- Email: [email protected]
Energy-dense meals improve energy intake in elderly residents in a nursing home
Clinical Nutrition (2003) 22(2): 125–131 r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1054/clnu.2002.0610
ORIGINAL ARTICLE
Energy-dense meals improve energy intake in elderly residents in a nursing home A. DLUND OLIN,*,z,** I. ARMYR,w,zz M. SOOP,*,8,z,**,zz S. JERSTR M,z,ww I. CLASSON,w,ww T. CEDERHOLM,z,},z,** G. LJUNGGREN,},z O. LJUNGQVIST*,8,z,**,zz *Department of Surgery, Center for Surgical Sciences, wDietitian Unit, zDepartment of Geriatrics, }Department of Clinical Neuroscience and OccupationalTherapy, and Elderly Care Research, 8Centre of Gastrointestinal Disease, zKarolinska Institutet, **Huddinge University Hospital, wwKarolinska Hospital, zzErsta Hospital, Stockholm, Sweden (Correspondence to: A O, Nursing Care Research and Development Unit, M 98, Huddinge University Hospital, S-14186 Stockholm, Sweden)
AbstractFBackground: Several studies have shown that malnutrition is common among nursing home residents. Aim: We hypothesized that addition of natural energy-dense ingredients to a standard diet would improve voluntary energy intake and ability to perform activities of daily living (ADL) and decrease infections in elderly residents under nursing home care. Methods:Thirty-five residents in a municipality nursing home (median age 83 years) were served either a standard diet (1600 kcal/day) (control group, n=18) or the same meals fortified with natural energy-dense ingredients (2100 kcal/day) (experimental group, n=17). Energy intake and ADL function were measured before and after the intervention. All episodes of infection were registered. Non-parametric statistics were used. Results: Energy intake increased in the experimental group from 23.5 (21.3--28.5) kcal/kg/body weight before, to 31.9 (29.7--33.7) kcal/kg/body weight during the intervention (Po0.001). There was no change in energy intake in the control group. ADL function in the experimental group was unchanged, while it decreased significantly in the control group (Po0.001). The number of infections tended to be lower in the experimental group than in the control group (5 vs 13). The cost for the energy-dense ingredients was approx. 0.11EUR per resident and day. Conclusions: Addition of natural energy-dense ingredients to regular meals is an inexpensive way to improve voluntary energy intake in elderly nursing home residents, a treatment that was accompanied by maintained ADL function. r 2003 Elsevier Science Ltd. All rights reserved.
experience that not all patients want oral nutritional supplements. Recent findings show that a substantial increase in EI can be achieved by simply increasing the energy density of regular hospital food. When elderly patients are served energy-dense meals they continue to consume the same volume of food and it is a simple and cheap method to increase EI (13). We hypothesized that addition of natural energydense ingredients to a standard diet would improve voluntary EI and ability to perform activities of daily living (ADL) and decrease infections in elderly residents under nursing home care over a prolonged period of time.
Key words: energy-dense meals; elderly residents; nursing home; energy intake; inexpensive
Introduction The elderly population is growing rapidly in the developed world. An increasing number of people are in need of nursing and hospital care. Several studies have shown that low energy intake (EI) and malnutrition is common among elderly hospitalized patients and nursing home residents (1–6). Malnutrition is associated with increased morbidity and mortality (7–9), prolonged hospital stay and increased costs of care (8–10). It has been suggested that elderly patients are especially vulnerable to nutritional deficiencies due to reduced nutrient reserves and an inability to respond adequately in periods of stress (11). The causes of malnutrition are multi-factorial and include anorexia, metabolic disorders, malabsorption, difficulties in swallowing, psychosocial factors and hospitalization per se (5, 12). Oral supplements have been shown to be able to improve functional condition and clinical outcome among elderly patients (7, 8). However, it is a common
Material and methods Subjects Thirty-five residents in a municipality nursing home (median age 83 years) were studied. Residents in one ward of the nursing home (n=18) were served regular hospital diet (1600 kcal/day) during a 15-week period (control group), while residents in another ward of the nursing home (n=17) were served the same meals 125
126
ENERGY-DENSE MEALS
fortified with natural energy-dense ingredients such as butter and cream (2100 kcal/day) (experimental group). The participating wards were similar in manning level and recruited the same kind of residents. The most common medical diagnoses were dementia and cerebrovascular lesions (Table 1). A total of 40 residents entered the study. Five subjects, two from the control group and three from the experimental group, died during the intervention period, and these five subjects were excluded from the study.
Table 2 Examples of meals in the two study groups during the intervention period Control diet Lunch Beef in horseradish sauce Boiled potatoes, mixed vegetables Fruit syrup dessert with milk
Experimental diet additions/changes 30 g whipping cream (40% fat) instead of milk 13 g hydrolysed starch 100 g cream (12% fat) instead of milk
Dinner Oven-baked sausage Mashed potatoes, boiled broccoli
20 g cheese (31% fat) 15 g margarine (80% fat)
Study design All residents were given the regular diet for a period of 2 weeks before the intervention, in order to establish a control period (pre-test period). Immediately following the pre-test period, residents in one ward of the nursing home (n=18) were served the regular diet prepared by the hospital kitchen with an energy content of 1600 kcal/ day or 6.7 MJ (control group) during 15 weeks (intervention period). Residents in another ward of the nursing home (n=17) were served energy-dense meals (experimental group), with a total of 2100 kcal/day, 8.8 MJ. Patients were studied again during a period of 1 week. This follow-up was performed 12 weeks after the intervention ended.
Intervention and meal composition Residents, relatives and staff were informed that evaluation of changes in food composition were about to take place, but no further details of changes in diet composition were given. The limited information was intended not to cause patient and staff bias. Meals were prepared in the kitchen of the nursing home and delivered to the wards in canteens. Hot meals (lunch and dinner) delivered to the experimental group were energy fortified with natural ingredients such as cream, oil, butter, etc., while desserts were energy fortified with cream, sour cream or hydrolysed starch. A menu for a 6 weeks period was elaborated (for examples, see Table 2). The nutritional composition for the meals was calculated (Table 3). Table 1 Diagnoses in the two study groups Main diagnosis Dementia Cerebro-vascular disorder Parkinson’s disease Multiple sclerosis Post-polio Rheumatoid arthritis Associated medical conditions Chronic heart failure Clinical oedema
Control group
Experimental group
10 4 1 1 1 1
9 4 3 1 – –
8 6
2 2
Table 3 Energy and macro-nutrient content of served main meals in control and experimental groups, and energy per cent of each macronutrientw
Energy, kcal Protein, g (energy %) Fat, g (energy %) Carbohydrate, g (energy %)
Control diet
Experimental diet
507 26(21%) 22(38%) 53(41%)
769 26(14%) 48(55%) 60(31%)
w The cost increase for the experimental diet was approx. 0.11 EUR per resident and day.
