Abstract
Background
Despite abundant evidence linking hyperglycemia to stroke, its heterogeneous effects across stroke etiologies and lesions remain unclear. We examined the associations between admission glycated hemoglobin (HbA1c) levels and stroke subtypes, lesions, and functional outcomes.
Methods
Adults with acute ischemic stroke admitted between 2016 and 2020 were analyzed using a Japanese nationwide registry. Patients were grouped by admission HbA1c (<6.0%, 6.0–6.9%, 7.0–7.9%, and ≥ 8.0%). Stroke subtypes were determined using the Trial of Org 10,172 in Acute Stroke Treatment criteria. Lesion was categorized as cerebral cortex, basal ganglia/corona radiata, thalamus, brainstem, or cerebellum. Unfavorable outcome was defined as discharge modified Rankin Scale (mRS) >2 for patients with premorbid mRS ≤2, or higher discharge mRS than premorbid mRS for others. Associations were assessed using multivariable mixed-effects logistic regression analysis.
Results
Among 13,569 patients, HbA1c was <6.0% in 7001 (51.6%), 6.0–6.9% in 4201 (31.0%), 7.0–7.9% in 1264 (9.3%), and ≥ 8.0% in 1103 (8.1%). Higher HbA1c levels were associated with a greater proportion of large artery atherosclerosis (adjusted OR for HbA1c ≥8.0% vs <6.0%, 1.40; 95% CI, 1.19–1.65) and lower proportion of cardioembolism (0.63; 0.50–0.79). Elevated HbA1c levels were associated with higher prevalence of brainstem lesions (2.35; 1.82–3.02) and lower prevalence of cerebral cortex and basal ganglia/corona radiata lesions, even after adjusting for stroke subtype. Unfavorable outcomes were more common in patients with higher HbA1c levels (overall, 1.56; 1.31–1.85), particularly in those with large artery atherosclerosis and cardioembolism.
Conclusions
Higher admission HbA1c levels were associated with a higher proportion of large artery atherosclerosis and a lower proportion of cardioembolism, with a higher proportion of brainstem lesions and a lower proportion of cortical and basal ganglia/corona radiata lesions, and with worse functional outcomes.
Keywords: Glycated hemoglobin (HbA1c), Ischemic stroke, Large artery atherosclerosis, Brainstem
Highlights
-
•
We examined the association between HbA1c levels and stroke characteristics.
-
•
Elevated HbA1c was associated with higher prevalence of brainstem lesions.
-
•
Higher admission HbA1c was linked to a shift toward large artery atherosclerosis.
1. Introduction
Stroke affects over 100 million people worldwide and is the third leading cause of death and disability [1]. Diabetes mellitus is a recognized risk factor for stroke and is associated with poor outcomes among patients with stroke [2]. Glycated hemoglobin (HbA1c), an index of long-term glycemic control, is commonly used to diagnose diabetes and assess the impact of hyperglycemia on cardiovascular outcomes [3], [4]. Elevated HbA1c levels are associated with poor stroke prognosis, and each 1.0% increase in HbA1c is associated with a 37% increase in post-stroke mortality [5].
Despite abundant evidence that hyperglycemia increases the risk of stroke and worsens outcomes, strict glycemic control aimed at lowering HbA1c levels has not clearly reduced stroke incidence or improved prognosis in individuals with acute stroke [6]. The unclear effects of glycemic control on ischemic stroke may be due to the disease's heterogeneity, which arises from variations in etiology and lesion location. The effect of glycemic control likely varies depending on these factors. Some studies have linked genetic susceptibility to type 2 diabetes with large artery atherosclerosis and small vessel occlusion subtypes of ischemic stroke [7]. However, the overall effect of diabetes on stroke etiology and lesions in clinical settings remains uncertain. In this study, we aimed to investigate how HbA1c levels affect the etiology, lesions, and functional outcomes of ischemic stroke.
