Masseter muscle parameters can function as an alternative for skeletal muscle mass assessments on cross-sectional imaging at lumbar or cervical vertebral levels

Introduction

Head and neck squamous cell carcinoma (HNSCC) is the seventh most common type of cancer worldwide with 890,000 new cases and 450,000 deaths in 2018 (1). About two thirds of patients present with advanced stage disease. HNSCC at this stage is associated with a poor 5-year overall survival (OS) of less than 50%. There is a need for accurate prognostic factors to tailor treatment for HNSCC patients, and sarcopenia is emerging as a novel candidate in HNSCC (2-4).

Sarcopenia is defined as the loss of skeletal muscle mass (SMM) and muscle function (5), although measurements of only SMM are often used in literature. Sarcopenia is often the result of cancer cachexia (6,7).

Patients with HNSCC are at an increased risk for cancer related cachexia and sarcopenia. This is partly due to dysphagia caused by tumor localization or its treatment and side effects thereof. Moreover, patients with HNSCC might present with underlying malnutrition caused by poor diet, tobacco use or alcohol abuse (8,9). Low SMM cancer patients treated with surgery are at risk for complications and decreased survival (10). In HNSCC, low SMM has been associated with and increased risk of surgical complications and cisplatin dose limiting toxicity and with decreased survival (11-13). Low SMM can be considered as an emerging biomarker for the clinical setting in HNSCC patients (14).

While the gold standard for total SMM assessment is full body imaging, earlier research has shown that the muscle cross-sectional area (CSA) measured on a single abdominal cross-sectional slice at the level of the third lumbar vertebra (L3) on computed tomography (CT) imaging can provide accurate estimates of patient’s total SMM (15). Unfortunately, patients treated for head and neck cancers do not usually have imaging performed at this level. Therefore, a method was developed to assess SMM on a single CT slide at the level of the third cervical vertebra (C3) in head and neck cancer patients (16). However, CSA assessment at this level may be impaired by extension of primary tumor and/or lymph nodes or previous treatment. Moreover, accurate assessment is time consuming (17,18). There is a need for a reliable index muscle that is consistently present on routine imaging, is rarely impacted by disease or treatment and is quick and easy to characterize using commonly used imaging software. For this purpose, we propose the masseter muscle. The masseter muscle has been shown to be adequate in determining SMM and predicting mortality in other fields of medicine (19-21).

The purpose of this study was firstly to investigate whether masseter muscle quantity measures correlate with the CSA at C3 and L3. Secondly, the study sought to investigate the association between these masseter muscle parameters and OS.

Methods

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Medical Ethical Research Committee of the University Medical Center Utrecht (approval ID 17-365/C) and individual consent for this retrospective analysis was waived.

Patient and study design

We reviewed patients with newly diagnosed, pathologically proven HNSCC who underwent a whole body FDG-PET/CT-scan between 2010 and 2018 at the University Medical Center Utrecht, the Netherlands (UMCU). Indications to perform a whole body FDG-PET/CT-scan in our institute were clinical suspicion of advanced (III/IV) stage at presentation, carcinoma of unknown primary tumor, recurrent disease and second primary tumor. Patients with previous HNSCC or second primary tumor were excluded. Patient scans who were incomplete, of insufficient quality or incompatible with current imaging software were excluded from further analysis.

Patient factors with known or expected relation to HNC outcome measures or development of sarcopenia were collected: age at diagnosis, gender, histological diagnosis, comorbidities scored using the Charlson Comorbidity Index (CCI) and the ACE-27 score, tumor site and tumor staging according to the 7th edition of the UICC TNM classification system, human papillomavirus (HPV) status for oropharyngeal carcinomas, weight loss 6 months before diagnosis and treatment regimens.

Radiological assessment

Segmentation of muscle tissue at the level of C3 and L3 was manually performed using the commercially available software package SliceO-matic (Tomovision, Canada). For analysis of the CSA at the level of C3, a standard method for slide selection was used, where the first slide to show the entire vertebral arc and the transverse and spinous process when scrolling in a cephalad to caudad direction was selected. Skeletal muscle tissue was identified using Houndsfield Unit (HU) range settings from −29 to 150 HU and the outer contours of the sternocleidomastoid and paravertebral muscles were traced manually. The CSA at the level of C3 was determined as the sum of delineated areas of the paravertebral muscles and both sternocleidomastoid muscles within a HU range of −29 to 150 HU in cm².

