Dit stuk gaat over het begrip "Frailty", dat wellicht overeenkomt met "kwetsbaarheid", en dat in een maat uitgedrukt aangeeft welke patiënten meer risico lopen dan anderen.

Het is uit 1999 en komt uit Annals of Internal medicine. De noten (141) zijn voor de leesbaarheid weggelaten.

Toward an Understanding of Frailty

Health care providers who serve an aging population inherently associate the word frailty with patients whom they perceive as frail. Although frailty is not confined to elderly persons, it is more frequently observed in persons older than 85 years of age (the "oldest old") (1). This is because limitations and diseases associated with aging are an inseparable part of frailty (2-5). As a result, frailty remains more a constellation of many conditions than a discrete clinical entity, and it certainly does not have a precise scientific meaning (6). Frailty is often identified by words that express a similar state, such as "feeble," or, as the French describe it, "fragilité" (7). Speechly and Tinetti (8) contrasted "frail" with "vigorous" in a study of elderly community dwellers.

These limitations in defining frailty make its actual prevalence uncertain. Strawbridge and colleagues (9) noted the breadth of the conceptual range of frailty: On the one hand, it might include one or more conditions associated with old age (and, thus, most older persons would be frail) (10); on the other hand, it might be limited to severely disabled persons (in which case only a small proportion of older persons would be frail). But even if frailty is not universal and represents a subset of the many persons who have attained advanced age and remain vigorous, it nevertheless assumes an importance out of proportion to its prevalence. There is a high toll in personal suffering, caregiver burden, costs of health care, medication use, and hospitalization, all of which signify enhanced health care demands by any measure of health resources utilization.

Defining frailty may be like attempting to "know the dancer from the dance"; nevertheless, I attempt to present an understanding of frailty by describing many of its clinical correlates and considering evidence for its biological underpinnings. Frailty might emerge more sharply as a clinical entity if it were described in biological terms. Evidence indicates that many of the clinical states associated with frailty are related to the pathophysiologic effects of an altered metabolic balance, manifested by cytokine overexpression and hormonal decline. Currently studied serum markers that reflect this imbalance lack specificity thus far but may provide a way to identify frailty at its incipient stages and differentiate it from associated comorbid conditions. This would stimulate new therapeutic interventions and ultimately promote primary prevention. The potential to ameliorate frailty is an important health care goal that has enormous societal implications; it could add to the quality of the later years of life and perhaps extend the life span.



Clinical Correlates

Comprehensive reviews of frailty have been published recently (2, 3, 5, 7, 9, 11-13). A MEDLINE search revealed that frailty was cited in 335 articles (13) from 1989 to 1992 and in 563 articles (3) from 1992 to January 1996. Although Brown and coworkers (7) defined frailty as "diminished ability to carry out the important practical and social activities of daily living," frailty is more often described in general terms: a product of "excess demand imposed upon reduced capacity" (3); a "precarious balance easily perturbed" (11); a "condition in individuals lacking in strength who are delicately constituted or fragile" (6); or a state that puts persons "at risk for adverse health outcomes" (2) or makes them "inherently vulnerable to challenge from the environment" (9, 14) or "unable to integrate responses in the face of stress" (15).

Table 1 shows an evolving geriatric functional continuum in which frailty is a midpoint between independence and pre-death. Primary intervention is currently limited because manifestations of frailty before functional decline are not always apparent. In an intermediate stage, manifestations are already evident, prompting secondary interventions; for persons in whom frailty is destined to progress to an end stage, interventions are essentially palliative. The designation of frailty overlaps with a range of other clinical correlates that broadly represent functional decline (12): being underweight or having anorexia or cachexia (29-33), moving slowly or taking to bed (28), responding only partially to nutrients or failing to thrive (20-23), coping poorly or not at all (acopia) (27), demonstrating sarcopenia and weakness (18, 19) with limited mobility and a tendency to fall (8), or requiring limited assistance or being dependent (12, 13, 26) (Table 1).

Stage for Potential InterventionClinical correlate
IntermediateIndependence with difficulty
The "dwindles"
Functional decline
Failure to thrive
Failure to cope
Taking to bed

Cognitive impairments may also appear as a slowing of mentation rather than as an acute change or a state of confusion, such as delirium. Frailty is sometimes but not always associated with dementia.

