Translate this page into:
Outcomes and predictors of early neurological deterioration among young adults with acute ischemic stroke
, Sankar P. Gorthi2, Chaitanya Prasad Kulkarni1, Rasika Sudip Sule1, Nitin Rana1, Priyesha Titre1, Shivakumar Iyer3
*Corresponding author: Dr. Vaibhav R. Suryawanshi, Department of Pharmacy Practice, Bharati Vidyapeeth Deemed University, Poona College of Pharmacy, Pune, Maharashtra, India. phdrvaibhav@gmail.com
-
Received: ,
Accepted: ,
How to cite this article: Suryawanshi VR, Gorthi SP, Kulkarni CP, Sule RS, Rana N, Titre P, et al. Outcomes and predictors of early neurological deterioration among young adults with acute ischemic stroke. J Neurosci Rural Pract. doi: 10.25259/JNRP_52_2026
Abstract
Objectives:
Despite the rising incidence of early-onset acute ischemic stroke (AIS) in India, evidence on outcomes and predictors of early neurological deterioration (END) among young adults remains scant. We aimed to evaluate the association of National Institutes of Health Stroke Scale (NIHSS)-defined END with mortality and functional dependency among young adults with AIS and to identify the clinical, radiological, laboratory, and treatment-related factors contributing to END.
Materials and Methods:
In this prospective observational cohort study of young adult (18–49 years) patients with AIS, we determined the association of END with functional dependency (assessed using modified Rankin scale [mRS]) and analyzed clinical, imaging, laboratory, and treatment-related data from 52 patients presenting within 4.5-hours of last known well. END was defined as a ≥2-point NIHSS increase or ≥1-point motor worsening within 7 days of the index event. Predictors were identified using modified Poisson multivariate and least absolute shrinkage and selection operator (LASSO) regression models.
Results:
The median interquartile range age of the cohort was 42 (31–46) years. END occurred in 21 (40.4%) patients and caused greater functional dependency (mRS, 3–6) compared to those without END (adjusted odds ratio 4.89 [95% confidence interval 1.48–11.10, p = 0.008]). Hypertension (adjusted risk ratios [aRRs] 2.98, p = 0.047), Glasgow Coma Scale ≤12 (aRR 2.56, p = 0.023), and Alberta stroke program early computed tomography score ≤7 (aRR 1.48, p = 0.051) were identified as independent predictors of END. Anterior cerebral artery involvement (β = +1.92), hyperhomocysteinemia >15 µmoL/L (β = +0.68), male sex (β = +0.59), hypoalbuminemia (β = +0.55), and middle cerebral artery involvement (β = +0.22) showed positive prediction scores for END in LASSO models.
Conclusion:
END occurred in 2 of every 5 young adult patients with AIS, with large-vessel occlusion as the major predictor, and poor short-term outcomes were documented in more than half. Hypertension and index-event hyperglycemia emerged as modifiable risk factors. Distinct predictors of END underscore the importance of early recognition for targeted acute interventions.
Keywords
Acute ischemic stroke
Early neurological deterioration
Outcomes
Predictors
Young adults
INTRODUCTION
Acute ischemic stroke (AIS) accounts for most stroke cases in India and remains a major cause of mortality and long-term disability. National estimates report 108–172 strokes/100,000 people annually, with AIS predominating.[1] Young adults (18–45 years) comprise 16.8% of cases, underscoring a significant early-onset burden.[1,2] Indian studies highlight the need to evaluate varied etiologies, including intracranial infections and coagulopathies, recognizing stroke in the young as a distinct entity.[3-5] Urbanization-driven lifestyle changes have shifted risk profiles. Unlike developed countries, where arterial dissection and cardioembolism are common,[6,7] limited evidence suggests atherothrombosis predominates among young adults in developing regions.[3,5,8]
Early neurological deterioration (END), a measurable decline in neurological status within hours to days after stroke, is a strong predictor of poor functional outcomes.[9,10] In AIS patients, END often arises from infarct progression, cerebral edema, hemorrhagic transformation, recurrent ischemia, and/or systemic complications that accelerate infarct progression.[10,11] Studies in younger adults show that those experiencing END are far more likely to remain functionally dependent at discharge and at 3-month follow-up.[9,11-13]Despite extensive characterization in older adults, data on END among young adult patients remain limited, especially in Indian tertiary care settings where delays and differences in acute management may influence clinical outcomes.
We hypothesize that END in young adults with AIS worsens functional outcomes, driven by diverse predictors. Identifying these predictors is essential for timely risk stratification, targeted monitoring, and early intervention. We aimed to evaluate the association of National Institutes of Health Stroke Scale (NIHSS)-defined END with mortality and functional dependency in young adults with AIS and to identify the clinical, radiological, laboratory, and treatment-related factors contributing to END.
