Cumulative Smoking Exposure and Carotid Plaque in Ischemic Stroke Without Traditional Risk Factors

Article information

J Neurosonol Neuroimag. 2026;18(1):26-33

Publication date (electronic) : 2026 June 30

doi : https://doi.org/10.31728/jnn.2026.00182

Hyo Jeong Choi, MD, Youngchan Jung, MD, Jin Kim, MD, Sang Won Han, MD

Department of Neurology, Sanggye Paik Hospital, Inje University College of Medicine, Seoul, Korea

Correspondence: Sang Won Han, MD Department of Neurology, Sanggye Paik Hospital, Inje University College of Medicine, 1342 Dongil-ro, Nowon-gu, 01757, Seoul, Korea Tel: +82-2-950-1090 Fax: +82-2-950-1955 E-mail: swhan@paik.ac.kr

Received 2026 May 9; Revised 2026 June 5; Accepted 2026 June 23.

Abstract

INTRODUCTION

Stroke remains a leading global cause of mortality and long-term disability. While the majority of cerebrovascular events are driven by standard modifiable stroke risk factors (SMoRFs) such as hypertension, diabetes mellitus (DM), dyslipidemia, and atrial fibrillation, a substantial proportion of strokes occurs in individuals entirely free of these traditional risk factors.1,2 Cigarette smoking is a well-established independent risk factor for stroke, increasing the risk by two- to four-fold compared to non-smokers.1,3 It is a potent, independent driver of oxidative stress, endothelial dysfunction, and chronic inflammation, and has been linked to accelerated atherogenesis.4-6 However, its specific pathophysiological footprint in the absence of other confounding metabolic risk factors remains incompletely characterized. Many SMoRF-free stroke patients present at a relatively younger age, particularly among active smokers. Carotid intima-media thickness (CIMT) and the presence of carotid plaques serve as validated, noninvasive surrogate markers for systemic subclinical atherosclerosis and are robust predictors of future ischemic stroke events.7,8 Because chronological age is the most powerful determinant of CIMT and carotid plaque, raw anatomical comparisons often yield milder vascular pathology in smokers at the time of stroke onset, given their younger age. This age discrepancy may mask the vascular impact of tobacco toxicity. Furthermore, while chronic smoking is known to induce a pro-thrombotic and pro-inflammatory state, the dose-dependent impact of cumulative smoking exposure on arterial wall integrity in SMoRF-free stroke patients is rarely quantified.

In this SMoRF-free population, uncovering covert drivers of vascular disease is clinically important. To address these knowledge gaps, this study aimed to investigate the relationship between smoking status, cumulative tobacco exposure, and carotid structural changes in patients presenting with a first-ever ischemic stroke or transient ischemic attack (TIA) who were free from SMoRF. We hypothesized that in the absence of other traditional risk factors, cumulative tobacco exposure is associated with CIMT progression and focal carotid plaque formation, and we sought to explore a potential dose-dependent relationship between smoking intensity and carotid atherosclerosis in this population.

SUBJECTS AND METHODS

This was a retrospective, cross-sectional observational study. Consecutive patients aged over 20 years who experienced a first-ever ischemic stroke or TIA within 5 days of symptom onset were screened between January 2020 and December 2024. For the study, patients who met the following criteria were included: 1) experienced a first-ever ischemic stroke or TIA, 2) were either current smokers or never smokers and had no other SMoRF.

Risk factors were collected through interviews of the patients and medical records. Hypertension was defined as prior diagnosis of hypertension, the use of antihypertensive medications, or systolic blood pressure (BP) ≥ 140 mmHg. Elevated BP during the acute phase did not constitute a diagnosis of hypertension. DM was defined as documented history, prior use of glucose-lowering agents, fasting blood glucose ≥126 mg/dL, or an admission HbA1c ≥6.5%. Dyslipidemia was defined as having history of dyslipidemia, lipid-lowering medications, total cholesterol ≥240 mg/dL, or low-density lipoprotein (LDL)-cholesterol ≥160 mg/dL. Cigarette smoking was categorized as never smoker or current smoker, and the pack-years of smoking (PYS) were calculated in current smoking patients. The demographic characteristics, clinical information, SMoRFs, and routine blood studies were collected. Prior medications regularly taken before admission were reviewed on the basis of the hospital records. Brain CT or MRI were performed in all patients, along with an angiographic study. Other diagnostic tests including echocardiography, 24-hour Holter, and thrombophilia screening were conducted in selected patients to identify possible other determined etiology. Stroke subtypes were classified in accordance with the trial of ORG 10172 in the acute stroke treatment (TOAST) classification system.