In the constellation of frailty, the health provider must make critical decisions that relate to "medicalization" of a condition that may not necessarily be disease based. For persons who seem frail, how extensive should the work-up, which is often poorly tolerated, be in an effort to uncover underlying, progressive disease (cancer is the most obvious in view of the invariable weight loss)? Should a condition that "tips the balance" (such as stroke or hip fracture after a fall) be considered an accountable, or "organic," basis on which to diagnose frailty (8, 11, 20, 22, 24)? When can a frail person, his or her family, or, more often, his or her physician consider frailty "nonorganic" (22)-a progressive loss of homeostatic and stabilizing mechanisms and a primary failure to thrive (25)-and accept or recommend comfort care? Clearly, these choices will require individual considerations and a review of the place of the frail person within the geriatric continuum from apparently early to very late. Proposed interventions aimed at improving or stabilizing the existing quality of life raise medical decisions, ethical issues, and, particularly in this era, aspects of cost and benefit. Defining a rational way to intervene may also be difficult in "many elderly patients where illness does not fit a classic disease model" (35) and when the presentation is atypical (36). Involuntary weight loss without apparent disease or vague symptoms alone may indicate risk for death in some older persons (28, 29). In one study, persons in an intermediate care facility who took to bed because of fatigue or nonlocalizing weakness had significantly shorter survival than the group as a whole (28).


Interactive conditions associated with frailty
Advanced age
Allostatic load score
Physical disability; functional decline; need for help with (I)ADLs
Falls and injury
Chronic diseases
Social starus: dependency or caregiver needs
Cognitive state: decline or depression
Physician visits:hospitalisation or institutionalisation
Nutritional impairment: reduced food intake; serum albumin level <35 g/l; unintentional weight loss; cachexia; BMI <24 kg/m2; increase in soluble interleukin-2 receptor level
Hip fracture


Despite the many reviews that seek to define frailty more precisely by patient assessment and identification of risk factors in order to promote early intervention (37-39), the concept of frailty remains diffuse. To some extent, it is similar to the term anemia, which implies a clinical perception; symptoms; and physical, laboratory, and biological findings that reflect many underlying conditions. Table 2 lists several conditions that interact with frailty and increase the likelihood of its development. Prospects for intervention would have to take these interactive conditions into account. Fried (2) proposed a model pathway of frailty that depends on whether the intervention occurs in the incipient phase of the alteration (primary strategy), during the clinical syndrome (secondary strategy), or at the time of the adverse outcome (tertiary strategy). Buchner and Wagner (6) discuss "risk factors predictive of future disability that can be modified to reduce subsequent risk (primary prevention) and "interventions to retard an early state of disablement (secondary prevention)." In a cohort of older persons living in the community, risk factors for frailty included leg and arm weakness, reduced vision or hearing, and anxiety (56). In Great Britain, the required annual health check for all patients 75 years of age or older revealed physical symptoms in about half, "for which some action was taken" (57). Certainly, a reservoir of unreported needs that can be addressed exists for elderly persons at home (58).

The benefits of comprehensive geriatric assessment generally seem more evident in frail populations (10, 11, 14, 59) as Rubenstein and colleagues (60) showed. In a groundbreaking study, they found that comprehensive geriatric assessment prolonged life in frail older men in a Veterans Administration facility (60). In the long run, however, the health outcomes that result from assessment of elderly frail persons are not likely to reduce morbidity (61) but rather to extend it (62, 63) because associated chronic, nonfatal conditions will be identified (63). In fact, as Boult and associates (64) point out, the current orientation aimed at reducing fatal conditions, such as cancer or coronary artery disease, would increase the number of older persons with functional limitations. In the next decade, reduction of chronic, nonfatal conditions would be a signal public health achievement in an aging society (63, 64). The prevalence of chronic disability and of many chronic degenerative diseases in the United States seemed to decrease from 1982 to 1994 (65). During the course of this continuing achievement, identification of the biological underpinnings of frailty would help dissociate this state from the constellation of associated and less remediable chronic conditions and facilitate early intervention.


