Clinical Nutrition xxx (2014) 1e5
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients Maria Cristina Gonzalez a, *, Rodrigo Roig Pureza Duarte a, Silvana Paiva Orlandi b, Renata Moraes Bielemann c, Thiago Gonzalez Barbosa-Silva c a b c
Post-graduation Program in Health and Behavior, Catholic University of Pelotas, Pelotas, RS, Brazil Nutrition College, Federal University of Pelotas, Brazil Post-graduation Program in Epidemiology, Federal University of Pelotas, Brazil
a r t i c l e i n f o
s u m m a r y
Article history: Received 26 September 2014 Accepted 9 November 2014
Background & aims: Body composition is important to identify malnutrition, and several anthropometric measurements are used to estimate muscle mass in the clinical practice. This study aimed to assess the adductor pollicis muscle thickness (APMT), its covariates and association with malnutrition in hospitalized surgical patients. Methods: APMT was measured in 361 surgical patients in both dominant (DAPMT) and non-dominant (NDAPMT) sides. APMT values below the 5th percentile of reference values provided by a healthy population were considered as malnourished. Nutritional status was assessed by Subjective Global Assessment (SGA). The difference in APMT values among nutritional status categories was evaluated, and the association between malnutrition by SGA and APMT was estimated using multivariate linear regression. Sensitivity, specificity and positive and negative predictive values were also calculated. Results: Most patients were women (60.4%). APMT values were significantly different among SGA categories. Well-nourished patients had APMT values significantly higher compared to the ones with moderate or severe malnutrition by SGA, with no significant difference between APMT values in moderate or severe malnourished patients. Statistically significant associations between both DAPMT and NDAPMT below the 5th percentile and malnutrition and were found (RR ¼ 3.99, CI 95% ¼ 3.19e5.00; p < 0.001; and RR ¼ 3.92; CI 95% ¼ 3.10e4.96; p < 0.001; respectively). Gender, age, estimated weight and nutritional status were considered associated factors for APMT. APMT showed low sensitivity (DAPMT: 34.9%, NDAPMT: 37.7%) but high specificity (DAPMT: 98.7%, NDAPMT: 97.8%) to identify malnutrition. Conclusions: APMT was significantly associated with nutritional status in a sample of surgical patients. The APMT seems to be a simple and useful anthropometric tool to confirm the diagnosis of malnutrition. © 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism.
Keywords: Adductor pollicis muscle Malnutrition Anthropometry Hospitalized patients
1. Introduction There is still lack of an adult malnutrition definition accepted in clinical practice. A new proposal published recently defines an etiology-based diagnosis, according to the presence of starvation with or without acute or chronic inflammation. The final effect of either starvation or disease-related malnutrition will be on lean body mass. This definition highlights the importance of body
* Corresponding author. Rua Ver. Ariano R. de Carvalho, 304, CEP 96055-800 Pelotas, RS, Brazil. Tel.: þ55 53 99821328. E-mail address: [email protected] (M.C. Gonzalez).
composition assessment to identify malnutrition, even in overweight or obese patients [1]. Although a number of body composition equipment is now available to analyze with precision muscle and body fat mass, high cost restrains its use in clinical practice. Bioelectrical impedance analysis is a simple and practical method of body composition assessment, but fluid and electrolyte imbalances and the presence of edema or ascites, usually found in hospitalized patients, restricts its use in this population [2,3]. Anthropometric measurements usually assess muscle mass indirectly, through arm-muscle circumference and muscle area of the arm [4]. The only muscle that can be almost directly measured is the adductor pollicis muscle. This muscle has an easily accessible
http://dx.doi.org/10.1016/j.clnu.2014.11.006 0261-5614/© 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism.
