Prevalence of Malnutrition in Patients with Hepatocellular Carcinoma: A Comparative Study of GLIM Criteria, NRS2002, and PG-SGA, and Identification of Independent Risk Factors

Abstract

Aim

Malnutrition is prevalent in hepatocellular carcinoma (HCC) patients, linked to poor outcomes, necessitating early intervention. This study aimed to investigate malnutrition in HCC patients, assess Nutrition Risk Screening 2002 (NRS-2002) and Patient-Generated Subjective Global Assessment (PG-SGA) vs. Global Leadership Initiative on Malnutrition (GLIM) criteria, and identify independent risk factors.

Method

A cross-sectional retrospective study was conducted on 207 patients with HCC. Nutritional screening/assessment results and blood samples were collected within 72 h of admission. This study assessed the prevalence of malnutrition using the NRS-2002 and PG-SGA and retrospectively using the GLIM criteria. The performance of the screening tools was evaluated using kappa (K) values. Logistic regression analyses were performed to determine whether laboratory parameters were associated with malnutrition as identified by the GLIM criteria.

Results

Of the participants, 30.4% were at risk of malnutrition according to NRS-2002. The agreement between the NRS-2002 and GLIM criteria was substantial. The GLIM criteria and PG-SGA diagnosed malnutrition in 43 and 54.6% of the participants, respectively. Age, anemia, and ascites correlated with malnutrition in regression.

Conclusion

The GLIM criteria, along with NRS-2002 and PG-SGA, aid in diagnosing malnutrition in HCC patients. Recognizing risk factors improves accuracy, enabling timely interventions for better outcomes.

Introduction

Hepatocellular carcinoma (HCC) is recognized globally as the sixth most commonly diagnosed cancer and the third leading cause of cancer-related deaths (1). Among cancer patients, malnutrition is a frequent complication that contributes significantly to cancer-related mortality, surpassing the impact of the cancer itself (2–4). Malnutrition has been found to hinder the effectiveness of cancer therapies, increase healthcare expenses, prolong hospital stays, diminish patients’ quality of life, and reduce their lifespan (5–7). Given these consequences, early detection of malnutrition is crucial to provide a foundation for nutritional therapy in patients with cancer and ensure optimal nutritional status (8,9).

In clinical practice, the Nutrition Risk Screening 2002 (NRS-2002) is a tool widely used in clinical practice for screening nutrition-related risks (10,11). Although primarily recognized as a screening method, NRS-2002 plays a crucial role in identifying patients at risk of malnutrition, and guiding subsequent nutritional interventions. Whereas the scored Patient-Generated Subjective Global Assessment (PG-SGA) is the recommended method for nutritional assessment in oncology (8,12). It provides a detailed evaluation of various aspects of a patient’s nutritional status, offering insights beyond the scope of screening tools. However, previous studies have revealed discrepancies in the outcomes of different nutritional risk screening tools (13). Despite these differences in primary purposes, both NRS-2002 and PG-SGA are frequently utilized in clinical settings for addressing nutritional concerns in patients with cancer. Nevertheless, it remains essential to assess and compare their diagnostic performance in the specific context of patients with HCC.

Moreover, the PG-SGA is time-consuming and relies on a complex design and detailed questioning. In view of the limitations of these tools, the Global Leadership Initiative on Malnutrition (GLIM) introduced the GLIM Criteria as a more comprehensive and convenient nutrition assessment tool (14,15). It incorporates essential nutritional criteria from various widely used assessments and screening tools, including the PG-SGA and NRS-2002 (16,17).

