Malnutrition by the GLIM criterion and reduced HGS are associated with anthropometric indicators in hospitalized patients

Abstract

In the hospital environment, anthropometry is commonly used to assess nutritional status. To assess whether muscle strength and malnutrition, as defined by the GLIM (global leadership initiative on malnutrition) criteria, are associated with anthropometric indicators in hospitalized patients. In a cross-sectional study, body mass index (BMI), mid-upper arm circumference (MUAC), triceps skinfold thickness (TSF), arm muscle circumference (AMC), calf circumference (CC), body weight loss, HGS, and GLIM criteria were investigated. The Chi-square, Fisher, and Mann-Whitney tests were used for comparisons, and the Kappa coefficient for agreement analyses. Univariate and multivariate logistic regressions were used to investigate the association between the GLIM criterion and FPM and anthropometric indicators. HGS (p = 0.0004, OR = 15.193, 95% CI = 3.397; 67,956) and the GLIM criterion (p = 0.0002, OR = 6,142, 95% CI = 2,352; 16,035) were associated with CC < 31 cm. Age (p = 0.0019, OR = 1.054, 95% CI = 1.020; 1,090), the mean HGS (p = 0.0064, OR = 4,094, 95%CI = 1,487; 11.265), and the GLIM criterion (p = 0.0005, OR = 4.798, 95% CI = 1.994; 11,548) were associated with BMI (a higher chance of low body weight by BMI). Reduced HGS and malnutrition, as defined by the GLIM criterion, are associated with CC. Age, HGS, and the GLIM criterion are associated with low body weight, as measured by BMI. These findings reinforce the relevance of using these tools for nutritional care in the hospital setting.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-50959-4.

Introduction

In the hospital setting, anthropometric parameters are used to assess nutritional status and guide therapeutic management, although their application is limited in patients who are bedridden or completely immobile.

A study indicates that, with the exception of height, anthropometric indicators differed significantly between patients with severe and moderate malnutrition¹. Changes in mid-upper arm circumference (MUAC) and triceps and biceps skinfolds have been observed in patients with chronic kidney disease undergoing hemodialysis following nutritional therapy².

MUAC, body mass index (BMI) and body weight were used in conjunction with laboratory tests and screening instruments in the evaluation of patients undergoing colorectal cancer surgery³ and in patients with laryngeal cancer screening instruments were used combined with BMI, hip circumference, calf circumference (CC), weight loss, laboratory tests and hand grip strength (HGS)⁴. Another study that used various tools to assess nutritional risk and malnutrition, such as BMI, CC, and HGS, showed different prevalences between the methods⁵; body weight, BMI, phase angle (PhA), CC, and HGS were lower in patients who presented malnutrition according to the GLIM criteria⁵.

HGS has already been reported as a viable method for estimating muscle mass and a marker of nutritional status^6,7. However, it has had very low diagnostic value and accuracy to identify severe malnutrition in hospitalized cancer patients⁶. It was also considered a marker of malnutrition, since muscle function could respond early to nutritional deprivation⁷. A recent study with hospitalized older adults reinforces its application as an alternative measure for estimating muscle mass and managing malnutrition⁸. It has also been reported that after hepatectomy for hepatocellular carcinoma, skeletal muscle mass and sarcopenia classified by HGS were associated factors with complications⁹.

Regarding the GLIM criteria, one study found that CC and arm muscle circumference, when used as phenotypic criteria of muscle mass, were effective for diagnosing malnutrition¹⁰.

This study evaluated whether hand grip strength (HGS) and the Global Leadership Initiative on Malnutrition (GLIM) criteria are associated with anthropometry indicators in hospitalized patients.

Method

Study design, location, and sample calculation

This was the second part of a cross-sectional study conducted in a university hospital, among patients hospitalized in surgery wards by the Unified Health System (SUS). Detailed methods have been described in a previous publication¹¹.

Based on estimates of hospital malnutrition (15–60%)^12,13, a sample size calculation was performed, adopting a significance level of 5% and a sampling error of 6–8%. Sample sizes of 137 (prevalence of 15%, error of 6%) and 145 (prevalence of 60%, error of 8%) were obtained. Considering the variability of the estimates, the larger value (n = 145) was selected, plus 10% (n = 160) to account for potential participant attrition or incomplete data.

