The prognostic role of prognostic nutritional score, geriatric nutritional risk index, and controlling nutritional status score in critically ischemic geriatric patients with acute ischemic stroke: A retrospective cohort study

Abstract

Acute ischemic stroke (AIS) is a major cause of functional disability and mortality in the geriatric population. This study aimed to evaluate the predictive performance of the prognostic nutritional index (PNI), geriatric nutritional risk index (GNRI), and the controlling nutritional status (CONUT) score for in-hospital mortality among critically ill geriatric patients. Critically ill AIS patients admitted to a tertiary hospital intensive care unit from January 2021 to January 2023 were retrospectively analyzed. Patients were classified into survivor and mortality groups. Nutritional scores were calculated: PNI = (10 × serum albumin [g/dL]) + (0.005 × total lymphocyte count); GNRI = (1.489 × albumin [g/dL]) + (41.7 × [body weight/ideal body weight]); CONUT (scored 0–12 based on serum lymphocyte count, cholesterol, and albumin). The predictive performances of the scores for in-hospital mortality were compared. A total of 142 patients were included, with 25 in the mortality group and 117 in the survivor group. The median age was 77 (range: 68–84) years, and 55.6% (n = 79) were female. Demographic characteristics (age, sex, body mass index) were similar between groups. Median PNI (36.3 vs 39.5, P = .017), GNRI (53.4 vs 56.6, P = .007), and CONUT (2 vs 2, P = .012) scores were significantly lower in the mortality group. Multivariate regression analysis showed that GNRI was an independent predictor of mortality (odds ratio = 0.935, 95% confidence interval = 0.877–0.998, P = .042). Receiver operating characteristic analyses showed a PNI cutoff value of ≤34.2 (area under curve [AUC] 0.653, 0.535–0.770), GNRI ≤56.2 (AUC 0.672, 0.577–0.767), and CONUT ≥1.5 (AUC 0.659, 0.545–0.773). Nutritional scores such as PNI, GNRI, and CONUT can predict mortality in critically ill geriatric AIS patients in the intensive care unit. Their prognostic performances were found to be similar.

1. Introduction

Stroke (cerebrovascular disease) is one of the leading causes of morbidity and mortality worldwide, particularly in the geriatric population. Globally, stroke represents the third most common cause of death.^[1] According to data from the Global Burden of Disease studies, approximately 7.8 million new ischemic stroke cases were reported worldwide in 2021, showing a substantial increase since 1990.^[2] As highlighted in the Global Burden of Disease reports, stroke accounts for approximately 10% to 12% of all-cause mortality, and ischemic stroke alone is responsible for nearly 3.6 million deaths annually.^[3] Strokes are classified into 2 main categories: ischemic and hemorrhagic. More than 70% of all strokes are acute ischemic strokes (AIS).^[4–6] Following AIS, the severe reduction in cerebral blood flow compromises perfusion and oxygenation, ultimately leading to neuronal injury. Early identification and treatment of patients at risk for poor outcomes are critical for preventing adverse prognoses. For this purpose, fast, reliable, and easily accessible biomarkers are increasingly being used and developed to predict poor outcomes.

AIS tends to follow a more complex clinical course in geriatric patients, and the recovery process is significantly influenced by individual factors. Comorbidities that increase with age, the chronic progression of inflammatory processes, and the deterioration of nutritional status markedly elevate the risk of mortality and morbidity in this population.^[7–9] Factors such as sarcopenia, low body mass index, and malnutrition emerge as key contributors to poor clinical outcomes following stroke in geriatric individuals. A strong association has been established between nutritional deficiencies and complications that develop, particularly after acute-phase management.^[10] In this context, scoring systems that allow an objective assessment of nutritional status at the time of hospital admission are gaining increasing importance. Several studies have reported that the prognostic nutritional index (PNI), geriatric nutritional risk index (GNRI), and controlling nutritional status (CONUT) score not only reflect nutritional status and systemic inflammation but are also easy to apply and possess prognostic value regarding mortality.^[11–13] However, studies evaluating the prognostic impact of these scores specifically in critically ill geriatric patients with AIS remain limited.