MATERIALS AND METHODS
Study design and settings
This single-center prospective cohort study was conducted at “Bharati Hospital and Research Center, Pune” for a period of 16-months (September 2024–December 2025). Ethical clearance was obtained from the study hospital (Ref., BVDU/MC/IEC/92/24–25, September 28th, 2024). Written informed consent was obtained from study participants and/or legally authorized representatives when neurological deficits impaired decision-making capacity.
Study patients
Consecutive young adults aged 18–49 years presenting with AIS were evaluated in the emergency department. All suspected cases underwent urgent clinical assessment, vital-sign evaluation, non-contrast computed tomography (CT) brain, CT angiography of carotid and vertebral circulation, and magnetic resonance imaging (MRI) when indicated, following the institutional stroke protocol. Patients presenting within 4.5-hours of last-known-well, with measurable neurological deficits (NIHSS ≥4) or major impairments such as severe aphasia, complete hemianopia, or marked limb weakness, and without intracranial hemorrhage on CT, received intravenous thrombolysis with alteplase or tenecteplase.
Inclusion criteria
(1) Age 18–49 years; (2) clinical diagnosis of AIS confirmed by CT or MRI with diffusion-weighted imaging (DWI); (3) acute onset of focal neurological deficits including altered consciousness, hemiparesis, speech disturbance, or other focal signs; (4) pre-stroke modified Rankin Scale (mRS) ≤1; and (5) receipt of at least one stroke-specific therapeutic intervention (antiplatelet therapy, anticoagulation, thrombolysis, or thrombectomy).
Exclusion criteria
(1) Pre-existing intracranial pathology such as vascular malformations, neoplasms, or other neurological diseases; (2) neuroimaging evidence of intracerebral hemorrhage, vascular malformation, tumor, or abscess; (3) non-stroke causes of neurological impairment such as traumatic brain injury or post-surgical complications; and (4) patients undergoing mechanical thrombectomy/endovascular therapy, as procedural factors may independently influence neurological worsening.
Sample size estimation
The primary objective was to assess mortality or functional dependency among young adults with AIS. Prior study showed rates of 94.1% among patients with END versus 30% in those without END.[11] Assuming that 94% of subjects in the reference population had the factor of interest, and applying a continuity correction with an expected response rate of 50%, the required sample size was estimated at 16 participants/group (32 in total, assuming equal allocation). This sample size provides 80% power to detect a difference in proportions of −0.64 between the test and reference groups at a two-sided significance level of 0.05.
Clinical assessments and data collection
Baseline variables
Demographic and clinical data were systematically extracted from medical records and case files, supplemented by direct patient or caregiver interviews. Baseline variables included age, sex, vascular risk factors (hypertension, diabetes mellitus, dyslipidemia, current smoking, alcohol use), cardiac conditions (ischemic heart disease, atrial fibrillation, valvular heart disease), prior stroke or transient ischemic attack, and other comorbidities.
Neurological assessment
Stroke severity was assessed at admission using the NIHSS, and consciousness using the Glasgow Coma Scale (GCS). Serial NIHSS evaluations were performed on admission, at 24-hours, at 72-hours, at 5th day, and at 7th day (during any clinical worsening to detect END). END was defined as a ≥2-point increase in total NIHSS or ≥1-point increase in the motor subscale within 7-days of the index event, allowing detection of complications such as delayed edema or infarct progression beyond the hyper-acute phase.[14] Causes of deterioration were determined through clinical assessment and repeat neuroimaging and categorized as hemorrhagic (hemorrhagic transformation or symptomatic intracerebral hemorrhage), ischemic (infarct extension, early recurrent stroke, malignant edema with mass effect or midline shift >5 mm, or unexplained decline), or systemic (aspiration pneumonia, sepsis, metabolic disturbances).
Neuroimaging protocol
All patients underwent non-contrast CT at presentation to exclude hemorrhage and assess for early ischemic changes. Alberta stroke program early CT score (ASPECTS) was assessed on initial CT to quantify middle cerebral artery (MCA)-territory ischemia, with scores ≤7 indicating extensive involvement. For posterior circulation strokes, pc-ASPECTS was used to evaluate the brainstem, thalamus, occipital lobes, and cerebellum. CT angiography of the carotid and vertebral systems was performed to identify large-vessel occlusion, including internal carotid artery (ICA), MCA (M1/M2), anterior cerebral artery (ACA), basilar, or posterior cerebral artery (PCA) occlusions. When feasible, MRI with DWI and magnetic resonance angiography was obtained to confirm infarct territory, assess penumbra, and characterize vascular pathology. Follow-up CT or MRI at 24–72 hours, or during clinical worsening, assessment of infarct evolution, hemorrhagic transformation, development of new ischemic lesions, cerebral edema with mass effect, and herniation. Infarcts were classified by vascular territory (anterior: ICA, MCA, ACA; posterior: PCA, basilar artery, vertebral). Hemorrhagic transformation was categorized as hemorrhagic infarction or parenchymal hematoma, and symptomatic intracerebral hemorrhage was defined by neurological worsening on NIHSS.[15]
Laboratory investigations
Key biochemical assessments included serum homocysteine, albumin, calcium, vitamin B12, cholesterol profile, C-reactive protein (CRP), and urea levels, as these markers were central to the study objectives. Random blood glucose was measured at presentation in the emergency department, and hemoglobin A1c was assessed during evaluation in the ward.