Ethical aspects, inclusion and exclusion criteria, and data collection

This study was approved by the Ethics and Research Committee of the Pontifical Catholic University of Campinas, SP, Brazil (protocol no. 5.728.982). Informed consent was obtained from all individuals who participated in the study. All methods were carried out in accordance with relevant guidelines and regulations.

Patients with preserved motor and cognitive function aged 18 years or older who agreed to participate in the study by signing the informed consent form (ICF) were included.

Patients with fractures of the upper or lower limbs were excluded (as these conditions made it impossible to measure HGS and anthropometry indicators), with any dementia (as these conditions made it impossible to provide adequate information), with edema or ascites (as these conditions could mask the real body weight and other anthropometry measurements), in isolation (patients with Covid or other infectious diseases), in terminal critical condition and with a hospitalization time of less than 48 h.

Clinical and demographic data were obtained by consulting the hospital’s electronic medical records. Nutritional status evaluations, anthropometric measurements, and HGS were performed at the bedside during patient care.

Nutritional assessment:- anthropometry, muscle strength, and GLIM criteria

Anthropometric indicators, including body mass index (BMI), mid-upper arm circumference (MUAC), triceps skinfold thickness (TSF), arm muscle circumference (AMC), calf circumference (CC), and recent body weight loss (RBWL), were evaluated.

BMI was assessed according to the cutoff points established for adults aged¹⁴ and the elderly aged¹⁵. The indicators of MUAC, TSF, and AMC were classified according to the reference values recommended for adults and older individuals, as specified in the standardization^16,17, and CC was evaluated according to the recommendations of the World Health Organization¹⁸. RBWL was investigated and found to be unintentional¹⁹, as reported by the patient at the time of admission.

Hand grip strength (HGS) was assessed using standardized procedures with the Jamar hydraulic dynamometer^7,8,20,21 and previously defined cutoff points (men: < 27 kg, women: < 16 kg)²². Detailed methods for implementing this measure have been previously presented in a publication¹¹.

The GLIM criterion was assessed based on nutritional screening and the presence of at least one phenotypic criterion (weight loss, low BMI, or reduced muscle mass: MUAC or CC and one etiological criterion (low food intake or disease severity)²³.

Statistical analysis

The population characteristics were described using absolute and percentage frequencies (categorical variables) and descriptive measures, including mean, standard deviation, and median (quantitative variables). Comparisons between groups were performed using the Mann-Whitney test (continuous variables), chi-square, or Fisher’s exact test (proportions). The variables age and FPPM were evaluated for normality using the Shapiro-Wilk test, considering the analysis subgroups. Due to the lack of adherence to normality in most subgroups, non-parametric tests were chosen for all comparisons.

Agreement between the GLIM criteria, HGS, and anthropometric indicators was assessed using the kappa coefficient. The discriminatory ability of the criteria and indicators within the study sample was assessed using univariate and multivariate logistic regression with stepwise variable selection. Factors with significant univariate associations were included in the multivariate model. A p < 0.05 significance level was adopted, without correction for multiple comparisons.

Given the exploratory nature of the bivariate analyses, no correction for multiple testing was applied. Consequently, the reported p-values ​​were interpreted cautiously, and the findings were considered hypothesis-generating rather than confirmatory. To minimize the risk of overinterpretation, the study’s main conclusions were based on the multivariable model rather than on individual bivariate comparisons. The studies were performed using the Statistical Analysis System (SAS) software.

We emphasize the exploratory nature of stepwise regression, aiming to identify potential associations among variables rather than to establish a definitive predictive model. We used a bidirectional stepwise method, with entry and retention criteria based on p-value thresholds (any value for entry and < 0.05 for retention).

Results

The study included 160 patients, with a mean age of 59.31 ± 16.14 years and a mean length of stay of 18.46 ± 22.19 days; 62.5% were male. Surgical procedures were performed in 61.3% of cases, 67.7% had no complications, and 1.9% resulted in death. The most frequent diseases in the sample were cardiac (25%), neoplasms (25%), renal (15.6%), and vascular (14.4%), followed by pulmonary (7.5%), orthopedic (5%), rheumatological (3.1%), and other (4.4%).

