Comparison of two different approaches for the application of the mini nutritional assessment in nursing homes: Resident interviews versus assessment by nursing staff

Abstract

BackgroundWhen the Mini Nutritional Assessment (MNA©) was developed, the authors did not specifically focus on the nursing home setting.
Due to a number of particularities of nursing home residents, such as cognitive and linguistic disabilities, a number of uncertainties
with regard to its application await clarification.

Aims and objectivesThe aim of this study was to compare the results of two different modes of MNA application in nursing homes: resident interviews
versus assessment by nursing staff.

MethodThe MNA was applied to 200 residents of two municipal nursing homes in Nuremberg, Germany. First one-on-one interviews of
the residents were conducted by two researchers from our group. Next, the MNA was applied by the attending nursing staff who
was blinded to the results of the first MNA. To evaluate the prognostic properties of the two different approaches, data on
mortality of the screened residents were collected during a six-month follow-up period.

ResultsAmong 200 residents (f 147 m 53, f 86.5±7.4 y. m 83.0±8.5 y.), the MNA could be applied t 138 residents (69. 0%) by one-on-one
interviews and to 188 residents (94.0%) by the nursing staff. 15.2% of the residents were categorised as malnourished by the
interviews and 8.7% by the nursing staff’s assessment. The agreement of the two forms was low for the MNA short form (weighted
kappa = 0.31; 95% CI: 0.14 − 0.47) as well as for the full MNA (weighted kappa = 0.35; 95% CI: 0.27 − 0.44). After exclusion
of residents with cognitive impairment (n=89), agreement for the full version increased (weighted kappa = 0.47, 95% CI 0.
25 − 0.68). 25 (12.5%) study participants deceased during the follow-up period. Mortality was significantly associated with
the mortality for both approaches, while the MNA application by the nursing staff proved to be superior (nursing staff p<0.001,
residents p<0.05).

ConclusionsThe results of the MNA in nursing home residents may differ substantially when resident interviews are compared to assessment
by nursing staff. The authors recommend that the MNA should be routinely applied by the nursing staff. The application rate
is higher and interference with cognitive as well as linguistic deficits is lower. In future studies, the mode of MNA application
in nursing home residents should be clearly stated to facilitate comparability of results.

Key wordsNutrional screening-Mini Nutritional Assessment-malnutrition-nursing home

