- Email: [email protected]
Prevalence and Causes of Undernutrition in Medical Outpatients
Prevalence and Causes of Undernutrition in Medical Outpatients Margaret-Mary G. Wilson, MB, BCh, Surender Vaswani, MD, David Liu, MD, John E. Morley, MB, BCh, Douglas K. Miller, MD PURPOSE: To assess the prevalence, common causes, and frequency of recognition and treatment of undernutrition in older and younger medical outpatients using a cross-sectional survey design with 2-year follow-up of undernourished subjects. PATIENTS AND METHODS: Charts of 1017 adult patients attending a hospital outpatient department were reviewed for the presence of undernutrition, and 85 patients meeting inclusion criteria for undernutrition were evaluated and followed for 2 years. An initial evaluation focused on nutritional, cognitive, and affective status and on nutritional attitudes using two subscales of the EAT-26 eating disorder inventory. After 2 years, initial data plus outpatient records were evaluated by 2 independent reviewers to determine a primary cause of undernutrition and to assess the recognition and treatment of undernutrition by the primary physician. RESULTS: Undernutrition was identified in 46 (11%) and 44 (7%) of older and younger subjects respectively; odds ratio (OR) (95% [confidence interval (CI)]) for older versus younger 5 1.65 (1.06 to 2.51). The primary cause of undernu-
trition differed between age groups but was deemed treatable in nearly 90% of all subjects. Undernutrition was recognized in 19 (43%) older subjects and 5 (12%) younger subjects (OR 5 5.47 [1.87 to 16.0]), and appropriate intervention(s) were instituted in 6 (14%) and 2 (5%) of older and younger subjects, respectively (OR 5 3.08 [0.668 to 14.2]). Older subjects scored higher on the EAT-26 oral control subscale than did younger subjects (4.7 versus 2.5, P 5 0.004) but similarly on the EAT-26 dieting subscale (5.2 versus 6.3, P 5 0.332); these relationships did not change with control for potentially confounding variables. CONCLUSIONS: In this study, undernutrition was relatively common, usually amenable to treatment, but frequently undetected and undertreated in both older and younger medical outpatients. Older undernourished subjects exhibited higher oral control needs than younger persons, which may have implications for the pathophysiology and treatment of their malnutrition. Further improvement in detection and intervention is warranted in both younger and older age groups. Am J Med. 1998;104:56 – 63. q1998 by Excerpta Medica, Inc.
U
ings, there has been a paucity of research directed towards determining the prevalence and etiology of undernutrition in older outpatients. Our previous study of older male veteran outpatients (12) demonstrated that undernourished patients had increased levels of oral control needs but similar dieting behaviors compared with normally nourished patients. An important minority of undernourished patients in that study also evidenced various degrees of a distorted body image that bore some similarity to anorexia nervosa in younger persons. The relative rarity of documentation of undernutrition as a medical diagnosis or contributory cause of death has been noted in several studies (1,5,13). This phenomenon may hinder practitioners’ appreciation of the frequency with which undernutrition occurs in medical outpatients and the impact of undernutrition on subsequent decrements in health. It may also hamper early initiation of appropriate therapeutic interventions that have the potential for preventing nutritionally related health declines. This study was designed to assess the prevalence, common causes, frequency of clinical recognition, and extent of therapy for undernutrition in community-dwelling medical outpatients. The results obtained in older subjects were compared with the findings in younger subjects.
ndernutrition is not a rare occurrence in older hospital patients and nursing home residents (1,2). Within acute care hospitals and chronic institutions, prevalence rates of undernutrition range from 17% to 65% (3). Available data suggests that the prevalence of diagnosed undernutrition in community dwelling senior citizens may be less than 5% (4). However, it has been estimated that one fifth to one third of older persons may have evidence of undernutrition, some of which might be considered ‘‘subclinical’’ (5,6). Existing data clearly demonstrate that undernutrition adversely affects both length and quality of life (7–11). In particular a positive correlation between undernutrition and poor functional status has been identified in older persons (10,11). Despite these find-
From the St. Louis University Health Sciences Center (M-MGW, SV, DL, JM, DK), and the St. Louis Veterans Affairs Medical Center (JEM, DKM), St. Louis, Missouri. Requests for reprints should be addressed to Douglas K. Miller, MD, Division of Geriatric Medicine, 1402 S. Grand, Rm. M238, St. Louis, Missouri 63104. Dr. Liu’s involvement in the study was supported in part by an educational grant from the John A. Hartford Foundation, which took no direct role in the investigation. Manuscript submitted January 30, 1997 and accepted in revised form August 9, 1997. 56
q1998 by Excerpta Medica, Inc. All rights reserved.
0002-9343/98/$19.00 PII S0002-9343(97)00279-9
Undernutrition in Medical Outpatients/Wilson et al
METHODS Study Site, Subjects, and Inclusion/Exclusion Criteria The faculty in the Division of General Internal Medicine at Saint Louis University (SLU) Health Sciences Center provides direct services to about 1600 general adult medical outpatients through 5 half-day sessions each week. From September 1990 to July 1993, all active patients 14 years of age and older attending any of these sessions were considered for entry into the study. Exclusion criteria included clinical features or a documented diagnosis of protein-losing syndromes (eg, nephrotic syndrome, protein-losing enteropathy or severe liver disease). Patients with a history of any cancer, other than skin cancer, within the preceding 10 years were also excluded. Each patient was screened only once. Charts of patients not meeting the exclusion criteria were screened for 1 of the following indices of undernutrition: (1) A body weight less than 90% of the appropriate median weight using age, gender, and height adjusted norms. Master’s reference tables (14) were used for those 65 years of age and older. Similarly adjusted norms from the First National Health and Nutritional Examination Survey were used for those younger than 65 years (15). (2) Documented weight loss, unexplained by fluid loss, in excess of 2.27 kilograms (5 pounds) over the preceding 6 months. (3) Serum albumin levels less than 3.5 g/dL.
Recruitment and Evaluation of Study Participants Chart reviews were performed on the afternoon prior to the patients’ scheduled visit. If the chart review identified one of the criteria for undernutrition, the patient was informed of the purposes, obligations, and risks of the study, and signed consent was obtained from willing participants. A trained research assistant obtained the following information from subjects and their charts using a standardized protocol: (1) demographic data, (2) socioeconomic circumstances, (3) past medical illnesses, (4) medication history, and (5) recreational and social habits. A nutritional assessment was carried out on each recruited subject. This included inquiry into dietary habits, baseline body weight, and time course of any changes in weight. Mini-mental State Examination (16) and Geriatric Depression Scale (GDS) (17) scores were obtained on each subject. The dieting and oral control subscales of the EAT-26 questionnaire were administered to each subject in an attempt to identify inappropriate nutritional attitudes (12,18). The EAT-26 was developed by Garner and colleagues as a multidimensional eating disorder inventory for facilitating identification of anorexia nervosa and bulimia in younger patients (18). It has also proven useful in
identifying abnormal eating attitudes and body image in undernourished older outpatients (12). The EAT-26 inventory has three subscales: dieting, oral control, and bulimia. The dieting subscale includes questions assessing perception of body image, eating habits, and nutritional attitudes. The oral control subscale evaluates the individual’s attempted control over the circumstances of eating. The bulimia subscale assesses purging behaviors, was not helpful in the prior study (12) and thus was omitted from the present investigation. After the initial assessment, enrolled subjects were monitored over the next several months by serial review of the outpatient records to identify the major cause of undernutrition in each subject, to assess the clinical detection rate and extent of therapeutic intervention by the primary physician, and to evaluate the course of weight change over time. The charts were reviewed independently by two reviewers (MMW and DKM), each of whom had access to all the data collated over the 2-year follow-up period. Criteria for identifying an underlying cause of undernutrition were established prior to the chart reviews by consensus conference among three of the investigators (MMW, DKM, and JEM). For example, the diagnosis of depression was made if chart review revealed: (1) Diagnostic Statistical Manual (DSM) IIIR (19) criteria for major depression or (2) a GDS score over 14 (17) in the presence of one or more DSM-IIIR symptoms. (The full list of diagnostic criteria is available from the first author on request.) A major cause of undernutrition was accepted as established when both reviewers agreed that it was the primary contributor to undernutrition. Any disagreements on blinded review between the two primary reviewers were presented to a third reviewer (JEM), who acted as adjudicator. Four cases required adjudication; in each case the final, adjudicated major cause had been identified as the primary contributor by one primary reviewer and as a contributing cause by the other primary reviewer. The syndrome of undernutrition was considered to have been recognized by the primary physician if reference was made to a nutritional problem in the clinic notes or problem list at the front of the chart. For the purpose of this study, appropriate therapeutic intervention was considered to be comprised of one or more of the following: (1) identification and treatment, where possible, of the identified major contributory cause of undernutrition; (2) a critical review of the subject’s diet and institution of appropriate therapeutic nutritional support; and (3) serial anthropometric or biochemical monitoring to assess the subject’s response to treatment. Approval was obtained from the SLU Health Sciences Center’s Institutional Review Board for all research procedures. Informed consent was obtained from all subjects in conformance with assurances given by SLU to the DeJanuary 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 57
Undernutrition in Medical Outpatients/Wilson et al
Table 1. Clinical Presentation of Undernourished Patients
Weight loss ($5 lb/2.27 kg) Low body weight Hypoalbuminemia
Older (n 5 46)
Younger (n 5 44)
21 (46%) 11 (24%) 14 (30%)
29 (66%) 10 (23%) 5 (11%)
partment of Health and Human Services to protect the rights of human subjects.
