Association between GLIM‐defined malnutrition and in‐hospital costs in adults with sepsis: A sub analysis of a prospective cohort study

Yuki Hashimoto; Minoru Yoshida; Tomoka Miyagi; Akane Kariya; Hitomi Tanoue; Naoki Kanda; Yusuke Hoshi; Takahiro Fujita; Kensei Togo; Tomoaki Obata; Kenzo Ishii; Katsuyuki Abe; Chihiro Sato; Yoshifumi Heshiki; Daisetsu Yasumura; Shizuko Gomi; Masafumi Takahashi; Takashi Hongo; Keiji Sakurai; Miho Nishihara; Yuko Matsuda; Kiyota Chinen; Akiko Arai; Michihiro Maeda; Yuhei Irie; Hiroshi Kamijo; Kana Hosokawa; Daisuke Kasugai; Michiko Higashi; Ayaka Matsuoka; Hirotaka Ogawa; Yuichi Araki; Kenji Oike; Daigo Shimakura; Hajime Katsukawa; Toru Kotani; Takayuki Ogura; Keibun Liu; Kensuke Nakamura

doi:10.1002/jpen.70120

. 2026 Jul 4;50(7):985–995. doi: 10.1002/jpen.70120

Association between GLIM‐defined malnutrition and in‐hospital costs in adults with sepsis: A sub analysis of a prospective cohort study

Yuki Hashimoto ¹, Minoru Yoshida ^1,^2,^✉, Tomoka Miyagi ³, Akane Kariya ⁴, Hitomi Tanoue ⁵, Naoki Kanda ⁶, Yusuke Hoshi ⁷, Takahiro Fujita ⁸, Kensei Togo ⁹, Tomoaki Obata ⁹, Kenzo Ishii ¹⁰, Katsuyuki Abe ¹¹, Chihiro Sato ¹¹, Yoshifumi Heshiki ¹², Daisetsu Yasumura ¹³, Shizuko Gomi ¹⁴, Masafumi Takahashi ¹⁵, Takashi Hongo ^16,¹⁷, Keiji Sakurai ¹⁸, Miho Nishihara ¹⁹, Yuko Matsuda ²⁰, Kiyota Chinen ²⁰, Akiko Arai ²¹, Michihiro Maeda ²¹, Yuhei Irie ²², Hiroshi Kamijo ²³, Kana Hosokawa ²⁴, Daisuke Kasugai ²⁵, Michiko Higashi ²⁵, Ayaka Matsuoka ²⁶, Hirotaka Ogawa ²⁶, Yuichi Araki ²⁷, Kenji Oike ²⁸, Daigo Shimakura ²⁹, Hajime Katsukawa ³⁰, Toru Kotani ³¹, Takayuki Ogura ³², Keibun Liu ³³, Kensuke Nakamura ³⁴

^¹Department of Clinical Data Management and Research, Clinical Research Center, National Hospital Organization Headquarters, Tokyo, Japan

^²Department of Health Data Science, Graduate School of Data Science, Yokohama City University, Yokohama, Japan

^³Anesthesiology and Critical Care Medicine, Master's Degree Program, Graduate School of Medicine, Yokohama City University, Yokohama, Japan

^⁴Department of Nursing, Yokosuka General Medical Center, Yokosuka, Japan

^⁵Department of Nutrition, St. Marianna University School of Medicine, Yokohama City Seibu Hospital, Yokohama, Japan

^⁶Division of General Internal Medicine, Jichi Medical University, Tochigi, Japan

^⁷Department of Nutrition, Hitachi General Hospital, Ibaraki, Japan

^⁸Department of Rehabilitation, Hitachi General Hospital, Ibaraki, Japan

^⁹Department of Emergency and Intensive Care Medicine, National Hospital Organization Nagoya Medical Center, Nagoya, Japan

^¹⁰Department of Anesthesiology, Intensive Care Unit, Fukuyama City Hospital, Hiroshima, Japan

^¹¹Department of Nutrition, Japanese Red Cross Maebashi Hospital, Gunma, Japan

^¹²Department of Critical Care, Naha City Hospital, Okinawa, Japan

^¹³Department of Rehabilitation, Naha City Hospital, Okinawa, Japan

^¹⁴School of Medicine, University of Tsukuba, Ibaraki, Japan

^¹⁵Department of Rehabilitation, University of Tsukuba Hospital, Ibaraki, Japan

^¹⁶Department of Emergency, Critical Care, and Disaster Medicine, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Okayama, Japan

^¹⁷Department of Emergency, Okayama Saiseikai General Hospital, Okayama, Japan

^¹⁸Department of Emergency Medicine, SUBARU Health Insurance Society Ota Memorial Hospital, Gunma, Japan

^¹⁹Department of Nutrition, SUBARU Health Insurance Society Ota Memorial Hospital, Gunma, Japan

^²⁰Department of Intensive Care Medicine, Okinawa Kyodo Hospital, Okinawa, Japan

^²¹Department of Emergency and Critical Care Medicine, Tsukuba Medical Center Hospital, Ibaraki, Japan

^²²Department of Emergency and Critical Care Medicine, Fukuoka University Hospital, Fukuoka, Japan

^²³Department of Emergency and Critical Care Medicine, Shinshu University School of Medicine, Nagano, Japan

^²⁴Division of Clinical Nutrition, Shinshu University Hospital, Nagano, Japan

^²⁵Department of Emergency and Critical Care Medicine, Nagoya University Graduate School of Medicine, Nagoya, Japan

^²⁶Department of Emergency and Critical Care Medicine Faculty, Saga University Hospital, Saga, Japan

^²⁷Department of Emergency and Critical Care Medicine, Tsuchiura Kyodo General Hospital, Ibaraki, Japan

^²⁸Department of Rehabilitation, Tsuchiura Kyodo General Hospital, Ibaraki, Japan

^²⁹Graduate School of Data Science, Shiga University, Shiga, Japan

^³⁰Japanese Society for Early Mobilization, Tokyo, Japan

^³¹Department of Intensive Care Medicine, Showa Medical University, Tokyo, Japan

^³²Department of Emergency Medicine and Critical Care Medicine, Tochigi Prefectural Emergency and Critical Care Center, Saiseikai Utsunomiya Hospital, Tochigi, Japan

^³³Non‐Profit Organization ICU Collaboration Network (ICON), Tokyo, Japan

^³⁴Department of Disaster and Emergency Medicine, Graduate School of Medicine, Kobe University, Hyogo, Japan

Correspondence Minoru Yoshida, MD, PhD, Department of Health Data Science, Graduate School of Data Science, Yokohama City University, 3‐9 Fukuura, Kanazawa‐ku, Yokohama, Kanagawa 236‐0004, Japan. Email: y265754g@yokohama-cu.ac.jp

^✉

Corresponding author.

PMCID: PMC13636117 PMID: 42400337

Abstract

Background

International statements suggest using the Global Leadership Initiative on Malnutrition (GLIM) criteria in intensive care units (ICUs); however, the economic impact of GLIM‐defined malnutrition in sepsis remains unclear. This study investigated the association between GLIM‐defined malnutrition and in‐hospital costs in sepsis.