Full text

Page 1

Introduction
Malnutrition has been identified as a frequent problem
among the frail elderly in most industrialized countries. It is
associated with loss of functionality, lower quality of life,
higher risk of morbidity and increased mortality (1-4, 7). The
highest prevalence rates for malnutrition were reported for
elderly patients in hospitals and those living in nursing homes,
with a range of between 5% and 70%, depending on the
definition of malnutrition applied (5, 6). Still, no gold standard
exists for the latter. In 2003, the European Society for Clinical
Nutrition and Metabolism (ESPEN) recommended three
nutritional screening tools to be applied according to the
population which is to be screened (4). These tools were the
Malnutrition Universal Screening Tool (MUST), the Nutritional
Risk Screening (NRS) and the Mini Nutritional Assessment
(MNA). The latter has been regarded as the most suitable
screening tool for the geriatric population. The original study,
published in 1994 by Guigoz, Vellas and Garry, was based on
three different cohorts of elderly persons from the Toulouse
area and from the New Mexico Aging Progress Study (8, 9).
The MNA was developed for practicable and rapid nutritional
screening in frail elderly people. It was designed to be
completed in 10 to 15 minutes, with the intention of targeting
early nutritional intervention (10). Meanwhile, the MNA has
become the best documented screening tool for malnutrition in
the elderly, with more than 10,000 study participants in
different settings (6).
With regard to its application in nursing homes, several
studies have been published over the last two years (11-19).
However, the application of the MNA may be hampered by
cognitive and linguistic disabilities, which are frequently
present among nursing home residents. Similar problems have
been described in an application study among hospitalized
geriatric patients (20). Under these circumstances, the MNA
was completed by the attending nurses in most cases. Potential
differences between these two approaches of MNA application
– one-on-one resident interview versus assessment by nursing
staff – have not been systematically analysed before for only
Tsai et al. published a comparison with caregivers` assessment,
COMPARISON OF TWO DIFFERENT APPROACHES FOR THE APPLICATION
OF THE MINI NUTRITIONAL ASSESSMENT IN NURSING HOMES:
RESIDENT INTERVIEWS VERSUS ASSESSMENT BY NURSING STAFF
R. KAISER1, K. WINNING1, W. UTER2, S. LESSER3, P. STEHLE3, C.C. SIEBER1, J.M. BAUER1
1. Department of Geriatric Medicine, University of Erlangen-Nuremberg, Germany; 2. Department of Medical Informatics, Biometry and Epidemiology, University of Erlangen-
Nuremberg, Germany; 3. Department of Nutrition and Food Sciences (IEL) - Nutritional Physiology, University of Bonn, Germany. Corresponding author: Juergen M. Bauer
Department of Geriatric Medicine, University of Erlangen-Nuremberg, Prof.-Ernst-Nathan-Str. 1, D-90419 Nuremberg, Germany, Phone: +49-398-2748, Fax: +49-398-2117,
Email: juergen.bauer@klinikum-nuernberg.de
Abstract: Background: When the Mini Nutritional Assessment (MNA©) was developed, the authors did not
specifically focus on the nursing home setting. Due to a number of particularities of nursing home residents, such
as cognitive and linguistic disabilities, a number of uncertainties with regard to its application await clarification.
Aims and objectives: The aim of this study was to compare the results of two different modes of MNA
application in nursing homes: resident interviews versus assessment by nursing staff. Method: The MNA was
applied to 200 residents of two municipal nursing homes in Nuremberg, Germany. First one-on-one interviews of
the residents were conducted by two researchers from our group. Next, the MNA was applied by the attending
nursing staff who was blinded to the results of the first MNA. To evaluate the prognostic properties of the two
different approaches, data on mortality of the screened residents were collected during a six-month follow-up
period. Results: Among 200 residents (f 147 m 53, f 86.5±7.4 y. m 83.0±8.5 y.), the MNA could be applied to
138 residents (69.0%) by one-on-one interviews and to 188 residents (94.0%) by the nursing staff. 15.2% of the
residents were categorised as malnourished by the interviews and 8.7% by the nursing staff’s assessment. The
agreement of the two forms was low for the MNA short form (weighted kappa = 0.31; 95% CI: 0.14 - 0.47) as
well as for the full MNA (weighted kappa = 0.35; 95% CI: 0.27 - 0.44). After exclusion of residents with
cognitive impairment (n=89), agreement for the full version increased (weighted kappa = 0.47, 95% CI 0.25 -
0.68). 25 (12.5%) study participants deceased during the follow-up period. Mortality was significantly associated
with the mortality for both approaches, while the MNA application by the nursing staff proved to be superior
(nursing staff p<0.001, residents p<0.05). Conclusions: The results of the MNA in nursing home residents may
differ substantially when resident interviews are compared to assessment by nursing staff. The authors
recommend that the MNA should be routinely applied by the nursing staff. The application rate is higher and
interference with cognitive as well as linguistic deficits is lower. In future studies, the mode of MNA application
in nursing home residents should be clearly stated to facilitate comparability of results.
Key words: Nutrional screening, Mini Nutritional Assessment, malnutrition, nursing home.
1
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
THE JOURNAL OF NUTRITION, HEALTH & AGING©
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 1

