International Journal of Nursing Studies 50 (2013) 83–89

Contents lists available at SciVerse ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

A comparison of the full Mini Nutritional Assessment, short-form Mini

Nutritional Assessment, and Subjective Global Assessment to predict the
risk of protein-energy malnutrition in patients on peritoneal dialysis:
A cross-sectional study
Alan C. Tsai a,b,*, Jiun-Yi Wang a, Tsui-Lan Chang c, Tsz-Yan Li d
a
Department of Healthcare Administration, Asia University, Wufeng, Taichung, Taiwan
b
Department of Health Services, School of Public Health, China Medical University, Taichung, Taiwan
c
Nursing Department, Hsin Yung Ho Hospital, Taoyuan, Taiwan
d
Department of Nursing, Tungs’ Taichung MetroHarbor Hospital, Wuchi, Taichung, Taiwan

A R T I C L E I N F O A B S T R A C T

Article history: Background: The full Mini Nutritional Assessment (full-MNA) and short-form MNA (MNA-
Received 1 January 2012 SF) are simple and effective nutrition screening scales, but their usefulness for identifying
Received in revised form 23 August 2012 patients with peritoneal dialysis (PD) at risk of protein-calorie malnutrition (PEM) has not
Accepted 26 August 2012 been investigated.
Objectives: This study was aimed to investigate the convergent validity of the full-MNA
Keywords: and MNA-SF for identifying patients with PD at risk of PEM.
Peritoneal dialysis Design: A cross-sectional study.
Nutritional screening
Setting: A hospital-managed dialysis center.
Mini Nutritional Assessment
Participants: 80 adult ambulatory PD patients.
Subjective Global Assessment
Methods: Patients were interviewed for personal data and rated with the full-MNA, MNA-
SF and the Subjective Global Assessment (SGA) for nutritional status. The consistency
among the scales was assessed with kappa coefficients. The ability of each scale to
differentiate undernutrition was evaluated with external standards including serum
albumin and creatinine concentrations, mid-arm and calf circumferences, and dialysis-
related indicators. Statistical significance was evaluated with Wilcoxon rank-sum test.
Results: The full-MNA and MNA-SF showed low agreements with the SGA (kappa = 0.346
and 0.185, respectively). The full-MNA and MNA-SF performed better than the SGA in
differentiating undernutrition according to the external standards. However, contrary to
general expectation, MNA-SF rated a significantly smaller proportion of subjects at risk of
undernutrition.
Conclusion: The full-MNA and MNA-SF are more able than the SGA in identifying PD
patients at risk of PEM. However, MNA-SF rates a smaller proportion of PD patients at risk
of undernutrition than the full-MNA. The use of MNA-SF as a stand-alone unit requires
further confirmation.
ß 2012 Elsevier Ltd. All rights reserved.

What is already known about the topic?

 Patients on dialysis are at an increased risk of protein-

* Corresponding author at: 3411 E. Dobson Place, Ann Arbor, MI 48105,
USA. Tel.: +1 734 761 2468; fax: +1 734 761 2468. energy malnutrition. Early identification of the risk is
E-mail address: atsai@umich.edu (A.C. Tsai). important for timely intervention.

0020-7489/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijnurstu.2012.08.015
84 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89
 The Subjective Global Assessment (SGA) is a tool
recommended by K/DOQI for identifying those who
are at risk of protein-energy-malnutrition (PEM).
 Other simple and non-invasive tools, such as the mini-
nutritional assessment (MNA) may also be appropriate
for rating the risk but have not been evaluated.
What this paper adds
 Predictions made with the SGA, the full-MNA and the
short-form-MNA do not agree well to each other.
 According to external standards including a list of
biochemical, anthropometric, health- and dialysis-
related indicators, the full-MNA and the MNA-SF were
more able to differentiate malnutrition in PD patients at
risk of PEM than the SGA.
 The full-MNA appears appropriate for rating the risk of
PEM in PD patients. The use of MNA-SF as a stand-alone
unit requires further confirmation because it rates a
smaller proportion of patients at risk of PEM than the
full-MNA.
1. Introduction
Protein-energy malnutrition (PEM, inadequate intake
of total energy and protein) or undernutrition is
prevalent and associated with increased risk of mortality
in patients with end stage renal disease (ESRD)
(CANADA-USA, 1996; Chung et al., 2000; Lacquaniti
et al., 2009). In these patients, dietary restrictions,
altered sensory functions, drug–nutrient interactions
and uremic conditions further contribute to poor
appetite and aggravate the risk of PEM. In maintenance
peritoneal dialysis (PD) patients, co-morbid diseases are
also believed to contribute to PEM, hypoalbuminemia,
muscle wasting and higher mortality (Chung et al.,
2003). Thus, in these patients, early identification of
those who are at risk of PEM may help prevent severe
malnutrition and muscle wasting. A simple, reliable and
easy-to-use tool is vitally important.
Nutritional status of dialysis patients can be assessed
with serum biochemical indicators, anthropometry, and
health- or dialysis-related indicators (National Kidney
Foundation, 2000). However, none of these parameters is
reliable enough to be universally accepted as a gold
standard. Instead, the nutritional status of ESRD patients is
generally rated with multiple indicators or scales that
consist of multi-dimensional nutritional and health
indicators. The U.S. National Kidney Foundation Kidney
Disease/Dialysis Outcomes and Quality Initiative (K/DOQI)
(National Kidney Foundation, 2000) and the Taiwan
Society of Nephrology (2004) recommend the Subjective
Global Assessment (SGA) as a regular tool for assessing
PEM in the adult dialysis population. The SGA determines
whether nutrient assimilation has been restricted because
of decreased food intake or poor digestion and absorption,
whether malnutrition has affected the function of organs,
and whether the patients’ disease process influences
nutrient requirements (National Kidney Foundation,
2000). However, there are other simple and effective tools
worth to be evaluated.
The Mini Nutritional Assessment (MNA), a widely used
geriatric nutritional screening/assessment scale, is well
known for its simplicity, portability, reliability and non-
invasiveness. It has been shown to be effective for
screening/identifying persons at risk of undernutrition
living in a variety of settings (Cabrera et al., 2007; Guigoz
et al., 2002; Tsai et al., 2010a) or having various health
conditions (Kuzuya et al., 2005; Tsai et al., 2009a) or
chronic diseases (Read et al., 2005; Tsai et al., 2009b,
2010b). The MNA has been shown to be appropriate for
rating the risk of undernutrition in hemodialysis (HD)
patients (Tsai and Chang, 2011; Tsai et al., 2009b, 2011),
but its use in PD patients is largely unexplored. The PD
patients are a select subgroup of the ESRD population. They
are usually younger and less likely to have primary
diagnosis of diabetes or hypertension (Flanigan et al.,
2001). However, they are also subject to nutritional risks.
Hence, the aim of the present study was to investigate the
convergent validity of the MNA, using biochemical,
anthropometric and dialysis-related measures as criteria
for comparison.
2. Methods
2.1. Design and sampling
We conducted a cross-sectional study and recruited
subjects from PD patients of a hospital-managed dialysis
center in central Taiwan. A clinical nurse at the center
(coauthor TYL) approached each patient during a routine
treatment visit, explained the study protocol and asked
him/her for voluntary participation. Adult PD patients
who had been patients of the center for three months or
longer, and were without acute disease or infection and
able to communicate verbally were qualified to partici-
pate. Among the 90 patients at the center, 80 met the
selection criteria and agreed to participate in the study;
8 were unable to communicate or had acute disease/
infection at the time of admission; and 2 chose not to
participate. All subjects had been patients of the center
for more than 90 days (a period for a new PD patient to
be stabilized).
The required sample size was estimated by the
formula used for planning studies in two-sample
comparisons: N = [(k + 1)2/k]  s2  [Z1 a/2 + Z1 b]2)/d2,
where k is the proportion of the two sample sizes, s2
is the common variance and d is the mean difference
between two samples (Rosner, 2006). We assumed that
the prevalence of undernutrition was 20% in PD patients
(k = 4), the mean difference in serum albumin between
two samples was d = 0.3 g/dL and common variance
s2 = (0.43)2. Under significance level a = 0.05 and 80%
power, the required sample size would be 101. For the
recruited sample size of 80, it could achieve around
70% power.
All those who qualified and agreed to participate signed
an informed consent prior to admission. The study protocol
was approved by the Institutional Review Board of the
hospital. Ethics guidelines and subjects’ confidentiality
were observed throughout the study. The study took place
during September to December 2010.
 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89 85

2.2. Assessment tools under investigation
Nutritional status was evaluated with the full-MNA, the
short-form MNA-Taiwan version-1 (MNA-T1-SF) and the
SGA.
2.2.1. Mini Nutritional Assessment (MNA)
The full-MNA consists of 18 items and evaluates
dietary, anthropometric, global and self-viewed aspects
of nutrition (Vellas et al., 1999). MNA-SF consists of the
first 6 items of the full-MNA and supposedly serves as a
pre-screening unit of the full-MNA (Rubenstein et al.,
2001) (Table 1). Individuals who are suspected of under-
nutrition are first evaluated with the SF and those who are
identified as possible undernutrition are further evalu-
ated with the second part of the MNA (12 additional items)
to confirm the diagnosis. MNA-SF generally rates a greater
proportion of subjects as at risk of undernutrition than the
full MNA. MNA-SF was found to have 90% agreement with
the full-MNA, 97.9% sensitivity, 100% specificity, and
98.7% diagnostic accuracy for predicting undernutrition
according to a study conducted in 881 elderly and using a
cutoff of >11 as normal nutrition (Rubenstein et al., 2001).
Because of high consistency with the full scale, MNA-SF
has been proposed to function as a stand-alone unit
(Kaiser et al., 2009).
The present study used MNA-T1 to rate the nutri-
tional status of the PD subjects. MNA-T1 is a normalized
version of the MNA for the Taiwanese by adopting
population-specific anthropometric cutoff points (Tsai
et al., 2007). In order to assure accuracy, the clinical
nurse (TYL) who had prior experience using the SGA
was trained by a clinical dietitian to rate the risk of
undernutrition of PD patients with the MNA. After
evaluating the first 26 patients of the study sample, good
consistency was established between the dietician and
the researcher (kappa = 0.806). The researcher rated the
rest of the patients independently. A final full-MNA score
23.50 or MNA-SF score 11 was considered as at risk of
undernutrition (Rubenstein et al., 2001).
2.2.2. Subjective Global Assessment (SGA)
We used a 7-point scale SGA modified from the original
Detsky’s version (Detsky et al., 1987; CANUSA, 1996). A
clinical nurse (TYL) who was well experienced in using the
SGA first rated each patient with the 7-point scale. After
considering the scores of all items and the past clinical
history of the patient, she then subjectively assigned an
overall SGA score of 1–7 to each subject to indicate the
overall nutritional status. An overall score of 1–2 suggests
severely malnourished; 3, 4, or 5, mild-moderate mal-
nourishment; and 6 or 7, very mild risk to well-nourished
(National Kidney Foundation, 2000) (Table 1). The 7-point
scale SGA has been shown to be reliable and valid for
nutritional assessment in adults on dialysis (Enia et al.,
1993). The SGA has also been shown to have fair reliability
(the intra-class correlation was 0.72 for inter-observer
reliability and 0.88 for intra-observer reliability) (Visser
et al., 1999). Statistically significant differences in mean
BMI and serum albumin across categories of SGA (all
p < 0.05) suggest good validity of the SGA (Steiber et al.,
2004, 2007).

Table 1
Content items and item scores of the full MNA, MNA-SF and the SGA.

Items Full-MNA MNA-T1-SF SGAa

A. Food intake status 2b 2b H

B. Weight loss 3 3 H
C. Mobility 2 2 H
D. Psychological stress or acute disease 2 2
E. Neuropsychological problems 2 2
F. Body mass index (BMI) 3 3
G. Able to live independently 1
H. Taking >3 prescribed drugs/day 1
I. Having pressure sores or skin ulcers 1
J. Full meals/day 2
K. Protein-rich foods 1
L. Fruit & vegetable intake 1
M. Water/fluid consumed/day 1
N. Mode of feeding 2
O. Self-view of nutritional status 2
P. Self-view of health status 2
Q. Mid-arm circumference 1
R. Calf circumference 1
S. Gastrointestinal symptoms H
T. Disease & its relation to nutritional requirements H
U. Physical loss of subcutaneous fat, muscle wasting, H
ankle edema, sacral edema, and ascites

Total or maximum score 30 14 SGA scorec

MNA, Mini Nutritional Assessment; SF, short-form; SGA, Subjective Global Assessment.
a
The status of each item was evaluated with a 3-level (A, B, C) rating.
b
Assigned maximum item score.
c
After evaluating above items, the rater then subjectively assign an overall SGA score of 1–7 to each subject. An overall score of 1–2 suggests severely
malnourished; 3, 4, or 5, mild-moderate malnourishment; and 6 or 7, very mild risk to well-nourished.
86 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89

2.3. External standards Table 2

Characteristics of subjects (N = 80).

Because of the lack of a universally accepted gold Item (n, %) Mean  SD

standard for grading the nutritional status of dialysis Men 35 (43.8)
patients, we used a batch of indicators of clinical relevance Women 45 (56.2)
as external standards. Biochemical indicators included Age (years) 52.9  14.4
serum albumin and creatinine. Anthropometric indicators Years on dialysis
<1 9 (11.3)
included MAC and CC. Dialysis-related indicators included
2–3 36 (45.0)
urea clearance (Kt/Vurea), weekly creatinine clearance rate 4–5 20 (25.0)
(wCCr), and protein equivalent of nitrogen appearance >6 15 (18.8)
(nPNA). Kt/Vurea (target value is 2.0/week or higher) and Body mass index (kg/m2) 24.1  3.7
Serum albumin (g/dL) 3.7  0.4
wCCr (recommended minimum value is 50 L/week) are
<3.5 21 (26.3)
dialysis performance indicators. nPNA (body size adjusted) Serum creatinine (mg/dL) 11.6  3.6
is a measure of net protein catabolism. It is a useful clinical <10 29 (36.3)
measure of nutrition that correlates with patient outcome. Serum total cholesterol (mg/dL)
In dialysis patients, the recommended level of nPNA is 1.0– <200 43 (55.1)
Dialysis-related parameters
1.2 g/kg body weight/day of protein intake, an intake of
Dialysis solution (mL/day) 8848  1552
<0.8 g/kg/day is considered malnutrition (National Kidney Kt/V urea 2.2  0.5
Foundation, 2000). In patients who are in nitrogen balance, <2.0a 28 (35.9)
nPNA is a reasonable estimate of dietary protein intake. We wCCr (L/week/1.73 m2)b 63.1  18.9
nPCR (g/kg/day) 0.98  0.26
compared the ability of the full-MNA, MNA-SF and the SGA
<0.8c 19 (24.4)
in differentiating the nutritional risk based on these Urine volume (mL/day) 407  472
external indicators. All these data were extracted from
Kt/V urea, urea clearance rate; wCCr, weekly creatinine clearance; nPCR,
patients’ routine records maintained at the center and
standardized protein catabolic rate.
updated every 3 months. a
In peritoneal dialysis the target is 2.0/week (National Kidney
Foundation, 2000).
b
2.4. Procedure and outcome measurements The recommended minimum wCCr is 50 L/week (National Kidney
Foundation, 2000).
c
Most guidelines specify maintaining the protein intake above 1.0–
The interviews were arranged to take place within the 1.2 g/kg/day in dialysis patients, with values less than 0.8 g/kg/day being
same week of the lab test which took place every 3 months equated with malnutrition (National Kidney Foundation, 2000).
for all PD patients at the center. Nutritional assessments
with the two scales took place in two separate occasions,
approximately 10 days apart and the sequence was by 2 had one or more comorbidities; average serum albumin
random. Computation of results was done only after both was 3.7 g/dL; and average creatinine was 11.6 mg/dL.
assessments were completed for each person. The same Table 3 shows the distribution of nutritional status
researcher conducted all interviews, measurements and rated with the MNA or the SGA. The SGA rated 1 person as
data collection. The predictive ability of each scale was malnourished, 12 persons (15%) as at risk of undernutrition
evaluated by comparing the ability to differentiate under- and 67 persons as normal; the full-MNA rated 1 person as
nutrition based on a batch of biochemical, anthropometric malnourished, 16 (20%) as at risk of undernutrition and 63
and dialysis-related variables (as external standards). as normal, whereas MNA-SF rated 1 person as malnour-
Height and weight (for computing body mass index, ished, 7 (8.7%) as at risk and 72 as normal.
BMI), mid-arm circumference (MAC) and calf circumfer- Table 4 shows the consistency among the scales.
ence (CC) were measured according to methods described The full-MNA showed a moderately low consistency
by Lee and Nieman (2003). with the SGA (kappa 0.346), where as MNA-SF showed a
low consistency with the SGA (kappa 0.185). The
2.5. Statistical analysis consistency between the two MNA forms was also
moderate (kappa 0.491).
Results were analyzed with Statistical Package for the Table 5 shows the ability of the scales to differentiate
Social Sciences, SPSS 15.0 (SPSS Inc., Chicago, IL). The patients’ nutritional status based on the seven indicators.
characteristics of subjects were computed with descriptive
statistics. Kappa statistic was used to determine the Table 3
consistency among the scales. The means of the biochem- Nutritional status (n, %) of 80 patients on peritoneal dialysis classified
ical and health- or dialysis-related indicators (external with the full-MNA, the short-form MNA and the SGA.
standards) were compared by Wilcoxon rank-sum test. Tool Malnourished At risk Normal
Significance level was set at alpha = 0.05.
Full-MNAa 1 (1.3) 16 (20.0) 63 (78.7)
MNA-SFa 1 (1.3) 7 (8.7) 72 (90.0)
3. Results SGA 1 (1.3) 12 (15.0) 67 (83.7)

SGA, Subjective Global Assessment; MNA, Mini Nutritional Assessment;

Table 2 shows the characteristics of subjects. Among 80 SF, short-form.
subjects (35 men and 45 women, 22–84 years old), 5 were a
A normalized version of the MNA that adopted the Taiwanese-specific
underweight, 75 were normal or overweight/obese; all but anthropometric cutoffs.
 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89 87

Table 4
Binary cross-tabulation test of the nutritional status rated with the full MNA, the short-form MNA and the SGA (N = 80).

SGA Full-MNA

Undernutr. Normal Undernutr. Normal

Full-MNAa
Undernutrition 7 10
Normal 6 57
Kappa (95% CI) 0.346 (0.093–0.599)**
MNA-SFa
Undernutrition 3 5 7 1
Normal 10 62 10 62
Kappa (95% CI) 0.185 ( 0.087 to 0.457) 0.491 (0.244–0.738)***

SGA, Subjective Global Assessment; MNA, Mini Nutritional Assessment; SF, short-form; Undernutr., undernutrition (malnourished + at risk of
malnutrition).
a
A normalized version of the MNA that adopted the Taiwanese-specific anthropometric cutoffs.
** p<0.01.
*** p<0.001.

The full-MNA and MNA-SF, but not the SGA showed good
convergent validity and significantly differentiated serum
albumin, serum creatinine, MAC and CC (all p < 0.05).
However, all three scales were not effective in differentiat-
ing the dialysis-related indicators including Kt/Vurea, wCCr
and nPNA.
4. Discussion
The SGA is a tool recommended by the U.S. K/DOQI
(National Kidney Foundation, 2000, 2006) and the Taiwan
Society of Nephrology (2004) for assessing PEM in the
adult dialysis population. However, when we tested the
MNAs and the SGA against the same set of external
standards which are also among those indicators recom-
mended for assessing the nutritional status of the PD
patients, we found that the MNAs but not the SGA showed
good convergent validity.
Results showed that the full-MNA and the MNA-SF have
better ability than the SGA in differentiating patients’
serum albumin, serum creatinine, mid-arm circumference
and calf circumference, but all the three tools were not
effective in differentiating patients in terms of their Kt/
Vurea-, wCCr- and nPNA-readings. Insufficient sample size
could be part of the reason but the major reason could be
that these dialysis-related indicators are less influenced by
nutritional status than by dialysis dose or adequacy of
dialysis. Perhaps, these results may be viewed as the
divergent validity of the MNA.
There are several potential reasons for the SGA to
function not as well as the MNA in predicting the
nutritional status in these PD patients. (a) The SGA and
the MNA are both global nutritional assessment tools, but
they were developed for different target populations. The
SGA was originally developed for predicting nutrition-
related complications in patients undergoing gastrointest-
inal surgery (Detsky et al., 1987), whereas the MNA was
developed specifically for evaluating the risk of malnutri-
tion in frail elderly (Guigoz et al., 1994). (b) Both tools
evaluate appetite/food intake, weight status, weight loss
and functional capacity/mobility, but they differ in other
aspects of evaluation. The SGA emphasizes clinical
evaluations in gastrointestinal symptoms, diseases and
comorbidities as related to nutritional needs, and physical

Table 5
Means and (SD) of serum biochemical, anthropometric, number of comorbidity and dialysis-related indicators stratified by nutritional status rated with the
SGA, the full-MNA and the short-form-MNA in patients on peritoneal dialysis (N = 80).

Item SGA Full-MNA MNA-SF

Undernutr. Normal Undernutr. Normal Undernutr. Normal

Serum indicators
Albumin (g/dL) 3.51 (0.57) 3.72 (0.41) 3.44 (0.49) 3.75 (0.40)**,a 3.31 (0.48) 3.72 (0.42)*
Creatinine (mg/dL) 11.17 (3.79) 11.67 (3.56) 9.59 (2.91) 12.13 (3.57)** 8.70 (2.92) 11.91 (3.52)*
Anthropometrics
MAC (cm) 26.3 (4.7) 27.3 (3.3) 25.6 (4.2) 27.6 (3.3)* 24.3 (4.0) 27.5 (3.4)*
CC (cm) 33.4 (4.7) 34.1 (4.1) 30.9 (4.9) 34.8 (3.5)*** 27.9 (4.1) 34.6 (3.6)***
Dialysis-related indicators
Kt/V urea 2.01 (0.45) 2.19 (0.54) 1.99 (0.46) 2.21 (0.53) 1.84 (0.61) 2.19 (0.51)
wCCr (L/week/1.73 m2) 59.47 (11.69) 63.81 (19.95) 58.97 (14.56) 64.22 (19.85) 57.58 (18.68) 63.69 (18.99)
nPCR (g/kg/day) 1.02 (0.30) 0.98 (0.26) 0.94 (0.30) 1.00 (0.26) 0.94 (0.28) 0.99 (0.26)

SGA, Subjective Global Assessment; MNA, Mini Nutritional Assessment; SF, short-form; Undernutr., undernutrition; TIBC, total iron-binding capacity; BMI,
body mass index; MAC, mid-arm circumference; CC, calf-circumference; Kt/V urea, urea clearance rate; wCCr, weekly creatinine clearance; nPCR,
standardized protein catabolic rate.
a
Significant differences between the ‘‘normal’’ and ‘‘undernutrition’’ values on the basis of Wilcoxon rank-sum test.
* p < 0.05.
** p < 0.01.
*** p < 0.001.
88 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89
appearance such as loss of subcutaneous fat, muscle
wasting and edema, whereas the MNA evaluates intakes of
protein foods and fluid, medication, psychological health,
neuropsychological problems and self-rated nutritional
and health status (Table 1). (c) The MNA items are based
mostly on measured values or well-defined criteria with a
weighted scoring system and well-defined cutoffs to define
the overall nutritional status of patients. The SGA
emphasizes subjectivity. Although it uses a 7-point scale
scoring system, the overall rating is assigned subjectively
by the rater after considering the item-scores and patient’s
clinical history.
The SGA has been observed to exhibit large inter-rater
variability. It was effective in differentiating the severely
malnourished patients from the nutritionally normal
individuals but not effective in differentiating the degree
of malnutrition in ESRD patients (Cooper et al., 2002). A
study that evaluated the two tools in elderly newly
admitted to municipal care found that the SGA was more
useful in detecting residents with established malnutrition
while the MNA was more useful in detecting residents who
needed preventive nutritional measures (Christensson
et al., 2002). Thus, the predictive ability of a tool is also
dependent on the nutritional condition of the study
sample. These differences in the property, structure,
method of rating and target population could affect the
differentiating ability of the tools. MNA-SF, which contains
the 6 key items of the full-MNA, predicts the full-MNA
well. Its predictive ability usually closely follows that of the
full-MNA.
The MNA has recently been evaluated for its usefulness
in rating the risk of undernutrition in HD patients (Afsar
et al., 2006; Tsai et al., 2011), and to the best of our
knowledge, the present study is the first time that it is
evaluated for rating the risk of undernutrition in PD
patients. Results of the present study suggest that the MNA
(both the full scale and the short-form) predicts the risk of
undernutrition more effectively than the SGA. However, it
should be mentioned that similar to our observation in HD
patients (Tsai et al., 2011), MNA-SF rates a smaller
proportion of PD patients as at risk of undernutrition
compared to the full-MNA (10% vs. 21.3%). This is not
consistent with the functioning principle of the MNA. As a
prescreening tool, MNA-SF is expected to predict a larger
proportion of patients as at risk of undernutrition than the
full-MNA. Whether this discrepancy is due to under-
estimation by the SF or over-estimation by the full-scale is
difficult to pinpoint. However, since the risk of under-
rating out-weighs the extra cost of confirming a diagnosis
of undernutrition, to be prudent, the full-MNA should be
used to avoid under-diagnosis.
The predictive ability of nutritional assessment tools is
relatively infrequently compared in PD patients. Results of
the present study suggest that the MNA may be a useful
tool for identifying PD patients at risk of undernutrition.
ESRD patients are at increased risk of PEM and premature
mortality. The availability of a simple, easy-to-use and
non-invasive tool may enhance routine screening and
early detection of emerging PEM in these patients.
However, the discrepancy between the full-MNA and
MNA-SF needs to be resolved. The impact of use of either of
these tools on nutritional status and health in PD patients
also needs to be evaluated.
The study has some limitations. (a) Sample size is
small relative to the observed prevalence of under-
nutrition. We had assumed the prevalence would have
been in the range of 20%, but it turned out to be lower.
Inadequate sample size could lead to biased result. Thus,
results should be interpreted with caution. Confirmation
of our finding with studies conducted with larger
samples is desired. (b) Subjects are recruited from one
dialysis center. It is possible that the study sample
may not represent the entire spectrum of PD patient
population in Taiwan. (c) Uremic conditions and
medication can disproportionately impact the scoring
of some scales. (d) Finally, given the difficulties to define
an optimal reference standard for measuring PEM, the
usefulness of the MNA in rating the nutritional risk of PD
patients needs to be further confirmed. Controlled trials
comparing use of the MNA to current practice in terms of
their impact on nutrition-related health care and health
outcomes in PD patients would allow more robust
conclusions about likely beneficial effects and may
provide reliable estimates.
5. Conclusions
Results of the present study suggest that the MNA may
be appropriate for rating PD patients at risk of PEM. Both
the full-MNA and MNA-SF appeared more able than the
SGA in differentiating undernutrition in PD patients.
However, because of limited sample size, further con-
firmation of results is desired. The impact of use of either of
these tools on nutritional status and health in PD patients
also needs to be evaluated. The use of MNA-SF as a stand-
alone unit requires further confirmation because it may
under-rate the risk of PEM in PD patients.
Acknowledgments
The authors wish to thank the hospital and the dialysis
center administrators and nursing staff for permitting and
assisting the interviews, and the participants for their
cooperation during the course of this study.
Conflicts of interest
None declared.
Funding
The present study received no specific grant from any
funding agency in the public, commercial or not-for-profit
sectors.
Ethical approval
Tungs’ Taichung MetroHarbor Hospital Institutional
Review Board.
References
Afsar, B., Sezer, S., Arat, Z., et al., 2006. Reliability of mini nutritional
assessment in hemodialysis compared with subjective global assess-
ment. Journal of Renal Nutrition 16, 277–282.
Cabrera, M.A., Mesas, A.E., Garcia, A.R., et al., 2007. Malnutrition and
depression among community-dwelling elderly people. Journal of
American Medical Directors Association 8, 582–584.
 A.C. Tsai et al. / International Journal of Nursing Studies 50 (2013) 83–89
CANADA-USA (CANUSA) Peritoneal Dialysis Study Group, 1996. Adequacy
of dialysis and nutrition in continuous peritoneal dialysis: association
with clinical outcome. Journal of the American Society of Nephrology
7, 198–207.
Christensson, L., Unosson, M., Ek, A.-C., 2002. Evaluation of nutritional
assessment techniques in elderly people newly admitted to muni-
ciple care. European Journal of Clinical Nutrition 56, 810–818.
Chung, S.H., Lindholm, B., Lee, H.B., 2000. Influence of initial nutritional
status on continuous ambulatory peritoneal dialysis patient survival.
Peritoneal Dialysis International 20, 19–26.
Chung, S.H., Lindholm, B., Lee, H.B., 2003. Is malnutrition an independent
predictor of mortality in peritoneal dialysis patients? Nephrology
Dialysis Transplantation 18, 2134–2140.
Cooper, B.A., Bartlert, L.H., Aslani, A., et al., 2002. Validity of subjective
global assessment in chroni renal failure. American Journal of Kidney
Disease 40, 126–132.
Detsky, A.S., McLaughlin, J.R., Baker, J.P., et al., 1987. What is subjective
global assessment? Journal of Parenteral and Enteral Nutrition 11, 8–13.
Enia, G., Sicuso, C., Alati, G., et al., 1993. Subjective global assessment of
nutrition in dialysis patients. Nephrology Dialysis Transplantation 8,
1094–1098.
Flanigan, M.J., Rocco, M.V., Prowant, B., et al., 2001. Clinical performance
measures: the changing status of peritoneal dialysis. Kidney Inter-
national 60, 2377–2384.
Guigoz, Y., Vellas, B.J., Garry, P.J., 1994. The Mini Nutritional Assessment
(MNA): a practical assessment tool for grading the nutritional state of
elderly patients. In: Vellas, B.J., Guigoz, Y., Garry, P.J., Albarede, J.L.
(Eds.), Facts and Research in Gerontology. Serdi Publishing Co., New
York, NU, (Supplement: Nutrition), pp. 15–60.
Guigoz, Y., Lauque, S., Vellas, B.J., 2002. Identifying the elderly at risk for
malnutrition. The mini nutritional assessment. Clinics in Geriatric
Medicine 18, 737–757.
Kaiser, M.J., Bauer, J.M., Rämsch, C., et al., 2009. Validation of the Mini
Nutritional Assessment1 Short-Form (MNA-SF): a practical tool for
identification of nutritional status. Journal of Nutrition, Health &
Aging 13 (9), 782–788.
Kuzuya, M., Kanda, S., Koike, T., et al., 2005. Evaluation of mini-nutritional
assessment for Japanese frail elderly. Nutrition 21, 498–503.
Lacquaniti, A., Bolignano, D., Campo, S., et al., 2009. Malnutrition in the
elderly patient on dialysis. Renal Failure 31, 239–245.
Lee, R.D., Nieman, D.C., 2003. Chapter 7: Assessment of the hospitalized
patient. In: Nutritional Assessment, 3rd ed. McGraw Hill, New York,
NY, pp. 216–250.
National Kidney Foundation (K/DOQI), 2000. Clinical practice guidelines
for nutrition in chronic renal failure. American Journal of Kidney
Disease 35 (Suppl. 2), S1–S140 (Erratum in: American Journal of
Kidney Disease 2001; 38:917).
National Kidney Foundation (K/DOQI), 2006. Clinical Practice Guidelines
for peritoneal dialysis Adequacy. Update 2006. http://www.kid-
ney.org/professionals/kdoqi/guideline_uphd_pd_va/index.htm
(accessed on 19.12.11).
Read, J.A., Crockett, N., Volker, D.H., et al., 2005. Nutritional assessment in
cancer: comparing the mini-nutritional assessment (MNA) with the
89
scored patient-generated subjective global assessment (PGSGA).
Nutrition and Cancer 53, 51–56.
Rosner, B., 2006. Fundamentals of Biostatistics, 6th ed. Thomson, Bel-
mont, CA.
Rubenstein, L.Z., Harker, J.O., Salva, A., et al., 2001. Screening for under-
nutrition in geriatric practice: developing the short-form mini-nutri-
tional assessment (MNA-SF). Journals of Gerontology Series A:
Medical Sciences 56A, M366–M372.
Steiber, A.L., Kalantar-Zadeh, K., Secker, D., et al., 2004. Subjective global
assessment in chronic kidney disease: a review. Journal of Renal
Nutrition 14 (4), 191–200.
Steiber, A., Leon, J.B., Secker, D., et al., 2007. Multicenter study of
the validity and reliability of subjective global assessment in
the hemodialysis population. Journal of Renal Nutrition 17 (5),
336–342.
Taiwan Society of Nephrology, . Clinical Guidelines for Hemodialysis in
Taiwan. http://www.tsn.org.tw/UI/F/F005.aspx (accessed 1.06.11).
Tsai, A.C., Ho, C.S., Chang, M.C., 2007. Population-specific anthropometric
cut-points improve the functionality of the mini nutritional assess-
ment (MNA) in elderly Taiwanese. Asia Pacific Journal of Clinical
Nutrition 16, 656–662.
Tsai, A.C., Chou, Y.T., Chang, T.L., et al., 2009a. A modified mini nutritional
assessment without BMI can effectively assess the nutritional
status of neuropsychiatric patients. Journal of Clinical Nursing 18,
1916–1922.
Tsai, A.C., Lu, S.J., Chang, T.L., 2009b. Modified mini-nutritional assess-
ment can effectively assess the nutritional status of patients on
hemodialysis. Journal of Renal Nutrition 19, 380–388.
Tsai, A.C., Chang, T.L., Wang, Y.C., et al., 2010a. Population-specific short-
form mini nutritional assessment with BMI or calf-circumference can
predict risk of malnutrition in community-living or institutionalized
elderly in Taiwan. Journal of the American Dietetic Association 110,
1328–1334.
Tsai, A.C., Chang, T.L., Yang, T.-W., et al., 2010b. A modified mini nutri-
tional assessment without BMI predicts nutritional status of com-
munity-living elderly in Taiwan. Journal of Nutrition, Health & Aging
14 (3), 183–189.
Tsai, H.J., Tsai, A.C., Hung, S.Y., Chang, M.Y., 2011. Comparing the pre-
dictive ability of population-specific Mini-Nutritional Assessment
with Subjective Global Assessment for Taiwanese patients with
hemodialysis: a cross-sectional study. International Journal of Nur-
sing Studies 48, 326–332.
Tsai, A.C., Chang, M.Z., 2011. Long-form but not short-form mini-nutri-
tional assessment is appropriate for grading nutritional risk of
patients on hemodialysis—a cross-sectional study. International Jour-
nal of Nursing Studies 48, 1429–1435.
Vellas, B., Guigoz, Y., Garry, P.J., et al., 1999. The mini nutritional
assessment (MNA) and its use in grading the nutritional state of
the elderly patients. Nutrition 15, 116–122.
Visser, R., Dekker, F.W., Boeschoten, E.W., Stevens, P., Krediet, R.T., 1999.
Reliability of the 7-point subjective global assessment scale in asses-
sing nutritional status of dialysis patients. Advances in Peritoneal
Dialysis 15, 222–225.