Association between energy adequacy and clinical outcomes in septic intensive care unit patients stratified by modified Nutrition Risk in the Critically Ill score Authors: Pi-Yi Lin, Yu-Tong Yen, Ming-Jen Lu, Han-Shui Hsu Published: 2026 ABSTRACT Background:Although malnutrition is common among critically ill patients with sepsis, it is unclear whether early nutritional supplementation is beneficial. We investigated the effects of early calorie and protein delivery in critically ill patients with sepsis.Methods:Adult patients with sepsis who required mechanical ventilation in a surgical intensive care unit (ICU) between January 2021 and December 2022 were enrolled. Nutritional risk in these patients was evaluated using the modified Nutrition Risk in the Critically ill (m-NUTRIC) score. Participants were divided into two groups based on whether they received <50% or ≥50% of their estimated calorie needs within the first 5 days of their ICU stay. The primary outcome was 28-day ICU mortality.Results:Among the 258 participants (195 men and 63 women; mean age, 64.8 ± 13.7 years), within the first 5 days of their ICU stay, 63.2% and 36.8% of the patients received <50% and ≥50% of their estimated calorie needs, respectively. The 28-day ICU mortality rate did not differ significantly between the two groups. Multivariate logistic regression revealed that the use of continuous renal replacement therapy (odds ratio [OR] = 4.07, 95% CI, 1.10-15.09; p = 0.036) and the duration of mechanical ventilation (OR = 1.05, 95% CI, 1.01-1.10; p = 0.020) were independent predictors of 28-day ICU mortality. Providing early nutritional support of ≤50% of the estimated calorie needs did not affect 28-day ICU mortality in these patients (p = 0.229).Conclusion:Achieving at least 50% of the estimated calorie needs within the first 5 days of ICU stay did not improve 28-day ICU survival in this cohort of critically ill patients with sepsis. Further investigations are needed to elucidate whether early nutritional interventions are beneficial for this population. 1. INTRODUCTION Malnutrition affects 30% to 69% of critically ill patients and is independently associated with higher mortality.1 Sepsis, a systemic inflammatory response to pathogens such as bacteria, fungi, or viruses, remains a leading cause of intensive care unit (ICU) mortality, which reaches 40.4% in cases of septic shock. The clinical course of sepsis is largely determined by the severity of inflammation, nutritional reserves, and immune competence.2,3 Among critically ill patients, complications such as sepsis and acute kidney injury are common and can exacerbate metabolic disturbances, which further increase the risk of malnutrition.4,5 Heyland et al6 developed the Nutrition Risk in the Critically ill (NUTRIC) score, which is the first risk-assessment tool specifically designed for ICU patients to identify those most likely to benefit from targeted nutritional support. Rahman et al7 introduced a modified NUTRIC (m-NUTRIC) score, which excludes interleukin-6 (IL-6) because of its limited clinical availability. The score includes age, Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment (SOFA), comorbidities, and the number of days from hospitalization to ICU transfer. The m-NUTRIC score ranges from 0 to 9, with scores ≥5 indicating high nutritional risk.7 The m-NUTRIC score is widely used to assess nutritional risk in ICU patients and is strongly associated with higher mortality among those identified as high risk.5 Patients with sepsis are particularly prone to malnutrition, making early nutritional risk assessment crucial for guiding and optimizing nutritional support. Recent evidence has indicated that high-dose nutritional support during the early phase of sepsis does not significantly improve outcomes. This may reflect anabolic resistance, impaired cellular repair, stress hyperglycemia, and increased insulin demand.8 Although insufficient protein and energy intake are associated with poor prognosis, early overfeeding may induce excessive metabolic stress. As metabolic derangements and severe inflammation alter nutritional needs in patients with sepsis, managing caloric intake remains challenging. Early near-target energy delivery does not reduce mortality or the duration of ventilation, which indicates that the optimal energy balance remains uncertain.9 Therefore, in this study, we evaluated the role of the m-NUTRIC score in assessing nutritional risk and examined the impact of adequate calorie intake during the first 5 days of stay on clinical outcomes in ICU patients with sepsis. 