Stroke etiology classification
Stroke etiology was classified using Trial of Org 10172 in Acute Stroke Treatment[16] criteria into large-artery atherosclerosis, cardioembolism, small-vessel occlusion, other determined causes (e.g., dissection, vasculitis, hypercoagulable states), and undetermined etiology, including cryptogenic stroke due to incomplete evaluation, multiple possible causes, or negative workup.
Treatment protocols
Treatment was individualized based on onset time, stroke severity, imaging, and contraindications. Eligible patients within 4.5-hours received IV-thrombolysis. Antiplatelet therapy was started in non-thrombolyzed patients and after 24-hours post-thrombolysis, once hemorrhage was ruled out. Anticoagulation was initiated only after ruling out hemorrhagic transformation. All patients received high-intensity statins. Hypertension was managed as per guidelines and hyperglycemia was managed on insulin to maintain blood sugar levels between 140 and 180 mg/dL.
Study endpoints: Outcome assessment
Patients were followed for 3 months from hospitalization for an index event. The primary outcome was the occurrence of END within 7 days of AIS onset. Secondary outcomes included mortality and functional dependency at 3 months. Functional outcomes were determined using mRS at 3 months, where favorable was deemed as ≤2. Activities of daily living were assessed using the Barthel Index, where ≥60 score indicated functional independence at 3 months. All the assessments were conducted in person and/or through structured telephone interviews.
Statistical analysis
Descriptive statistics summarized demographics, stroke etiology, risk factors, clinical parameters, vascular territories, treatments, and outcomes. Categorical variables were presented as frequencies and percentages and continuous variables as means ± standard deviation (SD) or medians with interquartile range (IQR) based on distribution. For group comparisons (age categories and END vs. No-END), we used Fisher’s exact test considering categorical variables (age, sex, risk factors, NIHSS, vascular territory, GCS, ASPECTS, and treatments) and the Wilcoxon two-sample (Mann–Whitney U) test considering continuous variables (pretreatment characteristics, biochemical markers, and functional outcomes) due to non-normality.
Modified Poisson univariate and multivariate regression models were used to study risk ratios with 95% confidence intervals (CIs) for potential predictors of END, with statistical significance set at p < 0.05. To identify the most robust predictors and reduce overfitting, the least absolute shrinkage and selection operator (LASSO) regression model was applied, considering automated variable selection. The optimal regularization parameter (λ) was determined using 10-fold cross-validation based on the minimum criteria. Variables with non-zero coefficients (β) were retained; coefficient magnitude reflected association strength, and sign indicated direction (positive for risk and negative for protection). Before model fitting, all predictor variables were standardized to have a mean of zero and an SD of one. This step ensures that the LASSO penalty is applied evenly, preventing variables with naturally larger numerical ranges from disproportionately influencing the selection process. After the model identified the most significant predictors, the coefficients were transformed back to their original scales to allow straightforward clinical interpretation. All analyses were performed using “SAS 9.4M8” and “R version 4.3.2.”
RESULTS
Of 66 eligible patients enrolled, 14 patients were excluded from the final analysis; 8 due to loss to follow-up, 3 with incomplete medical records, and 3 with missing outcome data. Fifty-two patients completed a 3-month follow-up in the neurology clinic and were included in the final analysis.
Demographics, clinical, and laboratory variables
Among the 52 young adult patients, END was observed in 21 (40.4%), while the remaining 31 (59.6%) had no occurrence of END. The median (IQR) age of the cohort was 42 (31–46) years [Table 1]. A male predominance was noted, with 38 patients (73.1%) being male. The incidence of END was higher among male patients 17 (80.9%) compared with 4 (19.1%) female patients. The median (IQR) body mass index (kg/m2) was 28.5 (24.4–33.2). Diabetes mellitus was a more common comorbidity in the END group (13 [61.9%] versus 7 [22.6%], p = 0.018). Blood glucose at presentation was also higher in END patients, with a mean ± SD (mg/dL) of 161 ± 16 compared with 147 ± 19 in the No-END group (p = 0.007). Similarly, a greater proportion of END patients had systolic blood pressure ≥160 mmHg and diastolic blood pressure ≥90 mmHg compared with the No-END cohort (p= 0.042 and 0.026, respectively).