The GLIM Criteria have been validated in various clinical oncology settings (18–20). However, the absence of consensus on malnutrition diagnostic criteria in clinical settings highlights the need for comparative studies that assess different tools within a single patient population (21–23). Furthermore, studies applying the GLIM Criteria to patients with HCC are limited. Moreover, limited information is available regarding the relationship between malnutrition prevalence and HCC risk factors, such as demographic parameters, clinical parameters, tumor-related parameters, and liver functional reserve. Hence, the objective of this study was to assess the diagnostic performance of the GLIM Criteria in comparison with the NRS-2002 and PG-SGA criteria in patients with HCC. Moreover, we aimed to identify independent risk factors for malnutrition in this specific population. By conducting this research, our goal was to contribute to the understanding of malnutrition in patients with HCC and provide valuable insights that can enhance nutritional management and improve patient outcomes. This comprehensive analysis will help elucidate the intricate relationship between malnutrition and various parameters in patients with HCC, shedding light on optimal strategies for early detection and intervention.

Study Design and Method

Patients and Exclusion Criteria

This cross-sectional retrospective study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board, and Ethics Committee of Liuzhou People’s Hospital, affiliated with Guangxi Medical University in China (Approval No. KY2022-030-02). The study included 207 treatment-naïve patients with a new diagnosis of primary HCC, following the criteria. These are outlined in the NCCN Guidelines (24) for Hepatobiliary Cancers between July 2019 and July 2022. The exclusion criteria included patients with a confirmed diagnosis of another cancer type based on postoperative pathology, those who had undergone preoperative treatment, individuals under 18 years of age, those with difficulties in understanding or completing the questionnaires, individuals with severe impairment of vital organ function, and pregnant women. All the participants provided written informed consent upon admission.

Data Collection

Before data collection, all clinical staff members underwent thorough training to ensure consistency and accuracy. Within the first 24 h of admission, our trained doctors, dietitians, and nurses conducted nutritional screening and assessments for all participants using the NRS-2002 and PG-SGA. This involved gathering general information and taking anthropometric measurements. Blood samples and imaging examinations were collected within 72 h of the patient’s admission to determine the necessary parameters. These included serum albumin (ALB), hemoglobin (HGB), lymphocyte (LYM) count, assessment of portal hypertension (PHT), and the presence of ascites (ASC). The assessment of portal hypertension, as outlined in the Chinese Expert Consensus on Clinical Diagnosis and Treatment of Portal Hypertension Complicated with Hepatocellular Carcinoma 2022, involves the evaluation of specific criteria. These include: (1) Splenomegaly, characterized by a longitudinal dimension exceeding 10–13 cm and a thickness >4.5 cm; (2) Persistent reduction in platelets (PLT) (<100 × 10^9/L) or white blood cells (WBC) (<4 × 10^9/L) occurring over three consecutive measurements; (3) Identification of esophageal or gastric varices through imaging or gastroscopy; and (4) Ultrasonography indicating an increased portal vein width (>14 mm) or splenic vein width (>10 mm). The diagnosis of portal hypertension was confirmed when at least two of these criteria were met.

For the assessment of ascites, according to the guidelines of The European Association for the Study of the Liver (EASL), a clinical examination is employed, for the severity of ascites to be graded. Ultrasound imaging was performed as part of the evaluation process.

Additionally, other parameters, such as tumor diameter, tumor numbers, Child-Pugh class, Barcelona Clinic Liver Cancer Stage (BCLC), and model for albumin-bilirubin (ALBI) scores were also recorded. These laboratory data, along with the nutritional screening and assessment results, were used in logistic regression analysis to assess the applicability of the GLIM criteria. Specifically, we extracted and analyzed BMI data, dietary intake reduction, and weight loss information obtained from the NRS-2002 and PG-SGA assessments within the context of the GLIM criteria.

Anthropometric Measurements

On the day of hospitalization, body weight and height were measured in participants who were instructed to wear light clothing and no shoes. Participants also self-reported their weight history for the previous one to six months. The percentage of unintentional weight loss was calculated as follows: [(Previous Weight − Current Weight)/Previous Weight].