2. METHODS This retrospective observational study was conducted in the surgical ICU of a single medical center. Patient data were retrieved from electronic medical records between January 2021 and December 2022. This study was approved by the institutional review board (IRB No. 20230201R). Eligible participants were adults (aged >18 years) with sepsis who were admitted to the ICU and received mechanical ventilation (MV) for >24 hours. An m-NUTRIC score assessment was required within the first 24 hours of ICU admission. Patients were excluded if they were pregnant, readmitted to the ICU during the same hospitalization period, or had incomplete data. 2.1. Nutritional assessment Energy requirements were estimated by a clinical dietitian using the patient’s body weight on admission and a standard predictive equation of 25 kcal/kg/d. Day-5 cumulative adequate calorie intake was defined as achieving at least 50% of the estimated calorie requirements during the first 5 days of ICU stay. Based on the population distribution and clinical feasibility, patients were stratified according to day-5 cumulative adequate calorie intake (≥50% vs <50%) (Supplementary Figure 1, https://links.lww.com/JCMA/A416). In addition to the m-NUTRIC score, additional nutritional status indicators were assessed, including the Prognostic Nutritional Index (PNI) and Controlling Nutritional Status (CONUT) score. The PNI was calculated using the serum albumin concentration and total lymphocyte count, and patients were categorized using a cutoff value of 40. The CONUT score was derived from serum albumin, total cholesterol, and lymphocyte count and was classified into absent to mild (0-4), moderate (5-8), and severe (9-12) nutritional risk categories according to established criteria.10–12 2.2. Data collection Clinicodemographic data, including sex, age, height, weight, body mass index (BMI), and APACHE II, SOFA, and m-NUTRIC scores, were collected. Laboratory parameters included hematologic and biochemical profiles, C-reactive protein (CRP), and serum albumin levels. Clinical complications during ICU stay, such as shock and acute kidney injury, as well as treatments including vasopressor use and continuous renal replacement therapy (CRRT), were recorded. Additional data included the ICU and hospital length of stay, MV duration, and 28-day ICU mortality. Nutritional intake data from the first 5 days of ICU admission were extracted from the feeding records. These encompassed feeding routes, including nothing per os (NPO), enteral nutrition (EN), parenteral nutrition (PN), combined EN and PN, and caloric contributions from medications such as propofol and dextrose-containing solutions. 2.3. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics version 27. A p value <0.05 was considered statistically significant. The Kolmogorov-Smirnov test was used to assess the normality of continuous variables. Data are presented as mean ± SD or median with interquartile range (IQR), as appropriate. Group comparisons for continuous variables were performed using the independent t test or Mann-Whitney U test, depending on the data distribution, whereas categorical variables are presented as percentages and analyzed using the chi-square or Fisher’s exact test. Spearman’s rank correlation was used to evaluate the associations between the variables. Univariate logistic regression was performed, and variables with p < 0.10 or clinical relevance were included in the multivariate logistic regression. The results are presented as adjusted odds ratios (ORs) with 95% CIs. Kaplan-Meier survival analysis was used to estimate survival curves, and group differences were compared using the log-rank test. Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the predictive performance, and the area under the curve (AUC) was calculated. The optimal cutoff value was determined using the Youden index. 3. RESULTS Upon screening, the records of 1267 patients admitted to the ICU were initially screened, and 1009 patients with incomplete m-NUTRIC data, no MV, repeated ICU admissions, or missing dietary records were excluded. The data of the remaining 258 patients with sepsis who were placed on MV were included in the final analysis. The participants were grouped based on their m-NUTRIC scores: m-NUTRIC <5 (n = 107) and m-NUTRIC ≥5 (n = 151) (Fig. 1). Flowchart of patient selection. In total, 1267 patients admitted to the SICU were screened. After excluding those without complete m-NUTRIC data, without mechanical ventilation, with repeated ICU admissions, or with incomplete dietary records, 258 patients with sepsis were included in the final analysis. Patients were stratified into two groups based on m-NUTRIC score: low-risk (<5, n = 107) and high-risk (≥5, n = 151). ICU, intensive care unit; m-NUTRIC, modified Nutrition Risk in the Critically Ill; SICU, surgical intensive care unit. Table 1 summarizes the baseline clinical demographic characteristics of the two groups. The mean age of the cohort was 64.8 ± 13.7 years, 75.6% (n = 195) were male, and the median BMI was 24.4 (21.8-27.1) kg/m2. The high-risk group had significantly higher rates of hypertension (68.9% vs 45.8%, p < 0.001), type 2 diabetes mellitus (47.7% vs 30.8%, p = 0.007), chronic kidney disease (35.1% vs 2.8%, p < 0.001), and liver cirrhosis (4.0% vs 0%, p = 0.037) than the low-risk group. However, the actual energy intake did not differ significantly between the groups (557.5 [368.3-880.8] vs 656.2 [396.6-1002.8] kcal/day; p = 0.096). Energy intake per kilogram of body weight was significantly higher for the high-risk group (7.9 [4.7-12.9] vs 9.9 [5.9-16.8] kcal/kg/d, p = 0.030); however, the protein intake per kilogram was comparable between the groups (1.0 [0.86-1.17] vs 1.0 [0.8-1.2] g/kg/d, p = 0.523). Table 1Baseline demographic and clinical characteristics stratified by m-NUTRIC score (<5 vs ≥5) in critically ill patients with sepsis (n = 258)VariableAll patients (n = 258)m-NUTRIC <5 (n = 107)m-NUTRIC≥ 5 (n = 151) p Age, y (SD)64.8 ± 13.756.8 ± 12.770.5 ± 11.5<0.001Sex (male), %195 (75.6)85 (79.4)110 (72.8)0.225Type of patient, %0.022 Medical25 (9.7)5 (4.7)20 (13.2) Surgical233 (90.3)102 (95.3)131 (86.8) Trauma12 (4.7)6 (5.6)6 (4.0)0.539Body weight, kg65.3 [57-75]67.5 [59.2-78]64.2 [56.2-72]0.020BMI, kg/m224.4 [21.8-27.1]24.3 [21.8-27.3]24.4 [21.8-26.9]0.448Comorbidities, % Cardiac disease147 (57.0)55 (51.4)92 (60.9)0.128 Hypertension153 (59.3)49 (45.8)104 (68.9)<0.001 Diabetes mellitus type 2105 (40.7)33 (30.8)72 (47.7)0.007 Chronic kidney disease56 (21.7)3 (2.8)53 (35.1)<0.001 Liver cirrhosis6 (2.3)0 (0.0)6 (4.0)0.037 Cancer56 (21.7)22 (20.6)34 (22.5)0.707Energy intake, kcal/d613.3 [378.4-967.9]557.5 [368.3-880.8]656.2 [396.6-1002.8]0.096Average energy delivered, kcal/kg/d9.3 [5.6-15.6]7.9 [4.7-12.9]9.9 [5.9-16.8]0.030Predicted protein delivered, g/d65 [57.4-72.2]65.9 [60-72]63.6 [55.5-72.6]0.055Protein delivered, g/kg/d1.0 [0.85-1.2]1.0 [0.86-1.17]1.0 [0.8-1.2]0.523Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant.BMI = body mass index; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill. Baseline demographic and clinical characteristics stratified by m-NUTRIC score (<5 vs ≥5) in critically ill patients with sepsis (n = 258) Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant. BMI = body mass index; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill. Table 2 summarizes the clinical characteristics of the patients stratified by their day-5 cumulative adequate calorie intake (Fig. 1). Patients were