Share

PEBT

Faculty Research Fellowship Abstracts

 

"Piloting an Adaptive Agent Based Model to evaluate Frailty in the pre-elderly and elderly populations" by Andrew B. Phillips, PhD, RN

The elderly population is the United States places an increasing demand on both our economic and healthcare resources. Those over age 65 account for approximately 34% of healthcare spending, but only 14% of the population1. This disproportionate cost is reflective of the increasing number of chronic conditions we experience as we age. Changing this dynamic is important to improving the quality of life of our aging population and reducing the demand on our economic and healthcare resources. Advances in technology nutrition, medication, and genomics have served to improve our management and understanding of these disease processes. Other influencers of health such as our environment, social well-being, cognitive health, physical activities, and access to care represent further influencers of the aging process and health. Less understood are the ways in which these influencers interact and contribute to aging. How much does our environment contribute our health later in life? Are there “points of no return” during middle age where behavioral, medical, or other interventions become less effective? How do our behaviors and choices interact with physical, mental, and other health indicators to predict future health?

Frailty has been identified as a key measure of the decreased “physiologic reserve” that results in cumulative declines and increased vulnerability to chronic illness. While there are several proposed measures of frailty that have been validated against specific data sets, there are minimal studies on the influence to be given to the criteria that assess overall frailty risk. The result is a need for a more dynamic rather than static model of frailty that reflects the interrelationships among the current five frailty criteria (Shrinking, Weakness, Poor endurance and energy, Slowness, and Low physical activity) as well as the many potential physiological, cognitive, environmental and other factors that in combination contribute to frailty.

The science of Adaptive Agent Models (AAMs) are one method of dynamic evaluation using complex systems science. AAM is a bottom-up methodology that considers individual agent attributes such as age, sex, socioeconomic status, physical and mental health, etc., and the multiple interactions that take place among multiple agents, the environment, and the other “agents” within a complex system. An AAM has the ability to examine both individual behaviors and environmental characteristics that contribute to frailty and identify those that have the greatest influence. Our growing understanding of aging, combined with an increasing availability of large data sets and new computational methods, create an ideal opportunity to develop an AAM at this time.

The first step in achieving this goal is development of a Pilot AAM utilizing the method through replication of well documented frailty measures and outcomes. Validation of a simple Pilot AAM is a critical starting point for future models that have the capability to improve or replace existing frailty instruments and identify opportunities for early intervention in the aging process at the primary care level. The results from this Pilot should provide initial data on the relative influence of each of the five frailty criteria on frailty prevalence and mortality.

1. National Health Expenditures Fact Sheet. 2016.

“Regulation of IL-1β Production by Mcl-1” by John Wong, PhD

Interleukin-1 beta (IL-1β) is a cytokine involved in the immune response to microbial infection and the development of many human diseases. IL-1β is synthesized as a precursor in immune cells. The precursor is processed by a multimeric protein complex called the NLRP3 inflammasome to yield mature IL-1β that is secreted from the cell. Different kinds of danger signals (microbes, DNA damaging agents, toxins, etc.) stimulate immune cells to produce IL-1β by activating the NLRP3-inflammasome, yet how these danger signals cause inflammasome activation is not known. I hypothesize that Mcl-1 regulates the NLRP3-mediated activation of IL-1β production in mammalian cells by acting as an inhibitor. Mcl-1 fits the criteria of such a candidate because it has a short half-life, plays a role in cell survival, and belongs to a family that binds NLRP3-like members. In addition, my preliminary data demonstrated that Mcl-1 is degraded in response to stimuli that activates NLRP3. To test whether the hypothesis of Mcl-1 being the regulatory protein of the NLRP3-inflammasome is true, I will artificially increase and decrease the expression of Mcl-1 in mouse bone marrow-derived macrophages using lentivirus and siRNA, respectively. Understanding how the production of IL-1β is regulated may lead to effective treatments to block IL-1β production in many diseases.

"Infusing Robot-Assisted Therapy with Motor Learning Principles: An Active Learning Program for Stroke (ALPS)" by Susan E. Fasoli, ScD, OTR/L

Rehabilitation robots and passive gravity-assist orthoses provide clinicians with new treatment options to improve upper extremity (UE) motor capacity and performance after stroke. Systematic reviews of robot assisted therapy for the paretic UE confirm gains in motor capacity as measured by clinical assessments, but provide little evidence of improved UE performance during daily tasks and occupations.1,2,3 These findings may be attributed to the limited availability of rehabilitation robots to train the paretic hand and a primary focus on intensity of practice with little regard for other principles of motor learning and experience-dependent neuroplasticity.4,5 These principles, including the salience of training tasks, transfer of acquired skills to similar activities, and active engagement and problem solving, are key to task-oriented training paradigms in stroke (e.g. constraint-induced movement therapy) but have not been well integrated into robot-assisted therapy protocols. The transfer of robot-trained movements to UE activities within the home and community needs further exploration before widespread use in rehabilitation practice is expected.

The objectives of this pilot project are to 1) generate and test a feasible and reproducible protocol in which motor learning principles are more fully infused into robot-assisted therapy and 2) examine participant outcomes via measures across domains of the International Classification of Functioning (ICF).6 Specifically, we will develop a structured Active Learning Program for Stroke (ALPS) that promotes identification of interfering and changeable UE impairments; establishes clear patient-centered goals; encourages active problem solving and UE self-management; and facilitates self-efficacy and confidence in UE use via challenging and meaningful practice. We will provide participants with a home action plan that will combine ALPS training with patient-targeted treatment activities to encourage the transfer of acquired skills to activities in the home and community. Persons with moderate UE impairments more than six months post stroke will be randomly assigned to one of two intervention groups in which they receive ALPS training with either 1) robot assisted therapy (RT) alone, or 2) robot-assisted therapy plus therapist-guided task-oriented training (RTTOT).

Back to Top