Nutritional practices and impact of feeding adequacy on clinical outcomes in Chinese respiratory intensive care units patients: a prospective observational study (ORIENT study)

TL;DR

This study examines nutritional practices in Chinese respiratory ICU patients and finds that overfeeding increases mortality risk in older patients.

Contribution

The study provides new insights into the relationship between energy overfeeding and mortality in older ICU patients in China.

Findings

Only 36.8% of patients received early enteral nutrition within 24 hours.

Overfeeding was linked to higher mortality risk in patients over 50 years old.

Protein intake did not significantly affect 28-day mortality outcomes.

Abstract

The current status of nutritional support for patients in Respiratory ICU across mainland China remains inadequately characterized. This multi-center study, conducted in RICUs, was designed to investigate nutritional practices in this specific patient population, focusing on initiation timing, energy/protein adequacy, and associations with clinical outcomes.

A prospective, observational study enrolled 1,026 patients (ICU stay >48 h) across 68 Chinese RICUs. We analyzed EN initiation rates (24 h/48 h), caloric/protein intake adequacy during the first 7 days, and outcomes via multivariable Cox regression.

EN initiation occurred in 36.8% (24 h) and 43.4% (48 h) of patients. Excluding patients who could take oral food, the proportion of patients who started EN within 24 and 48 h increased to 86.9 and 95.3%. Among 499 EN-fed patients, 26.1% developed EN complications. Caloric analysis (n = 317) identified three trajectories: underfeeding (<70% targets, 32.5%), adequate feeding (70–110, 43.2%), and overfeeding (>110, 24.3%). Overfeeding independently predicted higher non-social infections and significantly increased mortality risk in patients >50 years (HR = 1.83, 95% CI 1.02–3.28; p = 0.04). Mean protein intake was 0.9 g/kg/day, with no 28-day mortality benefit at higher thresholds (≥1.2 g/kg: p = 0.31; ≥1.3 g/kg: p = 0.42).

This multicenter study demonstrates optimal early EN initiation rates in Chinese RICUs. Energy overfeeding was associated with increased mortality risk only in patients >50 years and non-social infections risk, whereas protein adequacy showed no outcome associations. Protocolized EN delivery balancing adequacy and overfeeding risks is urgently needed in RICUs.

Identifier, NCT04958447.

Figures

Flow of patients recruitment.

Cox analysis of overfeeding and 28-day mortality in patients with respiratory critical illness.

Tables

Table 1. Baseline of patient characteristics.

Variable	n = 1,026(%)
Age, mean(SD), y	70 ± 14.1
Male	690 (67.3)
BMI, mean (SD), kg/m2	22.34 ± 4.19
BMI<18.5	170 (16.6)
18.5 ≤ BMI <30	812 (79.1)
BMI ≥ 30	44 (4.3)
APACHEII score, median (Q1, Q3)	14 (10.19)
SOFA score, median (Q1, Q3)	4 (2.6)
GCS score, median (Q1, Q3)	15 (11.15)
Simplified NUTRIC score	3 (2.4)
Simplified NUTRIC score ≥5	256 (25.0)
NRS-2002 score	6 (4.7)
NRS-2002 score (≥5)	686 (66.9)
NRS ≥ 5 and NUTRIC ≥5	213 (20.8)
NRS ≥ 5 or NUTRIC ≥5	729 (71.1)
Causes of ICU admission
Severe penumonia	407 (39.7)
Extrapulmonary sepsis	27 (2.6)
COPD	241 (23.5)
ARDS	18 (1.8)
Malignant tumor	86 (8.4)
Albumin	32.6 ± 5.8
Proalbumin*	142.9 ± 77.1
Patients not measured Proalbumin (%)	407 (39.7)
VFDs	28 (20.28)
ICU-free days	17 (3.22)
Number of deaths	141 (13.7)
Non-social infections	74 (7.2)

Table 2. The characteristics of initiating EN in patients in the RICU.