categorized into the <50% (n = 163) and ≥50% (n = 95) groups. Table 2Patient characteristics stratified by day-5 cumulative adequate calorie intake (<50% vs ≥50%) (n = 258)VariableDay-5 cumulative adequate calorie intake <50% (n = 163)Day-5 cumulative adequate calorie intake ≥50% (n = 95) p Age, y (SD)63.0 ± 12.867.8 ± 14.90.007Gender (male), %132 (81.0)63 (66.3)0.008Body weight, kg67.7 [60-77]61 [52-70]<0.001BMI, kg/m224.8 [22-27.9]23.4 [20.8-26.3]0.003CRP, mg/dL6.4 [4-11.6]7.4 [2.8-13.7]0.829Albumin, g/dL3.6 [3-4.9]2.8 [2.4-3.3]<0.001ICU complication, % S/P acute kidney injury49 (30.1)29 (30.5)0.937 Vasopressors106 (65.0)50 (52.6)0.049 CRRT9 (5.5)6 (6.3)0.793Length of ICU stay, d3 [2-5]5 [2-8]<0.001Total hospital days18 [13-27]29 [16-53]<0.001Duration of MV1 [1-3]4 [1-11]<0.001SOFA7 [5-9]7 [5-9]0.602APACHE II17 [14-21]20 [16-25]<0.001m-NUTRIC score5 [4-6]5 [4-6]0.065ICU 28-d mortality9 (5.5)9 (9.5)0.229PNI score36 [30-41]28 [24-33]<0.001PNI <40, n (%)116 (71.2)87 (91.6)<0.001PNI ≥40, n (%)47 (28.8)8 (8.4)CONUT score5 [4-7]8 [6-10]<0.001CONUT category, % Absent to mild (0-4)51 (31.3)9 (9.5)<0.001 Moderate (5-8)90 (55.2)40 (42.1) Severe (9-12)22 (13.5)46 (48.4)Nutrition supply, % PN8 (4.9)33 (34.7)<0.001 EN142 (87.1)52 (54.7)<0.001 NPO14 (8.6)0 (0)0.003 EN+PN1 (0.6)10 (10.5)0.001Energy intake, kcal/d429 [308-599.6]1042 [914-1254]<0.001Average energy delivered, kcal/kg/d6.3 [4.1-8.9]17.3 [14.3-21.6]<0.001Protein delivered, g/kg/d1 [0.8-1.1]1 [0.9-1.2]0.001Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant.APACHE-II = Acute Physiology and Chronic Health Evaluation II; BMI = body mass index; CONUT= Controlling Nutritional Status score; CRP = C-reactive protein; CRRT = continuous renal replacement therapy; EN = enteral nutrition; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; NPO = nothing per os; PN = parenteral nutrition; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. Patient characteristics stratified by day-5 cumulative adequate calorie intake (<50% vs ≥50%) (n = 258) Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant. APACHE-II = Acute Physiology and Chronic Health Evaluation II; BMI = body mass index; CONUT= Controlling Nutritional Status score; CRP = C-reactive protein; CRRT = continuous renal replacement therapy; EN = enteral nutrition; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; NPO = nothing per os; PN = parenteral nutrition; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. Significant between-group differences in demographic characteristics, including age, sex distribution, body weight, and BMI, were observed. The patients in the <50% group had a higher BMI than those in the ≥50% group (24.8 [22.0-27.9] vs 23.4 [20.8-26.3] kg/m2; p = 0.003). Participants with day-5 cumulative adequate calorie intake ≥50% had significantly lower serum albumin concentrations (2.8 [2.4-3.3] vs 3.6 [3.0-4.9] g/dL; p < 0.001), lower PNI scores, and a higher prevalence of PNI <40 (91.6% vs 71.2%; p < 0.001). The CONUT scores were significantly higher for the ≥50% group, with a greater proportion of patients classified as having severe nutritional risk (9-12 points; 48.4% vs 13.5%; p < 0.001). The APACHE II score was significantly higher in the ≥50% group than in the <50% group (20 [16-25] vs 17 [14-21], p < 0.001). However, SOFA scores did not differ significantly between the groups (p = 0.602). The 28-day ICU mortality rates were similar between the two groups (9.5% vs 5.5%, p = 0.229). However, the patients in the ≥50% group had longer ICU stays (5 [2-8] vs 3 [2-5] days, p < 0.001), longer total hospital stays (29 [16-53] vs 18 [13-27] days, p < 0.001), and longer durations of MV (4 [1-11] vs 1 [1-3] days, p < 0.001). EN was more frequently used for the <50% group, whereas PN and combined enteral-PN were more common for the ≥50% group (87.1% vs 54.7%, p < 0.001). The ≥50% group had a higher actual daily energy intake (1042 [914-1254] vs 429 [308-599.6] kcal/day, p < 0.001), average energy intake per kilogram (17.3 [14.3-21.6] vs 6.3 [4.1-8.9] kcal/kg/day, p < 