Energy and nutrient intake Before starting the study, the nursing staff were trained to perform the recording of food and fluid intake by the eye-measurement method (13). All oral intake was recorded Monday through Friday by the use of a 24 h dietary record chart. The nursing staff recorded the type and amount of food served, and the amount consumed, at breakfast, lunch and dinner. All additional intake, such as snacks and beverages, was also recorded by the staff. The nursing staff were given information each day at lunch as well as at dinner about the planned full-size portion, because of the fact that the hot meals were delivered in canteens and the staff were responsible for dimensioning the portions. Consumption of every single component of a meal, such as potatoes, meat, etc. was registered. Two research nurses and a dietitian visited the wards every week to double check portion sizes and food-intake registrations. Daily energy and nutrient intake was calculated using a computer program, Kostplan Dietist (AIVO AB, Spa˚nga, Sweden). The ratio between the estimated EI and the predicted basal metabolic rate (BMR) denoting the physical activity level (PAL), was calculated (14). The EI/BMR ratio was used to evaluate the reliability of the food and fluid records and was also related to the activity of daily living scores (see below). Subjects bedridden or bound to a wheelchair usually have a EI/BMR ratio 1.2. A sedentary life with low physical activity corresponds to ratios of 1.4–1.5 (15). The predicted BMR was calculated according to formulas recommended by the WHO (16).
CLINICAL NUTRITION
Energy expenditure was determined in eight subjects using indirect calorimetry (Deltatracr, Dansjo¨o¨, Stockholm, Sweden) (17). These eight subjects did not differ from the other 27 subjects in body mass index (BMI), body weight or total EI per day but they had less EI per kg body weight and were less dependent on assistance with primary ADL functions (data not shown). Anthropometric measurements Body weight was measured once during the pre-test, intervention and follow-up periods, respectively. Body height was measured in bed in the supine stretched position. Functional condition The functional condition was assessed by the resident assessment instrument (RAI) (18). The instrument incorporates an ADL scale. This scale includes four key ADL capabilities: mobility in bed, ability to transfer from bed to chair, capacity to use a toilet and the need of help to eat. These four capabilities are scored from 0 to 4, where a lower value indicates a better performance and a lower reliance on help from carers. With a simple algorithm, these scores are summarized to a total ADL score that runs from 4 to 18. The lower the score the better functional condition. Each subject was assessed by the registered nurse responsible for the care of the patient, together with staff who were well acquainted with the patient, and a research nurse. All episodes of infection during the intervention period were registered by staff.
127
Statistics and ethics Values are given as median and the 25th and 75th percentiles (P25–P75). Mann–Whitney U-test, Wilcoxon’s signed rank test, Spearman’s rank–order-correlation, simple regression analysis and Chi-squared exact test were used as appropriate. Informed consent was obtained from each resident and the next of kin. The protocol was approved by the Institutional Ethics Committee at the Karolinska Hospital.
Results The experimental group did not differ from the control group in initial body weight, BMI, EI or ADL score but they were slightly younger (Table 4). Four subjects in each group had a BMI below 20. The median BMI in all subjects was 23.1 (20.0–26.3): in men, 24.2 (22.7–26.3); and in women, 21.9 (19.7–23.6). Subjects in both groups were highly dependent on assistance with primary ADL functions. The median daily EI during the pre-test period for all subjects was 1353 (1159–1578) kcal/day or 23.5 (21.3–26.1) kcal/kg body weight. Twenty subjects (57%), 10 from each group, consumed less than 25 kcal/kg during the pre-test period. Serving of the experimental diet resulted in a significant increase in voluntary EI per day by 504 kcal (36% increase) (Table 5). EI remained stable in subjects served the control diet (Table 5). Fat and protein intake
Table 4 Demographic data of subjects before the start of the study Control group (n=18) Gender (females/males) Age (years) BMI (kg/m2) Body weight (kg) ADL score
83.5 23.5 63.1 15.5
Experimental group (n=17)
9/9 (79.0–89.0) (20.0–26.8) (50.9–72.3) (10.0–17.0)
80.0 22.7 56.0 16.0
P
9/8 (71.0–85.0) (20.1–24.4) (48.4–65.1) (14.0–18.0)
ns o0.05 ns ns ns
Median (P25P75). Control group vs experimental group, Mann–Whitney U-test.
Table 5 Daily intake of energy and nutrients during pre-test and intervention periods in control and experimental groups, and energy per cent of each macro-nutrient Control group (n=18)
EI (kcal/day) Energy per kg body weight (kcal/kg/day) Protein (g) Energy % Fat (g) Energy % Carbohydrates (g) Energy %
Experimental group (n=17)
Pre-test period
Intervention period
Pre-test period
Intervention period
1431 (1142–1564) 23.5 (21.9–25.4) 53.7 (42.7–58.7) 15 53.2 (41.1–61.4) 34 180 (128–204) 51
1437 (1252–1617) 24.9 (21.1–26.6) 54.7 (47.0–59.9) 15 57.6 (51.3–63.7) 34 175 (153–209) 50
1336 (1261–1578) 23.5 (21.3–28.5) 48.3 (41.8–54.3) 14 48.4 (43.8–56.4) 32 178 (152–198) 53
1840 (1497–2012)**,**** 31.9 (29.7–33.7)***,**** 57.9 (46.2–61.2)**** 12 93.1 (75.3–104.0)***,**** 45 183 (149–215) 41
Median (P25P75). **Po0.01, ***Po0.001, control group vs experimental group, Mann–Whitney U-test. ****Po0.001, pre-test vs intervention period, Wilcoxon’s signed rank test.
128
ENERGY-DENSE MEALS
increased during the intervention period in the experimental group, while carbohydrate intake remained unchanged (Table 5). Regardless of the diet during the study, residents with a lower BMI had a higher EI in relation to their body weight. Thus pre-test BMI was negatively correlated with EI during the study, r=0.60 (Po0.001). Despite differences in EI, body weights were stable in both groups throughout the intervention period. There was no correlation between EI and change in body weight during this period for the entire study population. In each group, the residents were split into subgroups, one with those with BMIo24, the other with those above this value. There was no significant change in body weight between those residents with a BMI below 24 and those above. However, for residents with a BMI below 24, served the energy-dense diet, there was a positive trend in weight gain (+0.3 kg), while those with higher BMI tended to lose weight (–1.6 kg). Subjects served the standard diet had a trend towards reduction in body weight. For residents with a low BMI served the energydense diet, EI per kg body weight increased by 34% (Po0.01). The same degree of change was found in those with BMI424 given the same diet. For residents with a BMIo24 served the standard diet, EI per kg body weight increased by 10% (Po0.01) (Table 6). Median estimated BMR, n=35 was for the entire study group 1273 (1140–1376) kcal/day. For eight subjects, the indirect calorimetry was 1015 (925–1180) kcal/day and for these eight subjects, the estimated BMR was 1272 (1198–1390) kcal/day (Table 7). The difference between methods was significant (Po0.005). However, regression analysis of indirect calorimetry vs BMR showed a high level of correlation (R2=0.86, Po0.001).
