Analysis of the predictive value of three nutritional risk screening methods for surgical prognosis and recurrence of chronic subdural hematoma: A retrospective study

Abstract

Objective

Patient nutritional status is an independent and crucial prognostic predictor; however, it is often overlooked in clinical practice. This study aimed to evaluate the value of three nutritional risk screening tools—Geriatric Nutritional Risk Index, Prognostic Nutritional Index, and Nutritional Risk Screening 2002—in predicting postoperative prognosis and recurrence in patients with chronic subdural hematoma.

Method

According to the inclusion and exclusion criteria, a total of 153 patients with chronic subdural hematoma who were hospitalized and underwent surgical treatment at our hospital between December 2017 and October 2022 were retrospectively enrolled. General clinical data and laboratory test results were collected. The Nutritional Risk Screening 2002, Geriatric Nutritional Risk Index, and Prognostic Nutritional Index scores were assessed. Based on the modified Rankin scale scores, patients were divided into poor prognosis (score ≥3) and good prognosis (score <3) groups. According to the presence or absence of recurrence within 6 months of surgery, patients were classified into recurrence and nonrecurrence groups. General clinical data, laboratory results, and scores of three nutritional risk screening tools were evaluated. Receiver operating characteristic curves were plotted, and multivariate logistic regression analysis was performed.

Results

Among the 153 patients, 13 (8.5%) had a poor prognosis. Compared with the good prognosis group, patients in the poor prognosis group were older, had higher fibrinogen levels, and had lower serum albumin and hemoglobin levels (p < 0.05). In total, 33 (21.6%) patients experienced recurrence. Compared with the nonrecurrence group, patients in the recurrence group had a higher percentage of monocytes (p < 0.05) and lower body mass index, serum albumin levels, and hemoglobin levels (p < 0.05). In the poor prognosis group, the rates of nutritional risk determined using the Geriatric Nutritional Risk Index, Prognostic Nutritional Index, and Nutritional Risk Screening 2002 were 46.2%, 69.2%, and 92.3%, respectively, which were significantly higher than those in the good prognosis group (p < 0.05). In the recurrence group, the nutritional risk rates determined using the abovementioned three tools were 42.4%, 51.5%, and 78.8%, respectively, all significantly higher than those in the nonrecurrence group (p < 0.05). Receiver operating characteristic curve analysis indicated that the Geriatric Nutritional Risk Index, Prognostic Nutritional Index, and Nutritional Risk Screening 2002 are predictive markers for both surgical prognosis and postoperative recurrence of chronic subdural hematoma. Moreover, multivariate logistic regression analysis identified the three tools as risk factors for poor prognosis and recurrence after surgery for chronic subdural hematoma (p < 0.05).

Conclusion

Three nutritional risk screening tools—the Geriatric Nutritional Risk Index, Prognostic Nutritional Index, and Nutritional Risk Screening 2002—were independently associated with both poor prognosis and recurrence of chronic subdural hematoma. Among them, the Geriatric Nutritional Risk Index showed the highest specificity in predicting poor prognosis, while the Nutritional Risk Screening 2002 demonstrated the highest specificity in predicting postoperative recurrence.

Keywords

Chronic subdural hematomas Geriatric Nutritional Risk Index Prognostic Nutritional Index Nutritional Risk Screening 2002 prognosis recurrence

Introduction

Chronic subdural hematoma (CSDH) is a common age-related condition. Owing to factors such as brain atrophy and the increasing use of anticoagulant and antiplatelet medications,¹ older populations face a growing risk of developing CSDH, with an annual rise in the associated mortality and morbidity rates.^2,3 Older adults are also more susceptible to malnutrition, which has been identified as a significant independent predictor of poor clinical outcomes in conditions such as stroke, heart failure, and cancer.^4–6 Nutritional status assessment should be regarded as an essential component of patient management because it can influence clinical outcomes, complication risks, and treatment response. Therefore, early nutritional risk screening deserves greater emphasis in clinical practice. This is particularly relevant for older CSDH patients, who are at higher risk of malnutrition. Preoperative nutritional risk screening, comprehensive assessment, and nutritional guidance are critical for predicting clinical prognosis and optimizing patient care.

The Geriatric Nutritional Risk Index (GNRI) is a screening tool for nutritional risk, calculated using the patient’s height, body weight, and serum albumin (ALB) level.⁷ It is strongly associated with malnutrition-related mortality and other adverse outcomes and has been established as a reliable predictor of clinical prognosis in various types of cancer. The GNRI has been validated as an important method for estimating survival in cancer patients,⁸ assessing prognosis in those with heart failure,⁹ and predicting postoperative complications among older trauma patients.¹⁰ The Nutritional Risk Screening 2002 (NRS 2002) is an evidence-based nutritional risk screening tool developed by the European Society for Clinical Nutrition and Metabolism in 2003; it is recommended for routine use in clinical practice.^11,12 Furthermore, it has proven useful in predicting postoperative complications,¹³ predicting adverse clinical outcomes in hospitalized adults,¹⁴ and guiding nutritional support strategies for intensive care unit patients.¹⁵ The Prognostic Nutritional Index (PNI) was originally developed by the Japanese researcher Onodera as a novel nutritional indicator for predicting disease progression.¹⁶ It has been reported to be closely associated with prognosis in numerous malignant tumors and has been validated in gastrointestinal cancer,¹⁷ breast cancer,¹⁸ and advanced malignancies.¹⁹ The PNI remains highly valuable in predicting postoperative outcomes for lung cancer, spinal surgeries, and cranial procedures.^20–22

CSDH occurs most frequently in older patients, a population that also exhibits a high prevalence of nutritional risk. Previous studies have indicated that a low PNI is associated with postoperative recurrence of CSDH;²³ however, the potential of the GNRI and NRS 2002 in predicting postoperative prognosis and recurrence of CSDH has not yet been thoroughly investigated. Therefore, this study aimed to evaluate and compare the predictive value of these three nutritional risk screening tools (GNRI, PNI, and NRS 2002) by analyzing their correlations with surgical prognosis and recurrence rates of CSDH in older patients.

