“Failure To Thrive” in Older Adults

Review of the Literature: An Intellectual Tension

A review of the complicated failure to thrive literature shows many synonyms and colloquialisms familiar to physicians [1-5] (Table 1). Pediatricians use the term “failure to thrive” when their patients fail to achieve height, weight, or behavioral milestones as determined from large populations that have generated normative data [6]. Braun [7] was the first to provide a justification for exporting the concept to geriatrics:

The clinical picture or symptom complex in the elderly failure-to-thrive person presents as a mirror image of the infant failure-to-thrive. The older person loses weight, declines in physical and cognitive function, and often exhibits signs of hopelessness and helplessness.

The verbs “lose” and “decline” in the above statement underscore the primary intellectual obstacle to a tidy transposition of the term “failure to thrive” from infant to octogenarian: The pediatrician uses the term to describe the patient who has not attained functional status; the geriatrician uses it for the patient who has not maintained functional status. The distinction is crucial because it forces the internist to reconfront a central challenge of geriatric medicine—distinguishing disease states prevalent in older adults from “normal aging” (for example, distinguishing dementing illness from “physiologic changes” in the aging brain [8]). If we were to take a cue from our pediatric colleagues and construct nomograms with “activities of daily living dependence” rather than height or weight as the measure of scrutiny, we would discover that the prevalence of at least one impairment increases with each decade of life and approaches 40% in nonagenarians [9]. If some degree of impairment is a “normative” milestone, it is certainly not a desirable one; should we express surprise or acceptance when our patients reach it?

Given the pervasive ageism in our society, in which older adults are often denied optimal care [10], geriatricians teach that it is a mistake to reflexively ascribe decline to “old age” and accept it as inevitable. It is with these same noble ideals that failure to thrive was first described by Hodkinson [11] in 1973:

Illness [in the elderly] often presents as insidious and progressive physical deterioration, for which the paediatric term “failure to thrive” is appropriate. Typically the patient's decline comprises deteriorating social competence, weight loss, loss of appetite, increasing frailty, and diminishing initiative, concentration, and drive. This general failure of the old person is all too often accepted as due to “old age” or senility or is regarded as a dementing process and the physical basis is overlooked. There are many diagnostic possibilities.

Seeking occult and treatable illnesses as the basis for decline resonates with modern geriatric practice. In contrast, Isaacs and colleagues [1] reported on the functional status of residents of Glasgow toward the end of life:

… a high proportion of all deaths in old age were preceded by a period of “pre-death” during which the patient was unable to care for himself in consequence of loss of mobility, incontinence, or mental abnormality … It seems that many of those who survive into old age enter a phase of “pre-death” in which they outlive the vigour of their bodies and the wisdom of their brains.

The juxtaposition of these two papers from the early 1970s emphasizes the balance between the thorough medical evaluations typical of the geriatrician and failure to thrive as a prelude to “natural” death. Could the symptom complex described as failure to thrive sometimes be simply the manifestations of “pre-death”? It has been proposed that in the absence of disease, very elderly patients will eventually undergo a process of progressive functional decline, apathy, and loss of willingness to eat and drink that culminates in death [12]. Clearly, in such a situation, an aggressive diagnostic approach would not only be futile but could contribute to suffering.

But which is the more appropriate vantage point from which to approach such patients? No published work addresses this question directly. Two systematic studies [13, 14], both retrospective and done in inpatient settings, have examined how failure to thrive is used as a diagnosis in adults. These studies show that the term is applied to a heterogeneous, incapacitated group of patients who are no longer able to function independently. Comorbid conditions, functional impairment, anorexia, depression, and dementia are prevalent. In the larger study [13], 45% of patients were incontinent, 35% were fallers, and 9.8% had decubiti. All of these problems are familiar geriatric syndromes in their own right. Although 13% to 15% of patients died during hospitalization and more than 30% were discharged to nursing homes, neither study addressed the extent and cost of the diagnostic procedures done with the hope of finding something “curable.”

Despite the heterogeneity of the patients studied (and the subsequent admonition [13] that the diagnosis “failure to thrive” should perhaps be abandoned), several subsequent review articles in the medical [2, 15, 16] and nursing [17-19] literature have conceptually approached failure to thrive as a clinically meaningful diagnosis. It has had its own International Classification of Diseases, Ninth Revision (ICD-9) code since 1979 and is used increasingly in the geriatric population as well as in patients positive for the human immunodeficiency virus (HIV). A recent evaluation of a geriatric inpatient unit showed that failure to thrive was a common admitting diagnosis, alongside more tangible entities, such as diabetes and gastrointestinal bleeding [20].

Unlike diabetes and gastrointestinal bleeding, however, the symptom complex inconsistently described as failure to thrive does not conform to any accepted model of disease (such as that of the New York Heart Association), in which a fully specified disease must include a clear definition of its cause, anatomy, pathophysiology, and functional effect [21]. Alternatively, one might conceptualize failure to thrive as a geriatric syndrome (such as falling and immobility) that may be defined as “a cluster of symptoms, conditions, and disabilities resulting in a variety of physiologic changes, pathologic conditions, comorbid conditions, and environmental challenges” [22]. With this approach, the National Institute on Aging in 1991 described failure to thrive as “a syndrome of weight loss, decreased appetite and poor nutrition, and inactivity, often accompanied by dehydration, depressive symptoms, impaired immune function, and low cholesterol” [22].

A Rational Approach to Management

Abandoning Failure To Thrive as a Disease Construct

How, then, should one proceed when confronted with a patient who, by anyone's definition, is failing? Ironically, we believe that the rather disorderly literature on failure to thrive—by the very nature of its disorganization—says something about the conceptual framework required to care for these patients. In our view, the label “failure to thrive” promotes an intellectual laziness—accompanied by a certain resignation, passivity, or fatalism—that needs to be balanced by a considered and thoughtful deconstructionist approach, wherein the areas of impairment would be carefully identified, quantified, and, most importantly, scrutinized for potential interactions. We therefore advocate the abandonment of the term “failure to thrive” and the adoption of a more measurement-oriented approach with particular attention to four major contributor domains that recur in this literature and are known to be morbid and mortal entities in older adults: impaired physical functioning, malnutrition, depression, and dementia (Figure 1).

Impaired Physical Functioning

Beyond the obvious relevance of functional impairment to quality of life, numerous studies have shown the prognostic power of functional status: Impaired patients are at high risk for many poor outcomes, including nursing home placement and death [23, 24]. Early measures focused on self-reported ability to perform activities of daily living [25]; these have since been supplemented by performance-based methods that might detect functional dependence in a “subclinical state” [26-28].

Careful measurement with standard assessment tools precisely defines the nature of impairments and forces the clinician to adopt a problem-solving approach. Does the patient who is unable to dress himself have a frozen shoulder from previous injuries, apraxia from dementia, hemi-neglect from stroke, or a combination of these? Does the patient who cannot use a telephone have visual disturbance from cataracts, memory loss from depression, or impaired manual dexterity from arthritis? Physical and occupational therapy consultations are invaluable in sorting out these contributors. The use of performance-based measures, wherein the patient is observed in typical real world maneuvers [29], may be especially useful, although there is no consensus on which tests should be selected for office-based practices or how they should be integrated.

Determining the acuity of any deficit is crucial. Using many data sources (including the patient and family) helps to determine the trajectory of decline. Rapid decline suggests new illness or acutely decompensating chronic disease, whereas insidious decline suggests the natural history of chronic processes such as dementing syndromes. If all treatable contributors have been addressed and impaired physical functioning still persists, patients may benefit from several interventions, such as high-intensity strength training to increase functional mobility [30] or multiple risk-factor reduction to decrease the risk for falling [31, 32]. Whether intervening in incipient impairment will avert illness and death is one of the most actively pursued questions in geriatric research.

Malnutrition

Even though 55% of geriatric inpatients have protein-calorie undernutrition, most cases of this condition are never diagnosed or treated [33]. This is alarming, because undernutrition is an independent predictor of death in older adults [34]. How best, then, to screen for malnutrition? Body weight alone may be unreliable, because pathologically low weight in an older person may be confused with lifelong low weight [3]. Furthermore, the weight at which a person has the lowest mortality increases with age [35]. Anthropometric measurements derived from midarm circumference and triceps skin-fold thickness [36] are time-consuming and lack interobserver reliability [3]. Although not diagnostic by themselves, low serum cholesterol and albumin levels may be indicators of undernutrition [37]. Hypocholesterolemia is a predictor of death in patients in nursing homes [38], and hypoalbuminemia is an independent risk factor for death in older persons [39]. A rational approach is to consider abnormal laboratory values in the context of body weight, weight trend, and the presence of other stigmata of malnutrition, such as muscle wasting. Of course, when the underlying causes of malnutrition are being considered, it is important to take into account not only the exacerbation of more serious chronic illness (the patient with chronic obstructive pulmonary disease who is so dyspneic that he cannot eat) but also easily remedied causes, including poorly fitting dentures, changes in food preference, and dysgeusia-causing medications.

Depression and Decreasing Socialization

Failure to diagnose depression in elderly patients is a well-documented phenomenon [40, 41]. The rates of major or minor depression among elderly persons range from 5% in primary care clinics to 15% to 25% in nursing homes, and all-cause mortality rates are higher among elderly persons with depression than among their nondepressed counterparts [42]. Similarly, social isolation in itself is a risk factor for poor outcomes, including institutionalization [43] and death [44]. Berkman and Syme [45] showed that persons who lacked social and community ties were more likely to die during the follow-up period than were those with more extensive contacts. An increasing body of research has shown that emotional support is predictive of recovery from specific illnesses, such as myocardial infarction [46].