Methods

We conducted a sub‐analysis of the prospective cohort study, Investing Long‐term Outcomes of Sepsis or Septic shock, among medically managed critically ill patients with sepsis in 15 Japanese ICUs. Associations between GLIM‐defined malnutrition and cost categories (low, moderate, and high) were assessed using multinomial logistic regression (reference: low‐cost group), adjusted for age, sex, Charlson Comorbidity Index, and Sequential Organ Failure Assessment score. Individual GLIM components were also examined. Total in‐hospital costs and secondary outcomes were compared between malnutrition and non‐malnutrition groups using inverse probability of treatment weighting.

Results

Among 259 patients, 111 had GLIM‐defined malnutrition. GLIM‐defined malnutrition was associated with both the moderate‐cost (adjusted odds ratio [AOR]: 2.03, 95% confidence interval [CI]: 1.05–3.89) and high‐cost classifications (AOR: 2.27, 95% CI: 1.20–4.29). Among GLIM components, reduced muscle mass (AOR: 1.95, 95% CI: 1.04–3.68) and reduced food intake (AOR: 2.61, 95% CI: 1.39–4.92) were associated with the high‐cost classification. After weighting, the malnutrition group had significantly higher median in‐hospital costs (24,959 vs. 18,651 USD), longer hospital stays, and longer ICU stays than the non‐malnutrition group.

Conclusions

GLIM‐defined malnutrition—particularly reduced muscle mass and reduced food intake—was associated with higher in‐hospital costs in sepsis and may be a potential indicator of high‐cost hospitalization.

Keywords: economic burden, hospital costs, intensive care unit, malnutrition, sepsis

graphic file with name JPEN-50-985-g003.webp

INTRODUCTION

Malnutrition is a major factor associated with increased mortality and complications among hospitalized patients. ¹ Beyond clinical outcomes, malnutrition imposes a substantial economic burden globally: in‐hospital costs increase by 1.3‐ to 1.7‐fold across multiple countries, ² , ³ based on traditional nutritional screening and assessment tools. To date, however, most of these economic evaluations have focused on general ward populations. Notably, an economic analysis estimated that the additional in‐hospital costs associated with malnutrition were approximately 6.5‐fold higher among ICU patients than among general ward patients. ⁴

In 2018, the Global Leadership Initiative on Malnutrition (GLIM) criteria were introduced as a standardized diagnostic framework for malnutrition. ⁵ Numerous studies have demonstrated that GLIM‐defined malnutrition is associated with higher mortality and prolonged hospitalization. ⁶ International statements have suggested using the GLIM criteria to diagnose malnutrition in ICUs. ⁷ Moreover, nutrition‐related risk has been associated with adverse outcomes among critically ill patients with sepsis. ⁸ Sepsis is one of the most costly conditions requiring hospitalization, ⁹ particularly in the ICU, where sepsis‐related costs are approximately 1.5‐fold higher than those of non‐sepsis admissions. ¹⁰ Given the substantial economic burden of sepsis in the ICU, identifying cost drivers is essential, ¹¹ making malnutrition a plausible and clinically important candidate. ⁴ However, evidence supporting the association between GLIM‐defined malnutrition and healthcare costs among critically ill patients with sepsis remains limited.

Therefore, we conducted a sub‐analysis of the multicenter prospective cohort study, Investing Long‐term Outcomes of Sepsis or Septic shock (ILOSS), ¹² to assess the association between GLIM‐defined malnutrition at ICU admission and total in‐hospital costs among critically ill patients with sepsis. We hypothesized that GLIM‐defined malnutrition at ICU admission would be associated with higher total in‐hospital costs in these patients. Clarifying this association may help inform future research on nutritional assessment and intervention strategies in critical care.

METHODS

Study design and ethics

This study was a sub‐analysis of the ILOSS study, ¹² a multicenter prospective observational study involving 21 ICUs in Japan. ¹³ For this sub‐analysis, we used data from 15 ICUs where detailed nutrition therapy information had been prospectively collected. The remaining 6 ICUs did not participate in nutritional data collection and were therefore not eligible for inclusion in this sub‐analysis. In the parent ILOSS study, a per‐site enrollment target of up to 20 cases was used to support balanced representation across participating sites and to limit center‐size imbalance. ¹⁴ The study protocol was approved by the central institutional review board (Saiseikai Utsunomiya Hospital, No. 2019‐72) and all local review boards. The parent ILOSS study was registered with the UMIN Clinical Trials Registry (UMIN000041433). Written informed consent was obtained from all patients or their legally authorized representatives. This study adhered to the Declaration of Helsinki and followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. ¹⁵

Participants

In the parent ILOSS study, patients admitted to the ICU between November 2020 and April 2022 were included if they met the Sepsis‐3 criteria for sepsis or septic shock at admission. ¹³ , ¹⁶ The ILOSS protocol excluded patients based on the following criteria: age <18 years; a diagnosis of COVID‐19; discharge from the ICU within 48 h; a central nervous system disorder not caused by sepsis; inability to communicate due to pre‐existing psychiatric symptoms; inability to walk independently before admission; or being in an end‐of‐life or terminal stage.

For our sub‐analysis, we further included only patients with detailed nutritional data. We also applied two additional exclusion criteria for this analysis: (1) patients who underwent surgery before ICU admission—to avoid the confounding influence of surgical costs on the primary outcome (total in‐hospital costs) ¹⁷ —and (2) patients with missing in‐hospital cost data, as these lacked the necessary information for our primary outcome analysis.

Outcome variables

The primary outcome was total in‐hospital costs derived from administrative claims data, converted to US dollars using the 2022 mean exchange rate (130 JPY/USD). Secondary outcomes included in‐hospital mortality; quality of life (QOL) score at discharge using the EuroQoL 5‐dimension 5‐level instrument, with death assigned a score of 0 ¹⁸ ; length of hospital stay after ICU admission (LOS); length of ICU stay; and duration of mechanical ventilation.

Covariates

The exposure was GLIM‐defined malnutrition, and the comparator was non‐malnutrition, both assessed at ICU admission. ⁵ We collected the following covariates at ICU admission: age; sex; employment status before admission; admission year; body mass index (BMI); Charlson Comorbidity Index (CCI) score for assessing comorbidities ¹⁹ ; and Sequential Organ Failure Assessment (SOFA) score ¹⁶ to evaluate illness severity. We also gathered information on treatments administered upon ICU admission, including mechanical ventilation, vasopressors, inotropes, and sedation. Additionally, we recorded whether septic shock was present and identified the primary site of infection.