Biological Underpinnings

Our understanding of a biological basis for many chronic conditions, such as osteoarthritis (66), osteoporosis (67, 68), and Alzheimer disease (69, 70), has advanced because of advocating personal health practices, defining genetic and biological markers, and evaluating diagnostic imaging techniques, where the ultimate intent is secondary or primary prevention. However, potential interventions for frailty based on understanding of its biological basis are confounded at the outset by the multifactorial nature of the condition, its manifestations within the spectrum of many chronic nonfatal conditions, the overlay of cognitive and mood alterations, and the interaction of crucial personal and social forces. Nevertheless, there is now an appreciation that many manifestations of frailty are observed in conditions where a metabolic balance seems to be impaired (71); for example, cytokines are overexpressed in immune disorders, in inflammatory states, and in debilitating conditions (30, 33) and hormonal mediators are deficient, especially in somatopause, during which production of growth hormone and insulin-like growth factor decreases (72-76). Evidence of a similar biological basis for the more precisely defined wasting syndromes in cancer and AIDS (77, 78) may also shed light on frailty in older persons.

To develop the biological underpinnings of frailty, I modified Fried's clinical model pathway (2) of stages related to primary, secondary, and tertiary prevention and substituted terms that describe evolving patterns of aging-successfull, usual and disease (79, 80)with the inevitable tissue modifications influenced by genetic and environmental factors. Cellular events linked to aging, such as cell senescence (81), programmed cell death (apoptosis) (82), and oxidative stress and free radical-mediated processes (70, 83), cannot yet be "translated" to define the clinical state of frailty. Some understanding may come from studies in the worm Caenorhabditis elegans, in which mutations in one of the many stress response genes led to stress tolerance and longevity (84).

More related to the clinical manifestations of frailty is evidence of a disturbed metabolic balance. On one side of the metabolic balance are cytokines. Increases in levels of interleukin-6, a "cytokine for gerontologists" (85), and tumor necrosis factor-α have been observed in aging as well as in conditions associated with anorexia, cachexia, chronic inflammatory diseases, and immune alterations (85-88). In a study of participants (mean age, 78 years) in the Framingham Heart Study, production of interleukin-6 and the anti-inflammatory cytokine interleukin-1 receptor antagonist by peripheral blood monocytes markedly increased with age. These data suggested to Roubenoff (30) that aging is associated with a dysregulation of cytokine formation that could lead to loss of muscle strength (sarcopenia), which may be one of the earliest manifestations of frailty (89-91). In a large community-based population studied by Cohen and associates (92), higher levels of serum and mononuclear cell production of interleukin-6 seemed to be related to participants' self-reports of fatigue and functional decline and were not necessarily associated with disease. Elevated plasma levels of interleukin-6 were found in communitydwelling persons 70 years of age or older who were depressed (93). Healing of punch-biopsy skin wounds was impaired in persons 60 to 90 years of age who did not have apparent disease compared with younger persons; modifications in cytokines released at the inflammatory site in older persons contributed to alterations in endothelial cell-adhesion molecules and the cellular infiltrate in the wound (94).

The evolution of successful aging into usual aging (79, 80) may be related to even higher levels of cytokines, especially tumor necrosis factor-α, interleukin-1, interleukin-6, and members of the interleukin-6 superfamily (33), that contribute to progression of disease, including bone resorption with osteoporosis (95, 96). Cytokines also induce hepatic synthesis of acute-phase response proteins (97), among which C-reactive protein has recently received particular attention. A sensitive assay revealed small increases in serum C-reactive protein levels in early osteoarthritis that predicted progressive disease (98). Elevated levels of C-reactive protein in persons without apparent clinical cardiovascular disease may be related to inflammation in the atheromatous plaque and may predict subsequent myocardial infarction (99, 100). Elevated levels of cytokines, especially tumor necrosis factor-α, have been observed in the final stages of advanced debilitating diseases that are often associated with frailty in elderly persons:

weight loss and malnutrition associated with chronic inflammation (101), cardiac cachexia (71), cardiomyopathy (102), emphysema (103), and infections and pressure sores (104).