In patients with advanced-stage liver disease, fluid retention, such as edema and ascites, is commonly observed. This can impose certain limitations on the use of Body Mass Index (BMI). For patients with advanced-stage liver disease and fluid retention, it is possible to calculate the dry body mass index (BMI = dry body mass/height squared, kg/m²). Dry body mass (DBM) can be calculated using several methods: (1) Body mass before the onset of fluid retention. (2) Body mass after paracentesis drainage. (3) Corrected body mass, adjusted based on the clinical assessment of ascites severity (mild, 5%; moderate, 10%; severe, 15%, with an additional 5% reduction if peripheral edema was present) (24).

All patients underwent abdominal CT examinations within 72 h after the day of admission. We evaluated the area of the total skeletal muscle (cm²), including the psoas major, erector spinae, transverse abdominis muscle, abdominal internal oblique muscle, abdominal external oblique muscle, and rectus abdominis, at the third lumbar (L3) level in computed tomography images. The skeletal muscle index (SMI) was determined using the following formula: SMI = total skeletal muscle area (cm²) at the L3 level/height squared (m²).

Food Intake

To assess the percentage reduction in food intake, we utilized the validated Simple Diet Self-Assessment Tool (SDSAT) (25) Participants were asked to rate their current food intake compared with their normal intake, on a scale ranging from 1 to 5, with 1 indicating an intake of <300 kilocalories (kcal); 2, intake between 300 and 600 kcal; 3, intake between 600 and 900 kcal; 4, intake between 900 and 1200 kcal; and 5, intake of more than 1200 kcal. The scale was converted to corresponding calorie intake ranges. We then calculated the percentage reduction in food intake by comparing the participants’ reported food intake with the expected or normal calorie intake.

Malnutrition Screening/Assessment Criteria

The NRS-2002 tool was used to determine the risk of malnutrition by considering factors, such as weight loss, reduced food intake, BMI, disease severity, and age (≥70 years). The score ranged from 0 to 7, with a score of ≥3 indicating a risk of malnutrition.

The PG-SGA consists of two components. The patient-generated component includes four sections that assess weight history, food intake, symptoms, activities, and function, which the patients complete independently. The professional component evaluates disease stages, metabolic stress, and physical components, and is completed by physicians, nurses, or dietitians. Based on the scoring criteria provided by the PG-SGA, patients are categorized as well-nourished (score <4) or malnourished (score ≥4).

The GLIM incorporates both phenotypic and etiological criteria to classify malnourished individuals in a two-step process. In the first step, Malnourished individuals were identified by a combination of at least one phenotypic and one etiological criterion. Relevant data, such as BMI measurements, information on dietary intake reduction, and details of weight loss, were obtained from the NRS-2002 and PG-SGA assessments. The phenotypic criteria, including unintentional weight loss, low BMI for age, and reduced muscle mass, were assessed. Unintentional weight loss (>5% body weight within 6 months or >10% over >6 months) and low BMI (<18.5 kg/m² for patients <70 years or <20 kg/m² for patients 70 years of age) were selected as the phenotypic criteria to define malnutrition. According to the Japan Society of Hepatology guidelines for sarcopenia in liver disease (1st edition) (26), reduced muscle mass is defined as an L3-SMI (Lumbar 3 Skeletal Muscle Index) <42 cm²/m² for males and <38 cm²/m² for females. The etiological criteria included reduced food intake and/or disease burden. As all patients were diagnosed with cancer, which is considered a chronic disease, and met the etiological components (disease burden), reduced food intake was not considered. Our study primarily focused on the initial step of the GLIM criteria, which involved the identification of malnutrition. Notably, the second step, which included further grading and assessing etiologic criteria, such as reduced food intake and disease burden, was not part of our study’s scope. As previous research has highlighted low BMI and inflammation as the most prevalent combination leading to malnutrition diagnosis in cirrhotic patients (27), we did not consider reduced muscle mass as part of the phenotypic criteria.

Statistical Analysis

All analyses were performed using the IBM SPSS Statistics (version 29.0) software package. Data are presented as mean and standard deviation (SD) or median and interquartile range (IQR) for continuous variables, and counts and percentages n (%) for categorical variables. The Mann-Whitney U test and Shapiro-Wilk test were used to compare the characteristics of participants with and without malnutrition as defined by the GLIM criteria. Categorical variables were compared using the Pearson’s chi-square test or Fisher’s exact test, as appropriate. The prevalence of malnutrition was compared in parallel with other tools used for the GLIM. Agreement between the GLIM criteria and NRS-2002 or PG-SGA was assessed using the kappa (k) statistic, with k values ranging from 0 to 1. Values <0.2 indicated poor agreement, 0.2–0.4 indicated fair agreement, 0.4–0.6 indicated moderate agreement, 0.6–0.8 indicated substantial agreement, and >0.8 indicated almost perfect concordance (28). Laboratory data were retrospectively analyzed using the GLIM criteria. Univariate and multivariate logistic regression analyses were used to identify independent risk factors associated with malnutrition. In the univariate analysis, variables with a p-value <0.05 were included in the subsequent multivariate analysis. A p-value <0.05 was considered statistically significant.

Results

Characteristics of the Participants

The participant characteristics are presented in Table 1. A total of 207 patients (82.1% males) with a mean age of 56.6 ± 11.3 years were included in this study. The characteristics of the patients are listed in Table 1. According to the GLIM criteria, patients in Child-Pugh Class B (73.7 vs. 26.3%) had a significantly higher incidence of malnutrition than those in Class A (35.8 vs. 64.2%). Further, those without ascites (65.3 vs. 34.7%) had a significantly higher rate of non-malnourished individuals (p = 0.009). The malnutrition group had the highest ALBI score (p = 0.001), oldest age (p = 0.004), lowest hemoglobin level (p < 0.001), lowest lymphocyte count (p = 0.004), and lowest serum albumin level (p = 0.001).

Table 1. Characteristics of the participants with and without malnutrition according to the GLIM criteria.

Variables	Total (n = 207)	The GLIM criteria		p-Value
		Malnourished (n = 89)	Non-malnourished (n = 118)
Sex (%)				0.136
Male	170 (82.1)	69 (40.1)	101 (59.9)
Female	37 (17.9)	20 (54.1)	17 (45.9)
Age (years), means	56.6 ± 11.3	59.3 ± 11.7	54.8 ± 10.7	0.004
BMI (kg/m^2), median (IQR)	22.7 (20.3, 25.3)	21.4 (18.7, 24)	23.8 (21.3, 25.6)	<0.001
L3-SMI	43.3 (41.5, 48.6)	42.1 (38.8, 43.6)	44.7 (42.5, 51.7)	<0.001
BCLC stage (%)				0.438
0–A	94 (45.9)	33 (35.1)	61 (64.9)
B	28 (13.5)	17 (60.7)	11 (39.3)
C	82 (39.6)	38 (46.3)	44 (53.7)
D	3 (1)	1 (33.3)	2 (66.7)
Child-Pugh class (%)				<0.001
A	165 (79.7)	59 (35.8)	106 (64.2)
B	38 (18.4)	28 (73.7)	10 (26.3)
C	4 (1.9)	2 (50)	2 (50)
PHT (%)				0.995
No	86 (41.5)	37 (43)	49 (57)
Yes	121 (58.5)	52 (42.9)	69 (57.1)
Tumor diameter (cm) (%)				0.519
≦5	110 (53.1)	45 (40.9)	65 (59.1)
>5	97 (46.9)	44 (45.4)	53 (54.6)
Tumor number (%)				0.174
<4	143 (69)	57 (39.9)	86 (60.1)
≧4	64 (31)	32 (50)	32 (50)
ALBI score median (IQR)	−2.40 (−2.70, −1.98)	−2.24 (−2.6, −1.8)	−2.5 (−2.8, −2.2)	0.001
ASC, n (%)				0.009
No	126 (60.5)	45 (35.7)	81 (64.3)
Yes	81 (39.5)	44 (54.3)	37 (45.7)
HGB (g/L), median (IQR)	126 (112, 139)	118 (103, 134)	131 (118, 142)	<0.001
LYM (g/L), median (IQR)	1.47 (1.06, 1.9)	1.3 (1, 1.6)	1.6 (1.1, 2.1)	0.004
ALB (g/dL), means	36.4 ± 5	35 ± 5	37.3 ± 5	0.001

