Servicios personalizados
Inglés (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Enviar articulo por email
Citado por SciELO 
Citado por Google 
Similares en
SciELO 
Similares en Google 
Permalink
INTRODUCTION
The prevalence of malnutrition in hospitalized elderly is high when compared to community elderly 1. Therefore, the investigation of adequate techniques for nutritional assessment in this population is paramount for nutritional care, in order to support diagnosis and nutritional management 2. In this context, the Mini Nutritional Assessment® (MNA®) applied in its integral form 3 has been widely used for the evaluation of nutritional status of hospitalized elderly 1,3. However, some difficulties are associated with its application, e.g., when the elderly presents physical and mental limitations 1,3, in the absence of a companion to provide information for the application of the instrument or when the elderly receives enteral nutrition therapy 3. As such, with the aim of searching nutritional assessment instruments that are independent of the patient collaboration 4 in conjunction with rapid application and low cost, the adductor pollicis muscle thickness (APMT) measurement emerges as an anthropometric parameter 5. The adductor pollicis muscle is the only muscle that allows its direct thickness measurement in function of its anatomic characteristic 6. Many studies have been using this measurement in different populations and scenarios, such as community individuals 6 7 8-9, institutionalized elderly 10; different clinical conditions as in oncologic patients 11,12,13, chronic kidney disease and hemodialysis 14 15-16; and hospitalized clinical 17, surgical 17,18,19,20,21 and intensive unit care patients 22,23-24.
Nonetheless, specific studies regarding hospitalized elderly are scarce 5, with only two available studies 25,26. None of the aforementioned studies analyzed the association between APMT and the MNA®, which is a specific nutritional assessment instrument for the elderly. Therefore, the aim of the present study was to analyze the association between the APMT and nutritional parameters in the hospitalized elderly.
METHODS
STUDY DESIGN, POPULATION AND SAMPLE
This is a transversal study. Three hundred and thirty-one elderly patients (≥ 60 years old) hospitalized at the Hospital São Lucas from the Pontifical Catholic University of Rio Grande do Sul (HSL/PUCRS) were evaluated from May 2014 to December 2015. Successful obtainment of the APMT and further assessed nutritional parameters were used as inclusion criterions.
Regarding the sample size for the APMT evaluation, with an error margin of 5 mm 17, power of 80% and significance level of 5%, and considering the frequency risk of malnutrition and malnutrition in the studied population, a minimal sample size of 130 elderly was required.
INVESTIGATED VARIABLES
The variables investigated were sex, age, nutritional state (MNA® and BMI), anthropometric variables (arm circumference [AC], calf circumference [CC] and APMT) and handgrip strength (HGS).
DATA COLLECTION LOGISTICS
Data was collected at the elderly admission through trained evaluators.
The MNA® was applied with the patient alone or with the aid of a companion, when there was lack of cognitive conditions to answer the questions independently 27.
Body weight was measured through the digital portable scale (Urâno, UPC 150 model, Brazil), and height was obtained with a portable stadiometer (Sanny®, Personal Caprice model, Brazil). When it was not possible to perform body and/or weight measurements, the information was collected with the patient or companion. The BMI was obtained as the body mass (kg)/squared estimated height (m2)28.
The elderly patients were classified according to the established cut-off values described by Lipschitz 29: < 22 kg/m² (underweight), between 22 and 27 kg/m² (eutrophic) and > 27 kg/m² (overweight).
The AC was measured with an inelastic tape, positioned in the arm midpoint, between the acromion and olecranon. CC was assessed with the leg of the elderly flexed in a 90° angle, with an inelastic tape positioned in the maximum perimeter of the muscle, with an uncovered calf 30. AC ≤ 22 cm and CC < 31 cm values were considered as decreased, for both genders, following the cut-off values described at the MNA® 27.
The APMT was obtained in the dominant hand, and in case of obtainment impossibility, in the non-dominant hand. Details regarding APMT measurement method were previously published by El Kik and Schwanke 25. APMT was classified as decreased according to the percentile 5 values (18 mm for men and 14 mm for women) 8.
HGS was measured with the patient sitting, on the dominant hand, or in case of obtainment impossibility, on the non-dominant hand. The elbow was flexed at 90°, the forearm was in a neutral position, the grip was adjusted in the second position, and the patient held the maximum grip force during three seconds with 30 seconds to one-minute interval 31. Three measurements were performed and the average was considered for the HGS. HGS was classified according to the percentile 5 (18 kgf for men and 11 kgf for women) 9.
STATISTICAL ANALYSIS
Data was computed on a Microsoft Excel database and further analyzed with SPSS 21.0 (SPSS Inc. Chicago, IL, USA). Normality distribution of continuous data was assessed by the Kolgomorov-Smirnov test, with all data variables presenting a normal distribution. Descriptive analysis was performed by frequency means, central tendency and dispersion. In order to compare mean average of APMT in relation to sex, the Student's t-test was applied. For the association between APMT classification with sex and age, the Pearson's Chi-square was used. Comparison between the APMT averages was conducted with the Student's t-test and analysis of variance (ANOVA), with a Bonferroni post-hoc correction. Multivariate analysis of Poisson regression was applied in order to assess the APMT effect over the investigated nutritional parameters due to the transversal study design and to correct for age and sex of the elderly. For this analysis, the polytomic variables were dichotomized.