For analysis of the CSA at level L3 the muscle groups analyzed were the psoas, paravertebral and the anterior abdominal wall.

For assessment of the masseter muscle, Intellispace (version 14, Phillips, Netherlands) was chosen for its ability to measure the volume of a selected structure (e.g., the masseter muscle) using the TumorTracking feature which allows for rapid tissue volume assessment. The following muscle parameters were chosen and measured: masseter cross-sectional area (MCSA), masseter muscle volume (MV), masseter muscle maximum thickness (MT). Furthermore, measurements of muscle quality defined by the Hounsfield unit (HU) and expressed as the average HU of all measured tissue (HU_tot) and in a region of interest (HU_ROI) were collected. MCSA was measured at the level of the dens of the second cervical vertebra (Figure 1).

Figure 1 Example of masseter delineation using Intellispace.

Coronal tilt alignment was made according to a tangent running through the dens and hard palate. MCSA was measured by outlining the outer surface of masseter after which IntelliSpace automatically calculated the surface area (mm²), a method independent of the HU value of the defined area. MV (cm³) and HU_tot were automatically calculated after segmenting the entire muscle (Figure 2).

Figure 2 A 3D example of the measuring of masseter volume using the TumorTracking feature in Intellispace.

MT of the masseter was determined using the measuring-tool included in Intellispace (mm). HU_ROI was determined in a 1-centimeter diameter circle on the same level as MCSA.

Since the state of a patient’s teeth may impact masseter function and size each patient was examined for the presence of dental elements (22). Dental status was scored as follows: [0] no missing dentition, [1] one or more missing teeth, [2] total absence of dentition.

Presence of scattering cause by (dental) implants was scored as follows: [0] no scattering present, [1] slight scattering present, [2] significant scattering present. Measurements were performed bilaterally for each patient and an average was calculated and used for further analysis.

Earlier research has shown that there is excellent agreeability between image scoring software programs used for measuring CSA (23). Therefore, we found it acceptable to use the two programs independently and compare the data.

Body composition measurements

Weight and height were recorded during a patient’s first consultation at our out-patient clinic and used to calculate body mass index (BMI) and body surface area (BSA) using the Mosteller formula (24). Lumbar skeletal muscle index (LSMI), cervical skeletal muscle index (CSMI) and masseter skeletal muscle index (MSMI) were calculated by dividing the corresponding patient’s CSA values by patient’s squared height. There is, to our knowledge, no scientific consensus on a cut-off point for MSMI. We therefore designated patients present in the lowest quartile of MCSA for their specific gender as “low MSMI”.

Overall survival

The status of the patient (alive/deceased) was acquired from the UMCU electronic patient data system on date of last follow-up. OS was defined as the time between the date of histologic diagnosis and death, or date of last follow-up. UMCU patient system is linked to the provincial government register and is updated continuously for patients living in the Utrecht province. Patients were considered alive if no date of death was available on date of last follow-up or if there was no physician note reporting on their death. Cause of date was determined by physician’s notes.

Statistical analysis

SPSS 26 for Windows (IBM, Armonk, NY, USA) was used for analysis. Descriptive statistics were calculated with the continuous variables presented as mean (standard deviation) or median (interquartile range). Discrete variables were displayed as counts (percentages). Normality was investigated by using the Kolmogorov-Smirnov test. Characteristics and muscle measurements were analyzed using independent-samples t-test for normally distributed variables, independent-samples test for skewed variables and Fisher’s exact test or Pearson’s chi-squared test for categorical variables. Spearman correlation coefficients were calculated to establish the relationship between L3 measurements, C3 measurements and masseter measurements. A correlation coefficient of (−)0.8 to (−)1 was interpreted as a very strong correlation, (−)0.6–0.8 as strong, (−)0.4 to (−)0.6 as moderate, and (−)0.2 to (−)0.4 as a low correlation (25). Radiological measurements and patient characteristics were analyzed using Cox regression proportional hazards first as univariate analysis. Variables with a P value lower than 0.05 alongside lumbar and cervical muscle cross sectional area (CSA) were included for multivariate analysis. The backward step-method was chosen for multivariate analysis. The influence of MSMI classification and low LSMI using the cut-off established by Wendrich et al. (11) on OS was evaluated using Kaplan-Meier curves and associated Log-Rank tests.