Please cite this article in press as: Gonzalez MC, et al., Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients, Clinical Nutrition (2014), http://dx.doi.org/10.1016/j.clnu.2014.11.006
2
M.C. Gonzalez et al. / Clinical Nutrition xxx (2014) 1e5
location in the hands. The first studies focused its evaluation for muscle function through ulnar electric stimulation, which is not ideal [5,6]. It was later used as an anthropometric parameter [7,8] and reference values were recently published for healthy subjects [9]. Subjective global assessment (SGA) [10] is one of the most used techniques for nutritional assessment in clinical practice. Due to the lack of a gold standard for nutritional assessment, it is also used to validate new methods [11]. The use of adductor pollicis muscle thickness (APMT) in the clinical practice is still scarce, but recent studies have shown a positive association of APMT with the estimation of the total muscle compartment [12] and also a negative relationship with mortality [13]. The relationship between APMT and nutritional parameters has been investigated in surgical patients [14,15]. However, both studies were carried out with small samples and no distinction was made between gender and APMT measurement. In addition, SGA was used as the reference method to assess malnutrition in only one of the studies [14]. The objective of this study is to test the validity of APMT as a nutritional parameter in a sample of hospitalized surgical patients, using SGA as the gold standard method to assess nutritional status. 2. Materials and methods A sample of 361 patients of both genders, older than 18 years, hospitalized in a surgical ward in an academic hospital in Southern Brazil was studied in 2009. Patients who had been hospitalized or submitted to surgery in the last 30 days or were not able to be evaluated due to the presence of local venous accesses were excluded. Patients were invited to participate in the study and signed an informed consent term. A questionnaire was applied to assess information such as: age, gender, marital status and height. Weight was obtained from patient's chart records or referred by patient or their relatives because most of patients were in postoperative period, and were not allowed to be weighted. Dominant hand side and presence of injury and/or fracture in any of the hands were also considered relevant data, as they could influence the final results of measurements. The socioeconomic level was evaluated according to the crite~o Brasileira de Empresas de Pesquisa rion adopted by the Associaça (ABEP) [16]. This criterion is based on the property of domestic consumption assets, the presence of house servants and the education level of the head member of the family, allowing the classification of the patient in five groups e from the richest (A) to the poorest (E). The medical history (changes in weight, dietary intake and functional capacity, gastrointestinal symptoms with nutritional impact, and metabolic stress of present disease) and a brief physical examination were also assessed to determine the nutritional assessment by SGA, as firstly described by Detsky [10]. Individuals were classified in one out of three groups: A e eutrophic; B e moderate or suspected malnourishment; and C e severely malnourished. A Lange® skinfold caliper was used to assess APMT. The APMT measurement was performed using a standardized position: patients sat in the bed with both arms relaxed and elbows in a 90degree angle, approximately, with the hands laying over the legs. Adductor pollicis muscle thickness was measured by skinfold caliper, in the vertex of an imaginary triangle formed by the extension of the thumb and the index finger, under a continuous pressure of 10 g/mm. According to previous study [9], the mean of three measurements was used. Dominant APMT (DAPMT) values were considered from the right hand of right-handed subjects, and from the left hand of the left-handed ones. Non-dominant APMT