Page 2

focussing exclusively on the two MNA questions that are
exploring self-assessment with regard to nutritional status and
general health (11). In the present study, we analysed these two
approaches for the short form and the full MNA. We also tested
their prognostic value with regard to mortality in a six-month
follow-up period.
Subjects
Two hundred residents from two municipal nursing homes in
Nuremberg, Germany, were recruited for this study from June
to December 2007. Residents were excluded if they were under
65 years of age, terminally ill or did not give informed consent
(resident or legal proxy, respectively). The study protocol was
approved by the ethics committees of the Friedrich-Alexander-
University of Erlangen-Nuremberg, Germany, and the
Rheinische Friedrich-Wilhelms-University of Bonn, Germany.
Method
A nutrition scientist (KR) and a physician (WK) approached
all residents in two Nuremberg nursing homes and/or their legal
proxy. After obtaining informed consent, all participants were
examined in an identical manner. First, the MNA was
completed with the help of the resident in an one-on-one
interview, conducted by either WK or KR. Afterwards, the
MNA for each subject was completed by the attending nursing
staff member, who was blinded with regard to the result of the
initial MNA. Here all results with the exception of the sub-
items F, Q and R (see below) were based on the estimation by
the nursing staff, their experience and their awareness of the
medical records. In six cases the length of stay was less than
three months (one month in two cases, two months in four
cases); here the nursing staff members completed the MNA
with the knowledge and data available to them by that time.
The two MNAs for each resident were always completed within
one week.
The MNA consists of two parts: the MNA-Short Form
(MNAsf) with six items (items A-G) and the full MNA which
comprises the MNAsf and twelve additional items. The MNAsf
was developed by Rubenstein in 2001 (27) and scores a
maximum of 14 points. The screened subject is classified as
“normal” with scores of 12 or more; below 12, the subject is “at
risk of malnutrition”. The full MNA (items G-R) contains 12
additional questions on the living situation, drug intake, skin
lesions, anthropometric parameters, dietary and subjective
assessment. The nutritional status is classified into three
categories: 24 - 30 points indicate that the subject is well-
nourished, 17 - 23.5 points indicate that the resident is at risk of
malnutrition and up to 16.5 points mean that he/she is
malnourished. In the present study, items F (BMI), Q (mid-arm
circumference) and R (calf circumference) were assessed by
KR and WK for both modes. The question about the residents`
living situation (item G) always scored 0 points, due to the
nursing home setting of the study.
Cognitive status was determined by the Mini Mental State
Examination by Folstein (MMSE) (21). The emotional status
was tested by the short form of the Geriatric Depression Scale,
validated by Lesher and Berryhill in 1994 (22). Anthropometric
measurements were weight, knee height, mid-arm
circumference (MAC) and calf circumference (CC). The
measurements of all subjects were performed by the nutrition
scientist (KR). The weight was measured to the nearest 0.1 kg
using a weigh chair (type Arjo CFA 2000). The knee height
was measured by a knee height sliding calliper provided by the
Austrian Society for Clinical Nutrition (AKE) (23) in a sitting
position. Bedridden residents were measured in a supine
position. The knee and the ankle had to be bent at 90°. The
knee height was recorded to the nearest 0.1 cm for all
participants. The measurement was carried out as described by
Chumlea et al. (24). The body height was calculated using the
equations developed by Chumlea et al. in a cohort beyond the
age of 60 (25):
white non-Hispanic male: 78.31 + (1.94 x knee height) – (0.14
x age)
white non-Hispanic female: 82.21 + (1.85 x knee height) –
(0.21 x age)
The Body Mass Index (BMI) was calculated by body weight
(kg) divided by squared body height (m2).
The measurements of the MAC and the CC were carried out
with a flexible measuring tape and were recorded to the nearest
0.1 cm. The MAC was measured at the mid-point between the
tip of the shoulder and the tip of the elbow between olecranon
process and the acromion on the left side. The CC was
measured at the widest point of the calf muscle without
compressing the subcutaneous tissue of the patient in a supine
or sitting position on the left side. The mean was taken from
two consecutive measurements. Individual variables, such as
gender, age, social and medical data, were collected from the
medical records of the nursing home or from the nursing staff
by way of a questionnaire. The Barthel’s Index, developed in
1965 by Mahoney and Barthel (ADL) (26), was carried out by
the nursing staff. Maximum score was 100 points; a score of
>64 points was regarded as independent in activities of daily
living, 35 – 64 points was indicative of being in need of
assistance, while a score <35 points indicated need of care.
During a six-month follow-up period, data on the residents’
mortality were collected.
Statistics
Statistical analysis was performed using SPSS© version 16.0
(SPSS for windows, SPSS Inc., Chicago, IL, USA) and SAS
(version 9.1, SAS Institute, Cary, NC, USA). Inter-rater
reliability - in this context meaning the agreement between one-
on-one interviews and assessment by nursing staff - was
quantified by Cohen’s kappa index (28) for the classified MNA
outcome and the individual items of the long version of the
MNA. In the case of ordinal items with more than two levels,
the weighted kappa value was estimated. To examine whether
RESIDENT INTERVIEWS VERSUS ASSESSMENT BY NURSING STAFF
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
2
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 2