2. Methods
2.1. Study design and data source
This retrospective observational cohort study used data from the Japan Stroke Data Bank (JSDB). Detailed information on the JSDB was described elsewhere [8], [9], [10]. The JSDB is an ongoing hospital-based, multicenter, prospective registry of patients with acute stroke or transient ischemic attack in Japan. The JSDB prospectively obtains information associated with the clinical diagnosis of acute stroke, based on professional evaluations by stroke experts from 130 certified stroke centers and academic teaching hospitals. Data on hospitalizations attributable to acute stroke were recorded by the treating physician or clinical report coordinators at each institute using a standardized online database. Routine blood tests, including HbA1c, were obtained at admission, starting in October 2016.
2.2. Standard protocol approvals, registrations, and patient consents
The Institutional Ethics Review Board of the National Cerebral and Cardiovascular Center approved the study protocol (M27–090-15). Owing to the anonymous nature of the data, the requirement for individual consent to enter the information into the database was waived by the institution. Instead, an opt-out consent method was used to notify or disclose research information.
2.3. Patients and data collection
Adult patients (≥18 years) diagnosed with acute ischemic stroke with available data on HbA1c at admission were eligible for inclusion into the study. The study cohort was recruited between October 2016 and December 2020. The database had clinical information on demographic variables, National Institutes of Health Stroke Scale (NIHSS) scores at admission, medical history, risk factors, medications, premorbid modified Rankin Scale (mRS) scores, and mRS scores at discharge. Stroke lesions were recorded according to the JSDB data input format as involvement of the cerebral cortex, basal ganglia/corona radiata, thalamus, brainstem, or cerebellum, based on the imaging diagnosis (CT or MRI) entered by the treating stroke physician at each institution. Because patients could have lesions in more than one region, each lesion location was analyzed as a separate, non-mutually-exclusive binary outcome. Stroke subtype was classified into the five subtypes according to the Trial of Org 10,172 in Acute Stroke Treatment criteria: large artery atherosclerosis, cardioembolism, small vessel occlusion, other determined etiology stroke, or undetermined etiology stroke [11]. Patients were categorized into four HbA1c categories as recommended for Japanese populations by the Japanese Clinical Practice Guideline for Diabetes 2019: (1) HbA1c <6.0%; (2) 6.0% ≤ HbA1c <7.0%; (3) 7.0% ≤ HbA1c <8.0%; (4) HbA1c ≥8.0% [12]. Unfavorable functional outcome was defined as discharge modified Rankin Scale (mRS) >2 for patients with premorbid mRS ≤2, or higher discharge mRS than premorbid mRS for others.
2.4. Statistical analysis
Baseline characteristics were summarized according to HbA1c levels using descriptive statistics. Continuous data are shown as medians (interquartile ranges), and categorical data as n (%). Differences between groups were assessed using Mann–Whitney U and χ2 tests, respectively. A mixed-effects binary logistic regression, with the treating institution included as a random intercept, was used to evaluate the associations between HbA1c levels and the three main outcomes (stroke subtype, lesion location, and unfavorable functional outcome). Restricted cubic spline analyses were also conducted with three knots (6.0, 7.0, and 8.0%), corresponding to the same clinically relevant thresholds used for the HbA1c categories described above [12], calculating the odds ratios (ORs) at each 0.1% increment in HbA1c. For categorical analyses, an HbA1c level of <6.0% was used as the reference, while 6.0% was used as the reference for the spline models. Covariates were selected for each outcome based on previous literature [8], [13]. The model for stroke subtype was adjusted for age, sex, calendar year, hypertension, previous statin use, chronic kidney disease, and atrial fibrillation. The analysis for stroke lesion location was conducted only in the subgroup with available lesion data after comparing the subgroup's baseline characteristics with those of the full cohort to assess potential selection bias. The model for stroke lesions was adjusted for the same covariates as the subtype model, plus stroke subtype, following the hypothesis that subtype, as the primary pathophysiology, influences lesion location. Finally, the model for unfavorable functional outcomes was adjusted for age, sex, calendar year, hypertension, history of stroke, previous statin use, initial NIHSS score, and acute reperfusion therapy. The calendar year was included to account for potential year-to-year variations in diagnostic and treatment practices, including those that occurred during the COVID-19 pandemic period. Outliers were not excluded owing to their potentially important behavior in our research question. The analyses were performed using complete-case data without imputation. As this was an analysis of all eligible patients in the registry, no a priori sample size calculation was performed. Patients were observed during the index hospitalization, with outcomes assessed at discharge; no longitudinal follow-up beyond discharge was conducted. Given the hypothesis-generating nature of this study, all analyses were considered exploratory. We therefore focused on point estimates with 95% confidence intervals rather than on hypothesis testing, and no correction for multiple testing was applied. STATA version 16 (StataCorp LP) was used to conduct analyses.