| Sr. no. | Characteristics/variables | Total n=52, n (%) | END n=21, n (%) | No END n=31, n (%) | p-value |
|---|---|---|---|---|---|
| A. Patient demographics and history | |||||
| 1. | Age (years) | ||||
| Median (IQR) | 42 (31–46) | 43 (31–47) | 41 (31–45) | 0.520 | |
| 18–39 | 17 (32.7) | 5 (23.8) | 12 (38.7) | 0.251 | |
| 40–49 | 35 (67.3) | 16 (76.2) | 19 (61.3) | 0.374 | |
| 2. | Gender | ||||
| Male | 38 (73.1) | 17 (80.9) | 21 (67.7) | ||
| Female | 14 (26.9) | 4 (19.1) | 10 (32.3) | 0.353 | |
| 3. | Body mass index (kg/m2) | ||||
| Median (IQR) | 28.5 (24.4–33.2) | 32.5 (27.5–35.8) | 24.5 (21.3–30.6) | 0.140 | |
| 4. | Known hypertension | 43 (82.7) | 19 (90.5) | 24 (77.4) | 0.270 |
| 5. | Diabetes mellitus | 20 (38.4) | 13 (61.9) | 7 (22.6) | 0.018 |
| 6. | Cardiac disease | 14 (26.9) | 7 (33.4) | 7 (22.6) | 0.532 |
| 7. | Atrial fibrillation | 7 (13.5) | 4 (19.1) | 3 (9.7) | 0.685 |
| 8. | Previous ischemic stroke/TIA | 3 (5.8) | 2 (9.5) | 1 (3.2) | 0.598 |
| 9. | Current/previous smoking | 28 (53.8) | 13 (61.9) | 15 (48.4) | 0.542 |
| B. Pretreatment characteristics | |||||
| 1. | Blood glucose at presentation (mg/dL)* | 153±20 | 161±16 | 147±19 | 0.007 |
| 2. | SBP ≥160 mmHg | 31 (59.6) | 16 (76.2) | 15 (48.4) | 0.042 |
| DBP ≥90 mmHg | 22 (42.3) | 13 (61.9) | 9 (29) | 0.026 | |
| 3. | IV antihypertensive before thrombolysis | 17 (32.7) | 10 (47.6) | 7 (22.6) | 0.103 |
| 4. | GCS | ||||
| Median (IQR) | 12 (11–15) | 10 (9–15) | 15 (13–15) | 0.021 | |
| GCS >12 | 32 (61.5) | 7 (33.4) | 25 (80.6) | 0.003 | |
| GCS ≤12 | 20 (38.5) | 14 (66.7) | 6 (19.4) | 0.001 | |
| 5. | NIHSS | ||||
| Median (IQR) | 9 (4–14) | 11 (6–14) | 8 (3–13) | 0.122 | |
| NIHSS ≤8 | 18 (34.6) | 5 (23.8) | 13 (41.9) | 0.184 | |
| NIHSS 9–13 | 23 (44.2) | 11 (52.4) | 12 (38.7) | 0.412 | |
| NIHSS >14 | 11 (21.2) | 5 (23.8) | 6 (19.4) | 0.740 | |
| 6. | ASPECTS | ||||
| Median (IQR) | 7 (5–9) | 6 (4–8) | 8 (6–9) | 0.007 | |
| ASPECTS >7 | 23 (44.2) | 4 (19.1) | 19 (61.3) | 0.004 | |
| ASPECTS ≤7 | 26 (50) | 15 (71.4) | 11 (35.5) | 0.011 | |
| 7. | Vascular territory involved | ||||
| Proximal arterial occlusion | 31 (59.6) | 18 (85.7) | 13 (41.9) | 0.001 | |
| Only ICA occlusion | 15 (28.8) | 9 (42.8) | 6 (19.3) | 0.072 | |
| 8. | Symptom onset to needle time (minutes)* | 219±53 | 211±48 | 226±57 | 0.343 |
| 9. | Door-to-needle time (minutes)* | 50±15 | 52±14 | 49±17 | 0.520 |
| C. TOAST classification | |||||
| 1. | Large artery atherosclerosis | 27 (51.9) | 14 (66.7) | 13 (41.9) | 0.414 |
| 2. | Cardioembolic | 6 (11.5) | 3 (14.3) | 3 (9.7) | |
| 3. | Small vessel disease | 10 (19.2) | 6 (28.6) | 4 (12.9) | |
| 4. | Other | 9 (17.3) | 5 (23.8) | 4 (12.9) | |
| D. Treatments received | |||||
| 1. | Thrombolyzed using rtPA | 29 (55.8) | 13 (61.9) | 16 (51.6) | 0.451 |
| 2. | Use of single antiplatelet | 16 (30.8) | 11 (52.4) | 5 (16.1) | 0.026 |
| 3. | Use of dual antiplatelet | 36 (69.2) | 10 (47.6) | 26 (83.8) | 0.026 |
| 4. | Use of anticoagulants | 21 (40.4) | 10 (47.6) | 11 (35.5) | 0.415 |
| E. Outcomes | |||||
| 1. | Length of stay >5 days | 28 (53.8) | 16 (76.2) | 12 (38.7) | 0.015 |
| 2. | Mortality (in-hospital and 3 months) | 5 (9.6) | 4 (19.1) | 1 (3.2) | 0.142 |