Methods

Patients

This retrospective study enrolled 153 consecutive patients with CSDH who were admitted to our hospital and underwent concurrent surgical treatment from December 2017 to October 2022. We conducted this study in accordance with the Helsinki Declaration of 1975, as revised in 2024. All procedures were approved by the Ethics Committee of the Affiliated Hospital of Chengde Medical University. Written informed consent was not required for this retrospective study. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.²⁴ Patients were recruited if they met the following inclusion criteria: (a) age 60–90 years; (b) preoperative craniocerebral computed tomography (CT) establishing a clear diagnosis of CSDH; (c) underwent minimally invasive puncture and drainage for subdural hematoma; and (d) available for postoperative follow-up. Patients were excluded if they met the following exclusion criteria: (a) transfusion of blood products such as human ALB before admission; (b) presence of infectious diseases; (c) history of chronic diseases such as liver cirrhosis, hyperlipidemia, and anemia; or (d) missing clinical data and inability or unwillingness to attend postoperative follow-up.

Data collection

Within 24 h after admission, the patients’ demographic characteristics and clinical information were recorded, including age, sex, height, weight, smoking history, alcohol consumption history, medical history (including hypertension and diabetes), use of anticoagulant and antiplatelet medications, and hematoma location. Nutritional risk screening was performed using the GNRI, PNI, and NRS 2002 scoring systems. Laboratory data were collected to obtain preoperative serum ALB, hemoglobin (HB), cholesterol, triglyceride, and fibrinogen levels, peripheral lymphocyte and neutrophil counts, and peripheral monocyte percentage. Information about the use of urokinase before, during, and after the surgery was obtained from the medical record documentation system. In addition, all patient details have been deidentified.

Operative procedure

The patient was placed in the supine position with the head tilted toward the healthy side. The puncture site was selected at the thickest region of the hematoma. Under local anesthesia, an adjusted YL-1 type puncture needle was vertically advanced into the cranium in a single motion using an electric drill. The dura was carefully ground in situ until perforation was achieved. The needle stylet was then withdrawn, and dark red liquid was observed flowing out. The hematoma cavity was irrigated with saline until the effluent became clear. An external anti-reflux sterile drainage device was connected, the drainage tube was clamped, and the surgical site was dressed with a sterile bandage.

Prognostic score

According to the modified Rankin scale (mRS) score (Table 1) at discharge, participants were categorized into the poor prognosis (score ≥3) and good prognosis (score <3) groups.²⁵

Table 1.

The modified Rankin scale (mRS).

The scale ranges from 0 to 6, with 0 indicating perfect health without symptoms and 6 representing death.
0 - No symptoms.
1 - No significant disability. Able to carry out all usual activities, despite some symptoms.
2 - Slight disability. Able to look after own affairs without assistance, but unable to carry out all previous activities.
3 - Moderate disability. Requires some help, but able to walk unassisted.
4 - Moderately severe disability. Unable to attend to own bodily needs without assistance, and unable to walk unassisted.
5 - Severe disability. Requires constant nursing care and attention, bedridden, incontinent.
6 - Dead.

Follow-up procedure and criteria for recurrence

Patients were followed up at the neurosurgery outpatient clinic 1 month after discharge, and head CT was performed. Recovery status, presence of reported symptoms, and neurological deficits were documented. From months 2 to 6 after discharge, follow-ups were conducted via telephone or at the neurosurgery outpatient clinic to record recovery progress, reported symptoms, and neurological deficits. Recurrence was defined as the re-accumulation of a subdural hematoma on the operative side (CT attenuation values >20 HU), showing an increase in volume compared with the imaging findings at discharge, accompanied with new neurological deficits such as headache, gait disturbance, or hemiparesis, within 6 months postoperatively.²⁶

NRS 2002

All patients were screened for nutritional risk using the NRS 2002 (Table 2); the total score was calculated using the following formula: total score = (highest score obtained from the nutritional status indicator) + (highest disease severity score obtained) + (age score). A score ≥3 was considered to indicate nutritional risk.

Table 2.

The NRS 2002 scoring criteria.

	0	1	2	3
Nutritional status indicator	Normal nutritional status	Weight loss >5% in the past 3 months or impaired dietary intake (50%–75% of normal) in the past week	Weight loss >5% in the past 2 months or impaired dietary intake (25%–50% of normal) in the past week	Weight loss >5% in the past 1 month (weight loss >15% in the past 3 months) or impaired dietary intake (0%–25% of normal) in the past week or BMI <18.5 and impaired dietary intake
Disease severity score	Without disease or disease with increased nutritional requirements	Hip fracture, chronic patients with acute complications (e.g. cirrhosis and COPD), long-term hemodialysis, diabetes, oncology patient	Major abdominal surgery, stroke, severe pneumonia, hematologic malignancy	Head injury, bone marrow transplantation, ICU patients (APACHE score >10)
Age score	Age < 70 years	Age ≥ 70 years

NRS 2002: Nutritional Risk Screening 2002; COPD: chronic obstructive pulmonary disease; ICU: intensive care unit; APACHE: Acute Physiology and Chronic Health Evaluation.

GNRI

The GNRI was calculated as follows: GNRI = 1.489 × serum ALB (g/L) + 41.7 ×(actual body weight/ideal body weight).⁷ Ideal body weight = patient’s height (m)² ×22 (body mass index (BMI)). When the actual preoperative body weight was higher than the ideal weight, the ratio was set to 1. GNRI ≥98 indicated no nutritional risk, while GNRI <98 indicated the presence of nutritional risk.

PNI

The PNI was calculated using the formula: PNI = ALB (g/L) + 5 × lymphocyte count (×10⁹/L).²⁷ According to the criteria established by Onodera et al.,¹⁶ PNI ≥45 indicated no nutritional risk, while PNI <45 indicated the presence of nutritional risk.

Statistical analyses

Continuous variables were presented as mean or median (interquartile range), and categorical variables were presented as frequencies (percentages).

We used t-test to determine differences between the poor and good prognosis groups as well as between the recurrence and nonrecurrence groups when the data were distributed normally. Analysis of baseline characteristics was performed using Student’s t-test or Mann–Whitney U test for continuous variables and chi-square test or Fisher exact test for categorical variables. In cases where the data exhibited a hierarchical distribution, the rank-sum test was applied to assess differences between the groups. Receiver operating characteristic (ROC) curves were generated to evaluate the value of the NRS 2002, GNRI, and PNI in predicting postoperative prognosis and recurrence in patients with CSDH. Additionally, multivariate logistic regression analyses were performed to identify risk factors associated with CSDH. A p-value <0.05 was considered statistically significant.