Index	Results (n = 1,026)
EN initiation within 7d	499 (48.6%)
EN initiation within 24 h of RICU admission	378 (36.8%)
EN initiation within 48 h RICU admission	445 (43.4%)
EN initiation within 24 h of RICU admission in high nutritional risk patients (mNUTRIC ≥5; 256)	138 (53.9%)
EN initiation within 48 h of RICU admission in high nutritional risk patients (mNUTRIC ≥5; 256)	154 (60.2%)
EN initiation within 24 h of RICU admission in high nutritional risk patients (NRS2002 ≥ 5; 686)	306 (44.6%)
EN initiation within 24 h of RICU admission in high nutritional risk patients (NRS2002 ≥ 5; 686)	337 (49.1%)
AGI on EN initiation	491
Gastrointestinal function intact	332 (67.6%)
Class I	114 (23.2%)
Class II	39 (7.9%)
Class III	6 (1.2%)
Class IV	0 (0.0%)
Shock on EN initiation within 24 h of RICU admission	46 (12.2%)
Shock on EN initiation within 48 h of RICU admission	8 (1.8%)
Reasons for no EN feeding within 48 h
Oral feeding	523 (90%)
Uncontrolled shock	21 (3.6%)
Uncontrolled upper gastrointestinal bleeding	21 (3.6%)
Threatening hypoxemia and acidosis	13 (2.2%)
Others	3 (0.5%)

Table 3. EN intolerance and discontinuation during EN feeding.

Variables	Number (%)
EN intolerance (130)
Abdominal distention	43 (33.1)
Abdominal pain	5 (3.8)
Diarrhea	41 (31.5)
Nausea	8 (6.2)
Vomiting	16 (12.3)
Reflux	18 (13.8)
Increased GRV(ml)	24 (18.5)
Aspiration	2 (1.5)
Abnormal bowel sound	5 (3.8)
Gastrointestinal hemorrhage	17(13.1)
Nasoenteric Tube Displacement	2 (1.5)
Nasogastric Tube	2 (1.5)
Constipation	1 (0.7)
EN discontinuation due to intolerance (68)
Abdominal distention	14 (20.6)
Abdominal pain	3 (4.4)
Diarrhea	9 (13.2)
Nausea	5 (7.4)
Vomiting	11 (16.2)
Reflux	13 (19.1)
Increased GRV	24 (35.3)
Aspiration	2 (2.9)
Abnormal or absence of bowel sounds	6 (8.8)
Gastrointestinal hemorrhage	13 (19.1)
Nasotube intolerance	3 (4.4)
Feeding tube obstruction or translocation	3 (4.4)

Table 4. Aspiration precautions for high aspiration risk patients.

Index	Results (%)
High aspiration risk
Bed to care ratio <1:2.5	85 (17.1)
Mechanical ventilation or prolonged horizontal supine >72 h	270 (54.2)
Age >70 years old	263 (52.8)
Vomit	13 (2.6)
Past history of aspiration	64 (12.9)
Neuromuscular disease or structural abnormalities of the Aerodigestive tract	54 (10.8)
Decreased level of consciousness	219 (44.0)
Aspiration precautions
Position	494 (99.2)
The head of the bed is raised ≥30°	398 (79.9)
The head of the bed is raised <30°	89 (17.9)
Horizontal Supine	7 (1.4)
Oral careFrequency (times/day)	490 (98.4)2 (2,3)
Prokinetic agents	142 (28.5)
Mosapride	90 (63.4)
Metoclopramide	15 (10.6)
Itopride	15 (10.6)
Monitor of cuff pressure (≥25 cmH2O)	211 (100.0)

Table 5. The correlation between feeding adequacy and non-social infections in patients with respiratory critical illness.

	Model 1			Model 2
	OR	95% CI	p	OR	95% CI	p
A. Comparison of the effects of overfeeding and underfeeding on non-social infections
Underfeeding	1			1
Overfeeding	2.48	1.12–5.50	0.02	2.73	1.09–6.85	0.03
B. Comparison of the effects of adequately feeding and overfeeding on non-social infections
Adequately feeding	1			1
Overfeeding	2.03	1.00–4.14	0.05	2.65	1.16–6.06	0.02

Full text

Background

Critically ill patients admitted to intensive care units (ICUs) frequently encounter nutritional challenges stemming from the hypercatabolic state induced by acute critical illness. During this acute phase, the body initiates tissue catabolism to generate metabolic substrates that support essential acute-phase physiological priorities, including hemostatic regulation and immune defense mechanisms.