To ensure diagnostic consistency, a single experienced clinician executed all carotid evaluations following a rigorous standardized protocol.7,9 High-resolution B-mode imaging was performed using ACOUSON S1000 ultrasound system equipped with a linear transducer (Siemens Healthineers, Erlangen, Germany), adhering to the Mannheim intima-media thickness (IMT) consensus guidelines.10 During the examination, patients were positioned supine with their necks extended and heads aligned to facilitate optimal bilateral visualization. IMT was defined as the distance between the luminal-intimal boundary and the medial-adventitial interface along the far wall of the artery. Measurements were focused on a longitudinal segment of the common carotid artery (CCA), specifically 10 mm proximal to the carotid bulb. This anatomical landmark, identified by the point where the artery’s parallel wall configuration begins to dilate, provided a stable and reproducible reference point for distal CCA assessment. For the analysis, the maximum IMT value obtained from either the left or right IMT was utilized. The mean IMT was calculated as a mean value of both CIMT. In accordance with established standards, a carotid plaque was identified as a focal wall thickening that either protruded at least 0.5 mm into the lumen, exceeded 50% of the adjacent IMT, or reached an absolute thickness of over 1.5 mm.

The present study was reviewed and approved by the Inje University Sanggye Paik Hospital Research Ethics Committee (IRB No. 2026-05-005). These data were collected by a retrospective review of institutional electronic database. Since all data had been acquired as part of routine clinical practice and was used exclusively for analytical purposes, the requirement for informed consent to participate was waived. Personal identifying information was excluded from the study, ensuring that researchers had access only to the aggregated results.

RESULTS

A total of 2,810 patients with a first-ever ischemic stroke or TIA were screened for enrollment, of whom 204 met the inclusion criteria for the analysis. Baseline demographic and clinical characteristics are summarized in Table 1. Seventy-two patients (35.3%) were current smokers, with a mean smoking intensity of 29.3±10.14 PYS. The current smoking group was significantly younger (p<0.001) and predominantly male (88.9% vs. 37.9%; p<0.001) compared to the never smoking group. There were no intergroup differences in body mass index, antiplatelet use, or TOAST classification. As detailed in Table 2, current smokers exhibited a pro-thrombotic and atherogenic profile, with significantly elevated hemoglobin (Hb), hematocrit (Hct), creatinine, white blood cell counts, platelet counts, and triglycerides, alongside reduced high-density lipoprotein cholesterol. Levels of international normalized ratio and high-sensitivity C-reactive protein did not differ between the groups.

Table 1.

Baseline demographic and clinical characteristics of the study population

Table 2.

Laboratory profiles of the study population

Regarding carotid parameters, both mean and maximum CIMT were significantly lower in current smokers in the unadjusted analysis (p=0.032 and p=0.027, respectively), a finding consistent with their younger age. The prevalence of carotid plaque did not differ significantly between current smokers and never-smokers (44.4% vs. 54.5%; p=0.168). After adjusting for age and sex, there was no significant effect of smoking status on mean or maximum CIMT (p=0.327 and p=0.276, respectively). Thus, a simple binary comparison of current smokers and never-smokers did not reveal a significant difference in carotid plaque burden or age-adjusted CIMT.