Definition of malnutrition

Malnutrition was defined at ICU admission using the GLIM criteria, which require the presence of at least one phenotypic and one etiologic criterion for a diagnosis. ⁵ We evaluated malnutrition at ICU admission because this time point is clinically relevant for early risk stratification and initial nutritional management. ⁷

(1)
Phenotypic criteria

The three phenotypic criteria were assessed as follows. Weight loss was defined as a reduction of ≥5% in 6 months. Low BMI was defined according to the GLIM criteria for Asian populations: BMI < 18.5 kg/m² for patients <70 years and <20 kg/m² for those ≥70 years. Reduced muscle mass was determined by registered dietitians using methods such as bioelectrical impedance analysis and computed tomography, ²⁰ with the final clinical judgment recorded.
(2)
Etiologic criteria

The two etiologic criteria were defined as follows. Given the difficulty of accurately quantifying dietary intake before ICU admission in critically ill patients, reduced food intake was defined as ≤200 kcal/day—a pragmatic threshold approximating a typical basic meal unit in Japan. This threshold is also consistent with the clinical concept of trophic feeding in ICU nutrition therapy, which refers to a minimal level of energy intake intended to maintain gut integrity. ²¹ This information was prospectively collected by ICU dietitians through interviews with patients or their families. Inflammation was defined as a C‐reactive protein level ≥5 mg/dL upon ICU admission, based on the GLIM guidance for severe acute inflammation. ²²

Statistical analysis

Baseline patient characteristics were summarized using descriptive statistics. Continuous variables were presented as means with standard deviations (SDs) or medians with interquartile ranges (IQRs), as appropriate. Differences between groups were assessed using the Welch's t‐test, Wilcoxon rank‐sum test, or the Kruskal‐Wallis test, depending on the number of groups and data distribution. Categorical variables were presented as numbers and percentages, and associations were examined using the chi‐square test. Missing values were imputed using a random forest‐based algorithm. ²³

(1)
Primary analysis

We evaluated whether GLIM‐defined malnutrition was associated with classification into higher‐cost groups. Because healthcare cost data typically exhibit a right‐skewed distribution, standard linear regression is often inappropriate. ²⁴ A common approach in health economics is to categorize costs based on percentiles to analyze the risk factors for high resource utilization. ¹⁷ , ²⁵ , ²⁶ We categorized costs into three groups: low (0–50th percentile), moderate (50–75th percentile), and high (75–100th percentile). We performed a multinomial logistic regression analysis to estimate the odds of being in the moderate‐ or high‐cost categories relative to the low‐cost reference category. The model was adjusted for key clinical confounders, including age, ²⁷ sex, CCI, ²⁸ and SOFA score ²⁹ at ICU admission, with adjusted odds ratios (AORs) and 95% confidence intervals (CIs). For comparison, the Wald test was used.

To confirm the robustness of our findings, we performed the following sensitivity analyses: a complete case analysis to assess the potential impact of missing‐data handling; an analysis among patients discharged alive to assess the potential influence of in‐hospital mortality on cost outcomes; an analysis additionally adjusted for early enteral nutrition and adequate energy intake (defined in Figure 2) ³⁰ ; and an analysis restricted to patients with LOS ≥ 7 days to verify that the overall results were consistent with the population receiving these nutrition therapies.

Furthermore, to examine the contribution of each GLIM component to the primary outcome, we conducted an additional multinomial logistic regression analysis. In this analysis, each of the five individual components (phenotypic criteria: weight loss, low BMI, reduced muscle mass; etiologic criteria: reduced food intake, inflammation) was separately entered as the main explanatory variable into the model. This model was adjusted for the same clinical confounders: age, sex, CCI, and SOFA score at ICU admission.
(2)
Secondary analysis

We used inverse probability of treatment weighting (IPTW) to balance measured baseline covariates between patients with and without GLIM‐defined malnutrition at ICU admission, thereby enabling comparison of the primary and secondary outcomes with reduced confounding. ³¹ The propensity score for the GLIM‐defined malnutrition at ICU admission was calculated using a logistic regression model, which included all baseline covariates except the components of the GLIM criteria. ⁵ We evaluated the balance between the groups using the standardized mean difference (SMD), with an SMD value ≤ 0.1 considered well balanced. ³² After weighting, we compared primary and secondary outcomes between the malnutrition and non‐malnutrition groups.

Overall and sensitivity analyses of the association between GLIM‐defined malnutrition and cost categories in sepsis. The forest plots show the AORs from the multinomial logistic regression. Panel (A) presents the results for the moderate‐cost group (50th–75th percentile), and Panel (B) for the high‐cost group (75th–100th percentile), both using the low‐cost group (0–50th percentile) as the reference. (a) “EEN (+)” indicates inclusion of the additional moderating variable for patients who were started on enteral nutrition within 3 days of ICU admission in the multinomial logistic regression model. (b) “AN (+)” indicates inclusion of the additional moderating variable for patients received energy intake of ≥20 kcal/kg/day at least once between Days 4 and 7 after ICU admission in the multinomial logistic regression model. *Based on the Wald test. AOR, adjusted odds ratio; GLIM, Global Leadership Initiative on Malnutrition; ICU, intensive care unit; LOS, length of stay after ICU admission.

All statistical analyses were performed using R (version 4.3.1). Significance testing was performed with the use of a two‐sided alpha level of 0.05.

RESULTS

Study population

The study included 259 eligible patients (Figure 1). Missing data were imputed for two covariates: malnutrition status and CCI (maximum missing rate <10%; Table S1). Baseline characteristics were similar between the overall cohort and the complete‐case cohort, with SMDs ≤0.1 (Table S2). The primary outcome, total in‐hospital costs, exhibited a right‐skewed distribution across the cohort (Figure S1).

Flow diagram of the enrollment of study participants. Three cost groups were defined: low (0–50th percentile [p0–50]), moderate (p50–75), and high (p75–100). (a) “Post surgery” indicates that the patients underwent surgery before ICU admission. (b) “Missing cost data” indicates the patients with missing in‐hospital cost data. COVID‐19, coronavirus disease 2019; ICU, intensive care unit; ILOSS study, Investing Long‐term Outcomes of Sepsis or Septic shock study.

Primary analysis

Patients were stratified into low‐ (n = 130), moderate‐ (n = 62), and high‐cost (n = 67) groups. The median total costs for the low‐, moderate‐, and high‐cost groups were 12,391 USD (IQR: 6,700–17,386), 27,762 USD (IQR: 24,702–29,662), and 46,283 USD (IQR: 40,758–60,887), respectively (Table 1).

Table 1.

Characteristics of patients with sepsis at ICU admission among cost groups.