(NDAPMT) values were considered from the measurements of the opposite hand. Examiners were trained and standardized using acceptable technical errors of measurement calculated based on Habicht's publication [17]. Sample size estimation was made accordingly to two previous Brazilian studies using SGA in hospitalized patients [18,19]. These studies found malnutrition prevalence from 33% to 38%. Using such values and a confidence level of 95%, it was established that the necessary number of patients to be evaluated was 362. Reference values used in this study were obtained from a sample of 300 healthy subjects, grouped by gender and age, published previously by our group [9]. The 5th percentile in each age and gender group was considered the cut-off value for healthy subjects. Values lower than the 5th percentile were considered depletion of muscle mass by APMT. The normal distribution of variables was tested through Skewness and Kurtosis tests. ANOVA test was performed, and Bonferroni's test was also used to verify the difference of APMT values among nutritional status categories. Association between malnutrition diagnosis by SGA and depletion of muscle mass by APMT was tested by c2 test. Pearson's correlation test was made to assess the correlation between the several variables and APMT, as well as a multivariate linear regression (backward selection) to identify possible determinant factors of APMT in this group of patients. Specificity and sensitivity of APMT to identify malnutrition (B or C in the SGA) were tested in comparison to SGA diagnosis (here, considered the gold standard). All analyses were performed by Stata program (version 12.0), and the significance level considered for all tests was 5%. The present study was submitted and approved by the Ethics rdia de Pelotas, a hospital in Committee of Santa Casa de Miserico southern Brazil. 3. Results A sample of 361 patients who had APMT evaluated in at least one hand was studied. The sample characteristics are presented in Table 1. Most of the patients (60.2%) were women. The average age of the sample was 49.6 ± 17.8 years old, with no significant difference between genders. Most of the patients were Caucasian (78.4%) and lived with a partner (53.5%). The mean usual weight was 68.4 ± 13.8 kg, and men showed higher weight than women (72.08 ± 14.4 kg and 65.9 ± 12.8 kg,
Table 1 Sociodemographic characteristics of 361 surgical patients. Variables Gender Male Female Age (years) 18e30 31e59 60 Socioeconomic status D/E C B Marital status Single Married/Committed Widow/Widower Separated Race Caucasian Non-caucasian
n (%) 144 (39.8) 217 (60.2) 68 (18.8) 184 (51.0) 109 (30.2) 117 (32.4) 200 (55.4) 44 (12.2) 92 193 40 36
(25.5) (53.4) (11.1) (10.0)
283 (78.4) 78 (21.6)
Please cite this article in press as: Gonzalez MC, et al., Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients, Clinical Nutrition (2014), http://dx.doi.org/10.1016/j.clnu.2014.11.006
M.C. Gonzalez et al. / Clinical Nutrition xxx (2014) 1e5 Table 2 Type of surgery performed in 361 patients of the sample. Number of patients (%) Digestive system surgery Gynecological and breast surgery Urgent abdominal surgery Post-surgery complications Wall surgery Genitourinary surgery Thoracic area surgery Other procedures
119 58 55 29 25 14 8 53
(33.0) (16.1) (15.2) (8.0) (6.9) (3.9) (2.21) (14.7)
3
Table 4 Pearson correlation between the adductor pollicis muscle thickness (dominant and non-dominant sides) and age and other anthropometric variables. All the correlations were significant (p < 0.05).
c
DAPMT NDAPMTd a b c d
Age
Weight
Height
BMIa
SGAb
0.28 0.28
0.52 0.56
0.19 0.18
0.44 0.50
0.61 0.60
BMI e body mass index. SGAe nutritional status according to subjective global assessment. DAPMT e dominant adductor pollicis muscle. NDAPMT e non-dominant adductor pollicis muscle 7.
respectively, p < 0.001). Most of the patients (68.4%) were considered well-nourished by SGA, but 87 patients (24.1%) were considered suspects or moderately malnourished (SGA B) and only 7.5% of the patients were considered severely malnourished (SGA C). Most of the patients in this ward were submitted to biliary tract surgery (14.4%). The most common procedure was cholecystectomy and exploratory laparotomy (14.1%). All procedures are specified in Table 2. APMT values were also compared according to the nutritional status categories (Table 3). DAPMT and NDAPMT values were significantly different according to the nutritional status categories (p < 0.001). According to Bonferroni test, it was found a significant difference in DAPMT and NDAPMT values among patients who were considered well-nourished (SGA A) when compared to patients with moderate (SGA B) and severe malnutrition (SGA C). However, no significant difference was found among values in patients who had moderate and severe malnutrition. It was also detected a significant linear trend in DAPMT and NDAPMT values, according to nutritional status categories: the worse the nutritional status was, the lower the DAPMT and NDAPMT values were found (p < 0.001). A strong association of malnutrition assessed by SGA and depletion of muscle mass by APMT was found. Patients with APMT below the 5th percentile had a greater risk of being malnourished, as assessed by SGA, for both dominant and non-dominant sides (RR ¼ 3.99; 95%CI: 3.19e5.00; p < 0.001; and RR ¼ 3.92; 95%CI: 3.10e4.96; p < 0.001; respectively). Sensitivity and specificity of DAPMT and NDAPMT were also tested, using their 5th percentile value to identify malnutrition and having SGA as the gold standard. APMT measurements presented low sensitivity (34.9% and 37.7% for DAPMT and NDAPMT, respectively), but a high specificity (98.7% and 97.8% for DAPMT and NDAPMT, respectively) to rule out malnutrition diagnosis in wellnourished patients. In patients with DAPMT higher than the healthy cut-off values, only 23.2% were malnourished (negative predictive value ¼ 76.8%). When a DAPMT lower than 5th
percentile value was found, 92.5% of patients were malnourished (positive predictive value ¼ 92.5%). For the NDAPMT, positive and negative predictive values were 88.9% and 77.3%, respectively. Correlations between DAPMT and NDAPMT and other variables, such as gender, age, weight, height, BMI and SGA were evaluated and presented in Table 4. All correlations with those variables were significant (p < 0.05). The highest positive correlations were obtained using DAPMT or NDAPMT associated with usual weight (r ¼ 0.52 and 0.56, respectively). The highest negative correlations were obtained using DAPMT or NDAPMT associated with SGA (r ¼ 0.61 and 0.60, respectively). Multivariate linear regression is presented in Table 5. Nutritional status was considered the most important determinant factor of DAPMT and NDAPMT, even after adjustment for other variables such as gender, age and weight. Patients classified as suspects or moderately malnourished (SGA B) presented a decrease of 4.59 and 4.56 mm in DAPMT and NDAPMT, respectively. An even larger decrease was found in patients classified as severely malnourished (SGA C): DAPMT and NDAPMT were 6.51 and 6.14 mm lower, respectively, when compared to values found in patients who were considered well-nourished.
Table 3 Median and interquartile range values of adductor pollicis muscle thickness (dominant and non-dominant sides) according to the nutritional status and gender.
Table 5 Multivariate linear regression of the adductor pollicis muscle thickness and gender, age, weight and nutritional status (backward selection).
DAPMTa Female c
e,f
4. Discussion This study assessed the relationship between APMT and nutritional parameters in surgical patients from a hospital located in Pelotas, Southern Brazil. It was found that patients with APMT below the 5th percentile in comparison to the general population had a greater risk of being malnourished according to SGA. APMT showed low ST in the prediction of malnutrition, though a high SP (more than 90%) was found. The adductor pollicis is the only muscle in the body which allows an almost direct and adequate assessment of its thickness, due to its easily accessible location, in a well-defined anatomic position, and minimal thickness of subcutaneous layers of adipose tissue surrounding it [7,8]. Most of previous studies use its function assessment through electric stimulation of the ulnar nerve.
NDAPMTb Male e,f
Female e,f
DAPMTa,b Male e,f
22 (20; 25) 25 (24; 28) 23 (21; 25) 26 (25; 28) SGA A (nd ¼ 148/84) e e SGA B (n ¼ 45/36) 18 (16; 20) 19.5 (16.5; 22) 18e (16; 20) 18e (15.5; 20) SGA C (n ¼ 11/14) 17f (15; 18) 18f (15; 20) 16f (12; 18) 16f (15; 20) e,f : In the same gender, same letter means a significant difference of DAPMT/ NDAPMT values among the SGA categories (p < 0.05). a DAPMT: dominant adductor pollicis muscle thickness of patients. b NDAPMT: non-dominant adductor pollicis muscle thickness of patients. c Subjective Global Assessment. d n: number of female/male.