BMI: body mass index; PHT: portal hypertension; ASC: ascites; HGB: hemoglobin; LYM: lymphocyte count; ALB: serum albumin; L3-SMI: Lumbar 3 Skeletal Muscle Index.

Nutrition Statues

Table 2 shows that 30.4% (63) of the participants were categorized as being at risk of malnutrition according to NRS-2002. The agreement between the GLIM criteria and NRS-2002 had a K-value of 0.61 (p < 0.01). NRS-2002 had low sensitivity and high specificity for identifying malnutrition according to the GLIM criteria (sensitivity: 64%, specificity: 94.9%). Of the participants, 43% (89) and 54.6% (113) were diagnosed as malnourished by the GLIM criteria and PG-SGA, respectively. The agreement between the GLIM criteria and PG-SGA had a K-value of 0.68 (p < 0.001). PG-SGA had high sensitivity and low specificity in identifying malnutrition according to the GLIM criteria (sensitivity: 94.4%, specificity: 75.4%).

Table 2. Cross tabulation of the results of NRS2002, PG-SGA, and the GLIM criteria for the diagnosis of malnutrition.

Category		NRS 2002		PG-SGA		Total
		≥3	<3	≥4	<4
		At risk	No risk	Malnourished	Non-malnourished
The GLIM criteria, n (%)	Malnourished	57 (65.9)	32 (34.1)	84 (94.4)	5 (5.6)	89
	Non-malnourished	6 (5.7)	112 (94.3)	29 (24.6)	89 (75.4)	118
K value (p-value)		0.61 (<0.001)		0.68 (<0.001)

K: kappa.

Clinical Parameters Associated with the GLIM Criteria

As summarized in Table 3, in logistic regression analysis, elderly patients were defined as those ≥65 years old (29), hypoalbuminemia was defined as serum albumin concentration <35 g/L (30), and anemia was defined as hemoglobin concentration <110 g/L for females, and <120 g/L for males (31). Lymphopenia was defined as an absolute count below 0.8 × 10⁹/L (32). BCLC staging was performed in subgroups of 0–A and B–C (33). The cutoff points of ALBI were as follows: grade 1 (≤ −2.60), grade 2 (> −2.60 to ≤ −1.39), and grade 3 (> −1.39). Age, Child-Pugh score, hemoglobin level, lymphocyte count, serum albumin level, and ascites differed significantly between patients with and without malnutrition in univariate analyses. After adjusting for sex, age, hemoglobin level, lymphocyte count, serum albumin level, and ascites variables in the multivariate logistic regression analysis, significant associations with malnutrition persisted for age, anemia, and ascites (Hosmer and Lemeshow goodness of fit x2 = 9.78, p = 0.28). In applying GLIM criteria, 89 cases were identified as malnutrition. Among these, 78 experienced weight loss, 23 had a low BMI, and 31 exhibited muscle depletion. Specifically, 62 cases were independently diagnosed as malnutrition due to weight loss, seven due to a low BMI, and four due to muscle depletion. Additionally, 16 cases were concurrently diagnosed with weight loss and low BMI, 11 with low BMI and muscle depletion, and 16 with weight loss and muscle depletion.

Table 3. Univariate and multivariate logistic regression analyses for the association of variables with malnutrition by GLIM criteria.