Variable	Low n = 130	Moderate n = 62	High n = 67	p
Age, years, median (IQR)	77.0 (67.2, 82.0)	72.0 (67.0, 81.8)	74.0 (64.5, 79.5)	0.35^*
Male sex, n (%)	70 (53.8)	38 (61.3)	46 (68.7)	0.13^†
Employed, n (%)	22 (16.9)	9 (14.5)	15 (22.4)	0.47^†
Admission year, n (%)				0.92^†
2020	12 (9.2)	6 (9.7)	9 (13.4)
2021	107 (82.3)	51 (82.3)	53 (79.1)
2022	11 (8.5)	5 (8.1)	5 (7.5)
GLIM‐defined malnutrition, n (%)	42 (32.3)	33 (53.2)	36 (53.7)	0.003^†
Phenotypic criteria, n (%)
Weight loss	21 (16.2)	13 (21.0)	13 (19.4)	0.69^†
Low BMI	22 (16.9)	17 (27.4)	12 (17.9)	0.21^†
Reduced muscle mass	38 (29.2)	21 (33.9)	31 (46.3)	0.06^†
Etiologic criteria, n (%)
Reduced food intake	37 (28.5)	24 (38.7)	34 (50.7)	0.008^†
Inflammation	89 (68.5)	50 (80.6)	51 (76.1)	0.17^†
BMI, kg/m², n (%)				0.09^†
<18.5	12 (9.2)	12 (19.4)	9 (13.4)
18.5–24.9	78 (60.0)	34 (54.8)	30 (44.8)
≥25.0	40 (30.8)	16 (25.8)	28 (41.8)
Charlson comorbidity index, median (IQR)	2.0 (1.0, 2.0)	2.0 (1.0, 3.0)	2.0 (1.0, 3.5)	0.03^*
SOFA score at ICU admission, median (IQR)	8.0 (5.0, 10.8)	9.0 (7.0, 12.0)	9.0 (7.0, 13.0)	0.004^*
Respiration, median (IQR)	1.0 (0.0, 2.0)	2.0 (1.0, 3.0)	2.0 (0.5, 3.0)	0.03^*
Coagulation, median (IQR)	1.0 (0.0, 2.0)	0.0 (0.0, 2.0)	0.0 (0.0, 1.0)	0.10^*
Liver, median (IQR)	0.0 (0.0, 1.0)	0.0 (0.0, 1.0)	0.0 (0.0, 1.0)	0.25^*
Cardiovascular, median (IQR)	3.0 (0.0, 4.0)	4.0 (3.0, 4.0)	4.0 (1.0, 4.0)	0.001^*
Central nervous system, median (IQR)	1.0 (0.0, 3.0)	1.0 (0.0, 4.0)	2.0 (1.0, 4.0)	0.11^*
Renal, median (IQR)	1.0 (0.0, 2.0)	1.0 (0.0, 2.0)	2.0 (1.0, 3.5)	0.04^*
Mechanical ventilation, n (%)	49 (37.7)	43 (69.4)	52 (77.6)	<0.001^†
Vasopressor, n (%)	101 (77.7)	56 (90.3)	64 (95.5)	0.002^†
Inotropes, n (%)	13 (10.0)	21 (33.9)	17 (25.4)	<0.001^†
Sedation, n (%)	58 (44.6)	47 (75.8)	57 (85.1)	<0.001^†
Septic shock, n (%)	89 (68.5)	49 (79.0)	52 (77.6)	0.20^†
Infection site, n (%)				0.001^†
Pulmonary	23 (17.7)	9 (14.5)	14 (20.9)
Intra‐abdominal	28 (21.5)	20 (32.3)	27 (40.3)
Genitourinary	39 (30.0)	8 (12.9)	3 (4.5)
Musculoskeletal	21 (16.2)	12 (19.4)	16 (23.9)
Others	19 (14.6)	13 (21.0)	7 (10.4)
Total cost, median (IQR), USD	12,391 (6,700, 17,386)	27,762 (24,702, 29,662)	46,283 (40,758, 60,887)	<0.001^*

Open in a new tab

Note: Three cost groups were defined: low [0–50th percentile (p0–50)], moderate (p50–75), and high (p75–100).

Abbreviations: BMI, body mass index; ICU, intensive care unit; IQR, interquartile range; SOFA, sequential organ failure assessment.

Based on the Kruskal‐Wallis test.

^{^†}

Based on the Chi‐squared test.

After adjustment, GLIM‐defined malnutrition was significantly associated with classification into both the moderate‐cost (AOR: 2.03, 95% CI: 1.05–3.89, p = 0.03) and high‐cost groups (AOR: 2.27, 95% CI: 1.20–4.29, p = 0.01) compared with the low‐cost group (Table S3). Sensitivity analyses were generally consistent with the overall analysis, although the complete‐case analysis showed a similar trend without reaching statistical significance (Figure 2).

We also examined the contribution of individual GLIM components to this association (Table 2). Among the phenotypic criteria, reduced muscle mass was significantly associated with the high‐cost group (AOR: 1.95, 95% CI: 1.04–3.68, p = 0.04). Among the etiologic criteria, reduced food intake was significantly associated with classification into the high‐cost group (AOR: 2.61, 95% CI: 1.39–4.92, p = 0.003). Neither of these components was significantly associated with the moderate‐cost group.

Table 2.

Adjusted odds ratio between GLIM‐defined malnutrition components and cost categories.

Variable	Moderate‐cost group		High‐cost group
Variable	Odds ratio (95% CI)	p ^*	Odds ratio (95% CI)	p ^*
GLIM‐defined malnutrition	2.03 (1.05, 3.89)	0.03	2.27 (1.20, 4.29)	0.01
Phenotypic Criteria
Weight loss	1.17 (0.53, 2.59)	0.69	1.07 (0.49, 2.36)	0.86
Low BMI	1.79 (0.85, 3.78)	0.13	1.10 (0.50, 2.45)	0.81
Reduced muscle mass	1.09 (0.56, 2.14)	0.80	1.95 (1.04, 3.68)	0.04
Etiologic Criteria
Reduced food intake	1.52 (0.79, 2.92)	0.21	2.61 (1.39, 4.92)	0.003
Inflammation	2.07 (0.97, 4.42)	0.06	1.57 (0.78, 3.16)	0.21

Open in a new tab

Note: Three cost groups were defined: low [0–50th percentile (p0–50)], moderate (p50–75), and high (p75–100). The adjusted odds ratio was calculated by the multinomial logistic regression on cost groups with the low‐cost group as reference group for patients with sepsis.

Abbreviations: 95% CI, 95% confidence interval; BMI, body mass index; GLIM, Global Leadership Initiative on Malnutrition; ICU, intensive care unit.

Based on the Wald test.

Secondary analysis

Patients were stratified into GLIM‐defined malnutrition (n = 111) and non‐malnutrition (n = 148) groups. Before weighting, several baseline characteristics were imbalanced between the groups; however, all covariates were well balanced after weighting, with SMDs ≤0.1 (Table S4).

The primary outcome, total in‐hospital costs, exhibited a right‐skewed distribution in both groups (Figure S2). The malnutrition group had higher median in‐hospital costs than the non‐malnutrition group (24,959 USD [IQR: 15,269–36,593] vs. 18,651 USD [IQR: 9,580–29,840], p = 0.01). Additionally, the malnutrition group had a longer median LOS (29.1 days [IQR: 16.3–46.8] vs. 21.5 days [IQR: 13.0–42.0], p = 0.04) and a longer median ICU stay (7.0 days [IQR: 4.6–11.0] vs. 5.0 days [IQR: 3.0–8.0], p = 0.01) than the non‐malnutrition group. There were no significant differences between the groups in in‐hospital mortality, QOL score at discharge, or duration of mechanical ventilation (Table 3).

Table 3.

Clinical outcomes of GLIM‐defined malnutrition for patients with sepsis at ICU admission after IPTW.

Outcomes	Non‐malnutrition n = 143.7	Malnutrition n = 112.1	p
Total cost, median (IQR), USD	18,651 (9,580, 29,840)	24,959 (15,269, 36,593)	0.01^*
In‐hospital mortality, n (%)	34.9 (24.3)	25.0 (22.3)	0.74^†
QOL score at discharge, mean (SD)	0.49 (0.36)	0.52 (0.34)	0.54^‡
LOS, median (IQR), days	21.5 (13.0, 42.0)	29.1 (16.3, 46.8)	0.04^*
Length of ICU stay, median (IQR), days	5.0 (3.0, 8.0)	7.0 (4.6, 11.0)	0.01^*
Duration of mechanical ventilation, median (IQR), days	1.0 (0.0, 4.6)	1.0 (0.0, 5.5)	0.89^*

Open in a new tab

Abbreviations: GLIM, Global Leadership Initiative on Malnutrition; ICU, intensive care unit; IPTW, inverse probability of treatment weighting; IQR, interquartile range; LOS, length of stay after ICU admission; QOL, quality of life; SD, standard deviation.