This hypercatabolic state, however, may precipitate significant protein-energy deficits in this population, which has been strongly correlated with detrimental clinical outcomes and substantial healthcare expenditures. This cascade manifests clinically as heightened susceptibility to non-social infections, prolonged ventilator dependence, extended ICU hospitalization, and elevated mortality risk (1–3).

Despite the pervasive issue of malnutrition in ICU settings, this metabolic derangement frequently remains underrecognized and persists as a global health challenge. Nutrition, especially enteral nutrition (EN), has been widely accepted as the standard care for critically ill patients (4–7). EN is considered more physiologically appropriate, helps maintain gastrointestinal integrity, is easy to administer, and is cost-effective (8, 9). Growing clinical evidence shows that targeted nutritional support can help patients in several ways. It maintains gut health, repairs tissue, regulates the immune system, reduces excessive metabolism after injury, and lowers infection risk. Together, these benefits help patients recover faster and shorten hospital stays (10).

The volume of scholarly publications in clinical nutrition has expanded considerably in recent years. Nevertheless, current evidence remains marked by persistent controversies and unresolved questions regarding critical care feeding protocols. Fundamental questions remain regarding the use of early enteral versus parenteral nutrition in high-risk patients, optimal feeding timing for unstable patients, and the value of gastric residual volume monitoring. Furthermore, there is a significant gap between guideline recommendations and clinical practice, as evidenced by inadequate enteral nutritional intake (11) and unnecessary feeding interruptions (12–14). While RICU patients demonstrate overlapping clinical profiles with general ICU populations, they present distinct pathophysiological characteristics that necessitate prioritized management of oxygenation deficits, carbon dioxide retention disorders, acid–base homeostasis disturbances, and severe pulmonary infections. These compounding barriers pose significant challenges to delivering effective nutritional interventions and achieving prescribed caloric and protein requirements. There is a lack of epidemiological data, specifically from RICUs, particularly in China.

This multicenter prospective observational study constitutes the largest systematic investigation to date specifically targeting nutritional support practices in RICUs across China. The primary objective is to establish comprehensive epidemiological data regarding EN administration protocols in Chinese RICU settings. A secondary analysis of this dataset specifically examined the relationship between nutritional adequacy levels during the first week following ICU admission and key clinical outcomes in critically ill respiratory patients.

Materials and methods

Study design

This prospective observational investigation was implemented across 68 RICUs spanning multiple provinces in mainland China from July 1 (00:00) to July 21 (24:00), 2021. Demographic and clinical parameters were systematically documented at RICU admission, with patient outcomes evaluated at 28 days post-enrolment. Data acquisition utilized standardized electronic case report forms (eCRFs) administered through a secure web-based platform.

The study was approved by the ethics committee of China-Japan Friendship Hospital (approval no. 2021–38-K22), and written informed consent was obtained from all patients or their legal representatives. The inclusion criteria were as follows:

Patients admitted to the RICU who survived for more than 48 h.Patients aged 18 years or older.

The exclusion criteria were as follows:

Severe trauma cases (defined as life-threatening trauma requiring emergency surgery and treatment, including criteria such as low systolic blood pressure, abnormal respiratory rate or heart rate, serious disturbance of consciousness, traumatic limb amputation, flail chest, multiple long bone fractures, or fall injuries from a height of more than 3 meters).Surgical patients admitted to the RICU.Patients or their legal representatives who declined to participate in the study.

Data collection

Standard demographic variables including age, sex, admission source, weight, and height were systematically documented. Within 24 h of RICU admission, Acute Physiologic and Chronic Health Evaluation II (APACHE II) scores and Sequential Organ Failure Assessment (SOFA) scores were calculated.

