In traumatic brain injury (TBI), the appropriate timing and route of feeding, and the efficacy of immune-enhancing formulae have not been well established. We performed this meta-analysis aiming to compare the effects of different nutritional support modalities on clinical outcomes of TBI patients.
We systematically searched Pubmed, Embase, and the Cochrane Library until October, 2012. All randomized controlled trials (RCTs) and non-randomized prospective studies (NPSs) that compared the effects of different routes, timings, or formulae of feeding on outcomes in TBI patients were selected. The primary outcomes included mortality and poor outcome. The secondary outcomes included the length of hospital stay, the length of ventilation days, and the rate of infectious or feeding-related complications.
13 RCTs and 3 NPSs were included. The pooled data demonstrated that, compared with delayed feeding, early feeding was associated with a significant reduction in the rate of mortality (relative risk [RR] = 0.35; 95% CI, 0.24–0.50), poor outcome (RR = 0.70; 95% CI, 0.54–0.91), and infectious complications (RR = 0.77; 95% CI, 0.59–0.99). Compared with enteral nutrition, parenteral nutrition showed a slight trend of reduction in the rate of mortality (RR = 0.61; 95% CI, 0.34–1.09), poor outcome (RR = 0.73; 95% CI, 0.51–1.04), and infectious complications (RR = 0.89; 95% CI, 0.66–1.22), whereas without statistical significances. The immune-enhancing formula was associated with a significant reduction in infection rate compared with the standard formula (RR = 0.54; 95% CI, 0.35–0.82). Small-bowel feeding was found to be with a decreasing rate of pneumonia compared with nasogastric feeding (RR = 0.41; 95% CI, 0.22–0.76).
Citation: Wang X, Dong Y, Han X, Qi X-Q, Huang C-G, Hou L-J (2013) Nutritional Support for Patients Sustaining Traumatic Brain Injury: A Systematic Review and Meta-Analysis of Prospective Studies. PLoS ONE 8(3): e58838. https://doi.org/10.1371/journal.pone.0058838
Editor: Anthony E. Kline, University of Pittsburgh, United States of America
Received: December 7, 2012; Accepted: February 7, 2013; Published: March 19, 2013
Copyright: © 2013 Wang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by Natural Science Foundation of China (81171130/H0908) and Natural Science Foundation from Science and Technology Commission of Shanghai Municipality (11ZR1448700). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Traumatic brain injury (TBI) remains a major worldwide health and socioeconomic problem. It is the most common cause of death and disability in people between 15 and 30 years of age . The weighted average mortality for severe TBI is 39%, and for unfavorable outcome on the Glasgow Outcome Scale (GOS) is 60% .
TBI has a dynamic pathophysiology that evolves in time, consisting of primary injury, followed by a combination of systemic disorders (hypoxia, hypotension, and hypercarbia) and local events, which together lead to secondary injury . As brain is the functional regulator for metabolic activities, a complex milieu of metabolic alterations may occur in TBI, consisting of hormonal changes, aberrant cellular metabolism, and inflammatory cascade . The abnormal metabolic processes, mainly including hypermetabolism, hypercatabolism, and glucose intolerance, have been recognized as incredibly essential elements of secondary injuries –. Not only can they complicate the initial period of hospitalization and stabilization, but also they may negatively impact rehabilitative treatments . Nutritional support, in addition to providing daily calories, has been appreciated as an important adjunctive therapy for metabolic disorders following TBI .
Nutritional support constitutes an important issue in intensive care for critically ill patients. However, it is generally neglected and underestimated in the subgroup of TBI population. In the recent most important trials in nutrition, Casaer et al. only included 0.6% of patients with neurological diseases . Heidegger et al. incorporated 15% of neurological patients, and only included those with functioning gastrointestinal tract . Interestingly, disagreement on the role of early parenteral nutrition even existed between the two trials. Currently, nutritional support for TBI patients, especially the appropriate timing, route, and formula of feeding, has not been well illustrated yet. In the earlier Cochrane Review, a trend towards better outcome with early nutritional support for TBI patients was shown, but without any statistically significant result . In addition to not including updated studies, it had several flaws of its own. Although the Brain Trauma Foundation has recommended achieving full caloric replacement by day 7 following TBI, no agreement has been reached in the optimal timing or route of feeding . In fact, nutritional support is frequently underestimated in the clinical management of severe TBI patients. Given insufficient previous evidence, as well as the introduction of recent randomized controlled trials (RCTs), we perform this meta-analysis and systematic review, aiming to compare the effects of different timings and routes of feeding, and to explore the effect of immune-enhancing formula on outcomes in TBI patients.
