Frailty: different tools for different purposes?
In this issue, Ravaglia et al. [1], report on ‘a frailty score’, which they have tested as a predictor of a range of adverse health outcomes which they characterise as ‘frailty outcomes’ in people aged 65+ years (mean age at baseline was 74.7) [1]. The outcomes were death (14.6% of the baseline sample), fractures, hospitalisation, and new onset or worsening of capacity to perform activities of daily living (ADL) during the next 4 years. Their model was arrived at through statistical exploration of data from a longitudinal population study of 1,016 individuals in Italy. The nine variables emerging after multivariate analysis were age >80 years, male gender, low physical activity, co-morbidity, sensory deficits, calf circumference <31 cm (taken as an indicator of sarcopenia), dependence in Instrumental ADL, a Tinetti gait and balance performance score
24, and pessimism about one's health compared to others. Together they produce a numerical score which at various cut-points tested was significantly predictive of the ‘frailty outcomes’. While acknowledging that their prognostic score is not yet adequately developed and requires at least to be tested in a remote cohort of older people, the authors suggest that their findings support further investigation of frailty scores for use in clinical geriatric practice. Consideration of the frailty concept and its utility for the geriatrician is certainly welcome. Indeed, in this journal, Rockwood challenged the clinical geriatrics community to recognise frailty as the central raison d'etre of the specialty [2] and went on to discuss the theoretical and empirical work needed to make a definition of frailty valid and useful [3]. Part of the work needed is to consider what frailty means from a range of perspectives. Within the clinical research community, particularly in the United States, considerable effort has been made in recent years to achieve a consensus [4, 5] and during 2005–06, the Journal of the American Geriatrics Society published 14 papers with frailty in the title. Further progress was reported from the Second International Working Meeting on Frailty and Aging held in Montreal in 2006, leading to several fascinating papers, setting out a spectrum of approaches, published in July 2007 in a special section of the Journals of Gerontology [6–9].
The shared central notion about frailty is of an older person who is at heightened vulnerability to adverse health status change. The phenotype fitting this has been operationalised with five components—unintentional weight loss, self-reported fatigue and diminished physical activity, and measured impairment (comparative to age-standardised norms) of grip strength and gait speed [10]. Although this omits cognitive and psychosocial features which are well known to be predictive of adverse health outcomes such as loss of functional ability, institutionalisation and death, this physical frailty phenotype has demonstrated predictive power for adverse health outcomes in several cohorts of older people [10, 11]. This highlights two of the contesting concepts about frailty.
The first concept is of a syndrome associated with underlying physiological and metabolic changes, which may be inter-related and which are responsible for driving progressive physical and cognitive impairments through to loss of functional capacity, often helped on the way by acute or chronic disease or injury. This approach was encapsulated in a definition of frailty proposed in this journal a decade ago—‘a condition or syndrome which results from a multi-system reduction in reserve capacity to the extent that a number of physiological systems are close to, or past, the threshold of symptomatic failure. As a result, the frail person is at increased risk of disability or death from minor external stresses’ [12]. This emphasises the pathophysiological causal pathway. Subsequent work has thrown up a number of candidates to occupy this pathway, such as cytokines and other components of the inflammatory response [13–15].
An alternative approach has been to treat frailty as a collective of risk factors for future adverse events, while not necessarily bearing any pathophysiological relationship to these outcomes. This more pragmatic approach is about identifying people at higher risk of pre-defined adverse outcomes. This could be done for a range of purposes and in a range of contexts and these considerations will influence the measures which are suitable, the thresholds for defining high risk (and therefore by implication frailty) and the time interval between the measurement and the outcome. Only the most general concept of frailty could retain this flexibility.
The first approach, of a pathophysiologically driven syndrome, is supported by evidence that differentiates it from ageing. Whether frailty changes are inevitable, as long as premature death does not intervene, remains uncertain [9] but it seems that frailty is also more than just the sum total of the effects of identifiable chronic diseases. While a number of studies have shown that frailty is more prevalent in the oldest of the old and in those with the greatest burden of co-morbidity, chronic disease and frailty do not map exactly within the older population and thus measure different ideas. Chronic disabling conditions such as cardiac failure and obstructive pulmonary disease are associated both with metabolic features proposed as components of the pathophysiological frailty model and with clinical features present in the operational model.
