Orthostatic Hypotension as a marker of Frailty in Older People

Abstract

This doctoral investigation endeavoured to explore the association between orthostatic hypotension (OH) and frailty in community-dwelling older people, in order to assess the validity of to-date untested, clinically grounded claims that OH could be a marker of frailty in older people. Both OH and frailty are complex, heterogeneous clinical entities without universally agreed definitions. The operationalisations and clinical significance of both entities are reviewed in the first part of the investigation.

The investigation is based on comprehensive geriatric assessment data collected by a multidisciplinary team (including the candidate) at the Technology Research for Independent Living (TRIL) Clinic at St James?s Hospital Dublin between August 2007 and May 2009. A convenience sample of 442 community-dwelling subjects aged ? 60 years (without dementia or risk factors for autonomic neuropathy) was cross-sectionally studied. The orthostatic hemodynamic assessments were conducted via active stand tests with Fimometer?, a validated non-invasive beat-to-beat blood pressure monitor.

Based on Fimometer? data, the sample was classified according to five OH definitions: consensus OH (COH), Fedorowski et al.?s modification of COH (i.e. FOH), initial OH (IOH), a novel 3-group morphological classification based on decreasing systolic blood pressure (SBP) recoverability after standing (MOH), and a clinical definition based on symptoms of orthostatic intolerance (OI). Individual orthostatic hemodynamic variables were also used in the analyses. The comprehensive geriatric assessment data were used to construct two ordinal frailty classifications (i.e. non-frail, pre-frail and frail), one based on a modification of Fried?s phenotypes and another one on Rockwood?s frailty index approach (TRIL-FI). Appropriate bivariate statistics were used to correlate frailty and the OH definitions, and multivariable structural equation models (SEM) were used to assess the extent to which postulated causal relationships between variables were supported by the data.

Amongst the OH definitions considered, OI was the only significant marker of frailty. IOH was also associated with frailty, but this may have been due to the inclusion of OI in its definition. Impaired SBP recoverability was found as the hemodynamic hallmark of OI. The degree of SBP drop (i.e. delta) was the main predictor of SBP recoverability, but delta SBP itself had no independent correlation with OI or falls. SEM supported OI as a mediator between orthostatic hemodynamic changes and previous falls, but did not find orthostatic hemodynamic variables in independent association with falls. In the face of frailty (which had a significant correlation with previous falls), OI had only a modest (P < 0.05) independent association with previous falls.

Considered as a screening tool for the presence of pre-frailty or frailty, the presence of OI after standing and at least one fall in the last six months had, in the sample, a positive predictive value of 88.9% (modified Fried?s classification) and 96.3% (TRIL-FI). If externally validated, such a screening tool could be useful in primary care.

The findings of this cross-sectional exploratory study represent an original contribution to the understanding of the clinical relevance of beat-to-beat orthostatic hemodynamics in older people, and a methodological advancement in the area. Given the limitations of the research setting, findings warrant confirmation in a longitudinal context such as The Irish Longitudinal Study on Ageing (TILDA).