The overview of this meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement (checklist S1) . A computerized bibliographic search for all relevant articles from 1980 to October 2012 was performed by Ovid Medline, EMBASE, and the Cochran database. We used the following search core terms: “traumatic brain injury”, “craniocerebral injury”, “head injury”, “head trauma”, “enteral nutrition”, “enteral tube feeding”, “parenteral nutrition”, “gastrointestinal intubation”, “nutritional support”, “prospective cohort”, and “randomized controlled trial”. The language was limited to English. We also manually searched the references of selective papers to identify additional potentially eligible studies.
Studies were selected into the meta-analysis if they: i.) were RCTs or non-randomized prospective studies (NPSs); ii.) investigated TBI patients; iii.) compared the effect of different feeding routes (enteral nutrition [EN] vs. parenteral nutrition [PN], or nasogastric enteral feeding vs. non-nasogastric enteral feeding), different feeding timings (early or delayed), or different immunonutritional elements (such as probiotics, arginine, glutamine, nucleotides, and ω-3 fatty acids), on the outcome variables; iv.) reported the number of outcome events in different interventions.
Two assessors (XW and YD) independently reviewed the full manuscripts of eligible studies. Data were extracted in standardized data-collection forms. Extracted data included first author’s name; year of publication; sample size; patients’ characteristics (mean age, gender); starting time of feeding; treatment arms; outcome variables and the score of quality assessment. Any discrepancy was resolved by discussion or a third author (CGH). Selected RCTs were critically appraised using the Jadad scale, which scores studies’ description of randomization (2 points), blinding (2 points) and attrition information (1 point) . The Newcastle–Ottawa scale (NOS) was used to evaluate the methodological quality of prospective cohort studies, as recommended by the Cochrane Non-Randomized Studies Methods Working Group . The quality of a study was judged on the selection of the study groups, the comparability of the groups, and the ascertainment of the outcome of interest.
Primary outcomes of clinical importance included mortality and functional outcome on GOS score. Secondary outcomes include the length of stay (LOS) in hospital or ICU, and major complications. Infectious complications and feeding-related complications were assessed, respectively. We defined infectious complications as pneumonia (ventilator or non-ventilator-associated pneumonia and other lower respiratory tract infections), central nervous system (CNS) infection, bloodstream infection (laboratory-confirmed-bloodstream infections and clinical sepsis), or urinary tract infection. Feeding-related complications include feeding intolerance, aspiration, diarrhea, constipation and vomiting, and abdominal distention.
Data relating to outcomes were combined from pertinent studies. We used risk ratios (RR) and the associated 95% confidence intervals (CIs) to pool binary outcomes, including mortality, poor outcome and complications. Mean differences (MDs) with 95% CIs were used for continuous outcomes, which included the LOS in hospital or ICU, and the length of ventilator days. Review Manager 5.1.7 (Cochrane Collaboration, 2012) was used to process the meta-analysis. The Mantel-Haenszel method was used to test the significance of treatment effect, and the random-effects model was used to estimate the overall RRs. Heterogeneity of treatment effects between studies was statistically explored by the I2 statistic. I2 statistic of 0%–40% indicates unimportant heterogeneity, 30%–60% indicates moderate heterogeneity, 50%–90% indicates substantial heterogeneity, and 75%–100% indicates considerable heterogeneity . Besides, we performed subgroup analyses based on the following factors which may contribute to the heterogeneity: study design, sample size, publication year, staring time of early nutrition, and different route of feeding. The sensitivity analyses were carried out by excluding studies one by one, or by employing a fixed effect model. All reported P values were two-sides, and P values less than 0.05 were deemed as statistically significant. The publication bias was examined visually by inspecting the funnel plots on Review Manager 5.1.7, and statistically by using the Egger’s regression model, calculated by Stata 12.0 (Stata Corporation, College Station, TX, USA).
709 articles were found in total from the initial search, of which 34 eligible articles were selected after screening of titles and abstracts. Further, 10 studies were excluded, with 5 of non-English language, 1 comparing combined EN and PN with PN , 1 comparing two different fat emulsions , 1 comparing essential amino acid with placebo , 1 comparing intermittent EN with continuous EN , and 1 compared different infusion speed of EN . In the remaining 24 studies included in qualitative synthesis, 8 articles lacked sufficient data relating to our outcomes –. Thus 16 studies were pooled into the meta-analysis, including 13 randomized controlled trials –, and 3 NPSs –. The search flow diagram was shown in Figure 1 and protocol S1. The characteristics of these studies were shown in Table 1. Of the 13 RCTs, the mean Jadad score was 2, ranging from 0 to 5. All NPSs have a high NOS score of 8.