There was no difference in pre-test-calculated EI/ BMR ratios between the two study groups (Table 7). The experimental group had a significant increase in calculated EI/BMR ratio during the intervention period (Po0.001) whereas the ratio remained the same and low in the control group. There was no difference in pre-test ADL score between the two study groups (Table 7). Nor were there any differences during the intervention period. However, the control group deteriorated in their ADL functions during the intervention period (Po0.001), whereas the experimental group remained stable (Table 7). The number of infection in the experimental group tended to be lower than in the control group. In the control group, eight patients had a total of 13 infections episodes while five patients in the experimental group had one infections episode each (Table 8a and 8b). In the follow-up period, it was found that six of the 35 residents had died, three from each group. Residents in the experimental group had returned to pre-test levels of EI (Fig. 1). Discussion The present study shows that addition of natural energydense ingredients to a standard diet significantly improves voluntary energy intake in elderly nursing home residents over a prolonged period of time. Sixteen of 17 of the subjects in the experimental group increased their EI during the study. The greatest effect was registered in residents with a low BMI. The increase in EI was accompanied by a maintenance of ADL function. In contrast, subjects served a standard diet had a decline in ADL status during the study period.
Table 6 Median differences between pre-test and intervention periods for subjects with BMIo24 or above this value in control and experimental groups
BMI Differences in energy per kg body weight (kcal/kg)
Control group (n=18)
Experimental group (n=17)
o24 n=10 3.0 (1.9–4.0)
o24 n=12 8.0 (4.6–10.3)*
424 n=8 0.7 (1.6–0.2)
424 n=5 7.6 (5.3–8.5)**
Median (P25P75). *Po0.05, **Po0.01, control group vs experimental group, Mann–Whitney U-test.
Table 7 Energy served, EI, BMR, estimated physical activity level (PAL) according to Nordic Nutrition Recommendations, estimated energy requirement (BMR estimated PAL), EI/BMR ratio and ADL-score during pre-test and intervention periods in control and experimental groups Control group (n=18)
Energy served (kcal) EI (kcal) Energy requirement BMR Estimated PAL EI/BMR ratio ADL score
Experimental group (n=17)
Pre-test period
Intervention period
Pre-test period
Intervention period
1600 1431 (1142–1564) 1569 (1476–1782) 1288 (1230–1361) 1.2 1.1 (0.9–1.2) 15.5 (10.0–17.0)
1600 1437 (1252–1617) 1558 (1460–1795) 1294 (1140–1376) 1.2 1.2 (1.1–1.2) 16.0 (15.0–18.0)****
1600 1336 (1261–1578) 1501 (1368–1690) 1225 (1140–1376) 1.2 1.1 (1.0–1.3) 16.0 (14.0–18.0)
2100 1840 (1497–2012)**,**** 1506 (1396–1662) 1230 (1164–1344) 1.2 1.5 (1.3–1.7) ***,**** 15.0 (14.0–18.0)
Median (P25P75). **Po0.01, ***Po0.001, control group vs experimental group, Mann–Whitney U-test. ****Po0.001, pre-test vs intervention period, Wilcoxon’s signed rank test.
CLINICAL NUTRITION
129
Table 8a Number of episodes of infections in control vs experimental groups Infection Pneumonia Urinary Skin Total
Control group (n=18)
Experimental group (n=17)
4 8 1 13
1 3 1 5
b Distribution of infections in subjects in the control vs experimental groups
Control group Experimental group
No infection
One infection
Two infections
Three infections
10 12
4 5
3 0
1 0
Comparisons with Chi-square exact test P=0.29. 60
50
Kcal/kg b.wt
40
30
20
10
0 Pre-test Experimental
Intervention
Follow up
Control
Fig. 1 Median EI related to body weight in the control group (n= 18) and the experimental group (n=17) during the pre-test, intervention and follow-up periods. (Median P25P25) and range.
The subjects in this study were elderly and suffered from a variety of chronic diseases which are often associated with malnutrition, and in high need of care. The high dependency on nursing care was confirmed by high ADL scores. The median BMI for all residents was 23.1 before the start of the project, similar to values reported in earlier studies in this category of patients (5, 6). A BMI between 24 and 29 has recently been suggested as optimal for elderly patients, especially those with a poor EI (19). Twenty-two of the 35 subjects (63%) had BMI values below this range. However, BMI may be unreliable as an indication of nutritional state in this population, as a result of diseases causing fluid retention. The subjects in our study were grouped according to wards rather than truly randomized, since it proved practically impossible to serve different diets within the same ward. Nevertheless, the study groups were well matched, with a slightly younger median age in the experimental group.
It was found that the experimental diet had the greatest effect on the residents with a low BMI. This is in line with findings reported by Gall et al. who offered patients an extra 966 kcal/day and 22.2 g protein in meal fortification and snacks. Intakes were registered over a 3-day period and the increase in EI among orthopaedic, medical and elderly care patients was 19.6–23% (20). The percentage increase in EI in the present study was even higher than that. The residents in the control group were served a diet according to Nordic Nutrition Recommendations (15). These recommendations are given for healthy elderly. However, there is an ongoing discussion about specific recommendations for elderly sick persons. In the present study, it was only possible to perform indirect calorimetry in eight subjects, since most subjects were incapable of going through the procedure properly. The findings from the eight subjects confirm that estimations of BMR are often higher than the measurements from indirect calorimetry (21). Nevertheless, these
130
ENERGY-DENSE MEALS
findings indicate that energy expenditure is low in elderly subjects under nursing care. The Nordic recommendations for dietary EI in the elderly are based on an estimated BMR and adjusted for activity. For the present population, the men had a median estimated BMR of 20.5 kcal/kg/day and with the suggested adjustment for activity an energy requirement of 24.6 kcal/kg/day. Similarly, the women had a BMR of 21.5 and 25.8 kcal/kg/day in requirements. The regular diet served in this study contained 26.0 kcal/kg. However, the determinations of indirect calorimetry suggest that needs may actually be lower in these frail elderly individuals. The energy-dense diet supplied well above the estimated needs of calories. One of the aims of the study was to investigate the long-term effects of energy-dense food. Surprisingly, the body weight of the patients in the experimental group did not increase, despite a substantially higher intake of fat during 15 weeks. In our previous study of geriatric patients, 6 weeks of high energy food intake resulted in a significant 3.4% increase in body weight (13). This suggests that only small weight changes are to be expected even during a prolonged treatment period. There may be a variety of reasons why the subjects in the experimental group in this study did not gain weight. Cardiac failure, renal insufficiency and medication may influence body water content and subsequently body weight. We were unable to control these factors in this study. Six patients in the control group had oedema on a clinical basis but only two patients in the experimental group. This may, to some extent, explain the stable weight in both groups. A more precise estimation of body composition could