On the other side of the metabolic balance are hormonal mediators. Their link with cytokines, especially tumor necrosis factor-α, interleukin-1, and interleukin-6, occurs by way of cytokine interactions with the central nervous system, resulting in various physiologic, neuroendocrine, and behavioral responses (105). Elevated levels of interleukin-6 represent an important stress response that activates the hypothalamic-pituitary-adrenal axis (105-109). Aging and disease may subsequently impair the neural-immune (105, 108), neuroendocrine (106, 107), and endocrine (74) pathways that control the production and peripheral actions of hormonal mediators, with the decline of endocrine functions described as menopause, andropause, adrenopause, and somatopause (74). The increase in abdominal fat mass that occurs with aging, which is thought to contribute to insulin resistance and cardiovascular diseases, may be an important source of leptin, another hormonal mediator (110, 111). Leptin is part of the host's cytokine response to inflammation and infection; interleukin-1 and tumor necrosis factor-α stimulate leptin synthesis in adipose tissue (112-115). Leptin acts on receptors in the hypothalamus and periphery to reduce food intake (satiety) and increase energy expenditure (116-118). In experimental animals, adiposity and levels of plasma leptin increase with age, but there is a cut-off in the oldest group of animals ("senescence") at which adiposity and plasma leptin levels decrease (119). Similarly, plasma levels of leptin were lower in persons older than 60 years of age than in a younger group; these results were attributed to an altered leptin signaling system in the central nervous system "similar to [that in] other endocrine systems" and to decreased adipose tissue and leptin production (120).

As the biological underpinnings of frailty continue to emerge, the "translational" potential for the health provider will perhaps be the ability to use serum markers to identify frailty more precisely. These serum markers are under investigation. The health provider can play a role (121) as an advocate for patients to adopt personal health practices to make "successful aging and usual aging synonymous" (80). lack of smoking, maintenance of a low body mass index, and exercise seemed to reduce cumulative disability in a study group now 75 years of age (122). Exercise in particular maintains muscle mass, increases strength and function, stabilizes bone mineral density, and favorably modifies glucose homeostasis and cardiovascular dynamics (123, 124). Caloric restriction in various animal species decreased body weight and fat mass, extended the life span, and retarded manifestations of aging (73, 125). The potential application of caloric restriction to the modification of human aging and longevity will be interesting. Other interventions to maintain successful aging that are under investigation include antioxidant therapies (70, 126) and hormone replacement protocols, most traditionally with estrogen for menopause with its attendant cardiovascular and skeletal risk (127), testosterone for andropause (128), growth hormone and insulin-like growth factor for somatopause (73, 129-132), and dehydroepiandrosterone for adrenopause (133, 134). Finally, new methods of inhibiting Inflammatory effects of cytokines are forthcoming, in particular by blocking cytokines themselves or cytokine receptors, especially those for tumor necrosis factor-α, interleukin-1, and interleukin-6 (78, 13-137).




Older persons tend to develop several age-related chronic conditions that interact with and contribute to frailty. As a result, the concept of frailty has not emerged as a well-defined clinical entity; preventive approaches remain tertiary, extend morbidity, and fail to improve long-term health outcomes. In its biological underpinnings, frailty may represent the complex and cumulative expression of altered homeostatic responses to multiple stresses. The resulting metabolic imbalance seems to be reflected by serum markers that document overproduction of catabolic cytokines and by diminished availability or action of anabolic hormones. At present, however, the serum markers under study, particularly in communitydwelling older persons, do not provide specificity for a biological basis for frailty; rather, they reflect aging itself and the presence of associated chronic conditions. Nevertheless, restoration of the metabolic balance remains a potential therapeutic goal. The health provider continues to play an important role in encouraging older persons to accept personal health practices-especially exercise, which provides overall systemic benefits-as a way of life. Interventions emerging from biological research and advocacy of health practices may ultimately promote secondary or primary preventive measures for frailty, permitting older persons to achieve successful aging and well-being longer (138-140). Perhaps in the next decade, the "new gerontology" (141) will become "preventive gerontology" (80), with reduction of morbidity in very late life and great societal benefits.

David Hamerman, MD

Albert Einstein College of Medicine

Bronx, NY 10467

Acknowledgments: The author thanks Drs. Nir Barzilai and Jomarie Zeleznik for helpful discussions and Dawn Bowen-Jenkins for preparation of the manuscript.


Requests for Reprints: Dr. David Hamerman, Resnick Gerontology Center, Albert Einstein College of Medicine, 111 East 210th Street, Bronx, NY 10467

Ann Intern Mdc. 1999;130:945-950.