Based on the Wilcoxon rank‐sum test.

^{^†}

Based on the Chi‐squared test.

^{^‡}

Based on the Welch's t‐test.

DISCUSSION

In this sub‐analysis of a multicenter prospective study, GLIM‐defined malnutrition was associated with classification into the moderate‐ and high‐cost groups in the primary analysis. Among the individual GLIM components, reduced muscle mass and reduced food intake before ICU admission were associated with classification into the high‐cost group. In the secondary analysis, after IPTW adjustment, the median in‐hospital cost in the GLIM‐defined malnutrition group was significantly higher by approximately 6000 USD than in the non‐malnutrition group.

Given that sepsis imposes one of the highest healthcare and economic burdens, particularly in the ICU, ⁹ , ¹⁰ there is a growing need to identify cost drivers and evaluate cost‐effective interventions. ¹¹ , ³³ To the best of our knowledge, this is the first study to examine the association between GLIM‐defined malnutrition and in‐hospital costs among critically ill patients with sepsis in the ICU. Previous studies, primarily from China and involving non‐sepsis populations, reported relative cost increases of 1.2‐ to 2.0‐fold between GLIM‐defined malnutrition and non‐malnutrition (Table S5). We observed a comparable relative increase (1.3‐fold) in our cohort. However, the absolute economic impact differed: the additional cost in our ICU sepsis cohort was approximately 6000 USD, exceeding the approximately 300 to 3500 USD reported previously (Table S5). These findings suggest that GLIM‐defined malnutrition may have a greater economic impact in more severe or clinically complex populations, including ICU sepsis, than in non‐sepsis cohorts.

GLIM‐defined malnutrition was associated with higher in‐hospital costs and longer hospital and ICU stays, consistent with previous studies suggesting that prolonged LOS is a key cost driver among patients with malnutrition (Table S5). However, GLIM‐defined malnutrition was not associated with in‐hospital mortality or discharge QOL score, suggesting that patients with malnutrition may have required longer hospitalization to reach a functional status sufficient for discharge.

Several mechanisms may contribute to the higher costs associated with malnutrition. Even after the acute inflammatory and catabolic phase has resolved, malnutrition may be accompanied by persistent immune dysfunction, impaired tissue repair, and anabolic resistance, ⁷ increasing susceptibility to infectious complications and often leading to prolonged antimicrobial therapy and longer hospital stays. ³⁴ , ³⁵ A food intake of ≤200 kcal/day, defined as reduced food intake in this study, ²¹ may be insufficient to maintain intestinal structural and functional integrity, potentially impairing gut‐associated lymphoid tissue function and promoting persistent systemic inflammation through bacterial and endotoxin translocation. ³⁶ Reduced muscle mass may reduce the secretion of anti‐inflammatory myokines, further contributing to a pro‐inflammatory state. ³⁷ Collectively, these mechanisms may facilitate the transition to chronic critical illness, a condition common among patients with sepsis, and requiring substantial healthcare resources. ³⁸

Future work is needed in larger and more diverse populations to confirm the association between GLIM‐defined malnutrition and healthcare expenditure, thereby helping to identify potential intervention targets that could shorten hospitalization among patients with sepsis and malnutrition. The association between GLIM‐defined malnutrition and higher cost categories remained consistent after adjustment for early enteral nutrition and adequate energy intake, suggesting that the excess costs may not be fully explained by differences in post‐admission nutritional management. Because pre‐admission nutritional optimization is difficult in emergency sepsis care, future studies should evaluate whether early multidisciplinary strategies, including individualized nutritional support and rehabilitation, can shorten hospitalization and reduce costs in this population. ³⁰ The observed cost difference of approximately 6000 USD may provide a preliminary benchmark for developing cost‐reduction strategies.

Limitations

This study has some limitations. First, the sample size was modest, which constrained the number of covariates that could be included in the multivariable adjustment and may have limited statistical power, particularly for subgroup analyses. To our knowledge, prior data on the association between GLIM‐defined malnutrition and in‐hospital costs were lacking, making it difficult to estimate an appropriate effect size for a priori sample size calculation in critically ill patients with sepsis. However, this modest sample size was a trade‐off for our focus on a specific, homogeneous cohort of medically managed critically ill patients with sepsis. This specificity is a methodological strength, as it reduces the cost variability introduced by different diagnoses and surgery. ⁹ , ¹⁷ Second, although the GLIM criteria distinguish between moderate and severe malnutrition, ⁵ the present analysis did not evaluate outcomes according to malnutrition severity. Because severity data for reduced muscle mass were unavailable and severity thresholds for low BMI in Asian populations are not well established, malnutrition severity could not be determined consistently across all GLIM phenotypic components, precluding a severity‐based analysis. Third, given the observational nature of this study, the influence of unmeasured confounding factors cannot be fully excluded. Our findings should be interpreted as exploratory and associative rather than causal. Nevertheless, we attempted to mitigate confounding as rigorously as possible by employing multiple analytic approaches, including multinomial logistic regression, IPTW, and sensitivity analyses. The consistency of the association between GLIM‐defined malnutrition and high hospitalization costs across these methods supports the robustness of our findings. Fourth, this sub‐analysis was restricted to 15 of the 21 ICUs in the ILOSS study, where detailed nutritional data were collected, and the number of patients enrolled per site was modest. Furthermore, cost analyses were based on the Japanese healthcare system using a fixed exchange rate for interpretability, and both this study and the majority of prior reports were conducted in Asian populations (Table S5). These factors may limit the generalizability of our findings beyond the ICUs included in this analysis and to other settings with different ethnic backgrounds, healthcare systems, and reimbursement structures. Fifth, only total in‐hospital cost data were collected, and itemized data on individual components such as pharmacy, imaging, and nursing were unavailable. Therefore, specific cost drivers could not be identified. Finally, nutritional assessment was not fully standardized across participating hospitals. Reduced muscle mass was assessed using locally available techniques, and pre‐ICU food intake was estimated from interview‐based nutritional assessments. Furthermore, the 200 kcal/day threshold in reduced food intake differs from the cutoff generally recommended in the GLIM framework, ⁵ which may limit comparability with prior research. These procedures may have introduced interinstitutional variability and limited precision in GLIM component assessment. Future studies should use standardized tools, such as bioelectrical impedance analysis, computed tomography, or anthropometry, ³⁹ to improve the validity and reproducibility of muscle mass assessment.

CONCLUSIONS

In medically managed patients with sepsis admitted to the ICU, GLIM‐defined malnutrition was associated with higher in‐hospital costs. These exploratory findings may support prioritizing patients with GLIM‐defined malnutrition as a target population for future studies examining cost‐reduction strategies in critical care.