have given information about any potential changes in these parameters, but was not performed presently for practical reasons. Another possible reason for weight stability despite increased EI could be an increased energy expenditure due to increased physical activity. There are several instruments for measuring ADL functions, all with specific limitations. In this study, we used the RAI for measuring functional condition. The RAI has been positively tested for validity and reliability (22). In the future, modifications in scales measuring ADL function may allow greater sensitivity in detecting changes caused by nutritional intake. In this study, the subjects in the control group had a small but significant decrease in ADL function score over time. For this population, this change is not unexpected. In contrast, the experimental group was able to maintain the ADL score, suggesting that non-energy-dense food may contribute to successive decline in function in an elderly frail population. This reduction in functional capacity may result in increased need for nursing care. The subjects in this study had pre-test very low EI/BMR ratios which may correspond to the low physical activity of the residents. Only 8 of 35 subjects (23%) had EI/BMR ratios above 1.2 during the pre-test period. After the intervention, 13 of 17 subjects (76%)
had EI/BMR ratios above 1.2. The EI/BMR ratio combined with the minimal weight change suggests that the physical activity in the experimental group increased during the intervention period. Nevertheless, the registered ADL score in the experimental group was not significantly improved. The number of infections tended to be lower in the experimental group than in the control group, but there was no significant difference between the two groups. However, earlier studies have described that the immune defence system is weakened in malnourished patients and that the risk for infections is increased (23–25). Energy-dense food needs to contain a high proportion of fat if it is to be fortified with natural ingredients. The risk for the elderly population in this study to develop secondary cardiovascular problems due to increased fat consumption seems less likely than complications from malnutrition, but is a question that needs to be addressed. There are alternative ways to improve EI. An energydense diet with natural ingredients has several advantages. It is readily accepted by elderly subjects who may not accept other nutritional supplements. Energy-dense meals can easily be prepared in any hospital or nursing home kitchen. It is also by far the cheapest way to increase the energy content of the food for these patients. By using fat as a means of fortification, the taste of food improves and this was commented upon by many subjects. In conclusion, this study shows that the daily EI in elderly nursing home residents can be improved over a prolonged period of time by the simple method of adding natural energy-dense ingredients to regular meals. Serving energy-dense food is an inexpensive way to improve EI in this population. Improving EI using this simple treatment seems to help maintain ADL function in elderly nursing home residents. Acknowledgements The nursing staff in the nursing home Hallen in Solna is gratefully acknowledged. We thank Elisabeth Ljungqvist, RN for her assistance in collecting the data. The study was supported by grants from the Swedish National Board of Health and Welfare and the National Institute of Public Health, Sweden.
References 1. McWhirter J P, Pennington C R. Incidence and recognition of malnutrition in hospital. Br Med J 1994; 308: 945–948 2. Cederholm T, Ja¨gre´n C, Hellstro¨m K. Nutritional status and performance capacity in internal medical patients. Clin Nutr 1993; 12: 8–14 3. Volkert D, Kruse W, Oster P, Schlierf G. Malnutrition in geriatric patients: diagnostic and prognostic significance of nutritional parameters. Ann Nutr Metab 1992; 36: 97–112 4. Thomas D R, Verdery R B, Gardner L, Kant A, Lindsay J. A prospective study of outcome from protein–energy malnutrition in nursing home residents. J Parenter Enteral Nutr 1991; 4: 400–404 5. Blaum C S, Fries B E, Fiatarone M A. Factors associated with low body mass index and weight loss in nursing home residents. J Gerontol Med Sc 1995; 3: M 162–168
CLINICAL NUTRITION
6. Saletti A, Yifter Lindgren E, Johansson L, Cederholm T. Nutritional status according to Mini Nutritional Assessment in an institutionalized elderly population in Sweden. Gerontology 2000; 46: 139–145 7. Larsson J, Unosson M, Ek A C, Nilsson L, Thorslund L. Effect of dietary supplementation on nutritional status and clinical outcome in 501 geriatric patientsFa randomised study. Clin Nutr 1990; 9: 179–184 8. Delmi M, Rapin C-H, Bengoa J M, Delmas P D, Vasey H, Bonjour J P. Dietary supplementation in elderly patients with fractured neck of the femur. Lancet 1990; 335: 1013–1016 9. Gallagher-Allred C R, Voss A C, Finn S C, McCamish M A. Malnutrition and clinical outcomes: the case for medical nutrition. J Am Diet Assoc 1996; 96: 361–366 10. Chima C S, Barco K, Dewitt M L A, Maeda M, Teran J C, Mullen K D. Relationship of nutritional status to length of stay, hospital costs, and discharge status of patients hospitalized in the medicine service. J Am Diet Assoc 1997; 97: 975–978 11. Rapin C H. Nutrition support and the elderly. In: Payne-James J, Grimle G, Silk D (eds). Artificial Nutrition Support in Clinical Practice. Filey, North Yorkshire: Edward Arnold, 1995; 535 12. Green C J. Existence, causes and consequences of disease-related malnutrition in the hospital and the community, and clinical and financial benefits of nutritional interventions. Clin Nutr 1999; 18 (suppl 2): 3–28 13. O¨dlund Olin A, O¨sterberg P, Ha˚dell K, Armyr I, Jerstro¨m S, Ljungqvist O. Energy-enriched hospital food to improve energy intake in elderly patients. J Parenter Enteral Nutr 1996; 2: 93–97 14. Hambreaus L. Dietary assessments: how to validate primary data before conclusions can be drawn. Scand J Nutr 1998; 42: 66–68 Submission date: 14 December 2001 Accepted: 30 September 2002
131
15. Nordic Nutrition Recommendations 1996: Nordic 1996: 28, Copenhagen: Nordic Council of Ministers, 1996 16. FAO/WHO/UNU. Report of a joint Expert Consultation: Energy and Nutrient requirements. Technical Report Series 724, WHO, Geneva, 1985 17. Frayn K N. Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl Physiol 1983; 55: 628–634 18. Morris J N, Hawes C, Fries B E et al. Designing the national resident assessment instrument for nursing homes. Gerontologist 1990; 30: 293–302 19. Beck A M, Ovesen L. At which body mass index and degree of weight loss should hospitalized elderly patients be considered at nutritional risk. Clin Nutr 1998; 17: 195–198 20. Gall M J, Grimble G K, Reeve N J, Thomas S J. Effect of providing fortified meals and between-meal snacks on energy and protein intake of hospital patients. Clin Nutr 1998; 6: 259–264 21. Beck AM, Ovesen L. Body mass index, weight loss and energy intake of old Danish nursing home residents and home-care clients. Scand J Caring Sci 2002; 16: 86–90 22. Sgadari A, Morris J N, Brant B E et al. Efforts to establish the reliability of the RAI. Age Ageing 1997; 26 (suppl 2): 27–30 23. Chandra R K. Nutrition and the immune system. Proc Nutr Soc 1993; 8: 243–248 24. Chandra R K. Nutritional regulation of immunity and risk of infection in old age. Immunology 1989; 67: 141–147 25. Eriksson K M, Cederholm T, Palmblad J E. Nutrition and acute leukaemia in adults: relation between nutritional status and infectious complications. Cancer 1998; 82: 1071–1077