Page 3

differences between the two ratings depart from chance
variation, i.e. whether there are systematic differences between
the two assessments, the McNemar and the Bowker test for
symmetry was used, as appropriate. Correlation between MNA
score values was quantified with the Spearman rank correlation
coefficient. Statistical significant differences were taken by p-
values <0.05.
Results
Study population
Out of 322 residents living in the two nursing homes, 122
persons (98 females and 24 males) were not eligible for the
present study. For 42 residents, their proxies did not agree in
their participation (34.4% of non-participating residents).
Proxies could not be contacted in another 14 cases (11.5%). 28
residents were personally unwilling to participate in the study
(23.0%). Ten residents were terminally ill (8.2%) and 10
residents were under age 65 (8.2%). Other residents did not
participate for various reasons such as acute hospitalisation,
infection with multi-resistant bacterial strains, etc. (18
residents, 14.8%). The mean age of the non-participating
residents (82.8±11.5 years) was significantly lower than that of
the participating residents.
Of the 200 residents included, 147 were female and 53 male
(mean age 85.5±7.8 years). All participants were Caucasians.
Women were significantly older than men (p<0.05). The
majority of included residents were multimorbid, with diabetes
mellitus being present in 36.0% of the residents, arterial
hypertension in 76.5%, chronic heart failure in 74.0%, residual
symptoms after cerebral ischemia in 29.0%, and osteoarthritis
in 31.0%. The mean number of drugs taken on a daily basis was
7.8±4.9 (range 0-24). The GDS was completed in 71.0% of the
residents. 50.0% of these were found to have a score indicative
of overt depression. The MMSE was applicable to 91.0% of the
residents. 75.5% of these showed results indicating relevant
cognitive impairment (<25 points). With regard to the activities
of daily living (ADL), 28.0% of the residents were independent
and 29.5% needed assistance. 42.5% required a high level of
care. The characteristics and anthropometric data of the
participants are shown in table 1.
Table 1
Characteristics of participants
WomenMen
Number of participants
Age [y]c
ADL [points/100]
GDS [points/15]
MMSE [points/30]
Body weight [kg]c
BMI [kg/m2]
MAC [cm]
CC [cm]
14753
86.5(± 7.4)b
40 [15-65]a
5 [3-8]a
19 [1-24.5]a
63.2 (± 13.9)b
26.5 (± 5.5)b
28.3 (± 4.5)b
32.6 (± 5.2)b
83.0 (± 8.5)b
45 [20-67.5]a
4 [2-8.5]a
19 [7.5-29]a
74.6 (± 15.5)b
25.6 (± 4.4)b
29.1 (± 4.5)b
33.0 (± 4.3)b
a. median (1. quartile-3. quartile), b. mean ± SD, c. (t-test p<0.05); ADL = Activities of
Daily Living, GDS = Geriatric Depression Scale, MMSE = Mini Mental State
Examination, MAC = Mid-arm circumference, CC = calf circumference
Application of the MNA
Application of the MNA by one-on-one interview of the
resident was possible in 69.0% (n=138). The reasons for non-
applicability included communicative deficits, such as aphasia
and cognitive impairment in 25%, tube feeding in 5% and
amputation of both lower limbs in 1%, as CC could not be
measured in these cases. The MNA could be completed by
assessment of the nursing staff in 94.0% of the cases (n=188).
Again, the MNA could not be applied due to tube feeding (5%)
or lower limb amputations (1%). The following analyses are
based on the subset of data that included only those residents
for whom the MNA was completed by both approaches
(n=138).
MNA short form (MNAsf)
Figure 1 shows a scatter plot of score points obtained by the
MNAsf for resident interviews versus assessment by nursing
staff. The size of bubbles indicates the frequency when scoring
results of both approaches met. The cut-offs of the two
categories of the MNAsf – “normal” and “at risk of
malnutrition” - are represented by horizontal and vertical lines.
According to the one-on-one interviews 76.1% of residents
were categorised as “at risk of malnutrition”, while 63.8% were
categorised in the same group by the nursing staff, respectively
(table 2). So, a higher number of residents was categorised as
“normal” when they were screened by the nursing staff.
Correlation between the two modes of application was
moderate (Spearman rank correlation coefficient 0.60,
p<0.001). The test of symmetry revealed a significant
difference (p<0.05, McNemar test). Agreement was only fair
(kappa 0.31; 95% CI: 0.14 - 0.47). In addition, we analysed five
sub-items of the MNAsf separately to identify the most obvious
differences between the two approaches (table 3). With regard
to item A, which addresses loss of appetite and reduced food
intake, it was noticed that 14 residents reported severe loss of
appetite, while only six residents were placed in this category
by the nursing staff. Agreement between the two approaches
was only low for this item. When confronted with the second
question of the MNAsf that explores ‘weight loss during the
last three months’ residents chose the option ‘does not know’ in
23.0%, while the nursing staff provided a definitive answer in
every case. There was also a considerable discrepancy with
regard to the option ‘no weight loss’ (57.9% resident vs. 78.2%
nursing staff). Item C ‘mobility’ showed the highest agreement
among the items of the short form (weighted kappa 0.52). For
this item, no significant difference was found between the two
approaches. With regard to item D, 29.8% of the residents
indicated that they suffered from acute disease or psychological
stress in the past three months, which is close to the 28.9%
determined by the nursing staff’s assessment, but again,
agreement was only fair (kappa = 0.28). For item E the rating
of presence of depression and dementia significantly skewed
(p<0.05) the way that residents estimated their
neuropsychological problems as less severe than the nursing
staff did. Anyhow, between the two approaches a fair but
THE JOURNAL OF NUTRITION, HEALTH & AGING©
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
3
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 3