This study aimed to investigate the prognostic value of the PNI, GNRI, and CONUT scores regarding mortality and functional outcomes in critically ill patients with AIS who were followed in the intensive care unit (ICU) of a tertiary hospital. We hypothesized that lower PNI and GNRI values and higher CONUT scores at ICU admission would be associated with increased mortality and poorer functional outcomes in critically ill geriatric patients with AIS.

2. Materials and methods

2.1. Ethics

This retrospective observational study was initiated following approval from the clinical trials review board and ethics committee (KAEK/2025.10.260). The study was conducted in accordance with the principles of the Declaration of Helsinki. The Research Ethics Committee waived informed consent from study participants due to the study’s retrospective nature and the use of de-identified data, thereby ensuring participant anonymity and confidentiality.

2.2. Study background and setting

Geriatric patients diagnosed with AIS who were treated and followed between January 2021 and January 2023 at the University of Health Sciences, Istanbul Kanuni Sultan Süleyman Training and Research Hospital were included. Patient data were obtained from follow-up records and the hospital’s electronic medical information system.

The inclusion criteria were as follows:

1. age ≥ 65 years;

2. first-time diagnosis of AIS;

3. presentation within 24 hours of symptom onset.

The exclusion criteria were:

1. history of infection, major trauma, or surgery within the 2 weeks preceding the AIS diagnosis;

2. presence of malignancy, autoimmune disease, or hematological disorders;

3. severe liver or kidney disease/failure;

4. use of medications that could affect platelet or lymphocyte counts (such as antiplatelet agents or glucocorticoids);

5. history of transient ischemic attack, cerebral infarction, intracranial hemorrhage, aneurysmal subarachnoid hemorrhage, or cerebral venous sinus thrombosis;

6. missing data.

According to the World Health Organization’s multinational monitoring of trends and determinants in cardiovascular diseases criteria, AIS is defined as a neurological deficit of cerebrovascular origin that persists for more than 24 hours or is interrupted by death within 24 hours.^[14] The diagnosis of AIS was confirmed by magnetic resonance imaging or computed tomography performed within 24 hours of hospital presentation.

Demographic data, comorbidities, need for mechanical ventilation, length of stay in the ICU and on mechanical ventilation, Glasgow Coma Scale (GCS) score, and acute physiology and chronic health evaluation II (APACHE II) score at ICU admission, as well as serum hemoglobin, albumin, lymphocyte, and cholesterol levels at admission, were recorded. Twenty-eight–day mortality and functional outcomes were also documented. PNI, GNRI, and CONUT scores were calculated using laboratory results obtained at admission. For patients who were discharged, 28-day mortality and modified Rankin Scale (mRS) scores were recorded.

2.2.1. PNI

The PNI is a simple parameter used to assess a patient’s nutritional status. It has been reported to be associated with postoperative morbidity, adverse clinical outcomes, and mortality in various malignancies, diseases, and geriatric trauma populations.^[13,15,16]

PNI is calculated using the following formula:

$${\displaystyle \begin{array}{lll}{\displaystyle \mathrm{P}\mathrm{N}\mathrm{I}}& =(10\times \mathrm{s}\mathrm{e}\mathrm{r}\mathrm{u}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{b}\mathrm{u}\mathrm{m}\mathrm{i}\mathrm{n}\left[\mathrm{g}/\mathrm{d}\mathrm{L}\right])& +(0.005\times \mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{y}\mathrm{m}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{c}\mathrm{y}\mathrm{t}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}).\end{array}}$$

