Frailty

The Broad View

Matteo Cesari, Olga Theou

Demographic trends show that independently of geographic regions and socioeconomic background, the absolute and relative number of older persons is increasing worldwide. It has been estimated that the prevalence of persons aged 65 years and older worldwide will increase from 7.7% in 2010 to 15.6% in 2050. This trend is evident even among the subpopulation of those aged 80 years and older, which is predicted to more than double in size from 2010 to 2050 (1.6% and 4.1%, respectively).This is not exclusive to more developed countries, because similar trends have even been reported in the least developed regions of the world.¹ The reduction of mortality risk at advanced age is largely the result of scientific advancements and improvement of life conditions.² At the same time, as the aging population seeks care, simply extending current practices is threatening the sustainability of the health care system. In particular, current models of care do not sufficiently take into account the new (and still unmet) needs of the changing population. Not surprisingly, a recent alert by the Royal College of Physicians (London) indicated the necessity of “more consultants with skills in acute, general and geriatric medicine to be able to cope with the ageing population.”³

One of the major challenges of health care systems is to face the severe burden imposed by disabling conditions of old age.⁴ Because disability in the older person has to be considered as an almost irreversible condition, because it is largely caused by lifelong accrual of deficits,⁵ attention must be focused on preventing the disabling cascade⁶ and on managing people in ways that aim to mitigate, or at least not add to, their level of dependence.

During the past 2 decades, a large body of evidence has been developed to identify biomarkers and instruments capable of measuring the biologic age of the older individual. Given the increasing number of older persons requiring assistance and the high economic burdens imposed by disabling conditions, the identification of “biologically aged” (rather than “chronologically aged”) subjects is today both necessary and urgent.⁴^,⁷

The need to reshape the traditional idea of the geriatric patient is obvious. The commonly adopted criterion of chronologic age based on the number of years lived is no longer sufficiently selective for identifying the right target population requiring adapted care and special resources. It is necessary to replace the term chronologic age with a parameter capable of measuring the biologic status of an individual more in-depth.

The age criterion is not the only thing that needs to be redefined in the clinical and research settings to develop the optimal care of older adults. A significant and important transition from disease-oriented toward function-oriented medicine is necessary. In fact, with advancing age, the meaning of traditional diseases becomes lost, because these conditions are largely modified by the effects of the aging process. This issue is at the very basis of the so-called evidence-based medicine issue in geriatrics, largely because the social, clinical, and biologic characteristics of older adults do not reflect those on which international recommendations and guidelines are developed.⁵

The frailty concept might offer a solution in this context. Fried and coworkers have provided a perfect description of the role played by frailty. They stated that “the cornerstone, even the raison d’etre, of geriatric medicine concerns the identification, evaluation, and treatment of frail older adults and prevention of loss of independence and other outcomes for which they are at risk.”⁸ It is on this type of patient that the field of geriatric medicine has built up its knowledge and developed its own specific methodology—the comprehensive geriatric assessment.⁹^,¹⁰ This chapter focuses on the how frailty is defined and assessed, how we can treat frailty, and why it needs to be considered within our health care system.

Frailty

To better appreciate the heterogeneous health status of the older persons, the frailty concept was introduced in geriatric and gerontology literature about 20 years ago. Frailty is now noncontroversially understood as the concept of increased vulnerability to adverse outcomes among people of the same chronologic age. It is the term used to indicate the geriatric syndrome or state characterized by a reduction of the organism’s homeostatic reserves. The lower capacity of the organism to face entropic forces (coming from endogenous and exogenous sources) exposes an individual to an increased risk of negative health-related events, including falls, hospitalizations, worsening disability, institutionalization, and mortality.⁸^,¹¹^–¹³ In a frail individual, a clinically irrelevant endogenous or exogenous stressor may become the trigger for the initiation of the burdening disabling cascade.¹³

Frailty has been defined by an international consensus of experts as “a multidimensional syndrome characterized by decreased reserve and diminished resistance to stressors.”¹¹ A widely accepted definition of frailty was provided in Orlando, Florida, by an international consensus group in 2012. It stated that frailty is “a medical syndrome with multiple causes and contributors that is characterized by diminished strength, endurance, and reduced physiologic function that increases an individual’s vulnerability for developing increased dependency and/or death.”⁶

In a recent review paper, Clegg and colleagues clearly depicted frailty not as a syndrome but as a state of vulnerability that challenges the maintenance of homeostasis in older persons.¹² Similarly, the deficit accumulation approach sees frailty as a multidimensional risk state that can be measured by the quantity rather than by the nature of health problems.¹⁴ This approach proposes that frail older adults have many things wrong with them; the more things they have wrong, the higher the likelihood that they will be frail and the greater their risk of adverse health outcomes. Based on this definition, frailty arises from a multisystem decline, which compromises the body’s ability to repair damage that arises externally or as the byproduct of internal processes, (e.g., metabolic, respiratory, inflammatory), including genetically induced damage.¹⁵