AUTHOR CONTRIBUTIONS

Yuki Hashimoto: Formal analysis; investigation; methodology; visualization; writing—original draft; writing—review and editing. Minoru Yoshida: Methodology; supervision; writing—original draft; writing—review and editing. Tomoka Miyagi: Investigation; supervision; writing—review and editing. Akane Kariya: Conceptualization; writing—review and editing. Hitomi Tanoue: Visualization; writing—review and editing. Naoki Kanda: Methodology; writing—review and editing. Yusuke Hoshi: Investigation; writing—review and editing. Takahiro Fujita: Investigation; writing—review and editing. Kensei Togo: Investigation; writing—review and editing. Tomoaki Obata: Investigation; writing—review and editing. Kenzo Ishii: Investigation; writing—review and editing. Katsuyuki Abe: Investigation; writing—review and editing. Chihiro Sato: Investigation; writing—review and editing. Yoshifumi Heshiki: Investigation; writing—review and editing. Daisetsu Yasumura: Investigation; writing—review and editing. Shizuko Gomi: Investigation; writing—review and editing. Masafumi Takahashi: Investigation; writing—review and editing. Takashi Hongo: Investigation; writing—review and editing. Keiji Sakurai: Investigation; writing—review and editing. Miho Nishihara: Investigation; writing—review and editing. Yuko Matsuda: Investigation; writing—review and editing. Kiyota Chinen: Investigation; writing—review and editing. Akiko Arai: Investigation; writing—review and editing. Michihiro Maeda: Investigation; writing—review and editing. Yuhei Irie: Investigation; writing—review and editing. Hiroshi Kamijo: Investigation; writing—review and editing. Kana Hosokawa: Investigation; writing—review and editing. Daisuke Kasugai: Investigation; writing—review and editing. Michiko Higashi: Investigation; writing—review and editing. Ayaka Matsuoka: Investigation; writing—review and editing. Hirotaka Ogawa: Investigation; writing—review and editing. Yuichi Araki: Investigation; writing—review and editing. Kenji Oike: Investigation; writing—review and editing. Daigo Shimakura: Investigation; writing—review and editing. Hajime Katsukawa: Project administration; writing—review and editing. Toru Kotani: Project administration; writing—review and editing. Takayuki Ogura: Project administration; writing—review and editing. Keibun Liu: Project administration; writing—review and editing. Kensuke Nakamura: Conceptualization; methodology; project administration; supervision; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

None declared.

Supporting information

Supplemental appendix 1.

JPEN-50-985-s002.docx^{(198.9KB, docx)}

Supplemental appendix 2.

JPEN-50-985-s001.docx^{(19.5KB, docx)}

ACKNOWLEDGEMENTS

Acknowledgements are provided in Supplementary Appendix 2.

REFERENCES

1. Norman K, Pichard C, Lochs H, Pirlich M. Prognostic impact of disease‐related malnutrition. Clin Nutr. 2008;27(1):5‐15. [DOI] [PubMed] [Google Scholar]
2. Guenter P, Abdelhadi R, Anthony P, et al. Malnutrition diagnoses and associated outcomes in hospitalized patients: United States, 2018. Nutr Clin Pract. 2021;36(5):957‐969. [DOI] [PubMed] [Google Scholar]
3. Álvarez‐Hernández J, Planas Vila M, León‐Sanz M, et al. Prevalence and costs of malnutrition in hospitalized patients; the PREDyCES® study. Nutr Hosp. 2012;27(4):1049‐1059. [DOI] [PubMed] [Google Scholar]
4. Correia MITD, Perman MI, Pradelli L, Omaralsaleh AJ, Waitzberg DL. Economic burden of hospital malnutrition and the cost‐benefit of supplemental parenteral nutrition in critically ill patients in Latin America. J Med Econ. 2018;21(11):1047‐1056. [DOI] [PubMed] [Google Scholar]
5. Jensen GL, Cederholm T, Correia MITD, et al. GLIM criteria for the diagnosis of malnutrition: a consensus report from the global clinical nutrition community. J Parenter Enteral Nutr. Jan 2019;43(1):32‐40. [DOI] [PubMed] [Google Scholar]
6. Yin L, Chong F, Huo Z, Li N, Liu J, Xu H. GLIM‐defined malnutrition and overall survival in cancer patients: a meta‐analysis. J Parenter Enteral Nutr. 2023;47(2):207‐219. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Compher CW, Fukushima R, Correia MITD, et al. Recognizing malnutrition in adults with critical illness: guidance statements from the Global Leadership Initiative on Malnutrition. J Parenter Enteral Nutr. 2025;49(4):405‐413. [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Lee JS, Choi HS, Ko YG, Yun DH. Performance of the Geriatric Nutritional Risk Index in predicting 28‐day hospital mortality in older adult patients with sepsis. Clin Nutr. 2013;32(5):843‐848. [DOI] [PubMed] [Google Scholar]
9. Liang L, Moore B, Soni A. National inpatient hospital costs: the most expensive conditions by payer, 2017. In: Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Statistical Brief #261. Agency for Healthcare Research and Quality; 2020. Accessed June 1, 2026. https://www.ncbi.nlm.nih.gov/books/NBK561141/. [PubMed]
10. Moerer O, Plock E, Mgbor U, et al. A German national prevalence study on the cost of intensive care: an evaluation from 51 intensive care units. Crit Care. 2007;11(3):R69. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Dugani S, Veillard J, Kissoon N. Reducing the global burden of sepsis. Can Med Assoc J. 2017;189(1):E2‐E3. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Liu K, Kotani T, Nakamura K, et al. Effects of evidence‐based ICU care on long‐term outcomes of patients with sepsis or septic shock (ILOSS): protocol for a multicentre prospective observational cohort study in Japan. BMJ Open. 2022;12(3):e054478. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Liu K, Watanabe S, Nakamura K, et al. One‐year outcomes in sepsis: A prospective multicenter cohort study in Japan. J Intensive Care. 2025;13(1):23. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Wynants L, Kent DM, Timmerman D, Lundquist CM, Van Calster B. Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting. Diagn Progn Res. 2019;3:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (strobe) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806‐808. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Singer M, Deutschman CS, Seymour CW, et al. The third international consensus definitions for sepsis and septic shock (sepsis‐3). JAMA. 2016;315(8):801‐810. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Ogola GO, Shafi S. Cost of specific emergency general surgery diseases and factors associated with high‐cost patients. J Trauma Acute Care Surg. 2016;80(2):265‐271. [DOI] [PubMed] [Google Scholar]
18. Shiroiwa T, Fukuda T, Ikeda S, et al. Japanese population norms for preference‐based measures: EQ‐5D‐3L, EQ‐5D‐5L, and SF‐6D. Qual Life Res. 2016;25(3):707‐719. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care. 2005;43(11):1130‐1139. [DOI] [PubMed] [Google Scholar]
20. Suganuma S, Nakamura K, Kanda N, et al. Clinical utility of measuring temporal muscle volume by head computed tomography for Global Leadership Initiative on Malnutrition phenotypic criteria in critically ill patients. Clin Nutr. 2025;53:133‐143. [DOI] [PubMed] [Google Scholar]
21. McClave SA, Taylor BE, Martindale RG, et al. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). J Parenter Enteral Nutr. 2016;40(2):159‐211. [DOI] [PubMed] [Google Scholar]
22. Jensen GL, Cederholm T, Ballesteros‐Pomar MD, et al. Guidance for assessment of the inflammation etiologic criterion for the GLIM diagnosis of malnutrition: a modified Delphi approach. J Parenter Enteral Nutr. 2024;48(2):145‐154. [DOI] [PubMed] [Google Scholar]
23. Stekhoven DJ, Bühlmann P. MissForest‐‐non‐parametric missing value imputation for mixed‐type data. Bioinformatics. 2012;28(1):112‐118. [DOI] [PubMed] [Google Scholar]
24. Veazie P, Intrator O, Kinosian B, Phibbs CS. Better performance for right‐skewed data using an alternative gamma model. BMC Med Res Methodol. 2023;23(1):298. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Ribbink ME, van Seben R, Reichardt LA, et al. Determinants of post‐acute care costs in acutely hospitalized older adults: the hospital‐ADL study. J Am Med Dir Assoc. 2019;20(10):1300‐1306.e1. [DOI] [PubMed] [Google Scholar]
26. Barrett KA, Sheikh F, Chechulina V, et al. High‐cost users after sepsis: a population‐based observational cohort study. Crit Care. 2024;28(1):338. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Kobayashi H, Tonai M, Karumai T, Shiraishi A, Fushimi K, Hayashi Y. Age‐related epidemiology and outcomes of sepsis in Japanese critical care units: a nationwide administrative claims database study. J Intensive Care. 2025;13(1):66. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Soley‐Bori M, Ashworth M, Bisquera A, et al. Impact of multimorbidity on healthcare costs and utilisation: a systematic review of the UK literature. Br J Gen Pract. 2021;71(702):e39‐e46. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Ohbe H, Sasabuchi Y, Doi K, Matsui H, Yasunaga H. Association between levels of intensive care and in‐hospital mortality in patients hospitalized for sepsis stratified by sequential organ failure assessment scores. Crit Care Med. 2023;51(9):1138‐1147. [DOI] [PubMed] [Google Scholar]
30. Yoshida M, Yamamoto R, Izumino H, et al. Nutrition therapy for physical impairment in critically ill patients: a narrative review of the latest evidence. Clin Nutr. 2025;52:134‐153. [DOI] [PubMed] [Google Scholar]
31. Desai RJ, Franklin JM. Alternative approaches for confounding adjustment in observational studies using weighting based on the propensity score: a primer for practitioners. BMJ (Clin Res ed.). 2019;367:l5657. [DOI] [PubMed] [Google Scholar]
32. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228‐1234. [Google Scholar]
33. Evans L, Rhodes A, Alhazzani W, et al. Surviving sepsis campaign: International guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181‐1247. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Wang W, Dai L, Ma J, Gu L, Xie H, Fu J. Malnutrition accelerates the occurrence of infectious complications in patients with chronic kidney disease: a cross‐sectional survey of 682 patients with chronic kidney disease. Nutr Clin Pract. 2023;38(5):1167‐1174. [DOI] [PubMed] [Google Scholar]
35. Inciong JFB, Chaudhary A, Hsu HS, et al. Economic burden of hospital malnutrition: a cost‐of‐illness model. Clin Nutr ESPEN. 2022;48:342‐350. [DOI] [PubMed] [Google Scholar]
36. Assimakopoulos SF, Triantos C, Thomopoulos K, et al. Gut‐origin sepsis in the critically ill patient: pathophysiology and treatment. Infection. 2018;46(6):751‐760. [DOI] [PubMed] [Google Scholar]
37. Saponaro F, Bertolini A, Baragatti R, et al. Myokines and microbiota: new perspectives in the endocrine muscle–gut axis. Nutrients. 2024;16(23):4032. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Kahn JM, Le T, Angus DC, et al. The epidemiology of chronic critical illness in the United States*. Crit Care Med. 2015;43(2):282‐287. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Yoshida M, Yamamoto R, Sagawa M, et al. Optimal route of supplementary protein delivery for physical recovery in critically ill patients: protocol for a multicenter randomized clinical trial. Clin Nutr ESPEN. 2026;73:103301. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental appendix 1.