Variables GLIM	Univariable analysis	Multivariable analysis
	OR (95%CI)	OR (95%CI)
Age (≧65/<65)	2.31 (1.24–4.30)	1.04 (1.01–1.07)
Sex (male/female)	0.58 (0.28–1.19)
BCLC stage (CD/0AB)	1.25 (0.71–2.19)
Child-Pugh grade (BC/A)	4.49 (2.14–9.43)
PHT (yes/no)	1.00 (0.57–1.74)
Tumor diameter (>5/≦5)	1.20 (0.69–2.08)
Tumor number (≧4/<4)	1.51 (0.83–2.73)
ALBI grade
2/1	0.51 (0.12–2.21)
3/1	0.92 (0.22–3.86)
ASC (yes/no)	2.14 (1.21–3.78)	2.01 (1.10–3.72)
Anemia (yes/no)	2.88 (1.61–5.14)	2.06 (1.01–4.23)
Hypoalbuminemia (yes/no)	2.51 (1.40–4.50)	1.33 (0.62–2.85)
Lymphopenia	1.76 (1.01–3.07)	0.49 (0.23–1.06)

PHT: portal hypertension; ASC: ascites; HGB: hemoglobin; LYM: lymphocyte count; ALB: serum albumin; OR: odds ratios; 95% CI: 95% confidence interval.

Discussion

Nutritional challenges are common in patients with chronic liver diseases. Previous studies have shown that the prevalence of malnutrition according to the GLIM criteria for chronic liver disease and cirrhosis was 21 and 33.3%, respectively (27,34). In this study, we primarily focused on patients with HCC and applied the GLIM criteria to assess their nutritional status. Our findings indicate a higher prevalence of malnutrition (41.1%) than that reported in previous studies. This difference could be attributed to the use of an alternative approach.

Previous research has revealed that more than 30% of patients at risk of malnutrition would have been missed if NRS-2002 screening had been introduced (35,36). Furthermore, the absence of prior screening significantly increased the prevalence of malnutrition diagnosed using the GLIM criteria (37,38). To avoid this potential bias, we adopted a different method by excluding the two-step approach in which the NRS 2002 was used as the initial screening tool (39,40). Instead, we independently evaluated and validated the performance of these tools. Our findings indicate that the NRS-2002 exhibited substantial agreement with the GLIM criteria. However, the overall detection rate of individuals at risk of malnutrition using the NRS-2002 was only 30.4% (63/207). In contrast, the GLIM criteria diagnosed 43% (89/207) of individuals as malnourished. This suggests that the NRS-2002 may have a high false-negative rate in patients with HCC, leading to the omission of cases. This observation aligns with previous studies, emphasizing the need for caution when relying on the NRS 2002 as an initial screening tool for malnutrition. It is crucial to acknowledge this limitation and report the incidence of malnutrition using the GLIM criteria regardless of whether previous screening has been conducted.

It is worth noting that without prior nutritional risk screening, the PG-SGA exhibited substantial agreement with the GLIM criteria in our study. The PG-SGA has been widely accepted as the most suitable approach for assessing the nutritional status of patients with cancer, therefore, it is noteworthy that the GLIM criteria had an accuracy of 78.8% (89/113) for detecting malnutrition and 95.7% (90/94) for identifying non-malnutrition. However, the number of patients diagnosed with malnutrition according to the GLIM criteria was lower than that in the PG-SGA. This difference can be attributed to the PG-SGA's incorporation of seven nutrition-related dimensions: weight, intake, symptoms, functional status, disease state, metabolic stress, and physical nutritional examination. By including these comprehensive dimensions, the PG-SGA assesses various aspects of nutritional status. These dimensions collectively contribute to a more thorough evaluation, potentially capturing a wider range of malnutrition indicators and leading to a higher number of malnutrition diagnoses than the GLIM criteria. Furthermore, among the total of 89 cases identified as GLIM positive, 78 exhibited unintentional weight loss, 23 presented a low BMI, and 31 showed reduced muscle mass. These findings align with the validated and recognized superior phenotypic criterion of unintentional weight loss in the GLIM criteria for patients with cancer (36,41). It is worth emphasizing that despite our efforts to minimize the impact of fluid retention by instructing patients to fast and measure their weight under specific conditions, we observed that 54.3% (44/81) of the patients with ascites were diagnosed as malnourished. This finding highlights a potential limitation when relying solely on weight loss and BMI, as the presence of ascites cannot be completely ruled out and sometimes obscures underlying weight changes, potentially leading to a falsely elevated BMI. Therefore, introducing additional criteria for assessing malnutrition, such as evaluating reduced muscle mass, is crucial, especially for patients with HCC. By incorporating these comprehensive dimensions into the assessment, as seen in the PG-SGA, a more thorough evaluation can be achieved. These dimensions collectively contribute to a more comprehensive assessment, potentially capturing a wider range of malnutrition indicators and leading to a higher number of malnutrition diagnoses than the GLIM criteria.

Additionally, in our study, the prevalence of malnutrition detected using the GLIM and PG-SGA criteria was 41.1 and 54.6%, respectively. These results differ from those of a previous study conducted in 22 cities including more than 80 hospitals in China. This study reported that the overall prevalence of malnutrition in patients with liver cancer was 79.9% (42). The lower prevalence of malnutrition may be attributed to the relatively lower proportion of advanced-stage cancer patients in our sample, with BCLC stages 0–B accounting for 59.4% and Child-Pugh class A accounting for 79.7%. In addition, our study only evaluated the nutritional status of patients on admission, before undergoing subsequent treatment, and before experiencing the potential side effects thereof. Therefore, the overall rates of malnutrition in our study may be significantly lower than those in other studies that included advanced-stage cancer patients and considered the impact of treatment and related side effects on nutritional status.

In the multivariate analysis, we identified several independent risk factors for malnutrition based on the GLIM criteria, including age >65 years, anemia, and ascites.

Age has long been recognized as a well-established risk factor for malnutrition and is considered a phenotype of the GLIM criteria (43,44). Our study found a 4% increase in the incidence of malnutrition among the elderly population. Therefore, healthcare providers should be vigilant when assessing malnutrition risk in elderly patients, as they may require more focused nutritional support and interventions.

Ascites, the accumulation of fluid in the abdominal cavity, is a common complication of liver disease, as well as it is an important indicator within the Child-Pugh score. It reflects the severity of liver disease and aids in the assessment of a patient’s liver function and prognosis. A significant body of literature has demonstrated that the Child-Pugh score and ascites are critical predictors of malnutrition in patients with liver disease (45–48). This finding emphasizes the need to incorporate liver disease severity into the malnutrition assessment process. Patients with more advanced liver disease are at greater risk of malnutrition, and early identification is crucial for providing appropriate nutritional support. In our study, patients with ascites had a 101% higher risk for malnutrition. Considering ascites in assessments using the GLIM criteria is essential as it reflects the impact of fluid retention on nutritional status. Identifying malnutrition in patients with ascites allows for tailored interventions that address specific needs by considering both fluid management and nutritional support. Moreover, the presence of ascites should serve as a reminder to health care professionals to consider the inclusion of muscle depletion as an essential criterion within the GLIM phenotype, potentially enhancing the accuracy of malnutrition diagnosis.

Anemia is a common complication of liver disease (49–51) and a significant concern in cancer patients (52,53). Iron deficiency, the most common type of anemia, has been observed in patients with compensated and decompensated cirrhosis. Intriguingly, administering blood transfusions to individuals with anemia can inadvertently lead to secondary iron overload, thereby elevating the risk of HCC and mortality. Notably, our study found that anemia substantially increased the incidence of malnutrition by ∼106%. This underscores the critical importance of considering anemia when applying the GLIM criteria. Recognizing the impact of anemia on malnutrition risk empowers healthcare providers to effectively tailor nutritional interventions to address this specific challenge.

Our findings underscore the substantial impact of age, anemia, and ascites as independent risk factors within the GLIM criteria, contributing significantly to malnutrition among patients with HCC. This recognition is crucial for healthcare professionals to optimize nutritional support and enhance patient outcomes.

