Location: Jean Mayer Human Nutrition Research Center On Aging
2023 Annual Report
Objectives
Objective 1: Continue to study the oldest old that survive in the existing Geisinger Rural Aging Study cohort in relation to longitudinal health outcomes and mortality. Perform sub-analyses on all of those who survive to 100 years of age or older, whether currently surviving or not, with a focus on nutrition risk, quality of life and health outcomes as predictors of living to become a centenarian. [NP107, C5, PS5A]
Sub-objective 1.A: Continue to study the oldest old that survive in the existing Geisinger Rural Aging Study cohort in relation to longitudinal health outcomes and mortality.
Sub-objective 1.B: Perform sub-analyses on all of those eligible to survive to 100 years of age or older, whether currently surviving or not, with a focus on nutrition risk, quality of life and health outcomes as predictors of living to become a centenarian.
Objective 2: Conduct secondary analyses of the entire existing Geisinger Rural Aging Study dataset to relate nutrition risk and other lifestyle variables at baseline enrollment with additional health outcomes not previously explored in our investigations (e.g., dementia). [NP107, C5, PS5A]
Sub-objective 2.A: Conduct secondary analyses of the entire existing Geisinger Rural Aging Study dataset to relate nutrition risk and other lifestyle variables at baseline enrollment to dementia outcomes.
Sub-objective 2.B: Conduct secondary analyses of the entire existing Geisinger Rural Aging Study dataset to relate nutrition risk and other lifestyle variables at baseline enrollment to additional health outcomes not previously explored.
Approach
The Geisinger Rural Aging Study (GRAS) was initiated between 1994-99 as a longitudinal study of health outcomes in relation to nutritional status among 21,645 individuals =65-years of age. Participants have been rescreened at 3-4-year intervals with questionnaires that encompass multiple domains of nutrition risk. Our investigations have found high prevalence of poor quality diets, obesity, and ill health. We continue to validate our diet quality screening questionnaire (DQSQ) in relation to health outcomes and mortality. We have found that DQSQ scores are significantly correlated with Healthy Eating Index (HEI) scores, which were based on 24-hour dietary recall. Those that were considered not at nutritional risk as determined by the DQSQ score had significantly higher HEI scores compared to those who were in the at-risk or possibly at risk groups. These results suggest that the DQSQ is a valid measure of diet quality even in the oldest old. Recent analysis has also revealed that higher diet quality is associated with lower mortality. Participants with the lowest diet quality scores had significantly increased risk of mortality compared to those with the highest scores. These findings suggest that diet quality may play an integral role in healthy aging with potential impact on dietary guidance for older adults. Further research will afford a unique opportunity to better characterize the impact of diet quality in a cohort of the oldest old which include a growing number of centenarians. Out of the original Geisinger Rural Aging Study (GRAS) cohort of 21,645 community-dwelling Pennsylvanians aged =65-years who completed a baseline nutrition risk screening at over 100 Geisinger clinic sites between 1994-1999, we have identified 4,245 participants that remain active in the Geisinger Healthcare System. Our analysis indicates that we already have 130 confirmed centenarians (living or expired) and we project that an additional 70-170 will become available for investigation over the coming 5-years. We will continue to study the oldest old in relation to health outcomes and mortality. By performing sub-analyses on those eligible to have survived =100 years, we will discern whether nutrition risk, quality of life and health outcomes are predictors of living to become a centenarian. In particular, our robust longitudinal dataset now spans more than two decades and will provide unparalleled opportunities for secondary analyses that explore nutrition risk in relation to additional health outcomes like dementia. By studying the oldest old we can identify potentially modifiable diet and lifestyle factors to promote healthy aging.
Progress Report
Mortality: Our most recent data confirmed 17,045 deaths out of the 21,645 (79%) original GRAS participants. Overall mortality was 76% at the time of our last reporting. This corresponds to an additional 542 participants with deceased status. There remain 328 original GRAS participants in active follow-up (defined as an encounter within the electronic health record occurring within the last two years).
Centenarians: As of last year’s report, there were 173 GRAS participants confirmed as centenarians and another 19 that were pending confirmation. We further validated our existing centenarians by comparing data within our electronic medical record to historical recruitment files. We were able to validate 184 centenarians that were selected for use in further analyses. We used a similar process to identify GRAS participants that could have reached 100 years old (born prior to 3/1/2023) but died before reaching 100 years old (n=5507). This cohort was selected for use as a resource for selecting a comparison group for use with the centenarian analyses.
In our recent exploratory analyses, we identified 178 centenarians and 5,431 non-centenarians without centenarian status at the study baseline (1994-1999), that also completed baseline Level II nutrition risk screening items. Unconditional logistic regression revealed that being male, gaining 10 or more pounds in the past 6 months, and using three or more prescription drugs daily, along with over-the-counter medications and/or vitamin/mineral supplements daily, had significantly lower odds of becoming a centenarian. However, eating alone and living alone at study baseline were associated with higher odds of reaching centenarian status. In sum, our preliminary analysis indicates that sex, weight gain, medication use, eating alone, and living alone are significant predictors of centenarian status among older adults in rural Pennsylvania. Further analyses will explore the relationship between nutrition risk variables and the likelihood of becoming a centenarian utilizing a case-control study design, matching centenarians with controls based on age and sex.
Other variables: Preliminary analysis using Fisher's exact tests were conducted to explore any meaningful associations between potential indicators of food insecurity in the Level II nutrition risk screening items obtained at study baseline and outcomes related to all-cause dementia and Parkinson’s disease. We identified significant associations between certain potential indicators including, "usually eating alone," "poor appetite," "housebound status," "lack of stove and/or refrigerator," "usually or always need assistance with preparing food," and "usually or always need assistance with shopping for food and other necessities" with all-cause dementia. Similarly, "usually eating alone" "usually or always need assistance with preparing food" and "usually or always need assistance with shopping for food and other necessities" demonstrated significant associations with Parkinson's disease. At the time the baseline data were collected (well over 20 years ago), valid measures of food security were not yet well developed. These variables, therefore, may not adequately reflect the current concept of food security. Future analysis of the association between food security and neurological outcomes in older adults using more rigorous methodology may be warranted.
We previously reported on 331 individuals that had a positive COVID-19 test of which 70 died within one month of the positive result. We continued review of these data but found them unsuitable for additional analyses, primarily due to implementation of vaccination programs part way through the data collection period and a lack of quality information on vaccination status. However, we will include the COVID data within our final GRAS research database and metadata document to be made available for data sharing (Sub-objective 2b).
Accomplishments
1. The key problem that our research has targeted is limited understanding of the relationship between nutrition risk and health outcomes among persons older persons of advanced age. For the accomplishment shared in this year’s annual report, our approach aimed to create a Data Discovery Team tasked to develop a process for making GRAS data available to external collaborators. The action taken was to constitute a Data Discovery Team comprised of representatives from Geisinger, USDA, and Tufts University. The Discovery Team conducted a series of meetings to discuss procedures for access to GRAS data, and helped to guide the development of a metadata document. Our results include a draft metadata document with data dictionaries and descriptive frequency tables for the data domains of the baseline questionnaires, the diet quality screening questionnaire, and the electronic medical record. The document also contains supplemental data reports to aid investigators in developing research questions and will eventually include details of the data sharing process. Customers of these findings include researchers, health professionals, government agencies, non-government agencies, and older persons.