Distinguishing frailty from both ageing and co-morbidity has potential clinical utility. If underlying metabolic processes could be identified, then would it become possible to intervene at a preclinical stage of frailty before the operational phenotypic criteria are evident? Further, would addressing the frailty aspects offer an additional clinical approach to the management of patients with chronic diseases? Disease-specific factors do not fully explain well-being and quality of life [16], and frailty may contribute independently of disease. Comprehensive geriatric assessment already encompasses an approach which combines disease specific and non-specific aspects to the assessment and treatment of older people. Frailty recognition would be a refinement of this approach.
Recognition of frailty through better definition may also improve clinical decision making by informing the prediction of benefit or the risk of adverse effects of clinical interventions including medications, surgical interventions and physical displacement etc. For example, the ability to improve prediction of post-operative functional recovery would be invaluable, as disease-based predictive models are far from perfect [17]. At the level of the individual patient, laboratory tests and clinical assessments might each contribute to a ‘risk profile’. The challenge then for 21st-century geriatrics would be to design and test clinical approaches which both modify standard interventions and address the specific underlying frailty processes.
The pragmatic risk factor approach could also be employed clinically at the individual patient level but it is a large step to translate prediction from large datasets which work epidemiologically at the population level to the precision needed in clinical practice. For planning health services and deciding in broad terms where and for whom to apply health preventative interventions, the larger population predictive ability may suffice. For this sort of approach, self-reported measures or simple assessments may be suitable, rather than laboratory investigations or individual detailed clinical measurements such as a gait speed. Clearly, this concept of frailty and the criteria necessary and feasible to define it are quite different from those which might be necessary in clinical practice.
So we end up with a range of perspectives. For the clinical geriatrician, identifying frailty could facilitate clinical decision making, improving estimates of benefit and risk through a quantitative assessment of frailty. This would be a refinement of what clinicians believe they already do intuitively. To improve on this, reliability and predictive accuracy are important and we are some way from achieving this. For the research clinical gerontologist, identifying the underlying biological drivers of frailty and their relationship to age-related usual changes could enable targeted research into the causes of frailty, perhaps with identification of a pre-frail population, and the possibility from this of developing preventative interventions. Animal studies may be a stage in this process. For public health purposes of planning future needs or designing population-level interventions, then once again the focus is different. Being able to identify, using simple methods, those among the older population who are at increased risk of repeat hospitalisations or functional dependency may enable more effective use of resources such as community-based case management. The assessments suitable for this purpose would likely involve different parameters from those which might be useful for individuals in clinical care, e.g. for pre-surgical assessment.
In the light of this, how does the article by Ravaglia et al. help? Their aim was to test a range of measures feasible in clinical practice to predict ‘frailty outcomes’ over a long timescale. The authors do not propose a pathophysiological line of causation between the attributes that they assessed at baseline and the outcomes experienced by the patients. That was not their purpose. So this study does not clarify the pathophysiological nature of the frailty syndrome. The variables do not seem readily applicable to population-level approaches in primary medical care or social care as a tool for identifying an at-risk population. Their use would be feasible in specialist clinical practice, but here we are usually interested in shorter-term outcomes, not those occurring over 4 years. Further, do the data suggest sufficient predictive accuracy? As a worked example, consider the cut-off score of three for predicting hospitalisation (Table 3). The sensitivity is 0.51 and specificity is 0.75, and positive predictive value is 25%. Whether these properties are useful would depend on the nature, expense, acceptability, etc of whatever intervention may be applied to the identified at-risk population. A first step would be to design an intervention and demonstrate some efficacy. A next step would be to establish whether the parameters used to predict the adverse outcomes also predicted the potential to benefit from an intervention.
So, as the authors conclude, there is need for ‘further investigation of frailty scores including easy-to-collect indicators for settings in which a complex and time-consuming examination might be impractical’. We suggest that a vital step in designing such work is to first consider carefully the purpose of the frailty score in question.