2.2.2. GNRI

The GNRI is used to identify patients at risk of malnutrition-related morbidity and mortality. This score is calculated using serum albumin and ideal body weight (IBW) according to the following formula:

$${\displaystyle \begin{array}{lll}{\displaystyle \mathrm{G}\mathrm{N}\mathrm{R}\mathrm{I}}& =(1.489\times \mathrm{a}\mathrm{l}\mathrm{b}\mathrm{u}\mathrm{m}\mathrm{i}\mathrm{n}\left[\mathrm{g}/\mathrm{L}\right])& +(41.7\times \left[\mathrm{b}\mathrm{o}\mathrm{d}\mathrm{y}\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}/\mathrm{I}\mathrm{B}\mathrm{W}\right]).\end{array}}$$

GNRI has been reported to possess long-term prognostic value in geriatric trauma patients and to be useful in predicting postoperative cognitive impairment and complications.^[13,17]

In accordance with the literature, IBW in our study was calculated separately for men and women using the following formulas.^[18]

IBW was calculated using the Lorentz formula as follows:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{I}\mathrm{B}\mathrm{W}\left(\mathrm{M}\mathrm{e}\mathrm{n}\right)=\mathrm{H}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\left(\mathrm{c}\mathrm{m}\right)-100-\left([\mathrm{H}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\left(\mathrm{c}\mathrm{m}\right)-150]/4\right).}\end{array}}$$

$${\displaystyle \begin{array}{lll}{\displaystyle \mathrm{I}\mathrm{B}\mathrm{W}\left(\mathrm{W}\mathrm{o}\mathrm{m}\mathrm{e}\mathrm{n}\right)}& =\mathrm{H}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\left(\mathrm{c}\mathrm{m}\right)-100& -\left([\mathrm{H}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\left(\mathrm{c}\mathrm{m}\right)-150]/2.5\right).\end{array}}$$

2.2.3. CONUT score

The CONUT score is a simple, objective nutritional index used to evaluate patients’ nutritional status. It is based on 3 fundamental laboratory parameters:

1. Serum albumin level – reflecting protein reserves;

2. Total lymphocyte count – representing immune competence;

3. Total cholesterol level – indicating energy reserves.

Each parameter is assigned a specific score range, and the sum of these values yields a total CONUT score between 0 and 12. Serum albumin was scored as follows: ≥3.5 g/dL (0 points), 3.0 to 3.49 g/dL (2 points), 2.5 to 2.99 g/dL (4 points), and <2.5 g/dL (6 points). Total lymphocyte count was scored as ≥1600/mm³ (0 points), 1200 to 1599/mm³ (1 point), 800 to 1199/mm³ (2 points), and <800/mm³ (3 points). Total cholesterol level was scored as ≥180 mg/dL (0 points), 140 to 179 mg/dL (1 point), 100 to 139 mg/dL (2 points), and <100 mg/dL (3 points). Higher total CONUT scores indicated increasingly severe malnutrition. The CONUT score has been reported to have prognostic value in various clinical conditions.^[19,20] It can be used to predict the risk of complications, length of hospital stay, and mortality – particularly in elderly patients, those undergoing surgery, and individuals with chronic diseases. Owing to its ease of scoring, low cost, and reliance on routine laboratory tests, the CONUT score has become increasingly utilized in clinical practice.

The mRS is one of the most widely used tools for assessing disability (or dependence) following a stroke. According to the mRS, scores of 0 to 2 indicate functional independence, 3 to 5 dependence, and 6 death.^[20] Functional independence (0–2 points) is defined as the ability of the patient to walk unassisted. In summary, on the mRS – which is known for its simplicity and ease of interpretation – the key determinant of independence is the ability to ambulate without assistance.

In our study, functional outcomes were evaluated on day 28 following ICU admission. Patients were classified into 2 groups based on 28-day mortality: the survivor group and the mortality group. In addition, according to the 28-day mRS scores, patients were categorized as having a good functional outcome (mRS ≤ 2) or a poor functional outcome (mRS > 2). The PNI, GNRI, and CONUT scores were analyzed and compared between these groups. Because the study was retrospective, no blinding procedure was used. Data extraction and nutritional score calculations were performed after patient outcomes were recorded, and outcome assessors were not involved in score computation.

