Is it possible using handgrip strength instead of body mass index in MNA-SF test to assess the nutritional status of geriatric patients?

Abstract

Introduction

In Mini-Nutritional Assessment-Short Form (MNA-SF) test, a practical and reliable alternative parameter is still necessary for patients with difficult body mass index evaluation. We aimed to show whether or not handgrip strength may be used instead of body mass index (BMI) in MNA-SF test.

Materials and Methods

MNA-SF test scores, calf circumferences (CC), handgrip strength (HGS), and BMI of 191 patients were evaluated. The first one of calculated MNA-SF tests was with BMI, the second one with CC, and the last one with HGS. Zero point was given if CC was <31 cm and 3 points were given if CC was ≥31 cm. Zero, 1, 2, and 3 points were given if the loss of HGS when compared to expected HGS were ≥%60, from ≥%30 to <%60, from ≥%10 to <%30, and <%10 or greater than expected HGS, respectively. MNA-SF scores and nutritional status according to these three measures were compared.

Results

Mean age and median MNA-SF scores of the patients were 75±7.6 years and 12 points (min-max: 0-14) respectively. There were strongly positive correlations between MNA-SF scores with BMI and CC, with BMI and HGS, and with CC and HGS (r=0.938 p<0.001, r=0.938 p<0.001, r=0.914 p<0.001, respectively). Substantial agreement in nutritional status of the patients were seen between MNA-SF groups with BMI and CC, with CC and HGS, and with BMI and HGS (kappa: 0.795 p<0.001, kappa: 0.709 p<0.001, and kappa: 0.760 p<0.001, respectively).

Conclusions

HGS might be considered instead of BMI in MNA-SF test to assess nutritional status of geriatric patients.

Introduction

Malnutrition is a major problem for the elderly people (1). Its association with poor clinical outcomes indicates that this situation should be considered for the evaluation of all elderly patients routinely. To assess the malnutrition, there are a number of tests available in the literature (2., 3., 4.). While some of them are easy and practical test, the others may be too complicated and time-consuming to perform in routine clinical practice.

The most commonly used test for the assessment of malnutrition in geriatric patients in daily practice is the MNA-SF test (5., 6., 7.). One of the domains of the MNA-SF test is BMI. If it is not possible to assess the BMI of the patients, the well validated parameter instead of BMI is CC (5). It is recommended to give 0 point if the CC is lower than 31 cm and 3 points if the CC is 31 cm or greater. There are some handicaps of using CC instead of BMI in this test. Major handicap is that without any other deficits found in remaining parameters, the patient is accepted as having malnutrition risk (with a total test score of 11), if he/she has CC lower than 31 cm. The second handicap is that some countries have the population with mean CC greater than 31 cm and in these countries, the patients may falsely be considered as having normal nutritional status and given greater points from MNA-SF test (8, 9).

Assessment of HGS is used for the evaluation of the elderly patients for sarcopenia (10, 11) all over the world and almost all geriatric departments have these devices to assess muscle strength. In hypothesis we can think that using HGS in MNA-SF test instead of BMI we may have a chance to assess the patients' nutritional status easily in geriatric practice for patient with unknown BMI and who may be considered to be in poor nutritional status in spite of having a CC of 31 cm or greater.

We aimed to assess the nutritional status of elderly patients by three MNA-SF tests (first one MNA-SF with BMI, second MNA-SF with CC and the third one MNA-SF with HGS) and to compare the reliability of this new, revised MNA-SF test containing HGS.

Material and Methods

Study population

This is a cross-sectional study conducted in the outpatient clinic of Gazi University School of Medicine, Division of Geriatrics between March 2013 and April 2014. One hundred ninety one patients were included in the study after their inform consent were obtained. The participants who could not use the HGS device because of upper extremity abnormalities or any other disorder affecting the bone or joints of the hands resulting in inability in grasping function such as rheumatologic disorders were excluded. We also excluded those with severe pretibial edema due to conditions such as heart failure, nephrotic syndrome, liver failure, or deep vein thrombosis of the lower extremities and also those with greater than 0.5 cm of difference between CC of two legs because these conditions could affect CC measurements. All participants underwent comprehensive geriatric assessment (CGA). CGA consisted of Barthel activities of daily living (ADL) (12), Lawton ADL (13), clock drawing test (CDT) (14), mini-mental state examination (MMSE) (15), Yesavage geriatric depression scale (16) and 6 m walking speed. The Clinical Investigation and Ethic Committee of Gazi University School of Medicine approved the study protocol.