Early VS Delayed
8 studies were available for the comparison of early nutrition with delayed nutrition, including 5 RCTs , , , , , and 3 NPSs –. Research conducted by Grahm et al. compared early jejunal feeding with delayed gastric feeding without data of mortality or functional outcome, and thus was excluded from the pooled analyses . The pooling data of the other seven studies indicated that early nutrition was associated with a significant reduction of the mortality rate compared with delayed nutrition, but with moderate heterogeneity (RR = 0.35; 95% CI, 0.24–0.50; P<0.05; I2 = 44%) (Figure 2A).
(A) Forest plot illustrates the different effects on mortality. (B) Forest plot shows the different effects on poor outcome. PO, poor outcome.
The heterogeneity was explored by subgroup analyses according to the sample size, publication year, study design, starting time of early feeding, and the route of feeding. Most studies have a small sample size below 100 except for two studies (Table 1). Five studies initiated the early nutrition within 72 hours after injury , , , , . In two NPSs, mortality was compared in patients who started feeding within/out of the first 7 days postinjury. Three studies compared the early enteral nutrition with delayed enteral nutrition , , . Two trials compared early PN with delayed EN , . The results of subgroup analyses were shown in Table 2. Notably, the RCT subgroup, the subgroup of publication year earlier than 2005, and the subgroup that compared early PN with delayed EN were not found to be with statistical significances (RR = 0.62 with 95% CI 0.32–1.22, RR = 0.54 with 95% CI 0.23–1.27, and RR = 0.57 with 95% CI 0.19–1.76, respectively), demonstrating that these subset factors might be the sources of heterogeneity. Meta-regression was not further performed due to the limited number of available studies.
Four studies reported the functional outcome on GOS. There was a significant lower risk of poor outcome in patients who received early nutrition compared with those who received delayed nutrition (RR = 0.70; 95% CI, 0.54–0.91; P<0.05) (Figure 2B). Four studies investigated the occurrence of infectious complications in both groups, in terms of pneumonia, CNS infection, bloodstream infection, and urinary tract infection. Generally, early nutrition was significantly associated with a lower risk of infectious complications compared with delayed nutrition (RR = 0.77; 95% CI, 0.59–0.99; P<0.05). In the subgroup analyses, however, no similar statistical significance was observed (Figure 3). Three studies reported the rate of feeding complications, including diarrhea and feeding intolerance. However, no statistical significance was revealed (figure S1). Two studies reported the mean value and standard deviation (SD) of length of stay (LOS) in intensive care unit (ICU) , , whereas no significant difference was observed between early and delayed groups (P = 0.68) (figure S2).
The funnel plots of data relating to mortality were found to be symmetrical, suggesting a low likelihood of having publication bias (Figure 4A). No publication bias was revealed by Egger test either (P = 0.239).
EN VS PN
Five RCTs involving a total of 215 patients evaluated delivery route of nutrition support (EN vs PN) in TBI patients (Table 1). As it showed before, two trials compared early PN with delayed EN , . In the other three trials, both PN and EN were started early after admission , , .
Aggregating data of the five studies demonstrated a trend toward a lower mortality rate associated with PN (RR = 0.61; 95% CI, 0.34–1.09; I2 = 8). Nevertheless, statistical significance was not revealed (P = 0.09) (Figure 5A). In order to investigate the impact of different starting time on the results, the subgroup analysis was performed. However, no statistically significant result was revealed in any subgroup (P>0.05) (Table 3). It is similar when subanalyzing the impact of publication year (Table 3). The low heterogeneity may justify the application of a fixed-effect model. In sensitivity analyses by using this model, a marginal statistical significance was revealed (RR = 0.56; 95% CI, 0.32–0.99; P = 0.05). Further, only by excluding the study by Young et al. , a statistical significance without heterogeneity was demonstrated (RR = 0.35; 95% CI 0.15–0.83; P = 0.02; I2 = 0).
(A) Forest plot illustrates the different effect on mortality. (B) Forest plot shows the different effect on poor outcome. PO, poor outcome.
Three of the five studies reported the functional outcome on GOS. Pooling data revealed a trend toward reducing the rate of poor outcome in PN groups (RR = 0.73; 95% CI, 0.51–1.04). Nevertheless, statistical significance was not observed in the overall analysis or in any subgroup analysis according to the timing of nutrition (P>0.05) (Figure 5B). The fixed effect model also failed to show any significant alteration. The reported infectious complications mainly included pneumonia, central nervous system (CNS) infection, bloodstream infection, and urinary tract infection. Pooling data suggested that PN patients may have a slight trend of lower rate of infection complications compared with EN patients (RR 0.89; 95% CI, 0.66–1.22), especially in reducing the occurrence of pneumonia, whereas statistical significance was not revealed (P = 0.48) (Figure 6). The heterogeneity was at an unimportant level across our pooled analyses.
The funnel plots of data relating to mortality were found to be symmetrical, suggesting a low likelihood of having publication bias (Figure 4B). No publication bias was revealed by Egger test either (P = 0.621).
Standard Formula VS Immune-enhancing Formula
Three trials compared the immune-enhancing formula (arginine, glutamine, probiotics, and ω-3 fatty acids et al.) with the standard formula of EN in TBI patients , , . Infection rate was reported in all trials, and the aggregating data revealed that immune-modulating formula was associated with a statistical significant reduction in infection rate in contrast with the standard formula (RR = 0.54; 95% CI, 0.35–0.82; P<0.05) (figure S3).
Non-nasogastric (NNG) Feeding VS NG Feeding
Five trials compared NNG feeding with NG feeding in EN support. Taylor et al. compared a mixed group (intestinal or gastric) with a standard gastric group, and thus was excluded . Minard et al. compared early nasoenteric feeding with delayed NG feeding . Grahm et al. compared nasojejunal feeding with NG feeding . Kostadima et al. compared percutaneous endoscopic gastrostomy with standard EN in the occurrence of ventilator-associated pneumonia . Escribano et al. randomized patients to receive transpyloric feeding or gastric feeding . The pooling data of 4 trials showed a significant reduction in the occurrence of pneumonia in patients receiving NNG feeding compared with those receiving NG feeding (RR = 0.62; 95% CI, 0.40–0.96; P = 0.03). The NNG group was associated with a trend to reduce mortality rate and shorten ventilator day, whereas failing to show statistical significances. Three trials additionally reported the LOS in ICU, and the aggregated results revealed no statistical significance in the two arms (figure S4).
We carried out a comprehensive literature search to detect prospective studies, including RCTs and NPSs, to compare the effects of different routes, timings and formulae of nutritional support on clinical outcomes in TBI patients. Explicit criteria were utilized for study selection and methodological quality assessment.
In comparison of different timing for nutritional support, our meta-analysis demonstrated beneficial effects of early nutrition on reducing mortality, improving functional outcome, and decreasing infectious complications. Although a trend was indicated for early nutrition in lowering the risk of stratified specific infections, no statistical significance was revealed. There was also no significant difference in feeding-related complications between the two arms. Notably, opposite to pooled data of NPSs, the data of RCTs failed to demonstrate statistical significance in mortality. The RCTs were largely carried out in early years with smaller sample sizes and limited nutritional support approaches, which might contribute to this discrepancy. Similar concern may also explain the heterogeneity from publication year. The earlier Cochrane review also suggested a trend toward using early nutrition to reduce mortality and improve functional outcome by analyzing fewer studies, whereas without statistical significance . Furthermore, several queries were concerned. For study by Hadley et al., the PN and EN were both started within 48 hours after admission, and thus it may not be reliable for considering it as the comparison of early nutrition with delayed nutrition , . Additionally, they included one study that compared different infusion speeds of EN . In light of the simultaneous initiation of EN in both groups, it is not likely to be justified to consider this trial as the comparison of nutritional timings, and thus we excluded it. The Brain Trauma Foundation has cautiously recommended achieving full caloric replacement by day 7 following TBI based on limited evidence . Taken together, our results reinforced the inclination of early nutrition for TBI patients.
When compared with EN, our results showed that there might be a trend toward using PN to reduce mortality and improve functional outcome, whereas statistical significance was only marginally revealed when the fixed effect model was used. In the subgroup analyses based on timing of feeding and publication year, no statistical significant findings were revealed. Moreover, significant difference in the rate of complications was not recognized in any subgroup analysis between the two arms. In fact, EN and PN show unique advantages, respectively. The use of EN is superior to PN in patients with functioning gastrointestinal tracts . Compared with PN, EN formulae may conveniently make use of more effective substrates to support cell and organ function, have a lower risk of hyperglycemia or hyperosmolarity, be administered at lower rates to avoid overfeeding, and better support the gut mass and barrier function. However, the use of enteral feeding in patients with gastrointestinal intolerance is associated with underfeeding and consequent malnutrition ,. Less than 70% of patients receive an adequate enteral caloric intake even in the most experienced and motivated ICUs . In comparison, PN patients have benefits in obtaining more dependable nutrient bioavailability, getting nutrition effects in a shorter period, requiring no functional GT tract, and staying away from satiety, abdominal distention or other enteral feeding complications . However, overfeeding (the administration of excess dextrose, fat, or calories) and refeeding syndrome (rapid feeding of patients with preexisting malnutrition) may occur, and thus induce a variety of metabolic complications, including hyperglycemia, hypertriglyceridemia, thiamine deficiency, hypervolemic, and hypercapnia . In our pooled studies, enrolled patients unanimously experienced moderate or severe TBI, mostly with GCS lower than eight –, , . They are always comatose, intubated or mechanically ventilated, with malfunctioning parasympathetic and sympathetic system, disturbed hypothalamic-pituitary axis, elevated intracranial pressure (ICP), increasing endogenous opioids and endorphins, and widespread prescription of narcotics. All of these unfavorable factors may contribute to impaired GT function, delayed gastric emptying, and increased risk of EN intolerance . In this context, PN might be superior to EN for initial life saving nutritional support. Notably, PN was unanimously initiated early after admission in related studies. Thus, our result should not be misunderstood as opposition to the suggestion that EN is preferable whenever possible with functional gastrointestinal tracts . However, data of feeding related complications, especially the data of hyperglycemia were insufficient across included studies, and further persuasive evidence is warranted.
Given the prevalence of inconveniences for routine nasogastric EN, other alternative EN routes have been attempted, including nasojejunal feeding, percutaneous endoscopic gastrotomy feeding, and transpyloric feeding , , . Pneumonia rate was shown to be significantly reduced by NNG, which may be associated with the prevention of aspiration by NNG feeding. There was also a trend toward reducing mortality rate and decreasing ventilator days, but failed to demonstrate statistical significances. It has been evaluated that at least 20% of TBI patients don’t tolerate enteral alimentation at all in the first week . By using NNG route, patients will probably tolerate enteral feeding as well as avoid the hyperalimentation brought about by PN. Although data on feeding complications were scant, in light of its well tolerance and prevention of pneumonia, we are inclined to side with ESPEN guidelines, which suggested that when jejunal feeding can be carried out easily, it should be given .
Furthermore, our results showed that immune-enhanced formulae were associated with a significant reduction in infectious complications compared with standard formulae. Although a growing number of studies emphasized the importance of nutrition content of foods on post-injury recovery, studies relating to TBI are scant, especially according to our criteria of study design. In fact, the effect of immune-enhanced formulae has been widely investigated in general population, but with confusing and undefined conclusions for critically ill patients . The ESPEN guideline has recommended that immune-modulating formulae, enriched with arginine, nucleotides, and ω-3 fatty acids, are superior to standard enteral formulae in trauma patients . In contrast, the updated guidelines from the Canadian Practice Group in 2009 and the American Dietetic Association evidence analysis library did not recommend the routine use of immune-modulating diets in critically ill patients . The most prominent controversy was the effect of immune-modulating diet on mortality and functional outcomes. For example, the use of arginine-contained formulae has shown greater mortality, and it is hypothesized that arginine may be converted to nitric oxide and thus contributed to hemodynamic instability . Our results could only suggest the benefits of immune-modulating diet for TBI patients in reducing infectious complications. The effects on mortality or functional outcome could not be elucidated by the few studies, and further studies are warranted to investigate the effects of particular diets on the outcome of TBI.
We are aware of the limitations of this meta-analysis. Trials with statistically significant results may be more likely to be published and cited, and are preferentially published in English language journals . We included only studies written in English language and therefore, may have missed relevant studies published in non-English language journals. Besides, several studies (e.g. Rapp et al. and Young et al.) are included in 2 or more of the reported analyses; this has a potential risk of overemphasizing positive results. Although results from Egger’s tests and funnel plots did not show evidence of publication bias, their capacities to detect bias was limited by the small number of studies . The studies included were of relatively poor quality, with most of RCTs having a Jadad score of <3 (13/15). Only a few trials described the method of randomization , . Blinded design was only described in two trials on immune-enhanced formulae , . Although it seems difficult to conceal the route of nutritional support, studies with inadequate or unclear concealment of allocation may overestimate the intervention effect. The sample sizes were relatively small across included studies, especially in the RCTs. Small sample size might contribute to the failure of randomization and imbalance between clinical variables, and thus failed to detect the statistically significant effects. In fact, the compared different arms were not well-controlled. For example, the comparison of route for nutritional support would be more convinced if started simultaneously. However, it is commonly seen that EN was initiated late until the recovery of gastroparesis , . In fact, in subgroup analyses of timing of feeding, two studies that compared early PN with late EN were both revealed to be the potential source of heterogeneity. Question may be raised that whether the effects found were related to the parenteral nutrition or perhaps more due to the early onset of nutrition versusing delayed onset. The impossibility to differentiate here was a substantial confounding factor in the interpretation of results. In the sensitivity analyses of route of feeding, the exclusion of study by Young et al. has led to a significant change of results. Notably, only this study initiated the EN support until the termination of low wall suction, which was non-conventionally performed and might contribute to the heterogeneity . Furthermore, our meta-analysis was absent of aggregating various nutrition indexes, such as caloric intake, nitrogen intake, and nitrogen balance. In previous systematic reviews, it has been revealed that the measurement methods, definitions of metabolic abnormalities, and energy expenditure following TBI varied greatly, which may restrict the incorporation , . Additionally, the included studies utilized different criteria for inclusion and exclusion. Especially, the differences in severity of disease across studies may explain some of the heterogeneity. For example, a number of studies specifically investigated patients with mechanical ventilation –, –. All of these limitations restrict the strength of conclusions drawn from our meta-analysis.
Though disputes of optimal nutritional support would continue, we postulate that the optimal clinical decision in nutritional support should be personalized, in terms of the individual profile, including nutritional status, severity, complications, feeding tolerance, and day-to-day changes in clinical conditions. Last but not least, greater multidisciplinary efforts from nutritionists and clinicians are required for better management of the nutritional support for TBI patients.
Our meta-analysis lends support to early initiation of nutritional support for TBI patients, which can decrease mortality, reduce complications and facilitate recovery. PN appears to be superior to EN in reducing mortality and improving outcome in the acute gut-intolerant phase of TBI. Immune-modulating formulae seem to be superior to standard formulae in reducing infectious complications. Small-bowel feeding was recommended if possible. However, our results should be interpreted with caution given the various limitations. Further well-designed RCTs are expected to clarify the optimal nutritional strategies for TBI patients.
Forest plot shows the effect of early nutrition and delayed nutrition on feeding compliations. (A) Forest plot illustrates the effect on diarrhea. (B) Forest plot illustrates the effect on feeding intolerance. I, intolerance.
Forest plot shows the effect of early nutrition and delayed nutrition on length of stay in the intensive care unit.
Forest plot shows the effect of standard and immuno-modulated nutritional formulae on infectious complications.
Forest plot shows the effect of non-nasogastric and nasogastric enteral feeding on outcomes in patients with TBI. (A) Forest plot illustrates the effect on pneumonia. (B) Forest plot shows the effect on mortality. (C) Forest plot shows the effect on ventilator days. (D) Forest plot shows the effect on length of stay in the intensive care unit. Pneumo, pneumonia.
Conceived and designed the experiments: XW YD CGH LJH. Performed the experiments: XW YD XH XQQ. Analyzed the data: XW YD XQQ XH. Contributed reagents/materials/analysis tools: XH CGH LJH. Wrote the paper: XW YD CGH LJH.
- 1. Giacino JT, Whyte J, Bagiella E, Kalmar K, Childs N, et al. (2012) Placebo-controlled trial of amantadine for severe traumatic brain injury. N Engl J Med 366: 819–826.
- 2. Rosenfeld JV, Maas AI, Bragge P, Morganti-Kossmann MC, Manley GT, et al. (2012) Early management of severe traumatic brain injury. Lancet 380: 1088–1098.
- 3. Cook AM, Peppard A, Magnuson B (2008) Nutrition considerations in traumatic brain injury. Nutr Clin Pract 23: 608–620.
- 4. Vizzini A, Aranda-Michel J (2011) Nutritional support in head injury. Nutrition 27: 129–132.
- 5. Gadisseux P, Ward JD, Young HF, Becker DP (1984) Nutrition and the neurosurgical patient. J Neurosurg 60: 219–232.
- 6. Bistrian BR, Askew W, Erdman JW, Oria MP (2011) Nutrition and traumatic brain injury: a perspective from the Institute of Medicine report. JPEN J Parenter Enteral Nutr 35: 556–559.
- 7. Casaer MP, Mesotten D, Hermans G, Wouters PJ, Schetz M, et al. (2011) Early versus late parenteral nutrition in critically ill adults. N Engl J Med 365: 506–517.
- 8. Heidegger CP, Berger MM, Graf S, Zingg W, Darmon P, et al. (2013) Optimisation of energy provision with supplemental parenteral nutrition in critically ill patients: a randomised controlled clinical trial. Lancet 381: 385–393.
- 9. Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, et al. (2007) Guidelines for the management of severe traumatic brain injury. XII. Nutrition. J Neurotrauma 24 Suppl 1S77–82.
- 10. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 6: e1000097.
- 11. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, et al. (1996) Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 17: 1–12.
- 12. Wells GA, Shea B O, Connell D, Peterson JVW (2009) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available: http://wwwohrica/programs/clinical epidemiology/oxfordhtm Accessed 9 May 2010.
- 13. Higgins JPT, Green S, Collaboration C (2008) Cochrane handbook for systematic reviews of interventions: Wiley Online Library.
- 14. Hausmann D, Mosebach KO, Caspari R, Rommelsheim K (1985) Combined enteral-parenteral nutrition versus total parenteral nutrition in brain-injured patients. A comparative study. Intensive Care Med 11: 80–84.
- 15. Calon B, Pottecher T, Frey A, Ravanello J, Otteni JC, et al. (1990) Long-chain versus medium and long-chain triglyceride-based fat emulsion in parental nutrition of severe head trauma patients. Infusionstherapie 17: 246–248.
- 16. Boselli M, Aquilani R, Baiardi P, Dioguardi FS, Guarnaschelli C, et al. (2012) Supplementation of essential amino acids may reduce the occurrence of infections in rehabilitation patients with brain injury. Nutrition in Clinical Practice 27: 99–113.
- 17. Maurya I, Pawar M, Garg R, Kaur M, Sood R (2011) Comparison of respiratory quotient and resting energy expenditure in two regimens of enteral feeding - Continuous vs. intermittent in head-injured critically ill patients. Saudi Journal of Anaesthesia 5: 195–201.
- 18. Taylor SJ, Fettes SB, Jewkes C, Nelson RJ (1999) Prospective, randomized, controlled trial to determine the effect of early enhanced enteral nutrition on clinical outcome in mechanically ventilated patients suffering head injury. Crit Care Med 27: 2525–2531.
- 19. Clifton GL, Robertson CS, Contant CF (1985) Enteral hyperalimentation in head injury. Journal of Neurosurgery 62: 186–193.
- 20. Twyman D, Young AB, Ott L, Norton JA, Bivins BA (1985) High protein enteral feedings: a means of achieving positive nitrogen balance in head injured patients. JPEN J Parenter Enteral Nutr 9: 679–684.
- 21. Sacks GS, Brown RO, Teague D, Dickerson RN, Tolley EA, et al. (1995) Early nutrition support modifies immune function in patients sustaining severe head injury. JPEN J Parenter Enteral Nutr 19: 387–392.
- 22. Suchner U, Senftleben U, Eckart T, Scholz MR, Beck K, et al. (1996) Enteral versus parenteral nutrition: effects on gastrointestinal function and metabolism. Nutrition 12: 13–22.
- 23. Berg A, Bellander BM, Wanecek M, Norberg A, Ungerstedt U, et al. (2008) The pattern of amino acid exchange across the brain is unaffected by intravenous glutamine supplementation in head trauma patients. Clin Nutr 27: 816–821.
- 24. Nataloni S, Gentili P, Marini B, Guidi A, Marconi P, et al. (1999) Nutritional assessment in head injured patients through the study of rapid turnover visceral proteins. Clin Nutr 18: 247–251.
- 25. Young B, Ott L, Haack D, Twyman D, Combs D, et al. (1987) Effect of total parenteral nutrition upon intracranial pressure in severe head injury. J Neurosurg 67: 76–80.
- 26. Rapp RP, Young B, Twyman D, Bivins BA, Haack D, et al. (1983) The favorable effect of early parenteral feeding on survival in head-injured patients. J Neurosurg 58: 906–912.
- 27. Hadley MN, Grahm TW, Harrington T, Schiller WR, McDermott MK, et al. (1986) Nutritional support and neurotrauma: a critical review of early nutrition in forty-five acute head injury patients. Neurosurgery 19: 367–373.
- 28. Young B, Ott L, Twyman D, Norton J, Rapp R, et al. (1987) The effect of nutritional support on outcome from severe head injury. J Neurosurg 67: 668–676.
- 29. Grahm TW, Zadrozny DB, Harrington T (1989) The benefits of early jejunal hyperalimentation in the head-injured patient. Neurosurgery 25: 729–735.
- 30. Borzotta AP, Pennings J, Papasadero B, Paxton J, Mardesic S, et al. (1994) Enteral versus parenteral nutrition after severe closed head injury. J Trauma 37: 459–468.
- 31. Minard G, Kudsk KA, Melton S, Patton JH, Tolley EA (2000) Early versus delayed feeding with an immune-enhancing diet in patients with severe head injuries. JPEN J Parenter Enteral Nutr 24: 145–149.
- 32. Falcao de Arruda IS, de Aguilar-Nascimento JE (2004) Benefits of early enteral nutrition with glutamine and probiotics in brain injury patients. Clin Sci (Lond) 106: 287–292.
- 33. Kostadima E, Kaditis AG, Alexopoulos EI, Zakynthinos E, Sfyras D (2005) Early gastrostomy reduces the rate of ventilator-associated pneumonia in stroke or head injury patients. Eur Respir J 26: 106–111.
- 34. Briassoulis G, Filippou O, Kanariou M, Papassotiriou I, Hatzis T (2006) Temporal nutritional and inflammatory changes in children with severe head injury fed a regular or an immune-enhancing diet: A randomized, controlled trial. Pediatr Crit Care Med 7: 56–62.
- 35. Khorana J, Rerkasem K, Apichartpiyakul C, Sakonwasun C, Watcharasakslip W, et al. (2009) Immunonutrition and cytokine response in patients with head injury. J Med Assoc Thai 92: 188–194.
- 36. Acosta-Escribano J, Fernandez-Vivas M, Grau Carmona T, Caturla-Such J, Garcia-Martinez M, et al. (2010) Gastric versus transpyloric feeding in severe traumatic brain injury: a prospective, randomized trial. Intensive Care Med 36: 1532–1539.
- 37. Chourdakis M, Kraus MM, Tzellos T, Sardeli C, Peftoulidou M, et al. (2012) Effect of early compared with delayed enteral nutrition on endocrine function in patients with traumatic brain injury: an open-labeled randomized trial. JPEN J Parenter Enteral Nutr 36: 108–116.
- 38. Justo Meirelles CM, de Aguilar-Nascimento JE (2011) Enteral or parenteral nutrition in traumatic brain injury: a prospective randomised trial. Nutr Hosp 26: 1120–1124.
- 39. Hartl R, Gerber LM, Ni Q, Ghajar J (2008) Effect of early nutrition on deaths due to severe traumatic brain injury. J Neurosurg 109: 50–56.
- 40. Chiang YH, Chao DP, Chu SF, Lin HW, Huang SY, et al. (2012) Early enteral nutrition and clinical outcomes of severe traumatic brain injury patients in acute stage: a multi-center cohort study. J Neurotrauma 29: 75–80.
- 41. Dhandapani S, Dhandapani M, Agarwal M, Chutani AM, Subbiah V, et al. (2012) The prognostic significance of the timing of total enteral feeding in traumatic brain injury. Surg Neurol Int 3: 31.
- 42. Perel P, Yanagawa T, Bunn F, Roberts I, Wentz R, et al.. (2006) Nutritional support for head-injured patients. Cochrane Database Syst Rev: CD001530.
- 43. Griffiths RD (2006) Parenteral nutrition in adults with a functional gastrointestinal tract. Lancet 368: 1868.
- 44. Ziegler TR (2009) Parenteral nutrition in the critically ill patient. N Engl J Med 361: 1088–1097.
- 45. Vincent JL, Preiser JC (2012) When should we add parenteral to enteral nutrition? Lancet 381: 354–355.
- 46. Zaloga GP (2006) Parenteral nutrition in adult inpatients with functioning gastrointestinal tracts: assessment of outcomes. Lancet 367: 1101–1111.
- 47. Kreymann KG, Berger MM, Deutz NE, Hiesmayr M, Jolliet P, et al. (2006) ESPEN Guidelines on Enteral Nutrition: Intensive care. Clin Nutr 25: 210–223.
- 48. Worthington ML, Cresci G (2011) Immune-modulating formulas: who wins the meta-analysis race? Nutr Clin Pract 26: 650–655.
- 49. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629–634.
- 50. Krakau K, Omne-Ponten M, Karlsson T, Borg J (2006) Metabolism and nutrition in patients with moderate and severe traumatic brain injury: A systematic review. Brain Inj 20: 345–367.
- 51. Foley N, Marshall S, Pikul J, Salter K, Teasell R (2008) Hypermetabolism following moderate to severe traumatic acute brain injury: a systematic review. J Neurotrauma 25: 1415–1431.