Nutritional risk status

Nutritional risk stratification was performed using the modified Nutrition Risk in Critically Ill (mNUTRIC) score and Nutritional Risk Screening 2002 (NRS 2002) instruments, with scores ≥5 defining the high-risk cohort according to standardized clinical thresholds. Routine laboratory tests were conducted to measure plasma indicators like albumin and proalbumin.

Feeding performance

The primary objective was to evaluate both the timing of nutritional support initiation (within 24 and 48 h) and the modalities employed, including oral intake versus medical nutrition therapy encompassing EN, total parenteral nutrition (TPN), or supplemental parenteral nutrition (SPN). Energy and protein requirements for each nutritional intervention were systematically recorded. EN was defined as feeding through tubes, excluding oral feeding, which was measured separately. PN is any peripheral or central intravenous administration of nutrition. Details of medical nutrition delivery were recorded within the first 7 days of RICU admission. Enteral feeding volume, energy density of EN formula, feeding routes, and feeding styles were documented. PN volume and composition of nutrients (carbohydrates, proteins, and fats) were also recorded.

Feeding tolerance and complications

Hemodynamic and gastrointestinal function parameters were evaluated before EN initiation. Measures to reduce the risk of aspiration during EN administration, such as bed head angle, GRV, use of gastrointestinal motility drugs, tracheal tube balloon pressure in patients with artificial airways, and nasojejunal tube usage, were recorded. Reasons for discontinuing EN within the first 7 days of RICU admission were documented. Intolerance to EN included symptoms like nausea, vomiting, aspiration, abdominal pain and distension, high GRV, and diarrhea.

Feeding adequacy

Current critical care protocols advocate early nutritional intervention initiation (typically within 24–48 h post-admission) for hemodynamically stabilized patients. Nevertheless, during the initial 48-h window following RICU admission, suboptimal nutrient delivery frequently occurs secondary to compromised hemodynamic profiles and disease severity. By the third post-admission day, when physiological parameters demonstrate progressive stabilization, both caloric and protein provisions achieve therapeutic consistency. Consequently, we designated the mean energy/protein delivery during days 3–5 as established clinical parameters for nutritional assessment.

Energy requirements were calculated based on body weight and adjusted for BMI to better reflect metabolic needs. The following calculations were applied: BMI 25–30 kg/m^2^: 25–30 kcal/kg/day; BMI 30–50 kg/m^2^: 11–14 kcal/kg/day; BMI > 50 kg/m^2^: 22–25 kcal/kg/day. Adequate feeding was defined as achieving 70–110% of the required energy intake. Energy underfeeding refers to intake below 70%, and overfeeding indicates intake above 110% (7). The recommended protein intake ranges from 1.2 to 2.0 g/kg/day (4).

Clinical outcomes

Clinical outcomes, recorded at 28 days after RICU admission, included survival status (alive, dead, or lost), non-social infections, days free from mechanical ventilation, and RICU-free days.

The dataset contained missing values, and single imputation (EM) was performed to address these missing values.

Statistical analysis

Descriptive analyses were conducted using appropriate statistical measures: categorical variables were expressed as frequency distributions (%) while continuous variables were reported as mean ± standard deviation or median with interquartile range (IQR). Missing data were handled without assumptions, and the characteristics of patients with missing outcomes were compared to those with complete data.

Proportions were compared using χ^2^ or Fisher’s exact tests for categorical variables, while the t-test or Wilcoxon rank sum test was used for continuous variables, as appropriate. Multivariable Cox proportional hazards analyses were performed to evaluate the association between feeding adequacy and 28-day mortality, while logistic regression models were employed to assess relationships with other clinical outcomes (e.g., hospital-acquired infections, ICU length of stay). A two-sided p-value of 0.05 or less was considered statistically significant. Unless otherwise specified, statistical analyses were performed using SPSS software, version SPSS 26.