Results

Search and inclusion

In total 139 patients who had undergone a CT-scan were screened for study viability. Of these patients 15.2% (n=21) had (partially) missing imaging and were subsequently excluded. Furthermore, in 3.6% (n=5) of included patients the available imaging was of insufficient quality for analysis either due to low resolution or poor image quality. In total, 113 whole body FDG-PET/CT-scans were included for further image analysis.

Patient, tumor and treatment characteristics

In total 99 patients were included, with a median age of 61.69 (IQR, 56.0–68.40) years. Of the included patients, 81 (71.7) were male. A minority of patients had no history of alcohol consumption (n=32, 28.3%) or smoking (n=13, 13.1%). Forty patients (40.4%) were categorized as having normal weight based on BMI score. Most patients presented tumorous disease localized in the oropharynx (n=66, 66.7%) of which 16 (16.2%) were HPV-positive. Most patients presented with a clinical stage IV disease (n=64, 64.6%) and patients were most commonly treated with a combination of radiotherapy and systemic therapy (n=53, 46.9%).

Twenty-six patients were designated as “Low MSMI” and seventy-three as “Normal MSMI”. There was a statistically significant difference between these groups for LSMI, C3 CSA, CSMI and BMI (P=0.015, 0.019 and 0.014, respectively).

See Tables 1,2 for patient, tumor and treatment characteristics and the comparison of these characteristics between the low MSMI and normal MSMI groups.

Body composition measurement

Table 3 shows a significant difference based on sex for HU_tot and HU_ROI (both P=0.049), BSA, L3 CSA, C3 CSA, MCSA, MV, MT, L3 SMI and MSMI (all P<0.001). There was no significant difference based on sex for BMI at time of diagnosis (Table 3).

Masseter left-right deviations, the effect of scattering on deviations and the impact of dental status on masseter parameters

Generally, there was some amount of left-right difference present. The deviations are shown as median (percentage of average masseter characteristic). The median left-right difference for MCSA, MV, MT, HU_tot and HU_ROI were 36.48 mm² (9.17%), 1.04 cm³ (5.76%), 1.0 mm (7.26%), 4.80 HU (4.15%) and 9.40 HU (16.80%) respectively (Table 4). There was a significant difference in left-right deviation of median MV and HU_tot for different scattering scores (P<0.001; Table 5). MV and HU_tot had a significant negative relationship with dental score (P=0.011 and 0.001, respectively; Table 6).

Correlation between masseter parameters and muscle mass measured at C3 and L3

All masseter and muscle mass parameters had a highly significant correlation with each other (P=0.001 to P<0.001). The strongest correlation was between L3 CSA - C3 CSA (r=0.708), followed by C3 CSA–MV (r=0.639) and L3 CSA–MV (r=0.586). MT was moderately correlated to C3 CSA (r=0.509) and L3 CSA (r=0.431) and had low correlation with LSMI (r=0.361). MCSA had low to moderate correlation with L3 CSA and C3 CSA (r=0.451, 0.586). MSMI had a low correlation with LSMI and L3 CSA (r=0.278, r=0.215), but a moderate correlation with C3 CSA, r=0.415, Table 7).

Univariable Cox regression analysis of the effect of body composition indicators on OS

A selection of body composition indicators were tested using Cox univariate regression analysis. In univariate analysis, MCSA, low MSMI classification and CCI were all significantly associated with OS (Table 8). For variables that were strongly correlated or dependent on each other (e.g., MSMI, low MSMI and MCSA) the variable with the lowest P value was included in the multivariate analysis. As to not exceed the >10 events per variable rule four variables were included (26). This left MSMI-classification, C3 CSA, LSMI and CCI as included variables. Low MSMI and CCI score remained as the only independent predictors of OS (HR 2.227, P=0.009 and HR 1.338, P<0.001, respectively; Table 9).

Overall survival

There was a significant difference in OS for patients with low MSMI compared to normal MSMI [17.92 months (IQR, 11.64–57.09) versus 34.10 months (IQR, 15.54–62.03) log rank P=0.015; Figure 3]. There was no significant difference in OS between patients with low and normal LSMI using the previously established cut-off value of LSMI <43.2 cm²/m² (Figure 4) (11).

Figure 3 Kaplan-Meier curve displaying differences in survival probability based on MSMI classification. MSMI, masseter skeletal muscle index.

Figure 4 Kaplan-Meier curve displaying differences in survival probability based on LSMI classification using the pre-defined cut-off of 43.2 cm². LSMI, lumbar skeletal muscle index.

