Location: Boston, Massachusetts2010 Annual Report
1a. Objectives (from AD-416)
1. Use population-based approaches to explore the role of dietary patterns and diet quality in relation to weight gain and obesity among rural older-aged adults. 2. Use population-based approaches to explore the role of dietary patterns and diet quality in relation to obesity-related chronic diseases, such as cardiovascular disease, metabolic syndrome, and diabetes, among rural older-aged adults. 3. Develop and validate GRAS dietary instruments with a population of rural adults aged 80 years of age or greater.
1b. Approach (from AD-416)
The Geisinger Rural Aging Study (GRAS) is a longitudinal study of health outcomes in relation to nutritional status among 21,645 community-dwelling Pennsylvanians aged greater than or equal to 65 years. Initial screenings for GRAS participants took place between 1994-1999. The baseline nutrition risk screening data includes height, weight, reported weight gain/loss, depression, polypharmacy, eating habits, food security, oral health, and functional status. Follow up rescreening at 3-4 year intervals has encompassed this information as well as additional queries of medical history, weight history, family history, dietary practices, medical co-morbidities, medication use, general health, depression, mobility/functional status, and healthcare resource use. Our prior investigations have highlighted the growing prevalence of obesity and ill health among these individuals. Findings suggest that many of these obese older persons consume poor quality diets and have associated micronutrient deficiencies. Dietary patterns/quality and their relationships with obesity and chronic diseases of older persons are poorly characterized; especially among “old older” persons 80 years of age or greater. Current investigations have supported the development of a Population Specific Food Frequency Questionnaire (PSFFQ) and a 37 item Diet Quality Screening Questionnaire (DQSQ). The DQSQ is an innovative instrument for assessment of diet quality that will be made widely available to investigators and practitioners. Our proposed next steps to further characterize the GRAS cohort include longitudinal follow up of obesity-related and other health outcomes in relation to dietary patterns and diet quality among 459 subjects for whom we have conducted comprehensive assessments 5-10 years previously. We will confirm and/or refine the results of our previous dietary patterns analyses with these data then determine associations among dietary intake, dietary patterns, diet quality lifestyle factors and anthropometric (height, weight, body-mass index) and available laboratory measures. We are particularly interested in evaluating associations with weight gain and obesity in older-aged adults. These analyses will be facilitated by use of the Geisinger Health Plan EPIC electronic database. Subsequent analyses will focus on the relationship between the aforementioned dietary variables and obesity-related chronic diseases, such as cardiovascular disease, metabolic syndrome, and diabetes. We will also administer the new DQSQ to the next wave of GRAS screening participants via mailing with telephone follow-up as needed. Based upon our rescreening experience we estimate that over the next two years we will have access to approximately 2,000 respondents (57% female; 53% 80 years of age or older). We will then monitor their health-related outcomes prospectively in relation to diet quality. Anticipated products of this research include better understanding of obesity-related and other health outcomes in relation to dietary patterns/quality among rural older persons and further validation of the DQSQ in relation to health outcomes among “old older” persons.
3. Progress Report
The Geisinger Rural Aging Study (GRAS) is a longitudinal study of health outcomes in relation to nutritional status among 21,645 community-dwelling Pennsylvanians aged =65-years. Our prior investigations have highlighted the surging prevalence of obesity and ill health among these individuals. Prior investigations have also supported the development of a Population Specific Food Frequency Questionnaire (PSFFQ) and a 37 item Diet Quality Screening Questionnaire (DQSQ). Our findings have suggested that many obese older persons consume poor quality diets and have associated micronutrient deficiencies. In the current phase of investigation this work is being extended by relating dietary patterns and diet quality to weight gain and obesity as well as to obesity-related chronic diseases in subjects for whom we have conducted comprehensive assessments 5-10 years previously. In addition, the application of the dietary instruments will be further validated in a population of rural adults aged 80 years of age or greater. Specific activities conducted during FY 2010 successfully addressed the milestones required to support the proposed population study approaches described in the first phase of the approved project plan objectives. Appreciable progress was made in the laying the foundation steps required for completion of all three objectives. Long-term data that encompasses study outcomes has been accessed and captured in new databases. A reliability study has been conducted contrasting manual individual record review versus electronic data abstraction. Our ability to access suitable long-term body weight data has been systematically evaluated. A scannable version of our DQSQ has been developed, tested, and disseminated to the screening cohort.
1. Long-term database of health outcomes created: In order to monitor the impact of dietary patterns upon weight gain, obesity, and obesity-related chronic diseases, it is necessary to access these long-term outcome measures from electronic medical record databases. Over the past year we have successfully obtained height, weight, and medical diagnosis codes for obesity-related chronic diseases for subjects who participated 5-10 years previously in comprehensive diet assessment studies. Analysis confirmed that the vast majority of the original enrollees from the earlier studies remained available for further investigation and that they had follow up weight measures. A high quality database that accesses these long-term study outcomes is a critical prerequisite for relating dietary patterns to the health outcomes of interest.
2. Study confirms reliability of data accessed from electronic medical record: Reliable capture of electronic medical record data is required to process large volumes of long-term health outcome information. A formal reliability study has been completed contrasting data from manual individual record review versus electronic abstraction of electronic medical records. Evaluated health outcomes included cardiovascular disease, diabetes, hypertension, obstructive sleep apnea, osteoarthritis, depression, liver disease, and metabolic syndrome; biochemical measurements (triglycerides, HDL-cholesterol, blood pressure, glucose); and measured height and weight. Medical diagnosis codes, laboratory data, past medical history, medications, and physician notes were used for the audit. Agreement between the manual record review and the electronic data abstraction were very strong. Highly reliable electronic medical record information is essential to address the relationships of dietary patterns to health outcomes.
5. Significant Activities that Support Special Target Populations
GRAS is the largest cohort of rural older persons in the US for the study of nutritional status in relation to health outcomes. Rural areas are often socio-economically disadvantaged with predominantly aging populations, lower educational levels, and lower health care utilization. Our technology transfers that include presentations at national meetings are examples of specific activities that target rural aging populations. At baseline, approximately 21,645 participants (split evenly by gender) 65 years of age or older were recruited. A significant proportion (72.5%) of the cohort was overweight (body mass index >25) or obese (BMI>30). At this point the majority of the cohort is now over 80 years of age with increasing frailty and mortality. This presents a very special opportunity to characterize the oldest old in regard to nutrition, obesity, and disease outcomes. Recent extraction of data for the GRAS cohort spanning the calendar years 2004-2008 revealed 21,578 GRAS participants in the data file. Of these participants, 3,245 died prior to 2004 and 3,956 died between 2004 and 2008.