2.3. Sample size

The sample size was calculated using GPower 3.1 (Heinrich-Heine-Universität Düsseldorf). The primary aim of the study was to evaluate the prognostic value of PNI, GNRI, and CONUT scores at ICU admission on 28-day mortality and functional outcomes in critically ill patients with AIS. Assuming P < .05, an effect size of 0.5, and a statistical power of 80% for t tests, a minimum of 140 patients would be required. All AIS patients who were followed in the ICU during the study period and had complete datasets were included in the analysis.

2.4. Statistical analysis

Data were analyzed using SPSS version 26.0 (SPSS Inc.). Descriptive statistics were presented as numbers and percentages for categorical variables, and as median and range for continuous variables. The Shapiro–Wilk test and histogram plots were used to assess the normality of distribution. Quantitative variables that did not show a normal distribution were compared using the Mann–Whitney U test. Qualitative variables were evaluated using the chi-square or Fisher exact test, as appropriate. Multivariate regression analysis was performed to determine whether PNI, GNRI, and CONUT scores differed significantly between groups and whether they were independent predictors of mortality. Logistic regression results were expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Receiver operating characteristic curve analysis was conducted to assess the prognostic value of PNI, GNRI, and CONUT. A P value of <.05 was considered statistically significant.

3. Results

A total of 142 critically ill AIS patients followed in the ICU were included in the study (Fig. 1). The median age of the overall population was 77 (68–84) years, and 55.6% (n = 79) were female. The median age was significantly higher in the mortality group compared with the survivor group (81 vs 75 years, P = .013). Overall, 45.1% (n = 64) of the patients had a good clinical outcome (mRS ≤ 2).

Figure 1..

Flow chart of the study. A total of 241 geriatric patients with acute ischemic stroke were assessed for eligibility between January 2023 and May 2025. The remaining 236 patients were included in the final analysis and were categorized into the mortality group (n = 29) and the survivor group (n = 207).

The most common comorbidities among patients with AIS were hypertension (67.6%), diabetes mellitus (35.2%), and coronary vascular disease/heart failure (33.8%). There were no significant differences between the groups regarding comorbid conditions. In the mortality group, median GCS scores at ICU admission were significantly lower (8 vs 12, P < .001), while APACHE II scores were significantly higher (22 vs 13, P < .001).

When laboratory parameters at ICU admission were evaluated, the mortality group had significantly lower mean hemoglobin levels (11.2 ± 2.2 vs 13.1 ± 2.1 g/dL, P = .001) and lower median albumin levels (3.4 vs 3.9 g/dL, P = .016). Conversely, the mean platelet count was significantly higher in the mortality group (263.2 ± 79.2 vs 226.4 ± 63.7 × 10⁹/L, P = .013). Among the nutritional indices at ICU admission, PNI (36.3 vs 39.5, P = .017) and GNRI (53.4 vs 56.6, P = .007) were significantly lower in the mortality group, whereas median CONUT scores were significantly higher (2 vs 2, P = .012; Table 1).

Table 1.

Demographic data and clinical characteristics of the groups.