Department of Ageing and Health, Guys and St Thomas' NHS Foundation Trust, London, UK
* To whom correspondence should be addressed E-mail: finbarr.martin{at}gstt.nhs.uk
References
- Ravaglia G, Forti P, Lucicesare A, Pisacane N, Rietti E, Patterson C. Development of an easy prognostic score for frailty outcomes in the aged. Age Ageing (2008) 37:161–66.
[Abstract/Free Full Text] - Rockwood K, Hubbard R. Frailty and the geriatrician. Age Ageing (2004) 33:429–30.
[Free Full Text] - Rockwood K. What would make a definition of frailty successful? Age Ageing (2005) 34:432–4.
[Abstract/Free Full Text] - Ferrucci L, Guralnik JM, Studenski S, et al, Interventions of Frailty Working Group. Designing randomized controlled trials aimed at preventing or delaying functional decline and disability in frail older persons: a consensus report. J Am Geriatr Soc (2004) 54:625–34.
- Walston J, Hadley EC, Ferrucci L, et al. Research agenda for frailty in older adults: Toward a better understanding of physiology and etiology: summary from the American Geriatrics Society/National Institute on Aging Research conference on frailty in older adults. J Am Geriatr Soc (2006) 54:991–1001.[CrossRef][Web of Science][Medline]
- Kuh D, Sayer AA, Ben Shlomo Y, et al. A life-course approach to health aging, frailty and capability. J Gerontol A Biol Sci Med Sci (2007) 62A:717–21.
[Free Full Text] - Whitson HE, Purser JL, Cohen HJ. Frailty Thy name is ... Phrailty? J Gerontol A Biol Sci Med Sci (2007) 62A:728–30.
[Free Full Text] - Bergman H, Furrucci L, Guralnil J, et al. Frailty: An emerging research and clinical paradigm—issues and controversies. J Gerontol A Biol Sci Med Sci (2007) 62A:731–7.
[Abstract/Free Full Text] - Rockwell K, Andrew M, Mitniski A. A comparison of two approaches to measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci (2007) 62A:738–43.
[Abstract/Free Full Text] - Ensrud KE, Ewing SK, Taylor BC, et al. Frailty and risk of falls, fracture and mortality in older women: The study of osteoporotic fractures. J Gerontol A Biol Sci Med Sci (2007) 62A:744–51.
[Abstract/Free Full Text] - Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci (2001) 56:M146–56.
[Abstract/Free Full Text] - Campbell AJ, Buckner DM. Unstable disability and the fluctuations of frailty. Age Ageing (1997) 26:315–8.
[Free Full Text] - Puts MTE, Visser M, Twisk JWR, et al. Endocrine and inflammatory markers as predictors of frailty. Clin Endocrinol (2005) 63:403–11.[CrossRef][Medline]
- Abate M, Di Iorio A, Di Renzo D, et al. Frailty in the elderly: the physical dimension. Eur Medicophysica (2006) 43:407–15.
- Leng SX, Xue QL, Tian J, et al. Inflammation and frailty in older women. J Am Geriatr Soc (2007) 55:864–71.[CrossRef][Web of Science][Medline]
- Yohannes AM, Roomi J, Waters K, et al. Quality of life in elderly patients with COPD: measurement and predictive factors. Respir Med (1998) 92:1231–6.[CrossRef][Web of Science][Medline]
- Hamel MB, Henderson WG, Khuri SF, et al. Surgical outcomes for patients aged 80 and older: Morbidity and mortality from major noncardiac surgery. J Am Geriatr Soc (2005) 53:424–9.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. E Dupre, D. Gu, D. F Warner, and Z. Yi Frailty and type of death among older adults in China: prospective cohort study BMJ, April 9, 2009; 338(apr09_2): b1175 - b1175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Hubbard, M. S. O'Mahony, and K. W. Woodhouse Characterising frailty in the clinical setting--a comparison of different approaches Age Ageing, January 1, 2009; 38(1): 115 - 119. [Full Text] [PDF]
|
|
|
|
|
|
G. Ravaglia, P. Forti, A. Lucicesare, N. Pisacane, E. Rietti, and C. Patterson Reply Age Ageing, July 1, 2008; 37(4): 484 - 485. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||