Results

Comparison of clinical data

In total, 153 patients with CSDH who were aged ≥60 years were included; of these, 13 (8.50%) had a poor prognosis and 140 (91.5%) had a good prognosis. Patients with a poor prognosis were significantly older than those with a good prognosis (p < 0.05). Thirty-three (21.6%) patients experienced recurrence within 6 months, while 120 (78.4%) did not; the average BMI of the recurrence group was significantly lower than that of the nonrecurrence group (p < 0.05) (Tables 3 and 4).

Table 3.

Comparison of clinical data between the good and poor prognosis groups.

Variable	Poor prognosis (n = 13)	Good prognosis (n = 140)	t/x ²	p
Age, mean (IQR), year	76.23 (72.07, 80.39)	71.29 (69.96, 72.61)	2.167	0.032
BMI, mean (IQR), kg/m²	22.19 (20.86, 23.52)	23.00 (22.67, 23.34)	−1.387	0.167
Sex, n (%)
Male	9 (69.2)	116 (82.9)	0.706	0.401
Female	4 (30.8)	24 (17.1)	0.706	0.401
Diabetes, n (%)
Yes	1 (7.7)	10 (7.1)	0.000	1.000
No	12 (92.3)	130 (92.9)	0.000	1.000
Hypertension, n (%)
Yes	7 (53.8)	48 (34.3)	1.218	0.27
No	6 (46.2)	92 (65.7)	1.218	0.27
History of smoking, n (%)
Yes	5 (38.5)	71 (50.7)	0.714	0.398
No	8 (61.5)	69 (49.3)	0.714	0.398
History of alcohol consumption, n (%)
Yes	4 (30.8)	67 (47.9)	1.397	0.237
No	9 (69.2)	73 (52.1)	1.397	0.237
Atorvastatin, n (%)
Yes	4 (30.8)	69 (49.3)	1.635	0.201
No	9 (69.2)	71 (50.7)	1.635	0.201
Antiplatelet, n (%)
Yes	1 (7.7)	5 (3.6)	0.000	1.000
No	12 (92.3)	135 (96.4)	0.000	1.000
Hypophosphatemia, n (%)
Yes	0 (0)	9 (6.4)	0.106	0.744
No	13 (100)	131 (93.6)	0.106	0.744
Hypokalemia, n (%)
Yes	0 (0)	18 (12.9)	0.858	0.354
No	13 (100)	122 (87.1)	0.858	0.354
Hypocalcemia, n (%)
Yes	1 (7.7)	5 (3.6)	0.000	1.000
No	12 (92.3)	135 (96.4)	0.000	1.000
Hyponatremia, n (%)
Yes	4 (30.8)	19 (13.6)	0.000	1.000
No	9 (69.2)	121 (86.4)	0.000	1.000
Location of hematoma, n (%)
Left	7 (53.8)	65 (46.4)	1.119	0.594
Right	3 (23.1)	51 (36.4)
Bilateral	3 (23.1)	24 (17.1)
Urokinase application, n (%)
Postoperative	10 (76.9)	106 (75.7)	1.098	0.529
Intraoperative + postoperative	0 (0)	12 (8.6)
No	3 (23.1)	22 (15.7)

BMI: body mass index; IQR: interquartile range.

Table 4.

Comparison of clinical data between the recurrence and nonrecurrence groups.

Variable	Recurrence (n = 33)	Nonrecurrence (n = 120)	t/x ²	p
Age, mean (IQR), year	73.76 (71.06, 76.45)	71.14 (69.69, 72.59)	1.681	0.095
BMI, mean (IQR), kg/m²	22.25 (21.74, 22.76)	23.12 (22.74, 23.51)	−2.764	0.007
Sex, n (%)
Male	28 (84.8)	97 (80.8)	0.279	0.597
Female	5 (15.2)	23 (19.2)	0.279	0.597
Diabetes, n (%)
Yes	2 (6.1)	9 (7.5)	0.000	1.000
No	31 (93.9)	111 (92.5)	0.000	1.000
Hypertension, n (%)
Yes	12 (36.4)	43 (35.8)	0.003	0.955
No	21 (63.6)	77 (64.2)	0.003	0.955
History of smoking, n (%)
Yes	19 (57.6)	57 (47.5)	1.051	0.305
No	14 (42.4)	63 (52.5)	1.051	0.305
History of alcohol consumption, n (%)
Yes	15 (45.5)	56 (46.7)	0.015	0.902
No	18 (54.5)	64 (53.3)	0.015	0.902
Atorvastatin, n (%)
Yes	16 (48.5)	57 (47.5)	0.01	0.920
No	17 (51.5)	63 (52.5)	0.01	0.920
Antiplatelet, n (%)
Yes	0 (0)	6 (5.0)	0.647	0.421
No	33 (100)	114 (95.0)	0.647	0.421
Hypophosphatemia, n (%)
Yes	1 (3.0)	8 (6.7)	0.136	0.712
No	32 (97.0)	112 (93.3)	0.136	0.712
Hypokalemia, n (%)
Yes	4 (12.1)	14 (11.7)	0.000	1.000
No	29 (87.9)	106 (88.3)	0.000	1.000
Hypocalcemia, n (%)
Yes	1 (3.0)	5 (4.2)	0.000	1.000
No	32 (97.0)	115 (95.8)	0.000	1.000
Hyponatremia, n (%)
Yes	7 (21.2)	16 (13.3)	0.717	0.397
No	26 (78.8)	104 (86.7)	0.717	0.397
Location of hematoma, n (%)
Left	14 (42.4)	58 (48.3)	0.399	0.819
Right	13 (39.4)	41 (34.2)
Bilateral	6 (18.2)	21 (17.5)
Urokinase application, n (%)
Postoperative	24 (72.7)	92 (76.7)	1.245	0.547
Intraoperative + postoperative	4 (12.1)	8 (6.7)
No	5 (15.2)	20 (16.7)

BMI: body mass index; IQR: interquartile range.

Comparison of laboratory indicators

Compared with the good prognosis group, ALB and HB levels decreased and fibrinogen level increased in the poor prognosis group (p < 0.05); compared with the nonrecurrence group, ALB and HB levels decreased, and percentage of monocytes increased in the recurrence group (p < 0.05) (Tables 5 and 6).

Table 5.

Comparison of laboratory indicators between the good and poor prognosis groups.