0.001), and protein intake (1.0 [0.9-1.2] vs 1.0 [0.8-1.1] g/kg/day, p = 0.001). Table 3 compares the clinical characteristics, nutritional parameters, and disease severity of survivors and non-survivors. Age, sex, body weight, and BMI did not differ significantly between the groups. The laboratory values, including those for CRP and serum albumin levels, were also comparable. More non-survivors required vasopressors (83.3% vs 58.8%, p = 0.040) and CRRT (27.8% vs 4.2%, p < 0.001) and had longer ICU stays (8.5 [5.8-14.0] vs 3 [2-6] days, p < 0.001) and MV duration (12 [6-23.3] vs 1 [1-4] days; p < 0.001) than survivors. Acute kidney injury was more common in non-survivors; however, the difference was not statistically significant. The median m-NUTRIC (6 [5-8] vs 5 [4-6], p = 0.004), SOFA (9.5 [8-11] vs 7 [5-9], p < 0.001), and APACHE II (21 [17-26.5] vs 18 [14-23], p = 0.023) scores of non-survivors were significantly higher. No significant differences in PNI scores, proportion of patients with PNI <40, CONUT scores, or CONUT risk categories were observed between survivors and non-survivors. No significant differences were observed between the groups in day-5 cumulative adequate calorie intake (p = 0.229), actual daily energy intake (p = 0.344), or protein intake (p = 0.712). Table 3Comparison of survival and non-survival with adequate calorie intake and clinical outcomes (n = 258)VariableSurvival (n = 240)Non-survival (n = 18) p Age, y (SD)64.6 ± 13.767.8 ± 14.90.333Gender (male), %184 (76.7)11 (61.1)0.138Body weight, kg65.6 [56.9-75]61 [59.2-68.6]0.398BMI, kg/m224.3 [21.8-27.1]24.5 [22.8-27.2]0.681CRP, mg/dL6.4 [3.8-11.7]4.7 [0.8-14.4]0.249Albumin, g/dL3.2 [2.8-3.9]3.2 [2.6-3.7]0.421ICU complication, % S/P acute kidney injury71 (29.6)7 (38.9)0.407 Vasopressors141 (58.8)15 (83.3)0.040 CRRT10 (4.2)5 (27.8)<0.001Length of ICU stay, d3 [2-6]8.5 [5.8-14.0]<0.001Total hospital days20 [14-32]17 [9.8-32.8]0.288Duration of MV1 [1-4]12 [6-23.3]<0.001m-NUTRIC score5 [4-6]6 [5-8]0.004SOFA7 [5-9]9.5 [8-11.3]<0.001APACHE II18 [14-23]21 [17-26.5]0.023PNI score32 [28-39]32 [26-37]0.430PNI <40, n (%)187 (77.9)16 (88.9)0.273PNI ≥40, n (%)53 (22.1)2 (11.1)CONUT score6 [5-9]7 [6-9]0.216CONUT category, n (%) Absent to mild (0-4)57 (23.8)3 (16.7)0.698 Moderate (5-8)121 (50.4)9 (50.0) Severe (9-12)62 (25.8)6 (33.3)Energy intake, kcal/d606.5 [378-955.2]843.4 [453.5-1019.4]0.344Average energy delivered, kcal/kg/d8.9 [5.6-15.5]13.8 [5.4-17.3]0.301Day-5 cumulative adequate calorie intake <50%, n (%)154 (64.2)9 (50.0)0.229Day-5 cumulative adequate calorie intake ≥50%, n (%)86 (35.8)9 (50.0)Protein delivered, g/kg/d1.0 [0.9-1.2]1.0 [0.8-1.2]0.712Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant.APACHE-II = Acute Physiology and Chronic Health Evaluation II; BMI = body mass index; CONUT = Controlling Nutritional Status score; CRP = C-reactive protein; CRRT = continuous renal replacement therapy; EN = enteral nutrition; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; NPO = nothing per os; PN = parenteral nutrition; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. Comparison of survival and non-survival with adequate calorie intake and clinical outcomes (n = 258) Data are presented as median, IQR, mean ± SD, or percentages. p < 0.05 was significant. APACHE-II = Acute Physiology and Chronic Health Evaluation II; BMI = body mass index; CONUT = Controlling Nutritional Status score; CRP = C-reactive protein; CRRT = continuous renal replacement therapy; EN = enteral nutrition; IQR = interquartile range; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; NPO = nothing per os; PN = parenteral nutrition; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. Table 4 presents the univariate and multivariate logistic regression analyses for 28-day ICU mortality of the overall cohort and the m-NUTRIC risk groups (low and high). The following clinical variables were significantly associated with ICU mortality in the overall cohort: CRRT use (OR = 8.85; 95% CI, 2.64-29.67; p < 0.001), MV duration (OR = 1.08; 