3. Results
Among the 48,961 patients with ischemic stroke admitted during the study window, admission HbA1c data were available for 13,633 (27.8%). The baseline characteristics were generally similar between patients with and without admission HbA1c (Table S1). After excluding 10 patients aged <18 years and 54 patients with missing etiological data, 13,569 patients were included in the main analysis (Fig. S1). The median patient age was 75 years, and 41% of the patients were women. Among these patients, 12,889 (95%) underwent brain magnetic resonance imaging (MRI) on admission. HbA1c was <6.0% in 7001 (51.6%), 6.0–6.9% in 4201 (31.0%), 7.0–7.9% in 1264 (9.3%), and ≥ 8.0% in 1103 (8.1%). Baseline characteristics of the four groups classified according to HbA1c levels are shown in Table 1. Across the HbA1c strata, patients with higher HbA1c levels were younger and less often female, with lower rates of pre-stroke disability. Hypertension and previous use of statins were more prevalent in patients with higher HbA1c levels, while atrial fibrillation was less common. The proportion of patients receiving reperfusion therapy was lower in the group with the highest HbA1c levels. The baseline NIHSS scores and the proportion of patients undergoing acute-phase MRI were similar across the HbA1c strata.
Table 1.
Baseline characteristics by HbA1c strata.
| HbA1c < 6.0% (n = 7001) |
6.0% ≤ HbA1c < 7.0% (n = 4201) |
7.0% ≤ HbA1c < 8.0% (n = 1264) |
HbA1c ≥ 8.0% (n = 1103) |
|
|---|---|---|---|---|
| HbA1c, % | 5.6 [5.4–5.8] | 6.3 [6.1–6.5] | 7.4 [7.1–7.6] | 9.0 [8.4–10.2] |
| Age, years | 77 [67–85] | 77 [70–84] | 75 [69–82] | 71 [63–79] |
| Woman | 3114 (44.5) | 1693 (40.3) | 346 (27.4) | 353 (32.0) |
| Prestroke disability⁎ | 1355 (19.4) | 703 (16.7) | 207 (16.4) | 142 (12.9) |
| Hypertension | 4466 (63.8) | 2982 (71.0) | 934 (73.9) | 752 (68.2) |
| Previous stroke | 1657 (23.7) | 1073 (25.5) | 331 (26.2) | 245 (22.2) |
| Previous statin use | 1097 (21.3) | 1065 (31.0) | 369 (32.9) | 257 (32.2) |
| Chronic kidney disease | 572 (8.2) | 335 (8.0) | 111 (8.8) | 63 (5.7) |
| Atrial fibrillation | 1181 (16.9) | 752 (17.9) | 173 (13.7) | 98 (8.9) |
| Initial NIHSS | 3 [1–9] | 3 [1–8] | 3 [1–7] | 3 [2–7] |
| MRI in acute phase | 6647 (94.9) | 3976 (94.6) | 1207 (95.5) | 1059 (96.0) |
| Alteplase treatment | 612 (8.7) | 414 (9.9) | 111 (8.8) | 71 (6.4) |
| Endovascular treatment | 569 (8.1) | 307 (7.3) | 73 (5.8) | 55 (5.0) |
| Reperfusion therapy | 934 (13.3) | 580 (13.8) | 149 (11.8) | 105 (9.5) |
Data are presented as median (interquartile range) or n (%).