ORIGINAL ARTICLE
Energy-dense meals improve energy intake in elderly residents in a nursing home A. DLUND OLIN,*,z,** I. ARMYR,w,zz M. SOOP,*,8,z,**,zz S. JERSTR M,z,ww I. CLASSON,w,ww T. CEDERHOLM,z,},z,** G. LJUNGGREN,},z O. LJUNGQVIST*,8,z,**,zz *Department of Surgery, Center for Surgical Sciences, wDietitian Unit, zDepartment of Geriatrics, }Department of Clinical Neuroscience and OccupationalTherapy, and Elderly Care Research, 8Centre of Gastrointestinal Disease, zKarolinska Institutet, **Huddinge University Hospital, wwKarolinska Hospital, zzErsta Hospital, Stockholm, Sweden (Correspondence to: A O, Nursing Care Research and Development Unit, M 98, Huddinge University Hospital, S-14186 Stockholm, Sweden)
AbstractFBackground: Several studies have shown that malnutrition is common among nursing home residents. Aim: We hypothesized that addition of natural energy-dense ingredients to a standard diet would improve voluntary energy intake and ability to perform activities of daily living (ADL) and decrease infections in elderly residents under nursing home care. Methods:Thirty-five residents in a municipality nursing home (median age 83 years) were served either a standard diet (1600 kcal/day) (control group, n=18) or the same meals fortified with natural energy-dense ingredients (2100 kcal/day) (experimental group, n=17). Energy intake and ADL function were measured before and after the intervention. All episodes of infection were registered. Non-parametric statistics were used. Results: Energy intake increased in the experimental group from 23.5 (21.3--28.5) kcal/kg/body weight before, to 31.9 (29.7--33.7) kcal/kg/body weight during the intervention (Po0.001). There was no change in energy intake in the control group. ADL function in the experimental group was unchanged, while it decreased significantly in the control group (Po0.001). The number of infections tended to be lower in the experimental group than in the control group (5 vs 13). The cost for the energy-dense ingredients was approx. 0.11EUR per resident and day. Conclusions: Addition of natural energy-dense ingredients to regular meals is an inexpensive way to improve voluntary energy intake in elderly nursing home residents, a treatment that was accompanied by maintained ADL function. r 2003 Elsevier Science Ltd. All rights reserved.
experience that not all patients want oral nutritional supplements. Recent findings show that a substantial increase in EI can be achieved by simply increasing the energy density of regular hospital food. When elderly patients are served energy-dense meals they continue to consume the same volume of food and it is a simple and cheap method to increase EI (13). We hypothesized that addition of natural energydense ingredients to a standard diet would improve voluntary EI and ability to perform activities of daily living (ADL) and decrease infections in elderly residents under nursing home care over a prolonged period of time.
Key words: energy-dense meals; elderly residents; nursing home; energy intake; inexpensive
Introduction The elderly population is growing rapidly in the developed world. An increasing number of people are in need of nursing and hospital care. Several studies have shown that low energy intake (EI) and malnutrition is common among elderly hospitalized patients and nursing home residents (1–6). Malnutrition is associated with increased morbidity and mortality (7–9), prolonged hospital stay and increased costs of care (8–10). It has been suggested that elderly patients are especially vulnerable to nutritional deficiencies due to reduced nutrient reserves and an inability to respond adequately in periods of stress (11). The causes of malnutrition are multi-factorial and include anorexia, metabolic disorders, malabsorption, difficulties in swallowing, psychosocial factors and hospitalization per se (5, 12). Oral supplements have been shown to be able to improve functional condition and clinical outcome among elderly patients (7, 8). However, it is a common
Material and methods Subjects Thirty-five residents in a municipality nursing home (median age 83 years) were studied. Residents in one ward of the nursing home (n=18) were served regular hospital diet (1600 kcal/day) during a 15-week period (control group), while residents in another ward of the nursing home (n=17) were served the same meals 125
126
ENERGY-DENSE MEALS
fortified with natural energy-dense ingredients such as butter and cream (2100 kcal/day) (experimental group). The participating wards were similar in manning level and recruited the same kind of residents. The most common medical diagnoses were dementia and cerebrovascular lesions (Table 1). A total of 40 residents entered the study. Five subjects, two from the control group and three from the experimental group, died during the intervention period, and these five subjects were excluded from the study.
Table 2 Examples of meals in the two study groups during the intervention period Control diet Lunch Beef in horseradish sauce Boiled potatoes, mixed vegetables Fruit syrup dessert with milk
Experimental diet additions/changes 30 g whipping cream (40% fat) instead of milk 13 g hydrolysed starch 100 g cream (12% fat) instead of milk
Dinner Oven-baked sausage Mashed potatoes, boiled broccoli
20 g cheese (31% fat) 15 g margarine (80% fat)
Study design All residents were given the regular diet for a period of 2 weeks before the intervention, in order to establish a control period (pre-test period). Immediately following the pre-test period, residents in one ward of the nursing home (n=18) were served the regular diet prepared by the hospital kitchen with an energy content of 1600 kcal/ day or 6.7 MJ (control group) during 15 weeks (intervention period). Residents in another ward of the nursing home (n=17) were served energy-dense meals (experimental group), with a total of 2100 kcal/day, 8.8 MJ. Patients were studied again during a period of 1 week. This follow-up was performed 12 weeks after the intervention ended.
Intervention and meal composition Residents, relatives and staff were informed that evaluation of changes in food composition were about to take place, but no further details of changes in diet composition were given. The limited information was intended not to cause patient and staff bias. Meals were prepared in the kitchen of the nursing home and delivered to the wards in canteens. Hot meals (lunch and dinner) delivered to the experimental group were energy fortified with natural ingredients such as cream, oil, butter, etc., while desserts were energy fortified with cream, sour cream or hydrolysed starch. A menu for a 6 weeks period was elaborated (for examples, see Table 2). The nutritional composition for the meals was calculated (Table 3). Table 1 Diagnoses in the two study groups Main diagnosis Dementia Cerebro-vascular disorder Parkinson’s disease Multiple sclerosis Post-polio Rheumatoid arthritis Associated medical conditions Chronic heart failure Clinical oedema
Control group
Experimental group
10 4 1 1 1 1
9 4 3 1 – –
8 6
2 2
Table 3 Energy and macro-nutrient content of served main meals in control and experimental groups, and energy per cent of each macronutrientw
Energy, kcal Protein, g (energy %) Fat, g (energy %) Carbohydrate, g (energy %)
Control diet
Experimental diet
507 26(21%) 22(38%) 53(41%)
769 26(14%) 48(55%) 60(31%)
w The cost increase for the experimental diet was approx. 0.11 EUR per resident and day.