Regarding anthropometric indicators, 60.6% reported recent weight loss. The mean BMI was 25.36 ± 5.69 kg/m2, the mean MUAC was 28.91 ± 4.83 cm, the mean TSF was 14.55 ± 7.93 mm, the mean AMC was 239.34 ± 46.68 mm, and the mean CC was 34.44 ± 5.03 cm.

Regarding the instruments evaluated, 44.4% of patients were classified as malnourished according to the GLIM criteria. The mean HGS was 21.82 ± 10.22 kg, with 57.5% presenting values below the reference values²². A description of the study population is available in a supplementary file.

Agreement between the GLIM criteria, HGS, anthropometry indicators, and RBWL

Weak or no agreement (kappa) was observed between the indicators (GLIM and MUAC = 0.1555, 95% CI = 0.0218; 0.2892, accuracy = 60.6%; GLIM and TSF = 0.1234, 95% CI=-0.0049; 0.2517, accuracy = 59.4%; GLIM and AMC = 0.1207, 95% CI=-0.0045; 0.2459, accuracy = 59.4%; GLIM and CC = 0.3184, 95% CI = 0.1861; 0.4507, accuracy = 68.1%; GLIM and BMI = 0.3671, 95% CI = 0.2294; 0.5047, accuracy = 70.0%); (HGS and MUAC = 0.0826, 95% CI= -0.0269; 0.1920, accuracy = 50.0%; HGS and TSF = 0.1116, 95% CI = 0.0101; 0.2132, accuracy = 51.3%; HGS and AMC = 0.1146, 95% CI = 0.0169; 0.2124, accuracy = 51.3%; HGS and CC = 0.2865, 95%CI = 0.1822; 0.3909, accuracy = 61.3%; HGS and BMI = 0.2870, 95% CI = 0.1699; 0.4041, accuracy = 61.9%; HGS and RBWL = 0.1091, 95% CI= -0.0455; 0.2637, accuracy = 56.9%). There was only moderate agreement between GLIM and RBWL (0.5849, 95% CI = 0.4674;0.7025, accuracy = 78.8%).

Comparison of variables between CC, BMI, and RBWL

A significant difference was found in the mean (or median) between age and CC (p = 0.0149) and between mean HGS and CC (p < 0.0001). It was observed that individuals with CC < 31 cm were older and had lower mean HGS values. Regarding diseases, there was a predominance of lung diseases and neoplasms among individuals with CC < 31 cm (p = 0.0188). A significant association was observed between the GLIM criterion and CC (p<0.0001), with a higher percentage of malnourished patients in those with CC<31 cm. In addition, patients with HGS below the reference values had a higher frequency of CC < 31 cm (p<0.0001) (Table 1).

Table 1.

Descriptive analysis and comparison of variables between calf circumference (< 31 / ≥31 cm).

Variables	Categories	Calf circumference		P value
		< 31 cm (N = 34)	≥ 31 cm (N = 126)
Age	Mean ± SD Median (min.-max.)	65.38 ± 14.99 67.00 (25.00–87.00)	57.67 ± 16.10 59.50 (19.00–95.00)	0.0149¹
mHGS	Mean ± SD Median (min.-max.)	15.67 ± 6.89 15.00 (3.60–32.10)	23.48 ± 10.36 21.67 (4.33-48.00)	< 0.0001¹
Sex	Female Male	16 (47.1%) 18 (52.9%)	44 (34.9%) 82 (65.1%)	0.1945²
Disease	Cardiac Orthopedic Pulmonary Renal Rheumatology Vascular Neoplasm Other	5 (14.7%) 0 (0.0%) 5 (14.7%) 5 (14.7%) 3 (8.8%) 3 (8.8%) 13 (38.2%) 0 (0.0%)	35 (27.8%) 8 (6.3%) 7 (5.6%) 20 (15.9%) 2 (1.6%) 20 (15.9%) 27 (21.4%) 7 (5.6%)	0.0188³
GLIM	Malnourished Not malnourished	27 (79.4%) 7 (20.6%)	44 (34.9%) 82 (65.1%)	< 0.0001²
C mHGS	< reference ≥ reference	32 (94.1%) 2 (5.9%)	60 (47.6%) 66 (52.4%)	< 0.0001²

mHGS: mean handgrip strength. C mHGS: classified mean handgrip strength. GLIM: global leadership initiative on malnutrition.