Mini-nutritional assessment tests

MNA-SF is a test evaluating nutritional status of the patients and validated for Turkish elderly people (2). Classic form contains six domains including appetite, weight loss during the last 3 months, mobility, whether the patients are suffering from psychological stress of acute disease in the past 3 months, neuropsychological problems and BMI. Total points of this test may be from 0 to 14 points and the points from 12 to 14 are considered as normal nutritional status, from 8 to 11 points as risk of malnutrition and from 0 to 7 as malnourished. BMI is divided into 4 groups and given 0-3 points according to the patients BMI.

We conducted three MNA-SF tests in this study. First one was the classic MNA-SF with BMI, second one was MNA-SF with CC instead of BMI and the third one was MNA-SF with HGS instead of BMI. MNA-SF with CC was scored similarly to the classic MNA-SF except the BMI. The patients having less than 31 cm of CC were given 0 points and the patients having 31 or greater cm of CC were given 3 points.

MNA-SF with HGS was also scored similarly as the classic MNA-SF test except the BMI. HGS was divided into four groups like BMI. Zero, 1, 2, and 3 points were given if the loss of HGS when compared to expected HGS (expected HGS was 20 kg and 30 kg in female and male patients, respectively (17) were ≥ 60%, from 30% to 60%, from 10% to <30%, and lower than %10 or no loss of the HGS, respectively.

Anthropometric measurements

Measurements of CC, height and bodyweight were conducted in the morning after fasting for at least 8 hours by a geriatrician trained for the measurement of the calf circumference. To calculate the BMI (kg/m2), the patients' height and body weights were evaluated by a standard weighbridge and length measuring device. CC was measured in the standing position using a standard, non-elastic tape measure. To catch the maximal circumference, the measure was moved along the calf without compressing the subcutaneous tissue after surrounding the calf. The measurement of the CC was carried out for both legs. The patients having greater 0.5 cm differences of the CC between two legs were excluded and the greater CC was considered and recorded as the patients' CC result.

Assessment of handgrip strength

Evaluation of the HGS of the patients was carried out with a standard HGS device “grip strength dynamometer, Grip D produced by Takei / made in Japan” by experienced geriatrician. The measurement was applied in the standing position or while sitting in a chair while their hands were in neutral position. The dominant hands of the patients were used for the measurement. After given information how to use the device, the patients were asked to compress the dynamometer with a maximal isometric contraction for 3 seconds. The patients were given a standardized encouragement by the sentence that “Squeeze as hard as you can. If you are ready, we can start. Start, tighten, tighten, tighten, relax”. Three measurements were conducted with the intervals of 10 seconds rest between two measurements. The maximum HGS was used in the analyses.

Statistical analyses

All variables were entered into the statistical program of Statistical Package for Social Sciences (SPSS) 16.0 version. First of all, the numerical parameters were assessed by the test of Kolmogorov-Smirnov and histogram to decide whether they had skewed or normal distribution. The variables showing normal distribution were presented as mean ± standard deviation while the ones showing skewed distribution were as median (minimum - maximum). The categorical variables were shown as number and percentage. Spearman's correlation test was used for the assessment of correlations between three MNA-SF tests scores. To analyze the inter-test reliability, the Cohen's Kappa analysis was used. The p value <0.05 was considered as statistically significance.

Results

The mean age of the participants was 75 ± 7.6 years and 51.8% of them were female. When the nutritional status of the participants evaluated by standard MNA-SF, it was found that 58.6% were in normal nutritional status, 24.6% were at malnutrition risk and 16.8% were malnourished. Smoking and alcohol intake was seen in 11.5% and 3.6% of the participants, respectively. General characteristics of the participants are shown in Table 1.

Table 1.