To explore the impact of smoking intensity, current smokers were stratified by cumulative exposure: light (<23 PYS, n=17), moderate (23 to <30 PYS, n=18), heavy (30 to 36 PYS, n=19), and extreme (≥36 PYS, n=18) (Table 3). Importantly, mean age increased markedly with smoking intensity, ranging from 40.7 years in the light group to 69.6 years in the extreme group (p<0.001), indicating strong collinearity between cumulative pack-years and chronological age. In the unadjusted analysis, CIMT increased across strata (mean IMT from 0.56 mm to 0.76 mm, p=0.001; maximum CIMT from 0.60 mm to 0.81 mm, p<0.001), and carotid plaque prevalence rose in a dose-dependent manner from 11.8% in light smokers to 77.8% in extreme smokers (p<0.001). However, these dose-dependent associations did not persist after age adjustment. In the age-adjusted multivariable logistic regression, the overall categorical smoking-intensity variable was not statistically significant (p=0.312). Using the light smoking group as the reference, the point estimates for the odds of plaque presence increased across strata—moderate (odds ratio [OR], 6.72; 95% confidence interval [CI], 0.88–51.13; p=0.073), heavy (OR, 10.30; 95% CI, 0.78–135.65; p=0.082), and extreme (OR, 20.78; 95% CI, 0.67–644.48; p=0.084)—but none reached statistical significance, and all confidence intervals were very wide and crossed 1.0. Age was not significantly associated with plaque presence in this model (OR, 0.99; 95% CI, 0.90–1.09; p=0.836). Likewise, after adjusting for age, no differences were found in mean or maximum CIMT across the smoking categories (p=0.882 and p=0.725, respectively). Taken together, these data show a dose-dependent trend in the unadjusted analysis that did not remain statistically significant after age adjustment, and should therefore be regarded as exploratory and hypothesis-generating.

Table 3.

Descriptive statistics of the enrolled patients stratified by cumulative smoking exposure

While the distribution of TOAST stroke subtypes was not statistically significant across the subgroups (p=0.278), observational trends indicated that lacunar infarctions increased along with smoking intensity, representing 41.2% of strokes in the lightest smokers and rising to 66.7% in the extreme smokers (Table 4). Conversely, strokes of undetermined etiology were most prevalent in the youngest, light smoking group (41.2%) and steadily declined to 5.6% in the extreme smoking group. Up to the moderate smoking group, Hb and Hct levels increased as a physiological compensatory response to chronic hypoxia, which stimulates erythropoiesis. However, in the extreme smoking group, these levels declined despite the highest cumulative tobacco exposure. This phenomenon is likely attributable to the advanced chronological age of this group.

Table 4.

Laboratory profiles stratified by cumulative smoking exposure

DISCUSSION

This study explored the association between cigarette smoking and carotid atherosclerosis in patients with acute ischemic stroke in the absence of SMoRFs. Traditionally, macroscopic plaque formation is conceptualized as a late-stage, culminating manifestation of atherosclerosis, heavily dependent on age-related exposure to multiple risk factors.1,6,11 Within the current smoking group, several younger patients harbored a plaque burden comparable to that of older never-smokers, an observation that suggests cumulative tobacco exposure may contribute to earlier focal atheroma. Two findings should be distinguished. First, in the direct comparison of current smokers and never-smokers, carotid plaque prevalence did not differ significantly (44.4% vs. 54.5%; p=0.168), and although unadjusted CIMT was lower in smokers, this difference disappeared after age adjustment. This pattern is most plausibly explained by the substantially younger age of the smoking group rather than by a protective effect of smoking. Second, within current smokers, plaque prevalence increased with cumulative exposure in the unadjusted analysis, but this dose-dependent association did not remain statistically significant after age adjustment, with wide confidence intervals that crossed 1.0. We therefore interpret the within-smoker dose-response as a trend that is hypothesis-generating rather than as evidence of an independent causal effect.