Energy and nutrient intake Before starting the study, the nursing staff were trained to perform the recording of food and fluid intake by the eye-measurement method (13). All oral intake was recorded Monday through Friday by the use of a 24 h dietary record chart. The nursing staff recorded the type and amount of food served, and the amount consumed, at breakfast, lunch and dinner. All additional intake, such as snacks and beverages, was also recorded by the staff. The nursing staff were given information each day at lunch as well as at dinner about the planned full-size portion, because of the fact that the hot meals were delivered in canteens and the staff were responsible for dimensioning the portions. Consumption of every single component of a meal, such as potatoes, meat, etc. was registered. Two research nurses and a dietitian visited the wards every week to double check portion sizes and food-intake registrations. Daily energy and nutrient intake was calculated using a computer program, Kostplan Dietist (AIVO AB, Spa˚nga, Sweden). The ratio between the estimated EI and the predicted basal metabolic rate (BMR) denoting the physical activity level (PAL), was calculated (14). The EI/BMR ratio was used to evaluate the reliability of the food and fluid records and was also related to the activity of daily living scores (see below). Subjects bedridden or bound to a wheelchair usually have a EI/BMR ratio 1.2. A sedentary life with low physical activity corresponds to ratios of 1.4–1.5 (15). The predicted BMR was calculated according to formulas recommended by the WHO (16).
CLINICAL NUTRITION
Energy expenditure was determined in eight subjects using indirect calorimetry (Deltatracr, Dansjo¨o¨, Stockholm, Sweden) (17). These eight subjects did not differ from the other 27 subjects in body mass index (BMI), body weight or total EI per day but they had less EI per kg body weight and were less dependent on assistance with primary ADL functions (data not shown). Anthropometric measurements Body weight was measured once during the pre-test, intervention and follow-up periods, respectively. Body height was measured in bed in the supine stretched position. Functional condition The functional condition was assessed by the resident assessment instrument (RAI) (18). The instrument incorporates an ADL scale. This scale includes four key ADL capabilities: mobility in bed, ability to transfer from bed to chair, capacity to use a toilet and the need of help to eat. These four capabilities are scored from 0 to 4, where a lower value indicates a better performance and a lower reliance on help from carers. With a simple algorithm, these scores are summarized to a total ADL score that runs from 4 to 18. The lower the score the better functional condition. Each subject was assessed by the registered nurse responsible for the care of the patient, together with staff who were well acquainted with the patient, and a research nurse. All episodes of infection during the intervention period were registered by staff.
127
Statistics and ethics Values are given as median and the 25th and 75th percentiles (P25–P75). Mann–Whitney U-test, Wilcoxon’s signed rank test, Spearman’s rank–order-correlation, simple regression analysis and Chi-squared exact test were used as appropriate. Informed consent was obtained from each resident and the next of kin. The protocol was approved by the Institutional Ethics Committee at the Karolinska Hospital.
Results The experimental group did not differ from the control group in initial body weight, BMI, EI or ADL score but they were slightly younger (Table 4). Four subjects in each group had a BMI below 20. The median BMI in all subjects was 23.1 (20.0–26.3): in men, 24.2 (22.7–26.3); and in women, 21.9 (19.7–23.6). Subjects in both groups were highly dependent on assistance with primary ADL functions. The median daily EI during the pre-test period for all subjects was 1353 (1159–1578) kcal/day or 23.5 (21.3–26.1) kcal/kg body weight. Twenty subjects (57%), 10 from each group, consumed less than 25 kcal/kg during the pre-test period. Serving of the experimental diet resulted in a significant increase in voluntary EI per day by 504 kcal (36% increase) (Table 5). EI remained stable in subjects served the control diet (Table 5). Fat and protein intake
Table 4 Demographic data of subjects before the start of the study Control group (n=18) Gender (females/males) Age (years) BMI (kg/m2) Body weight (kg) ADL score
83.5 23.5 63.1 15.5
Experimental group (n=17)
9/9 (79.0–89.0) (20.0–26.8) (50.9–72.3) (10.0–17.0)
80.0 22.7 56.0 16.0
P
9/8 (71.0–85.0) (20.1–24.4) (48.4–65.1) (14.0–18.0)
ns o0.05 ns ns ns
Median (P25P75). Control group vs experimental group, Mann–Whitney U-test.
Table 5 Daily intake of energy and nutrients during pre-test and intervention periods in control and experimental groups, and energy per cent of each macro-nutrient Control group (n=18)
EI (kcal/day) Energy per kg body weight (kcal/kg/day) Protein (g) Energy % Fat (g) Energy % Carbohydrates (g) Energy %
Experimental group (n=17)
Pre-test period
Intervention period
Pre-test period
Intervention period
1431 (1142–1564) 23.5 (21.9–25.4) 53.7 (42.7–58.7) 15 53.2 (41.1–61.4) 34 180 (128–204) 51
1437 (1252–1617) 24.9 (21.1–26.6) 54.7 (47.0–59.9) 15 57.6 (51.3–63.7) 34 175 (153–209) 50
1336 (1261–1578) 23.5 (21.3–28.5) 48.3 (41.8–54.3) 14 48.4 (43.8–56.4) 32 178 (152–198) 53
1840 (1497–2012)**,**** 31.9 (29.7–33.7)***,**** 57.9 (46.2–61.2)**** 12 93.1 (75.3–104.0)***,**** 45 183 (149–215) 41
Median (P25P75). **Po0.01, ***Po0.001, control group vs experimental group, Mann–Whitney U-test. ****Po0.001, pre-test vs intervention period, Wilcoxon’s signed rank test.
128
ENERGY-DENSE MEALS
increased during the intervention period in the experimental group, while carbohydrate intake remained unchanged (Table 5). Regardless of the diet during the study, residents with a lower BMI had a higher EI in relation to their body weight. Thus pre-test BMI was negatively correlated with EI during the study, r=0.60 (Po0.001). Despite differences in EI, body weights were stable in both groups throughout the intervention period. There was no correlation between EI and change in body weight during this period for the entire study population. In each group, the residents were split into subgroups, one with those with BMIo24, the other with those above this value. There was no significant change in body weight between those residents with a BMI below 24 and those above. However, for residents with a BMI below 24, served the energy-dense diet, there was a positive trend in weight gain (+0.3 kg), while those with higher BMI tended to lose weight (–1.6 kg). Subjects served the standard diet had a trend towards reduction in body weight. For residents with a low BMI served the energydense diet, EI per kg body weight increased by 34% (Po0.01). The same degree of change was found in those with BMI424 given the same diet. For residents with a BMIo24 served the standard diet, EI per kg body weight increased by 10% (Po0.01) (Table 6). Median estimated BMR, n=35 was for the entire study group 1273 (1140–1376) kcal/day. For eight subjects, the indirect calorimetry was 1015 (925–1180) kcal/day and for these eight subjects, the estimated BMR was 1272 (1198–1390) kcal/day (Table 7). The difference between methods was significant (Po0.005). However, regression analysis of indirect calorimetry vs BMR showed a high level of correlation (R2=0.86, Po0.001).