MRI, magnetic resonance imaging; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.
Prestroke disability was defined as a premorbid mRS score of ≥2.
3.1. Association between HbA1c and stroke subtype
The distribution of stroke subtypes based on the HbA1c level is shown in Fig. 1. In the unadjusted comparisons, large artery atherosclerosis became more frequent, while cardioembolism became less common, as HbA1c increased. In the adjusted analyses, the proportion of large artery atherosclerosis was higher across all HbA1c groups ≥6.0% than it was in HbA1c ≤6.0% (adjusted OR for HbA1c ≥8.0%, 1.40; 95% CI, 1.19–1.65), while the proportion of cardioembolism was lower across all groups ≥7.0% (adjusted OR for HbA1c ≥8.0%, 0.63; 0.50–0.79) (Fig. 2). In the adjusted restricted cubic spline analysis, higher HbA1c levels were linearly associated with a higher incidence of large artery atherosclerosis and a lower incidence of cardioembolism (Fig. 3, Fig. S2).
Fig. 1.
Stacked bar charts of stroke subtypes across HbA1c strata.
Values represent the unadjusted percentage of each stroke subtype within each HbA1c stratum; the total number of patients per stratum is shown on the left, from which absolute counts can be derived.
Fig. 2.
Associations between HbA1c categories and stroke subtypes.
Odds ratio of high glycated hemoglobin A1c (HbA1c) for the proportion of ischemic stroke subtypes, with HbA1c <6.0% as reference. Risk estimates were adjusted for age, sex, calendar year, hypertension, previous statin use, chronic kidney disease, and atrial fibrillation.
Fig. 3.
Restricted cubic spline analysis of the associations between HbA1c levels and stroke subtype (large artery atherosclerosis and cardioembolism).
Odds ratios (ORs) and 95% confidence intervals (CIs) for the proportion of (A) large artery atherosclerosis and (B) cardioembolism by glycated hemoglobin A1c (HbA1c, %) using restricted cubic splines. Risk estimates were adjusted for age, sex, calendar year, hypertension, previous statin use, chronic kidney disease, and atrial fibrillation. Solid lines represent the adjusted OR, and dashed lines represent the 95% CI. Knots were placed at 6.0, 7.0, and 8.0% of HbA1c. A HbA1c of 6.0% was used as a reference. Confidence intervals widen at the extremes of the HbA1c distribution, where fewer patients were observed; estimates in these regions should therefore be interpreted with caution.
3.2. Association between HbA1c and stroke lesions
Among the 13,569 patients included in the main analysis, 7252 (53.5%) had stroke lesion information and were included in the exploratory analysis of stroke lesions (Fig. S1). Patients with available lesion information generally had similar baseline characteristics compared to those with missing lesion data, with only a slightly lower proportion receiving reperfusion therapy (Table S2). Among them, 1042 patients (14.4%) had multifocal lesions involving more than one region. Lesions involved the cerebral cortex in 3621 patients (49.9%), basal ganglia/corona radiata in 2600 (35.9%), thalamus in 501 (6.9%), brainstem in 986 (13.6%), and cerebellum in 753 (10.4%) (not mutually exclusive). The distribution of stroke lesions according to HbA1c level is shown in Fig. S3. In the unadjusted comparisons, brainstem lesions became more frequent, whereas cerebral cortex and basal ganglia/corona radiata lesions became less common as HbA1c levels increased. In the adjusted analysis, even after adjusting for stroke subtypes, compared with the reference group (HbA1c <6.0%), the proportion of brainstem lesions was higher (adjusted OR for HbA1c ≥8.0%, 2.35; 95% CI, 1.82–3.02) and that of basal ganglia/corona radiata lesions was lower (0.67; 0.53–0.84) in all groups with HbA1c ≥7.0%. The proportion of cerebral cortex lesions was lower in the HbA1c ≥8.0% group (0.73; 0.59–0.90) (Fig. 4). Higher HbA1c levels in the adjusted restricted cubic spline analysis were linearly associated with a higher proportion of brainstem lesions and a lower proportion of lesions in the cerebral cortex and basal ganglia/corona radiata (Fig. 5, Fig. S4).