Statistical Analysis Significant differences between groups was evaluated using Sato’s method to determine odds ratios (ORs) and 95% confidence intervals (CIs) for dichotomous variables (20) and Student’s t-test with two-tailed tests of significance for continuous variables. Analysis of covariance was used to evaluate for the possibility that confounding variables explained the apparent association between age and high EAT-26 oral control subscale scores (21).
RESULTS The charts of 1017 patients were screened for undernutrition; 408 patients were age 65 or older, and 609 patients were age 18 to 64. Using the criteria as previously defined, 46 (11%) older persons comprising 30 women and 16 men and 44 (7%) younger persons comprising 30 women and 14 men were found to be undernourished (OR [95% CI] for older versus younger 5 1.65 [1.05 to 2.51]; OR for female versus male 5 1.08 [0.689 to 1.71]). Five patients (aged 72, 69, 55, 42, and 37 years) screened positive but refused to participate. The 72-year-old presented with low weight and was thought likely to have abnormal eating attitudes based on screening chart review. He refused to participate on two separate office visits, declining to give a reason. The four other patients presented with weight loss and cited lack of time as their reason for not wanting to participate. Weight loss was the most common method of presentation for both age groups (Table 1). Younger persons were more likely than older persons to present with weight loss and were less likely to present with hypoalbuminemia (OR for younger versus older with weight loss 5 2.30 [0.997 to 5.31] and with hypoalbuminemia 5 0.293 [0.099 to 0.865]). Forty-four older and 41 younger subjects were interviewed at baseline and followed for 2 years. The primary cause of undernutrition was identified in 93% of older persons and 90% of younger persons (Table 2). Of the 78 patients with an explanation for their weight loss, 69 (88.5%) were felt to have potentially treatable causes for their undernutrition. Occult malignancies were impli58
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
cated in only 5% of the total sample, occurring only in older subjects. Two subjects were eventually diagnosed with gastric carcinoma and the other two subjects were found to have adenocarcinoma of the pancreas and lung. Depression was identified as the commonest cause of undernutrition in older persons, occurring in 30% of the group and was the apparent cause of undernutrition in 15% of younger subjects. The difference was not statistically significant. Undernutrition resulting from physician intervention was identified in association with prescribed medication and therapeutic diets. Persistent anorexia associated with oral digoxin therapy resulted in undernutrition in one older subject based on high serum level plus improvement in appetite and weight gain with reduction in digoxin dosage. Thyrotoxicosis resulting from inappropriately high doses of thyroxine was implicated as the primary cause of undernutrition in a younger subject. Two older persons on a low salt diet and a third on an antidiabetic diet developed undernutrition as a direct consequence of these medically prescribed therapeutic diets. In the younger subjects, low cholesterol diets were implicated in three subjects and low salt diets in two subjects. Intentional weight loss resulting from weight reducing diets, the commonest cause of undernutrition in younger subjects, was relatively rare in the older group. Intentional weight reduction was the only cause of undernutrition that exTable 2. Primary Causes of Weight Loss in Investigated Subjects
Depression Intentional* Therapeutic diet Oropharyngeal disease Chronic pain Cancer Poorly controlled diabetes Alcohol Anorexia nervosa Cardiac failure Cholelithiasis Obstructive airway disease Myopathy Dementia Difficulty with food preparation Postsurgery Thyrotoxicosis Medication Post menopausal syndrome Unexplained * P 5 0.01
Older (n 5 44)
Younger (n 5 41)
13 (29.5%) 1 (2.3%) 3 (6.8%) 3 (6.8%) 1 (2.3%) 4 (9.1%) 4 (9.1%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 2 (4.5%)
6 (14.6%) 8 (19.5%) 5 (12.2%) 2 (4.8%) 4 (9.7%) 0 (0%) 3 (7.3%) 2 (4.8%) 1 (2.4%) 2 (4.8%) 0 1 (2.4%) 0 0
1 (2.3%) 1 (2.3%) 0 1 (2.3%) 1 (2.3%) 3 (6.8%)
1 (2.4%) 0 1 (2.4%) 1 (2.4%) 0 4 (9.7%)
Undernutrition in Medical Outpatients/Wilson et al
Figure 1. Percentage of older and younger subjects whose undernutrition was recognized (open bars) and who received appropriate treatment (closed bars).
hibited a significant difference in prevalence between younger and older subjects (OR for older versus younger 5 0.096 [0.015 to 0.624]) (Table 2). In only 3 older (6.8%) and 4 younger (9.7%) subjects was no major cause of undernutrition identified. Figure 1 highlights the clinical recognition rate of undernutrition by primary physicians and the subsequent therapeutic intervention rate within the undernourished group. Primary physicians recognized undernutrition in 19 (43%) older persons. Undernutrition was recognized in 5 (12%) younger subjects. The rate of clinical recognition in older undernourished persons was significantly higher than that identified in the younger group (OR for older versus younger 5 5.47 [1.87 to 16.0]). Appropriate intervention was instituted in 6 (14%) older subjects and 2 (5%) younger subjects. However, the difference between therapeutic intervention rates in both groups was not significant (OR for older versus younger 5 3.08 [0.668 to 14.2]). Table 3 shows the pattern of weight changes documented in the charts during the 2-year follow-up period. The predominant weight trends exhibited in both age groups were progressive weight loss and failure to gain weight. No differences in weight trends were identified between age groups or among the three clinical presentations of undernutrition (Table 1). Four older subjects died during follow-up. They comprised 3 women and 1 man with a mean age of 77.7 years. Two subjects presented with hypoalbuminemia and subsequently developed progressive weight loss which persisted until death. The underlying causes of undernutrition were identified as chronic obstructive airway disease and carcinoma of the lung. One male subject with carcinoma of the pancreas presented with progressive weight loss and continued to lose weight till death. The fourth subject with congestive cardiac failure and low body weight, gained 2.9 kilograms (8 pounds) of dry weight prior to death, but still failed to achieve normal weight. There were no deaths in the younger study group. Compared with younger subjects, older subjects scored significantly higher on the oral control subscale but similarly on the dieting subscale (Figure 2). Subjects who intentionally decreased their nutritional intake
scored lower on the oral control subscale and higher on the dieting subscale than did those with unintended undernutrition. Subjects with a clinically significant number of depressive symptoms (11 or more on the GDS) demonstrated somewhat higher scores on both EAT-26 subscales, although neither difference was statistically significant. To evaluate whether the association of high oral control scores with increased age was due to confounding variables, analysis of covariance was used to control for GDS, history of eating disorder or life-long low weight, and whether the undernutrition was intentional. Controlling for these variables did not substantially change the relationship between age group and high oral control subscale scores (4.5 for older subjects versus 2.7 for younger subjects, P 5 0.018).
DISCUSSION Reported prevalence rates for undernutrition in older ambulant persons exhibit a wide variability. This may be partially attributed to the lack of consistency in the diagnostic criteria for undernutrition. In an earlier study, Manson et al identified undernutrition in 3.25% of older ambulatory patients (4). However, the sole diagnostic criteria in this study was weight loss. A comprehensive diagnostic approach may be crucial to the assessment of undernutrition, as highlighted by the findings of Miller et al. (5). In the latter study the prevalence of undernutrition in older male outpatients within the same group increased from 11% to 22% when hypoalbuminemia was included as an additional diagnostic criteria. We used similar criteria to those of Miller et al and identified undernutrition in 11% of older subjects. Fewer (7%) younger patients were found to be undernourished. As with any other syndrome, identification of the underlying cause forms the cornerstone of management. Earlier retrospective and cross sectional studies have reported a failure to identify the underlying cause of undernutrition in one quarter to one third of cases (22,23). Thompson et al carried out a case series chart review study of subjects with previously unexplained weight loss; after 2 years of follow-up, they were able to identify the cause of weight loss in 76% of subjects (24). In our study, only 7% of older persons and 10% of younger persons
Table 3. Longitudinal Weight Trends in Investigated Undernourished Subjects
Progressive weight loss Stable weight Weight gain
January 1998
Older (n 5 44)
Younger (n 5 41)
23 (52%) 14 (32%) 7 (16%)
17 (41%) 14 (34%) 10 (25%)
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 59
Undernutrition in Medical Outpatients/Wilson et al
Figure 2. EAT-26 oral control (A) and dieting subscale (B) scores in younger versus older subjects, patients with intentional versus unintentional weight loss, and patients with Geriatric Depression Scale (GDS) scores less than 11 versus 11 or more.
were considered to have unexplained undernutrition. These findings suggest that a detailed initial evaluation and close follow-up will reveal the underlying cause of undernutrition in most cases. In support of previous studies, we found that most of the common causes of undernutrition were amenable to treatment. Contrary to popular but unfounded belief, occult malignant disease is not a common cause of undernutrition in older persons when the cause is not readily evident. Earlier studies that targeted only unexplained and unintentional weight loss identified malignant disease in 16% to 36% of subjects (22–24). We examined all cases of undernutrition regardless of cause and identified malignant disease in only 5% of undernourished subjects. Our findings are similar to the findings of Morley et al who used a similar approach in older undernourished nursing home residents (2). They implicated malignant disease in only 7% of the undernourished patients. It is indisputable that cancers stand as a differential diagnosis when evaluating the older person with significant weight loss. However, recognition that this diagnosis is infre60
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
quently substantiated in patients without a history of cancer should serve to negate the unjustified pessimistic attitudes that may inhibit aggressive nutritional intervention. Depression and dysphoria are common in older patients but frequently remain unrecognized and untreated (25,26). Studies have also shown that anorexia and weight loss are more likely to occur as manifestations of depression in older persons (27). In earlier studies carried out by Rabinovitz et al and Marton et al, depression was implicated in the etiology of undernutrition in fewer than 10% of cases (22,23). Thompson et al identified depression in 18% of their study group (24). We identified an underlying affective disorder in 30% of our older subjects, rendering this the most common cause of undernutrition in our group. The significant role of depression in the etiogenesis of undernutrition has previously been identified in an earlier study by Morley et al, in which depression was noted to be the commonest cause of undernutrition, occurring in 36% of undernourished nursing home residents (2). We obtained GDS scores on all subjects in-
Undernutrition in Medical Outpatients/Wilson et al
cluded in our study, which probably enhanced our ability to identify an affective disorder as a cause of undernutrition. Recognition was further facilitated by 2 years of follow-up and additional use of DSM-IIIR criteria. About 50% of older Americans are completely edentulous, rendering them susceptible to masticatory difficulty and painful gums from poorly fitted dentures (28). However, few studies have examined the role of dental and oropharyngeal disease in undernourished older persons. We identified oropharyngeal disease as the underlying cause of undernutrition in 7% of older subjects. Heightened awareness of dental hygiene and health in older persons is a vital component of adequate nutritional maintenance. The role of iatrogenesis is often ignored in the evaluation of undernutrition. Intentional dietary modification geared specifically at weight reduction, the most common cause of undernutrition in the younger group, emerged as an uncommon cause of undernutrition within our older group. In support of our findings, the study by Wallace et al (29) of older male veteran outpatients found that the 2-year mortality in subjects with voluntary weight loss (36%) was similar to that seen in patients with involuntary weight loss (28%) and significantly higher than the mortality rate in patients with stable or increasing weight (11%). These results suggest that progressive weight loss may be dangerous in older persons even if it occurs ‘‘voluntarily.’’ Thus, it is of some concern that medically advised therapeutic diets were implicated in 7% of older persons and 12% of the younger group. It has been suggested in earlier studies that elimination of therapeutic diets in older persons may have a favorable impact on nutritional health (2). Most therapeutic diets used in geriatric medicine are recommended on the basis of data extrapolated from studies carried out in younger persons. Recent evidence suggests that the adoption of low cholesterol diets as primary prevention against atherosclerotic heart disease may be without benefit in persons over the age of 70 years (30). These findings may warrant a review of the clinical applicability and nutritional risk of dietary restrictions in older persons. The inappropriate and unsupervised use of digoxin and thyroxine were implicated as iatrogenic causes of undernutrition in 2 (2.6%) of our subjects. The pharmacokinetics of digoxin may be significantly altered by agerelated changes in renal function. Furthermore, the relative sarcopenia that occurs with aging decreases the volume of distribution of hydrophilic drugs, such as digoxin. These two factors, acting in concert, may account for the increased propensity of older persons to develop digoxin toxicity. Thompson et al identified four patients in their series with medication related weight loss (24). In their study theophylline, procainamide, thyroxine, and nitrofurantoin were identified as inducing significant weight loss as an adverse effect of their use. The
older adult is at an increased risk of drug toxicity due to unpredictable age related pharmacokinetic and pharmacodynamic changes (31). The coexistence of several diseases and polypharmacy may further increase the risk of adverse drug effects, such as anorexia, nausea, vomiting, and diarrhea. Other drugs in common use may result in gastrointestinal distress and consequent nutritional compromise in older persons. These include anticholinergic agents, psychotropic agents, antidepressants and nonsteroidal antiinflammatory agents. Recreational drugs, such as nicotine and alcohol may also contribute to undernutrition. Nicotine has been shown to depress the appetite and alter taste perception, thus detracting from the hedonic qualities of food (32,33). However, these agents did not emerge as causative factors in our study. Several diagnostic biochemical parameters have been proposed to facilitate accurate nutritional evaluation. The serum albumin level is used routinely, despite its low sensitivity, specificity and long half life of 10 to 14 days (34). However, serum albumin levels may be altered by stress related cytokine and counter regulatory hormone release (35,36). Direct down-regulation of albumin gene expression and transfer may also occur during acute illnesses, resulting in hypoalbuminemia (37). Prealbumin, transferrin, retinol-binding protein, fibronectin, and insulin-like growth factor-1 are other nutritional biochemical markers that have all been shown to have a relatively low diagnostic specificity and are often neither routinely available nor cost effective (34,38,39). The results of our study tend to downplay the significance of biochemical markers in the evaluation of undernutrition because most patients presented with weight problems, namely low body weight or clinically significant progressive weight loss, the latter being the most common. This trend suggests that body weight may be the single most important criterion in evaluation of the older person at risk for undernutrition. In this study, undernutrition in older subjects was associated less often with intentional weight loss and with increased levels of oral control compared with younger individuals. It would appear that many undernourished seniors have a form of eating disorder associated with high needs for oral control that will not be revealed by inquiry as to history of anorexia or bulimia nervosa. High oral control was associated to some degree with depressive symptoms, life-long low weight, and unintentional undernutrition, and these factors co-varied to some extent with age. However, after controlling for these covariates, age remained strongly related to high oral control in the undernourished subjects. These results suggest that some form of eating disorder may play a significant role in the nutritional problems of seniors and that this problem can be seen in those with ‘‘explained’’ as well as ‘‘unexplained’’ undernutrition. January 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 61
Undernutrition in Medical Outpatients/Wilson et al
This syndrome requires further clarification as it has some similarities to and some differences from classical anorexia nervosa of younger person (12). However, the findings of this study suggest that issues of oral control represent a major and perhaps a central feature of undernutrition in older persons. The benefits of aggressive early intervention in the management of undernutrition in hospitalized patients is well established and has been shown to result in improved outcomes in older patients (40). Despite these findings, available studies confirm the low clinical recognition rate of the syndrome of undernutrition. In the study by Manson et al, the diagnosis of undernutrition was overlooked in 48% of older patients with significant weight loss (4). Within the same patient group, there was a failure to institute therapeutic intervention in over three quarters of undernourished subjects. The results of our study support these findings. Undernutrition was identified by the primary physician in only 43% of the older subjects, only one third of whom received appropriate intervention. However, the rate of clinical recognition of undernutrition and subsequent treatment was much lower in younger persons. This may indicate some degree of recognition by the health care providers of the increased vulnerability of older patients to declines in health from undernutrition. The results of our study highlight the fact that in the absence of meaningful intervention specifically directed at improving nutritional status, the majority of older patients with weight problems will fail to attain their ideal body weight or continue to lose weight. In an earlier study targeting a general medical group, only 60% of undernourished subjects responded favorably following institution of appropriate therapy. The prognosis was significantly worse in older subjects (23). These findings in addition to the high mortality identified in the study by Wallace et al (29) support the need for increased nutritional surveillance and early, aggressive nutritional therapy within the geriatric group. Undernutrition may bear a more grave prognosis in older persons as evidenced by the pattern of mortality noted in our study group, with all deaths occurring in older persons. It is well recognized that low body weight is highly predictive of mortality in older persons (6,41). The mortality of 9% in our study contrasts notably with several previous studies with mortality rates ranging from 25% to 38% (22,23,29). However review of these studies reveal a much higher proportion of patients with malignant neoplastic disease. Our mortality rate is identical to that of Thompson et al who targeted a similar group comprising older community-dwelling citizens (24). 62
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
CONCLUSION Overall, our findings demonstrate the relatively high prevalence of unrecognized undernutrition in both older and younger persons with the phenomenon occurring more often in the older subjects. The common causes of undernutrition identified appeared to be readily amenable to treatment, with depression ranking as the most common cause overall. It is evident from our findings that recognition of undernutrition and therapeutic nutritional intervention are suboptimal. The routine incorporation of a detailed nutritional history and targeted nutritional intervention should form part of a proper evaluation of older outpatients. Health care workers and other professionals should be trained in the rudiments of nutritional screening to facilitate the identification and monitoring of high risk groups within the community. As the geriatric group expands, the syndrome of undernutrition will be encountered more often by health care providers. An increase in the level of nutritional awareness among physicians and other health care workers is crucial to the prevention and early recognition of undernutrition in older persons. As interventions are designed and tested, the high oral control needs in older undernourished subjects should be kept in mind. These may require modifications in the usual therapeutic approach that accommodate the patients’ oral control needs. Alternatively, this finding may indicate a type of eating disorder that needs to be addressed directly via behavioral, psychotherapeutic, or pharmacological means.