There was no difference in pre-test-calculated EI/ BMR ratios between the two study groups (Table 7). The experimental group had a significant increase in calculated EI/BMR ratio during the intervention period (Po0.001) whereas the ratio remained the same and low in the control group. There was no difference in pre-test ADL score between the two study groups (Table 7). Nor were there any differences during the intervention period. However, the control group deteriorated in their ADL functions during the intervention period (Po0.001), whereas the experimental group remained stable (Table 7). The number of infection in the experimental group tended to be lower than in the control group. In the control group, eight patients had a total of 13 infections episodes while five patients in the experimental group had one infections episode each (Table 8a and 8b). In the follow-up period, it was found that six of the 35 residents had died, three from each group. Residents in the experimental group had returned to pre-test levels of EI (Fig. 1). Discussion The present study shows that addition of natural energydense ingredients to a standard diet significantly improves voluntary energy intake in elderly nursing home residents over a prolonged period of time. Sixteen of 17 of the subjects in the experimental group increased their EI during the study. The greatest effect was registered in residents with a low BMI. The increase in EI was accompanied by a maintenance of ADL function. In contrast, subjects served a standard diet had a decline in ADL status during the study period.
Table 6 Median differences between pre-test and intervention periods for subjects with BMIo24 or above this value in control and experimental groups
BMI Differences in energy per kg body weight (kcal/kg)
Control group (n=18)
Experimental group (n=17)
o24 n=10 3.0 (1.9–4.0)
o24 n=12 8.0 (4.6–10.3)*
424 n=8 0.7 (1.6–0.2)
424 n=5 7.6 (5.3–8.5)**
Median (P25P75). *Po0.05, **Po0.01, control group vs experimental group, Mann–Whitney U-test.
Table 7 Energy served, EI, BMR, estimated physical activity level (PAL) according to Nordic Nutrition Recommendations, estimated energy requirement (BMR estimated PAL), EI/BMR ratio and ADL-score during pre-test and intervention periods in control and experimental groups Control group (n=18)
Energy served (kcal) EI (kcal) Energy requirement BMR Estimated PAL EI/BMR ratio ADL score
Experimental group (n=17)
Pre-test period
Intervention period
Pre-test period
Intervention period
1600 1431 (1142–1564) 1569 (1476–1782) 1288 (1230–1361) 1.2 1.1 (0.9–1.2) 15.5 (10.0–17.0)
1600 1437 (1252–1617) 1558 (1460–1795) 1294 (1140–1376) 1.2 1.2 (1.1–1.2) 16.0 (15.0–18.0)****
1600 1336 (1261–1578) 1501 (1368–1690) 1225 (1140–1376) 1.2 1.1 (1.0–1.3) 16.0 (14.0–18.0)
2100 1840 (1497–2012)**,**** 1506 (1396–1662) 1230 (1164–1344) 1.2 1.5 (1.3–1.7) ***,**** 15.0 (14.0–18.0)
Median (P25P75). **Po0.01, ***Po0.001, control group vs experimental group, Mann–Whitney U-test. ****Po0.001, pre-test vs intervention period, Wilcoxon’s signed rank test.
CLINICAL NUTRITION
129
Table 8a Number of episodes of infections in control vs experimental groups Infection Pneumonia Urinary Skin Total
Control group (n=18)
Experimental group (n=17)
4 8 1 13
1 3 1 5
b Distribution of infections in subjects in the control vs experimental groups
Control group Experimental group
No infection
One infection
Two infections
Three infections
10 12
4 5
3 0
1 0
Comparisons with Chi-square exact test P=0.29. 60
50
Kcal/kg b.wt
40
30
20
10
0 Pre-test Experimental
Intervention
Follow up
Control
Fig. 1 Median EI related to body weight in the control group (n= 18) and the experimental group (n=17) during the pre-test, intervention and follow-up periods. (Median P25P25) and range.
The subjects in this study were elderly and suffered from a variety of chronic diseases which are often associated with malnutrition, and in high need of care. The high dependency on nursing care was confirmed by high ADL scores. The median BMI for all residents was 23.1 before the start of the project, similar to values reported in earlier studies in this category of patients (5, 6). A BMI between 24 and 29 has recently been suggested as optimal for elderly patients, especially those with a poor EI (19). Twenty-two of the 35 subjects (63%) had BMI values below this range. However, BMI may be unreliable as an indication of nutritional state in this population, as a result of diseases causing fluid retention. The subjects in our study were grouped according to wards rather than truly randomized, since it proved practically impossible to serve different diets within the same ward. Nevertheless, the study groups were well matched, with a slightly younger median age in the experimental group.
It was found that the experimental diet had the greatest effect on the residents with a low BMI. This is in line with findings reported by Gall et al. who offered patients an extra 966 kcal/day and 22.2 g protein in meal fortification and snacks. Intakes were registered over a 3-day period and the increase in EI among orthopaedic, medical and elderly care patients was 19.6–23% (20). The percentage increase in EI in the present study was even higher than that. The residents in the control group were served a diet according to Nordic Nutrition Recommendations (15). These recommendations are given for healthy elderly. However, there is an ongoing discussion about specific recommendations for elderly sick persons. In the present study, it was only possible to perform indirect calorimetry in eight subjects, since most subjects were incapable of going through the procedure properly. The findings from the eight subjects confirm that estimations of BMR are often higher than the measurements from indirect calorimetry (21). Nevertheless, these
130
ENERGY-DENSE MEALS
findings indicate that energy expenditure is low in elderly subjects under nursing care. The Nordic recommendations for dietary EI in the elderly are based on an estimated BMR and adjusted for activity. For the present population, the men had a median estimated BMR of 20.5 kcal/kg/day and with the suggested adjustment for activity an energy requirement of 24.6 kcal/kg/day. Similarly, the women had a BMR of 21.5 and 25.8 kcal/kg/day in requirements. The regular diet served in this study contained 26.0 kcal/kg. However, the determinations of indirect calorimetry suggest that needs may actually be lower in these frail elderly individuals. The energy-dense diet supplied well above the estimated needs of calories. One of the aims of the study was to investigate the long-term effects of energy-dense food. Surprisingly, the body weight of the patients in the experimental group did not increase, despite a substantially higher intake of fat during 15 weeks. In our previous study of geriatric patients, 6 weeks of high energy food intake resulted in a significant 3.4% increase in body weight (13). This suggests that only small weight changes are to be expected even during a prolonged treatment period. There may be a variety of reasons why the subjects in the experimental group in this study did not gain weight. Cardiac failure, renal insufficiency and medication may influence body water content and subsequently body weight. We were unable to control these factors in this study. Six patients in the control group had oedema on a clinical basis but only two patients in the experimental group. This may, to some extent, explain the stable weight in both groups. A more precise estimation of body composition could have given information about any