	Survivor group (n = 117)	Mortality group (n = 25)	Overall (n = 142)	P value
Age (yr)	75 (67–83)	81 (76–89)	77 (68–84)	.013
Sex, n (%)				.170
Female	62 (53)	17 (68)	79 (55.6)
Male	55 (47)	8 (32)	63 (44.4)
BMI	25.9 (24.2–29.2)	26.5 (23.2–28.2)	26.1 (24.2–28.7)	.772
Functional outcome				<.001
Good (mRS ≤ 2)	64 (54.7)	0	64 (45.1)
Poor (mRS 3–6)	53 (45.3)	25 (100)	78 (54.9)
Comorbidity
Hypertension	77 (65.8)	19 (76)	96 (67.6)	.323
Diabetes	40 (34.2)	10 (40)	50 (35.2)	.581
CVD/HF	36 (30.8)	12 (48)	48 (33.8)	.098
Atrial fibrillation	26 (22.2)	6 (24)	32 (22.5)	.847
Asthma/COPD	9 (7.7)	1 (4)	10 (7)	1.000
GCS	12 (10–14)	8 (6–9)	12 (9–14)	<.001
APACHE II	13 (10–17)	22 (16–25)	14 (11–20)	<.001
Duration of ICU (d)	7 (3–13)	9 (7–21)	7 (3–13)	.058
Duration of Mv (d)	0	8 (5–17)	0 (0–2)	<.001
Hemoglobin (g/dL)	13.1 ± 2.1	11.2 ± 2.2	12.7 ± 2.3	.001
Albumin (g/dL)	3.9 (3.5–4.3)	3.4 (2.9–3.9)	3.9 (3.5–42)	.016
Lymphocyte, ×109/L	1.8 (1.2–2.4)	1.5 (0.8–2.3)	1.7 (1.1–2.3)	.371
Platelet, ×109/L	226.4 ± 63.7	263.2 ± 79.2	232.9 ± 67.8	.013
Cholesterol (mg/dL)	174.9 ± 46.2	165.9 ± 57.1	173.5 ± 48.2	.432
PNI	39.5 (34.9–42.7)	36.3 (30.2–40.5)	39.1 (34.7–42.3)	.017
GNRI	56.6 (47.5–59.6)	53.4 (46.2–54.4)	54.6 (46.3–58.6)	.007
CONUT	2 (0–3.5)	2 (2–6.5)	2 (1–4)	.012

Data are expressed as number of patients, percentage, and median (interquartile range = Q1–Q3).

Bold values indicate statistically significant.

APACHE II = acute physiology and chronic health evaluation II, BMI = body mass index, CONUT = controlling nutritional status score, COPD = chronic obstructive pulmonary disease, CVD/HF = coronary vascular disease/heart failure, GCS = Glasgow Coma Scale, GNRI = geriatric nutritional risk index, ICU = intensive care unit, mRS = modified Rankin Scale, Mv = mechanical ventilation, PNI = prognostic nutritional index.

Functional outcome was evaluated on day 28 after admission using the mRS. Patients with poor functional outcomes (mRS > 2) had significantly lower GCS scores and higher APACHE II scores at ICU admission (both P < .001). In this group, mean hemoglobin levels (12.1 ± 2.3 vs 13.4 ± 2.0 g/dL, P = .001), median albumin levels (3.7 vs 4.1 g/dL, P < .001), and lymphocyte counts (1.2 vs 2.1 × 10⁹/L, P < .001) were all significantly reduced compared with those with good functional outcomes. Among the nutritional indices, median PNI values were significantly lower in the poor functional outcome group (36.8 vs 41.4, P < .001), whereas median CONUT scores were significantly higher (3 vs 1, P < .001; Table 2).

Table 2.

Analysis of patients according to their functional outcome.

	Good functional outcome	Poor functional outcome	P value
	mRS ≤ 2	mRS > 2
	(n = 64)	(n = 78)
Age (yr)	73 (66–82)	80 (72–87)	.001
Sex, n (%)			.376
Female	33 (51.6)	46 (59)
Male	31 (48.4)	32 (41)
BMI	25.7 (24.1–29.3)	26.4 (24.1–28.3)	.815
Comorbidity
Hypertension	42 (65.6)	54 (69.2)	.648
Diabetes	26 (40.6)	24 (30.8)	.221
CVD/HF	19 (29.7)	29 (37.2)	.348
Atrial fibrillation	10 (15.6)	22 (28.2)	.074
Asthma/COPD	5 (7.8)	5 (6.4)	.754
GCS	14 (12–15)	9 (8–11)	<.001
APACHE II	11 (10–14)	17 (13–24)	<.001
Hemoglobin (g/dL)	13.4 ± 2.0	12.1 ± 2.3	.001
Albumin, ×109/L	4.1 (3.7–4.4)	3.7 (3.1–4.1)	<.001
Lymphocyte, ×109/L	2.1 (1.6–2.6)	1.2 (1.6–2.6)	<.001
Platelet, ×109/L	233.2 ± 65.7	232.7 ± 70.0	.968
Cholesterol (mg/dL)	173.7 ± 44.3	170.8 ± 48.8	.719
PNI	41.4 (36.8–43.7)	36.8 (31.3–40.6)	<.001
GNRI	55.4 (44.9–61.3)	54.3 (48.6–57.8)	.500
CONUT	1 (0–2)	3 (1–6)	<.001