Discussion

Patients with head and neck cancers are at an increased risk of sarcopenia compared to patients with other types of cancer (8,9,27). Previous reports have established that measuring muscle mass at level L3 on CT-scans is a reliable method for assessing total body SMM. Unfortunately, scans at this lumbar level are rarely available in patients with HNSCC. Previously published findings by Swartz et al. (16) show that the CSA of skeletal muscles at level C3 strongly correlate with the CSA of skeletal muscle at level L3, indicating that this is a viable alternative method. However, determining the CSA at C3 is time consuming and can be impacted by either treatment (e.g., neck dissection) or disease (e.g., invading lymph node in the SCM). We therefore investigated to what degree masseter muscle parameters are associated with levels L3 and C3, and their relationship on OS. We found moderate to strong associations for most masseter parameters with muscle mass on level L3 and C3, with MV being the strongest followed by MCSA. Low MSMI was shown to be an independent predictor for OS in multivariate analysis.

We found that the scatter-score had a significant impact on MV and Masseter HU measurements. It stands to reason that scattering results in unreliable masseter HU-measurements, as scattering generally causes a larger spread of pixel values shown on imaging. The method we used to determine MV used the TumorTracking feature included in IntelliSpace which utilizes the pixel values recorded and inputs them into an algorithm to determine whether certain areas are related to each other. It follows that a larger spread in pixel-values decreases the reliability of the algorithm. Manual adjustment of the measured area was often required to fully include all masseter muscle tissue, although this too becomes unreliable when significant scattering is present. However, we found no significant relationship between scatter-score and MT, HU_ROI and MCSA (and subsequently MSMI) leaving these as viable options when significant scattering is present. Our included patient group had 3 (4.1%) patients with tumors in the oral cavity. These tumors were always unilateral with no significant ingrowth into the masseter muscle. Consequently, they did not significantly impact the masseter muscles. If significant impairment would be present, one solution could be that in the rare cases where the muscle is unilaterally significantly affected, a contralateral masseter measurement is counted twice.

Our findings are consistent with other studies which determine that MCSA predicts mortality in patients suffering from blunt trauma, traumatic brain injury or survival after carotid endarterectomy (19-21). However, differences between our study and earlier scientific reports should be noted. Oksala et al., Wallace et al. and Hu et al. all used the MCSA measured at 2 cm below the arcus zygomaticus. In our study we chose the first slice showing the dens of the C2 vertebra as our landmark as this was easily identifiable when scrolling in cephalad-to-caudad fashion. Secondly, whereas Wallace et al. and Hu et al. did not correct for head tilt, Oksala et al. adjusted their CT-scans for both sagittal and coronal head tilt. Based on a consensus discussion we chose to only adjust for coronal head tilt. Using our center’s patient positioning protocol, we expected very little to no sagittal tilt in our imaging.

We corrected the observed MCSA by dividing by squared body height to determine a masseter muscle mass index (MSMI). The masseter muscle characteristics are dependent on various factors such as dental status and craniofacial structure (22,28). MCSA was adjusted by body height, as it has been established that muscle mass corrected by body height is an accurate adjustment method for other CSA measurements (29). We found a significant difference in muscle mass and body composition indicators for groups based on MSMI. Furthermore, we found a significant difference in OS between patients classified as normal or low MSMI classification (P=0.015). In multivariate analysis low MSMI classification significantly predicted OS.

Another limitation of our retrospective design is that patient frailty and sarcopenia as defined by the European Working Group on Sarcopenia in Older People (EWGSOP) could not be assessed. Sarcopenia is diagnosed by evaluating muscle mass, muscle quality, muscle strength and physical performance (6). Further prospective studies are needed that correlate masseter findings with muscle strength (e.g., by grip strength) and physical performance (e.g., by the Short Physical Performance Battery and the Timed Up and Go-test).

Finally, whole-body PET-CT-scans are only performed in patients with clinical suspicion of advanced (III/IV) stage at presentation. We reason that this does not cause any significant bias in our study as Swartz et al. found no significant difference in C3 or L3 CSA between patients with traumatic injury and head and neck cancer allowing for extrapolation to both healthy patients and patients with malignant disease (14).

