Reserve and resilience in CKD: concept introduction and baseline results from the Physical REsilience Prediction in Advanced REnal Disease (PREPARED) study

In contrast to the well-established construct of psychological resilience, defined as adaptive attitudes and behaviors in response to stressful life events [12, 13], physical resilience is an emerging area of aging research [10]. Physical resilience characterizes one’s physical ability to overcome health stressors and is supported by the recognition that while some older adults “snap back” after an acute health event, many others do not recover physical function quickly or at all [11, 14]. An existing model of physical resilience describes the role of both reserve and resilience in recovering from health events (Fig. 1). In this model, health stressors include acute health events that may impact function, such as an illness or injury that results in a hospitalization or ED visit. Health stressors often result in increased physical and cognitive demands. To meet the added demands of a health stressor, older adults may draw on their reserve defined as one’s physical, cognitive, or psychological capacity and the social and environmental resources available before a health stressor. The term reserve emphasizes that these abilities are not already in use, but are available when needed. In the model used here, resilience is defined by the trajectory of recovery after a health stressor. Resilience is evident when older adults avoid health stressor-related functional loss (i.e., resist) or recover (i.e., snap back) to pre-health stressor levels. Common clinical scenarios in which resilience, or lack of thereof, is evident include when some older adults recover from very severe illness, or some recover better than expected given an apparent limited pre-stressor reserve, or yet others have unexpected functional decline after minor health events. These scenarios suggest that all three factors – health stressors, reserve, and resilience – play a role in functional decline and recovery. These insights informed the PREPARED study of reserve and resilience that includes 1) measures of multiple domains of reserve before a health stressor, 2) prospective identification of health stressors with information on timing and severity, and 3) longitudinal measures of function, both before and after a health stressor, to capture functional decline and recovery [15].

PREPARED is a national, prospective cohort study of Veterans with advanced CKD designed to characterize physical resilience. PREPARED uses data from the Department of Veterans Affairs (VA) Health Administrations health system data paired with telephone surveys. The VA Corporate Data Warehouse (CDW) is a national repository of clinical and administrative data that is updated daily and provides detailed information on health care utilization. Primary data were collected through telephone surveys occurring at the time of study enrollment (baseline) to assess reserve and every 8 weeks for up to 32 weeks (8-, 16-, 24-, and 32-week calls) to characterize changes in physical function. On a weekly basis, the study team queried CDW to identify enrolled participants who experienced a health stressor defined as an acute care encounter (ED visit or hospitalization; Fig. 2). Those who had an ED visit without an associated hospitalization were flagged as having a health stressor with the date corresponding to the ED visit. For those hospitalized following the ED visit, we used the hospital discharge date so that the ED visit and hospitalization were considered a single health stressor event. Participants were then contacted within 3 to 14 days of the health stressor and every 8 weeks for 2 additional calls in order to capture their post-stressor functional trajectory.

The study cohort comprised older adult Veterans with CKD and at risk for functional decline. Eligible Veterans were identified via CDW who met the following inclusion criteria [1]: age ≥ 70 years old [2], estimated glomerular filtration rate (eGFR) < 30 ml/min/1.73 m² on two occasions at least 90 days apart without an intervening eGFR ≥30 ml/min/1.73 m² [3], prior nephrology consultation or outpatient referral, and [4] high risk for hospitalization based on the VA Care Assessment Need score. We restricted our study population to those with VA laboratory-based eGFR < 30 ml/min/1.73 m² estimated using the CKD-EPI equation (Stage G4 or greater) and prior nephrology consultation in order to identify those most likely to experience CKD-related complications or have had discussions about kidney failure treatment. We restricted our study population to those with a 90-day hospitalization CAN score ≥ 95 in order to identify a group at higher risk for experiencing a qualifying health encounter during follow-up [16, 17].

To focus our study on Veterans with pre-dialysis kidney disease, we excluded those with a history of kidney transplant or those on dialysis. Those with cognitive impairment based on a 6-item telephone screen [18], and those residing in a nursing home or receiving hospice care prior to enrollment were also not eligible for participation. We followed a pre-determined protocol to identify and address hearing limitations in order to avoid excluding those with hearing impairment [19]. We asked all participants to identify and provide permission to contact a proxy if they were no longer able to participate during follow-up.

Enrollment began in October 2019. This study was approved by the Durham VA Institutional Review Board (IRB #2205).

We calculated counts (frequencies) and medians (25th, 75th percentiles) for participant characteristics and baseline measures of reserve. To describe overlapping prevalence of low reserve across the many measures we collected (i.e., from cognitive, functional, and psychological domains), we created an UpSet plot for the following seven measures: life-space mobility, ADLs, IADLs, cognition, depressive symptoms, symptom burden, and social support. Results from an UpSet plot are similar to intersecting circles of a Venn diagram, but the UpSet plot is easier to interpret when there are more than three variables included. For each measure, we used the quartile to create a cut-point for the lowest level of reserve (e.g., 75th percentile for ADLs indicates greater difficulty and the 25th percentile for TICS-m indicates worse cognitive function) relative to other PREPARED participants. The UpSet visualization uses dots connected by a line to show what domains are in each intersection. This is displayed at the bottom of the figure. For example, those with only one domain of low reserve will have a single dot. In contrast those with the combination of all 7 domains of low reserve will have seven dots connected by a solid line. The UpSet visualization also displays the number of participants within each intersection (vertical bars labeled “Number of Participants”).