This study involved a prospective analysis of consecutive patients who presented with an acute ischemic stroke within 7 days of symptom onset at Asan Medical Center (Seoul, South Korea) between July 2021 and July 2022. We excluded patients who exhibited severe neurological deficits that precluded completion of study questionnaires. The patients with the following conditions were excluded: (1) altered mental status (drowsiness or more severe decreased consciousness, or disorientation and confusion as diagnosed by attending neurologists); (2) severe motor weakness (Medical Research Council score ≤3); (3) significant hemispheric cortical signs, such as aphasia, neglect, or visual disturbances (including diplopia or hemianopia); (4) severe dementia (14); (5) a previous modified Rankin score greater than 3; (6) no acute lesion on diffusion-weighted imaging (DWI); and (7) active systemic diseases (such as infection or active cancer) (15). Eligible patients completed the State–Trait Anxiety Inventory (STAI) and Beck Depression Inventory (BDI) questionnaires after neurological stabilization. We further excluded patients with lesions involving both anterior and posterior circulations, absence of acute lesions on imaging, or incomplete questionnaire data.

The study was approved by the Institutional Review Board of Asan Medical Center (approval number: 2021–0901), and written informed consent was obtained from all participants. All study procedures were conducted in accordance with the Declaration of Helsinki.

Demographic information and clinical variables were collected. Neurological examinations at enrollment were conducted by an experienced neurologist. The cause of stroke was categorized according to the Trial of Org 10,172 in Acute Stroke Treatment classification system (16). Furthermore, neurological deficits were evaluated using the National Institutes of Health Stroke Scale (NIHSS).

All patients underwent MRI on a 3.0 T Philips scanner (Philips Healthcare, Eindhoven, Netherlands), including DWI and MR angiography. The locations of acute lesions were assessed on DWI and classified as follows: PCS lesions, including those in the cerebellum [superior cerebellar, anterior inferior cerebellar, and posterior inferior cerebellar (PICA) artery territories], medulla oblongata, pons, midbrain, thalamus, and medial temporo-occipital areas; ACS lesions, including those in the frontal cortex/subcortex, parietal cortex/subcortex, temporal cortex/subcortex, deep structures, and insula. Multiple lesion locations were recorded if a patient had ischemic lesions in more than one region, with each classified as a separate variable; thus, a single patient could contribute multiple lesion sites (5). Dorsal brainstem stroke was defined as involvement of the dorsal medulla oblongata or the tegmentum of the pons and midbrain, regions implicated in the central vestibular pathway (17). Lesion locations were determined by consensus between two stroke experts (SH Ha and E-J Lee) who were blinded to clinical information.

Small vessel disease burden was evaluated on fluid-attenuated inversion recovery (FLAIR) imaging. White matter hyperintensities (WMHs) were defined as periventricular hyperintense lesions on T2-weighted FLAIR and graded according to the modified Fazekas scale: 0 = absent; 1 = pencil-thin lining; 2 = halo ≥5 mm in thickness; 3 = irregular WMHs extending into the deep white matter. Patients were subsequently dichotomized into two groups: grades 0–1 versus 2–3 (18). Cerebral microbleeds (CMBs) were defined as small perivascular hemosiderin deposits, appearing as rounded, homogeneous, hypointense lesions on T2*-weighted gradient-recalled echo or susceptibility-weighted MRI, and both lobar and deep CMBs were included. Lacunes were defined as chronic small cavities, presumed to represent the healed stage of a lacunar infarct (19).

We investigated the occurrence of new-onset dizziness following stroke, focusing on whether patients experienced any dizziness symptom regardless of its qualitative presentation (20) (Figure 1). Specifically, dizziness was assessed using a structured patient interview; patients were first asked whether they had experienced any abnormal sensations related to dizziness or balance since stroke onset. If present, they were asked to select symptom descriptors that best characterized their experience, including spinning sensation, unsteadiness or imbalance, lightheadedness or near-faintness, and vague or difficult-to-describe dizziness. The presence of dizziness was defined as a positive response to any of these symptom descriptors. This broad, symptom descriptor–based definition was adopted to minimize under-detection of post-stroke dizziness by encompassing both vertiginous and non-vertiginous manifestations. To ensure clinical consistency, the presence of dizziness was verified through structured interviews conducted by two trained neurological nurse specialists (GY Park and DY Seo), who were blinded to all clinical information. In cases where the nature of the symptoms was ambiguous, the final determination of whether the patient was symptomatic was made by consensus between the two raters.