Page 4

significant agreement (weighted kappa = 0.19) was observed.
Figure 1
Scatter plot of for the MNAsf: resident interview versus
screening by nursing staff. The size of the bubbles indicates the
frequency when scoring results of both approaches met. The
cut-offs of the two categories of the MNAsf are represented by
the horizontal and vertical lines
Table 2
Contingency table for MNAsf
Nursing staff
Risk of Well-
nourished malnutrition
Total
Resident
interviews Well-nourished
Risk of malnutrition n
Total
n 21
29
50
12
76
88
33
105
138n
Table 3
Comparison of selected items of the MNA by resident
interviews versus assessment by nursing staff
(weighted)§ kappa
(95% CI)
Bowker (McNemar)$
test of symmetry
(difference)
A Anorexia
B Weight loss
C Mobility
D Acute disease
E Neuropsychological
problems
O Nutritional status
P Health status
§0.12 (0-0.25)
§0.19 (0.07-0.32)
§0.52 (0.40-0.63)
0.28 (0.11-0.46)
§0.19 (0.06-0.32)
p<0.001
p<0.001
n.s.
$n.s.
p<0.05
§0.03 (0-0.18)
§0.03 (0-0.14)
n.s.
n.s.
n.s. = not significant
Full MNA
Resident interviews yielded a mean score of 21.4 points
(median 22, Interquartile range [IQR] 19 - 24.5). 15.2% of the
residents were categorised as malnourished, 52.9% as being at
risk of malnutrition and 31.9% as being well-nourished. For the
same residents, based on the assessment by the nursing staff,
the mean score of the total MNA was 21.3 points (median 22,
IQR 18.5 - 24). 8.7% were categorised as malnourished, 54.3%
as being at risk of malnutrition and 37.0% as being well-
nourished. The correlation between the two scores was
moderate (Spearman rank correlation coefficient 0.58), albeit
significant (p<0.001) (figure 2). Individual agreement of the
two approaches based on the sub-items was fair (weighted
kappa 0.35, 95% CI: 0.27 - 0.44) (table 4).
Table 4
Comparison of the MNA categorisation obtained by the two
approaches
Nursing staff
Risk of
malnutrition
Well-nourishedMalnutrition Total
Resident
interviews
Well-nourished
Risk of malnutrition
Malnutrition
Total
26
25
0
51
17
44
14
75
1
4
7
12
44
73
21
138
Figure 2
Scatter plot of the full MNA score resident vs. nursing staff.
The size of bubbles indicates the frequency when scoring
results of both approaches met. The lines represent the cut-off
values of the three MNA categories
Agreement between the two approaches was, apart from
item H ‘prescription drugs intake’ (kappa = 0.38) and item I
‘skin problems’ (kappa = 0.21), for all additional items of the
long form rated as not higher than would have been by chance
distribution. Symmetry of classification distribution tests
4
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
RESIDENT INTERVIEWS VERSUS ASSESSMENT BY NURSING STAFF
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 4