Variable	Poor prognosis	Good prognosis	t/x²	p
Blood glucose (IQR), mmol/L	6.26 (5.27, 7.25)	6.66 (6.30, 7.02)	−0.65	0.517
Serum ALB (IQR), g/L	36.07 (32.56, 39.58)	40.80 (40.17, 41.43)	−2.877	0.013
Total cholesterol (IQR), mmol/L	3.68 (2.96, 4.39)	3.93 (3.79, 4.08)	−0.765	0.458
Triglycerides (IQR), mmol/L	1.05 (0.83, 1.27)	1.44 (1.26, 1.62)	−1.267	0.207
High-density lipoprotein (IQR), mmol/L	1.08 (0.80, 1.36)	1.14 (1.09.1.19)	−0.443	0.665
HB (IQR), g/L	126.38 (119.63, 133.14)	137.99 (135.62, 140.35)	−2.867	0.005
Lymphocyte (IQR), 10⁹/L	1.19 (0.94, 1.43)	1.45 (1.36, 1.53)	−1.758	0.081
Neutrophil (IQR), 10⁹/L	6.03 (4.08, 7.98)	4.80 (4.47, 5.13)	1.345	0.202
PLT (IQR), 10⁹/L	203.54 (166.55, 244.52)	224.30 (211.40, 234.20)	−1.194	0.234
Percentage of monocytes % (IQR)	7.64 (5.82, 9.46)	7.28 (6.84, 7.72)	0.465	0.642
Fibrinogen (IQR), g/L	4.21 (3.36, 5.06)	3.15 (3.01, 3.29)	2.687	0.019

ALB: albumin; HB: hemoglobin; PLT: platelet; IQR: interquartile range.

Table 6.

Comparison of laboratory indicators between the recurrence and nonrecurrence groups.

Variable	Recurrence	Nonrecurrence	t/x²	p
Blood glucose (IQR), mmol/L	6.41 (5.78, 7.04)	6.68 (6.29, 7.08)	−0.652	0.515
Serum ALB (IQR), g/L	38.15 (36.63, 39.67)	41.01 (40.30, 41.72)	−3.621	<0.001
Total cholesterol (IQR), mmol/L	3.66 (3.38.3.93)	3.99 (3.82, 4.15)	−1.859	0.065
Triglycerides (IQR), mmol/L	1.14 (0.95, 1.34)	1.48 (1.27, 1.69)	−1.651	0.101
High-density lipoprotein (IQR), mmol/L	1.12 (1.01, 1.22)	1.14 (1.08, 1.20)	−0.379	0.705
HB (IQR), g/L	131.55 (127.15, 135.94)	138.50 (135.88, 141.11)	−2.52	0.013
Lymphocyte (IQR), 10⁹/L	1.36 (1.16, 1.56)	1.44 (1.35, 1.53)	−0.808	0.42
Neutrophil (IQR), 10⁹/L	4.94 (4.11, 5.78)	4.90 (4.52, 5.27)	0.103	0.918
PLT (IQR), 10⁹/L	209.45 (188.73, 230.18)	226.13 (215.25, 237.02)	−1.418	0.158
Percentage of monocytes (IQR), %	8.26 (7.37, 9.14)	7.05 (6.57, 7.53)	2.357	0.02
Fibrinogen (IQR), g/L	3.26 (2.91, 3.59)	3.23 (3.07, 3.40)	0.117	0.907

ALB: albumin; HB: hemoglobin; PLT: platelet; IQR: interquartile range.

Value of GNRI, PNI, and NRS 2002 in predicting the postoperative prognosis of CSDH

ROC analysis²⁸ showed that the area under the curve (AUC) of the GNRI for predicting postoperative prognosis of CSDH was 0.768 (95% confidence interval (CI): 0.637–0.899, p = 0.047). The maximum Youden’s index²⁹ was achieved at a GNRI cutoff value of 98.59, yielding a sensitivity of 79.3% and specificity of 69.2%. Regarding the PNI, the AUC was 0.762 (95% CI: 0.619–0.904, p = 0.002), with an optimal cutoff value of 42.93, corresponding to a sensitivity of 87.1% and specificity of 61.5%. The NRS 2002 yielded an AUC of 0.803 (95% CI: 0.687–0.920, p < 0.001) for predicting surgical prognosis, with a maximum Youden’s index achieved at a score of 2.5, demonstrating a sensitivity of 92.3% and specificity of 65.0%. Among the three indices, the GNRI demonstrated the highest specificity and lowest false-positive rate (Figure 1 and Table 7).

Figure 1.

ROC curves for the GNRI, NRS 2002, PNI to predict postoperative prognosis of CSDH.

Table 7.

Comparison of sensitivity, specificity, false-positive rate, and false-negative rate among three nutritional risk screening methods for predicting prognosis.

Screening tool	Nutritional risks	Poor prognosis, n (%)	Good prognosis, n (%)	Sensitivity, %	Specificity, %	False-positive rate, %	False-negative rate, %	x²	p
GNRI	Yes	6 (46.2)	26 (18.6)	79.3	69.2	30.8	20.7	3.931	0.047
	No	7 (53.8)	114 (81.4)
PNI	Yes	9 (69.2)	35 (25.0)	87.1	61.5	38.5	12.9	9.303	0.002
	No	4 (30.8)	105 (75.0)
NRS 2002	Yes	12 (92.3)	49 (35.0)	92.3	65.0	35.0	7.7	16.29	<0.001
	No	1 (7.7)	91 (65.0)

GNRI: Geriatric Nutritional Risk Index; PNI: Prognostic Nutritional Index; NRS 2002: Nutritional Risk Screening 2002.

Multivariate analysis of poor prognosis after surgery for CSDH

Nutrition-related factors with a significance level of p < 0.1 in the univariate analysis of clinical information and laboratory tests, including age, ALB and HB levels, lymphocyte count, GNRI, PNI, and NRS 2002, were incorporated into a multivariate logistic regression analysis. After adjusting for age, ALB and HB levels, and lymphocyte count, the multivariate logistic regression analysis indicated that the GNRI (odds ratio (OR) = 1.141, 95% CI: 1.008–1.292), PNI (OR = 1.286, 95% CI: 1.001–1.651), and NRS 2002 (OR = 14.878, 95% CI: 1.766–125.349) were predictors of poor prognosis after surgery for CSDH (Tables 8 to 10).

Table 8.