Fig. 4.
Associations between HbA1c categories and stroke lesions.
Odds ratio of glycated hemoglobin A1c (HbA1c) (6.0–7.0, 7.0–8.0, ≥8.0%) vs HbA1c < 6.0% for stroke lesions. Risk estimates were adjusted for stroke subtype, age, sex, calendar year, hypertension, previous statin use, chronic kidney disease, and atrial fibrillation. These lesion-topography analyses were exploratory and based on the 7252 patients (approximately half of the analyzed cohort) with documented lesion location.
Fig. 5.
Restricted cubic spline analysis of the associations between HbA1c levels and stroke lesions (cerebral cortex, basal ganglia/corona radiata, and brainstem).
Odds ratios (ORs) and 95% confidence intervals (CIs) for the proportion of (A) cerebral cortex, (B) basal ganglia/corona radiata, and (C) brainstem by glycated hemoglobin A1c (HbA1c, %) using restricted cubic splines. Risk estimates were adjusted for stroke subtype, age, sex, calendar year, hypertension, previous statin use, chronic kidney disease, and atrial fibrillation. Solid lines represent the adjusted OR, and dashed lines represent the 95% CI. Knots were placed at 6.0, 7.0, and 8.0% of HbA1c. A HbA1c of 6.0% was used as a reference. Confidence intervals widen at higher HbA1c levels, where fewer patients were observed; these trends should therefore be interpreted with caution.
3.3. Association between HbA1c and unfavorable functional outcome
Fig. S5 shows the unadjusted proportions of unfavorable functional outcomes in each ischemic stroke subtype. In general, and in all subtypes except small vessel occlusion, higher HbA1c levels showed a linear increase in rates of unfavorable functional outcomes. In the adjusted analysis, when compared with the reference group (HbA1c <6.0%), a higher proportion of patients had unfavorable functional outcomes in the following strata: the HbA1c ≥7.0% stratum for overall ischemic stroke (adjusted OR for HbA1c ≥8.0%, 1.56; 95% CI, 1.31–1.85); the HbA1c ≥8.0% stratum for large artery atherosclerosis and undetermined etiology stroke; and the HbA1c 6.0%–6.9% and ≥ 8.0% strata for cardioembolism (Fig. 6).
Fig. 6.
Associations between HbA1c categories and unfavorable functional outcomes in ischemic stroke subtypes.
Odds ratio of glycated hemoglobin A1c (HbA1c) (6.0–7.0, 7.0–8.0, ≥8.0%) vs HbA1c < 6.0% for unfavorable functional outcomes in ischemic stroke subtypes. Risk estimates were adjusted for age, sex, calendar year, hypertension, history of stroke, previous statin use, initial National Institutes of Health Stroke Scale score, and acute reperfusion therapy. Confidence intervals are wide in the smaller subgroups, particularly undetermined etiology stroke, reflecting limited precision in these subtypes.
4. Discussion
In this retrospective analysis of Japan's stroke registry, we demonstrated that higher HbA1c levels at admission were associated with different patterns of ischemic stroke subtypes, lesion locations, and functional outcomes. Our primary findings showed that elevated admission HbA1c levels were associated with a higher proportion of large artery atherosclerosis and a lower proportion of cardioembolism. A higher prevalence of brainstem lesions and a lower prevalence of lesions in the cerebral cortex and basal ganglia/corona radiata were also observed. These specific patterns of stroke etiology and topography are associated with a greater likelihood of unfavorable functional outcomes, especially in large arteries and cardioembolism. Given the multiple subtype, lesion, and subgroup analyses performed without correction for multiple comparisons, these findings should be regarded as exploratory and hypothesis-generating.