Conclusions

We conclude that several masseter muscle parameters, namely MV, MCSA and MT, are significantly correlated (varying from moderate to strong) with cross-sectional muscle area at cervical and lumbar level. Additionally, MSMI, defined as MCSA divided by the squared patient’s length in meters, proved to be an independent predictor for overall survival (HR 2.227), with other covariates for survival being: Low MSI-classification, C3 CSA, L3 CSA and CCI. Patients classified as having low MSMI had significantly decreased overall survival. In patients without cross-sectional imaging at level L3 or C3 or with impaired C3 measurements, masseter muscle parameters could serve as a (swifter assessable) alternative for SMM assessed by cross-sectional muscle area measurements at these vertebral levels. We recommend further studies to determine factors influencing masseter parameters as to formulate an improved method to correct for individual patient factors, e.g., dental status, previous dental disease, previous cancer treatment and facial morphologic features. Subsequently this research should correlate masseter parameters with muscle strength and physical performance.

Acknowledgments

Funding: None.

Footnote

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://dx.doi.org/10.21037/qims-21-43). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Medical Ethical Research Committee of the University Medical Center Utrecht (approval ID 17-365/C) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chow LQM. Head and Neck Cancer. N Engl J Med 2020;382:60-72. [Crossref] [PubMed]
2. Leemans CR, Snijders PJF, Brakenhoff RH. The molecular landscape of head and neck cancer. Nat Rev Cancer 2018;18:269-82. [Crossref] [PubMed]
3. Elicin O, Mahmut Ozsahin E. Head and neck cancer. Med Radiol 2018;91-126. [PubMed]
4. Wong A, Zhu D, Kraus D, Tham T. Radiologically Defined Sarcopenia Affects Survival in Head and Neck Cancer: A Meta-Analysis. Laryngoscope 2021;131:333-41. [Crossref] [PubMed]
5. Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Cederholm T, Cooper C, Landi F, Rolland Y, Sayer AA, Schneider SM, Sieber CC, Topinkova E, Vandewoude M, Visser M, Zamboni M. Writing Group for the European Working Group on Sarcopenia in Older People 2 (EWGSOP2), and the Extended Group for EWGSOP2. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing 2019;48:16-31. [Crossref] [PubMed]
6. Cruz-Jentoft AJ, Sayer AA. Sarcopenia. Lancet 2019;393:2636-46. [Crossref] [PubMed]
7. Fearon K, Strasser F, Anker SD, Bosaeus I, Bruera E, Fainsinger RL, Jatoi A, Loprinzi C, MacDonald N, Mantovani G, Davis M, Muscaritoli M, Ottery F, Radbruch L, Ravasco P, Walsh D, Wilcock A, Kaasa S, Baracos VE. Definition and classification of cancer cachexia: an international consensus. Lancet Oncol 2011;12:489-95. [Crossref] [PubMed]
8. Chasen MR, Bhargava R. A descriptive review of the factors contributing to nutritional compromise in patients with head and neck cancer. Support Care Cancer 2009;17:1345-51. [Crossref] [PubMed]
9. Almada-Correia I, Neves PM, Mäkitie A, Ravasco P. Body Composition Evaluation in Head and Neck Cancer Patients: A Review. Front Oncol 2019;9:1112. [Crossref] [PubMed]
10. Weerink LBM, van der Hoorn A, van Leeuwen BL, de Bock GH. Low skeletal muscle mass and postoperative morbidity in surgical oncology: a systematic review and meta-analysis. J Cachexia Sarcopenia Muscle 2020;11:636-49. [Crossref] [PubMed]
11. Wendrich AW, Swartz JE, Bril SI, Wegner I, de Graeff A, Smid EJ, de Bree R, Pothen AJ. Low skeletal muscle mass is a predictive factor for chemotherapy dose-limiting toxicity in patients with locally advanced head and neck cancer. Oral Oncol 2017;71:26-33. [Crossref] [PubMed]
12. Rier HN, Jager A, Sleijfer S, Maier AB, Levin MD. The Prevalence and Prognostic Value of Low Muscle Mass in Cancer Patients: A Review of the Literature. Oncologist 2016;21:1396-409. [Crossref] [PubMed]