Multivariate logistic regression analysis of the association between GNRI and prognosis.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	−0.046	0.041	1.242	0.265	0.955 (0.882–1.035)
ALB	−0.713	0.854	0.697	0.404	0.490 (0.092–2.613)
HB	0.499	0.916	0.296	0.586	1.646 (0.274–9.909)
Lymphocyte count	0.440	0.753	0.342	0.559	1.553 (0.355–6.798)
GNRI	0.132	0.063	4.322	0.038	1.141 (1.008–1.292)

ALB: albumin; HB: hemoglobin; GNRI: Geriatric Nutritional Risk Index; OR: odds ratio; CI: confidence interval; SE: standard error.

Table 10.

Multivariate logistic regression analysis of the association between NRS 2002 and prognosis.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	−0.022	0.043	0.255	0.613	0.979 (0.900–1.064)
ALB	−1.288	0.714	3.256	0.071	0.276 (0.068–1.117)
HB	−1.653	1.128	2.147	0.143	0.351 (0.064–1.934)
Lymphocyte count	−1.048	0.871	1.448	0.229	0.448 (0.066–3.019)
NRS 2002	2.700	1.087	6.165	0.013	14.878 (1.766–125.349)

ALB: albumin; HB: hemoglobin; NRS 2002: Nutritional Risk Screening 2002; OR: odds ratio; CI: confidence interval; SE: standard error.

Table 9.

Multivariate logistic regression analysis of the association between PNI and prognosis.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	−0.062	0.041	2.296	0.130	0.940 (0.867–1.018)
ALB	−0.244	1.005	0.059	0.808	0.783 (0.109–5.617)
HB	0.525	0.921	0.325	0.569	1.690 (0.278–10.277)
Lymphocyte count	−0.803	0.974	0.681	0.409	0.448 (0.066–3.019)
PNI	0.251	0.128	3.872	0.049	1.286 (1.001–1.651)

ALB: albumin; HB: hemoglobin; PNI: Prognostic Nutritional Index; OR: odds ratio; CI: confidence interval; SE: standard error.

Value of GNRI, PNI, and NRS 2002 in predicting postoperative recurrence of CSDH

ROC analysis indicated that the AUC of the GNRI for predicting postoperative recurrence of CSDH was 0.753 (95% CI: 0.659–0.847, p < 0.001). The maximum Youden’s index was achieved at a GNRI cutoff value of 100.05, resulting in a sensitivity of 75.8% and specificity of 69.7%. Regarding the PNI, the AUC was 0.666 (95% CI: 0.560–0.772, p = 0.001), with an optimal cutoff value of 44.41, corresponding to a sensitivity of 80.0% and specificity of 51.5%. The NRS 2002 yielded an AUC of 0.734 (95% CI: 0.636–0.832, p < 0.001) for predicting recurrence, with the highest Youden’s index achieved at a score of 2.5, showing a sensitivity of 78.8% and specificity of 70.8%. Among the three tools, the NRS 2002 exhibited the highest specificity and lowest false-positive rate (Figure 2 and Table 11).

Figure 2.

ROC curve for the GNRI, NRS 2002, PNI to predict postoperative recurrence of CSDH.

Table 11.

Comparison of sensitivity, specificity, false-positive rate, and false-negative rate among three nutritional risk screening methods for predicting postoperative recurrence.

Screening tool	Nutritional risks	Recurrence, n (%)	Nonrecurrence, n (%)	Sensitivity, %	Specificity, %	False-positive rate, %	False-negative rate, %	x ²	p
GNRI	Yes	14 (42.4)	18 (15.0)	75.8	69.7	30.3	24.2	11.768	<0.001
	No	19 (57.6)	102 (85.0)
PNI	Yes	17 (51.5)	27 (22.5)	80.0	51.5	48.5	20.0	10.635	0.001
	No	16 (48.5)	93 (77.5)
NRS 2002	Yes	26 (78.8)	35 (29.2)	78.8	70.8	29.2	21.2	26.583	<0.001
	No	7 (21.2)	85 (70.8)

GNRI: Geriatric Nutritional Risk Index; PNI: Prognostic Nutritional Index; NRS 2002: Nutritional Risk Screening 2002.

Multivariate analysis of postoperative recurrence in CSDH

Nutrition-related factors with a significance level of p < 0.1 in the univariate analysis of clinical information and laboratory tests, including age, total cholesterol, BMI, ALB and HB levels, PNI, NRS 2002, and GNRI, were included in the multivariate logistic regression analysis. After adjusting for age, total cholesterol, BMI, and ALB and HB levels, the multivariate logistic regression analysis showed that the GNRI (OR = 0.167, 95% CI: 0.059–0.475), PNI (OR = 1.142, 95% CI: 1.024–1.275), and NRS 2002 (OR = 7.751, 95% CI: 2.672–22.485) were predictors of postoperative recurrence of CSDH (Tables 12 to 14).

Table 12.

Multivariate logistic regression analysis of the association between GNRI and recurrence.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	−0.004	0.028	0.020	0.888	0.996 (0.942–1.053)
Total cholesterol	0.145	0.270	0.290	0.590	1.156 (0.681–1.963)
BMI	−0.418	0.516	0.655	0.418	0.658 (0.239–1.811)
ALB	0.129	0.748	0.030	0.863	0.138 (0.263–4.930)
HB	0.375	0.664	0.319	0.573	0.455 (0.396–5.344)
GNRI	−1.790	0.534	11.260	0.001	0.167 (0.059–0.475)

BMI: body mass index; ALB: albumin; HB: hemoglobin; GNRI: Geriatric Nutritional Risk Index; OR: odds ratio; CI: confidence interval; SE: standard error.

Table 14.

Multivariate logistic regression analysis of the association between NRS 2002 and recurrence.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	0.021	0.031	0.450	0.502	1.021 (0.961–1.085)
Total cholesterol	0.098	0.292	0.112	0.737	1.103 (0.622–1.954)
BMI	−0.406	0.558	0.529	0.467	0.666 (0.223–1.990)
ALB	−1.621	0.590	7.559	0.006	0.198 (0.062–0.628)
HB	−0.412	0.510	0.654	0.419	0.662 (0.244–1.798)
NRS 2002	2.048	0.543	14.203	<0.001	7.751 (2.672–22.485)

BMI: body mass index; ALB: albumin; HB: hemoglobin; NRS 2002: Nutritional Risk Screening 2002; OR: odds ratio; CI: confidence interval; SE: standard error.