| 3. | Functional outcome (mRS at 3 months) (dependency, mRS 3–6) | 23 (44.2) | 14 (66.7) | 9 (29.1) | 0.008 |
| 4. | Functional dependence (BI <40) | 18 (34.6) | 12 (57.2) | 6 (19.4) | 0.004 |
END: Early neurological deterioration, AIS: Acute ischemic stroke, IQR: Interquartile range, GCS: Glasgow coma score, NIHSS: National Institute of health stroke severity score, ASPECTS: Alberta stroke program early computed tomography score, SBP: Systolic blood pressure, DBP: Diastolic blood pressure, mRS: Modified Rankin scale. *Note, values are mentioned as mean±standard deviation, The values mention in the bold text are statistically significant (i.e., p < 0.05). rtPA: Recombinant tissue plasminogen activator, TIA: Transient ischemic attack, TOAST: Trial of Org 10172 in acute stroke treatment classification, IV: Intravenous, ICA: Internal carotid artery, BI: Barthel index
The END cohort had significantly lower GCS scores compared with the No-END group (10 [9–15] vs. 15 [13–15], p = 0.021). NIHSS scores were higher in the END cohort 11 (6–14) versus 8 (3–13) although this difference was not statistically significant (p = 0.122). ASPECTS was lower in the END cohort (6 [4–8]) compared with the no-END cohort (8 [6–9]) (p = 0.007). Regarding vascular involvement, proximal arterial occlusion was more frequent in END patients (18 [85.7%] vs. 13 [41.9%], p = 0.001) [Supplementary Files]. A similar trend was observed for isolated ICA occlusion (9 [42.8%] versus 6 (19.3%), p = 0.072). The END cohort also showed higher use of single antiplatelet therapy (SAPT) (11 [52.4%] vs. 5 [16.1%], p = 0.026) and lower use of dual antiplatelet therapy (DAPT) (10 [47.6%] vs. 26 [83.8%], p = 0.026) compared with the No-END group [Table 1]. Length of stay >5 days was more common in the END cohort (16 [76.2%] vs. 12 [38.7%], p = 0.015). Poor functional outcomes on mRS were observed in 14 patients (66.7%) with END, significantly higher than in the No-END group (9 [29.1%], p = 0.008).
CRP (mg/L) levels were elevated in END patients (17 [10–23]) compared with no-END patients (11 [4–17]) (p = 0.034). Similar findings were noted for homocysteine levels (µmoL/L) (25 [15–69] vs. 14 [10–33], p = 0.015) [Table 2]. Hypoalbuminemia (g/dL) was also more common in the END cohort (3.0 [2.3–3.7] vs. 3.8 [3.2–4.1], p = 0.023).
| Parameters | Total n=52, n (%) | END n=21, n (%) | No END n=31, n (%) | p-value |
|---|---|---|---|---|
| Vitamin B12 (pg/mL), median (IQR) | 200 (149–320) | 179 (150–262) | 215 (149–358) | 0.320 |
| Urea (mg/dL), median (IQR) | 19 (14–32) | 20 (14–37) | 18 (15–29) | 0.483 |
| CRP (mg/L), median (IQR) | 14 (7–20) | 17 (10–23) | 11 (4–17) | 0.034 |
| Homocysteine (µmoL/L), median (IQR) | 18 (12–45) | 25 (15–69) | 14 (10–33) | 0.015 |
| Calcium (mg/dL), median (IQR) | 8.72 (7.90–9.90) | 8.76 (7.46–10.10) | 8.68 (8.24–9.72) | 0.621 |
| Albumin (g/dL), median (IQR) | 3.4 (2.9–4.0) | 3.0 (2.3–3.7) | 3.8 (3.2–4.1) | 0.023 |
| Cholesterol (mg/dL), median (IQR) | 195 (168–245) | 210 (171–276) | 182 (165–225) | 0.064 |
CRP: C-reactive protein, IQR: Interquartile range, END: Early neurological deterioration, AIS: Acute ischemic stroke, The values mention in the bold text are statistically significant (i.e., p < 0.05).