Data are expressed as number of patients, percentage, and median (interquartile range = Q1–Q3).

Bold values indicate statistically significant.

APACHE II = acute physiology and chronic health evaluation II, BMI = body mass index, CONUT = controlling nutritional status score, COPD = chronic obstructive pulmonary disease, CVD/HF = coronary vascular disease/heart failure, GCS = Glasgow Coma Scale, GNRI = geriatric nutritional risk index, mRS = modified Rankin Scale, PNI = prognostic nutritional index.

In the multivariate regression analysis of nutritional scores that showed significant differences between groups, GNRI at ICU admission was identified as an independent predictor of mortality in geriatric patients with AIS (OR = 0.935, 95% CI = 0.877–0.998, P = .042; Table 3).

Table 3.

Multivariate regression analysis of factors affecting mortality.

Variables	OR	95% CI (min–max)	P value
PNI	0.970	0.867–1.085	.598
GNRI	0.935	0.877–0.998	.042
CONUT	1.142	0.865–1.507	.348
Constant	14.896		.367

Bold values indicate statistically significant.

CI = confidence interval, CONUT = controlling nutritional status score, GNRI = geriatric nutritional risk index, OR = odds ratio, PNI = prognostic nutritional index.

In the receiver operating characteristic curve analysis of nutritional scores that demonstrated significant differences between the groups, the cutoff value for GNRI was ≤56.2 with an area under curve (AUC) of 0.672 (0.577–0.767), the cutoff value for PNI was ≤34.2 with an AUC of 0.653 (0.535–0.770), and the cutoff value for the CONUT score was ≥1.5 with an AUC of 0.659 (0.545–0.773). The prognostic performance for predicting mortality ranked as GNRI > CONUT > PNI (Table 4).

Table 4.

Mortality prediction performance of prognostic nutritional index, geriatric nutritional risk index, and controlling nutritional status score.

	Cutoff	Sensitivity	Specificity	AUC (95% CI)
PNI	34.2	0.440	0.829	0.653 (0.535–0.770)
GNRI	56.2	0.960	0.513	0.672 (0.577–0.767)
CONUT	1.5	0.840	0.462	0.659 (0.545–0.773)

AUC = area under curve, CI = confidence interval, CONUT = controlling nutritional status score, GNRI = geriatric nutritional risk index, PNI = prognostic nutritional index.

In the multivariate regression analysis of factors associated with functional outcomes, the CONUT score was identified as an independent predictor of 28-day functional outcome (OR = 1.410, 95% CI = 1.105–1.800, P = .006; Table 5).

Table 5.

Multivariate regression analysis of factors affecting functional status.

Variables	OR	95% CI (min–max)	P value
PNI	0.932	0.850–1.023	.138
GNRI	1.005	0.964–1.048	.819
CONUT	1.410	1.105–1.800	.006
Constant	6.382		.433

Bold values indicate statistically significant.

CI = confidence interval, CONUT = controlling nutritional status score, GNRI = geriatric nutritional risk index, OR = odds ratio, PNI = prognostic nutritional index.

4. Discussion

In this study, investigating the prognostic value of nutritional scores in critically ill geriatric patients with AIS, PNI, GNRI, and CONUT scores at ICU admission, it was found that these scores differed significantly between survivors and those who experienced in-hospital mortality. GNRI was identified as an independent predictor of in-hospital mortality, whereas the CONUT score was an independent predictor of 28-day functional outcome. Although GNRI demonstrated the highest AUC among the 3 indices, the overall discriminative ability of PNI, GNRI, and CONUT was modest. Therefore, these scores should not be considered as standalone prognostic tools, but rather as adjunctive markers that may assist in early risk stratification in critically ill geriatric patients with AIS.