However, frailty is not all or none; grades of frailty make a difference. Still, many studies classify people simply as frail or nonfrail. In some settings, such as comparing frailty prevalence in different samples, this can be useful; however, even in this case, important information gets lost. Many clinical decisions require greater precision than a nonfrail-frail status.¹⁶ In addition, frailty is a dynamic process where transitions across states of frailty are common. On average, health tends to decline with age, and the population-based trajectories of frailty are consistent, showing acceleration in deficit accumulation. The frailty index increases, on average, tenfold between 20 and 90 years. Even so, individual trajectories of the frailty index are generally irregular, showing that frailty reflects a stochastic dynamic process. For an individual, most transitions are gradual, and the likelihood to change their frailty level is largely conditioned on their previous frailty level. Therefore, transitioning from a nonfrail state to a severely frail state (and vice versa) is not very common.¹⁷ In individuals, including older adults, frailty levels increase nonmonotonically over time; however, health status can improve, which will result in a transition from a higher to a lower frailty level state.¹⁸

In a study conducted by Gill and associates,¹⁹ nondisabled subjects aged 70 years and older were followed over time to monitor changes in their frailty status. Among the 754 participants, more than half (57.5%) experienced at least one transition across any of the frailty states during the 54-month follow-up period. It was also reported that 44.3% of robust participants at the baseline transitioned to a prefrailty (40.1%) or frailty (4.2%) condition during the first 18 months of follow-up. Among participants presenting as frail at baseline, most (63.9%) remained frail, whereas 23.0% improved to a healthier condition, and 13.1% died. Consistent results showing the positive and negative changes of the frailty condition over time have also been reported in data collected as part of the Survey of Health, Ageing, and Retirement in Europe (SHARE).²⁰ Recent studies have started exploring which conditions may be associated with improvement or worsening of the frailty condition. For example, Lee and coworkers²¹ reported the significant association of specific characteristics with negative modifications (e.g., older age, history of cancer, hospitalization events, chronic obstructive pulmonary disease, cerebrovascular disease, osteoarthritis) and positive modifications (e.g., higher cognitive function, absence of diabetes, higher socioeconomic status, no history of cerebrovascular disease) across frailty states. Even though these characteristics can accelerate frailty, typically frailty develops slowly, even insidiously, and can vary in important ways among individuals.

It is well established that although they frequently coexist in the older person, frailty, comorbidity (the concurrent presence of two or more medically diagnosed diseases), and disability (the difficulty or dependency in carrying out activities of daily life) are distinct conditions.²² For example, in the Cardiovascular Health Study, Fried and colleagues²² reported that comorbidity, disability, and the two together coexist in 67.7%, 27.2%, and 21.5% of frail participants, respectively. On the other hand, people who meet the criteria for the frailty might do so in the absence of comorbidity and disability, as reported in 26.6% of cases in the same study.

Recently, the concept of “resilience” (the individual’s ability to adapt in the face of stresses and adversities) has become increasingly used in parallel with frailty.²³ Although resilience is still is not adequately framed and defined, it is possible that it might be a field of promising research in the near future. Low versus high resilience may indeed make a difference and explain why individuals with the same frailty status follow opposite trajectories (i.e., toward disability and robustness, respectively). Although resilience is still only a briefly stated concept, its quantification may provide important insights in the assessment of the older person’s risk profile. As with the frailty syndrome, resilience also results from the complex network of biologic, clinical, social, and environmental factors characterizing each individual. We also need to consider that some degree of variability in health outcomes is to be expected in relation to frailty. Some people who are severely frail can survive in highly protective environments, whereas some nonfrail people will die. This could be related to the degree of damage to which an individual is exposed or to the resources available to assist with the repair of that damage. Even in Canada, a higher income country with a universal health care system, among individuals who are considered the fittest (lowest level of frailty), 5-year mortality is doubled among those classified with high social vulnerability levels compared to those with the lowest social vulnerability levels.²⁴

