that are available suggests that many obese older adults consume poor quality diets and that dietary practices may influence health outcomes even in populations of old-older individuals. Advances in dietary pattern research have focused on investigation of the association between specific dietary patterns and disease outcomes. Commonly used methodologies for deriving dietary patterns include both a priori (e.g., diet indices, such as the Healthy Eating Index-2005) and a posteriori methods (e.g., factor or cluster analysis). More recently, other techniques such as reduced rank regression and finite mixture modeling have also been employed. The objectives for this dissertation were three-fold: 1) to explore the association between dietary patterns and obesity-related health outcomes; 2) to examine the association between dietary patterns and diet quality in relation to obesity and weight change; and 3) to explore the use of a novel approach, finite mixture modeling, in determining dietary patterns in a sample of older adults. The first two objectives were carried out using data from the Geisinger Rural Aging Study (GRAS) and the third using data from the University of Alabama at Birmingham (UAB) Study of Aging. The GRAS is a longitudinal cohort of > 20,000 community-dwelling older adults (≥ 65 years old) living in rural Pennsylvania, initiated to examine the relationship between health outcomes and nutritional status. Cluster analysis, utilizing data from 24-hour dietary recalls, was used to derive dietary patterns in a subset of 449 participants from the GRAS (mean age: 76.5 ± 5.1 years). Prevalence (5-year follow-up) of CVD, type 2 diabetes mellitus, hypertension, and MetSyn and weight measurements were extracted from the outpatient electronic medical records using a validated data extraction process. Logistic regression, adjusting for relevant covariates, was used to examine the associations between dietary patterns and health outcomes. Cox proportional hazards regression models were used to examine the relationship between weight change and dietary patterns. The ‘Sweets and dairy’, ‘Health-Conscious’, and ‘Western’ dietary patterns were identified at baseline. Compared to the ‘Health-Conscious’ pattern, those in the ‘Sweets and dairy’ pattern had increased odds of hypertension; adjusted odds ratio (95% CI) was 2.17 (1.11-4.27). Only after stratification by gender was there a significant association between weight change and dietary pattern. Women characterized by the ‘Sweets and dairy’ and the ‘Western’ dietary pattern were three and two times more likely to lose 10 pounds, respectively, compared to those in the ‘Health-conscious’ dietary pattern. In the UAB Study of Aging (n = 416), finite mixture modeling identified three dietary patterns: a ‘Low produce, high sweets’, a ‘Western-like’, and a ‘More healthful’ dietary pattern. The most notable finding was the significant interaction found between body mass index (BMI) and gender for probability of dietary pattern membership which suggested that there was a stronger relationship between BMI and dietary pattern for women compared to men. In conclusion, finite mixture modeling was able to identify dietary patterns in a sample of older adults. Additionally, dietary patterns in the GRAS subset were significantly associated with hypertension, but not the other obesity-related outcomes of interest. The GRAS dietary patterns were also significantly related to weight loss, but only after stratifying by gender. The gender interaction seen in both the GRAS and UAB Study of Aging samples highlight that the relationship between weight and dietary pattern varies between the sexes. Unfortunately, due to sample size, we were limited in the ability to analyze meaningful subgroups for dietary pattern identification, such as gender. Additionally, our study was unable to address the potential benefits of adopting a prudent diet earlier in life. Research presented in this dissertation furthers the body of knowledge on older adults, dietary patterns and relationship to health outcomes and weight. Findings suggest that clinicians may be warranted to consider prescription of more liberalized, rather than overly-restrictive diets, for some old-older persons, especially when food intake may be inadequate. Future research should ensure investigation of larger samples of old-older adults, paying special attention to analysis by gender and measurement of dietary patterns longitudinally.