Nutritional status and clinical outcomes in the intensive care unit: A global systematic review and meta-analysis

Abstract

Background

The severity of illness and inability to eat while in the intensive care unit put ICU patients at high risk of malnutrition. Nutritional status assessment is vital for managing patients’ morbidity, length of stay, and mortality. This systematic review aimed to assess the nutritional status of adult ICU patients and evaluate the influence of malnutrition on clinical outcomes.

Methods

A comprehensive systematic review of articles published 2010 to 2024 was conducted Google Scholar, PubMed, Scopus, Web of Science, Cochrane, and Embase included, studies focusing on nutritional status in adult ICU patients. A random-effects meta-analysis model was performed via STATA 17 to determine the overall prevalence of malnutrition among the ICU patients and a forest plot was generated to visually depict the pooled estimate and individual study results. Publication bias was assessed using Egger's test and funnel plots. Heterogeneity was evaluated using I2 and Baujat plots.

Results

This systematic review included prospective and retrospective cohort and cross-sectional studies, 31 publications with a sample size of 21,413 patients. The pooled prevalence of malnutrition was 32.74 (95% confidence interval: 19.9--45.5,I² = 0.0%, τ² = 0.00) and the average length of stay for malnourished patients was 5.6 (2.920, 8.265) days. Malnourished patients had an overall risk of mortality of 1.45 (95% CI: 0.76–2.13).

Conclusion

Malnutrition is prevalent among adult ICU patients, with wide variation across studies, and the situation becomes worse after admission. Additionally, malnourished patients have longer hospital stays and are more likely to die than well-nourished patients.

1. Introduction

The nutritional status of patients in the intensive care unit (ICU) plays a crucial role in their ability to recover from critical illnesses and achieve positive clinical outcomes [1]. This is because nutrition and disease are closely interconnected [2]. Adequate nutritional intervention has been shown to attenuate the metabolic response to stress and favorably modulate immune responses. Nutritional support in critically ill patients prevents further metabolic deterioration and loss of lean body mass [3].

Malnutrition in critical care settings is a universal problem, with rates as high as 78.1% in developing nations and 50.8% in industrialized ones [4]. Depending on the severity of their organ failure, patients typically lose five to twenty-five percent of their lean body mass within the first 10 days of being admitted to the intensive care unit. Previous studies show that malnourished patients are more likely to experience worse results, including mortality, a longer hospital stay, and a higher prevalence of overall problems [5]. In critically ill patients, nutritional therapy can mitigate the effects of malnutrition. In order to initiate appropriate nutritional therapy in a timely manner, patients who are malnourished or at risk of malnutrition must be recognized. Therefore, in order to prevent adverse effects, patients' nutritional status should be continuously evaluated and appropriate nutritional support should be put in place as soon as feasible [6,7].

According to studies, malnutrition is highly prevalent among ICU patients and is associated with worse outcomes (longer stays, mortality) in LMIC settings. For instance, a prospective cohort in Sub-Saharan Africa discovered that hunger dramatically raised ICU patients' 30-day mortalit [8]. Malnutrition at hospital admission was linked to a higher likelihood of ICU admission and adult death, according to research from Ethiopia, illustrating the effects of poor nutritional status in a setting with inadequate resources [9].

Nutrition screening and valuation within intensive care units (ICU) can be commenced with general screening within 48 h post-admission employing methods such as modified Nutrition Risk in the Critically ill (mNUTRIC) score ( ≥ 5 for high-risk patients), Nutrition Risk Screening 2002 (NRS-2002) ( ≥ 5 for high-risk patients), and NRS combined with SOFA for calculation of risks like death and for prolonged ventilation. Detailed valuation can also include Subjective Global Assessment (SGA) for nutrition typing (well nourished & malnourished severe and moderate) and Global Leadership Initiative for Malnutrition with greater emphasis on etiology like inflammation. For this review studies used SGA as an assessment tool mostly included. Bedside quantification of muscle atrophy (such as a decrease of 16% within a week) within quadriceps femoris and diaphragm thickness via point-of-care ultrasonography (PU) provides accurate assessments for patients with edema than subjective determination due to edema within ICU [10,11].