Results

ICU and hospital characteristics

The multicenter investigation encompassed 68 RICUs across general hospitals (see Supplementary Table 1). Notably, only 13.2% of participating RICUs employed professional dietitians, while 32.4% maintained established multidisciplinary nutritional support teams.

Characteristics and outcomes of enrolled patients

From an initial screening cohort of 1,297 patients, 1,026 met inclusion criteria (Figure 1). Table 1 summarizes baseline characteristics and clinical parameters of the study cohort. Among them, the mean age was 70.0 ± 14.1 years. The median APACHE II score and SOFA scores were 14 (10.19) and 4 (2.6), respectively. The median ventilator-free days (VFDs) and ICU-free days at 28 days were 28 days (20.28) and 17 days (3.22), respectively. The 28-day mortality rate was 13.7% (141/1026), and non-social infections occurred in 7.2% (74/1026) of patients.

Flow of patients recruitment.

Nutrition risk score

The median mNUTRIC and NRS2002 scores were 3 (2.4) and 6 (4.7), respectively. Based on these assessments, 256 patients (25.0%) and 686 patients (66.9%) met criteria for high nutritional risk via the mNUTRIC and NRS2002 scales, respectively. Within the cohort of 1,026 critically ill patients, 170 (16.6%) exhibited malnutrition upon ICU admission, operationally defined as a body mass index (BMI) < 18.5 kg/m^2^. Biochemical analysis revealed mean serum albumin and prealbumin concentrations of 32.6 ± 5.8 g/L and 142.9 ± 77.1 g/L, respectively (Table 1).

Nutritional support

Within the initial 7 days following RICU admission, nutritional support was initiated in 1,018 patients (99.2% of the cohort). Of these, 97.8% received oral intake and/or EN, with SPN administered in combined regimens. Timely nutritional intervention (≤48 h post-admission) was implemented in 88.6% of cases (909/1026). The modalities of nutritional support during the critical 48-h window demonstrated the following distribution: exclusive oral feeding (37.5%), EN monotherapy (35.2%), PN alone (4.8%), combined EN + PN (9.1%), oral+EN coadministration (2.3%), oral+PN supplementation (6.7%), and oral intake with oral nutritional supplements (4.3%).

EN initiation

During the initial 7-day period following RICU admission, 499 patients received EN, with 378 (75.8%) and 445 (89.2%) initiating EN therapy within 24 and 48 h of RICU admission, respectively. Among patients stratified as high nutritional risk by mNUTRIC scoring, EN initiation occurred within 24 h in 138 cases (53.9%) and within 48 h in 154 cases (60.2%). Conversely, for those categorized as high nutritional risk through NRS2002 assessment, timely EN initiation was achieved in 306 cases (44.6%) within 24 h and 337 cases (49.1%) within 48 h.

Shock was observed in 12.8% of the cohort upon EN initiation. Among patients who commenced EN within the initial 7-day period, 67.6% had intact gastrointestinal (GI) function, 23.2% were in acute gastrointestinal injury grade (AGI) I, 7.9% were in AGI II, 1.2% were in AGI III, and none were in AGI IV at the time of EN initiation. The most common reasons for not initiating EN within 48 h were oral feeding, uncontrolled shock, and uncontrolled upper gastrointestinal bleeding (Table 2).

Type of EN route and EN formulation

Among 499 patients requiring EN support, nasogastric tube placement was utilized in 457 cases (91.6%), while a nasojejunal tube was inserted for 42 patients (8.4%). Continuous pump infusion constituted the predominant EN delivery method (82.1%), followed by intermittent pump-driven administration (7.6%) and gravity-controlled feeding protocols (4.4%).

Among the patients, 15.1% initially opted for elemental formulas, 68.3% received intact protein formulas, and 6.6% chose a homogenized diet. The prevalent choice among intact protein formulations was high-energy EN formulae (32.9%), followed by normal-energy formulae (26.9%) and low-energy formulae (11.8%). Only in 36 cases were changes in their EN formulation performed.