Page 5

5
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
THE JOURNAL OF NUTRITION, HEALTH & AGING©
revealed, that for item K ‘consumption markers for protein
intake’ nursing staff provided more points than residents did,
while for items H ‘prescription drugs intake’, L ‘consumption
of fruits and vegetables’ and item M ‘fluid consumption’, the
residents interviews scored higher. With regard to item O
addressing the estimation of the residents’ nutritional status,
results showed a similiar distribution for the self-estimation and
the estimation by the nursing staff. Here, only 2.2% (nursing
staff) and 2.9% (residents) were viewed as malnourished,
respectively. Nevertheless, agreement was poor (kappa = 0.02).
With regard to the estimation of the general health status in
item P, more residents than nursing staff assessed their health
status as better than that of their peers (29.0% vs. 16.7%),
depicting a poor agreement (kappa = 0.03) (table 3).
Cognitive status and MNA application
A subanalysis on categorisation by the full MNA was
performed for those 49 residents with a MMSE score of at least
25 points, indicating the absence of dementia. The agreement
between the two approaches increased to weighted kappa =
0.47 (95% CI: 0.25-0.68).
MNA application and prognosis
The prognostic value of the two modes of application were
evaluated with regard to mortality during a six-month follow-
up period. In total, 25 (12.5%) study participants deceased
during this period. Table 5 shows the mortality rate for the
different MNA categories (n=138). 19.0% of the residents
categorised as malnourished by one-on-one inerviews and
33.3% of those placed in this category by assessment of the
nursing staff residents deceased during the follow-up. The
association between mortality and the categorisation was
statistically stronger for the nurses’ assessment (nursing staff
p<0.001, residents p<0.05; Chi2test).
Table 5
Six-month mortality rate in the three categories of the MNA –
resident interviews vs. assessment by nursing staff
Nursing staffa
Residenta
Well-nourished
At risk of malnutrition
Malnutrition
3.9%
9.3%
33.3%
0%
12.3%
19.0%
a. Significant trend of increasing mortality across the three ordinal MNA score categories
(Mantel-Haenszel-Chi²: p < 0.05)
Discussion
This is the first study in nursing home residents that
compared the results of the MNA applied by one-on-one
interviews with the results obtained when the MNA was
performed solely by the attending nursing staff. Applicability
was different for these two approaches, with a lower rate for the
one-on-one interviews (69.0% vs. 94.0%). The agreement of
the results of the two approaches of application was tested for
the MNA short form and for the full MNA. It was low for both
methods, but improved when cognitively impaired participants
were excluded from analysis. While mortality after six months
was associated with the MNA results of both approaches, this
relationship was statistically stronger for the MNA completed
by the nursing staff. Notably, residents that were unable to
complete the MNA in an one-to-one interview were in a
markedly worse health situation, as 11 of them died during the
follow-up period (44.0% of the 25 deceased participants).
The low rate of applicability of one-on-one interviews was
mainly due to the high prevalence of cognitive and linguistic
deficiencies that interfered with the completion of the MNA.
As this screening tool was not designed with a special focus on
nursing home residents but for frail elderly people in general,
this problem was not specifically addressed when the MNA
was developed in the early nineties (8, 9). In a relevant number
of subsequent studies, residents with cognitive deficits were
excluded from study participation (16, 29, 30). Alternatively,
MNA completion relied on information obtained from close
relatives (29), while some authors did not even disclose how
they dealt with this problem (32, 37). In a study on geriatric
hospital patients that did not exclude those with dementia, an
applicability rate of 66.1% was reported (20), causes for non-
applicability including confusion, advanced dementia, post
stroke aphasia and apraxia. With regard to the long term care
setting, Donini et al. (33) reported that all MNA questions were
answered completely only in 17.5% of cases. Additionally, the
authors of this study could not perform the anthropometric
measurements correctly in 6.2% of cases. In our study, this
problem was present in only 1% of participants.
In the present study, we aimed at the inclusion of a
representative sample of the nursing home population. In
addition, dementia has to be regarded as one of the most
relevant comorbidities with regard to malnutrition in the
elderly. Therefore, we did not exclude residents with dementia
from our analyses despite the above presented difficulties. The
agreement between the interviews and the assessment by
nursing staff was low with regard to the results of sfMNA as
well as those of the full MNA. Likewise, the agreement was
low for several of the MNA items that were analysed separtely.
The observation that the agreement between the two
approaches improved when those residents were excluded that
showed dementia according to the MMSE illustrates the
relevance of this frequent comorbidity for the reliability of
MNA categorisation.
While it is obvious that the impairment of memory seen in
dementia will generally affect the reliability of answers
provided by the resident, the MNA questions that address the
residents` self assessment (items O and P) may be seen as
especially crucial for the application of the MNA in the nursing
home setting. As their living space and their personal contacts
usually become limited, many nursing home residents loose the
capability to self-assess their nutritional and general health
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 5