These findings align with previous research showing that diabetes and chronic hyperglycemia disproportionately affect large artery atherosclerosis but less so cardioembolic mechanisms [14], [15]. While prior studies have predominantly examined dichotomous diabetes status, our analysis extends this evidence by evaluating HbA1c as a continuous exposure and by incorporating radiologically confirmed lesion topography and functional outcomes within a large nationwide cohort.
Previous studies have linked higher HbA1c levels to large artery atherosclerosis and small vessel occlusion [7], [16], [17]. However, in this study, higher HbA1c levels were not associated with small vessel occlusion when the stroke subtype was modeled as the proportion of each subtype relative to the total. Since the overall proportion of large artery atherosclerosis increased with higher HbA1c levels, smaller changes in small vessel occlusion may have been diluted in this parts-of-a-whole outcome. Consistently, the unadjusted proportion of small vessel occlusion increased only slightly across the HbA1c strata (Fig. 1). Taken together, these findings suggest a stronger association between chronic hyperglycemia and large artery atherosclerosis than between chronic hyperglycemia and small vessel occlusion. This pattern may reflect population characteristics: intracranial atherosclerotic disease (ICAD), a major large artery atherosclerosis phenotype, is more prevalent in East Asia [18], [19], [20]. Given this epidemiology, the rise in large artery atherosclerosis with higher HbA1c in our cohort may be partly related to ICAD, which is more prevalent in this population; however, our data did not distinguish intracranial from extracranial disease, and this interpretation remains speculative.
A key contribution of this study is the observed association between HbA1c levels and ischemic stroke lesions. A few small studies have indicated a link between diabetes and brainstem infarction; nevertheless, their conclusions were limited by small sample sizes and unmeasured confounding factors, including stroke subtypes [21], [22], [23]. In this analysis, elevated HbA1c was independently associated with brainstem involvement, even after adjusting for stroke subtype. One speculative explanation may involve impaired sympathetic innervation and atherosclerosis of the basilar arteries related to chronic hyperglycemia; however, this mechanism is hypothesis-generating and could not be examined in our data, as posterior circulation atherosclerosis, intracranial atherosclerotic disease, and perforator burden were not directly modeled [24]. These associations may also reflect vascular distribution patterns rather than regional neurovascular susceptibility.
Regarding functional outcomes, this analysis showed that higher HbA1c levels were associated with unfavorable outcomes, especially among patients with large artery atherosclerosis, cardioembolism, and stroke of an undetermined etiology. This trend was not observed in patients with small vessel occlusion, indicating that the prognostic impact of glycemic control may differ based on the stroke subtype. The association between elevated HbA1c levels and poor functional outcomes has previously been reported in studies, including the UKPDS trial [5]. However, it remains unclear whether this relationship holds consistently across ischemic stroke subtypes. Some observational studies have reported associations primarily in small vessel occlusion, with no significant link in other subtypes [25], [26], while others have found no association even in small vessel occlusion [27]. This may indicate the limitations of our outcome definition, as marked functional deterioration captured by the mRS is less common in this subtype. The impact of chronic hyperglycemia on the mRS score at discharge might be too small to detect using our measures.
The strength of this study is in its use of a nationwide multicenter registry, which enabled a comprehensive evaluation of the association between HbA1c levels and stroke etiology, outcome, and lesions using a large sample size. The detailed neuroimaging data and experts' diagnoses of stroke subtypes further enhanced the reliability of our findings.