Univariate and multivariate analysis of potential factors associated with END
Results are presented as adjusted risk ratios with 95% CIs and p-values for each variable. Of the 14 potential factors evaluated for END in the univariate models, six showed statistical significance [Figure 1]. These included GCS score ≤12 (3.20 [1.51–6.50], p = 0.001), NIHSS >8 (1.86 [1.20–2.13], p = 0.037), ASPECTS ≤7 (2.49 [1.50–5.42], p = 0.024), involvement of the MCA territory (1.57 [1.15–2.83], p = 0.046), involvement of the ACA territory (2.01 [1.64–6.45], p = 0.001), and use of SAPT over DAPT (2.92 [1.18–7.55], p = 0.021).
Of the 9 potential factors assessed in the multivariate models, three remained statistically significant [Figure 2]. Hypertension (2.98 [0.99–6.01], p = 0.047), GCS score ≤12 (2.56 [1.14–5.75], p = 0.023), and ASPECTS ≤7 (1.48 [0.99–3.52], p = 0.051) were identified as independent factors associated with END.
LASSO regression curve and prediction scores of END with beta coefficients
ACA territory involved (β = +1.92), hyperhomocysteinemia >15 µmoL/L (β = +0.68), male gender (β = +0.59), hypoalbuminemia (β = +0.55), and MCA territory involved (β = +0.22) showed positive prediction scores for END, whereas DAPT use (β = −1.76), high-intensity statin use (β = −1.17), history of cardiac disease (β = −0.76), GCS score >12 (β = −0.24), and ASPECTS >7 (β = −0.08) showed negative prediction scores for END [Figure 3].
Time-dependent changes in the incidence rates of END
Kaplan–Meier curves demonstrate the cumulative probability of END rising over time, referenced either from stroke onset (Graph - A) or hospital admission (Graph - B), underscoring how risk varies by time metric [Figure 4].
END incidence declined over time from both stroke onset and hospital admission, “peaking within 24 hours” (9 and 10 patients, respectively) and reaching “lowest rates at 5–7 days” (2 and 1 patients), highlighting the importance of early monitoring and intervention [Supplementary Files].
Comparison of functional outcomes using mRS
For the 52 patients, the median (IQR) mRS at baseline was 3 (1–5). A shift toward better functional status was observed at 3 months, with a median (IQR) mRS of 2 (0–4). Overall, 29 patients (56%) showed functional improvement (favorable outcome, mRS ≤2) at 3-months, of whom 23 (44%) achieved good outcomes (mRS ≤1) [Figure 5a].
At 3 months, the median (IQR) mRS was 4 (2–4) in the END cohort compared with 1 (0–3) in the No-END cohort (p = 0.004). Favorable outcomes (mRS ≤2) were observed in 7 patients (33.4%) with END versus 22 patients (70.9%) with no-END (p = 0.008) [Figure 5b]. Odds for functional dependency at 3 months (mRS, 3–6) among the END cohort were 4.89 (95% CI, 1.48–11.10, p= 0.008). Similarly, using barthel index (BI), odds for functional dependency at 3-months (BI < 40) were 5.56 (95% CI, 1.61–13. 23, p= 0.004).
DISCUSSION
This prospective cohort study yielded several key findings, outlined as follows: (1) END was observed in 2 of every 5 young adults with AIS. (2) END incidence peaked within the initial 24-hours of admission and declined markedly over subsequent time intervals. (3) Hypertension, baseline hyperglycemia, poor GCS score, higher NIHSS, low ASPECTS score, involvement of ACA and/or MCA territory, proximal arterial occlusion, use of SAPT over DAPT, hyperhomocysteinemia, and hypoalbuminemia are strongly associated with END. (4) Hypertension, GCS ≤12, and ASPECTS ≤7 were independent predictors of END. (5) END was associated with increased odds of poor outcome at 3 months.
Applying the definition of END as a ≥2-point increase in total NIHSS or ≥1-point motor worsening within 7-days, we identified END in 40% of patients. This was comparable to prior studies,[12,17-19] systematic review,[20] and meta-analysis[21] using a ≥2-point NIHSS-worsening threshold, which have reported END rates of 5–40% depending on follow-up duration. Although a ≥4-point NIHSS increase within 24–74 hours remains the most commonly used criterion, evidence increasingly suggests that a ≥2-point deterioration may better predict in-hospital mortality.[14]
The greater incidence of END in late young adults (40– 49 years) aligns with young adult stroke studies,[3-5,7,12,17,21] which consistently show higher vascular risk burden and early vascular aging in this subgroup compared with early young adults. The male predominance similarly reflects established epidemiological patterns as they exhibit a more aggressive vascular risk profile,[4,5,12,22] though interpretation is tempered by our limited sample size. Hypertension, diabetes, and admission hyperglycemia have been widely recognized as key contributors to END, and our findings are consistent with this established evidence.[6,7,12,19,21] Chronic hypertension and acute hypertensive surges are known to disrupt cerebral autoregulation, while elevated diastolic pressure has been linked to reduced collateral perfusion and accelerated ischemic progression. Similarly, the association between diabetes, hyperglycemia, and oxidative injury is well documented, reinforcing their role in worsening early stroke outcomes.[6,8,17,23,24]
Prior stroke studies across both young and elderly population consistently identify hypertension,[3,5,7] index hyperglycemia,[6-7,10,12,17,20] poor GCS score,[10,12,17-20] higher NIHSS,[10,12,17-21,24,25] low ASPECTS score,[20,25] involvement of ACA and/or MCA territory,[11,12,13,18-22] cortical location of infarcts and proximal arterial occlusion,[13,20-22,24] use of SAPT over DAPT,[26-28] hyperhomocysteinemia,[8,12,21,29] and hypoalbuminemia[8,21,29] as robust predictors of END. Our literature review did not identify any link between END and earlier ischemic insults in prior studies. Large-artery atherosclerosis and shared vascular risks may explain this association. Having said this, we did not find the relationship between END and atrial fibrillation, congestive heart failure, or other cardiac diseases.