13. Ansari E, Chargi N, van Gemert JTM, van Es RJJ, Dieleman FJ, Rosenberg AJWP, Van Cann EM, de Bree R. Low skeletal muscle mass is a strong predictive factor for surgical complications and a prognostic factor in oral cancer patients undergoing mandibular reconstruction with a free fibula flap. Oral Oncol 2020;101:104530 [Crossref] [PubMed]
14. Economopoulou P, de Bree R, Kotsantis I, Psyrri A. Diagnostic Tumor Markers in Head and Neck Squamous Cell Carcinoma (HNSCC) in the Clinical Setting. Front Oncol 2019;9:827. [Crossref] [PubMed]
15. Shen W, Punyanitya M, Wang Z, Gallagher D, St-Onge MP, Albu J, Heymsfield SB, Heshka S. Total body skeletal muscle and adipose tissue volumes: estimation from a single abdominal cross-sectional image. J Appl Physiol (1985) 2004;97:2333-8. [PubMed]
16. Swartz JE, Pothen AJ, Wegner I, Smid EJ, Swart KM, de Bree R, Leenen LP, Grolman W. Feasibility of using head and neck CT imaging to assess skeletal muscle mass in head and neck cancer patients. Oral Oncol 2016;62:28-33. [Crossref] [PubMed]
17. Ufuk F, Herek D, Yüksel D. Diagnosis of Sarcopenia in Head and Neck Computed Tomography: Cervical Muscle Mass as a Strong Indicator of Sarcopenia. Clin Exp Otorhinolaryngol 2019;12:317-24. [Crossref] [PubMed]
18. Bozkurt G, Elhassan HA, Mahmutoğlu AS, Çelebi İ, Mcleod RWJ, Soytaş P, Erol ZN, Sözen E. Neck Muscle Mass Index as a Predictor of Post-Laryngectomy Wound Complications. Ann Otol Rhinol Laryngol 2018;127:841-7. [Crossref] [PubMed]
19. Hu P, Uhlich R, White J, Kerby J, Bosarge P. Sarcopenia Measured Using Masseter Area Predicts Early Mortality following Severe Traumatic Brain Injury. J Neurotrauma 2018;35:2400-6. [Crossref] [PubMed]
20. Oksala NKJ, Lindström I, Khan N, Pihlajaniemi VJ, Lyytikäinen LP, Pienimäki JP, Hernesniemi J. Pre-Operative Masseter Area is an Independent Predictor of Long-Term Survival after Carotid Endarterectomy. Eur J Vasc Endovasc Surg 2019;57:331-8. [Crossref] [PubMed]
21. Wallace JD, Calvo RY, Lewis PR, Brill JB, Shackford SR, Sise MJ, Sise CB, Bansal V. Sarcopenia as a predictor of mortality in elderly blunt trauma patients: Comparing the masseter to the psoas using computed tomography. J Trauma Acute Care Surg 2017;82:65-72. [Crossref] [PubMed]
22. Newton JP, Yemm R, Abel RW, Menhinick S. Changes in human jaw muscles with age and dental state. Gerodontology 1993;10:16-22. [Crossref] [PubMed]
23. van Vugt JL, Levolger S, Gharbharan A, Koek M, Niessen WJ, Burger JW, Willemsen SP, de Bruin RW, IJzermans JN. A comparative study of software programmes for cross-sectional skeletal muscle and adipose tissue measurements on abdominal computed tomography scans of rectal cancer patients. J Cachexia Sarcopenia Muscle 2017;8:285-97. [Crossref] [PubMed]
24. Verbraecken J, Van de Heyning P, De Backer W, Van Gaal L. Body surface area in normal-weight, overweight, and obese adults. A comparison study. Metabolism 2006;55:515-24. [Crossref] [PubMed]
25. Akoglu H. User's guide to correlation coefficients. Turk J Emerg Med 2018;18:91-3. [Crossref] [PubMed]
26. Rowe AK, Kleinbaum DG, Koplan JP. Practical methods for public health practitioners. Am J Prev Med 2004;26:252-3. [Crossref] [PubMed]
27. Pressoir M, Desné S, Berchery D, Rossignol G, Poiree B, Meslier M, Traversier S, Vittot M, Simon M, Gekiere JP, Meuric J, Serot F, Falewee MN, Rodrigues I, Senesse P, Vasson MP, Chelle F, Maget B, Antoun S, Bachmann P. Prevalence, risk factors and clinical implications of malnutrition in French Comprehensive Cancer Centres. Br J Cancer 2010;102:966-71. [Crossref] [PubMed]
28. Grünheid T, Langenbach GE, Korfage JA, Zentner A, van Eijden TM. The adaptive response of jaw muscles to varying functional demands. Eur J Orthod 2009;31:596-612. [Crossref] [PubMed]
29. Han DS, Chang KV, Li CM, Lin YH, Kao TW, Tsai KS, Wang TG, Yang WS. Skeletal muscle mass adjusted by height correlated better with muscular functions than that adjusted by body weight in defining sarcopenia. Sci Rep 2016;6:19457. [Crossref] [PubMed]