JPEN-50-985-s002.docx^{(198.9KB, docx)}

Supplemental appendix 2.

JPEN-50-985-s001.docx^{(19.5KB, docx)}

[jpen70120-bib-0001] 1. Norman K, Pichard C, Lochs H, Pirlich M. Prognostic impact of disease‐related malnutrition. Clin Nutr. 2008;27(1):5‐15. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0002] 2. Guenter P, Abdelhadi R, Anthony P, et al. Malnutrition diagnoses and associated outcomes in hospitalized patients: United States, 2018. Nutr Clin Pract. 2021;36(5):957‐969. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0003] 3. Álvarez‐Hernández J, Planas Vila M, León‐Sanz M, et al. Prevalence and costs of malnutrition in hospitalized patients; the PREDyCES® study. Nutr Hosp. 2012;27(4):1049‐1059. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0004] 4. Correia MITD, Perman MI, Pradelli L, Omaralsaleh AJ, Waitzberg DL. Economic burden of hospital malnutrition and the cost‐benefit of supplemental parenteral nutrition in critically ill patients in Latin America. J Med Econ. 2018;21(11):1047‐1056. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0005] 5. Jensen GL, Cederholm T, Correia MITD, et al. GLIM criteria for the diagnosis of malnutrition: a consensus report from the global clinical nutrition community. J Parenter Enteral Nutr. Jan 2019;43(1):32‐40. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0006] 6. Yin L, Chong F, Huo Z, Li N, Liu J, Xu H. GLIM‐defined malnutrition and overall survival in cancer patients: a meta‐analysis. J Parenter Enteral Nutr. 2023;47(2):207‐219. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0007] 7. Compher CW, Fukushima R, Correia MITD, et al. Recognizing malnutrition in adults with critical illness: guidance statements from the Global Leadership Initiative on Malnutrition. J Parenter Enteral Nutr. 2025;49(4):405‐413. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0008] 8. Lee JS, Choi HS, Ko YG, Yun DH. Performance of the Geriatric Nutritional Risk Index in predicting 28‐day hospital mortality in older adult patients with sepsis. Clin Nutr. 2013;32(5):843‐848. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0009] 9. Liang L, Moore B, Soni A. National inpatient hospital costs: the most expensive conditions by payer, 2017. In: Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Statistical Brief #261. Agency for Healthcare Research and Quality; 2020. Accessed June 1, 2026. https://www.ncbi.nlm.nih.gov/books/NBK561141/. [PubMed]