We additionally administered standardized anxiety and depression questionnaires, given that dizziness is often nonspecific and may be influenced by emotional symptoms. The State–Trait Anxiety Inventory (STAI), a validated instrument for assessing trait and state anxiety, includes 20 items for trait anxiety, with scores ranging from 20 to 80 (higher scores indicating greater anxiety) (21). The Beck Depression Inventory (BDI) is a 21-item self-report scale for depressive symptoms, with responses from 0 (not at all) to 3 (very much) and total scores ranging from 0 to 63 (22).

Baseline characteristics were compared between ACS and PCS patients. Continuous variables were expressed as mean ± standard deviation or median [interquartile range] after testing for normality using the Shapiro–Wilk test. Comparisons between groups were performed using Student’s t-test or the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. To identify independent factors associated with dizziness, we performed multivariable logistic regression analyses. Variables that reached a entry threshold of p < 0.1 in univariable analysis, as well as age, were included as covariates. For the ACS group, two distinct models were constructed to avoid multicollinearity between the presence and the count of cerebral microbleeds. All analyses were performed using IBM SPSS Statistics, version 21.0 (IBM Corp., Armonk, NY, USA). Statistical significance was set at p < 0.05.

During the study period, a total of 976 patients with acute ischemic stroke were admitted to our center, and 265 patients completed the survey (Figure 2). All questionnaires were administered within a median of 1 day (range, 1–3 days) after stroke onset. Compared with the non-survey group (n = 711), survey completers were younger (64 ± 13 vs. 68 ± 14 years; p < 0.001) and had lower initial NIHSS scores [2 (0–4) vs. 3 (1–8); p < 0.001] (Supplementary Table S1). Of the surveyed patients, 4 with both anterior and posterior circulation stroke, 37 without an acute lesion on DWI, and 55 with incomplete survey data were excluded. The final analysis included 169 patients (17.3%), comprising 98 with ACS (58.0%) and 71 with PCS (42.0%).

The demographic and baseline characteristics are summarized in Table 1. The mean age was 63 ± 13 years, 112 patients (66.3%) were male, and the median initial NIHSS score was 2 (0–4). The mean BDI and STAI scores were 11 ± 9 and 41 ± 8, indicating mild depressive symptoms and moderate anxiety (21, 22). No significant differences in baseline characteristics were observed between the ACS and PCS groups, including stroke etiology (TOAST classification) and neurological severity (NIHSS score).

Dizziness was identified in 45 patients with ACS (45.9%) and 43 patients with PCS (60.6%) (Table 2). Although dizziness tended to be more frequent in PCS, the prevalence in ACS was also substantial, affecting nearly half of the patients (p = 0.060). Acute reperfusion therapy was administered in a small proportion of patients (4.1% received intravenous thrombolysis and 7.7% underwent intra-arterial thrombectomy), with similar rates between ACS and PCS. Importantly, the frequency of dizziness did not differ according to thrombolytic treatment status.

We assessed clinical factors associated with post-stroke dizziness in ACS and PCS separately (Table 3). No significant differences in age or neurological severity were observed between patients with and without dizziness. Similarly, depressive (BDI) and anxiety (STAI) symptoms did not differ by dizziness status in either group.

MRI findings are summarized in Table 3. In ACS, the most frequent acute lesion site was the frontal cortex and subcortex (50.0%), whereas in PCS it was the pons (25.3%). Insular involvement was observed in 16.3% of patients. The median Fazekas grade was 1 [1–2] in ACS and 1 [0–1] in PCS, with 30 (30.6%) and 14 (19.7%) showing moderate leukoaraiosis (Fazekas ≥2). CMBs were detected in 19 (19.3%) of ACS patients and 9 (12.6%) of PCS patients. Between patients with and without dizziness, those with dizziness were more likely to have microbleeds in the ACS group (28.9% vs. 11.3%; p = 0.028), whereas in the PCS group, dizziness was significantly associated with medullary or cerebellar involvement (46.5% vs. 21.4%; p = 0.032).

Multivariable logistic regression was performed to determine whether MRI findings significant in univariate analyses remained after adjustment for potential confounders, including age. In the ACS group, Model 1 revealed that the presence of microbleeds was independently associated with dizziness [adjusted odds ratio (aOR) = 3.19, 95% confidence interval (CI) 1.09–9.32; p = 0.034], and Model 2 showed that the number of microbleeds was significantly associated with dizziness (aOR = 2.38, 95% CI 1.10–5.15; p = 0.026) (Table 4). In the PCS group, dizziness was independently associated with medullary or cerebellar involvement (aOR = 3.13, 95% CI 1.01–9.74; p = 0.048) (Table 5).