Stroke is a major cause of functional disability and mortality worldwide. Age is a non-modifiable risk factor for stroke, and it has been reported that approximately 70% of individuals who experience a stroke are over 65 years old.^[21] Previous studies have indicated that sex does not show a significant difference in the development of stroke.^[21,22] Although higher 28-day mortality rates have been reported in women with AIS, this difference has not been found to be statistically significant.^[7] Hypertension is considered one of the most important modifiable risk factors for both ischemic and hemorrhagic stroke.^[23] In the literature, ischemic stroke has been reported in approximately 90% of patients admitted to stroke units, with a mean age of 71 years, and hypertension is present in 77% as a comorbidity.^[15] In the present study, only geriatric patients with AIS were included. Age was significantly higher in the mortality group, whereas sex did not show a significant association with mortality. Hypertension was present in 66% of the total population, but comorbidities did not differ significantly between the groups.

Aging is an irreversible biological process characterized by a gradual decline in the functional reserves of organ systems. In the geriatric population, deterioration in organ and body functions, the presence of comorbidities, and nutritional deficiencies can significantly affect quality of life. Malnutrition may lead to reduced physical activity, impaired immune function, and progressive decline in organ performance. It has been reported that stroke patients with nutritional impairment experience longer hospital stays and poorer clinical outcomes.^[24] Therefore, assessing nutritional status in the geriatric population is essential for improving prognosis. However, there is currently no universally accepted nutritional screening tool specifically for patients with AIS. Although biochemical markers such as serum albumin, prealbumin, and total cholesterol have been proposed for evaluating nutritional status, they are not widely accepted as standalone indicators. In contrast, PNI, GNRI, and CONUT are practical scoring systems that can be rapidly calculated using serum albumin, total lymphocyte count, cholesterol levels, and body weight measurements. Their prognostic value has been demonstrated in various patient populations in the literature.^{[11,13,25–27]}

Malnutrition in stroke patients is associated with poor prognosis in both the short and long term, with functional outcomes and mortality commonly evaluated as key prognostic indicators. Studies conducted particularly in older populations have suggested that GNRI may serve as a prognostic marker in stroke patients, similar to its established predictive value in infections, cardiovascular diseases, and malignancies.^[28,29] Serum albumin and body weight, which are used in the calculation of GNRI, are fundamental parameters that reflect the nutritional status of elderly individuals. Moreover, recent studies have reported that GNRI is significantly associated with cognitive function and quality of life following AIS.^[30,31] In our study, GNRI was found to be an independent predictor of in-hospital mortality in patients with AIS. Although GNRI demonstrated slightly better discriminative performance than PNI and CONUT, the overall predictive performance of all three indices was similar.

The CONUT score is a nutritional assessment tool based on total lymphocyte count, serum albumin level, and total cholesterol level, reflecting the patient’s immunonutritional status. Vincenzo et al^[20] reported that the CONUT score is an independent prognostic indicator in patients with stroke (both ischemic and hemorrhagic). The authors noted that a higher CONUT score – indicating malnutrition – is associated with increased functional disability, a higher incidence of in-hospital or early infections, and elevated mortality. Hu et al^[28] further demonstrated that malnutrition, defined by NRS-2002 ≥ 3, CONUT ≥ 5, and GNRI < 98, is strongly associated with early death and poor recovery after stroke. Similarly, Huang et al^[32] reported that assessing malnutrition at admission using any of the 3 nutritional indices (CONUT, GNRI, and PNI) is useful, as each shows significant associations with survival and functional outcomes in stroke patients. However, the authors also emphasized that there is no clear consensus regarding which nutritional assessment tool is the most appropriate. Consistent with the literature, the present study found that the CONUT score was associated with 28-day mortality and poor functional outcomes. Although the prognostic performances of the nutritional indices were similar, their predictive value for mortality ranked as GNRI > CONUT > PNI.

