| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||
GUEST EDITORIAL |
1 Cumberland Infirmary, Carlisle, England.
2 Dalhousie University, Halifax, Canada.
Please address correspondence to Kenneth Rockwood, Centre for Health Care of the Elderly, Capital Health, 1421-5955 Veterans Memorial Lane, Halifax, NS, Canada, B3H 2E1. E-mail: kenneth.rockwood{at}dal.ca
THE ancient European theories on aging were based on the hypothesis that all life is "wet" and that aging is caused by dehydration—hence the custom for older people to take special baths to become rehydrated and more youthful (1). We now know that dehydration is not inevitable with aging but is an association with acute illness in older people that results in a high mortality (2). The cause of dehydration is multifactorial, related to swallowing difficulty, lack of thirst, cognitive impairment, physical limitations (including restraints), and misuse of diuretics (3). Dehydration is both a predisposing and precipitating factor for delirium or acute confusional state (4). Dehydration generally is a predictor of impaired cognitive status, the mechanism for which has been postulated to be due to a reduction of nitric oxide synthase that occurs during aging (5). Patients in nursing homes may be especially vulnerable to dehydration because of a high prevalence of predisposing factors. In addition, dehydration may indicate substandard care, if it arises due to lack of shift time, or even neglect (6). Recognition of dehydration can be challenging, as physical signs of dehydration have a poor sensitivity and specificity in older people in comparison with laboratory measures (3). Although a common clinical problem, we have no easy method to detect it early and possibly prevent hospital admission, delirium, and even death.
In this issue, Culp and colleagues report the results of an analysis of bioimpedance and other hydration parameters as risk factors for delirium in rural nursing home residents (7). They found no relationship between bioimpedance measurements of hydration and delirium, but there was a statistically significant relationship between blood urea nitrogen: creatinine ratio and low hematocrit and delirium. This result is disappointing for anyone who might have hoped for a sensitive test from a practicable instrument. Experience with bioimpedance in elderly people is comparatively limited especially in dynamic scenarios and when there may be electrolyte disturbances (1). In addition, the bioimpedance measurements may have influenced the behavior of the nurses or even the patients, by consciously or subconsciously resulting in increased fluid intake. Although this effect might have been detected had fluid balance charts been kept, fluid intake records can be unreliable in nursing homes, and have their own potential for inducing an observation effect.
As difficult as the study of dehydration might be, the study of delirium can be even more complicated. At present, delirium remains a clinical diagnosis, without an adequately sensitive and specific biological marker. The authors usefully employed several measures to yield a reproducible estimate of whether delirium was present. Their counting as a second episode of delirium recurrence of symptoms after only 24 hours of cognitive stability might overestimate the incidence of delirium. Still, their protocol offers a practicable means to detect delirium in studies in which not everyone can be seen by a clinician expert in delirium diagnosis.
Despite these few reservations, the study by Culp and colleagues raises many fascinating questions (7). Is bioimpedance a practical method to detect dehydration in older people, or is it confounded by factors, such as changes in body composition and electrolyte balance, that accompany frailty? How well does it compare with other methods? Delirium is not a diagnosis but a multifactorial syndrome and a manifestation of frailty, itself a multiply-determined dynamic state (8). Does this study indicate that delirium is not as strongly associated with dehydration as is commonly thought? For example, Wilson and Morley have noted the lack of data that demonstrate a firm correlation between tissue hydration and cognitive function (5). Might the real association be with markers of frailty? That is, are hematocrit and blood urea nitrogen/creatinine ratio better indicators of frailty than of dehydration? If so, delirium might not be amenable to a simple "find a cause and then fix it" approach, but rather would require multifactorial interventions, even in the face of a clear precipitant. The situation is further complicated by the fact that dehydration is a common complication of delirium as well as a possible precipitant. Previous studies have documented other stronger associations with delirium including immobility, malnutrition, and medication (4). Despite this, prevention and treatment of delirium is still difficult in practice (9). Furthermore, the studies in prevention have concentrated on hospital patients, and their results may not be generalizable to nursing homes. Further research is needed, and the authors are to be congratulated on stimulating readers to question basic assumptions. This study illustrates how a seemingly simple clinical scenario (dehydration and delirium) in older people in nursing homes can pose a complex research challenge—therein lies the unique appeal of research in geriatric medicine.
Acknowledgments
James George received a BUPA Foundation award for sabbatical study during which this was completed. Kenneth Rockwood is supported by an Investigator Award from the Canadian Institutes of Health Research and by the Dalhousie Medical Research Foundation as the Kathryn Allen Weldon Professor of Alzheimer Research.
Received February 26, 2004
Accepted March 4, 2004
References
| ||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