ACKNOWLEDGMENTS The authors gratefully acknowledge the cooperation of the physicians and nurses in the General Internal Medicine office practice and the assistance of Dale Kraenzle, Peter Jiang, John Caster, and Robert Hagan in data collection. Any errors or omissions are solely those of the authors.
REFERENCES 1. Mowe-Morten KI, Bohmer T. The prevalence of protein calorie undernutrition in a population of hospitalized elderly patients. J Am Geriatr Soc. 1991;39:1089 –1092. 2. Morley JE, Kraenzle D. Causes of weight loss in a community nursing home. J Am Geriatr Soc. 1994;42:583–585. 3. Morley JE, Silver AJ, Miller DK, Rubenstein LZ. The anorexia of the elderly. NY Acad Sci. 1989;575:50 –59. 4. Manson A, Sheer S. Malnutrition in elderly ambulatory patients. Am J Pub Health. 1991;81:1195–1197. 5. Miller DK, Morley JE, Rubenstein LZ, et al. Formal geriatric assessment instruments and the care of older general medical outpatients. J Am Geriatr Soc. 1990;38:645– 651. 6. Morley JE, Mooradian AD, Silver AJ, et al. Nutrition in the elderly. Ann Intern Med. 1988;109:890 –904. 7. Tayback M, Kumanyika S, Chee E. Body weight as a risk factor in the elderly. Arch Intern Med. 1990;150:1065–1072.
Undernutrition in Medical Outpatients/Wilson et al 8. Harris T, Cook F, Garrison R, et al. Body mass index and mortality for nonsmoking older persons. JAMA. 1988;259:1520 –1524. 9. Rudman D, Feller AG, Nagray HS, et al. Relationship of serum albumin concentration to death rate in nursing home men. JPEN. 1987;11:360 –363. 10. Galanos AN, Pieper CF, Cornoni-Huntley JC, et al. Nutrition and function: is there a relationship between body mass index and the functional capabilities of community-dwelling elderly? J Am Geriatr Soc. 1994;42:368 –373. 11. Hubert HB, Block DA, Fries JF. Risk factors for physical disability in an aging cohort: The NHANES I epidemiologic follow-up study. J Rheumatol. 1993;20:480 – 488. 12. Miller DK, Morley JE, Rubenstein LZ, Pietruszka FM. Abnormal eating attitudes and body image in older undernourished individuals. J Am Geriatr Soc. 1991;39:462– 466. 13. Pinholt EM, Kroenke K, Hanley JF, et al. Functional assessment of the elderly: a comparison of standard instruments with clinical judgment. Arch Intern Med. 1987;147:484 – 488. 14. Master AM, Lasser RP, Beckman G. Table of average weight and height of Americans aged 65–94 years. JAMA. 1960;172:658 – 662. 15. Anon. Weight by height and age of adults 18 –74 years: United States, 1971–1974. DHEW Publication No. (PHS) 79-1656, Series II, No. 208. Hyattsville, Maryland: US Department of Health, Education, and Welfare; 1979:34. 16. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state.’’ A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189 –198. 17. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res. 1982–3;17:37– 49. 18. Garner DM, Olmsted MP, Polivy J. Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. Int J Eating Disorders. 1983;2(2):15–34. 19. American Psychiatric Association, Committee on Nomenclature and Statistics. Diagnostic and Statistical Manual of Mental Disorders, Revised. 3rd ed. Washington, DC: American Psychiatric Association; 1987. 20. Sato T. Confidence limits for the common odds ratio based on the asymptotic distribution of the Mantel-Haenszel estimator. Biometrics. 1990;46:71– 80. 21. Norusis MJ. SPSS/PC1 Advanced Statistics 4.0. Chicago, Illinois: SPSS Inc.; 1990. 22. Rabinovitz M, Pitlik SD, Leifer M, et al. Unintentional weight loss: a retrospective analysis of 154 cases. Arch Intern Med. 1986;146: 186 –187. 23. Marton KI, Sox HC, Krupp JR. Involuntary weight loss: diagnostic and prognostic significance. Ann Intern Med. 1981;95:568 –574. 24. Thompson MP, Morris LK. Unexplained weight loss in the ambulatory elderly. J Am Geriatr Soc. 1991;39:497–501. 25. Blazer D, Bachar JR, Hughes DC. Major depression with
26. 27. 28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38. 39.
40.
41.
melancholia: a comparison of middle-aged and elderly adults. J Am Geriatr Soc. 1987;35:927–932. Blazer D, Williams CD. Epidemiology of dysphoria and depression in an elderly population. Am J Psychiatry. 1980;137:439 – 444. Blazer D. The diagnosis of depression in the elderly. J Am Geriatr Soc. 1980;28:52–58. Weintrab JA, Burt BA. Oral health status in the United States: Tooth loss and edentulism. Dent Educ. 1985;49:368 –376. Wallace JI, Schwartz RS, LaCroix AZ, et al. Involuntary weight loss in older outpatients: incidence and clinical significance. J Am Geriatr Soc. 1995;43:329 –337. Krumholz HM, Seeman TE, Merrill SS. Lack of association between cholesterol and coronary heart disease mortality and morbidity and all cause mortality in persons older than 70 years. JAMA. 1994;272: 1335–1340. Schwartz JB. Clinical pharmacology. In: Hazzard WR, Bierman EL, Blass JP, et al., eds. Principles of Geriatric Medicine and Gerontology. 3rd ed. New York: McGraw-Hill; 1994:259 –275. Gross J, Stitzer ML. Nicotine replacement: Ten weeks effect on tobacco withdrawal symptoms. Psychopharmacology. 1989;98:334 – 341. Millie A, Berriau T, Paget D, et al. Can weight gain during weaning from smoking be limited using nicotine gum? Revue de Pneumologie Clinique. 1989;45:243–249. Heymsfield SB, Tighe A, Wang Z. Nutritional assessment by anthropometric and biochemical methods. In: Shils ME, Olson JA, Shike M, eds. Modern Nutrition in Health and Disease. 8th ed. Philadelphia: Lea & Ferbiger; 1994:812– 841. Pomposelli JJ, Flores EA, Bistrian BR. Role of biochemical mediators in clinical nutrition and surgical metabolism. JPEN. 1988;12: 212–218. Boosalis MG, Ott L, Levine AS, et al. Relationship of visceral proteins to nutritional status in chronic and acute stress. Crit Care Med. 1989;17:741–747. Perlmutter DH, Dinarello CA, Punsal PI, et al. Cachectin/tumor necrosis factor regulates hepatic phase gene expression. J Clin Invest. 1986;78:1349 –1354. Fleck A. Clinical and nutritional aspects of changes in acute phase proteins during inflammation. Proc Nutr Soc. 1989;48:347–354. Lipschitz DA. Nutrition and aging. In: Evans GJ, Williams FT, eds. Oxford Textbook of Geriatric Medicine. New York: Oxford University Press; 1992. Zaloga GP. Timing and route of nutritional support. In: Zaloga GP, ed. Nutrition in Critical Care. St. Louis: Mosby Year Book; 1994: 297–330. Andres R. Mortality and obesity: The rationale for age-specific height-weight tables. In: Andres R, Bierman EL, Hazzard WR, eds. Principles of Geriatric Medicine. New York: McGraw-Hill; 1984:31– 38.