Table 13.

Multivariate logistic regression analysis of the association between PNI and recurrence.

	β	SE	Wald χ²	p value	OR (95% CI)
Age	−0.006	0.028	0.044	0.835	0.994 (0.941–1.050)
Total cholesterol	0.266	0.259	1.056	0.304	1.305 (0.786–2.167)
BMI	−1.086	0.468	5.387	0.020	0.338 (0.135–0.845)
ALB	0.682	0.875	0.607	0.436	1.977 (0.356–10.991)
HB	0.682	0.875	0.607	0.436	1.385 (0.379–5.056)
PNI	−1.373	0.491	7.837	0.005	1.142 (1.024–1.275)

BMI: body mass index; ALB: albumin; HB: hemoglobin; PNI: Prognostic Nutritional Index; OR: odds ratio; CI: confidence interval; SE: standard error.

Discussion

CSDH is a common neurological condition. With progressive aging of the population, its current incidence ranges from 1.7 to 20.6 per 100,000 people and is increasing annually.^1,30 As the hematoma expands and the disease progresses, clinical manifestations vary widely and may include nausea, vomiting, dizziness, headache, limb motor deficits, speech disturbances, seizures, impaired consciousness, and even death. For patients with severe symptoms, burr-hole drainage is currently the most frequently performed surgical intervention.³¹ However, high recurrence rates, which can range from 5% to 30%, pose a significant clinical challenge. Recurrence often leads to prolonged hospitalization or readmission, increases the economic burden on patients, and is associated with poor clinical outcomes. Postoperative recurrence of CSDH is influenced by multiple factors, including patient characteristics, choice of anesthesia, surgical technique, and medication-related issues.^32–34 Previous studies have demonstrated a strong correlation between nutritional status (and associated scores) and clinical outcomes across various neurosurgical conditions, including traumatic craniofacial injuries, gliomas,^35,36 spinal surgery in older adults, and subarachnoid hemorrhage.^37,38 Both CSDH and malnutrition are common clinical conditions among the older population. In a prospective study, Scerrati et al.³⁹ identified a correlation between malnutrition and poor postoperative prognosis in patients undergoing surgery for CSDH, as assessed through detailed micronutritional evaluation. Therefore, this study conducted a retrospective analysis of clinical data obtained from patients with CSDH at our hospital in recent years, aiming to investigate the risk factors for surgical prognosis and recurrence of CSDH and explore the association of the GNRI, PNI, and NRS 2002 with poor postoperative outcomes and recurrence in these patients.

Serum ALB level is commonly used in clinical practice as a laboratory indicator to assess patient nutritional status. In recent years, it has been recognized that serum ALB levels can serve as a predictor of postoperative complications and mortality following surgical procedures.⁴⁰ Malnutrition may exacerbate neuroinflammatory responses by inhibiting the expression of axonal terminal proteins, thereby impairing brain functional recovery.⁴¹ This study confirmed that serum ALB levels were significantly lower in the recurrence group than in the nonrecurrence group. These findings suggest that lower serum ALB levels are associated with a higher nutritional risk and an increased rate of postoperative recurrence. Among 153 patients who underwent surgery for CSDH in this study, 33 (21.6%) experienced recurrence within 6 months. According to GNRI assessments, 42.4% of patients in the recurrence group and 15% in the nonrecurrence group were classified as malnourished. Well-nourished patients exhibited a lower recurrence rate than malnourished or at-risk patients. In the poor prognosis group, six patients (46.2%) exhibited nutritional risk, whereas only twenty-six patients (18.6%) in the good prognosis group exhibited nutritional risk. These differences were statistically significant, indicating that malnourished nutritional status is associated with a higher rate of poor prognosis.

The PNI is a nutritional indicator that reflects the body’s inflammatory and immune status. It is calculated using serum ALB levels and lymphocyte counts, both of which can be easily obtained from routine laboratory tests, making PNI a simple and practical tool. Inflammation plays a critical role in CSDH formation, and the PNI has been shown to possess prognostic value in several diseases.²³ HB, as the primary oxygen-carrying molecule in the body, has also been reported to have prognostic significance; elevated HB levels have been associated with improved outcomes in malignant tumors.^42,43 Although the predictive value of the PNI and HB levels in recurrence and prognosis following CSDH has not been widely reported, our study revealed several significant findings. According to PNI-based nutritional assessment, malnutrition was identified in 51.5% of the patients in the recurrence group, compared with 22.5% in the nonrecurrence group, suggesting that malnourished patients are at higher risk of recurrence after CSDH surgery. Furthermore, nutritional risk was observed in 69.2% of patients in the poor prognosis group versus 25% in the good prognosis group, indicating that malnutrition is associated with a higher likelihood of poor postoperative outcomes. Additionally, HB levels were significantly lower in both recurrence and poor prognosis groups than in their respective control groups. This suggests that higher HB levels exert a protective effect against postoperative recurrence of CSDH.

Sensitivity and specificity are widely utilized for comparing diagnostic tests, assessing the validity of methods, and conducting clinical evaluations. The results of this study demonstrated that the GNRI, NRS 2002, and PNI exhibit high sensitivity in nutritional risk screening among patients hospitalized for surgical treatment of CSDH. In predicting postoperative prognosis, the GNRI showed the highest specificity (69.2%) and lowest false-positive rate (30.8%). In contrast, the PNI and NRS 2002 were associated with higher false-positive rates, indicating a greater likelihood of misdiagnosis in nutritional risk screening. Furthermore, the NRS 2002 yielded the largest AUC in the ROC analysis, suggesting a strong correlation between nutritional risk identified using the NRS 2002 and poor prognosis in CSDH patients. The NRS 2002 demonstrated the highest specificity (70.8%) among the three tools in predicting postoperative recurrence, while the PNI showed a high false-positive rate (48.5%), implying a considerable misdiagnosis rate. The GNRI achieved the largest AUC value in the ROC curve, supporting its utility in establishing a strong association between nutritional risk and recurrence in CSDH patients. Furthermore, multivariate analysis indicated that the GNRI, PNI, and NRS 2002 are independent predictors of both prognosis and recurrence.

Malnutrition represents a systemic state of frailty that compromises a patient’s fundamental physiological capacity to cope with disease and trauma, including surgery; it is more likely to affect older patients.⁴⁴ It is primarily characterized by diminished tissue repair and healing ability, impaired immune function accompanied with chronic inflammation, decreased plasma ALB levels that disrupt internal homeostasis and metabolic waste clearance, and reduced colloidal osmotic pressure leading to cerebral edema, which may impede postoperative brain re-expansion and hinder closure of the hematoma cavity.^45,46 Through these mechanisms, malnutrition collectively contributes to an increased risk of hematoma recurrence and poorer overall prognosis in CSDH patients who have undergone surgery. The three nutritional risk screening scores (GNRI, PNI, and NRS 2002) can be obtained using admission clinical data and laboratory tests. These tools are suitable for routine nutritional risk screening in hospitalized patients; they help determine whether the nutritional status of a patient is adequate to tolerate surgery and provide a basis for nutritional support therapy. Those identified at nutritional risk warrant enhanced perioperative surveillance and systematic follow-up. Perioperative nutritional support (including protein supplementation or enteral feeding) should be considered in malnourished patients to improve outcomes and reduce recurrence. Further studies are needed to establish the effectiveness of tailored nutritional interventions in this population.

Limitations of the study

This study has several limitations, including a relatively small sample size, single-center design, and constraints inherent to a retrospective study. These factors may have introduced selection bias, made the conclusions vulnerable to extreme values, and led to an underestimation of risk among lost-to-follow-up cases. Additionally, the study did not assess other markers related to immune–nutritional status, such as C-reactive protein or globulin levels. Only a single assessment of the PNI, NRS 2002, and GNRI was conducted at admission, without tracking dynamic changes. This may have compromised accurate evaluation of their impact on outcomes. The relatively short follow-up period also restricts prognostic evaluation, hindering the assessment of long-term efficacy, determination of whether the observed positive outcomes persist over time, or evaluation of potential long-term recurrence risks. Furthermore, although objective assessment tools such as PNI and GNRI offer certain advantages in nutritional risk screening, they should be combined with subjective measures such as the NRS 2002 to comprehensively evaluate nutritional risk in patients with CSDH. The analysis of recurrence risk is limited by the omission of several pertinent pre- and postoperative radiographic factors, such as hematoma density and the extent of postoperative subdural air, which are potential predictors of recurrence. Additionally, the accurate assessment of nutritional status is inherently constrained by the retrospective study design; precise evaluation would require prospective cohort studies implementing standardized nutritional risk screening protocols. Therefore, future research should involve larger sample sizes, multicenter prospective designs, and long-term follow-ups to validate the findings of this study and further refine the methods and indicator system for nutritional risk screening in this patient population.

Conclusion

Admission GNRI, PNI, and NRS 2002 scores are closely associated with poor postoperative prognosis and recurrence in CSDH patients. This facilitates the early identification of nutritional risk in CSDH patients and enables subsequent implementation of personalized enteral or parenteral nutrition support protocols to reduce the rates of poor postoperative outcomes and recurrence.

Footnotes

Acknowledgements

We are deeply grateful to our colleagues for contributing to this thesis in various ways.

Authors’ contribution

Yu Tian: Conceived and designed the study, performed the analysis, wrote the manuscript, and verified the accuracy of the study.

Ling Li: Reviewed and provided final approval of the manuscript.

Shuang Sun, Hui Jiang, and Huisong Chu: Contributed to conceptual design of the study, edited the manuscript, collected and compiled the data, and provided final approval of the manuscript.

Data availability statement

The dataset presented in the study is available on request from the corresponding author during submission or after publication.

Declaration of conflicting interests

The authors declare that they have no competing interests.

Ethical approval

This study was approved by the Ethics Committee of the Affiliated Hospital of Chengde Medical University (NO.CYFYLL2023415).

Funding

This study was supported by the Medical Science Research Project of Hebei (NO:20242041).

ORCID iD

Yu Tian

References

Feghali

Yang

Huang

Updates in chronic subdural hematoma: epidemiology, etiology, pathogenesis, treatment, and outcome. World Neurosurg 2020; 141: 339–345.

Adhiyaman

Chatterjee

Increasing incidence of chronic subdural haematoma in the elderly. QJM 2017; 110: 775.

Amaral

Filho

Fernandes

, et al. Effectiveness of invasive interventions for chronic subdural hematoma: a systematic review. J Neurosurg Sci 2025; 69: 200–209.

Cai

Chen

, et al. Being at risk of malnutrition predicts poor outcomes at 3 months in acute ischemic stroke patients. Eur J Clin Nutr 2020; 74: 796–805.

Watanabe

Horiuchi

Nakase

, et al. Malnutrition, hemodynamics and inflammation in heart failure with reduced, mildly reduced and preserved ejection fraction. Heart Vessels 2022; 37: 1841–1849.

Chen

A novel nomogram based on the nutritional risk screening 2002 score to predict survival in hepatocellular carcinoma treated with transarterial chemoembolization. Nutr Hosp 2022; 39: 835–842.

Bouillanne

Morineau

Dupont

, et al. Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr 2005; 82: 777–783.

Fujiwara

Sato

Hayashi

, et al. The Geriatric Nutritional Risk Index as a prognostic factor in older adult patients with locally advanced head and neck cancer receiving definitive chemoradiotherapy with tri-weekly cisplatin. J Geriatr Oncol 2023; 14: 101523.

Dong

Chen

Zeng

, et al. Geriatric nutritional risk index predicts all-cause mortality in patients with heart failure: a systematic review and meta-analysis. Clinics (Sao Paulo) 2021; 76: e2258.

10.

Kregel

Murphy

Attia

, et al. The Geriatric Nutritional Risk Index as a predictor of complications in geriatric trauma patients. J Trauma Acute Care Surg 2022; 93: 195–199.

11.

Kondrup

Rasmussen

Hamberg

; Ad Hoc ESPEN Working Groupet al. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 2003; 22: 321–336.

12.

Hersberger

Bargetzi

, et al. Nutritional risk screening (NRS 2002) is a strong and modifiable predictor risk score for short-term and long-term clinical outcomes: secondary analysis of a prospective randomised trial. Clin Nutr 2020; 39: 2720–2729.

13.

Zang

Qiu

, et al. Association between risk of malnutrition defined by the Nutritional Risk Screening 2002 and postoperative complications and overall survival in patients with cancer: a meta-analysis. Nutr Cancer 2023; 75: 1600–1609.

14.