Gender Age (years) Weight (kg) SGAe B SGAe C a b c d e
NDAPMTc,d
b
EP (b)
p
b
EP (b)
p
1.55 0.02 0.10 4.59 6.51
0.36 0.01 0.01 0.45 0.69
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients Maria Cristina Gonzalez a, *, Rodrigo Roig Pureza Duarte a, Silvana Paiva Orlandi b, Renata Moraes Bielemann c, Thiago Gonzalez Barbosa-Silva c a b c
Post-graduation Program in Health and Behavior, Catholic University of Pelotas, Pelotas, RS, Brazil Nutrition College, Federal University of Pelotas, Brazil Post-graduation Program in Epidemiology, Federal University of Pelotas, Brazil
a r t i c l e i n f o
s u m m a r y
Article history: Received 26 September 2014 Accepted 9 November 2014
Background & aims: Body composition is important to identify malnutrition, and several anthropometric measurements are used to estimate muscle mass in the clinical practice. This study aimed to assess the adductor pollicis muscle thickness (APMT), its covariates and association with malnutrition in hospitalized surgical patients. Methods: APMT was measured in 361 surgical patients in both dominant (DAPMT) and non-dominant (NDAPMT) sides. APMT values below the 5th percentile of reference values provided by a healthy population were considered as malnourished. Nutritional status was assessed by Subjective Global Assessment (SGA). The difference in APMT values among nutritional status categories was evaluated, and the association between malnutrition by SGA and APMT was estimated using multivariate linear regression. Sensitivity, specificity and positive and negative predictive values were also calculated. Results: Most patients were women (60.4%). APMT values were significantly different among SGA categories. Well-nourished patients had APMT values significantly higher compared to the ones with moderate or severe malnutrition by SGA, with no significant difference between APMT values in moderate or severe malnourished patients. Statistically significant associations between both DAPMT and NDAPMT below the 5th percentile and malnutrition and were found (RR ¼ 3.99, CI 95% ¼ 3.19e5.00; p < 0.001; and RR ¼ 3.92; CI 95% ¼ 3.10e4.96; p < 0.001; respectively). Gender, age, estimated weight and nutritional status were considered associated factors for APMT. APMT showed low sensitivity (DAPMT: 34.9%, NDAPMT: 37.7%) but high specificity (DAPMT: 98.7%, NDAPMT: 97.8%) to identify malnutrition. Conclusions: APMT was significantly associated with nutritional status in a sample of surgical patients. The APMT seems to be a simple and useful anthropometric tool to confirm the diagnosis of malnutrition. © 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism.
Keywords: Adductor pollicis muscle Malnutrition Anthropometry Hospitalized patients
1. Introduction There is still lack of an adult malnutrition definition accepted in clinical practice. A new proposal published recently defines an etiology-based diagnosis, according to the presence of starvation with or without acute or chronic inflammation. The final effect of either starvation or disease-related malnutrition will be on lean body mass. This definition highlights the importance of body
* Corresponding author. Rua Ver. Ariano R. de Carvalho, 304, CEP 96055-800 Pelotas, RS, Brazil. Tel.: þ55 53 99821328. E-mail address: [email protected] (M.C. Gonzalez).
composition assessment to identify malnutrition, even in overweight or obese patients [1]. Although a number of body composition equipment is now available to analyze with precision muscle and body fat mass, high cost restrains its use in clinical practice. Bioelectrical impedance analysis is a simple and practical method of body composition assessment, but fluid and electrolyte imbalances and the presence of edema or ascites, usually found in hospitalized patients, restricts its use in this population [2,3]. Anthropometric measurements usually assess muscle mass indirectly, through arm-muscle circumference and muscle area of the arm [4]. The only muscle that can be almost directly measured is the adductor pollicis muscle. This muscle has an easily accessible
http://dx.doi.org/10.1016/j.clnu.2014.11.006 0261-5614/© 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism.
Please cite this article in press as: Gonzalez MC, et al., Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients, Clinical Nutrition (2014), http://dx.doi.org/10.1016/j.clnu.2014.11.006
2
M.C. Gonzalez et al. / Clinical Nutrition xxx (2014) 1e5
location in the hands. The first studies focused its evaluation for muscle function through ulnar electric stimulation, which is not ideal [5,6]. It was later used as an anthropometric parameter [7,8] and reference values were recently published for healthy subjects [9]. Subjective global assessment (SGA) [10] is one of the most used techniques for nutritional assessment in clinical practice. Due to the lack of a gold standard for nutritional assessment, it is also used to validate new methods [11]. The use of adductor pollicis muscle thickness (APMT) in the clinical practice is still scarce, but recent studies have shown a positive association of APMT with the estimation of the total muscle compartment [12] and also a negative relationship with mortality [13]. The relationship between APMT and nutritional parameters has been investigated in surgical patients [14,15]. However, both studies were carried out with small samples and no distinction was made between gender and APMT measurement. In addition, SGA was used as the reference method to assess malnutrition in only one of the studies [14]. The objective of this study is to test the validity of APMT as a nutritional parameter in a sample of hospitalized surgical patients, using SGA as the gold standard method to assess nutritional status. 2. Materials and methods A sample of 361 patients of both genders, older than 18 years, hospitalized in a surgical ward in an academic hospital in Southern Brazil was studied in 2009. Patients who had been hospitalized or submitted to surgery in the last 30 days or were not able to be evaluated due to the presence of local venous accesses were excluded. Patients were invited to participate in the study and signed an informed consent term. A questionnaire was applied to assess information such as: age, gender, marital status and height. Weight was obtained from patient's chart records or referred by patient or their relatives because most of patients were in postoperative period, and were not allowed to be weighted. Dominant hand side and presence of injury and/or fracture in any of the hands were also considered relevant data, as they could influence the final results of measurements. The socioeconomic level was evaluated according to the crite~o Brasileira de Empresas de Pesquisa rion adopted by the Associaça (ABEP) [16]. This criterion is based on the property of domestic consumption assets, the presence of house servants and the education level of the head member of the family, allowing the classification of the patient in five groups e from the richest (A) to the poorest (E). The medical history (changes in weight, dietary intake and functional capacity, gastrointestinal symptoms with nutritional impact, and metabolic stress of present disease) and a brief physical examination were also assessed to determine the nutritional assessment by SGA, as firstly described by Detsky [10]. Individuals were classified in one out of three groups: A e eutrophic; B e moderate or suspected malnourishment; and C e severely malnourished. A Lange® skinfold caliper was used to assess APMT. The APMT measurement was performed using a standardized position: patients sat in the bed with both arms relaxed and elbows in a 90degree angle, approximately, with the hands laying over the legs. Adductor pollicis muscle thickness was measured by skinfold caliper, in the vertex of an imaginary triangle formed by the extension of the thumb and the index finger, under a continuous pressure of 10 g/mm. According to previous study [9], the mean of three measurements was used. Dominant APMT (DAPMT) values were considered from the right hand of right-handed subjects, and from the left hand of the left-handed ones. Non-dominant APMT