January 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 63
trition differed between age groups but was deemed treatable in nearly 90% of all subjects. Undernutrition was recognized in 19 (43%) older subjects and 5 (12%) younger subjects (OR 5 5.47 [1.87 to 16.0]), and appropriate intervention(s) were instituted in 6 (14%) and 2 (5%) of older and younger subjects, respectively (OR 5 3.08 [0.668 to 14.2]). Older subjects scored higher on the EAT-26 oral control subscale than did younger subjects (4.7 versus 2.5, P 5 0.004) but similarly on the EAT-26 dieting subscale (5.2 versus 6.3, P 5 0.332); these relationships did not change with control for potentially confounding variables. CONCLUSIONS: In this study, undernutrition was relatively common, usually amenable to treatment, but frequently undetected and undertreated in both older and younger medical outpatients. Older undernourished subjects exhibited higher oral control needs than younger persons, which may have implications for the pathophysiology and treatment of their malnutrition. Further improvement in detection and intervention is warranted in both younger and older age groups. Am J Med. 1998;104:56 – 63. q1998 by Excerpta Medica, Inc.
U
ings, there has been a paucity of research directed towards determining the prevalence and etiology of undernutrition in older outpatients. Our previous study of older male veteran outpatients (12) demonstrated that undernourished patients had increased levels of oral control needs but similar dieting behaviors compared with normally nourished patients. An important minority of undernourished patients in that study also evidenced various degrees of a distorted body image that bore some similarity to anorexia nervosa in younger persons. The relative rarity of documentation of undernutrition as a medical diagnosis or contributory cause of death has been noted in several studies (1,5,13). This phenomenon may hinder practitioners’ appreciation of the frequency with which undernutrition occurs in medical outpatients and the impact of undernutrition on subsequent decrements in health. It may also hamper early initiation of appropriate therapeutic interventions that have the potential for preventing nutritionally related health declines. This study was designed to assess the prevalence, common causes, frequency of clinical recognition, and extent of therapy for undernutrition in community-dwelling medical outpatients. The results obtained in older subjects were compared with the findings in younger subjects.
ndernutrition is not a rare occurrence in older hospital patients and nursing home residents (1,2). Within acute care hospitals and chronic institutions, prevalence rates of undernutrition range from 17% to 65% (3). Available data suggests that the prevalence of diagnosed undernutrition in community dwelling senior citizens may be less than 5% (4). However, it has been estimated that one fifth to one third of older persons may have evidence of undernutrition, some of which might be considered ‘‘subclinical’’ (5,6). Existing data clearly demonstrate that undernutrition adversely affects both length and quality of life (7–11). In particular a positive correlation between undernutrition and poor functional status has been identified in older persons (10,11). Despite these find-
From the St. Louis University Health Sciences Center (M-MGW, SV, DL, JM, DK), and the St. Louis Veterans Affairs Medical Center (JEM, DKM), St. Louis, Missouri. Requests for reprints should be addressed to Douglas K. Miller, MD, Division of Geriatric Medicine, 1402 S. Grand, Rm. M238, St. Louis, Missouri 63104. Dr. Liu’s involvement in the study was supported in part by an educational grant from the John A. Hartford Foundation, which took no direct role in the investigation. Manuscript submitted January 30, 1997 and accepted in revised form August 9, 1997. 56
q1998 by Excerpta Medica, Inc. All rights reserved.
0002-9343/98/$19.00 PII S0002-9343(97)00279-9
Undernutrition in Medical Outpatients/Wilson et al
METHODS Study Site, Subjects, and Inclusion/Exclusion Criteria The faculty in the Division of General Internal Medicine at Saint Louis University (SLU) Health Sciences Center provides direct services to about 1600 general adult medical outpatients through 5 half-day sessions each week. From September 1990 to July 1993, all active patients 14 years of age and older attending any of these sessions were considered for entry into the study. Exclusion criteria included clinical features or a documented diagnosis of protein-losing syndromes (eg, nephrotic syndrome, protein-losing enteropathy or severe liver disease). Patients with a history of any cancer, other than skin cancer, within the preceding 10 years were also excluded. Each patient was screened only once. Charts of patients not meeting the exclusion criteria were screened for 1 of the following indices of undernutrition: (1) A body weight less than 90% of the appropriate median weight using age, gender, and height adjusted norms. Master’s reference tables (14) were used for those 65 years of age and older. Similarly adjusted norms from the First National Health and Nutritional Examination Survey were used for those younger than 65 years (15). (2) Documented weight loss, unexplained by fluid loss, in excess of 2.27 kilograms (5 pounds) over the preceding 6 months. (3) Serum albumin levels less than 3.5 g/dL.
Recruitment and Evaluation of Study Participants Chart reviews were performed on the afternoon prior to the patients’ scheduled visit. If the chart review identified one of the criteria for undernutrition, the patient was informed of the purposes, obligations, and risks of the study, and signed consent was obtained from willing participants. A trained research assistant obtained the following information from subjects and their charts using a standardized protocol: (1) demographic data, (2) socioeconomic circumstances, (3) past medical illnesses, (4) medication history, and (5) recreational and social habits. A nutritional assessment was carried out on each recruited subject. This included inquiry into dietary habits, baseline body weight, and time course of any changes in weight. Mini-mental State Examination (16) and Geriatric Depression Scale (GDS) (17) scores were obtained on each subject. The dieting and oral control subscales of the EAT-26 questionnaire were administered to each subject in an attempt to identify inappropriate nutritional attitudes (12,18). The EAT-26 was developed by Garner and colleagues as a multidimensional eating disorder inventory for facilitating identification of anorexia nervosa and bulimia in younger patients (18). It has also proven useful in
identifying abnormal eating attitudes and body image in undernourished older outpatients (12). The EAT-26 inventory has three subscales: dieting, oral control, and bulimia. The dieting subscale includes questions assessing perception of body image, eating habits, and nutritional attitudes. The oral control subscale evaluates the individual’s attempted control over the circumstances of eating. The bulimia subscale assesses purging behaviors, was not helpful in the prior study (12) and thus was omitted from the present investigation. After the initial assessment, enrolled subjects were monitored over the next several months by serial review of the outpatient records to identify the major cause of undernutrition in each subject, to assess the clinical detection rate and extent of therapeutic intervention by the primary physician, and to evaluate the course of weight change over time. The charts were reviewed independently by two reviewers (MMW and DKM), each of whom had access to all the data collated over the 2-year follow-up period. Criteria for identifying an underlying cause of undernutrition were established prior to the chart reviews by consensus conference among three of the investigators (MMW, DKM, and JEM). For example, the diagnosis of depression was made if chart review revealed: (1) Diagnostic Statistical Manual (DSM) IIIR (19) criteria for major depression or (2) a GDS score over 14 (17) in the presence of one or more DSM-IIIR symptoms. (The full list of diagnostic criteria is available from the first author on request.) A major cause of undernutrition was accepted as established when both reviewers agreed that it was the primary contributor to undernutrition. Any disagreements on blinded review between the two primary reviewers were presented to a third reviewer (JEM), who acted as adjudicator. Four cases required adjudication; in each case the final, adjudicated major cause had been identified as the primary contributor by one primary reviewer and as a contributing cause by the other primary reviewer. The syndrome of undernutrition was considered to have been recognized by the primary physician if reference was made to a nutritional problem in the clinic notes or problem list at the front of the chart. For the purpose of this study, appropriate therapeutic intervention was considered to be comprised of one or more of the following: (1) identification and treatment, where possible, of the identified major contributory cause of undernutrition; (2) a critical review of the subject’s diet and institution of appropriate therapeutic nutritional support; and (3) serial anthropometric or biochemical monitoring to assess the subject’s response to treatment. Approval was obtained from the SLU Health Sciences Center’s Institutional Review Board for all research procedures. Informed consent was obtained from all subjects in conformance with assurances given by SLU to the DeJanuary 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 57
Undernutrition in Medical Outpatients/Wilson et al
Table 1. Clinical Presentation of Undernourished Patients
Weight loss ($5 lb/2.27 kg) Low body weight Hypoalbuminemia
Older (n 5 46)
Younger (n 5 44)
21 (46%) 11 (24%) 14 (30%)
29 (66%) 10 (23%) 5 (11%)
partment of Health and Human Services to protect the rights of human subjects.