Consistent with prior AIS studies in young adults, cortical infarct location was linked to worse outcomes despite IV-thrombolysis.[18,21,30] END after IV recombinant tissue plasminogen activator (r-tPA) may result from failed recanalization, thrombus extension or migration, arterial re-occlusion, or recurrent emboli, suggesting these patients could benefit from endovascular therapy.[31] Patients with END had markedly poorer outcomes in the present study (66.7% showed mRS 3–6 at 90-days). Evidence consistently links neurological deterioration to long-term disability, increasing the odds nearly fourfold for poor outcomes due to END and producing deficits that persist beyond the acute phase, thereby complicating rehabilitation.[32,33]
For logistic regression, the effective sample size was determined by the number of events. With 21 events (END) and 9 candidate predictors, the events-per-variable (EPV) ratio was approximately 3.3, which was well below the conventional threshold of 5–8 EPV recommended for stable multivariable modeling. Under these conditions, the study was only powered to detect large effect sizes (odds ratios ≥2), while smaller associations are unlikely to be reliably identified. Therefore, the multivariable analysis should be interpreted with caution. The results’ robustness and wider applicability may be further limited by the use of NIHSS-based END criteria and the potential for residual confounding.
CONCLUSION
The occurrence of END was observed in 2 of every 5 young adults with AIS in the present study, with large-vessel occlusion emerging as the strongest predictor and associated with poor outcomes in over half of affected individuals. Hypertension and index-event hyperglycemia were identified as important modifiable risk factors. Early recognition of these factors can guide risk stratification and ensure timely support with targeted acute interventions in young adult patients with AIS and END.
Ethical approval:
The research/study was approved by the Institutional Review Board at Bharati Vidyapeeth Deemed University Medical College, Pune, number BVDU/MC/IEC/92/24- 25, dated 28th September 2024.
Declaration of patient consent:
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient has given consent for clinical information to be reported in the journal. The patient understands that the patient’s names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Conflicts of interest:
There are no conflicts of interest.
Use of artificial intelligence (AI)-assisted technology for manuscript preparation:
The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.
Financial support and sponsorship: Nil.
References
- Stroke incidence, mortality, and subtypes in rural and urban populations in five geographic areas of India (2018-2019): Results from the national stroke registry programme. Lancet Reg Health Southeast Asia. 2023;12:100152.
- [Google Scholar]
- Incidence and prevalence of stroke in India: A systematic review. Indian J Med Res. 2017;146:175-85.
- [CrossRef] [PubMed] [Google Scholar]
- Risk factors for acute ischaemic stroke in young adults in South India. J Neurol Neurosurg Psychiatry. 2007;78:959-63.
- [CrossRef] [PubMed] [Google Scholar]
- Stroke in young in India. Stroke Res Treatment. 2011;2011:368629.
- [CrossRef] [PubMed] [Google Scholar]
- Risk factors and etiologies of ischemic strokes in young patients: A tertiary hospital study in north India. J Stroke. 2014;16:173.
- [CrossRef] [PubMed] [Google Scholar]
- Risk factors for ischemic stroke in younger adults. Stroke. 2020;51:729-35.
- [CrossRef] [PubMed] [Google Scholar]
- Clinical outcome in 287 consecutive young adults (15 to 45 years) with ischemic stroke. Neurology. 2002;59:26-33.
- [CrossRef] [PubMed] [Google Scholar]
- Comparative evaluation of risk factors, outcome and biomarker levels in young and old acute ischemic stroke patients. Ann Neurosci. 2015;22:70.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurological deterioration (END) after stroke: The END depends on the definition. Curr Neurol Neurosci Rep. 2011;11:42-9.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurological deterioration in acute ischemic stroke: Predictors, mechanisms and management. J Neurol. 2018;265:216-30.