Prevalence of Frailty

In a recent systematic review,²⁵ Collard and associates provided estimates of frailty prevalence, analyzing data from 21 cohort studies (>61,500 community-dwelling older persons). The reported prevalence substantially varied across the studies examined, ranging from 4% to 59.1%, according to the adopted operational definition of frailty and the characteristics of the studied sample. Nevertheless, when analyses were restricted to studies using the most common model of operationalization (i.e., the frailty phenotype proposed by Fried and coworkers²⁶), the weighted average prevalence rate was 9.9% (95% confidence interval [CI], 9.6% to 10.2%) for frailty and 44.2% (95% CI, 44.2% to 44.7%) for prefrailty, whereas in the only study included that used the frailty index, the prevalence was 22.7%.²⁵ Similarly, using data from SHARE, it was reported that 11% of Europeans older than 50 years were identified as frail based on the frailty phenotype approach, and 21% were frail based on the frailty index approach.²⁷

Available evidence also tends consistently to report different prevalence of frailty according to age,²⁸^–³⁰ gender (e.g., higher estimates in women compared to men),²⁵^,²⁸ ethnic groups (e.g., higher prevalence in Hispanic and African Americans),²⁶^,³¹ migrant groups,³² individual socioeconomic characteristics (e.g., poor education and poverty are closely associated with frailty),²⁶^,³³^,³⁴ and macro socioeconomic factors (e.g., gross domestic product and health care expenditures of country of residence).³⁵

Biology of Frailty

Frailty has been described as a phase of acceleration occurring during the aging process due to endogenous and exogenous stimuli.³⁶ It results from the age-related cumulative declines occurring across multiple physiologic systems.⁵ The biology of frailty has its origins in the most intimate roots of the aging process. The parallelism between aging and frailty implicitly leads to the existence of a shared pathophysiologic substrate between the aging process and frailty.⁵

Such hypotheses can easily find support in the growing body of evidence showing that the same pathways indicated as crucial for the aging process (e.g., inflammation, oxidative damage, immune function, telomeres, natural selection) also represent key determinants in the development and maintenance of the frailty phenotypic syndrome.³⁷^–⁴¹ Furthermore, it cannot be ignored that specific innate capacities (e.g., mobility⁴²) characteristic of living beings across species (from Drosophila to humans) are strongly correlated with frailty and age-related conditions.⁷

Based on the deficit accumulation approach, frailty arises from the accumulation of microscopic damage (cellular and subcellular deficits) that are not repaired or removed and may reach macroscopic deficits—clinical detectable deficits at the organ and system levels. As organ level deficits accumulate, they may give rise to symptoms or signs, thereby presenting as clinically evident disease.⁴³ Also, damage in one organ system may predispose to damage in another organ system, showing that deficit accumulation and repair are intertwined. A recent study showed this association between the clinical macroscopic and subclinical microscopic deficit accumulation using a frailty index constructed by routine laboratory data.⁴⁴ This supports the notion that frailty that is macroscopically detectable represents the buildup of subcellular, tissue, and organ deficits from damage that is not removed or repaired.

Assessment of Frailty

Although the theoretical foundations of frailty are well established and largely agreed, controversies exist about its operational assessment. Multiple instruments have been developed over the years to capture in a standardized way the presence of frailty objectively. Unfortunately, the available instruments (stemming from different perspectives and purposes) are all predictive of negative health-related outcomes, but have modest agreement among them.²⁷^,⁴⁵^,⁴⁶ Analyses conducted by van Iersel and colleagues⁴⁷ compared the prevalence of frailty using four different defining tools—frailty phenotype, frailty index, usual gait speed, and hand grip strength. Results showed that the prevalence of frailty was different using different criteria. Moreover, there was only partial overlapping among the populations classified by these scales as frail. A study using SHARE data compared eight frailty scales and showed that among all scales, about half of participants (49.3%) were categorized as nonfrail and 2.4% as frail, and for 48.3% of participants, the eight scales did not classify them identically as frail or nonfrail.²⁷ In other words, every assessment tool was capturing a different risk profile, and none of them was completely exhaustive by itself. Such heterogeneity of results is in line with the nature of frailty but, at the same time, may require special caution when asked to choose one single operational definition. It is possible that the choice of the most appropriate frailty instrument should rely on the purpose of the evaluation, outcome for which the definition was originally validated, validity of the tool, population studied, and setting where the assessment will be conducted.

The most commonly used and widely diffused instruments for measuring frailty are the frailty index¹⁴ and the frailty phenotype.²⁶ Building on these two models, several other frailty scales have been proposed during the past several years.⁴⁸^–⁵³ Also, performance-based measures have been used to predict adverse health outcomes in older adults, suggesting that they could also be used as frailty screening tools.

Frailty Index

The frailty index was proposed by Rockwood and associates and was initially validated in the Canadian Study of Health and Aging.¹⁴ This instrument is designed to measure the age-related deficit accumulation using a mathematical approach. It is the ratio between the number of deficits that a person may experience (e.g., signs, symptoms, diseases, disabilities) and the total number of considered deficits (e.g., someone with 20 deficits out of 40 counted has a frailty index score of 20/40 = 0.5). In this way, the frailty index score is continuous (ranging from 0 to 1), and the higher the score of an individual, the more likely that this person is vulnerable to adverse health outcomes.

This approach proposes that in terms of understanding system behavior, knowing what exactly is wrong is less important than knowing how many things people have wrong. The index can be built up without the need of special resources, and not every frailty index needs to include the same items. In this way, the items included in a frailty index should not be included or excluded a priori as long as items are age-related, associated with adverse outcomes and, when combined, capable to cover several organ systems.⁵⁴ Given that the frailty index is not based on preset items, it is particularly useful for being applied to existing cohort data collected for different purposes a retrospective according to available data. At least 20 items should be considered to generate a frailty index; 30 or more is preferred to achieve stable estimates. Regardless of the nature and number of items included in the frailty index and whether the sample included community, institutionalized, or hospitalized older adults, it has remarkably similar measurement properties and substantive results.

As discussed earlier, one strength of the frailty index approach is that it can be developed from any existing biomedical database and even constructed solely by self-reported items. Because it requires the inclusion of at least 20 items, clinicians are often skeptical about its feasibility. However, in the era of the electronic health record, frailty indices can be constructed in clinical settings using routinely collected data, without requiring additional assessment. For example, a frailty index can be constructed based on a standard comprehensive geriatric assessment derived from a clinical examination⁵⁵ or a questionnaire that can be completed by care partners.⁵⁶ Also, recently, a frailty index was constructed using routine blood tests plus measured systolic and diastolic blood pressures.⁴⁴ Interestingly, the frailty index has been used as a model for developing frailty instruments in animal models. Following the approach of counting the accumulation of deficits for measuring the frailty condition, studies have replicated the frailty index in mice.⁵⁷^–⁵⁹ Such an extension of the frailty index implicitly provides the opportunity to develop frailty research in the preclinical setting and in translational research. Moreover, it provides further proof of the biologic substrate characterizing the validity of the instrument.⁶⁰

Rockwood and coworkers also generated a screening instrument (largely based on the clinical judgement of the physician) to indicate the different stages of the patient’s frailty, intended as poor health status.⁶^,⁶¹ The Clinical Frailty Scale is a seven- or nine-item scale that can indicate the global clinical status of the older person based on the evaluation of his or her status in the domains of mobility, energy, physical activity, and function. Future studies need to investigate how these frailty scales compare to other commonly used clinical instruments, such as the Karnofsky Performance Status⁶² or Eastern Cooperative Oncology Group (ECOG) Performance Status⁶³ scales.

Frailty Phenotype

Comparatively, the frailty phenotype was designed by Fried and colleagues and was initially validated in the Cardiovascular Health Study.²⁶ It is based on the evaluation of five defining criteria—involuntary weight loss, exhaustion, slow gait speed, muscle weakness, and sedentary behavior. Each criterion was operationalized following an epidemiologic approach that took advantage of the Cardiovascular Health Study database. The phenotype distinguished three consecutive states, robustness (defined as the absence of criteria), prefrailty (presence of one or two criteria), and frailty (when three or more criteria are present).

In theory, the frailty phenotype finds its best application in nondisabled older persons. Looking at signs and symptoms, it generates a preclinical risk estimation primarily aimed at indicating the individual in need of a comprehensive geriatric assessment.⁶⁰ Therefore, its design closely resembles that of a screening tool. However, such a characteristic also represents a major limitation failing of not being informative about the potential causes of the condition of interest. For example, the involuntary weight loss criterion might be due to social isolation, unknown diseases, or unhealthy behavior, with different causes requiring completely different interventions. As is evident, the phenotype is particularly focused at exploring the physical domain of the older person to capture the overall risk profile for negative outcomes. Such an approach has been proposed in the literature, because potentially limiting to a specific function or domain is thought to be insufficient for capturing the heterogeneity, multidimensionality, and complexity of the frailty condition. For example, several groups of researchers have discussed the need for extending the frailty phenotype to other health domains, which may improve the assessment of the risk profile in the older adults (e.g., cognition, mood, social status).⁶⁴ Although the frailty phenotype is commonly referred to as the instrument most adopted for assessment of the syndrome, such a statement might be arguable. It is very rare to have the original phenotype applied without adaptations in the definition of the five criteria and/or modifications of the thresholds of risk. Such qualitative and quantitative deviations from the version proposed by Fried and associates²⁶ are often necessary for adapting the assessment to the available resources and data, and/or the population being studied.

Only gold members can continue reading. Log In or Register to continue