Page 6

status in comparison to their peers. This observation can be
applied not only to those that are affected by early stages of
dementia, but also to residents with residues after cerebral
ischemia and to those with Parkinson`s disease. The
aforementioned problems with regard to the MNA self-
assessment items of institutionalized elderly persons have been
addressed only by few studies yet. In a recent study by Tsai et
al., the answers provided by interviews with the residents were
compared to those provided by the attending nurses in a long-
term care centre (11), but only in the sub-group of cognitively
competent residents. Results showed that residents assessed
their nutritional and health status more often as poor and that
they were frequently insecure about their status than when
compared to the nursing staff’s assessment. In another study,
substitution of the two questions addressing self-assessment
was suggested and tested (33).
In the MNA user guide (38), it is recommended to ask the
patient’s caregiver or the attending nursing staff member, if the
resident himself is not capable to provide an adequate answer to
MNA questions. But it is not an easy task to decide when to
shift the MNA application from the interview type to the
exclusive assessment by the attending nurse or caregiver. In
most cases, the resident’s capability to complete the MNA in a
reliable way will decrease gradually and so it will not be
possible to provide a clear threshold when the mode of
application should be changed. Therefore, we suggest that in
nursing homes with a high prevalence of dementia, the MNA
should routinely be completed by the attending nurses.
The inter-rater reliability of this mode of MNA application
has been tested in two studies (31, 35). It was proven moderate,
respectively good with kappa values of 0.53 and 0.78. The
results for several MNA sub-items were more diverse.
In the original publication on the MNA, all participants
classified as “normal” were still alive after one year of follow-
up, while 24% of the patients “at risk” and 48% of the
“malnourished” died during this period (36). When assessing
the ability of the full MNA to predict in-hospital and long-term
mortality in old hospitalised patients, Kagansky and co-workers
found a significant correlation between the full MNA and
survival rates after 2.7 years (30). Mortality in malnourished
participants was 3-fold higher than in well-nourished ones
(38.7% vs. 12.5%). Persson et al. (34) reported a 50% mortality
rate in malnourished elderly after one year. In the study by Tsai
et al. (11), it was shown that the MNA by caregiver’s
assessment predicted six-month follow-up mortality in
cognitively impaired elderly. The results of our study add to the
scarce longitudinal data currently available for MNA
application in the nursing home setting. In the present study,
association with mortality was highest in those categorised as
malnourished which could be shown for both approaches (table
5). Nevertheless, this relationship was statistically more
profound for the assessment by the nursing staff with a higher
positive predictive value. As malnutrition is closely related to
mortality among the elderly, this observation enforces that the
assessment by nursing staff is superior.
The lack of applicability of the MNA in residents with
percutaneous endoscopic gastrostomies (PEG) should not be
interpreted as a fault of this screening instrument as these
residents are commonly regarded as being at risk of
malnutrition in any case. As a consequence, they are supposed
to have a nutritional plan with fixed intakes and to be subjected
to regular monitoring of their nutritional status.
Conclusion
The applicability of the MNA in nursing homes with a high
percentage of cognitively impaired residents has to be
considered as limited if performed by one-on-one interviews.
As the risk of malnutrition is high in this sub-group and
nutritional screening in this population has therefore to be
regarded as mandatory, we suggest that the MNA should be
performed routinely by the nursing staff. In this regard, it has to
be considered that a valid MNA can only be provided by the
nursing staff when there is thorough knowledge of the different
aspects of the resident’s life. With regard to the use of the
MNA in scientific research, it has to be stressed that authors
should regularly provide information on the method used for
MNA application. This will be necessary to ensure the
comparability of epidemiological and interventional data in the
nursing home setting.
The present study was supported by a research grant of the Else Kröner-Fresenius-
Foundation, Bad Homburg.
Financial disclosure: None of the authors had any financial interest or support for this
paper.
References
1.Cederholm T, Jagren C, Hellstrom K: Outcome of protein-energy malnutrition in
elderly medical patients. Am J Med 1995;98:67-74.
Dey DK, Rothenberg E, Sundh V, Bosaeus I, Steen B: Body Mass Index, weight
change and mortality in the elderly. A 15 year longitudinal population study of 70 y
olds. Eur J Clin Nutr 2001;55:482-92.
Amarantos E, Martinez A, Dwyer J: Nutrition and Quality of Life in older Adults. J
Gerontol 2001;56A:54-64.
Kondrup J, Allison SP, Elia M, Vellas B, Plauth M.: ESPEN Guidelines for Nutrition
Screening 2002. Clin Nutr 2002;22(4):415-421.
Pauly L, Stehle P, Volkert D: Nutritional Situation of elderly nursing home residents.
Z Gerontol Geriat 2007;40: 3-12.
Bauer JM, Kaiser M, Anthony P, Guigoz Y, Sieber CC: The Mini Nutritional
Assessment-Its History, Today’s Practice, and Future Perspectives. Nutr Clin Pract
2008;23(4):388-96.
Morley JE and Thomas DR: Anorexia and aging: pathophysiology. Nutrition
1999;15:499-503.
Guigoz Y, Vellas B, Garry PJ: Assessing the Nutritional Status of the elderly. The
Mini Nutritional Assessment as Part of the Geriatric Evaluation. Nutrition reviews
1996;54:59-65.
Guigoz Y, Vellas B, Garry PJ: Mini Nutritional Assessment: a practical assessment
tool for grading the nutritional state of elderly patients. Facts Res Gerontol
1994;(2):15-60.
Vellas B, Garry PJ, Guigoz Y: Mini Nutritional Assessment (MNA): Research and
Practice in the Elderly. Nestle nutrition Workshop Series Clinical & Performance
Programme, Nestec Ltd; Vervey/S. Karger AG, Basel ©,1999;1:3-12.
Tsai A and Ku PY: Population-specific mini Nutritional assessment effectively
predicts the nutritional state and follow-up mortality of institutionalized elderly
Taiwanese regardless of cognitive status. Br Nutr 2007;1-7.
Wibky K, Ek AC, Cristensson L: The two-step Mini Nutritional Assessment
procedure in community resident homes. J Clin Nurs 2008,1211-1218.
Essed NH, van Staveren WA, Kok FJ, de Graaf C: No effect of 16 weeks flavor
enhancement on dietary intake and nutritional status of nursing home elderly.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
RESIDENT INTERVIEWS VERSUS ASSESSMENT BY NURSING STAFF
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
6
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 6