Patients who are declining in any of the four contributor domains should be screened for depression; several instruments such as the Geriatric Depression Scale [47] and the Center for Epidemiologic Studies Depression Scale [48] are available for this purpose. Geriatric psychiatry consultation can be useful in establishing a diagnosis and assisting with management. If depression persists despite treatment of all potentially contributing factors, pharmacologic treatment should be started. Although a detailed discussion of the pharmacotherapy of depression in older adults is beyond the scope of this discussion, readers are referred to recently published expert guidelines on the topic [49].

Cognitive Impairment

About 11% of adults in the United States older than 65 years of age [50] have dementing syndromes, and the prevalence of dementia increases dramatically with each successive decade [51]. Frequently overlooked by clinicians [52], dementia leads to progressive impairment, institutionalization, and death [53]. Standardized tests such as the Mini-Mental Status Examination [54] are useful for both screening and diagnosis and should be administered to all older adults. After cognitive impairment is documented, establishing a diagnosis of dementia requires additional criteria for which several algorithms have been established, including those in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). The evaluation should include a search for reversible or treatable causes of cognitive impairment, although only a few dementias are reversible and even fewer reverse completely [55]. Although comprehensive management of the patient with dementia is also beyond the purview of our discussion, clinicians are referred to other articles on medical management of the demented patient [56-58].

Abandoning Parsimony of Diagnosis

The notion that all signs and symptoms in a patient's presentation are referable to a single pathophysiologic process served internists well at the turn of the century, when the octogenarian was the exception rather than the rule in medical practice. By 75 years of age, however, most adults have two to three chronic medical conditions, with substantial implications for the clinician. Whereas the 55-year-old patient with dyspnea and back pain may have bronchoalveolar carcinoma metastatic to the spine, the 85-year-old patient is more likely to have a combination of decompensated congestive heart failure and osteoporotic compression fractures. In contrast, a single chief symptom in the elderly patient is more likely to have several contributing causes, and attempts to find a unifying diagnosis may be futile.

Furthermore, deficits frequently coexist in more than one of the contributor domains cited above. Both depression [59] and dementia [60] may initially present with weight loss. New depression may be the first sign of impending cognitive dysfunction [61]. Physically impaired patients are at greater risk for becoming malnourished [37] and are more likely to have decubitus ulcers [62], which may lead to infection and worsening nutritional status and thus perpetuate a vicious cycle of functional decline. Similarly, a higher level of depressive symptoms in patients with dementia [63, 64] and chronic medical conditions [65] correlates with poorer functional status. Clearly, attempts to determine which deficits are primary and which are secondary may be an academic exercise that is not clinically useful.

Thus, the linear or vectorial reasoning that is applicable to the care of the younger patient may actually be an intellectual obstacle in the evaluation of the older patient who is declining [66, 67]. A more rational approach is to quantify the nature and level of the patient's impairments, recognize that deficits interact in mutual and complex ways to create further impairment, and treat the easily remedied contributors. By modifying the predisposing risk factors that many geriatric syndromes share, it may be possible to restore compensatory ability and delay the progression of functional dependence [68].

Conclusions and Areas for Future Study

Failure to thrive should be abandoned as a diagnostic entity. This stigmatizing label distracts the clinician from a systematic evaluation of the combination of interacting deficits known to be prevalent in patients who are said to have failure to thrive: impaired physical functioning, malnutrition, depression, and cognitive impairment. Like many problems in geriatric medicine, failure to thrive is more clinically approachable when broken down into measurable domains.

At the same time, good geriatric practice requires careful balance and a broad perspective. Thus, common areas of impairment have similar predisposing factors, some treatable and some not. Although many predisposing factors are easily corrected, the treatment for others can be resource-intensive and may be poorly tolerated by the patient. New depression may indicate underlying pancreatic cancer, but few experienced clinicians would advocate routine computed tomographic scans as part of the initial workup. Similarly, a patient who is dying should not be forced to endure a battery of multiple assessment scales when palliation would be more appropriate.

We have deliberately avoided making specific recommendations about how aggressively to pursue underlying occult illness as a factor contributing to a patient's decline; this crucial question has profound ethical and policy implications and cannot be answered without better data. At this point, the clinician must rely on the limited information that is available and, most importantly, respect the personal preferences of the patient. Studies identifying the patients who are more likely to benefit from interventions at presentation, the diagnostic maneuvers that are the most effective and the most tolerable, and the therapeutic approaches that improve the outcomes that really matter to patients and families should help clinicians and patients make these difficult decisions. Although some geriatricians who are committed to the comprehensive care of older adults may find it objectionable to identify a group of older adults who should not receive aggressive diagnostic evaluations, sparing patients from undignified, painful, and ineffective therapies is an important goal of modern geriatric medicine.

Dr. Lachs: New York Hospital-Cornell University Medical Center, 515 East 17th Street #912, New York, NY 10021.