Cortés-Aguilar

Malih

Abbate

, et al. Validity of nutrition screening tools for risk of malnutrition among hospitalized adult patients: a systematic review and meta-analysis. Clin Nutr 2024; 43: 1094–1116.

15.

Kondrup

Nutritional-risk scoring systems in the intensive care unit. Curr Opin Clin Nutr Metab Care 2014; 17: 177–182.

16.

Onodera

Goseki

Kosaki

Prognostic nutritional index in gastrointestinal surgery of malnourished cancer patients. Nihon Geka Gakkai Zasshi 1984; 85: 1001–1005.

17.

Zhang

Qiu

, et al. Prognostic nutritional index as a prognostic biomarker for gastrointestinal cancer patients treated with immune checkpoint inhibitors. Front Immunol 2023; 14: 1219929.

18.

Zhang

Liu

Influence of Prognostic Nutritional Index on the surveillance after surgery-based systematic therapy for breast cancer. Am Surg 2023; 89: 6157–6171.

19.

Huang

Ding

, et al. Prognostic Nutritional Index predicts response and prognosis in cancer patients treated with immune checkpoint inhibitors: a systematic review and meta-analysis. Front Nutr 2022; 9: 823087.

20.

Shen

Liu

, et al. Prognostic value of pretreatment prognostic nutritional index in non-small cell lung cancer: a systematic review and meta-analysis. Int J Biol Markers 2018; 33: 372–378.

21.

Guo

Yang

Liu

, et al. Prognostic nutritional index and the risk of postoperative complications after spine surgery: a meta-analysis. World Neurosurg 2024; 185: e572–e581.

22.

Hung

Sun

Chang

, et al. Association of prognostic nutritional index with prognostic outcomes in patients with glioma: a meta-analysis and systematic review. Front Oncol 2023; 13: 1188292.

23.

Zhu

Guo

, et al. A low prognostic nutritional index is associated with chronic subdural hematoma recurrence. Neurosurg Rev 2023; 46: 139.

24.

Von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

25.

Broderick

Adeoye

Elm

Evolution of the modified Rankin Scale and its use in future stroke trials. Stroke 2017; 48: 2007–2012.

26.

Hori

Ebisudani

Aoi

, et al. Elevated serum fibrinogen degradation products on admission is a novel predictive factor for recurrence of chronic subdural hematoma. World Neurosurg 2018; 118: e753–e757.

27.

Peng

Chen

Shen

, et al. Prognostic Nutritional Index (PNI) and Controlling Nutritional Status (CONUT) score for predicting outcomes of breast cancer: a systematic review and meta-analysis. Pak J Med Sci 2023; 39: 1535–1541.

28.

Linden

Measuring diagnostic and predictive accuracy in disease management: an introduction to receiver operating characteristic (ROC) analysis. J Eval Clin Pract 2006; 12: 132–139.

29.

Fluss

Faraggi

Reiser

Estimation of the Youden index and its associated cutoff point. Biom J 2005; 47: 458–472.

30.

Maroufi

Farahbakhsh

Macdonald

, et al. Risk factors for recurrence of chronic subdural hematoma after surgical evacuation: a systematic review and meta-analysis. Neurosurg Rev 2023; 46: 270.

31.

Yang

Huang

Chronic subdural hematoma: epidemiology and natural history. Neurosurg Clin N Am 2017; 28: 205–210.

32.

Zhu

Wang

, et al. Factors correlated with the postoperative recurrence of chronic subdural hematoma: an umbrella study of systematic reviews and meta-analyses. EClinicalMedicine 2022; 43: 101234.

33.

Seo

Choo

, et al. Clinical characteristics and current managements for patients with chronic subdural hematoma: a retrospective multicenter pilot study in the Republic of Korea. J Korean Neurosurg Soc 2022; 65: 255–268.

34.

Jiang

Zhao

Wang

, et al. Safety and efficacy of atorvastatin for chronic subdural hematoma in Chinese patients: a randomized clinicaltrial. JAMA Neurol 2018; 75: 1338–1346.

35.

Liu

Luo

, et al. Controlling nutritional status score and prognostic nutrition index predict the outcome after severe traumatic brain injury. Nutr Neurosci 2022; 25: 690–697.

36.

Huq

Khalafallah

Botros

, et al. The prognostic impact of nutritional status on postoperative outcomes in glioblastoma. World Neurosurg 2021; 146: e865–e875.

37.

Acarbaş

Baş

NS.

Which objective nutritional index is better for the prediction of adverse medical events in elderly patients undergoing spinal surgery?

World Neurosurg 2021; 146: e106–e111.

38.

Yang

Hao

, et al. Clinical value of controlling nutritional status score in patients with aneurysmal subarachnoid hemorrhage. World Neurosurg 2019; 126: e1352–e1358.

39.

Scerrati

Pangallo

Dughiero

, et al. Influence of nutritional status on the clinical outcome of patients with chronic subdural hematoma: a prospective multicenter clinical study. Nutr Neurosci 2022; 25: 1756–1763.

40.

Jiang

Zhu

, et al. Value of postoperative serum albumin to predict postoperative complication severity in spinal tuberculosis. Biomed Res Int 2022; 2022: 4946848.

41.

Smith

Figley

Schreyer

, et al. Protein-energy malnutrition developing after global brain ischemia induces an atypical acute-phase response and hinders expression of GAP-43. PLoS One 2014; 9: e107570.

42.

Wang

Zhang

, et al. The effect of albumin and hemoglobin levels on the prognosis of early-stage cervical cancer: a prospective, single-center-based cohort study. BMC Womens Health 2023; 23: 553.

43.

Huang

Yang

Chen

, et al. Preoperative anemia or low hemoglobin predicts poor prognosis in gastric cancer patients: a meta-analysis. Dis Markers 2019; 2019: 7606128.

44.

Díaz de Bustamante

Alarcón

Menéndez-Colino

, et al. Prevalence of malnutrition in a cohort of 509 patients with acute hip fracture: the importance of a comprehensive assessment. Eur J Clin Nutr 2018; 72: 77–81.

45.

Ghaly

Iliopoulos

Ahmad

The role of nutrition in wound healing: an overview. Br J Nurs 2021; 30: S38–S42.

46.

Chiang

Chao

Chu

, et al. Early enteral nutrition and clinical outcomes of severe traumatic brain injury patients in acute stage: a multi-center cohort study. J Neurotrauma 2012; 29: 75–80.