Although our study conducted GLIM criteria-based assessments after HCC diagnosis, it is important to acknowledge the timing of these assessments as a relevant consideration. The optimal timing of the GLIM-based nutritional evaluation remains an area for further investigation. For instance, early assessment, even before treatment initiation, may facilitate timely intervention, possibly reducing the risk of malnutrition. Incorporating these risk factors into the GLIM criteria enhances its accuracy in identifying patients at nutritional risk. Healthcare providers can implement more effective nutritional interventions through timely recognition and addressing of these independent risk factors. This, in turn, can lead to improved patient outcomes and better chances of recovery in patients with HCC.

Limitations

Our study has several noteworthy limitations that should be taken into account. First, this was a single-center cross-sectional study with a relatively small sample size. This may limit the generalizability of our findings to broader patient populations, especially those undergoing treatment. Additionally, our retrospective analysis introduced the potential for bias due to missing data. It is important to note that the complete GLIM criteria involves a comprehensive two-step approach, including further grading. Meanwhile, our study focused solely on the initial step of identifying individuals at risk for malnutrition based on phenotypic and etiologic criteria.

To address these limitations and advance our understanding of nutritional care in HCC patients, future research should aim to conduct prospective studies with larger and more diverse participant groups. Such efforts could enhance the reliability of our findings and facilitate a more in-depth exploration of specific phenotypes associated with HCC. Furthermore, investigating the timing of nutritional assessments at various points along the patient’s journey could refine nutritional care strategies and ultimately contribute to improved patient well-being.

Conclusion

The GLIM criteria can be a valuable tool for diagnosing malnutrition in HCC patients, with NRS 2002 and PG-SGA as complementary options. Understanding independent risk factors, such as age, ascites, Child-Pugh score, and anemia when using the GLIM criteria enhances the accuracy of identifying malnutrition and underscores the significance of managing these factors to optimize patient outcomes.

Ethical Approval

This cross-sectional study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board and Ethics Committee of Liuzhou People’s Hospital affiliated with Guangxi Medical University in China (KY2022-030-02). Written informed consent was obtained from all participants upon admission.

Consent for Publication

Written informed consent for publication was obtained from all participants included in this study. The consent form explicitly included consent for the publication of anonymized data. All personal identification information was removed to protect the privacy and confidentiality of participants.

Author Contributions

ST and JJ designed the study, recruited participants, and collected data. JJ analyzed data. ST and JJ wrote the paper. LQ, YL, and JM participated in data collection. BX supervised the study and critically revised the article. Ning Tan supported the depiction of skeletal muscle mass, supplementing the missing data in GLIM. All authors reviewed and edited the final article. Approve of the final version of the manuscript and declare that the content has not been published elsewhere.

Acknowledgments

The authors would like to thank the Department of Scientific Research at Liuzhou People’s Hospital for the literature review and all patients for their participation.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

Data and materials used in this study are available upon request. Interested researchers may contact the corresponding author or the Institutional Review Board and Ethics Committee of Liuzhou People’s Hospital affiliated with Guangxi Medical University in China (email:[email protected]) to initiate the data sharing process. Owing to privacy and confidentiality concerns, access to raw data and specific patient information can only be granted after obtaining the necessary ethical and legal approval.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by Clinical Application Research of Whole Course Nutrition Management Based on ERSA in Hepatic Surgery, 2020 Guangxi Zhuang Autonomous Region Health Commission self-funded project (Z20200517), Clinical study of sarcopenia on the prognosis of hepatocellular carcinoma with transarterial chemoembolization (TACE) treatment, 2020 Liuzhou Science and Technology Plan Project (2020NBAB0813), Association between nutritional status and sarcopenia in patients with liver cancer according to GLIM criteria. Guangxi Zhuang Autonomous Region Health Commission self-funded project (Z-B20221396), National Natural Science Foundation of China (82160699), and Guangxi Natural Science Foundation of China (2023GXNSFAA026477).