Association of Adductor Pollicis Muscle Thickness and Handgrip Strength with nutritional status in cancer patients

Background and aim Malnutrition is common in patients with cancer, and its early diagnosis can reduce or prevent further complications and improve the clinical and nutritional prognosis. Adductor Pollicis Muscle Thickness (APMT) and Handgrip Strength have been explored in this population to identify a reduction in strength and muscle mass prior to the use of conventional methods. We aimed to correlate APMT and Handgrip Strength with conventional anthropometric variables in cancer patients and verify their association with nutritional status as determined by the Patient-Generated Subjective Global Assessment (PG-SGA). Methods A cross-sectional study was conducted with 80 patients diagnosed with cancer who were candidates for surgery. Nutritional status was obtained from the PG-SGA. Conventional anthropometric measurements were taken, as well as APMT and Handgrip Strength. Pearson’s correlation analysis and multivariate linear regression were applied to detect the influence of variables on APMT and HGS. A significance level of 5.0% was considered. Results A high prevalence of malnutrition and the need for dietotherapic intervention was found, identified by the PG-SGA. Correlations between APMT and Handgrip Strength with anthropometric variables and with the PG-SGA score were observed. After regression adjustments, the variables that interacted with APMT were TSF and AC, and the PG-SGA score, corrected Muscle Arm Area (CAMA), and age interacted with Handgrip Strength. Conclusion Correlations between anthropometric measurements and the PG-SGA score with APMT and Handgrip Strength were observed, even after adjusting for age and sex. These associations demonstrate that APMT and Handgrip Strength can be used with criterion in patients with cancer as complementary methods to evaluate nutritional risk and the need for nutritional intervention.

both sexes, aged �20 years, with a confirmed clinical diagnosis of cancer (ICD: C00 A C97), regardless of the type and location of the tumor, who were candidates for surgery, could be evaluated within 48 hours of hospital admission, and were able to answer the PG-SGA and perform the handgrip strength test. We evaluated those patients admitted to the General Surgery Unit from March 2017 to April 2018.
Patients were excluded if they had edema in the hands, were isolation by aerosols, could not walk independently, or did not submit complete and/or reliable information. A total of nine patients were excluded.
Handgrip Strength. For the evaluation of handgrip strength, we used the Jamar Hydraulic Hand Dynamometer with a scale from 0 to 90 kg/f and a resolution of 2 kg/f, and both handles were adjusted in the second position. The patient was instructed to sit in a chair with arms, place the assessed arm beside the body with the elbow at a 90˚angle, and lean the trunk in the chair without resting the body or receiving help from the assessor. Three measurements were performed on the dominant handgrip strength and non-dominant handgrip strength for approximately 5 seconds with a 1-minute interval between them [31]. The assessors provided motivational stimulus throughout the test. The maximum measure of both hands was considered, and the cutoff point was proposed by the European Working Group on Sarcopenia in Older People (EWGSO), according to sex (men: <27 kg/f; women: <16 kg/f) [32].

Statistical analysis
The normality of the distribution of quantitative variables was tested using the Kolmogorov-Smirnov test. Correlations among variables were analyzed by the Pearson correlation coefficient. Multivariate linear regression was used to detect the influence of selected independent variables on APMT and handgrip strength (dependent variables) in both hands. The variables sex, AC, TSF, dominant handgrip strength, non-dominant handgrip strength, and the PG-SGA score were included in the APMT model. In the HGS model, the variables sex, age, DAPMT, NDAPMT, CAMA, AC, CC, and the PG-SGA score were included. Sex was included for both APMT and handgrip strength, due to differences in strength and muscle mass. Data were analyzed with the software SPSS 21.0. A 5.0% significance level was used for all tests.

Ethical aspects
This study was approved by the Research Ethics Committee of the Federal University of Espirito Santo, under CAAE no. 27954014.0.0000.5060. Patients participated voluntarily and provided written informed consent.