95% CI, 1.03-1.13; p < 0.001), m-NUTRIC score (OR = 1.67; 95% CI, 1.20-2.33; p = 0.003), ICU length of stay (OR = 1.05; 95% CI, 1.00-1.09; p = 0.037), APACHE II score (OR = 1.10; 95% CI, 1.02-1.18; p = 0.012), and SOFA score (OR = 1.44; 95% CI, 1.19–1.74; p < 0.001). However, day-5 cumulative adequate calorie intake of ≥50% was not significantly associated with 28-day ICU mortality (OR = 1.79; 95% CI, 0.69-4.68; p = 0.235). The m-NUTRIC score was the only disease severity control variable included in the multivariate model to ensure coefficient stability and model robustness, given the potential collinearity among the ICU length of stay and SOFA, APACHE II, and m-NUTRIC scores. The multivariate logistic regression identified CRRT (OR = 4.07; 95% CI, 1.10-15.09; p = 0.036) and MV duration (OR = 1.05; 95% CI, 1.01-1.10; p = 0.020) as independent predictors of 28-day ICU mortality, whereas the m-NUTRIC score was not significantly associated with mortality after adjustment (OR = 1.36; 95% CI, 0.95-1.93; p = 0.091). Each additional day of MV was associated with an approximately 5% increase in the mortality risk. Table 4Univariate and multivariate logistic regression analyses of 28-d mortality stratified by m-NUTRIC score (n = 258)Risk groupUnivariateMultivariateOR (95% CI) p OR (95% CI) p Overall patients Age, y1.02 (0.98-1.06)0.332 BMI, kg/m21.01 (0.92-1.10)0.902 Vasopressors3.51 (0.99-12.45)0.052 CRRT8.85 (2.64-29.67)<0.0014.07 (1.10-15.09)0.036 Duration of MV1.08 (1.03-1.13)<0.0011.05 (1.01-1.10)0.020 m-NUTRIC score1.67 (1.20-2.33)0.0031.36 (0.95-1.93)0.091 Length of ICU stay, d1.05 (1.00-1.09)0.037 APACHE II1.10 (1.02-1.18)0.012 SOFA score1.44 (1.19-1.74)<0.001 Day-5 cumulative adequate calorie intake ≥50%1.79 (0.69-4.68)0.235Low-NUTRIC score Age, y0.94 (0.86-1.02)0.112 BMI, kg/m20.95 (0.73-1.23)0.702 Duration of MV1.35 (1.11-1.65)0.0031.32 (1.07-1.64)0.010 PNI ≥400.00 (0.00-0.00)0.998CONUT category Absent to mild (0-4)0.00 (0.00-0.00)0.998 Moderate (5-8)0.48 (0.04-5.47)0.555 Severe (9-12)11.00 (0.94-128.61)0.0563.62 (0.19-67.42)0.388Day-5 cumulative adequate calorie intake ≥50%4.93 (0.43-56.47)0.199High-NUTRIC score Age, y1.03 (0.93-1.15)0.548 BMI, kg/m21.02 (0.97-1.07)0.482 Duration of MV1.05 (1.01-1.10)0.0181.06 (1.01-1.10)0.010 PNI ≥401.15 (0.24-5.59)0.859CONUT category Absent to mild (0-4)1.30 (0.34-4.97)0.706 Moderate (5-8)1.14 (0.39-3.33)0.8070.89 (0.21-3.78)0.870 Severe (9-12)0.71 (0.22-2.36)0.5780.45 (0.08-2.44)0.352Day-5 cumulative adequate calorie intake ≥50%1.25 (0.43-3.65)0.683The concordance index (C-index) of the final multivariable model was 0.70. p < 0.05 was significant.APACHE-II = Acute Physiology and Chronic Health Evaluation II; CONUT = Controlling Nutritional Status score; CRRT = continuous renal replacement therapy; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; OR = odds ratio; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. Univariate and multivariate logistic regression analyses of 28-d mortality stratified by m-NUTRIC score (n = 258) The concordance index (C-index) of the final multivariable model was 0.70. p < 0.05 was significant. APACHE-II = Acute Physiology and Chronic Health Evaluation II; CONUT = Controlling Nutritional Status score; CRRT = continuous renal replacement therapy; m-NUTRIC = modified Nutrition Risk in the Critically Ill; MV = mechanical ventilation; OR = odds ratio; PNI = Prognostic Nutritional Index; SOFA = Sequential Organ Failure Assessment. The MV duration was independently associated with 28-day mortality in the low- and high-risk groups (low m-NUTRIC: OR = 1.32; 95% CI, 1.07-1.64; p = 0.010; high m-NUTRIC: OR = 1.06; 95% CI, 1.01-1.10; p = 0.010). The day-5 cumulative adequate calorie intake (≥50%), PNI (≥40), and CONUT categories were not significantly associated with 28-day mortality in either stratum. Fig. 2 shows the Kaplan-Meier survival curves for 28-day ICU mortality stratified by day-5 cumulative adequate calorie intake of ≥50% and <50%. The log-rank test revealed no significant between-group difference in