Nutrition is a key component of nursing home resident care. This article highlights the importance of anorexia and weight loss in the nursing home and discusses the problems associated with diagnosing protein-energy malnutrition. Weight loss and albumin are the key indications of malnutrition in the nursing home environment, and the MiniNutritional Assessment appears to be the most appropriate screening instrument for this condition.
Weight loss is a key indicator of poor performance in the nursing home. The majority of persons with weight loss in a nursing home either have protein-energy malnutrition or dehydration. The appropriate recognition and management of protein-energy malnutrition in nursing homes remains one of the major challenges for all health professionals. The diagnosis of vitamin B12 deficiency can explain alterations in mental status and hematocrit in nursing home residents. In some residents, calcium and vitamin D replacement reduces the risk of hip fracture. Zinc deficiency can be associated with poor healing of pressure ulcers.
In this series of three articles, we will discuss the common nutritional problems that occur in nursing home residents and the approach to their diagnosis (Part I), the management of nutritional problems in the nursing home (Part II), and the requirements of state and national regulations on approaches to nutritional problems in nursing homes (Part III).
It is now clearly established that there is a decline in food intake throughout the life span.[1] This decline occurs despite the fact that weight increases in middle age, suggesting that much of this weight gain is due to the decline in resting metabolic rate and physical activity that occurs with aging. In the old-old (over age 85), there is a tendency to lose weight and adipose tissue mass.[2] These physiologic changes mean that older persons are particularly at risk for developing severe anorexia and weight loss when they contract diseases.
Changes in the hedonic qualities of food occur universally with aging. These changes are due particularly to declines in olfaction[3] and taste[4] that occur with aging. Whereas the ability to smell declines in all individuals, the changes in taste are more variable. Individuals who have smoked are more likely to experience declines in taste. The major change in taste is the increase in the threshold at which one can recognize a taste. The primary factors involved in altering taste with aging are the effects of drugs and diseases on taste rather than the physiologic changes of aging. Schiffman et al[5] demonstrated that nursing home residents preferred foods that had flavor enhancers added, which produced a tendency for them to ingest greater quantities of food.
These changes in taste and smell are extremely important in nursing home residents. Residents commonly complain about the quality of food in nursing homes. Much of this dissatisfaction is due to the physiologic alterations in taste and smell that make food appear less "tasty" as humans age. Alterations in the ability to appreciate the taste of food (most of which are due to decreased olfaction) mean that in the nursing home, food presentation and food choice play a more important role than the actual taste of the food. In Finnish nursing homes, residents are involved in the preparation of their own food. Such an approach is further likely to decrease the complaints about food quality (personal communication, 1997).
Many older persons are unable to eat the same quantity of food at a single meal as they ate when they were younger. This early satiation appears to be secondary to a diminished ability of the fundus of the stomach to display appropriate adaptive relaxation at the presence of food.[6] This results in food passing more rapidly from the fundus to the antrum of the stomach. Food in the antrum causes increased antral stretch, which appears to be the major signal for fullness.[7]
With aging, there is an increase in the levels of cholecystokinin (CCK), a gastrointestinal hormone involved in producing physiologic satiation.[8] This increase in CCK levels is more marked in malnourished older individuals. In addition, animal studies have suggested that CCK may be more effective at producing satiation in older compared to younger rodents.[6]
Previous human studies have suggested that older persons are less likely than young individuals to be satiated when food is delivered directly into the duodenum.[9] This finding can be important in the management of malnutrition because it suggests that the liquid caloric supplements that pass rapidly into the duodenum may be better for caloric supplementation than caloric-fortified solid foods in this population. Preliminary data suggest that having healthy older persons ingest a liquid supplement 60 minutes before a meal does not alter the number of calories eaten at the subsequent meal (Wilson MM and Morley JE, unpublished data, 1997). In addition, liquid supplements where the calories are supplied by glucose rather than fat are less likely to interfere with subsequent satiation because fat, but not glucose, slows gastric emptying.[10]
Leptin is a hormone produced by fat cells. It decreases food intake and increases metabolism. Leptin levels decline with age in women but not in men. The failure of leptin to decline with age in men is most probably due to the decline in testosterone levels with aging.[11] Whether or not the increased leptin levels in males play a role in the greater degree of physiologic anorexia seen in older males compared to females has not yet been elucidated. In addition to leptin, circulating cytokines, such as tumor necrosis factor alpha (cachectin), also reduce food intake, produce muscle wasting, and inhibit albumin synthesis.[12]
Within the central nervous system, numerous neutrotransmitters regulate food intake. At present, no human studies have been undertaken to determine whether alterations in these neurotransmitters caused by aging play a role in anorexia associated with aging. Animal studies have suggested that alterations in the endogenous opioid feeding drive may result in a decline in fat intake with aging.[13] A single human study found that older persons lose their endogenous opioid drive to drink and that this loss may play a role in the hypodipsia of aging.[14]
Overall, the accumulated data suggest that aging is associated with declines in the drive to eat and drink. Numerous factors appear to be involved in producing these physiologic age-related changes. In part, they occur to offset the decrease in resting metabolic rate and physical activity that occur as people age. Whatever the physiologic mechanisms responsible for these changes, they place older people at major risk for developing malnutrition and dehydration when they are in the nursing home.