Statistical Analysis Significant differences between groups was evaluated using Sato’s method to determine odds ratios (ORs) and 95% confidence intervals (CIs) for dichotomous variables (20) and Student’s t-test with two-tailed tests of significance for continuous variables. Analysis of covariance was used to evaluate for the possibility that confounding variables explained the apparent association between age and high EAT-26 oral control subscale scores (21).
RESULTS The charts of 1017 patients were screened for undernutrition; 408 patients were age 65 or older, and 609 patients were age 18 to 64. Using the criteria as previously defined, 46 (11%) older persons comprising 30 women and 16 men and 44 (7%) younger persons comprising 30 women and 14 men were found to be undernourished (OR [95% CI] for older versus younger 5 1.65 [1.05 to 2.51]; OR for female versus male 5 1.08 [0.689 to 1.71]). Five patients (aged 72, 69, 55, 42, and 37 years) screened positive but refused to participate. The 72-year-old presented with low weight and was thought likely to have abnormal eating attitudes based on screening chart review. He refused to participate on two separate office visits, declining to give a reason. The four other patients presented with weight loss and cited lack of time as their reason for not wanting to participate. Weight loss was the most common method of presentation for both age groups (Table 1). Younger persons were more likely than older persons to present with weight loss and were less likely to present with hypoalbuminemia (OR for younger versus older with weight loss 5 2.30 [0.997 to 5.31] and with hypoalbuminemia 5 0.293 [0.099 to 0.865]). Forty-four older and 41 younger subjects were interviewed at baseline and followed for 2 years. The primary cause of undernutrition was identified in 93% of older persons and 90% of younger persons (Table 2). Of the 78 patients with an explanation for their weight loss, 69 (88.5%) were felt to have potentially treatable causes for their undernutrition. Occult malignancies were impli58
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
cated in only 5% of the total sample, occurring only in older subjects. Two subjects were eventually diagnosed with gastric carcinoma and the other two subjects were found to have adenocarcinoma of the pancreas and lung. Depression was identified as the commonest cause of undernutrition in older persons, occurring in 30% of the group and was the apparent cause of undernutrition in 15% of younger subjects. The difference was not statistically significant. Undernutrition resulting from physician intervention was identified in association with prescribed medication and therapeutic diets. Persistent anorexia associated with oral digoxin therapy resulted in undernutrition in one older subject based on high serum level plus improvement in appetite and weight gain with reduction in digoxin dosage. Thyrotoxicosis resulting from inappropriately high doses of thyroxine was implicated as the primary cause of undernutrition in a younger subject. Two older persons on a low salt diet and a third on an antidiabetic diet developed undernutrition as a direct consequence of these medically prescribed therapeutic diets. In the younger subjects, low cholesterol diets were implicated in three subjects and low salt diets in two subjects. Intentional weight loss resulting from weight reducing diets, the commonest cause of undernutrition in younger subjects, was relatively rare in the older group. Intentional weight reduction was the only cause of undernutrition that exTable 2. Primary Causes of Weight Loss in Investigated Subjects
Depression Intentional* Therapeutic diet Oropharyngeal disease Chronic pain Cancer Poorly controlled diabetes Alcohol Anorexia nervosa Cardiac failure Cholelithiasis Obstructive airway disease Myopathy Dementia Difficulty with food preparation Postsurgery Thyrotoxicosis Medication Post menopausal syndrome Unexplained * P 5 0.01
Older (n 5 44)
Younger (n 5 41)
13 (29.5%) 1 (2.3%) 3 (6.8%) 3 (6.8%) 1 (2.3%) 4 (9.1%) 4 (9.1%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 1 (2.3%) 2 (4.5%)
6 (14.6%) 8 (19.5%) 5 (12.2%) 2 (4.8%) 4 (9.7%) 0 (0%) 3 (7.3%) 2 (4.8%) 1 (2.4%) 2 (4.8%) 0 1 (2.4%) 0 0
1 (2.3%) 1 (2.3%) 0 1 (2.3%) 1 (2.3%) 3 (6.8%)
1 (2.4%) 0 1 (2.4%) 1 (2.4%) 0 4 (9.7%)
Undernutrition in Medical Outpatients/Wilson et al
Figure 1. Percentage of older and younger subjects whose undernutrition was recognized (open bars) and who received appropriate treatment (closed bars).
hibited a significant difference in prevalence between younger and older subjects (OR for older versus younger 5 0.096 [0.015 to 0.624]) (Table 2). In only 3 older (6.8%) and 4 younger (9.7%) subjects was no major cause of undernutrition identified. Figure 1 highlights the clinical recognition rate of undernutrition by primary physicians and the subsequent therapeutic intervention rate within the undernourished group. Primary physicians recognized undernutrition in 19 (43%) older persons. Undernutrition was recognized in 5 (12%) younger subjects. The rate of clinical recognition in older undernourished persons was significantly higher than that identified in the younger group (OR for older versus younger 5 5.47 [1.87 to 16.0]). Appropriate intervention was instituted in 6 (14%) older subjects and 2 (5%) younger subjects. However, the difference between therapeutic intervention rates in both groups was not significant (OR for older versus younger 5 3.08 [0.668 to 14.2]). Table 3 shows the pattern of weight changes documented in the charts during the 2-year follow-up period. The predominant weight trends exhibited in both age groups were progressive weight loss and failure to gain weight. No differences in weight trends were identified between age groups or among the three clinical presentations of undernutrition (Table 1). Four older subjects died during follow-up. They comprised 3 women and 1 man with a mean age of 77.7 years. Two subjects presented with hypoalbuminemia and subsequently developed progressive weight loss which persisted until death. The underlying causes of undernutrition were identified as chronic obstructive airway disease and carcinoma of the lung. One male subject with carcinoma of the pancreas presented with progressive weight loss and continued to lose weight till death. The fourth subject with congestive cardiac failure and low body weight, gained 2.9 kilograms (8 pounds) of dry weight prior to death, but still failed to achieve normal weight. There were no deaths in the younger study group. Compared with younger subjects, older subjects scored significantly higher on the oral control subscale but similarly on the dieting subscale (Figure 2). Subjects who intentionally decreased their nutritional intake
scored lower on the oral control subscale and higher on the dieting subscale than did those with unintended undernutrition. Subjects with a clinically significant number of depressive symptoms (11 or more on the GDS) demonstrated somewhat higher scores on both EAT-26 subscales, although neither difference was statistically significant. To evaluate whether the association of high oral control scores with increased age was due to confounding variables, analysis of covariance was used to control for GDS, history of eating disorder or life-long low weight, and whether the undernutrition was intentional. Controlling for these variables did not substantially change the relationship between age group and high oral control subscale scores (4.5 for older subjects versus 2.7 for younger subjects, P 5 0.018).
DISCUSSION Reported prevalence rates for undernutrition in older ambulant persons exhibit a wide variability. This may be partially attributed to the lack of consistency in the diagnostic criteria for undernutrition. In an earlier study, Manson et al identified undernutrition in 3.25% of older ambulatory patients (4). However, the sole diagnostic criteria in this study was weight loss. A comprehensive diagnostic approach may be crucial to the assessment of undernutrition, as highlighted by the findings of Miller et al. (5). In the latter study the prevalence of undernutrition in older male outpatients within the same group increased from 11% to 22% when hypoalbuminemia was included as an additional diagnostic criteria. We used similar criteria to those of Miller et al and identified undernutrition in 11% of older subjects. Fewer (7%) younger patients were found to be undernourished. As with any other syndrome, identification of the underlying cause forms the cornerstone of management. Earlier retrospective and cross sectional studies have reported a failure to identify the underlying cause of undernutrition in one quarter to one third of cases (22,23). Thompson et al carried out a case series chart review study of subjects with previously unexplained weight loss; after 2 years of follow-up, they were able to identify the cause of weight loss in 76% of subjects (24). In our study, only 7% of older persons and 10% of younger persons
Table 3. Longitudinal Weight Trends in Investigated Undernourished Subjects
Progressive weight loss Stable weight Weight gain
January 1998
Older (n 5 44)
Younger (n 5 41)
23 (52%) 14 (32%) 7 (16%)
17 (41%) 14 (34%) 10 (25%)
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 59
Undernutrition in Medical Outpatients/Wilson et al
Figure 2. EAT-26 oral control (A) and dieting subscale (B) scores in younger versus older subjects, patients with intentional versus unintentional weight loss, and patients with Geriatric Depression Scale (GDS) scores less than 11 versus 11 or more.
were considered to have unexplained undernutrition. These findings suggest that a detailed initial evaluation and close follow-up will reveal the underlying cause of undernutrition in most cases. In support of previous studies, we found that most of the common causes of undernutrition were amenable to treatment. Contrary to popular but unfounded belief, occult malignant disease is not a common cause of undernutrition in older persons when the cause is not readily evident. Earlier studies that targeted only unexplained and unintentional weight loss identified malignant disease in 16% to 36% of subjects (22–24). We examined all cases of undernutrition regardless of cause and identified malignant disease in only 5% of undernourished subjects. Our findings are similar to the findings of Morley et al who used a similar approach in older undernourished nursing home residents (2). They implicated malignant disease in only 7% of the undernourished patients. It is indisputable that cancers stand as a differential diagnosis when evaluating the older person with significant weight loss. However, recognition that this diagnosis is infre60
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
quently substantiated in patients without a history of cancer should serve to negate the unjustified pessimistic attitudes that may inhibit aggressive nutritional intervention. Depression and dysphoria are common in older patients but frequently remain unrecognized and untreated (25,26). Studies have also shown that anorexia and weight loss are more likely to occur as manifestations of depression in older persons (27). In earlier studies carried out by Rabinovitz et al and Marton et al, depression was implicated in the etiology of undernutrition in fewer than 10% of cases (22,23). Thompson et al identified depression in 18% of their study group (24). We identified an underlying affective disorder in 30% of our older subjects, rendering this the most common cause of undernutrition in our group. The significant role of depression in the etiogenesis of undernutrition has previously been identified in an earlier study by Morley et al, in which depression was noted to be the commonest cause of undernutrition, occurring in 36% of undernourished nursing home residents (2). We obtained GDS scores on all subjects in-
Undernutrition in Medical Outpatients/Wilson et al
cluded in our study, which probably enhanced our ability to identify an affective disorder as a cause of undernutrition. Recognition was further facilitated by 2 years of follow-up and additional use of DSM-IIIR criteria. About 50% of older Americans are completely edentulous, rendering them susceptible to masticatory difficulty and painful gums from poorly fitted dentures (28). However, few studies have examined the role of dental and oropharyngeal disease in undernourished older persons. We identified oropharyngeal disease as the underlying cause of undernutrition in 7% of older subjects. Heightened awareness of dental hygiene and health in older persons is a vital component of adequate nutritional maintenance. The role of iatrogenesis is often ignored in the evaluation of undernutrition. Intentional dietary modification geared specifically at weight reduction, the most common cause of undernutrition in the younger group, emerged as an uncommon cause of undernutrition within our older group. In support of our findings, the study by Wallace et al (29) of older male veteran outpatients found that the 2-year mortality in subjects with voluntary weight loss (36%) was similar to that seen in patients with involuntary weight loss (28%) and significantly higher than the mortality rate in patients with stable or increasing weight (11%). These results suggest that progressive weight loss may be dangerous in older persons even if it occurs ‘‘voluntarily.’’ Thus, it is of some concern that medically advised therapeutic diets were implicated in 7% of older persons and 12% of the younger group. It has been suggested in earlier studies that elimination of therapeutic diets in older persons may have a favorable impact on nutritional health (2). Most therapeutic diets used in geriatric medicine are recommended on the basis of data extrapolated from studies carried out in younger persons. Recent evidence suggests that the adoption of low cholesterol diets as primary prevention against atherosclerotic heart disease may be without benefit in persons over the age of 70 years (30). These findings may warrant a review of the clinical applicability and nutritional risk of dietary restrictions in older persons. The inappropriate and unsupervised use of digoxin and thyroxine were implicated as iatrogenic causes of undernutrition in 2 (2.6%) of our subjects. The pharmacokinetics of digoxin may be significantly altered by agerelated changes in renal function. Furthermore, the relative sarcopenia that occurs with aging decreases the volume of distribution of hydrophilic drugs, such as digoxin. These two factors, acting in concert, may account for the increased propensity of older persons to develop digoxin toxicity. Thompson et al identified four patients in their series with medication related weight loss (24). In their study theophylline, procainamide, thyroxine, and nitrofurantoin were identified as inducing significant weight loss as an adverse effect of their use. The
older adult is at an increased risk of drug toxicity due to unpredictable age related pharmacokinetic and pharmacodynamic changes (31). The coexistence of several diseases and polypharmacy may further increase the risk of adverse drug effects, such as anorexia, nausea, vomiting, and diarrhea. Other drugs in common use may result in gastrointestinal distress and consequent nutritional compromise in older persons. These include anticholinergic agents, psychotropic agents, antidepressants and nonsteroidal antiinflammatory agents. Recreational drugs, such as nicotine and alcohol may also contribute to undernutrition. Nicotine has been shown to depress the appetite and alter taste perception, thus detracting from the hedonic qualities of food (32,33). However, these agents did not emerge as causative factors in our study. Several diagnostic biochemical parameters have been proposed to facilitate accurate nutritional evaluation. The serum albumin level is used routinely, despite its low sensitivity, specificity and long half life of 10 to 14 days (34). However, serum albumin levels may be altered by stress related cytokine and counter regulatory hormone release (35,36). Direct down-regulation of albumin gene expression and transfer may also occur during acute illnesses, resulting in hypoalbuminemia (37). Prealbumin, transferrin, retinol-binding protein, fibronectin, and insulin-like growth factor-1 are other nutritional biochemical markers that have all been shown to have a relatively low diagnostic specificity and are often neither routinely available nor cost effective (34,38,39). The results of our study tend to downplay the significance of biochemical markers in the evaluation of undernutrition because most patients presented with weight problems, namely low body weight or clinically significant progressive weight loss, the latter being the most common. This trend suggests that body weight may be the single most important criterion in evaluation of the older person at risk for undernutrition. In this study, undernutrition in older subjects was associated less often with intentional weight loss and with increased levels of oral control compared with younger individuals. It would appear that many undernourished seniors have a form of eating disorder associated with high needs for oral control that will not be revealed by inquiry as to history of anorexia or bulimia nervosa. High oral control was associated to some degree with depressive symptoms, life-long low weight, and unintentional undernutrition, and these factors co-varied to some extent with age. However, after controlling for these covariates, age remained strongly related to high oral control in the undernourished subjects. These results suggest that some form of eating disorder may play a significant role in the nutritional problems of seniors and that this problem can be seen in those with ‘‘explained’’ as well as ‘‘unexplained’’ undernutrition. January 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 61
Undernutrition in Medical Outpatients/Wilson et al
This syndrome requires further clarification as it has some similarities to and some differences from classical anorexia nervosa of younger person (12). However, the findings of this study suggest that issues of oral control represent a major and perhaps a central feature of undernutrition in older persons. The benefits of aggressive early intervention in the management of undernutrition in hospitalized patients is well established and has been shown to result in improved outcomes in older patients (40). Despite these findings, available studies confirm the low clinical recognition rate of the syndrome of undernutrition. In the study by Manson et al, the diagnosis of undernutrition was overlooked in 48% of older patients with significant weight loss (4). Within the same patient group, there was a failure to institute therapeutic intervention in over three quarters of undernourished subjects. The results of our study support these findings. Undernutrition was identified by the primary physician in only 43% of the older subjects, only one third of whom received appropriate intervention. However, the rate of clinical recognition of undernutrition and subsequent treatment was much lower in younger persons. This may indicate some degree of recognition by the health care providers of the increased vulnerability of older patients to declines in health from undernutrition. The results of our study highlight the fact that in the absence of meaningful intervention specifically directed at improving nutritional status, the majority of older patients with weight problems will fail to attain their ideal body weight or continue to lose weight. In an earlier study targeting a general medical group, only 60% of undernourished subjects responded favorably following institution of appropriate therapy. The prognosis was significantly worse in older subjects (23). These findings in addition to the high mortality identified in the study by Wallace et al (29) support the need for increased nutritional surveillance and early, aggressive nutritional therapy within the geriatric group. Undernutrition may bear a more grave prognosis in older persons as evidenced by the pattern of mortality noted in our study group, with all deaths occurring in older persons. It is well recognized that low body weight is highly predictive of mortality in older persons (6,41). The mortality of 9% in our study contrasts notably with several previous studies with mortality rates ranging from 25% to 38% (22,23,29). However review of these studies reveal a much higher proportion of patients with malignant neoplastic disease. Our mortality rate is identical to that of Thompson et al who targeted a similar group comprising older community-dwelling citizens (24). 62
January 1998
THE AMERICAN JOURNAL OF MEDICINEt Volume 104
CONCLUSION Overall, our findings demonstrate the relatively high prevalence of unrecognized undernutrition in both older and younger persons with the phenomenon occurring more often in the older subjects. The common causes of undernutrition identified appeared to be readily amenable to treatment, with depression ranking as the most common cause overall. It is evident from our findings that recognition of undernutrition and therapeutic nutritional intervention are suboptimal. The routine incorporation of a detailed nutritional history and targeted nutritional intervention should form part of a proper evaluation of older outpatients. Health care workers and other professionals should be trained in the rudiments of nutritional screening to facilitate the identification and monitoring of high risk groups within the community. As the geriatric group expands, the syndrome of undernutrition will be encountered more often by health care providers. An increase in the level of nutritional awareness among physicians and other health care workers is crucial to the prevention and early recognition of undernutrition in older persons. As interventions are designed and tested, the high oral control needs in older undernourished subjects should be kept in mind. These may require modifications in the usual therapeutic approach that accommodate the patients’ oral control needs. Alternatively, this finding may indicate a type of eating disorder that needs to be addressed directly via behavioral, psychotherapeutic, or pharmacological means.
ACKNOWLEDGMENTS The authors gratefully acknowledge the cooperation of the physicians and nurses in the General Internal Medicine office practice and the assistance of Dale Kraenzle, Peter Jiang, John Caster, and Robert Hagan in data collection. Any errors or omissions are solely those of the authors.
REFERENCES 1. Mowe-Morten KI, Bohmer T. The prevalence of protein calorie undernutrition in a population of hospitalized elderly patients. J Am Geriatr Soc. 1991;39:1089 –1092. 2. Morley JE, Kraenzle D. Causes of weight loss in a community nursing home. J Am Geriatr Soc. 1994;42:583–585. 3. Morley JE, Silver AJ, Miller DK, Rubenstein LZ. The anorexia of the elderly. NY Acad Sci. 1989;575:50 –59. 4. Manson A, Sheer S. Malnutrition in elderly ambulatory patients. Am J Pub Health. 1991;81:1195–1197. 5. Miller DK, Morley JE, Rubenstein LZ, et al. Formal geriatric assessment instruments and the care of older general medical outpatients. J Am Geriatr Soc. 1990;38:645– 651. 6. Morley JE, Mooradian AD, Silver AJ, et al. Nutrition in the elderly. Ann Intern Med. 1988;109:890 –904. 7. Tayback M, Kumanyika S, Chee E. Body weight as a risk factor in the elderly. Arch Intern Med. 1990;150:1065–1072.
Undernutrition in Medical Outpatients/Wilson et al 8. Harris T, Cook F, Garrison R, et al. Body mass index and mortality for nonsmoking older persons. JAMA. 1988;259:1520 –1524. 9. Rudman D, Feller AG, Nagray HS, et al. Relationship of serum albumin concentration to death rate in nursing home men. JPEN. 1987;11:360 –363. 10. Galanos AN, Pieper CF, Cornoni-Huntley JC, et al. Nutrition and function: is there a relationship between body mass index and the functional capabilities of community-dwelling elderly? J Am Geriatr Soc. 1994;42:368 –373. 11. Hubert HB, Block DA, Fries JF. Risk factors for physical disability in an aging cohort: The NHANES I epidemiologic follow-up study. J Rheumatol. 1993;20:480 – 488. 12. Miller DK, Morley JE, Rubenstein LZ, Pietruszka FM. Abnormal eating attitudes and body image in older undernourished individuals. J Am Geriatr Soc. 1991;39:462– 466. 13. Pinholt EM, Kroenke K, Hanley JF, et al. Functional assessment of the elderly: a comparison of standard instruments with clinical judgment. Arch Intern Med. 1987;147:484 – 488. 14. Master AM, Lasser RP, Beckman G. Table of average weight and height of Americans aged 65–94 years. JAMA. 1960;172:658 – 662. 15. Anon. Weight by height and age of adults 18 –74 years: United States, 1971–1974. DHEW Publication No. (PHS) 79-1656, Series II, No. 208. Hyattsville, Maryland: US Department of Health, Education, and Welfare; 1979:34. 16. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state.’’ A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189 –198. 17. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res. 1982–3;17:37– 49. 18. Garner DM, Olmsted MP, Polivy J. Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. Int J Eating Disorders. 1983;2(2):15–34. 19. American Psychiatric Association, Committee on Nomenclature and Statistics. Diagnostic and Statistical Manual of Mental Disorders, Revised. 3rd ed. Washington, DC: American Psychiatric Association; 1987. 20. Sato T. Confidence limits for the common odds ratio based on the asymptotic distribution of the Mantel-Haenszel estimator. Biometrics. 1990;46:71– 80. 21. Norusis MJ. SPSS/PC1 Advanced Statistics 4.0. Chicago, Illinois: SPSS Inc.; 1990. 22. Rabinovitz M, Pitlik SD, Leifer M, et al. Unintentional weight loss: a retrospective analysis of 154 cases. Arch Intern Med. 1986;146: 186 –187. 23. Marton KI, Sox HC, Krupp JR. Involuntary weight loss: diagnostic and prognostic significance. Ann Intern Med. 1981;95:568 –574. 24. Thompson MP, Morris LK. Unexplained weight loss in the ambulatory elderly. J Am Geriatr Soc. 1991;39:497–501. 25. Blazer D, Bachar JR, Hughes DC. Major depression with
26. 27. 28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38. 39.
40.
41.
melancholia: a comparison of middle-aged and elderly adults. J Am Geriatr Soc. 1987;35:927–932. Blazer D, Williams CD. Epidemiology of dysphoria and depression in an elderly population. Am J Psychiatry. 1980;137:439 – 444. Blazer D. The diagnosis of depression in the elderly. J Am Geriatr Soc. 1980;28:52–58. Weintrab JA, Burt BA. Oral health status in the United States: Tooth loss and edentulism. Dent Educ. 1985;49:368 –376. Wallace JI, Schwartz RS, LaCroix AZ, et al. Involuntary weight loss in older outpatients: incidence and clinical significance. J Am Geriatr Soc. 1995;43:329 –337. Krumholz HM, Seeman TE, Merrill SS. Lack of association between cholesterol and coronary heart disease mortality and morbidity and all cause mortality in persons older than 70 years. JAMA. 1994;272: 1335–1340. Schwartz JB. Clinical pharmacology. In: Hazzard WR, Bierman EL, Blass JP, et al., eds. Principles of Geriatric Medicine and Gerontology. 3rd ed. New York: McGraw-Hill; 1994:259 –275. Gross J, Stitzer ML. Nicotine replacement: Ten weeks effect on tobacco withdrawal symptoms. Psychopharmacology. 1989;98:334 – 341. Millie A, Berriau T, Paget D, et al. Can weight gain during weaning from smoking be limited using nicotine gum? Revue de Pneumologie Clinique. 1989;45:243–249. Heymsfield SB, Tighe A, Wang Z. Nutritional assessment by anthropometric and biochemical methods. In: Shils ME, Olson JA, Shike M, eds. Modern Nutrition in Health and Disease. 8th ed. Philadelphia: Lea & Ferbiger; 1994:812– 841. Pomposelli JJ, Flores EA, Bistrian BR. Role of biochemical mediators in clinical nutrition and surgical metabolism. JPEN. 1988;12: 212–218. Boosalis MG, Ott L, Levine AS, et al. Relationship of visceral proteins to nutritional status in chronic and acute stress. Crit Care Med. 1989;17:741–747. Perlmutter DH, Dinarello CA, Punsal PI, et al. Cachectin/tumor necrosis factor regulates hepatic phase gene expression. J Clin Invest. 1986;78:1349 –1354. Fleck A. Clinical and nutritional aspects of changes in acute phase proteins during inflammation. Proc Nutr Soc. 1989;48:347–354. Lipschitz DA. Nutrition and aging. In: Evans GJ, Williams FT, eds. Oxford Textbook of Geriatric Medicine. New York: Oxford University Press; 1992. Zaloga GP. Timing and route of nutritional support. In: Zaloga GP, ed. Nutrition in Critical Care. St. Louis: Mosby Year Book; 1994: 297–330. Andres R. Mortality and obesity: The rationale for age-specific height-weight tables. In: Andres R, Bierman EL, Hazzard WR, eds. Principles of Geriatric Medicine. New York: McGraw-Hill; 1984:31– 38.
January 1998
THE AMERICAN JOURNAL OF MEDICINEt
Volume 104 63