The PNI score showed prognostic value similar to that of GNRI and CONUT in our study and has become an increasingly noteworthy marker in the stroke literature. PNI is calculated based on serum albumin level and lymphocyte count, thereby reflecting both nutritional and immunological status.^[33] Several studies, particularly those involving patients receiving thrombolytic therapy for AIS, have shown that lower PNI scores are associated with poor 3-month functional outcomes.^[34] In another prospective cohort study, a nonlinear association was observed between PNI and short-term stroke mortality, with lower PNI levels linked to increased mortality and higher complication rates.^[35] PNI has also been proposed as a predictor of infectious complications. Patients with low PNI scores have been reported to exhibit significantly higher rates of in-hospital infections.^[36] Moreover, even in younger adult populations, PNI has been shown to correlate with functional recovery following stroke.^[37] Although GNRI and CONUT incorporate more complex parameters, the simplicity, low cost, and strong integration with routine laboratory data make PNI particularly suitable for early triage and rapid risk assessment. For these reasons, despite its relatively lower sensitivity, the incorporation of PNI into clinical management algorithms is recommended.

This study has several limitations. First, it was a single-center, retrospective investigation with a relatively small sample size. Multicenter, prospective studies with larger patient cohorts may yield more robust, generalizable findings. Second, although patients with AIS presented to the emergency department within 24 hours of symptom onset, the exact time of stroke onset could not always be clearly determined. Third, the nutritional scores were calculated using laboratory parameters obtained at ICU admission, and dynamic or serial measurements were not assessed. Fourth, the prognostic outcomes in this study were limited to 28-day mortality and functional status; long-term effects were not evaluated. Finally, stroke severity was not assessed using a stroke-specific scale such as the National Institutes of Health Stroke Scale. Although GCS and APACHE II scores were recorded, these tools are not specifically designed to capture focal neurological deficits. Because stroke severity is associated with nutritional status, residual confounding cannot be excluded, and the independent prognostic value of nutritional indices should be interpreted with caution.

The discriminative ability of PNI, GNRI, and CONUT scores was modest in the present study. Accordingly, these indices should not be considered as standalone prognostic tools, but rather as adjunctive markers that may support early risk stratification in critically ill geriatric patients with AIS. Given their low cost and ease of calculation, they may be useful in routine clinical practice; however, their prognostic value should be interpreted with caution. Future prospective multicenter studies incorporating stroke-specific severity scales and larger cohorts are warranted to validate these findings and to assess the incremental value of nutritional indices beyond established prognostic models.

5. Conclusion

Nutritional scores calculated from routine laboratory parameters at ICU admission are associated with mortality and functional outcomes in critically ill geriatric patients with AIS. Among these indices, GNRI showed the highest discriminative ability, although the overall prognostic performances of PNI, GNRI, and CONUT were comparable.

Author contributions

Conceptualization: Kadir Arslan, Hale Cetin Arslan, Ayca Sultan Sahin.

Data curation: Kadir Arslan, Hale Cetin Arslan.

Formal analysis: Kadir Arslan, Hale Cetin Arslan, Ayca Sultan Sahin.

Funding acquisition: Kadir Arslan.

Investigation: Kadir Arslan, Hale Cetin Arslan.

Methodology: Kadir Arslan, Hale Cetin Arslan, Ayca Sultan Sahin.

Resources: Kadir Arslan.

Software: Kadir Arslan, Hale Cetin Arslan.

Supervision: Kadir Arslan, Hale Cetin Arslan, Ayca Sultan Sahin.

Visualization: Kadir Arslan, Hale Cetin Arslan.

Writing – original draft: Kadir Arslan, Hale Cetin Arslan.

Writing – review & editing: Kadir Arslan, Hale Cetin Arslan, Ayca Sultan Sahin.

Project administration: Hale Cetin Arslan.