Results
Eighty patients were assessed, with an average age of 60.8 ± 13.5 years. Of these, 56.3% (n = 45) were men, 60.0% (n = 48) were older adults, 51.2% (n = 41) were non-white, and 76.2% (n = 61) had tumors in the gastrointestinal tract (GIT). According to the BMI, 36.3% (n = 29) of patients had eutrophy, while PG-SGA identified 60.0% (n = 48) with some degree of malnutrition (B+C). The PG-SGA score showed that 70.0% (n = 56) of patients had a score equal to or above 9 points (Table 1). Table 2 shows the frequency of adequacy of APMT and handgrip strength. As for the APMT of both hands, most patients were classified with malnutrition (>40.0%). Dominant handgrip strength proved suitable for most patients (60.0%). The non-dominant handgrip strength was adequate for 50.0% of the patients and inadequate for the other 50.0%.
Significant correlations of the DAPMT with AC, TSF, PG-SGA score, dominant handgrip strength, and non-dominant handgrip strength were found. NDAPMT was correlated significantly with the PG-SGA score, dominant handgrip strength, and non-dominant handgrip strength. As for handgrip strength, we observed significant correlations with DAPMT, NDAPMT, age, CAMA, and the PG-SGA score. Non-dominant handgrip strength was correlated significantly with DAPMT, NDAPMT, age, CAMA, AC, CC, and the PG-SGA score ( Table 3). Table 4 shows the results of linear regression for APMT of both hands. The choice to keep TSF in the NDAPMT model was made because this measure is more preserved on the nondominant side. It is a measure used to assess energy reserve; therefore, its reduction indicates depletion of muscle reserves. For the DAPMT, after adjustment for age and sex, AC remained in the final model (β 0.61, 95% CI 0.15-0.49, p < 0.001), explaining 54% of the measure For the NDAPMT, TSF remained in the final model (β 0.33, 95% CI 0.02-0.34, p = 0.023), explaining 44% of value.  The results of linear regression for APMT of both hands indicated that after adjustment for age and sex, the variables CAMA, PG-SGA score, and age remained in the final model, explaining 81% of result. As to the non-dominant handgrip strength, age remained in the model, explaining 77% of the measure (Table 5).