Malnutrition is associated with higher mortality rates and increased complications and has a 28-day mortality rate that is significantly higher than that of their well-nourished counterparts during hospitalization [4,12]. The study revealed that the mean length of stay for patients with malnutrition was extended by 1.1--12.8 days, depending on the diagnosis-related group [13]. Another study reported an increased hospital stay of 1.43 days on average for malnourished patients [14]. Research indicates a significant correlation between early and appropriate nutritional support and a reduction in infections, delayed wound healing, and muscle wasting—factors known to prolong ICU stays for critically ill patients [13].

Nutritional management in the intensive care unit (ICU) setting is a complex challenge that requires a multidisciplinary approach. Barazzoni et al. (2020) identified three major challenges: lack of knowledge and skills among healthcare professionals, lack of agreement on how to provide optimal nutrition, and the cost of providing nutrition [15].

The problem of nutrition management in intensive care units (ICUs) is a growing concern due to the increased rate of malnourishment in intensive care unit (ICU) patients. As indicated by R Dhaliwal et al. in their 2017 study published in Nutrition in Clinical Practice, malnourishment can increase the risk of morbidity and mortality in ICU patients, resulting in an increase in the length of their stay and healthcare costs. The purpose of this meta-analysis and comprehensive review was to identify the overall nutritional status of intensive care unit patients and evaluate how it affects their outcomes.

1.1. Review question

• ⁃What is the overall prevalence of malnutrition among patients admitted to the adult ICU?
• ⁃What is the overall length of stay among malnourished patients admitted to the ICU?
• ⁃What is the overall risk of mortality among malnourished patients in the ICU?

2. Materials and methods

This systematic review were performed on cross-sectional, observational cohort and RCT studies aimed at assessing the nutritional status and clinical outcomes of adult ICU patients. We used the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses Checklist (PRISMA 2020 Statement) to organize the entire content [16].

2.1. Information search strategy

Articles published from January 2010 to June 1, 2024, were searched electronically. Systematic searches of articles published from January 2010 to June 1, 2024, were performed electronically via keywords and medical subject headings (MeSHs). We searched for published articles on Google Scholar, PubMed, Scopus, Web of Science, Cochrane, Embase, Global Health, Cumulative Index to Nursing and Allied Health Literature (CINAHL) (EBSCO) and references from previously published articles (snowball technique). The selected studies were managed, and duplicates were removed via EndNote version 20. The keywords and MeSH terms used in different databases are summarized to assess the nutritional status of adult ICU patients at admission and determine their associations with clinical outcomes in different groups of patients.

2.2. Eligibility criteria

This study included only published studies. The review included studies with prospective cohort, cross-sectional, case‒control, and RCT designs that assessed the nutritional status of adult ICU patients. For inclusion in the pooled prevalence analysis, studies needed to report both the sample size and the number of observations or the frequency of malnourished individuals. Additionally, to assess the overall association between malnutrition in the ICU and all‐cause mortality (overall survival), data on the HR with a 95% confidence interval (CI) were reported.

Our study had no geographical limitations. The systematic review included papers focusing on adult ICU patients’ nutritional status and clinical outcomes. Studies published before January 2010, as well as those in which treatment outcomes were not measured after nutritional status was assessed at admission, were excluded.

• ⁃Population (P): Adults (≥18 years).
• ⁃Intervention or exposure (I): Admission to the ICU
• ⁃Comparison (C): In this study, adults admitted to the ICU for any admission reason were included. For the risk of mortality, those whose length of stay in the ICU was well‐nourished were the reference group.
• ⁃O: Malnutrition as defined by the SGA. (primary outcome), mortality, and length of stay in the ICU (secondary outcome).
• ⁃Study (S): Observational studies (cross‐sectional and cohort studies).
• ⁃Time frame (T): 2010–2024.
• ⁃Language: Written in English.

2.3. Study selection and data extraction

The search was limited to primary research articles published in the English language. Reviews, case reports, conference abstracts, short communications, unpublished studies, theses and dissertations were excluded. The process of study selection was as follows: first, duplicate studies were removed via EndNote version 20. Second, we screened the articles on the basis of their titles, and if appropriate, we proceeded to screen the abstracts. Finally, articles that passed through the title and abstract screening processes were eligible for full‐text reading. In each process, we removed articles that were apparently irrelevant to our initial review questions.

Data were extracted via Microsoft Excel. All relevant data were extracted. Disparities between the reviewers at the time of data abstraction were resolved through discussion. The data extraction sheet, including the author's name, year of publication, study design, sample size, and nutritional assessment tool used to classify patients as well-nourished or malnourished, was extracted from the studies that met the inclusion criteria. Moreover, the proportion of malnourished patients, length of stay (mean ± SD), disease severity measurement tools, and proportion of mortality among malnourished patients were also extracted for systematic review and meta-analysis.

2.4. Statistical analysis

The overall prevalence of malnutrition among ICU patients was analyzed via STATA 17 to determine the effect size and heterogeneity across studies. A forest plot was generated to visually depict the pooled estimate and individual study results. Additionally, significant heterogeneity was assessed via the I2 statistic.

Next, the two malnutrition categories—moderate and severe—were analyzed in stratification. The distribution and weight proportions of moderately and severely malnourished patients were visualized via Python.

To assess the impact of malnutrition on clinical outcomes, the length of stay in the ICU was determined with a precomputed effect size of moderately and severely malnourished patients. A forest plot was generated in STATA 17 to determine the effect size and 95% confidence intervals. Finally, the distribution of mortality rates was compared between well-nourished and malnourished patients. Python was used to visualize the differences in mortality between the two groups. By employing these statistical techniques and data visualization methods, they were able to analyze the prevalence, severity, and association of malnutrition in ICU patients comprehensively. The forest plots generated via STATA 17 provided a concise summary of the effect sizes, whereas the Python visualizations offered an intuitive way to explore the distributions and proportions of malnutrition status and mortality rates.

3. Results

3.1. Study selection

The literature search generated 20389 articles, from which 4456 duplicates were removed, as shown in Fig. 1. The duplicates included articles with similar methodologies and subject characteristics. The screening based on article titles and abstracts generated 8325 papers for full-text review, but 8294 papers were removed because they did not meet the inclusion criteria, as indicated in Fig. 1. Finally, 31 papers were included for qualitative and quantitative synthesis (Fig. 1).

Fig. 1.

Flow diagram depicting the stages of the systematic review with the numbers of records identified, screened, excluded, and included in the final analysis, following PRISMA guidelines.

3.2. Characteristics of the studies

The total sample size across all the studies mentioned in the search results was 21,413 ICU patients. The studies cover a wide range of sample sizes, from as few as 57 patients to as many as 6518 patients. Most studies had sample sizes between 100 and 1000 patients. The majority of the studies were prospective cohort studies (16 studies). There were also 4 retrospective studies, 3 cross-sectional studies, and 3 randomized controlled trials. The mean age of the patients across the studies ranged from 36.0 to 74.2 years. Most studies included both medical and surgical ICU patients (19 studies). A few studies focused on specific patient populations, such as COVID-19 patients (2 studies), trauma patients (1 study), and kidney injury patients (1 study). Details of the variables are shown in Supplementary Table 1.

3.2.1. Measurement of disease severity

The two most frequently employed measures of illness severity were the APACHE II score, which was utilized in 19 studies, and the SOFA score, which was used in 9 studies. Additionally, a small number of studies have incorporated alternative scores, such as the Charlson Comorbidity Index, and have predicted mortality risk.

3.2.2. Assessment tools to measure nutritional status in the ICU

The subjective global assessment (SGA) was the most frequently used tool (16 studies). Other tools included the NRS-2002 (5 studies), the mNUTRIC score (7 studies), and various anthropometric and laboratory methods.

The Subjective Global Assessment (SGA) was prioritized due to its validation and acceptance as a method of nutritional assessment in critically ill patients and its use being the most frequently reported method in the included studies, thus allowing for the greatest comparability of the pooled estimates.

3.2.3. Outcome measures

The primary outcome measures evaluated in this meta-analysis were the nutritional status of ICU patients and the secondary outcome length of stay in the intensive care unit (ICU) and mortality rates within the ICU.

3.3. Quality of the articles

The assessment of the study quality via the NOS indicates that most of the studies in this meta-analysis, which assess nutritional risk in critically ill patients, exhibited good to excellent quality in their design and reporting. The majority of studies effectively addressed selection bias, comparability of cohorts, and the assessment of outcomes, making their findings more reliable and applicable to clinical practice. However, some studies have shown room for improvement, particularly in terms of the clarity and reliability of outcome measures. Overall, the findings provide a strong foundation for further understanding the impact of nutritional risk on clinical outcomes in critically ill patients, with implications for future research and clinical interventions.

The total scores for the studies ranged from 6 to 8. Table 1 shows the detailed information of the studies.

Table 1.

The Newcastle‒Ottawa Scale (NOS) quality assessment of the included studies.

Study	Selection (0-4)	Comparability (0-2)	Outcome (0-3)	Total Score (0-9)
Charles Chin Han Lew et al., 2017	4	2	2	8
Sungurtekin et al., 2014	3	2	2	7
Johane P. Allard et al., 2014	4	2	2	8
Daniel Fontes et al., 2013	3	2	2	7
ChandrashishChakravarty et al., 2013	3	2	1	6
Andrés Luciano NicolásMartinuzzi et al., 2021	4	2	2	8
Patricia M. Sheean et al., 2013	3	2	2	7
Bernadette Chimera-Khombe et al., 2022	3	2	1	6
F S Caporossi et al., 2012	3	2	2	7
SavitaBector et al., 2015	3	2	2	7
Anne Coltman et al., 2015	3	2	2	7
NajmehHejazi et al., 2016	3	2	1	6
ShaahinShahbazi et al., 2021	3	2	2	7
Suzie Ferrie et al., 2022	3	2	2	7
AdityaRameshbabuDevalla et al., 2020	3	2	1	6
SornwichateRattanachaiwong et al., 2020	3	2	1	6
ParsaMohammadi et al., 2022	3	2	1	6
H. G. Valente da Silva et al., 2012	3	2	1	6
RanimKaddoura et al., 2020	3	2	1	6
Smith et al., 2015	3	2	2	7
Arabi YM et al., 2015	3	2	2	7
Michael P. Casaer et al., 2011	4	2	2	8
Thain'aGattermann Pereira et al., 2018	3	2	1	6
LS Chapple et al., 2017	3	2	2	7
Kris M. Mogensen et al., 2015	4	2	2	8
GuilhermeDupratCeniccola et al., 2020	3	2	1	6
Rosa Mendes, 2019	3	2	1	6
Na Wang et al., 2023	3	2	2	7
Manon CH de Vries, 2018	3	2	1	6
OmidMoradiMoghaddam et al., 2024	3	2	1	6
DenizAvci, 2019	3	2	1	6

3.4. Publication bias

Visual assessment of the funnel plot, in addition to Egger's regression test, indicated the presence of small study effects and publication bias. This might be attributed to the publication bias of studies with high prevalence estimates, especially those with smaller sample sizes. This might have resulted in an inflated pooled prevalence. However, the similarity in effect estimates and the narrow confidence interval for the pooled effect suggest that the overall effect is not significantly altered. These findings should be interpreted with caution, especially for extrapolation of the prevalence estimates for the wider ICU population, as presented in Fig. 2.

Fig. 2.

Funnel plot asymmetry for publication bias.

The funnel plot of malnutrition status in ICU patients suggests potential publication bias or other types of bias, as presented in Fig. 3.

Fig. 3.

Funnel plot asymmetry for publication bias prevalence of malnutrition.

3.5. Nutritional status of ICU patients

In 30 studies, according to a meta-analysis, the pooled figure of the percentage of ICU patients suffering from a form of malnutrution was 32.7% using a random effects model, with a 95% confidence interval of 19.9-45.5% (see Fig. 4). Higher figures than the control trials, in most cases, were generally reported in low- and middle-income countries. Higher figures than in randomized trials, in most cases, were generally reported in observational studies. Although wide confidence intervals exist in these studies, there was virtually negligible variation in their results, which was confirmed by a very low percentage of 0%, denoted by Iˆ2 = 0.0%. Because of the vital significance of this analysis, such figures of the percentage of ICU patients suffering from a form of malnutrution, according to a systematic review, in low-and middle-income countries, tend to be substantially elevated.

Fig. 4.

Forest plot demonstrating the prevalence of all types of malnutrition among ICU patients.

3.6. Stratified malnutrition status

The forest plot (Fig. 5) shows a bar comparison of moderate and severe malnutrition among the examined studies. The x-axis represents the percentage of the study population that experienced malnutrition, and the y-axis represents the name of the initial author and the study year. The plot shows the abundance of proportions for both categories with increased rates of malnutrition toward the right. In the present study, the weighted mean proportions for moderate and severe malnutrition were 36.34% and 14.91%, respectively, indicating considerable variability in the prevalence of malnutrition across populations and methodologies.

Fig. 5.

Forest plot showing the proportions of two nutritional statuses, “moderate or suspected malnutrition” and “severely malnourished”, for each study.

3.7. Length of stay among malnourished patients

Length of stay was analyzed in 24 studies Fig. 6, and a meta-analysis revealed the indisputable impact of malnutrition on ICU stays. Employing the REML method and a random-effects model, minimal variations between studies and no observed heterogeneity were found. Malnourished patients were hospitalized in the intensive care unit for an average of 5.593 days longer, indicating significant effects. A p value of 0.0000, a Z value of 4.10, and a 95% confidence interval of [2.920, 8.265] support this. The confidence intervals and individual effect sizes revealed that smaller standard errors and larger sample sizes were more significant. Interestingly, the homogeneity test revealed no heterogeneity (p value of 0.9993), indicating that patients who are malnourished are more likely to require lengthier stays in the intensive care unit.

Fig. 6.

Forest plot showing the length of stay in malnourished patients.

3.8. Mortality of ICU PATIENTS

Malnourished patients appeared to have a greater risk of death than well-nourished patients did, as shown by the overall mortality odds ratio of 1.446 (95% CI: 0.761–2.130). In particular, those who were malnourished had 44.6% higher risks of death. This increased mortality risk is unlikely to have occurred by accident, according to the statistically significant p value (Prob> |z| = 0.0000). Malnutrition was also associated with a higher point estimate for mortality, with a RR of 1.45; however, the 95% CI crossed unity from 0.76 to 2.13, suggesting a lack of statistical significance for the finding.

The impact size may be constant among the included studies because of the lack of significant heterogeneity (Prob > Q = 0.9995). The confidence in the estimated overall mortality odds ratio is increased by the data’ homogeneity (see Fig. 7).

Fig. 7.

Forest plot of the overall mortality odds ratio of malnourished patients.

3.9. Mortality rate among well-nourished and malnourished ICU patients

As shown in a box plot (Fig. 8) comparing the mortality rate between the malnourished and well-nourished patients, how nutritional status affects patient survival was highlighted. The median figures revealed that patients who met the definition of malnutrition had significantly greater mortality than did those who were well nourished. The mean mortality proportion of the malnourished subgroup was 30.49, which was significantly greater than the mortality proportion for well-nourished patients of 14.89. Interestingly, the values of the IQRs are also given, showing that they are rather close to each other, while a difference in the higher 50% of the data is apparent. For SD, there are several varieties in both groups; however, the “malnourished” group has one rather high outlier above 60%. This graphic representation clearly defines the position that diet has in determining a patient's status as well as the fact that more research and consideration are still needed.

Fig. 8.

Box plots for mortality rates for well-nourished and malnourished adult ICU patients.

4. Discussion

It is important to consider the nutritional status of patients in the intensive care unit (ICU) since it is a key element in the ability to overcome and survive critical illnesses and clinical outcomes. According to the above studies, malnutrition is clearly a more common problem in the ICU than in other hospital wards [36]. This can be attributed to the strenuous physiological and psychological demands of the disease, which thereby activate cytokine hormones and other chemical mediators. These compounds increase the metabolic rate and oxidize substances such as fats and proteins [12,35]. Nutritional status influences the capacity to address crucial circumstances and clinical consequences among patients admitted to the intensive care unit (ICU) [15].

In this meta-analysis the pooled prevalence of malnutrition among critically ill adult ICU patients was 32.7% (95% CI: 19.9–45.5%), indicating that malnutrition is a common and clinically important burden in ICU populations. Our prevalence estimate is broadly compatible with previous systematic reviews, although some differences in magnitude are apparent. A systematic review of adult ICU patients conducted between 2015 and 2019 reported a pooled malnutrition proportion of approximately 51%, with moderate and severe malnutrition accounting for 46% and 20%, respectively, reflecting higher estimates possibly driven by heterogeneity in diagnostic criteria and study populations [17]. Similarly, narrative syntheses and meta-analyses using broader definitions have reported malnutrition prevalence ranging from 38% to 78% in critically ill populations depending on the assessment tool used (e.g., GLIM, SGA), highlighting the variable burden across settings and criteria [18]. Differences in pooled prevalence between our analysis and prior reviews may be attributed to several factors, including variations in nutritional assessment methodologies, thresholds for defining malnutrition, timing of assessment relative to ICU admission, and geographic or socioeconomic context.

The forest plot demonstrates substantial variability in mortality prevalence among malnourished ICU patients, with moderate malnutrition (blue diamonds) clustering tightly around 20-40% and severe malnutrition (orange diamonds) showing wider dispersion up to 60%, reflecting a dose-response relationship where severe cases face heightened mortality risk. The central diamond's pooled estimate (∼30-35%) confirms malnutrition as a consistent adverse prognostic factor, though high heterogeneity (evident from minimal CI overlap and scattered points) underscores diverse assessment tools (SGA vs. mNUTRIC), patient mixes (medical/surgical), and regional differences (e.g., higher rates in developing countries). Compared to prior meta-analyses, our moderate malnutrition mortality aligns with Lew et al.'s 28% but severe exceeds Sheean's 41% cohorts, emphasizing urgent need for standardized screening and targeted nutrition to mitigate this gradient effect in critical care [4,19].

This study also indicated that the subjective global assessment (SGA) was the most frequently used method for the nutritional status assessment of ICU patients for this systematic review and for that conducted by Charles Chin Han Lew et al., 2017 [37]. Additionally, the most frequently employed disease severity assessment tools in most of the studies were the Acute Physiology and Chronic Health Evaluation II (APACHE II) and Sequential Organ Failure Assessment (SOFA) scores. This finding aligns with the results reported by Charles Chin Han Lew et al., where the APACHE II score was the predominant scoring system used to evaluate disease severity in the included studies [6,38,39].

The mean duration of stay in the ICU of the malnourished patients included in the present meta-analysis ranged from 3.0 ± 2.125 to 22.2 ± 19.5 days. The results presented and discussed above are similar to those of Charles Chin Han Lew et al., in 2017 [45], who reported that the malnutrition group had a mean length of ICU stay ranging from 0.7 to 28.5 days. The results of this meta-analysis indicate that the ICU length of stay range is similar; therefore, the patients in both meta-analyses were similar in terms of their malnutrition status and disease severity and required a longer intensive care unit stay [46–48]. the mortality rate among malnourished patients has been reported to vary from as low as 7.4% to as high as 66% according to different authors [48, 50, 51]. In a study conducted by charles chin Hanlew et al. (2016), the mortality rate among ICU patients was slightly lower than that reported in this study and ranged from 11.9% to 23.4%. This inconsistency may be attributed to the different characteristics of the studies included in the meta-analysis. Moreover, differences in results may vary by region and country, where the level of mortality depends on the quality of medical care. The mortality risk ratio's confidence interval includes 1, this non-significant finding should be interpreted cautiously and does not exclude a clinically meaningful effect, particularly given the consistent direction of effect across studies.

5. Conclusion

Malnutrition is common among ICU patients. To assess the nutritional status of ICU patients, this review revealed that the commonly used tool is the subjective global assessment (SGA), while the APACHE II and SOFA scores are used to assess disease severity. Research has also shown that the general incidence of malnutrition among patients admitted to the ICU is quite high, ranging from as low as 13.9% to as high as 85%. For ICU patients, the average length of stay among malnourished patients increased the odds of a longer ICU stay by 5.21 times. Mortality rates have also been shown to differ from those of patients suffering from malnutrition in the ICU, with rates ranging between 7.4% and 66%. These numbers also demonstrate the variability of care among different settings.

6. Recommendations

Early nutritional screening at ICU admission is essential, with a focus on identifying moderate malnutrition, the most prevalent and potentially reversible stage.

Standardized assessment tools (e.g., SGA, GLIM, NRS-2002, mNUTRIC) should be routinely applied to guide risk stratification and nutritional therapy.

Patients with severe malnutrition should be treated as a high-risk group, requiring prompt, intensive nutritional support and close monitoring.

Targeted early interventions may prevent progression from moderate to severe malnutrition, potentially improving ICU outcomes.

Strengthening nutrition-focused ICU guidelines in LMICs may substantially improve patient outcomes.

6.1. Study limitations

Several limitations are observed in interpreting these findings. First, the included studies were methodologically heterogeneous, encompassing cross-sectional, cohort, and randomized controlled designs, which may introduce variability in pooled prevalence estimates despite subgroup analyses. Second, nutritional status was assessed using different tools (e.g., SGA, NRS-2002, mNUTRIC, GLIM), each with distinct diagnostic thresholds, Subjective Global Assessment (SGA) was the primary nutritional assessment tool, variability across alternative assessment instruments may influence outcome estimates and limits direct comparability across studies.

Third, the majority of included studies were observational in nature, limiting causal inference between malnutrition severity and clinical outcomes. Fourth, geographic representation was uneven, with limited data from low-income settings, which may restrict the generalizability of the findings to resource-constrained ICUs. Finally, several studies lacked detailed reporting on nutritional interventions, timing of assessment, and longitudinal changes in nutritional status, precluding assessment of treatment effects and temporal relationships.

CRediT authorship contribution statement

Misrak Weldeyohanes: Writing – review & editing, Writing – original draft, Software, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Zelalem Debebe: Resources, Investigation, Data curation, Conceptualization. Zeweter Abebe: Writing – review & editing, Validation, Supervision, Methodology, Formal analysis, Data curation, Conceptualization.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

All the data generated or analyzed during this study are included in this published article.

Funding

The authors received no specific funding for this work.

Conflict of interests

The authors declare that they have no competing interests.

Appendix A. Supplementary data

The following is/are the supplementary data to this article:

Supplementary Table 1

Characteristics of the included studies [6,10,[20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36],[38], [39]].