- [CrossRef] [PubMed] [Google Scholar]
- Outcomes after early neurological deterioration and transitory deterioration in acute ischemic stroke patients. Cerebrovasc Dis. 2016;42:378-86.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurological deterioration in patients with acute ischemic stroke: A prospective multicenter cohort study. Ther Adv Neurol Disord. 2023;16:17562864221147743.
- [CrossRef] [PubMed] [Google Scholar]
- Demographic, clinical, and radiologic predictors of neurologic deterioration in patients with acute ischemic stroke. J Stroke Cerebrovasc Dis. 2013;22:205-10.
- [CrossRef] [PubMed] [Google Scholar]
- What change in the National Institutes of Health Stroke Scale should define neurologic deterioration in acute ischemic stroke? J Stroke Cerebrovasc Dis. 2013;22:675-82.
- [CrossRef] [PubMed] [Google Scholar]
- Hemorrhagic transformation of ischemic brain tissue: Asymptomatic or symptomatic? Stroke. 2001;32:1330-5.
- [CrossRef] [PubMed] [Google Scholar]
- Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial, TOAST. Trial of org 10172 in acute stroke treatment. Stroke. 1993;24:35-41.
- [CrossRef] [PubMed] [Google Scholar]
- The etiologies of early neurological deterioration after thrombolysis and risk factors of ischemia progression. J Stroke Cerebrovasc Dis. 2016;25:383-8.
- [CrossRef] [PubMed] [Google Scholar]
- Predictors of early neurologic deterioration (END) following stroke thrombectomy. J Neurointervent Surg. 2023;15:584-8.
- [CrossRef] [PubMed] [Google Scholar]
- Predictors of early neurological deterioration in patients with acute ischemic stroke. Front Neurol. 2024;15:1433010.
- [CrossRef] [PubMed] [Google Scholar]
- Incidence, causes and predictors of neurological deterioration occurring within 24 h following acute ischaemic stroke: A systematic review with pathophysiological implications. J Neurol Neurosurg Psychiatry. 2015;86:87-94.
- [CrossRef] [PubMed] [Google Scholar]
- The rate of early neurological deterioration occurring after thrombolytic therapy: A meta-analysis. Brain Behav. 2019;9:e01210.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurologic deterioration after stroke depends on vascular territory and stroke etiology. J Stroke. 2016;18:203-10.
- [CrossRef] [PubMed] [Google Scholar]
- Association between neurological deterioration and outcomes in patients with stroke. Ann Transl Med. 2020;8:4.
- [CrossRef] [PubMed] [Google Scholar]
- Early Neurological Deterioration following acute stroke: Association with reperfusion therapies and national institutes of health stroke scale score. Front Stroke. 2025;4:1518685.
- [CrossRef] [PubMed] [Google Scholar]
- Predictive value of the alberta stroke program early CT score (ASPECTS) in the outcome of the acute ischemic stroke and its correlation with stroke subtypes, NIHSS, and cognitive impairment. Stroke Res Treat. 2021;2021:1-10.
- [CrossRef] [PubMed] [Google Scholar]
- Clopidogrel plus aspirin vs aspirin alone in patients with acute mild to moderate stroke: The ATAMIS randomized clinical trial. JAMA Neurol. 2024;81:450.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurological deterioration and time to start dual antiplatelet therapy in patients with acute mild-to-moderate ischemic stroke: A pre-specified post hoc analysis of the ATAMIS trial. J Stroke. 2024;26:403-14.
- [CrossRef] [PubMed] [Google Scholar]
- Efficacy and safety of dual-antiplatelet therapy with high-intensity statin versus single-antiplatelet therapy with high-intensity statin in patients with stroke or high-risk mini-stroke of atherosclerotic origin: A cohort study. J Neurosci Rural Pract. 2025;16:44-54.
- [CrossRef] [Google Scholar]
- Predictors of early neurological deterioration in patients with acute ischaemic stroke with special reference to blood urea nitrogen (BUN)/creatinine ratio and urine specific gravity. Indian J Med Res. 2015;141:299-307.
- [CrossRef] [PubMed] [Google Scholar]
- Predictors of early neurological deterioration following intravenous thrombolysis: Difference between risk factors for ischemic and hemorrhagic worsening. Ann Indian Acad Neurol. 2022;25:627-33.
- [CrossRef] [PubMed] [Google Scholar]
- Endovascular thrombectomy after large-vessel ischaemic stroke: A meta-analysis of individual patient data from five randomised trials. Lancet. 2016;387:1723-31.
- [CrossRef] [PubMed] [Google Scholar]
- Early neurological deterioration during the acute phase as a predictor of long-term outcome after first-ever ischemic stroke. Medicine (Baltimore). 2017;96:e9068.
- [CrossRef] [PubMed] [Google Scholar]
- Effects of early neurological deterioration in acute ischemic stroke on multiple dimensions of functional outcomes using propensity score matching analysis. Medicine (Baltimore); 2020 [preprint]
- [CrossRef] [Google Scholar]