EN complications

During EN administration, gastrointestinal complications occurred in 26.1% (130/499) of patients, with 5.0% (25/499) developing two distinct complications and 1.6% (8/499) manifesting three concurrent adverse events. The principal etiologies of enteral feeding intolerance comprised abdominal distension (33.1%), diarrhea (31.5%), and elevated gastric residual volumes (18.5%). EN discontinuation was required in 52.3% (68/130) of affected patients, with the predominant indications being persistent gastric residual volume elevation (35.3%) and refractory abdominal distension (20.6%; Table 3).

Aspiration precautions

Among the cohort of patients initiating EN within the initial 7-day period, 79.8% (398/499) received feedings in the semi-recumbent position (head-of-bed elevation ≥30°), with the remaining cohort receiving feedings in supine positioning or suboptimal elevation angles. Gastrointestinal prokinetic agents were used in 28.5% of patients. Among 266 patients requiring invasive mechanical ventilation, 211 underwent endotracheal cuff pressure monitoring, with all measurements maintained within the clinically recommended range of 25–30 cmH2O. Standardized oral care protocols were administered to 98.4% of the cohort, with standard implementation twice daily throughout the intensive care stay (Table 4).

Nutritional adequacy in RICU patients

The mean nutritional adequacy for energy and protein provisions during post-RICU admission days 3–5 served as the primary metric for evaluating first-week nutritional attainment. In the analytical cohort of 317 patients with validated energy intake records, nutritional status stratification revealed: 103 underfed (32.5%), 137 adequately nourished (43.2%), and 77 overfed (24.3%) individuals. Multivariable logistic regression demonstrated significantly elevated non-social infection risks in overfed patients compared to both underfed counterparts (adjusted OR = 2.73, 95% CI 1.09–6.85; p = 0.03) and adequately feeding patients (adjusted OR = 2.65, 95% CI 1.16–6.06; p = 0.02; Table 5). No statistically significant difference was observed between patients with adequately feeding and underfeeding. However, no significant differences were detected among the three groups regarding 28-day mortality, duration of invasive ventilator use, or length of non-ICU hospital stay. Notably, in patients aged ≥50 years, after adjusting for age, APACHE II score, NRS2002 score, ALB and non-social infections, patients with overfeeding demonstrated a significantly increased risk of 28-day mortality compared to those with non-overfeeding (HR = 1.83, 95% CI [1.02–3.28], p = 0.04; Figure 2).

Cox analysis of overfeeding and 28-day mortality in patients with respiratory critical illness.

Protein intake analysis in 298 evaluable patients revealed a cohort-wide mean provision of 0.9 g/kg/day. Suboptimal protein delivery (<1.2 g/kg/day) was observed in 218 cases (73.2%), while 80 patients (26.8%) met or exceeded the 1.2 g/kg/day threshold. Multivariable regression models identified no significant associations between protein adequacy levels and primary clinical endpoints (Supplementary Table 2; Supplementary Figure 1).

For the purpose of statistical analysis, patients were categorized into three groups based on their primary diagnosis: chronic obstructive pulmonary disease (COPD), severe pneumonia (including sepsis, septic shock, or ARDS), and others. The results indicated no significant intergroup differences in enteral nutrition intolerance rate, in-hospital infection rate, 28-day mortality, invasive mechanical ventilation, age, BMI, APACHE II, or NUTRIC scores (Supplementary Table 3).

Discussion

This study, conducted in mainland China, represents the most extensive prospective observational study in RICUs. In 1,026 critically ill respiratory patients, this study assessed nutrition therapy delivery and the association of feeding adequacy with clinical outcomes during the first 7 days in the RICU.

Numerous studies highlight the benefits of early nutrition for high-risk patients (15–19). Doig et al.’s meta-analysis of six randomized controlled trials showed that early enteral nutrition (within 24 h of ICU admission) was associated with significant reductions in mortality (OR = 0.34, 95% CI 0.14–0.85) and pneumonia (OR = 0.31, 95% CI 0.12–0.78). Based on US costs, economic analyses revealed meaningful reductions in total care costs, with acute hospital care costs decreasing by US$14,462 per patient (95$5,464 to US$23,669). ASPEN and the Chinese expert consensus on nutrition advocated for early initiation of enteral nutrition within 24 or 48 h of ICU admission (6, 7). However, real-world epidemiological studies have shown less satisfactory results. A global study involving 9,777 patients from 46 countries reported that only 10% received enteral nutrition on the first day, increasing to over 40% after 5 days. The most extensive epidemiological study in Chinese general ICUs revealed suboptimal delivery of enteral nutrition, with proportions of patients starting enteral nutrition within 24, 48, and 72 h of ICU admission at 24.8, 32.7, and 40.0%, respectively (20).

Within our cohort of 1,026 RICU patients, EN was initiated in 378 cases (36.8%) within the initial 24-h period post-admission, increasing to 445 patients (43.4%) by 48 h, with cumulative EN implementation reaching 499 patients (48.6%) within the first week. Excluding patients who could take oral food, the proportion of patients who started EN within 24 and 48 h increased to 86.9 and 95.3%. Notably, complete nutritional deprivation persisted in only 14.8% (152/1,026) and 11.5% (118/1,026) of patients during the first 24-h and 48-h post-admission periods, respectively. The high percentage of patients in good nutritional condition within 48 h can be attributed to several factors. Patients admitted to RICUs were not severely ill, with only a quarter requiring invasive mechanical ventilation and over a third being able to eat orally. Additionally, the recognition and understanding of the importance of nutrition among RICU doctors likely contributed, partly due to the publication and promotion of the Chinese expert consensus on nutrition for critically ill respiratory patients (6).

In our study, the incidence of shock was 12.2% in patients who initiated EN within 24 h. Despite this, EN tolerance is acceptable. 26.1% of patients receiving EN developed complications, with 12.4% requiring discontinuation due to intolerance (e.g., high gastric residuals and abdominal distension). These patterns mirror findings from Chinese ICU cohort studies (20).

Previous studies have shown that both overfeeding and underfeeding negatively impact critically ill patients, particularly those at high nutritional risk (21–25). Inadequate calorie and protein intake have been reported in previous observational studies, with variations observed globally (26–28). For instance, a Canadian ICU survey found that 16% of patients received no nutrition support, and those who did received only 56 to 62% of estimated calorie and protein needs during the first 12 days (28). Similarly, a cross-sectional study in 116 Chinese ICUs showed that a minority of patients received over 80% of the estimated energy target within 24, 48, 72 h, and 7 days after ICU admission (20).

Within the analyzed cohort of patients with complete energy intake documentation (n = 317), nutritional delivery patterns during the initial 7-day post-RICU admission period comprised three distinct categories: suboptimal (32.5%, n = 103), adequate (43.2%, n = 137), and excessive (24.3%, n = 77). Multivariable analysis identified energy-overfed status as demonstrating significant associations with both non-social infection susceptibility and elevated mortality risk in geriatric subgroups. This pathophysiological mechanism appears mediated through nutrient restriction-induced autophagy upregulation, an evolutionarily conserved cytoprotective process that augments innate antimicrobial defenses. Conversely, excessive caloric provision may attenuate these protective pathways, potentially explaining the observed mortality risk elevation in overfed cohorts (29, 30). Stress-induced hyperglycemia represents a pathophysiologically conserved adaptive metabolic mechanism in critical illness, evolutionarily optimized to prioritize fuel substrate mobilization during systemic homeostatic challenges (31, 32). Overfeeding potentiates hyperglycemia and insulin resistance via impaired glucose homeostasis, consequently increasing susceptibility to non-social infections. This pathophysiological correlation was substantiated in the EPANIC trial, with mechanistic analysis suggesting caloric excess as a contributing determinant to infectious complications (33).

Whereas seminal works by Zusman et al. and Weijs et al. established both underfeeding and overfeeding as independent mortality predictors in ICU patients (34), our analysis revealed no significant mortality association with underfeeding status. This discordance likely reflects fundamental divergences in pathophysiological characteristics and illness severity between RICU and general ICU patient cohorts.

No statistically significant differences in clinical outcomes were observed between adequately fed and underfed patients, which is consistent with previous studies (35, 36). Notably, two large-scale randomized controlled trials (RCTs), the TARGET and PERMIT studies, demonstrated no significant differences in clinical outcomes among patients receiving low, normal, or high caloric intake (37, 38). Contrastingly, emerging evidence suggested reduced mortality risk in adequately feeding subgroups relative to underfed cohorts, potentially mediated by baseline nutritional risk stratification discrepancies (5, 39).

Critically ill patients typically require enhanced protein provision, with clinical studies indicating that protein intake exceeding 1.3 g/kg/day correlates with reduced 28-day mortality in mechanically ventilated patients. In contrast, standard protein supplementation at 0.8 g/kg/day shows no comparable survival benefit (40). In our cohort of patients with validated protein intake documentation, the mean protein provision during post-RICU admission days 3–5 was 0.9 g/kg/day. Crucially, our analysis demonstrated no significant mortality reduction with protein delivery exceeding either 1.2 g/kg/day (p = 0.31) or 1.3 g/kg/day (p = 0.42), replicating outcomes from the EFFORT-Protein trial (41).

Our study’s strengths include its prospective design, large sample size across 68 RICUs in China, and comprehensive parameter consideration. The detailed description of EN delivery within the first 7 days after ICU admission provides profound insights into the current EN status in Chinese RICUs. RICU patients often present with respiratory failure, pulmonary infections, sepsis, or septic shock, which significantly elevate their metabolic rate and place them in a hypercatabolic state. This leads to a sharp increase in energy and protein consumption. While inadequate nutritional support can result in rapid muscle wasting, overfeeding also poses significant risks. For patients with respiratory failure, diaphragmatic function is critical, especially in those aiming for extubation. Insufficient energy and protein intake can lead to weaning failure and impede recovery from respiratory failure. Conversely, excessive carbohydrate intake generates substantial CO₂ production. In patients with compromised ventilatory function, this increased CO₂ load can significantly elevate respiratory effort, potentially causing respiratory muscle fatigue, difficulty in weaning, or hypercapnia. This is particularly critical in patients with chronic obstructive pulmonary disease (COPD). Therefore, it is essential to develop individualized nutritional support strategies for RICU patients. These findings can inform improvement strategies across various RICUs, addressing existing issues in EN feeding.

However, there are several limitations to acknowledge. Firstly, this study was cross-sectional, which prevents establishing causal relationships. Secondly, as the study was conducted during the summer season, it may not reflect the typical patient types and severity in RICUs. Consequently, the study results might underestimate the number of high nutritional risk patients, the severity of patients, and the proportion of EN patients, showing a relatively higher proportion of oral feeding patients. Thirdly, the study’s methodological constraints stem from the limited cohort size of enterally-fed patients, restricting statistical power for analyzing EN initiation timelines, caloric/protein delivery accuracy, goal attainment, and feeding tolerance. Consequently, potential associations between underfeeding regimens and clinical prognosis remain undetermined, necessitating prospective investigations with expanded sample sizes to validate these preliminary findings. Additionally, in this investigation, caloric and protein requirements were calculated using established body weight-based equations to examine correlations between nutritional adequacy and clinical outcomes, without employing indirect calorimetry for target validation. While weight-based modeling offers practical advantages in clinical settings, subsequent research should prioritize validating these associations through indirect calorimetry-guided nutritional protocols. Lastly, it is essential to note that the findings of this study may not apply to other types of ICUs, as it was specifically conducted in RICUs.

In conclusion, early nutritional intervention initiation has become standard practice in Chinese RICUs, and the proportion of EN initiated within 48 h of RICU admission was optimistic. Our multicenter analysis substantiates that energy-overfed status were associated with elevated non-social infection incidence in critically ill respiratory patients and increased mortality risk only in subgroup over 50 years. These findings underscore the critical need for precision nutritional protocols incorporating individualized caloric requirement quantification and real-time delivery surveillance systems to balance therapeutic efficacy with iatrogenic overfeeding mitigation.