ETHICAL ASPECTS
The data presented in this article were obtained from the doctorate project entitled "Handgrip strength and adductor pollicis muscle thickness as nutritional status, functional capacity and mortality assessment instruments in hospitalized elderly patients", approved by the PUCRS Research Ethics Committee (CAAE: 26825614.7.0000.5336, assessment 636.335 of 29/04/2014). Determinations from the National Heath Council resolution no. 466/2012 32 were followed and all the participants were instructed regarding the procedures involved in this research and signed the free and informed consent form.
RESULTS
The sample majority was composed by women (56.8%). Mean age was 78.4 ± 9.7 years old, ranging from 60 to 105 years old. Women are more frequently in the age group ≥ 80 years. The most frequent causes of hospitalization were respiratory diseases (n = 60, 17.9%) and gastroenterological (n = 53, 15.8%) and cardiac (n = 43, 12.8%) conditions.
Regarding nutritional parameters, MNA® nutritional status was normal (45.7%); patients were considered as eutrophic as measured by the BMI (40.1%), AC > 22 cm (87.6%), CC > 31 cm (71.5%) and a HGS without depletion risk (64.8%) (Table 1).
Table I. Nutritional status according to nutritional parameters of hospitalized elderly patients
< P5:lower than percentile 5;
≥ P5:equal or higher than percentile 5.
Note: 40 missing data were present for the MNA® and arm circumference analysis, two for the BMI, 43 for the calf circumference and 70 for the handgrip strength.
On Table 2, depicted average values of APMT were significantly lower in women as compared to men (12.67 ± 4.13 mm vs 15.26 ± 4.01 mm, p < 0.001). Elderly with over 80 years old had lower APMT values than younger elderly patients (12.62 ± 4.38 mm vs 14.83 ± 3.90 mm, p < 0.001), in men (14.36 ± 4.10 mm vs 15.76 ± 3.89 mm, p = 0.044) as well as in women (11.78 ± 4.28 mm vs 13.80 ± 3.66, p = 0.001).
Table II. Adductor pollicis muscle thickness classification of elderly hospitalized patients according to sex and age
APMT: adductor pollicis muscle thickness;
< P5:lower than percentile 5;
≥ P5:equal or higher than percentile 5.
*Student t test;
†Pearson's Chi-square test.
As for the APMT classification, women were more frequently below the P5 when compared to men (89.9% vs 37.1%), regardless of age (Table 2).
At the univariate analysis, APMT was associated with all investigated nutritional parameters, independent of age (Table 3).
Table III. Comparison of average adductor pollicis muscle thickness between sex, according to the classification of nutritional parameters of hospitalized elderly patients
MNA®: Mini Nutritional Assessment; APMT: adductor pollicis muscle thickness; HGS: handgrip strength
*Analysis of variance-ANOVA and Bonferroni post-hoc test (the mean values with different superscript letters are significally different).
†Student's t-test.
‡Information presented only between sex, since there is no cut-off value for the total sample (handgrip strength P5: for men = 18 kgf and for women = 11 kgf).
On Table 3, the comparison between the APMT according to the nutritional parameters classification is described. The APMT average value of elderly patients with normal nutritional status was higher than in elderly patients classified as malnourished or with risk of malnourishment (p < 0.001). No significant difference was observed between the APMT average values of malnourished elderly and at risk of malnourishment elderly patients. Elderly classified as underweight by the BMI presented a lower mean APMT as compared to eutrophic and overweight elderly patients.
Elderly patients with AC > 22 mm presented a higher APMT than elderly with AC < 22 cm, as well as patients with CC ≥ 31 cm presented higher APMT as compared to CC < 31 cm.
Men with HGS ≥ 18 kgf had superior APMT in contrast with elderly patients with HGS < 18 kgf (p = 0.021).
At the multivariate analysis (Table 4), APMT proved to be an independent measurement associated to all nutritional parameters, even when adjusted for age and sex (p < 0.05). Increased APMT was related to a 4% lower chance of malnutrition/risk of malnutrition as classified by the MNA®, 11% in the BMI, 21% in the AC < 21 cm, 12% in the CC < 31 cm and 6% in the HGS < P5.
DISCUSSION
The present study evaluated the association between the APMT and nutritional parameters in hospitalized elderly in Porto Alegre, a city located in the south of Brazil. It is one of the few studies performed only in hospitalized elderly patients 25,26. Another important characteristic is that this study is the first one to evaluate the association between the APMT and diagnosed malnutrition by the MNA®, whereas other studies analyzed the relationship between APMT and the Subjective Global Assessment, and other nutritional parameters 13,15,20 21-22,24,33. It was possible to observe that the APMT was associated to all investigated nutritional parameters such as the MNA®, BMI, AC, CC and HGS.
According to the different nutritional parameters, the investigated elderly patients presented a risk of malnutrition/malnutrition frequency of 54.3% in the MNA®, 26.1% in the BMI, 12.3% in the AC, 64.4% in the CC and 35.2% in HGS.
Recent literature has demonstrated that malnutrition in the hospitalized elderly is frequent 1. The present results were similar to those of a study conducted with 89 elderly patients in a medium complexity hospital of Rio Grande do Sul, where 58% of malnutrition/risk of malnutrition was observed according to the MNA® 34. A systematic review with meta-analysis included 66 studies that evaluated hospitalized elderly patients by the MNA® and demonstrated higher malnutrition/risk of malnutrition values (67.5%) in this population 35. Even higher values were identified in a study conducted in a high complexity hospital of Rio Grande do Sul including 131 elderly patients 36, observing malnutrition in 73.7% of women and 89.1% of men.
Other similar studies with hospitalized elderly demonstrated higher values as compared to the present study concerning the BMI underweight frequency, according to Lipschitz. We identified that 26.1% of the investigated elderly patients were underweight, whereas other studies observed values of 36.6% (36), 38% 37 and 41% 38.
The frequency of CC < 31 cm in the present study was 28.5%, whereas other studies identified even higher inadequacy values, such as 38.2% 36 and 59% 37.
Regarding AC, only 12.3% of the elderly patients were classified below the recommended values, while others verified higher values as 63% 39 and 67.2% (36).
The difference between the obtained values might be attributed to the epidemiological profile of the studied population. In the study performed by Zanchim et al. 36, the main cause of hospitalization was cancer. As for the study by Morais, Campos and Lessa 37, the most frequent cause of hospitalization was respiratory diseases.
Related to the anthropometric parameters, there was a perceived limitation regarding availability of data for comparative analysis. According to Fidelix, Santana and Gomes 39 in a narrative review, there are scarce number of studies providing updated data regarding hospitalized elderly population.
Besides being a functional capacity indicator and predictor of aggravated health status in the elderly, the HGS has been considered to be a nutritional state indicator 40. However, it was not possible to identify studies which used the same cut-off value as the one applied in the current study, which was based on a population research that determined the cut-off value for Brazilian elderly population 9.
Differences in the APMT related to sex 6 7-8,14,16 and age 6,16 have been previously described. According to a previous published study with a lower sample size 25, women and elderly had a reduced APMT. Still, in the present study, women had a higher frequency of APMT < P5, whereas men had APMT > P5 (p < 0.001), even when correcting for age < 80 or ≥ 80 years old. A study with hospitalized candidates to surgical procedure also found a higher frequency of malnutrition through the APMT in women. Such finding can be explained by the bodily composition differences among men and women 19, and also due to the fact than women are older than men in this sample. Although women live longer, they experience worse health and present a greater chance than men to experience the typical diseases of the last phase of life 41.
The present study demonstrated higher mean values of APMT in elderly classified with a normal nutritional status (MNA®); eutrophic by BMI, with a AC > 22 mm, CC ≥ 31 cm and HGS ≥ 18 kgf in men and FPP ≥ 11 kgf in women. Another study conducted in surgical patients observed a higher mean APMT value in patients classified as eutrophic by means of AC and BMI 19.
APMT demonstrated to be an independent factor associated to all parameters, even when corrected by age and sex, reinforcing its applicability in the clinical practice. Several studies have associated APMT to classic nutritional assessment parameters, such as anthropometric measures, highlighting its relation with malnutrition 6,14,15,18,19,21,24, as well as the HGS 15,40.
Karst, Vieira and Barbiero 24 assessed the correlation between APMT, BMI and CC in patients in a cardiologic intensive care unit. Oliveira et al. 14 observed a positive correlation between the APMT and BMI and also APMT and AC in hemodialysis patients.
In the study conducted by Pereira et al. 15 in hemodialysis patients, a positive correlation between the APMT and HGS was observed. This is in accordance with the study by Guerra et al. 40, conducted in hospitalized patients, where an association between APMT and HGS was also found.
Interestingly, one study conducted in hospitalized patients (17) and other in breast cancer patients 13 found a weak association between the APMT and anthropometrics parameters, and both studies related this finding to the fact that the studied population was predominantly overweight/obese.
Another study in hemodialysis patients 15 did not find a correlation among APMT and anthropometric parameters, relating this finding to possible hydric alterations in this population, as well as possible errors in the reproducibility of anthropometric measures related to the observer's variability.
Even though several studies pointed towards an association between APMT measurement and classic nutritional parameters, results are still insufficient to promote APMT as a unique indicator of lean mass reserve 7. The nutritional assessment is considered as a complex process and it is wise to associate different methods in the pursuit of a more trustworthy and precise diagnosis 2.
The present study has some limitations, especially regarding its design (cross-sectional) and the low malnutrition frequency. In this context, further research is required with matching nutritional status condition.
In conclusion, APMT was associated with all the investigated nutritional parameters, independent of sex and age, which reinforces its applicability in the nutritional assessment of elderly patients.