survival (p = 0.880). An exploratory subgroup analysis stratifying patients by m-NUTRIC score (<5 vs ≥5) and day-5 cumulative adequate calorie intake (<50% vs ≥50%) showed no significant difference in 28-day survival among the four groups (log-rank p = 0.521; Supplementary Figure 2 https://links.lww.com/JCMA/A417). Kaplan-Meier survival curves stratified by day-5 cumulative adequate calorie intake. Kaplan-Meier analysis of 28-d ICU survival based on day-5 cumulative adequate calorie intake at ICU admission. Patients were classified into two groups: day-5 cumulative adequate calorie intake ≥50% and <50%. The survival curves show no significant difference between the groups (log-rank test, p = 0.880). ICU, intensive care unit. Fig. 3 shows the results of ROC curve analysis. The m-NUTRIC score showed acceptable predictive performance for 28-day ICU mortality (AUC = 0.70; Fig. 3A). The optimal cutoff value was ≥5, and the Youden index was 0.27 (sensitivity = 83.3%, specificity = 43.3%; p < 0.001). The MV duration had the highest discriminative ability (AUC = 0.80; Fig. 3B). The optimal cutoff was ≥6 days, and the Youden index was 0.61 (sensitivity = 83.3%, specificity = 77.9%; p < 0.001). ROC curves of mechanical ventilation duration and m-NUTRIC score for predicting 28-d ICU mortality. ROC curves illustrating the predictive ability of the m-NUTRIC score and duration of MV for 28-d ICU mortality. A, The m-NUTRIC score showed an AUC of 0.70 (95% CI, 0.62-0.78), with an optimal cutoff of ≥5 and a Youden index of 0.27 (sensitivity = 83.3%, specificity = 43.3%, p < 0.001). B, Duration of MV showed the highest performance (AUC = 0.80; 95% CI, 0.73-0.87), with a cutoff of ≥6 d and a Youden index of 0.61 (sensitivity = 83.3%, specificity = 77.9%, p < 0.001). AUC, area under the curve; ICU, intensive care unit; m-NUTRIC, modified Nutrition Risk in the Critically Ill; MV, mechanical ventilation; ROC, receiver operating characteristic. 4. DISCUSSION This study evaluated the utility of the m-NUTRIC score in assessing the nutritional risk of critically ill patients with sepsis and examined the association between adequate caloric intake and clinical outcomes. Cumulative adequate calorie intake of ≥50% on day 5 was not significantly associated with a lower 28-day mortality rate among patients with high nutritional risk (m-NUTRIC score ≥5). Multivariate logistic regression revealed that MV duration and CRRT use were independent predictors of 28-day ICU mortality. ROC curve analysis further demonstrated that MV duration had the highest discriminative ability among all the variables examined (AUC = 0.80). A prolonged MV duration was significantly associated with increased short-term mortality and may serve as a prognostic indicator in these patients. There is no consensus on the optimal energy-provision strategy during the acute phase of sepsis. Several studies and clinical guidelines recommend that energy delivery during the early ICU period (within the first 3-5 days) should not exceed approximately 70% of the estimated requirements. A target range of 50% to 70% is generally considered safe.13,14 Achieving ≥50% of the estimated adequate calorie intake by day 5 was not associated with a reduction in 28-day ICU mortality, even for patients at high nutritional risk. This finding is consistent with evidence from large randomized controlled trials, including the NUTRIREA-3 and PermiT trials, which demonstrated that the early aggressive pursuit of full caloric targets does not improve survival in critically ill populations. The NUTRIREA-3 trial reported that a lower-energy and lower-protein strategy was not associated with reduced 90-day mortality but rather was linked to improved clinical recovery and fewer gastrointestinal complications.15,16 Lower-calorie or trophic feeding strategies are safe and feasible during the early phase of septic shock (within 48-72 hours).4 Additional stratified analyses of multiple nutritional assessment indicators, including adequate calorie intake, PNI, and CONUT score, showed