Page 7

Appetite 2007;48(1):29-36.
Souminen MH, Sandelin E, Soini H, Pitkala KH: How well do nurses recognize
malnutrition in elderly patients? Eur J Clin Nutr. Epub 2007;Sep 19.
Dion N, Cotart JL, Rabilloud M: Correction of nutrition test errors for more accurate
quantification of the link between dental health and malnutrition. Nutrition
2007;23(4):301-307.
Norman K, Smoliner C, Valentini L, Lochs H, Pirlich M: Is bioelectrical impedance
vector analysis of value in the elderly with malnutrition and impaired functionality?
Nutrition 2007;23(7-8):564-569.
Nijs KA, de Graaf C, Siebelink E et al.: Effect of family-style meals on energy intake
and risk of malnutrition Dutch nursing home residents: a randomized controlled trial.
J Gerontol A Biol Sci Med Sci 2006;61(9):935-942.
Soini H, Muurinen S, Routasalo P, et al.: Oral and nutritional status-is the MNA a
useful tool for dental clinics? J Nutr Health Aging 2006;10(6):495-499; discussion
500-501.
Wojszel ZB.: Determinants of nutritional status of older people in long-term care
settings on the sample of nursing homes in Bialystok. Adv Med Sci 2006;51:168-
173.
Bauer JM, Vogl T, Wicklein S, Trögner J, Mühlberg W, Sieber CC: Comparison of
the Mini Nutritional Assessment, Subjective Global Assessment, and Nutritional Risk
Screening (NRS 2002) for nutritional screening and assessment in geriatric hospital
patients. Z Gerontol Geriatr. 2005;38:322-327.
Folstein MF, Folstein S, McHugh PR: Mini Mental State: a practical method for
grading cognitive state of patients for he clinician. J Psychiatr Res 1975;12:189-198
Lesher EL and Berryhill JS: Validation of the Geriatric Depression Scale-short form
amoung inpatients. J Clin Psychol1994;50:256-60.
Arbeitsgemeinschaft für Klinische Ernährung. www.ake-nutrition.at: 22.07.2008
Chumlea W, Roche A, Steinbaugh M: estimating stature from knee height for
persons 60-90 years of age. J Am Geriatr Soc 1985;33:116-120.
Chumlea WW, Guo S, Wholihan K, Cockram D, Kuczmarski RJ, Johnson CL:
Stature predictions equations for elderly non-Hispanic white, non-Hispanic black and
Mexican American persons developed from NHANES-III data. J Amer dietet Assoc
1998;98:137-142.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26. Mahoney FI and Barthel DW: Functional Evaluation: The Barthel Index. Maryland
State Medical Journal 1965;14: 61-65.
Rubenstein LZ, Harker JO, Salva A, Guigoz Y, Vellas B: Screening for
undernutrition in geriatric practice: developing the short-form Mini-Nutritional
Assessment (MNA-SF). J Gerontol A Biol Sci Med Sci 2001;56:M366-M372.
Cohen A: Coefficient of agreement for nominal scales. Educ Psychol Meas
1960;19:3-11.
Hengstermann S, Nieczaj R, Steinhagen-Thiessen E, Schulz R-J: Which are the most
efficient items of the Mini Nutritional Assessment in multimorbid patients? J Nutr
Health Aging 2008;12(2):117-122.
Kagansky N, Berner Y, Koren-Morag N, Perelman L, Knobler H, Levy S: Poor
nutritional habits are predictors of poor outcome in very old hospitalized patients.
Am J Clin Nutr 2005;82:784-91.
Bleda MJ, Bolibar I, Parés R, Salvá A: Reliability of the Mini Nutritional Assessment
(MNA) in institutionalized elderly people, J Nutr Health Aging 2002;6(2):134-137.
Ödlund Olin A, Koochek A, Ljungqvist O, Cederholm T: Nutritional status, well-
being and functional ability in frail elderly service flat residents. Eur J Clin Nutr
2005;59:263-270.
Donini LM, de Felice MR, Tassi L, de Bernadini L, Pinto A, Giustini AM, Cannella
C: A “Proportional and objective score” for the Mini Nutritional Assessment in a
long term geriatric care. J Nutr Health Aging 2002;6(2):141-146.
Persson MD, Brismar KE, Katzarski KS, Nordenström J, Cederholm TE: Nutritional
status using mini nutritional assessment and subjective global assessment predict
mortality in geriatric patients. J Am Geriatr Soc 2002;50;1996-2002.
Neumann SA, Miller MD, Daniels LA, Ahern M, Crotty M: Mini Nutritional
Assessment in geriatric rehabilitation: Inter-rater reliability and relationship to body
composition and nutritional biochemistry 2007;64:179-185.
Guigoz Y and Vellas B: Test d’evaluation de l’etat nutritionnel de la personne agée:
le Mini Nutirtionel Assessment (MNA). Med Hyg. 1995;53:1965-1969.
Kulnik D, Elmadfa I: Assessment of the Nutritional Situation of elderly Nursing
home Residents in Vienna. Ann Nutr Metab 2008;52:51-53.
www.mna.elderly.com;10/2008.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
THE JOURNAL OF NUTRITION, HEALTH & AGING©
The Journal of Nutrition, Health & Aging©
Volume , Number , 2009
7
BAUER/c/PPPPP:04 LORD_c 8/10/09 11:47 Page 7