The prevalence of protein-energy malnutrition (PEM) varies with the population observed and the definition of malnutrition. In the United States, health care professionals estimate that 40% of nursing home patients and 50% of hospitalized patients over the age of 65 are malnourished. Forty-four percent of home health patients are estimated to be malnourished.[15] These subjective estimates are close to prevalence results reported in clinical trials. Among patients newly admitted to a long-term care setting, a point prevalence of 54% malnutrition was observed.[16] In a Swedish study, 29% of newly admitted patients at a long-term care geriatric hospital were malnourished on admission.[17] The range for PEM in nursing home residents varies from 23% to 85%.[18,19] By comparison, the prevalence of PEM ranges from 32% to 50% in acutely hospitalized patients.[20,21] Other reports confirm that malnutrition is a major problem among residents in long-term care facilities.[22-24] The high prevalence of malnutrition in nursing homes may in part reflect transfer of malnourished patients from acute care hospitals to long-term care facilities following an acute illness.
The prevalence varies also with the criteria used to define malnutrition. The diagnosis of PEM in elderly populations is difficult. Anthropometric and biochemical measurements are usually performed to define type and severity of malnutrition, but there is no "gold standard" for diagnosis. Body weight, weight/height (body mass index), triceps skinfold thickness, arm circumference, arm muscle area, and arm fat area are the most commonly used anthropometric variables.[16] A broad panel of biochemical variables has been advocated to provide useful nutritional information. Serum albumin concentration is the single most commonly recommended parameter,[25] although lymphocyte count and concentrations of hemoglobin, prealbumin, transferrin, and retinol binding protein are also recommended. No single biologic parameter is satisfactory as a predictor of residents at risk for PEM.[26] The discriminant cutoffs for each variable continue to be disputed.[27-29]
Little is known about whether PEM persists or improves after admission to a long-term care facility. Studies in an academic nursing home have shown that 60% of residents experienced a net weight loss following admission.[18] Dietary interventions and nutritional supplements may improve malnutrition in long-term care settings. Weight gain occurred in 50% of malnourished patients, compared to weight gain in 58% of nonmalnourished patients (odds ratio, 0.70; 95% confidence; limits, 0.14, 3.46). Improvement in PEM occurred in 63% of the initially malnourished residents. However, 37% of residents remained malnourished.[16]
The number of malnourished patients in hospital settings may be decreasing over time as nutritional awareness increases. Using the same assessment scale at one institution, 38% of hospitalized patients were found to be at risk for malnutrition in 1988, compared to 48% of patients in 1976.[30] Similar data for patients in long-term care are not available.
As alluded to in the previous section, making the diagnosis of malnutrition in an older person is often extremely difficult. All the so-called "gold standards" have ultimately had an element of "fool's gold" mixed in. Thus, the eye of the astute, nutritionally aware physician remains perhaps the best means of recognizing impaired nutritional assessment.
Jeejeebhoy et al [31-33] have attempted to quantify the factors that a nutritionist uses to recognize malnutrition. This attempt has led to the development of the subjective global assessment (SGA). Persons with severe nutritional deficits (grade C) are those with changes in dietary intake and body mass (greater than 10% weight change over the last 6 months) and poor functional status. Grade B is scored when there is evidence of food restriction and functional changes but minimal weight change. Grade A is minimal or no changes in food intake, improving body weight, and minimal change in function. This method has a reasonable interobserver agreement rate of 81% to 91%. Grade C is associated with a 7-fold increase in the likelihood of complications in patients undergoing gastrointestinal surgery. The SGA appeared to be better than any single objective nutritional parameter in assessing the likelihood that a person will develop nutrition-related complications. However, the SGA has not been validated in the nursing home environment.
Food intake represents a potentially important tool in monitoring persons at risk for malnutrition in the nursing home. Unfortunately, recent studies have suggested that the recording of the amount of food eaten in nursing homes is highly inaccurate.[34] Although counting calories by weighing all food before and after the meal would be more accurate, it is rarely feasible in the long-term care setting. Nurse's aides can be trained to be more accurate in estimating the amount of food ingested, but doing so requires a substantial time investment.
Weight loss remains one of the best indications of nutritional risk in nursing homes. All nursing homes should have a flow chart giving monthly weights in each resident's record. Because scales in nursing homes often malfunction, it is helpful if the persons responsible for recording residents' weights actually weigh themselves each day on each scale. Residents need to be weighed at the same time of the day each month, dressed in a minimal amount of clothing and without shoes. Obviously, both congestive heart failure and dehydration can alter weights. Height needs to be obtained on admission and reobtained yearly to allow identification of height loss due to osteoporosis.