potential changes in these parameters, but was not performed presently for practical reasons. Another possible reason for weight stability despite increased EI could be an increased energy expenditure due to increased physical activity. There are several instruments for measuring ADL functions, all with specific limitations. In this study, we used the RAI for measuring functional condition. The RAI has been positively tested for validity and reliability (22). In the future, modifications in scales measuring ADL function may allow greater sensitivity in detecting changes caused by nutritional intake. In this study, the subjects in the control group had a small but significant decrease in ADL function score over time. For this population, this change is not unexpected. In contrast, the experimental group was able to maintain the ADL score, suggesting that non-energy-dense food may contribute to successive decline in function in an elderly frail population. This reduction in functional capacity may result in increased need for nursing care. The subjects in this study had pre-test very low EI/BMR ratios which may correspond to the low physical activity of the residents. Only 8 of 35 subjects (23%) had EI/BMR ratios above 1.2 during the pre-test period. After the intervention, 13 of 17 subjects (76%)
had EI/BMR ratios above 1.2. The EI/BMR ratio combined with the minimal weight change suggests that the physical activity in the experimental group increased during the intervention period. Nevertheless, the registered ADL score in the experimental group was not significantly improved. The number of infections tended to be lower in the experimental group than in the control group, but there was no significant difference between the two groups. However, earlier studies have described that the immune defence system is weakened in malnourished patients and that the risk for infections is increased (23–25). Energy-dense food needs to contain a high proportion of fat if it is to be fortified with natural ingredients. The risk for the elderly population in this study to develop secondary cardiovascular problems due to increased fat consumption seems less likely than complications from malnutrition, but is a question that needs to be addressed. There are alternative ways to improve EI. An energydense diet with natural ingredients has several advantages. It is readily accepted by elderly subjects who may not accept other nutritional supplements. Energy-dense meals can easily be prepared in any hospital or nursing home kitchen. It is also by far the cheapest way to increase the energy content of the food for these patients. By using fat as a means of fortification, the taste of food improves and this was commented upon by many subjects. In conclusion, this study shows that the daily EI in elderly nursing home residents can be improved over a prolonged period of time by the simple method of adding natural energy-dense ingredients to regular meals. Serving energy-dense food is an inexpensive way to improve EI in this population. Improving EI using this simple treatment seems to help maintain ADL function in elderly nursing home residents. Acknowledgements The nursing staff in the nursing home Hallen in Solna is gratefully acknowledged. We thank Elisabeth Ljungqvist, RN for her assistance in collecting the data. The study was supported by grants from the Swedish National Board of Health and Welfare and the National Institute of Public Health, Sweden.
References 1. McWhirter J P, Pennington C R. Incidence and recognition of malnutrition in hospital. Br Med J 1994; 308: 945–948 2. Cederholm T, Ja¨gre´n C, Hellstro¨m K. Nutritional status and performance capacity in internal medical patients. Clin Nutr 1993; 12: 8–14 3. Volkert D, Kruse W, Oster P, Schlierf G. Malnutrition in geriatric patients: diagnostic and prognostic significance of nutritional parameters. Ann Nutr Metab 1992; 36: 97–112 4. Thomas D R, Verdery R B, Gardner L, Kant A, Lindsay J. A prospective study of outcome from protein–energy malnutrition in nursing home residents. J Parenter Enteral Nutr 1991; 4: 400–404 5. Blaum C S, Fries B E, Fiatarone M A. Factors associated with low body mass index and weight loss in nursing home residents. J Gerontol Med Sc 1995; 3: M 162–168
CLINICAL NUTRITION
6. Saletti A, Yifter Lindgren E, Johansson L, Cederholm T. Nutritional status according to Mini Nutritional Assessment in an institutionalized elderly population in Sweden. Gerontology 2000; 46: 139–145 7. Larsson J, Unosson M, Ek A C, Nilsson L, Thorslund L. Effect of dietary supplementation on nutritional status and clinical outcome in 501 geriatric patientsFa randomised study. Clin Nutr 1990; 9: 179–184 8. Delmi M, Rapin C-H, Bengoa J M, Delmas P D, Vasey H, Bonjour J P. Dietary supplementation in elderly patients with fractured neck of the femur. Lancet 1990; 335: 1013–1016 9. Gallagher-Allred C R, Voss A C, Finn S C, McCamish M A. Malnutrition and clinical outcomes: the case for medical nutrition. J Am Diet Assoc 1996; 96: 361–366 10. Chima C S, Barco K, Dewitt M L A, Maeda M, Teran J C, Mullen K D. Relationship of nutritional status to length of stay, hospital costs, and discharge status of patients hospitalized in the medicine service. J Am Diet Assoc 1997; 97: 975–978 11. Rapin C H. Nutrition support and the elderly. In: Payne-James J, Grimle G, Silk D (eds). Artificial Nutrition Support in Clinical Practice. Filey, North Yorkshire: Edward Arnold, 1995; 535 12. Green C J. Existence, causes and consequences of disease-related malnutrition in the hospital and the community, and clinical and financial benefits of nutritional interventions. Clin Nutr 1999; 18 (suppl 2): 3–28 13. O¨dlund Olin A, O¨sterberg P, Ha˚dell K, Armyr I, Jerstro¨m S, Ljungqvist O. Energy-enriched hospital food to improve energy intake in elderly patients. J Parenter Enteral Nutr 1996; 2: 93–97 14. Hambreaus L. Dietary assessments: how to validate primary data before conclusions can be drawn. Scand J Nutr 1998; 42: 66–68 Submission date: 14 December 2001 Accepted: 30 September 2002
131
15. Nordic Nutrition Recommendations 1996: Nordic 1996: 28, Copenhagen: Nordic Council of Ministers, 1996 16. FAO/WHO/UNU. Report of a joint Expert Consultation: Energy and Nutrient requirements. Technical Report Series 724, WHO, Geneva, 1985 17. Frayn K N. Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl Physiol 1983; 55: 628–634 18. Morris J N, Hawes C, Fries B E et al. Designing the national resident assessment instrument for nursing homes. Gerontologist 1990; 30: 293–302 19. Beck A M, Ovesen L. At which body mass index and degree of weight loss should hospitalized elderly patients be considered at nutritional risk. Clin Nutr 1998; 17: 195–198 20. Gall M J, Grimble G K, Reeve N J, Thomas S J. Effect of providing fortified meals and between-meal snacks on energy and protein intake of hospital patients. Clin Nutr 1998; 6: 259–264 21. Beck AM, Ovesen L. Body mass index, weight loss and energy intake of old Danish nursing home residents and home-care clients. Scand J Caring Sci 2002; 16: 86–90 22. Sgadari A, Morris J N, Brant B E et al. Efforts to establish the reliability of the RAI. Age Ageing 1997; 26 (suppl 2): 27–30 23. Chandra R K. Nutrition and the immune system. Proc Nutr Soc 1993; 8: 243–248 24. Chandra R K. Nutritional regulation of immunity and risk of infection in old age. Immunology 1989; 67: 141–147 25. Eriksson K M, Cederholm T, Palmblad J E. Nutrition and acute leukaemia in adults: relation between nutritional status and infectious complications. Cancer 1998; 82: 1071–1077