Min.-Max.: Minimum-Maximum. ¹ Mann-Whitney test, ² Chi-square test, ³ Fisher test.

A difference was also observed in the mean (or median) age (p < 0.0001) and mean HGS (p = 0.0141) between BMI classifications. Older age and lower mean HGS values were observed in patients with low body weight. Regarding diseases, there was a predominance of lung disease and neoplasms in the low body weight classification (p = 0.0239). A significant association was found between the GLIM criterion and BMI (p < 0.0001), with a higher percentage of malnourished patients having low body weight. Patients with HGS below the reference values had a higher rate of underweight BMI (p<0.0001) (Table 2).

Table 2.

Descriptive analysis and comparison of variables between classified BMI body mass index (underweight and adequate weight/overweight).

Variables	Categories	Body mass index		P value
		Low weight (N = 43)	Adequate weight/overweight (N = 117)
Age	Mean ± SD Median (min.-max.)	69.51 ± 12.72 71.00 (34.00–87.00)	55.56 ± 15.67 57.00 (19.00–95.00)	< 0.0001¹
mHGS	Mean ± SD Median (min.-max.)	18.17 ± 7.49 16.67 (5.17–34.33)	23.17 ± 10.78 21.00 (3.60–48.00)	0.0141¹
Sex	Female Male	13 (30.2%) 30 (69.8%)	47 (40.2%) 70 (59.8%)	0.2497²
Disease	Cardiac Orthopedic Pulmonary Renal Rheumatology Vascular Neoplasm Other	8 (18.6%) 0 (0.0%) 6 (14.0%) 5 (11.6%) 2 (4.7%) 5 (11.6%) 17 (39.5%) 0 (0.0%)	32 (27.4%) 8 (6.8%) 6 (5.1%) 20 (17.1%) 3 (2.6%) 18 (15.4%) 23 (19.7%) 7 (6.0%)	0.0239³
GLIM	Malnourished Not malnourished	33 (76.7%) 10 (23.3%)	38 (32.5%) 79 (67.5%)	< 0.0001²
C mHGS	< Reference ≥ Reference	37 (86.0%) 6 (14.0%)	55 (47.0%) 62 (53.0%)	< 0.0001²

mHGS: mean handgrip strength. C mHGS: classified mean handgrip strength. GLIM: Global Leadership Initiative on Malnutrition.

Min.-Max.: Minimum-Maximum. ¹Mann-Whitney test, ²Chi-square test, ³Fisher test.

A difference was found in age in relation to RBWL (p = 0.0435). Patients with RBWL were older (Table 3).

Table 3.

Descriptive analysis and comparison of variables between recent body weight loss.

Variables	Categories	Recent weight loss		P value
		No (N = 63)	Yes (N = 97)
Age	Mean ± SD Median (min.-max.)	56.32 ± 15.37 59.00(21.00–81.00)	61.26 ± 16.40 64.00 (19.00–95.00)	0.0435¹
mHGS	Mean ± SD Median (min.-max.)	23.08 ± 10.27 20.00 (7.67-48.00)	21.01 ± 10.16 19.33 (3.60–43.50)	0.2868¹
Sex	Female Male	26 (41.3%) 37 (58.7%)	34 (35.1%) 63 (64.9%)	0.4273²
Disease	Cardiac Orthopedic Pulmonary Renal Rheumatology Vascular Neoplasm Other	21 (33.3%) 3 (4.8%) 4 (6.3%) 10 (15.9%) 2 (3.2%) 9 (14.3%) 10 (15.9%) 4 (6.3%)	19 (19.6%) 5 (5.2%) 8 (8.2%) 15 (15.5%) 3 (3.1%) 14 (14.4%) 30 (30.9%) 3 (3.1%)	0.3569³
C mHGS	< reference ≥ reference	32 (50.8%) 31 (49.2%)	60 (61.9%) 37 (38.1%)	0.1667²

mHGS: mean handgrip strength. C mHGS: classified mean handgrip strength.

¹Mann-Whitney test, ²Chi-square test, ³Fisher test.

Discriminatory ability within the study sample of the GLIM and HGS criteria: univariate and multiple logistic regression analysis in relation to CC, BMI, and RBWL

The mean HGS (p = 0.0004; OR = 15.19; 95%CI = 3.40; 67.96) and the GLIM criterion (p = 0.0002; OR = 6.14; 95%CI = 2.35; 16.04) were factors associated with CC. HGS below the reference values increased the chance of CC < 31 cm by 15.2 times. And malnutrition, according to the GLIM criterion, was 6.1 times more likely to be associated with CC < 31 (Table 4).

Table 4.

Results of univariate and multiple logistic regressions to study the discriminatory ability of the GLIM criterion and mHGS in relation to CC.

Variables	Categories	Probability of CC < 31 cm
		P value	OR	95% CI
Univariate analysis
Age	Years	0.0154	1.034	1.006; 1.062
mHGS		0.0002	0.906	0.859; 0.954
Sex	Female vs. Male	0.1968	1.657	0.770; 3.566
Disease	Cardiac Orthopedic Pulmonary Renal Rheumatology Vascular Neoplasm	0.9603 1.0000 0.9544 0.9582 0.9516 0.9601 0.9558
GLIM	Malnourished vs. Not Malnourished	< 0.0001	7.188	2.898; 17.830
c mHGS	< reference vs. ≥ reference	0.0001	17.600	4.044; 76.601
Multiple analysis*
GLIM	Malnourished vs. Not Malnourished	0.0002	6.142	2.352; 16.035
c mHGS	< reference vs. ≥ reference	0.0004	15.193	3.397; 67.956

CC: calf circumference. mHGS: mean handgrip strength. c mHGS: classified mean handgrip strength.

GLIM: Global Leadership Initiative on Malnutrition. OR: Odds Ratio, 95% CI: 95% Confidence Interval.

*Hosmer and Lemeshow Goodness-of-Fit Test, p-value = 0.7259; accuracy c = 0.831.

Age (p = 0.0019, OR = 1.054, 95% CI = 1.020; 1,090), the mean HGS classified (p = 0.0064, OR = 4,094, 95% CI = 1,487; 11.265), and the GLIM criterion (p = 0.0005, OR = 4.798, 95% CI = 1.994; 11,548) were factors associated with BMI.

The 1-year increase in age increased the chance of underweight by 5.4%. HGS below the reference values increased the opportunity of underweight by 4.1 times. And malnutrition, according to the GLIM criterion, was 4.8 times more likely to be underweight according to BMI (Table 5).

Table 5.

Results of univariate and multiple logistic regressions to study the discriminatory ability of the GLIM criterion and mHGS in relation to BMI.

Variables	Categories	Probability of low weight
		P-value	OR	95% CI
Univariate analysis
Age	Years	< 0.0001	1.076	1.042; 1.111
mHGS		0.0074	0.947	0.911; 0.986
Sex	Female vs. Male	0.2516	1.549	0.733; 3.275
Disease	Cardiac vs. other Orthopedic vs. other Pulmonary vs. other Renal vs. other Rheumatology vs. other Vascular vs. other Neoplasm vs. other	0.9549 1.0000 0.9493 0.9549 0.9510 0.9545 0.9506
GLIM	Malnourished vs. Not Malnourished	< 0.0001	6.859	3.062; 15.362
c mHGS	< reference vs. ≥ reference	< 0.0001	6.952	2.727; 17.723
Multiple analysis*
Age		0.0019	1.054	1.020; 1.090
GLIM	Malnourished vs. Not Malnourished	0.0005	4.798	1.994; 11.548
c mHGS	< reference vs. ≥ reference	0.0064	4.094	1.487; 11.265

BMI: body mass index. mHGS: mean handgrip strength. c mHGS: classified mean handgrip strength. GLIM: Global Leadership Initiative on Malnutrition. OR: Odds Ratio, 95%CI: 95% Confidence Interval.

*Hosmer and Lemeshow Goodness-of-Fit Test, p-value = 0.3645; accuracy c = 0.850.

In Table 6, HGS was only mentioned because weight loss is one of the criteria that score GLIM. No variable was significant at the 5% level for the study of the discriminatory ability of the mean HGS in relation to the RBWL (Table 6).

Table 6.

Results of univariate and multiple logistic regressions to study the discriminatory ability of mHGS in relation to recent weight loss.

Variables	Categories	Likelihood of recent weight loss
		P-value	OR	95% CI
Univariate analysis
Age	Years	0.0603	1.019	0.999; 1.040
mHGS		0.2113	0.980	0.950; 1.011
Sex	Female vs. Male	0.4277	1.302	0.678; 2.500
Disease	Cardiac vs. other Orthopedic vs. other Pulmonary vs. other Renal vs. other Rheumatology vs. other Vascular vs. other Neoplasm vs. other	0.8205 0.4499 0.3164 0.4235 0.5603 0.4045 0.1015	1.206 2.222 2.667 2.000 2.000 2.074 4.000	0.239; 6.099 0.280; 17.631 0.391; 18.166 0.366; 10.919 0.194; 20.614 0.373; 11.528 0.761; 21.021
c mHGS	< reference vs. ≥reference	0.1677	1.571	0.827; 2.984
Multiple analysis
Stepwise selection process – no variable selected at the 5% level

mHGS: mean handgrip strength. c mHGS: classified mean handgrip strength.

OR: Odds Ratio, 95%CI: 95% Confidence Interval.

Discussion

This study is part of a research project that investigated various indicators and tools associated with nutritional diagnosis in hospitalized patients. The first stage of this project focused on the relationship between nutritional screening and subjective assessment instruments, with results already published¹¹; the second stage investigates the association with anthropometry and constitutes the subject of the present study.

In the present study, there was no agreement between the anthropometry indicators and the GLIM criterion and the HGS, and only moderate agreement was observed between the GLIM and the loss of body weight. On the other hand, a recent study showed moderate agreement between muscle depletion and isolated CC²⁴, whereas another study showed strong agreement between the instruments used¹.

Significant associations were found between age and CC, as well as between HGS and CC. CC less than 31 cm was associated with older age and lower muscle strength, suggesting loss of muscle mass and functional fragility, common characteristics observed in hospitalized patients^7,8. The association between CC and malnutrition, as determined by the GLIM criterion, reinforces the use of CC as a useful indicator for hospital nutritional diagnosis¹⁰. Another study⁵ also showed that patients with severe or moderate malnutrition, as defined by GLIM, had lower CC than non-malnourished patients.

Another interesting finding in this study was the association between low HGS. It reduced CC, highlighting a link between the two variables and suggesting they could be useful for assessing the nutritional status of hospitalized patients. In a cross-sectional study of mostly hospitalized patients undergoing surgical procedures, HGS values were also significantly associated with CC²⁵. And sarcopenia, assessed by low HGS and reduced CC, has been associated with cognitive impairment in elderly Chinese patients²⁶. It has been reported that cancer patients with sarcopenia have lower values ​​for BMI, CC, HGS, skeletal muscle area and mass index, and muscle radiodensity²⁷. The authors suggested that, in the absence of other methods, CC could replace computed tomography for assessing lean mass²⁷.

In this study, significant differences were also observed between BMI and HGS in relation to age. Patients with low body weight had lower mean HGS values and higher age, suggesting reduced muscle strength, highlighting a possible relationship between muscle function, aging, and nutritional deficit²⁸. This finding points to the need to explore these relationships in future studies.

Another association found was between the GLIM criterion and BMI, with a higher proportion of patients diagnosed with malnutrition among those with low body weight. It is essential to note that although BMI is not considered a definitive marker of malnutrition, its evaluation in conjunction with other criteria and/or indicators can contribute to a more accurate nutritional diagnosis²³. Another relevant finding was the association with older age in patients who had recently experienced weight loss.

The results of the regression analysis showed that muscle strength, as assessed by HGS, and the diagnosis of malnutrition, as determined by GLIM criteria, were associated factors for reduced CC. The results showed an association between low muscle strength and a smaller CC (< 31 cm), suggesting muscle loss or frailty. The GLIM criterion was also relevant, but with a lower impact than HGS.

In a study of geriatric patients, CC was positively associated with HGS in women²⁹. The authors reported that although measures such as MUAC and CC are practical for estimating skeletal muscle mass, their relationship with physical function was weak, and other anthropometric measures showed no association with HGS²⁹. CC could be used as an alternative to assess muscle mass when more valuable research techniques are not available²².

The regression results also revealed that age, HGS, and malnutrition diagnosis by the GLIM criterion were significantly associated with the risk of underweight assessed by BMI. The 5.4% annual increase in the risk of being underweight aligns with previous studies showing a tendency toward weight loss with aging, especially among elderly individuals³⁰.

In the present study, reduced HGS increased the likelihood of low weight by more than four times, indicating a relationship between low muscle mass and compromised nutritional status. Individuals with malnutrition, according to the GLIM, were almost five times more likely to be underweight, as determined by BMI, which may reinforce the discriminatory ability of this criterion in nutritional diagnosis and its relationship with anthropometry.

It is important to highlight that the purpose of this investigation was to evaluate whether these two instruments (HGS and GLIM criterion), individually or in combination, would be associated with the results of anthropometry. Although we acknowledge some overlap between the indicators, which may introduce a risk of conceptual circularity, we emphasize that the study’s objective was not to evaluate discriminative independence, but rather to explore the degree of agreement/association between nutritional metrics.

Recent studies assessing the association between HGS and anthropometry in elderly individuals have shown that the values of all anthropometric parameters increased with increasing HGS, and that increasing age was associated with a higher percentage of weak HGS³⁰. It was also reported that BMI and waist circumference were the best predictors of weak HGS in both sexes³⁰. In another study investigating the association between HGS and anthropometry, multivariate regression analysis showed that HGS was independently associated with sex, age, waist-to-height ratio, waist-to-hip ratio, and BMI³¹. There are also studies reporting HGS as a poor predictor of nutritional status in hospitalized patients⁶.

The findings of this investigation indicated that HGS and the GLIM criterion are effective methods for identifying patients with malnutrition in relation to CC and BMI, highlighting the importance of HGS and GLIM as complementary tools for assessing nutritional status. The objective of this study was to demonstrate that these instruments, indicators, and tools facilitate the diagnosis of malnutrition in a straightforward and easily applicable manner, provided they are utilized by properly trained professionals. Another contribution was to point out that these instruments and tools present a practical and valuable approach in hospital environments, particularly in places that lack advanced resources for this diagnosis.

This study suggests that these tools can be highly useful for routine nutritional care in hospital settings, especially in specific clinical situations involving hospitalized patients, such as those who are bedridden and cannot be weighed due to mobility limitations. In such cases, these tools can contribute to the early assessment of nutritional status and the timely implementation of nutritional interventions.

Conclusion

Reduced HGS and malnutrition, as defined by the GLIM criterion, are associated factors of CC. Age, HGS, and the GLIM criterion are associated factors of low body weight by BMI. These findings reinforce the relevance of using these tools for nutritional care in the hospital setting.

Highlights and limitations of the study

Some important aspects of this study, which investigated various anthropometric indicators in a clinically relevant sample of patients, should be highlighted. A group of low-income individuals hospitalized through the Unified Health System was analyzed. However, the study has limitations related to the adopted exclusion criteria. Most excluded patients did not present adequate conditions for the measurement of anthropometric indicators. Furthermore, due to the cross-sectional design, the results should be interpreted as associations and exploratory findings, rather than as causal or predictive relationships.

Although we acknowledge some overlap between the indicators, which may introduce a risk of conceptual circularity, we emphasize that the study’s objective was not to evaluate discriminative independence, but rather to explore the degree of agreement and association between nutritional metrics.

Another limitation of this study concerns the outcome of recent body weight loss. The analysis of recent body weight loss did not identify statistically significant associations. Although the analytical approach was appropriate, the number of informative events was limited.

Another limitation of this study concerns recent body weight loss as an outcome. The analysis did not identify statistically significant associations. Although the analytical approach was appropriate, the number of informative events was limited.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

VALM was principal investigator of the study. LRM actively participated in data collection. VACB, RIS, JPT, LSS contributed to data collection. The authors approved the final version of the manuscript.

Funding

Grant process nº 2023/01435-2 (Fundação de Amparo à Pesquisa do Estado de São Paulo-FAPESP, Brazil).

Data availability

“The data that support the findings of this study are available from the corresponding author upon reasonable request”.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This project was approved by the Ethics Committee of the Pontifical Catholic University of Campinas, SP, Brazil. CAAE: 64107922.5.0000.5481, No. 5.728.982.

Patient consent

Written informed consent was obtained from all individual participants included in the study. All methods were carried out in accordance with relevant guidelines and regulations.