that none of these indicators was independently associated with 28-day ICU mortality after multivariate adjustment. Our ROC curve and multivariate analyses further quantified this impact, demonstrating that MV duration held the highest discriminative power among all examined variables (AUC = 0.80; OR = 1.05; 95% CI, 1.01-1.10; p = 0.020). This prolonged dependence on ventilatory support likely reflects the combined burden of severe respiratory failure and refractory systemic inflammation, underscoring its value as a robust prognostic indicator in this critically ill population. Notably, patients who achieved ≥50% day-5 cumulative adequate calorie intake paradoxically had longer ICU stays and MV durations. This likely reflects greater illness severity and higher organ support requirements, rather than a direct causal effect of increased caloric intake. Similar patterns have been reported in subgroup analyses of the NUTRIREA-3 and PermiT trials, where higher caloric intake was associated with greater illness severity and prolonged ICU stay but not with reduced mortality.15,16 Consistent with these findings, MV duration demonstrated stable and clinically relevant predictive performance in the logistic regression and ROC analyses. Yoo et al17 reported a substantial decline in the predictive accuracy of the m-NUTRIC score in patients requiring prolonged MV, which supports the concept that baseline nutritional risk tools may lose prognostic precision as clinical courses become increasingly prolonged and complex. Relying on a single caloric threshold, such as ≥50% day-5 cumulative adequate calorie intake, may not adequately capture the effects of early nutritional intervention during the acute phase of critical illness. An individualized approach that integrates nutritional risk stratification with dynamic adjustment of caloric targets based on clinical progression is warranted.18,19 Among patients with sepsis at high nutritional risk (m-NUTRIC score ≥5), the m-NUTRIC score demonstrated acceptable discriminative ability for 28-day ICU mortality, with high sensitivity despite limited specificity. These findings support the use of the m-NUTRIC score as an early risk-screening tool rather than a precise predictor of short-term mortality and underscore the importance of incorporating nutritional assessment into a broader, clinically driven management strategy. This study had some limitations. First, its single-center, retrospective design limits the generalizability of the findings. Second, the energy requirements were estimated using weight-based predictive equations rather than indirect calorimetry, which may have reduced the accuracy of the energy estimates. Third, the incomplete documentation of oral intake in some patients further limited the analyzable sample size. Despite these limitations, our findings suggest that full caloric intake (≥70%-100% of resting energy expenditure) during the acute phase of sepsis may not provide measurable clinical benefits. A more conservative nutritional approach, such as providing 50% to 60% of the estimated energy requirements or trophic feeding during the early phase of critical illness, may be appropriate. This should be followed by the gradual advancement toward guideline-recommended targets as patients stabilize to reduce the risk of overfeeding. Further prospective multicenter studies using indirect calorimetry and standardized nutritional assessment tools are needed to refine the evidence-based nutritional protocols for critically ill patients with sepsis. In conclusion, achieving at least 50% of the estimated calorie needs within the first 5 days of ICU stay did not improve 28-day ICU survival in this cohort of critically ill patients with sepsis. Further investigations are needed to elucidate whether early nutritional interventions are beneficial for this population. APPENDIX A. SUPPLEMENTARY DATA Supplementary data related to this article can be found at https://links.lww.com/JCMA/A416 and https://links.lww.com/JCMA/A417. Supplementary Material