General characteristics and co-morbidities of the participants

	Number of participants	Percentage
Gender
Male	92	48.2
Female	99	51.8
Marital status
Married	101	52.9
Widow	57	29.8
Single	3	1.6
Not available	30	15.7
Educational status
Illiterate	28	14.7
Literate	12	6.3
Primary school	52	27.2
Middle school	8	4.2
High school	7	3.7
University	19	9.9
Not available	65	34.0
Smoking
Yes	22	11.5
No	106	55.5
Not available	63	33.0
Alcohol
Yes	7	3.6
No	121	63.4
Nutritional status
Normal	112	58.6
Malnutrition risk	47	24.6
Malnutrition	32	16.8
Hypertension
Yes	130	68.1
No	61	31.9
Diabetes mellitus
Yes	50	26.2
No	141	73.8
Alzheimer disease
Yes	27	14.1
No	164	85.9
Bronchial asthma
Yes	15	7.9
No	175	91.6
Not available	1	0.5
Heart failure
Yes	13	6.8
No	177	92.7
Not available	1	0.5
Cancer
Yes	73	38.2
No	117	61.3
Not available	1	0.5
Cerebrovascular disease
Yes	9	4.7
No	181	94.8
Not available	1	0.5
Pretibial edema
Yes	23	12.0
No	145	76.0
Not available	23	12.0

The most common co-morbidities of the participants were hypertension with 68.1%, and diabetes mellitus with 26.2%. The details of the comorbidities are presented in Table 1. All of the participants were evaluated with comprehensive geriatric assessment tests. According to these evaluations, it was detected that mean levels of the BMI, CC and HGS were 28.3 ± 5.6 kg/m2, 34.2 ± 4.6 cm and 23.1 ± 9.5 kg, respectively. The results of other CGA tests are listed in Table 2.

Table 2.

Age, measurements and comprehensive geriatric assessment test results

Body mass index, kg/m2	28.3 ± 5.6
Calf circumference, cm	34.2 ± 4.6
Handgrip strength, kg	23.1 ± 9.5
Age, years	75 ± 7.6
Basic activities of daily living, median (min-max.)	0 (0 - 15)
Instrumental activities of daily living, median (minmax.)	24 (1 - 24)
Clock drawing test, median (min-max.)	6 (0 - 6)
Mini-mental state examination, median (min-max.)	29 (10 - 30)
Yesavage geriatric depression scale, median (minmax.)	1 (0 - 11)
6 meter walking speed, m/sn median (min-max.)	1.1 (0.1 - 2.5)

After analyzing the three MNA tests with Cohen's kappa analysis, substantial agreement in nutritional status of the patients were seen between MNA-SF groups with BMI and CC, with CC and HGS, and with BMI and HGS (kappa: 0.795 p<0.001, kappa: 0.709 p<0.001, and kappa: 0.760 p<0.001, respectively) (Table 3). Similar results were detected when the analyses applied to men and women separately. For men, the kappa analyses results were as kappa: 0.867 p<0.001 for BMI vs. CC, kappa: 0.741 p<0.001 for CC vs. HGS, and kappa: 0.786 p<0.001 for BMI vs. HGS. For women, the kappa analyses results were as kappa: 0.719 p<0.001 for BMI vs. CC, kappa: 0.676 p<0.001 for CC vs. HGS, and kappa: 0.733 p<0.001 for BMI vs. HGS. After excluding the patients with pretibial edema, substantial agreement in nutritional status of the patients were also seen between MNA-SF groups with BMI and HGS (kappa: 0.761 p<0.001).

Table 3.

Kappa analysis test results between three MNA-SF test groups

n (%)			MNA-BMI		Kappa coefficient	p value
		Normal	Malnutrition risk	Malnutrition
MNA-HGS	Normal	99 (88.4)	0 (0.0)	0 (0.0)	0.760	<0.001
	Malnutrition risk	13 (11.6)	36 (76.6)	3 (9.7)
	Malnutrition	0 (0.0)	11 (23.4)	28 (90.3)
	Total	112 (100.0)	47 (100.0)	31 (100.0)
n (%)			MNA-BMI
		Normal	Malnutrition risk	Malnutrition
MNA-CC	Normal	93 (86.9)	0 (0.0)	0 (0.0)	0.795	<0.001
	Malnutrition risk	14 (13.1)	38 (88.4)	3 (9.7)
	Malnutrition	0 (0.0)	5 (11.6)	28 (90.3)
	Total	107 (100.0)	43 (100.0)	31 (100.0)
n (%)			MNA-CC
		Normal	Malnutrition risk	Malnutrition
MNA-HGS	Normal	86 (92.5)	9 (16.7)	0 (0.0)	0.709	<0.001
	Malnutrition risk	7 (7.5)	34 (63.0)	5 (15.2)
	Malnutrition	0 (0.0)	11 (20.3)	28 (84.8)
	Total	93 (100.0)	54 (100.0)	33 (100.0)

There were strongly positive correlations between MNA-SF scores with BMI and CC, with BMI and HGS, and with CC and HGS (r=0.938 p<0.001, r=0.938 p<0.001, r=0.914 p<0.001, respectively) (figure 1).

Figure 1.

There were strongly positive correlations between MNA-SF scores with BMI and CC, with BMI and HGS, and with CC and HGS (r=0.938 p<0.001, r=0.938 p<0.001, r=0.914 p<0.001, respectively)

Discussion

In our study, we have examined the reliability of MNA-SF with HGS and shown that adding HGS in MNA-SF test instead of BMI may be as reliable as the other MNA-SF tests with BMI or CC. To our knowledge, this is the first study showing this relationship.

Hand grip strength is generally used to reveal the patients' muscle function and especially to evaluate the patients for sarcopenia (18). Beside this benefit, recently, its importance in exposing the patients' nutritional status is showing that HGS may be used as a predictor to show the nutritional status of the patients (19). In a study by Flood et al., 217 hospitalized patients were investigated and found that HGS and Patients Generated Subjective Global Assessment score were significantly correlated and also HGS was found to be independently predictive factor to clarify nutritional status of the patients (20).

Mini nutritional assessment-SF test is a simple and highly reliable test for assessing the nutritional status of the patients especially in the elderly (21, 22). If the patient's BMI is not known or cannot be measured, the commonly advised way to assess the patient is that CC may be used instead of BMI (5). But, the major problem with this new advised way is that the conditions that may have an effect on measurement of CC by increasing the measurements of CC such as pretibial edema may cause false results. Moreover, we know that the CC may change between different ethnic groups. For example, in our country, the studies showing CC of the patients have demonstrated that our patients, even if they have sarcopenia, have CC greater than 31 cm (8, 9). Supporting these findings we have found in recent study that the mean CC is 34.2 ± 4.6 cm. So, it means that in our country in condition that we use CC in the assessment of nutritional status of the patients in MNA-SF test, we may be mistaken. Our findings suggest that HGS may be a reliable way for adding in MNA-SF instead of BMI. However, in this study MNA-SF using CC had a good agreement with classical MNA-SF. We suggest that our exclusion criteria regarding drawbacks of CC like severe pretibial edema may have resulted in accurate classification of the nutritional status of the patients using MNA-SF with CC.

Body mass index measurement is not available in many elderly people and height loss due to kyphosis and vertebral fractures result in overestimation of BMI and using CC has many drawbacks, few studies have been conducted to find alternatives for BMI and CC in MNA-SF test such as mid-upper arm circumferences (23). But, best of our knowledge, no studies is found in the literature using HGS in MNA-SF test and our study is the first study investigating this subject. A potential strength of using HGS is that HGS changes more easily in conditions precipitating malnutrition. For example, a fall in BMI category generally occurs after substantial weight loss. This is even further late in CC. It could also be noted that using HGS instead of BMI in MNA-SF could provide relevant information in longitudinal studies and such studies are needed

This study has some limitations. The sample size is not so large. Nonetheless, we could find statistically significant findings supporting our hypothesis with this sample size. But anyway, our findings should be repeated by studies with larger sample size. We have tried to find out the best cut off values for HGS to add MNA-SF test as 0, 1, 2 or 3 points according to the loss of expected HGS. There are no widely accepted and standard cut off values for HGS scores to use in MNA-SF test. One may repeat the analysis with different cut off levels. And lastly, our findings only reflect results of Turkish elderly population. Studies conducted in other ethnic groups should be designed. On the other hand, there are some limitations of using HGS in MNA-SF test. The most important one is that it may be difficult to use this device cause of the hand deformities or any other reasons affecting hand grip strength such as acute inflammatory arthritis on the hand of the patients.

Conclusion

As a conclusion, it has been shown in recent study that instead of BMI, HGS may be considered as a good, practical and reliable alternative in MNA-SF test to assess nutritional status of geriatric patients.

Conflict of interest

None

Financial disclosure:

None

Ethical Standards

The Clinical Investigation and Ethic Committee of Gazi University School of Medicine approved the study protocol. The study was conducted in accordance with Helsinki Declaration.