Strokes in SMoRF-free patients are frequently categorized as cryptogenic or Embolic Strokes of Undetermined Source.12 Many of these cryptogenic cases may be attributed to non-obstructive, highly unstable carotid atherosclerosis.6,12 A high-risk, complicated carotid plaques, even those causing less than 50% luminal stenosis are highly prevalent on the ipsilateral side of cryptogenic strokes.13 These non-stenosing plaques can easily rupture or undergo superficial ulceration, serving as a potent nidus for localized thrombosis and subsequent artery-to-artery embolism. The cumulative tobacco exposure induces severe oxidative stress within vascular endothelial cells and diminishes the bioavailability of nitric oxide, culminating in endothelial dysfunction.5,14 This initial vascular injury upregulates the expression of pro-inflammatory cytokines and adhesion molecules within the arterial wall, which accelerates the subendothelial infiltration of monocytes and their subsequent transformation into lipid-laden foam cells.14 Ultimately, this inflammatory cascade promotes thickening of the intima-media layer and the advancement of atherosclerotic plaques. These mechanisms provide biological plausibility for the dose-dependent trend observed in our unadjusted analysis, although our study was not designed or powered to confirm an independent effect.6,13 Unlike the carotid plaque prevalence, a direct correlation between smoking status and CIMT was not clearly observed in this study, likely due to the specific age distribution of patients. However, prior epidemiological studies consistently identify smoking as a major risk factor for the initiation of carotid atherosclerosis and the increased CIMT.15-18 The absence of a pronounced effect of smoking on CIMT in our study was most likely attributable to the masking effect of age-related confounding variables.

The primary strength of this study was its highly specific focus on the SMoRF-free stroke population, effectively isolating the distinct vascular impact of smoking. Nonetheless, several important limitations temper our interpretation. First, and most importantly, there was a large and significant age difference between current smokers and never-smokers (mean 55.4 vs. 67.3 years), and within the smoking group mean age increased steeply with smoking intensity (from 40.7 years in light smokers to 69.6 years in extreme smokers). Because pack-years of smoking is an inherently time-accumulating variable, it is strongly collinear with chronological age, and age is itself the most powerful determinant of structural vascular remodeling, increased CIMT, and carotid plaque progression. Consequently, the crude dose-dependent association between cumulative smoking and carotid plaque cannot be confidently separated from the effect of chronological vascular aging. Although we applied age-adjusted multivariable models, the association did not remain statistically significant after adjustment, and residual confounding cannot be excluded. The findings should therefore be interpreted with caution. Second, this study was cross-sectional design, which inherently restricts our ability to establish definitive causality or precise temporal sequences. Additionally, reliance on self-reported smoking histories may introduce bias. Third, the wide confidence intervals and marginal p values in our logistic regression model reflected the limited statistical power inherent in our relatively small subgroup sample sizes; the age-adjusted regression did not reach statistical significance, and the point estimates should not be over-interpreted. There was a disproportionate sex distribution, with the smoking group being predominantly male (88.9%) compared to never smokers (37.9%). This imbalance, largely driven by historical demographic smoking patterns, precluded a fully powered analysis of sex-specific interactions regarding smoking and plaque formation. Despite these limitations, the progressive increase in the odds ratios (from approximately 6.7 to 20.8) is consistent with, but does not confirm, a dose-dependent trend of cumulative smoking on carotid plaque formation in SMoRF-free patients; the association did not reach statistical significance after age adjustment.

In conclusion, in patients with ischemic stroke free of standard modifiable risk factors, cumulative smoking exposure showed a dose-dependent trend toward higher carotid plaque prevalence in unadjusted analysis. However, because smoking intensity was strongly confounded by age and the association did not remain statistically significant after age adjustment, these findings are hypothesis-generating and do not establish smoking as an independent driver of plaque formation in this model. Even so, the consistent direction of the trend suggests that conventional stroke risk calculators may underestimate risk in seemingly healthy SMoRF-free smokers. The detection of carotid plaque in a SMoRF-free smoker is a reasonable trigger for smoking cessation counseling and consideration of disease-modifying therapies, such as statin therapy for plaque stabilization. Future prospective, adequately powered studies using strictly age-matched stroke patients are warranted to isolate the vascular damage attributable to smoking independent of normal aging.

Article information Continued