[jpen70120-bib-0010] 10. Moerer O, Plock E, Mgbor U, et al. A German national prevalence study on the cost of intensive care: an evaluation from 51 intensive care units. Crit Care. 2007;11(3):R69. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0011] 11. Dugani S, Veillard J, Kissoon N. Reducing the global burden of sepsis. Can Med Assoc J. 2017;189(1):E2‐E3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0012] 12. Liu K, Kotani T, Nakamura K, et al. Effects of evidence‐based ICU care on long‐term outcomes of patients with sepsis or septic shock (ILOSS): protocol for a multicentre prospective observational cohort study in Japan. BMJ Open. 2022;12(3):e054478. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0013] 13. Liu K, Watanabe S, Nakamura K, et al. One‐year outcomes in sepsis: A prospective multicenter cohort study in Japan. J Intensive Care. 2025;13(1):23. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0014] 14. Wynants L, Kent DM, Timmerman D, Lundquist CM, Van Calster B. Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting. Diagn Progn Res. 2019;3:6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0015] 15. Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (strobe) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806‐808. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0016] 16. Singer M, Deutschman CS, Seymour CW, et al. The third international consensus definitions for sepsis and septic shock (sepsis‐3). JAMA. 2016;315(8):801‐810. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0017] 17. Ogola GO, Shafi S. Cost of specific emergency general surgery diseases and factors associated with high‐cost patients. J Trauma Acute Care Surg. 2016;80(2):265‐271. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0018] 18. Shiroiwa T, Fukuda T, Ikeda S, et al. Japanese population norms for preference‐based measures: EQ‐5D‐3L, EQ‐5D‐5L, and SF‐6D. Qual Life Res. 2016;25(3):707‐719. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0019] 19. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care. 2005;43(11):1130‐1139. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0020] 20. Suganuma S, Nakamura K, Kanda N, et al. Clinical utility of measuring temporal muscle volume by head computed tomography for Global Leadership Initiative on Malnutrition phenotypic criteria in critically ill patients. Clin Nutr. 2025;53:133‐143. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0021] 21. McClave SA, Taylor BE, Martindale RG, et al. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). J Parenter Enteral Nutr. 2016;40(2):159‐211. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0022] 22. Jensen GL, Cederholm T, Ballesteros‐Pomar MD, et al. Guidance for assessment of the inflammation etiologic criterion for the GLIM diagnosis of malnutrition: a modified Delphi approach. J Parenter Enteral Nutr. 2024;48(2):145‐154. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0023] 23. Stekhoven DJ, Bühlmann P. MissForest‐‐non‐parametric missing value imputation for mixed‐type data. Bioinformatics. 2012;28(1):112‐118. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0024] 24. Veazie P, Intrator O, Kinosian B, Phibbs CS. Better performance for right‐skewed data using an alternative gamma model. BMC Med Res Methodol. 2023;23(1):298. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0025] 25. Ribbink ME, van Seben R, Reichardt LA, et al. Determinants of post‐acute care costs in acutely hospitalized older adults: the hospital‐ADL study. J Am Med Dir Assoc. 2019;20(10):1300‐1306.e1. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0026] 26. Barrett KA, Sheikh F, Chechulina V, et al. High‐cost users after sepsis: a population‐based observational cohort study. Crit Care. 2024;28(1):338. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0027] 27. Kobayashi H, Tonai M, Karumai T, Shiraishi A, Fushimi K, Hayashi Y. Age‐related epidemiology and outcomes of sepsis in Japanese critical care units: a nationwide administrative claims database study. J Intensive Care. 2025;13(1):66. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0028] 28. Soley‐Bori M, Ashworth M, Bisquera A, et al. Impact of multimorbidity on healthcare costs and utilisation: a systematic review of the UK literature. Br J Gen Pract. 2021;71(702):e39‐e46. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0029] 29. Ohbe H, Sasabuchi Y, Doi K, Matsui H, Yasunaga H. Association between levels of intensive care and in‐hospital mortality in patients hospitalized for sepsis stratified by sequential organ failure assessment scores. Crit Care Med. 2023;51(9):1138‐1147. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0030] 30. Yoshida M, Yamamoto R, Izumino H, et al. Nutrition therapy for physical impairment in critically ill patients: a narrative review of the latest evidence. Clin Nutr. 2025;52:134‐153. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0031] 31. Desai RJ, Franklin JM. Alternative approaches for confounding adjustment in observational studies using weighting based on the propensity score: a primer for practitioners. BMJ (Clin Res ed.). 2019;367:l5657. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0032] 32. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228‐1234. [Google Scholar]

[jpen70120-bib-0033] 33. Evans L, Rhodes A, Alhazzani W, et al. Surviving sepsis campaign: International guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181‐1247. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0034] 34. Wang W, Dai L, Ma J, Gu L, Xie H, Fu J. Malnutrition accelerates the occurrence of infectious complications in patients with chronic kidney disease: a cross‐sectional survey of 682 patients with chronic kidney disease. Nutr Clin Pract. 2023;38(5):1167‐1174. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0035] 35. Inciong JFB, Chaudhary A, Hsu HS, et al. Economic burden of hospital malnutrition: a cost‐of‐illness model. Clin Nutr ESPEN. 2022;48:342‐350. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0036] 36. Assimakopoulos SF, Triantos C, Thomopoulos K, et al. Gut‐origin sepsis in the critically ill patient: pathophysiology and treatment. Infection. 2018;46(6):751‐760. [DOI] [PubMed] [Google Scholar]

[jpen70120-bib-0037] 37. Saponaro F, Bertolini A, Baragatti R, et al. Myokines and microbiota: new perspectives in the endocrine muscle–gut axis. Nutrients. 2024;16(23):4032. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0038] 38. Kahn JM, Le T, Angus DC, et al. The epidemiology of chronic critical illness in the United States*. Crit Care Med. 2015;43(2):282‐287. [DOI] [PMC free article] [PubMed] [Google Scholar]

[jpen70120-bib-0039] 39. Yoshida M, Yamamoto R, Sagawa M, et al. Optimal route of supplementary protein delivery for physical recovery in critically ill patients: protocol for a multicenter randomized clinical trial. Clin Nutr ESPEN. 2026;73:103301. [DOI] [PubMed] [Google Scholar]

PERMALINK

Association between GLIM‐defined malnutrition and in‐hospital costs in adults with sepsis: A sub analysis of a prospective cohort study

Yuki Hashimoto, RPh, PhD

Minoru Yoshida, MD, PhD

Tomoka Miyagi, RD

Akane Kariya, RN, MPH

Hitomi Tanoue, RD

Naoki Kanda, MD, MPH, PhD

Yusuke Hoshi, RD

Takahiro Fujita, PT

Kensei Togo, MD

Tomoaki Obata, MD

Kenzo Ishii, MD, PhD

Katsuyuki Abe, RD

Chihiro Sato, RD

Yoshifumi Heshiki, RN

Daisetsu Yasumura, PT, MHA

Shizuko Gomi, PhD

Masafumi Takahashi, PT

Takashi Hongo, MD, PhD

Keiji Sakurai, MD

Miho Nishihara, RD

Yuko Matsuda, RD

Kiyota Chinen, RD

Akiko Arai, MD

Michihiro Maeda, MD

Yuhei Irie, MD, PhD

Hiroshi Kamijo, MD, PhD

Kana Hosokawa, RD

Daisuke Kasugai, MD, PhD

Michiko Higashi, MD

Ayaka Matsuoka, MD

Hirotaka Ogawa, MD

Yuichi Araki, MD

Kenji Oike, PT

Daigo Shimakura, PT

Hajime Katsukawa, PT, PhD

Toru Kotani, MD

Takayuki Ogura, MD, PhD

Keibun Liu, MD, PhD

Kensuke Nakamura, MD, PhD

Abstract

Background

Methods

Results

Conclusions

INTRODUCTION

METHODS

Study design and ethics

Participants

Outcome variables

Covariates

Definition of malnutrition

Statistical analysis

Figure 2.

RESULTS

Study population

Figure 1.

Primary analysis

Table 1.

Table 2.

Secondary analysis

Table 3.

DISCUSSION

Limitations

CONCLUSIONS

AUTHOR CONTRIBUTIONS

CONFLICT OF INTEREST STATEMENT

Supporting information

ACKNOWLEDGEMENTS

REFERENCES

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases