The effect of perioperative oral carbohydrate loading on postoperative outcomes in patients undergoing orthopedic surgeries: a systematic review and meta-analysis

Abstract

Introduction

Oral carbohydrate loading is a core element used in Enhanced Recovery after Surgery (ERAS) protocols and has shown benefits in several surgical fields. However, evidence in orthopedic surgery is mixed. This systematic review and meta-analysis of randomized controlled trials aims to evaluate the impact of preoperative oral carbohydrate loading on postoperative outcomes in orthopedic surgery patients.

Method

Studies comparing preoperative oral carbohydrate loading versus fasting in orthopedic surgery measuring outcomes of hospital stay, pain, nausea, vomiting, or thirst were searched in PubMed, Cochrane, and Scopus database from 2013 up to July 2025. The review was registered in Prospero and follows PRISMA guidelines. Meta-analysis was performed using Review Manager 5.4 and R statistical software with a random effects model, and heterogeneity was assessed using the I² statistic. Risk of bias was evaluated using Cochrane ROB2.

Results

A total of 10 randomized controlled trials were included in this meta-analysis. A significant reduction was observed in postoperative thirst among patients receiving CHO compared with fasting (MD = − 2.58, 95% CI − 4.10 to − 1.05; P = 0.0009) among 3 studies (n = 190). Pooled analysis of six RCTs (n = 472) showed that preoperative oral carbohydrate loading also led to significant reduction in hospital stay compared with fasting (MD = − 0.36 days, 95% CI − 0.70 to − 0.02; P = 0.0391). For postoperative vomiting, data from four studies (n = 268) indicated no significant difference between groups (RR = 0.71, 95% CI of 0.12 to 4.05; P = 0.6960), similarly, pooled analysis of four studies (n = 268) assessing postoperative nausea showed no significant effect (RR = 0.72, 95% CI 0.30 to 1.70; P = 0.4506). For postoperative pain, seven RCTs (n = 512) demonstrated a non-significant overall reduction in pain scores (SMD = − 0.31, 95% CI − 0.65 to 0.04; p = 0.0797) and the subgroup analysis revealed a significant benefit in studies using spinal anesthesia (SMD = − 0.05, 95% CI: − 0.29 to 0.19, P = 0.007).

Conclusion

Preoperative oral carbohydrate loading consistently reduced postoperative thirst in patients undergoing orthopedic surgery. It also was associated with a modest reduction in length of hospital stay; however, this effect was may not be consistently observed across all clinical settings. Although trends toward reduced pain, nausea and vomiting were observed, these did not reach statistical significance. The overall certainty of evidence is limited by moderate to substantial heterogeneity across studies. Despite limitations, the intervention appeared safe, adheres to ERAS standards, and necessitates further exploration through multicenter trials especially in high risk groups.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-026-06809-0.

Introduction

Preoperative fasting has long been standard practice to reduce the risk of pulmonary aspiration during anesthesia. Yet prolonged fasting imposes measurable physiologic and patient reported burdens, heightened thirst, hunger, anxiety, and discomfort while exacerbating preoperative insulin resistance and catabolic responses that may impede recovery [1]. Reflecting this evidence, contemporary guidance from the American Society of Anesthesiologists (ASA) and the European Society for Clinical Nutrition and Metabolism (ESPEN) permits ingestion of carbohydrate drinks (e.g.Nutricia PreOp) up to two hours before elective procedures in appropriately selected patients, without increasing aspiration risk [1–3].

Preoperative oral carbohydrate loading seeks to mitigate the metabolic stress of surgery by attenuating perioperative insulin resistance and promoting glycemic control, while improving patient comfort parameters such as thirst and hunger. Within enhanced recovery after surgery (ERAS) pathways, abbreviated fasting with preoperative carbohydrate loading is widely recommended and has been integrated into perioperative care particularly for joint arthroplasty and other orthopedic procedures [4].

Despite biologic plausibility and inclusion within ERAS protocols, the clinical effects of preoperative carbohydrate loading in orthopedic surgery remain variably reported. Randomized controlled trials across lower limb procedures have shown mixed outcomes, with some studies suggesting reductions in postoperative pain and length of hospital stay, while others report no significant differences in outcomes such as postoperative nausea and vomiting (PONV) [5, 6]. These inconsistent findings are mirrored in a recent systematic review and meta-analysis of preoperative carbohydrate loading for elective hip and knee arthroplasty. While the study found no statistically significant reductions in length of stay, pain scores, or postoperative nausea and vomiting (PONV), it did note trends toward improved subjective outcomes alongside significant heterogeneity across the included trials [7]. Differences in patient characteristics (e.g., diabetic versus non-diabetic), types of procedures (fracture fixation versus arthroplasty), carbohydrate dosing and timing, anesthetic approaches, and outcome definitions may partly explain this heterogeneity [8].

Given the high volume and clinical impact of orthopedic surgery, a focused synthesis of orthopedic-specific randomized evidence is warranted to better define the role of preoperative carbohydrate loading. Accordingly, this systematic review and meta-analysis evaluates the effects of oral carbohydrate loading compared with standard fasting on key postoperative outcomes in orthopedic surgical patients, including length of hospital stay, postoperative nausea and vomiting, postoperative pain, and postoperative thirst.

Method

Study design

This is a systematic review and meta-analysis study that reports findings from 10 randomized controlled trials (RCTs). The protocol of this study has been published on PROSPERO (CRD420251112790) and followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [9] [Supplementary 1].

Eligibility and exclusion criteria

We included randomized controlled trials that compared preoperative oral carbohydrate loading to a fasting state in patients undergoing any type of orthopedic surgery to reflect clinical reality. Since preoperative fasting is the traditional standard of care, this comparison directly measures the effectiveness of shifting to an ERAS-based protocol (Tables 1 and 2). We acknowledge that while this prevents isolating the metabolic effects of carbohydrates from simple hydration, it provides findings more applicable to the surgical practice.

Table 1.

Comprehensive characteristics of included studies (Total = 708)

References	Country	Sample size	Type of orthopedic surgery	Gender N (%)		Age (M ± SD) years		BMI (M ± SD) kg/m²
				Male	Female	Carbohydrate group	Fasting group	Carbohydrate group	Fasting group
Yap [10]	Malaysia	34	Total arthroplasty, hemiarthroplasty, internal fixation	Fasting = 9 (52.9%) CHO = 7 (41.2%)	Fasting = 8 (47.1%) CHO = 10 (58.8%)	78 ± 1.5	78 ± 2.2	23.7 ± 0.61	23.1 ± 0.78
Chaudhary [5]	Nepal	66	Femur fracture surgery under spinal anesthesia	Fasting = 17 (47.7%) CHO = 14 (45.2%)	Fasting = 16 (45.7%) CHO = 19 (45.3%)	66.4 ± 11.8	69.3 ± 13.9	N/A	N/A
Lee [11]	South Korea	46	Elective Total Knee Arthroplasty (TKA) or Total Hip Arthroplasty (THA)	Fasting = 8 (33.3%) CHO = 6 (27.3%)	Fasting = 16 (66.7%) CHO = 16 (72.7%)	70.3 ± 6.8	70.5 ± 6.0	26.5 ± 3.17	25.73 ± 2.68
Ertural [12]	Turkey	50	Elective Hip Arthroplasty	Fasting = 8 (32%) CHO = 8 (32%)	Fasting = 17 (68%) CHO = 17 (68%)	59.24 ± 15.24	64.40 ± 15.41	27.75 ± 4.15	28.29 ± 4.79
Kadado [13]	USA	102	Elective Primary Total Knee Arthroplasty	Fasting = 17(32.7%) CHO = 15 (30% )	Fasting = 38 (74.5) CHO = 35 (67.3)	67.1 ± 7.4	67.7 ± 9.9	33.0 ± 5.4	32.5 ± 6.2
Ghaffari [14]	Iran	70	Elective orthopedic surgeries of the hip or lower extremities under spinal anesthesia	Fasting = 13 (37.1%) CHO = 21 (60%)	Fasting = 22 (62.9%) CHO = 14 (40%)	64.74 ± 18.36	61.26 ± 20.6	26.67 ± 4.66	27.11 ± 5.27
Luo-ting [6]	China	89	Lower extremity surgery	Fasting = 14 (35%) CHO = 19 (47.5%)	Fasting = 26 (65%) CHO = 21 (52.5%)	72.55 ± 3.56	73.62 ± 4.07	23.59 ± 1.79	23.31 ± 2.93
Dilmen [15]	Turkey	40	Lumbar disc surgery	Fasting = 10 (50%) CHO = 10 (50%)	Fasting = 10 (50%) CHO = 10 (50%)	48.95 ± 11.51	45.25 ± 7.23	26.57 ± 1.18	26.63 ± 1.68
Akbuğa [16]	Turkey	61	Arthroscopic surgery	Fasting = 15 (50%) CHO = 17 (54.8%)	Fasting = 15 (50%) CHO = 14 (45.2%)	47.58 ± 11.24	45.30 ± 9.43	N/A	N/A
Haselton [17]	USA	150	Total hip arthroplasty	Fasting = 30 (40%) CHO = 35 (46.7%)	Fasting = 45 (60%) CHO = 40 (53.3%)	68.2 ± 8.3	69.58 ± 9.3	30.7 ± 6.2	33.2 ± 5.7

Table 2.

Intraoperative characteristics of included studies

References	Intraoperative bleeding (M ± SD) / ml		Duration of surgery (M ± SD) / min
	CHO group	Fasting group	CHO group	Fasting group
Lee [11]	66.67 ± 39.62	66.67 ± 39.4	83.67 ± 32.49	94 ± 37.04
Ertural [12]	Not reported		109.79 ± 32.487	122.83 ± 38.882
Kadado [13]	88.8 ± 59.7	79.4 ± 46.4	138.8 ± 18.3	129.7 ± 20.5
Ghaffari [14]	658.29 ± 184.41	725.43 ± 188.88	132.46 ± 17.17	136.29 ± 19.86
Luo-ting [6]	175 ± 38.45	158.33 ± 76.89	120.4 ± 18.79	121.38 ± 15.65
Dilmen- [15]	Not reported		72.75 ± 29.06	80 ± 32.11
Akbuğa [16]	Not reported		74.19 ± 13.6	80.16 ± 20.1
Chaudhary [5]	Less than 500 ml: 31 (54.4%) More than 500 ml: 2 (22.2%)	Less than 500 ml 26 (45.6%) More than 500 ml: 7 (77.8%)	Less than one hour: 1 (20.0%) More than one hour: 32 (52.5%)	Less than one hour: 4 (80.0%) More than one hour: 29 (47.5%)

The studies had to report at least one of the following postoperative outcomes: length of hospital stay, pain, nausea, vomiting, or thirst level. We did not set any restrictions on the patients’ age or on the dose or regimen of the carbohydrate intervention. We excluded any studies that compared preoperative oral carbohydrates to a placebo or any other control group. We also excluded reviews, conference papers, expert opinions, case reports, non-English language publications, and any study with incomplete outcome data.

Information sources and search strategy

Studies in PubMed, Cochrane and Scopus databases were searched systematically covering all relevant studies published up to July 2025. The search strategy incorporated a combination of Medical Subject Headings (MeSH) terms and keywords such as “Dietary Carbohydrates, carbohydrate loading, oral carbohydrate, carbohydrate drink, oral nutrition, Nutrient, Dietary Supplements, Sugar, Orthopedic Procedures, Orthopedic surgery and orthopedic operation. Boolean operators were used to optimize the search and the full search strategy for at least one database (PubMed) is provided in the supplementary materials to ensure reproducibility and transparency [Supplementary 2].

Also additional studies were identified by manually reviewing the reference lists of included studies and relevant systematic reviews.

Selection process and data extraction

All results were exported into EndNote to remove duplicates. The remaining studies were then uploaded to an Excel sheet for title and abstract screening against predefined inclusion and exclusion criteria. This screening process was conducted independently by two researchers, with any disagreements resolved through discussion with a third team member. Full-text articles of potentially relevant studies were retrieved and assessed independently (Fig. 1).

Fig. 1.

PRISMA flow chart for the screening process

Data were then extracted into three separate sheets: a summary, a baseline, and an outcomes sheet. All baseline and outcome data were standardized, including operative time (minutes), blood loss (milliliters), BMI (kg/m2), duration of surgery (minutes), and hospital stay (days). For study that reported data as median and interquartile range (IQR) [11], the information was converted to mean and standard deviation (SD) using the method described by Wan et al., [18].

Methods of data synthesis and statistical analysis

A meta-analysis was conducted using Review Manager 5.4 and R statistical software. The effect measures for continuous variables were the mean difference (MD) or standard mean difference (SMD), while the risk ratio (RR) was used for dichotomous variables. All effect measures were reported with a 95% confidence interval (CI). To assess the heterogeneity among studies, the I² statistic was used and its classification based on the guidelines in the Cochrane Handbook for Systematic Reviews [19]. A random-effects model, sensitivity and subgroup analysis were employed to minimize high heterogeneity.For rare events, sensitivity analyses were performed using the risk difference metric within a Mantel–Haenszel random-effects model, allowing inclusion of studies with zero events in both arms.

Assessment of publication bias using funnel plots and formal testing was not performed, as each meta-analysis included fewer than 10 studies.

Risk of bias assessment

Risk of bias assessment was conducted using the Cochrane Risk of Bias tool (RoB 2) [20, 21]. It focuses on five key domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. In these domains, each study was assigned an overall risk of bias rating: low, some concerns, or high risk. The results of this assessment are detailed in Fig. 2.

Fig. 2.

Risk of bias assessment of the included RCTs (n = 10)

Results

A total of 10 RCTs were included in this meta-analysis. The mean age of participants varied between 45.3 and 78 years. The mean BMI was reported in 8 studies and ranged from 22.7 to 33.2 kg/m². Intraoperative bleeding was reported in 5 studies ranging from 66.67 to 725.43 ml. The duration of the surgeries was also documented and ranged from 72.75 to 138.8 min. The studies represent diverse geographic and clinical settings from Malaysia, Nepal, South Korea, Turkey, USA, Iran, and China. (Tables 1 and 2)

Results of meta-analysis

1/Postoperative thirst levels (oral carbohydrate vs. fasting status)

Three RCTs including 190 participants (95 in the preoperative oral carbohydrate group and 95 in the fasting group) were pooled using a random-effects model. Preoperative oral carbohydrate loading was associated with a statistically significant reduction in postoperative thirst levels compared with fasting status (mean difference − 2.58, 95% CI − 4.10 to − 1.05; Test for overall effect: Z = − 3.31 (P = 0.0009). Statistical heterogeneity was substantial (I² = 85.3%, chi² = 13.64, df = 2, P = 0.0011) and Prediction interval of -8.82 to 3.67 (Fig. 3).

Fig. 3.

Postoperative thirst levels in oral carbohydrate versus fasting groups (random-effects model) (n = 190)

After conducting a sensitivity analysis by excluding Ghaffari et al. [17], heterogeneity decreased to 0%, suggesting that this study was the primary contributor to the observed heterogeneity. This heterogeneity may be attributed to a potential dose-response relationship between the volume of preoperative intake and the reduction of postoperative thirst. As, Ghaffari et al. utilized a significantly lower dose of only 200 ml, whereas other included trials administered between 400 and 800 ml of carbohydrate loading.

2/Length of hospital stay (oral carbohydrate vs. fasting status)

6 RCTs including 472 participants (217 in the preoperative oral carbohydrate group and 225 in the fasting group) were pooled using a random-effects model. Preoperative oral carbohydrate loading was associated with a statistically significant reduction in length of hospital stay compared with fasting (mean difference − 0.36 days, 95% CI [− 0.70, − 0.02] and test for overall effect: Z = − 2.06 (P = 0.0391).

Regarding heterogeneity, the studies demonstrated moderate statistical heterogeneity, with Tau² = 0.0361, Chi² = 7.93 (df = 4, P = 0.0941), and I² = 49.6% and the Prediction interval was reported as − 1.07 to 0.36 (Fig. 4).

Fig. 4.

Hospital length of stay in oral carbohydrate versus fasting groups (random-effects model) (n = 472)

3/Vomiting (oral carbohydrate vs. fasting status)

A total of four studies were included. However, the study by Dilmen et al. [15] was not estimable due to zero events in both the experimental and control groups, and excluded from the pooled risk ratio calculation.

The pooled analysis of the remaining three studies [6, 11, 13] using a random-effects model yielded a risk ratio of 0.71 with a 95% CI of 0.12 to 4.05. The result was not statistically significant (P = 0.6960), indicating no clear evidence of a difference in the risk of vomiting between the preoperative oral carbohydrate and fasting status groups. There was considerable heterogeneity among the included studies (Tau² = 1.71, Chi² = 7.71, df = 2, P = 0.0211, I² = 74.1%) indicating substantial heterogeneity and the Prediction interval was calculated as 0.00 to 639.29 (Fig. 5).

Fig. 5.

postoperative vomiting in oral carbohydrate versus fasting groups (random-effects model) (n = 268)

Regarding the rare event, sensitivity analysis using risk difference produced results consistent with the primary risk ratio analysis, with no statistically significant difference between groups.

A sensitivity analysis excluding the study by Lee et al. [14] reduced heterogeneity to I² = 0%. The higher risk observed in this trial is likely attributable to differences in patient characteristics, since the study population consisted of diabetic patients, who may respond differently compared with the general population included in the other studies.

4/Nausea (oral carbohydrate vs. fasting status)

Four RCTs including 268 participants (132 in the experimental group and 136 in the control group) were pooled using a random-effects model. Preoperative intervention was associated with a non-significant reduction in the risk of nausea compared with the control group (risk ratio 0.72, 95% CI 0.30 to 1.70; P = 0.4506) Statistical heterogeneity was moderate to substantial (I² = 64.8%, chi² = 8.52, df = 3, P = 0.0365) and prediction interval of 0.05 to 9.43 (Fig. 6). After conducting a sensitivity analysis by excluding Lee et al. [14], heterogeneity decreased substantially (I² = 18%), suggesting that this study was a potential source of the observed heterogeneity. This difference may be attributed to variations in the study population, as Lee et al. [11] included diabetic patients, unlike the other trials.

Fig. 6.

Postoperative nausea in oral carbohydrate vs. fasting groups (random-effects model) (n = 268)

5/Post-operative pain (oral carbohydrate vs. fasting status):

Seven RCTs including 512 participants (252 in the preoperative oral carbohydrate group and 260 in the fasting group) were pooled using a random effects model. Preoperative oral carbohydrate loading was associated with a non-significant reduction in postoperative pain scores compared with fasting status (Std. mean difference − 0.31, 95% CI − 0.65 to 0.04; p = 0.0797). Statistical moderate heterogeneity was observed (I² = 70.3%, chi² = 20.21, df = 6 (P = 0.0025) and the prediction interval was calculated as − 1.34 to 0.73 (Fig. 7).

Fig. 7.

Postoperative pain levels in oral carbohydrate versus fasting groups (random-effects model) (n = 512)

A subgroup analysis was performed according to the type of anesthesia (Fig. 8). In the spinal anesthesia subgroup (3 studies), the intervention demonstrated a significant reduction in the outcome compared with controls (SMD = − 0.72, 95% CI: − 1.15 to − 0.29, p = 0.0010), with moderate heterogeneity (I² = 51%). However, this finding should be interpreted cautiously given the small sample size and the variability in time points across studies and should therefore considered as an exploratory subgroup. In contrast, in the mixed/general anesthesia subgroup (4 studies), no significant difference was observed between groups (SMD = − 0.05, 95% CI: − 0.29 to 0.19, P = 0.69), and heterogeneity was negligible (I² = 12%). The test for subgroup differences was statistically significant (Chi² = 7.22, df = 1, P = 0.007; I² = 86.2%), suggesting a potential difference between anesthesia types. However, due to the small number of studies and possible confounding factors such as surgical procedure type and analgesic protocols, these results should be regarded as hypothesis-generating rather than confirmatory.

Fig. 8.

Subgroup analysis based on anesthesia type in the postoperative pain levels in oral carbohydrate versus fasting Groups (random-effects model)

GRADE assessment

Certainty of evidence was assessed using the GRADE approach across five domains for each critical outcome. All outcomes started at high certainty, as all included studies were randomized controlled trials. Evidence was downgraded for moderate to serious risk of bias, as most studies had some concerns in the RoB2 assessment. Most outcomes were further downgraded for inconsistency due to substantial heterogeneity and for imprecision based on wide confidence or prediction intervals crossing the line of no effect. The effect estimates were direct and matched the population, intervention, comparator, and outcomes of interest. Publication bias was not formally assessed due to the small number of studies per outcome (Tables 3 and 4)

Table 3.

Summary of findings and certainty of evidence for preoperative oral carbohydrate versus fasting (GRADE assessment)

Outcome	Studies numbers (n intervention / n control)	Effect estimate (95% CI) & Prediction Interval	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias	Overall certainty
Postoperative thirst	3 (95 CHO/95 fasting)	MD = − 2.58 [− 4.10, − 1.05] PI = − 8.82 to 3.67	Serious¹	Very serious² (I² = 85.3%)	Direct evidence3 (No downgrade)	No downgrade4	Not assessed5	Low
Length of hospital stay	6 (217 CHO, 225 Fasting)	MD = − 0.36 [− 0.70, − 0.02] PI = − 1.07 to 0.36	Serious¹	Not serious² (I² = 49.6%, moderate heterogeneity)	Direct evidence 3 (No downgrade)	No downgrade4	Not assessed5	Moderate
Vomiting	4 (132 CHO/136 fasting)	RR = 0.71 [0.12, 4.05] PI = 0.00 to 639.29	Serious¹	Serious² (I² = 74.1%)	Direct evidence 3 (No downgrade)	Serious4 (wide CI crossing null, PI very wide)	Not assessed5	Low
Nausea	4 (132 CHO/136 fasting)	RR= 1.72 [0.30, 1.70] PI = 0.05 to 9.43	Serious¹	Serious² (I² = 64.8%)	Direct evidence 3 (No downgrade)	Serious4 (CI crosses null)	Not assessed5	Low
Postoperative pain	7 (252 CHO, 260 Fasting)	SMD = − 0.31 [− 0.65, 0.04] PI = − 1.34 to 0.73	Serious¹	Serious² (I² = 73.9%)	Direct evidence 3 (No downgrade)	Serious4 (CI crosses null, PI wide)	Not assessed5	Low

¹ Risk of bias: Downgraded one level because most studies contributing to each outcome were judged as having some concerns in the RoB2 assessment; for pain and hospital stay, one study was at high risk of bias, but high-risk studies were not dominant. ² Inconsistency: Downgraded when substantial heterogeneity was present. Heterogeneity was assessed using I² statistic and forest plots: - I² > 60–70% → serious inconsistency (downgrade 1 level) - I² ≥ 80% → very serious inconsistency (downgrade 2 levels) Hospital stay I² = 49.6%) was not downgraded. ³Indirectness: Evidence was direct, meaning the populations, interventions, comparators, and outcomes of the included studies matched the review question. According to GRADE: - Direct evidence → no downgrade - Indirect evidence → downgrade 1 level (serious) -Severe indirectness → downgrade 2 levels (very serious) ⁴Imprecision: Downgraded when confidence intervals were wide or crossed the line of no effect, or when prediction intervals indicated substantial uncertainty in the true effect. According to GRADE guidance: Narrow CI excluding null → no downgrade Wide CI crossing null → downgrade 1 level (serious) Very wide CI with small sample → downgrade 2 levels (very serious) ⁴Publication bias: Not assessed due to the small number of studies (< 10) per outcome

Table 4.

Study-level characteristics of ERAS and fasting protocols in the included trials

References	Intervention group	Control  group	Fasting duration before surgery	Clear fluids/water allowed	IV maintenance fluids standardized
Yap [10]	3 servings of a CHO ( two drinks on the day before surgery at 4pm and 8pm and the third, to be consumed 2–6 h prior to the scheduled time of surgery as recommended by the ERAS protocol	Conventional fasting	Solid food up to 6 h and clear fluids up to 2 h before surgery.	Clear fluids up to 2 h before surgery.	Not reported
Chaudhary [5]	CHO according to the ERAS protocol ( glucose-D as a carbohydrate-rich drink of Nepali product	Fasting control	From midnight to the next morning as in existence	Not reported	Not reported
Ertural [12]	OCS in 10 min, which contained 12.5% of glucose that the administration of 800 ml at midnight the night before the surgery and 400 ml 2 h before the surgery	Fasting control	fasted overnight from midnight, one night before surgery	No	No
Lee [11]	received a 400-mL carbohydrate drink (12.8%  carbohydrate, 0.5 kcal/mL, 14% monosaccharides,3% disaccharides, 83% polysaccharides, and 265 mOsm/ kg; NO-NPO, Daesang WelLife, Seoul, Korea) 2–3 h  before anesthesia	Fasting control	fasted overnight from midnight, one night before surgery	Not reported	Not reported
Kadado [13]	Received a clear, noncarbonated lemon-flavored, iso-osmolar carbohydrate drink (Nutricia preOp; Nutricia; 12.5% carbohydrates, 50 kcal/100 mL, 260 mOsm/kg, and pH 5.0).	Fasting control	fasted overnight from midnight, one night before surgery	No	No
Ghaffari [14]	patients received 200 mL of 10% oral dextrose  solution (10 g glucose/100 mL, osmolality: 555.2 mOsm/L, delivering 34 kcal/100 mL) 2–3 h before	Fasting control	Adhered to standard preoperative fasting guidelines, refraining from solids and liquids after 10:00 p.m.)	No	No
Luo-ting [6]	200 mL carbohydrate drink (Fuan [12.6% carbohydrate]; 214.2 kJ/100 mL, 280 mOsm/kg and 200 mL per bottle) 2 h before surgery.	Fasting Control	From midnight before the surgery	No	No
Dilmen [18]	12.5% oral carbohydrate solution (285 mOsm kg − 1, Nutricia Preop) 800 mL at between 21:00 and 24:00 the night before and 400 mL at 06:00 which was 2 h prior to operation	Fasting Control	8 h prior to operation	No	No
Akbuğa [16]	400 mL of grain-free sour cherry juice from a fixed brand was given to the patients in the intervention group 2 h before surgery	Fasting control	Not reported	No	No
Haselton [17]	ERAS group was provided with two 8 oz. bottles of a clear carbohydrate-enriched nutritional supplement (Ensure Pre-Surgery™, Abbott Laboratories, Coumbus, OH), containing a complex carbohydrate formulation primarily composed of matodextrin, with zinc and seenium as antioxidant components. Patients consumed the first dose prior to midnight and the second 2 h before surgery.	Traditional fasting protocols	12 h preoperatively	Limited sips of water allowed only for essential medications.	No

• For postoperative thirst, certainty was rated as low due to very serious inconsistency and serious imprecision.

• For length of hospital stay, certainty was rated as moderate due to serious risk of bias and moderate heterogeneity.

• For postoperative vomiting and postoperative nausea, certainty was rated as low due to serious inconsistency and serious imprecision.

• For postoperative pain, certainty was rated as low due to serious inconsistency and serious imprecision; exploratory subgroup analyses suggested potential benefit with spinal anesthesia, but these findings should be interpreted cautiously.

Discussion

This meta-analysis compares preoperative oral carbohydrate (CHO) loading with fasting across key postoperative outcomes, including hospital length of stay, postoperative nausea and vomiting, thirst, and pain. By synthesizing data from randomized controlled trials, the study provides an overview of the available evidence and highlights potential benefits and limitations of CHO intake within preoperative care.

The most definitive effect was observed is the postoperative thirst, where carbohydrate loading significantly reduced thirst scores compared to fasting. This finding aligns with earlier systematic reviews that highlighted improvements in both thirst and overall well-being among patients who consumed CHO drink [22]. Similarly, a clinical trial examining the effects of carbohydrate use on preoperative thirst reported that groups receiving the carbohydrate loading experienced less thirst discomfort than those in the fasting group [23].In contrast, A meta-analysis of 19 randomized trials in patients undergoing laparoscopic cholecystectomy found that preoperative CHO did not significantly reduce postoperative thirst when compared with fasting or placebo [24].

Although initial analysis show high heterogeneity among the included studies, a sensitivity analysis showed that excluding the study by Ghaffari et al. [14] reduced heterogeneity to zero, indicating that this study was the main contributor to the variability observed. This heterogeneity may be attributed to a potential dose-response relationship between the volume of preoperative intake and the reduction of postoperative thirst. Ghaffari et al. utilized a significantly lower dose of only 200 ml, whereas other included trials administered between 400 and 800 ml of carbohydrate loading. These findings suggest that a minimum threshold volume may be necessary to achieve a consistent physiological effect. Future research should focus on standardizing study protocols to reduce variability and investigate the mechanisms through which CHO intake alleviates thirst.

Regarding hospital stay, our findings indicate a statistically significant reduction in length of stay among patients receiving preoperative carbohydrate loading compared with those who fasted. However, the magnitude of this effect was modest (MD − 0.36 days, approximately 8.6 h), and its clinical relevance should be interpreted with caution. In many hospital settings, discharge timing is influenced by fixed ward rounds, institutional discharge pathways, and administrative factors; therefore, a reduction of less than one day may not consistently translate into earlier discharge for individual patients. This result is consistent with other meta-analyses in major non-cardiac surgeries that have reported reductions in hospital stay with CHO loading, suggesting that procedure type and perioperative protocols may influence outcomes [25]. In contrast, Pradhan et al. reported no significant difference in length of hospitalization following lower limb arthroplasty [7].

For postoperative vomiting and nausea, our analysis demonstrated no clear difference between CHO and fasting groups. Comparable inconsistencies have been reported in prior review, where the benefits of CHO on nausea and vomiting were minimal, although improvements in hunger, anxiety, and overall comfort were more consistently observed [22]. Other systematic reviews reported that preoperative carbohydrate loading was associated with lower postoperative nausea incidence compared with fasting [26].

Considerable heterogeneity was also observed and largely attributable to the trial by Lee et al. [11], which specifically included diabetic patients. This subgroup may respond differently to preoperative CHO intake and introduces a confounding variable that impacts the overall results and necessitates a cautious interpretation of the pooled data.

The safety profile of preoperative CHO loading in high-risk orthopedic population such as the diabetic patients or requires careful scrutiny. While generally considered safe for the general population, evidence regarding metabolic stability in these groups is conflicting. Notably, Lee et al. [11] observed increased glycemic variability among diabetic patients receiving preoperative carbohydrates, suggesting that the insulin response may not be sufficient to maintain euglycemia in all cases. This is particularly relevant in the context of orthopedic trauma, where the physiological stress of injury already induces a state of diabetes of injury or insulin resistance [27]. So CHO loading should not be a universal standard for patients with impaired glucose metabolism. Clinical protocols should prioritize individualized risk stratification and rigorous preoperative monitoring to balance patient comfort with metabolic stability.

Regarding the postoperative pain, our exploratory subgroup analysis suggests that the impact of preoperative carbohydrate loading may be influenced by the anesthetic technique. While no significant difference was observed in the general anesthesia group, a notable reduction in pain was found in patients undergoing spinal anesthesia. This potential benefit may be theoretically linked to the afferent neural blockade provided by spinal anesthesia, which effectively blunts the hypothalamus-pituitary-adrenal (HPA) axis and mitigates the systemic surgical stress response. By maintaining metabolic stability and preventing the overwhelming rise in cortisol typically associated with surgical trauma, neuraxial blocks create a physiological environment where the anti-insulin resistance effects of preoperative CHO drinks become more readily apparent [27, 28]. This aligns also with prior evidence suggesting that carbohydrate loading before surgery can attenuate stress responses and enhance postoperative comfort [29].

Comparable results were also observed reduced pain levels among colorectal surgery patients receiving preoperative CHO drinks [30]. Similarly, a systematic review and meta-analysis involving laparoscopic cholecystectomy patients demonstrated significantly lower postoperative pain in the carbohydrate group than in those who fasted [26]. In contrast, a systematic review and meta-analysis of preoperative carbohydrate loading in hip and knee arthroplasty patients found no statistically significant differences in pain scores between the CHO and control groups [7]. However, our results should be interpreted with caution due to the small total sample size in this subgroup, the variability in pain assessment scales, time points across the included trials, potential differences in surgical procedures and analgesic protocols. Therefore, these findings are hypothesis-generating and not conclusive.

Given that preoperative oral carbohydrate loading is low-cost, easy to implement, and already recommended in ERAS protocols, even modest improvements in patient comfort, such as reduced thirst, justify its use in orthopedic ERAS pathways. Our review, however, does not demonstrate strong benefits regarding reduce hospital stay, postoperative nausea and vomiting (PONV) and postoperative pain in this population. Future research should focus on adequately powered multicenter RCTs, especially in subgroups such as patients with diabetes, to clarify the extent of benefits and optimize preoperative perioperative carbohydrate protocols.

Conclusion

This meta-analysis demonstrates that preoperative oral carbohydrate (CHO) loading is a safe strategy that improves patient-reported comfort, specifically by consistently reducing postoperative thirst and modestly shortening hospital stay. However, the observed reduction in length of stay was very small. While no significant differences were observed in postoperative nausea and vomiting, exploratory subgroup analyses suggest a possible reduction in postoperative pain in patients receiving spinal anesthesia but it based on a small number of studies and participants and should be interpreted cautiously as hypothesis-generating. These findings reinforce the value of the preoperative CHO as part of Enhanced Recovery after Surgery pathways, where it contributes to greater comfort and possibly smoother recovery.

Limitations

This review has several limitations. The number of included trials and sample sizes for some outcomes were relatively small, limiting statistical power. Considerable heterogeneity was observed, partly due to differences in study populations, surgical procedures, and carbohydrate protocols. Some studies had concerns regarding selective reporting, raising the risk of bias. The generalizability of the results is constrained by geographical and linguistic factors as the majority of included studies originated from Asia and the Middle East, with a notable paucity of data from European and North American cohorts. This geographical concentration, combined with the exclusion of non-English language studies, introduces a potential language and publication bias that may limit the global applicability of the conclusions.

Finally, this review primarily focused on trials comparing CHO loading to standard fasting rather than placebos. While this approach directly evaluates the intervention against the traditional standard of care, it limits our ability to distinguish the specific metabolic benefits of carbohydrates from the physiological effects of preoperative fluid ingestion alone. Furthermore, our search strategy was limited to major peer-reviewed databases (PubMed, Cochrane, and Scopus) and did not encompass a comprehensive search of grey literature or clinical trial registries (e.g., https://clinicaltrials.gov/).

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

N. E. A. H. designed the research study and supervised the overall project. N. A., K. A. and M. A. work on the screening. N. A., K. (A) and E. A. (B) extracted the data. N. E. A. H. performed the data analysis and the analysis reviewed by A. H. A. A. A. I. M. A. and K. A. did the risk of bias assessment. A. I. M. A., H. A., and J. M. A. contributed to data interpretation and manuscript writing. D. A. wrote the abstract and revised the manuscript. All authors read and approved the final manuscript.

Funding

This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and analyzed during this study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study is a systematic review of previously published research. No new data were collected from human participants or animals. Therefore, ethical approval and informed consent were not required.

Human and animal participants

Not applicable.

Competing interests

The authors declare no competing interests.