A variety of other anthropometric tools are available to measure nutritional status. Overall, these tools have not been proven very useful in the nursing home. Of the skinfold thicknesses, the triceps measurement is most useful in females, and the subscapular measurement is more accurate in males. Mid-arm circumference or mid-arm muscle circumference can be a useful measurement in residents with major alterations in water metabolism. In these residents, mid-arm circumference needs to be recorded alongside the resident's weight. The former will be a more accurate indicator of protein loss from muscle.
Whereas measurements of circulating proteins can be useful to judge the degree of protein malnutrition, multiple nonnutritional factors can interfere with their levels. Albumin has a long half-life (21 days), making it less useful as a measurement of acute nutritional status. Two factors often associated with illness can, however, produce acute changes in albumin levels. Recumbency is associated with an increased intravascular volume, and the dilutional effect can lower serum albumin levels by as much as 0.5 mg/dl.[35] Cytokine release not only inhibits albumin synthesis but also causes extravasation of albumin from the intravascular to the extravascular space.[36] These 2 factors explain the rapid decline in albumin levels that are often experienced when older patients are admitted to the hospital. Nevertheless, albumin levels of 3.2 g/dl or less remain an excellent predictor of morbidity and mortality in older persons.[37]
Proteins with a shorter half-life, such as prealbumin (2 days) and retinol binding protein (RBP; 2 hours), are occasionally useful to determine response to nutritional supplementation. Prealbumin levels are subject to all the vagaries experienced by albumin. In addition, levels are increased with decreased creatinine clearance because the kidney is the major metabolic site. RBP is a glycoprotein that has its levels altered by vitamin A, zinc, or carbohydrate in the diet and by renal disease.
Acute phase reactants, such as fibronectin, have also been used to identify malnourished patients, but they are clearly more related to disease than to nutritional status. Recently, soluble interleukin-2 receptors have been shown to be a good correlate of outcomes in malnourished hospitalized patients.[38] They appear to be a good marker of catabolic states.
Low total cholesterol levels are also a measure of nutritional status. However, like albumin, the cholesterol level is altered by cytokines. Levels of cholesterol below 156 mg/dl are highly predictive of poor outcomes in nursing homes.[39]
Leptin levels are an excellent marker of total body fat.[40] As such, they have major potential as a nutritional marker. Anemia is often due to protein-energy malnutrition, and successful nutritional rehabilitation can reverse much of the anemia of chronic disease. Lymphocytopenia and, in particular, low CD(4) cell levels are good indicators of malnutrition.[41] Anergy to delayed cutaneous hypersensitivity testing for common antigens, such as Candida, is seen in malnourished persons. It can be reversed with nutritional support[42] and is related to increased septicemia and mortality.[43]
Malnutrition results in atrophy of muscle fibers and 2-band degeneration, which presents physiologically as the inability to maintain tetonic contractions, delayed relaxation rate, and reduced force generation. Clinically, it can be examined by measuring grip strength with a dynamometer. A decline in respiratory muscle function can be suspected in persons who have a weak cough.
In the nursing home, measurements of body composition can be obtained utilizing bioelectrical impedance with appropriate formulae.[44] However, other factors, such as dehydration and altered height secondary to osteoporosis, make the reliability of this technique highly suspect in the nursing home. The use of other techniques for measuring body composition are either not suitable for the majority of nursing home residents or are predominantly used for research purposes (eg, stable isotopes or underwater weighing).
Two types of protein-energy malnutrition exist--namely, marasmus and kwashiorkor. Marasmus is characterized by weight loss, whereas kwashiorkor shows a specific rapid decline in serum albumin levels. Marasmus is predominately due to poor food intake, whereas kwashiorkor is usually precipitated by cytokine release associated with an acute stressor.
The Nutritional Screening Index was developed to identify persons at risk for malnutrition.[45] It has poor sensitivity and specificity.[46] It should not be used in the nursing home environment.
The MiniNutritional Assessment (MNA) is the best validated of the nutritional screening tools, [47,48] and it is appropriate for use in nursing homes. Its major advantage is that it does not use laboratory tests, and so it is highly cost-effective. In the authors' experience, it is an excellent tool for screening persons on admission to nursing homes.
SCALE was developed by the authors for identification of malnutrition in the outpatient setting.[49] SCALE has been cross-validated with the MNA (Miller DK and Morley JE, unpublished data, 1997). This assessment tool is appropriate for use in the nursing home, although the shopping/food preparation criteria are dropped. SCALE appears to be a useful method for detecting early malnutrition risk in nursing homes.
It should be clear that the assessment of nutritional deficiency is extremely difficult. No single measurement is ideal. Clinical judgment remains the "gold standard." The second part of this series will discuss the causes of nutritional deficiencies and their appropriate management.
From the Geriatric Research, Education and Clinical Center, St. Louis VAMC, St. Louis, Missouri, and Division of Geriatric Medicine, St. Louis University Medical School, Address for correspondence: John Morley, MB, BCh, Dammert Professor of Gerontology, St. Louis University Health Sciences Center, 1402 S. Grand, Room M238, St. Louis, MO 63104.