Discussion
The main findings of this study were a high prevalence of malnutrition, indicated by PG-SGA and APMT, and the need for dietotherapic intervention according to the PG-SGA score. Correlations of APMT and handgrip strength with classic anthropometric variables and with the PG-SGA score were observed. In the regression model, AC was associated with DAPMT and TSF with non-dominant handgrip strength when adjusted for age and gender. The dominant handgrip strength was associated with CAMA, PG-SGA score, and age; however, nondominant handgrip strength was only associated with age. High rates of malnutrition are found in cancer patients, mainly located in the GIT [4,10,22,33], the region most affected in this study. Our results showed 60.0% of patients with some degree of malnutrition, (B+C) by PG-SGA, which corroborates previous studies [4,5,10,22,33].
In addition to metabolic changes generated by the tumor, patients with tumors in the GIT often show increased symptoms with a nutritional impact, significant weight loss, reduction in food consumption, and reduced functional capacity, conditions that raise the PG-SGA score, which indicates the need for nutrition intervention [10,34] observed in this study.
The results were already expected due to the severity of the disease and because the patients were evaluated in a tertiary referral hospital with late diagnosis and treatment. Other factors related to the high prevalence of malnutrition and the need for nutrition intervention (score �9) were the advanced age of most of the group, location of cancer in the GIT, presence of inflammation, and cancer staging (the latter factors were not assessed in this study).
Malnutrition measured by objective methods was also confirmed. Significant correlations were observed between the DAPMT, AC, TSF, PG-SGA score, and handgrip strength of both hands, while the NDAPMT was correlated significantly with the PG-SGA score and handgrip strength of both hands. After the regression adjustments, the results indicated that the variables that most interacted with DAPMT and NDAPMT were AC and TSF, respectively.  These findings agree with other studies that assessed candidates for surgery [13,14,35]. APMT has been indicated as a promising measure in the diagnosis of malnutrition for being able to reveal changes in the muscle composition of the whole body, indicating early changes related to malnutrition and the recovery of nutritional status [16,26].
As to the results found after regression analysis, it is possible that they were achieved because TSF and AC are indicative of peripheral fat mass and total-body skeletal muscle mass [35,36,37], besides being measures of the same nature [13]. Patients with cancer tend to have a highly catabolic metabolism, which would result in decreased AC and TSF due to the increase in proteolysis and lipolysis, rapid weight loss, and severity of the disease.
However, APMT should not be used in isolation due to the absence of a cutoff point for this population, but as a direct measure, it has the advantages of not requiring adjustment formulas, being the only muscle that allows proper thickness evaluation for its anatomical definition, and being flat, which may facilitate nutritional evaluation by a trained assessor and its inclusion in clinical practice [15,16,30,38].
Dominant handgrip strength was associated with age, CAMA, PG-SGA score, and APMT of both hands, while non-dominant handgrip strength was correlated with age, CAMA, AC, CC score, and APMT of both hands, with age, PG-SGA score, and CAMA for dominant handgrip strength and age for non-dominant handgrip strength remaining in the model after adjustments for regression. Differences between dominant handgrip strength and non-dominant handgrip strength are expected and have already been described. In general, the dominant hand performs, on average, 10.0% better then the non-dominant hand in both sexes [39].
Handgrip strength is a frequently used, validated, non-invasive, rapid, simple, and clinical method for the measurement of muscle activity [19,20]; however, there is also a cutoff point defined for this population and standardization of the measurement technique, which can affect the comparison of results. The relationship between age and handgrip strength is already known and appears with the loss of strength and lean body mass as age progresses, causing older adults to present typically lower handgrip strength than young and middle-aged adults [40,41], which is justified in this study, since it has a larger number of older adults. Zhang et al. [21] found that the handgrip strength decreased as age increased and that the decrease in handgrip strength was twice as fast in older adults.
Although handgrip strength, PG-SGA score, and CAMA assess different parameters, they are related to strength, lean body mass, and nutritional status, since malnutrition generates changes in the muscle compartment, measured here by CAMA, and these changes can bias the estimation of functional capacity by PG-SGA, which may explain the results found. Thus, other studies have shown the association of low functionality, evaluated by handgrip strength, with nutritional status [29][30][31].
These findings may be influenced by reduced muscle mass and increased body fat, which occur throughout the aging process and with excess weight gain, in addition to changes in body composition in patients with cancer [42,43,44]. Reductions in muscle strength, mass, and function are usually attributable to a decrease in muscle size; however, evidence has shown a new scenario, known as myosteatosis, characterized by fat infiltration into the muscle [43,45,46].
It is possible that the increase in corporal fat reduces the capacity for muscular power generation, which is more closely related to functional capacity than muscular force [45].
This hypothesis should be considered since there was a predominance of older adults, and a significant proportion of the patients were classified as well-nourished and/or overweight by PG-SGA and BMI. Thus, the absence of a cutoff point for patients with cancer limits the interpretation of the results beyond the absence of an evaluation by computed tomography or magnetic resonance imaging that could safely indicate the body composition of the patients evaluated [47].
Our study is relevant for using the PG-SGA score, as the global score can exhibit greater interobserver variability, while the PG-SGA score is an objective and continuous method comprising the sum of all the questions.
The study has some limitations because it is transverse, includes unique measures, and includes patients with several types of tumors; therefore, it is not possible to determine the causal relationship between the variables or to extrapolate the results. Another limitation is the lack of data on cancer staging, which is because the hospital is not specialized in the treatment of patients with cancer, mainly receiving patients for surgical correction.
Another possible limitation is the use of calipers to take measurements. Discrepancies can be associated with the error at the moment the correct anatomical point is pinched, or in the calibration of the apparatus, as well as in the variability between evaluators. To correct this problem, the evaluators were well trained, and the caliper was calibrated often.
However, studies that can confirm and indicate the use of APMT and handgrip strength in surgical patients with cancer are necessary. The results found here clarify associations of APMT and handgrip strength with the instruments used in hospitals, suggesting their implementation in the clinical routine.

Conclusion
Correlations between anthropometric measurements and the PG-SGA score with APMT and handgrip strength were observed, even after adjusting for age and sex. These associations demonstrate that APMT and handgrip strength can be used with criterion in patients with cancer and can complement the evaluation of nutritional status and the need for nutritional intervention.
However, new studies must be carried out with this population to define specific cutoff points for adults and older adults, as well as longitudinal studies to indicate causal relationships and the changes in measures of APMT and handgrip strength that occur during the hospital stay.