(NDAPMT) values were considered from the measurements of the opposite hand. Examiners were trained and standardized using acceptable technical errors of measurement calculated based on Habicht's publication [17]. Sample size estimation was made accordingly to two previous Brazilian studies using SGA in hospitalized patients [18,19]. These studies found malnutrition prevalence from 33% to 38%. Using such values and a confidence level of 95%, it was established that the necessary number of patients to be evaluated was 362. Reference values used in this study were obtained from a sample of 300 healthy subjects, grouped by gender and age, published previously by our group [9]. The 5th percentile in each age and gender group was considered the cut-off value for healthy subjects. Values lower than the 5th percentile were considered depletion of muscle mass by APMT. The normal distribution of variables was tested through Skewness and Kurtosis tests. ANOVA test was performed, and Bonferroni's test was also used to verify the difference of APMT values among nutritional status categories. Association between malnutrition diagnosis by SGA and depletion of muscle mass by APMT was tested by c2 test. Pearson's correlation test was made to assess the correlation between the several variables and APMT, as well as a multivariate linear regression (backward selection) to identify possible determinant factors of APMT in this group of patients. Specificity and sensitivity of APMT to identify malnutrition (B or C in the SGA) were tested in comparison to SGA diagnosis (here, considered the gold standard). All analyses were performed by Stata program (version 12.0), and the significance level considered for all tests was 5%. The present study was submitted and approved by the Ethics rdia de Pelotas, a hospital in Committee of Santa Casa de Miserico southern Brazil. 3. Results A sample of 361 patients who had APMT evaluated in at least one hand was studied. The sample characteristics are presented in Table 1. Most of the patients (60.2%) were women. The average age of the sample was 49.6 ± 17.8 years old, with no significant difference between genders. Most of the patients were Caucasian (78.4%) and lived with a partner (53.5%). The mean usual weight was 68.4 ± 13.8 kg, and men showed higher weight than women (72.08 ± 14.4 kg and 65.9 ± 12.8 kg,
Table 1 Sociodemographic characteristics of 361 surgical patients. Variables Gender Male Female Age (years) 18e30 31e59 60 Socioeconomic status D/E C B Marital status Single Married/Committed Widow/Widower Separated Race Caucasian Non-caucasian
n (%) 144 (39.8) 217 (60.2) 68 (18.8) 184 (51.0) 109 (30.2) 117 (32.4) 200 (55.4) 44 (12.2) 92 193 40 36
(25.5) (53.4) (11.1) (10.0)
283 (78.4) 78 (21.6)
Please cite this article in press as: Gonzalez MC, et al., Adductor pollicis muscle: A study about its use as a nutritional parameter in surgical patients, Clinical Nutrition (2014), http://dx.doi.org/10.1016/j.clnu.2014.11.006
M.C. Gonzalez et al. / Clinical Nutrition xxx (2014) 1e5 Table 2 Type of surgery performed in 361 patients of the sample. Number of patients (%) Digestive system surgery Gynecological and breast surgery Urgent abdominal surgery Post-surgery complications Wall surgery Genitourinary surgery Thoracic area surgery Other procedures
119 58 55 29 25 14 8 53
(33.0) (16.1) (15.2) (8.0) (6.9) (3.9) (2.21) (14.7)
3
Table 4 Pearson correlation between the adductor pollicis muscle thickness (dominant and non-dominant sides) and age and other anthropometric variables. All the correlations were significant (p < 0.05).
c
DAPMT NDAPMTd a b c d
Age
Weight
Height
BMIa
SGAb
0.28 0.28
0.52 0.56
0.19 0.18
0.44 0.50
0.61 0.60
BMI e body mass index. SGAe nutritional status according to subjective global assessment. DAPMT e dominant adductor pollicis muscle. NDAPMT e non-dominant adductor pollicis muscle 7.
respectively, p < 0.001). Most of the patients (68.4%) were considered well-nourished by SGA, but 87 patients (24.1%) were considered suspects or moderately malnourished (SGA B) and only 7.5% of the patients were considered severely malnourished (SGA C). Most of the patients in this ward were submitted to biliary tract surgery (14.4%). The most common procedure was cholecystectomy and exploratory laparotomy (14.1%). All procedures are specified in Table 2. APMT values were also compared according to the nutritional status categories (Table 3). DAPMT and NDAPMT values were significantly different according to the nutritional status categories (p < 0.001). According to Bonferroni test, it was found a significant difference in DAPMT and NDAPMT values among patients who were considered well-nourished (SGA A) when compared to patients with moderate (SGA B) and severe malnutrition (SGA C). However, no significant difference was found among values in patients who had moderate and severe malnutrition. It was also detected a significant linear trend in DAPMT and NDAPMT values, according to nutritional status categories: the worse the nutritional status was, the lower the DAPMT and NDAPMT values were found (p < 0.001). A strong association of malnutrition assessed by SGA and depletion of muscle mass by APMT was found. Patients with APMT below the 5th percentile had a greater risk of being malnourished, as assessed by SGA, for both dominant and non-dominant sides (RR ¼ 3.99; 95%CI: 3.19e5.00; p < 0.001; and RR ¼ 3.92; 95%CI: 3.10e4.96; p < 0.001; respectively). Sensitivity and specificity of DAPMT and NDAPMT were also tested, using their 5th percentile value to identify malnutrition and having SGA as the gold standard. APMT measurements presented low sensitivity (34.9% and 37.7% for DAPMT and NDAPMT, respectively), but a high specificity (98.7% and 97.8% for DAPMT and NDAPMT, respectively) to rule out malnutrition diagnosis in wellnourished patients. In patients with DAPMT higher than the healthy cut-off values, only 23.2% were malnourished (negative predictive value ¼ 76.8%). When a DAPMT lower than 5th
percentile value was found, 92.5% of patients were malnourished (positive predictive value ¼ 92.5%). For the NDAPMT, positive and negative predictive values were 88.9% and 77.3%, respectively. Correlations between DAPMT and NDAPMT and other variables, such as gender, age, weight, height, BMI and SGA were evaluated and presented in Table 4. All correlations with those variables were significant (p < 0.05). The highest positive correlations were obtained using DAPMT or NDAPMT associated with usual weight (r ¼ 0.52 and 0.56, respectively). The highest negative correlations were obtained using DAPMT or NDAPMT associated with SGA (r ¼ 0.61 and 0.60, respectively). Multivariate linear regression is presented in Table 5. Nutritional status was considered the most important determinant factor of DAPMT and NDAPMT, even after adjustment for other variables such as gender, age and weight. Patients classified as suspects or moderately malnourished (SGA B) presented a decrease of 4.59 and 4.56 mm in DAPMT and NDAPMT, respectively. An even larger decrease was found in patients classified as severely malnourished (SGA C): DAPMT and NDAPMT were 6.51 and 6.14 mm lower, respectively, when compared to values found in patients who were considered well-nourished.
Table 3 Median and interquartile range values of adductor pollicis muscle thickness (dominant and non-dominant sides) according to the nutritional status and gender.
Table 5 Multivariate linear regression of the adductor pollicis muscle thickness and gender, age, weight and nutritional status (backward selection).
DAPMTa Female c
e,f
4. Discussion This study assessed the relationship between APMT and nutritional parameters in surgical patients from a hospital located in Pelotas, Southern Brazil. It was found that patients with APMT below the 5th percentile in comparison to the general population had a greater risk of being malnourished according to SGA. APMT showed low ST in the prediction of malnutrition, though a high SP (more than 90%) was found. The adductor pollicis is the only muscle in the body which allows an almost direct and adequate assessment of its thickness, due to its easily accessible location, in a well-defined anatomic position, and minimal thickness of subcutaneous layers of adipose tissue surrounding it [7,8]. Most of previous studies use its function assessment through electric stimulation of the ulnar nerve.
NDAPMTb Male e,f
Female e,f
DAPMTa,b Male e,f
22 (20; 25) 25 (24; 28) 23 (21; 25) 26 (25; 28) SGA A (nd ¼ 148/84) e e SGA B (n ¼ 45/36) 18 (16; 20) 19.5 (16.5; 22) 18e (16; 20) 18e (15.5; 20) SGA C (n ¼ 11/14) 17f (15; 18) 18f (15; 20) 16f (12; 18) 16f (15; 20) e,f : In the same gender, same letter means a significant difference of DAPMT/ NDAPMT values among the SGA categories (p < 0.05). a DAPMT: dominant adductor pollicis muscle thickness of patients. b NDAPMT: non-dominant adductor pollicis muscle thickness of patients. c Subjective Global Assessment. d n: number of female/male.
Gender Age (years) Weight (kg) SGAe B SGAe C a b c d e
NDAPMTc,d
b
EP (b)
p
b
EP (b)
p
1.55 0.02 0.10 4.59 6.51
0.36 0.01 0.01 0.45 0.69