However, this study had some limitations. First, HbA1c levels were only measured upon admission. Second, this measurement may not fully reflect the chronic glycemic status in patients with anemia or renal failure [28]. Nonetheless, our results show that a single baseline HbA1c measurement can still provide valuable insights into the ischemic stroke profile. Third, the mRS score was recorded only at discharge because of the data acquisition system, which is less robust than 90-day outcomes and prevented us from assessing long-term functional recovery. In addition, our composite definition of unfavorable outcome combined a discharge mRS >2 in premorbidly independent patients with any worsening from baseline in previously disabled patients; although intended to capture clinically meaningful decline across the full range of premorbid function, these components are not strictly equivalent in severity. Fourth, HbA1c was available for only 27.8% of eligible patients, and its availability varied across calendar years (16.4% in 2016 to 25.9–31.7% in 2017–2020; p for trend = 0.025). Patients without HbA1c data more frequently had atrial fibrillation and received reperfusion therapy, raising the possibility of selection bias, although standardized mean differences between included and excluded patients were below 0.10 for nearly all variables (Table S1). Similarly, the lesion-topography analyses were restricted to the 7252 patients with documented lesion location and should be regarded as exploratory. Fifth, the impact of residual confounding cannot be ruled out completely. In particular, detailed information on glucose-lowering and stroke-related medications, as well as diabetes-related variables such as diabetes duration, insulin use, microvascular complications, obesity, smoking, and dyslipidemia severity, was unavailable, so HbA1c may partly act as a surrogate for broader cardiometabolic burden rather than isolated glycemic exposure. Although socioeconomic status was likewise unavailable, HbA1c itself may reflect socioeconomic conditions, so adjusting for it may be less critical when evaluating the association of elevated HbA1c with stroke profiles. Finally, the generalizability of our findings to other ethnic groups is limited since this study population solely involves Japanese patients.
In conclusion, higher HbA1c at admission was associated with a higher proportion of large artery atherosclerosis and a lower proportion of cardioembolism. This pattern was associated with a higher proportion of brainstem lesions and a lower proportion of cortical and basal ganglia/corona radiata lesions, and with poorer functional outcomes, especially in cases involving large arteries and cardioembolism. These findings highlight the subtype- and lesion-specific effects of chronic hyperglycemia on ischemic stroke presentation and prognosis.
CRediT authorship contribution statement
Shuhei Egashira: Writing – original draft, Visualization, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Kaori Miwa: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Conceptualization. Masatoshi Koga: Writing – review & editing, Investigation, Conceptualization. Sohei Yoshimura: Writing – review & editing, Project administration, Data curation. Akiko Ishigami: Writing – review & editing, Data curation. Yoshihiro Miyamoto: Writing – review & editing, Data curation. Tatsushi Toda: Writing – review & editing. Yuichi Imanaka: Writing – review & editing. Kazuo Minematsu: Writing – review & editing. Kazunori Toyoda: Writing – review & editing, Project administration, Methodology, Conceptualization.
Sources of funding
None.
Declaration of competing interest
Toyoda Kazunori reports the receipt of personal fees from Janssen, Otsuka Pharmaceutical Co Ltd., and Daiichi Sankyo. None of the other authors declare any potential conflicts of interest associated with this work; Masatoshi Koga reports honoraria from AstraZeneca, Bayer Yakuhin, Daiichi-Sankyo, Pfizer, BMS/Janssen Pharmaceuticals, and Otsuka Pharmaceutical, and research supports from Boston Scientific and Daiichi-Sankyo, all of which are outside of the submitted work.
Acknowledgements
None.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ensci.2026.100618.
Appendix. Supplementary data
Appendix A. Japan Stroke Data Bank investigators.
Appendix B. Supplemental materials.
References
- 1.Feigin V.L., Brainin M., Norrving B., Martins S., Sacco R.L., Hacke W., et al. World stroke organization (WSO): global stroke fact sheet 2022. Int. J. Stroke. 2022;17:18–29. doi: 10.1177/17474930211065917. [DOI] [PubMed] [Google Scholar]
- 2.Kaarisalo M.M., Räihä I., Sivenius J., Immonen-Räihä P., Lehtonen A., Sarti C., et al. Diabetes worsens the outcome of acute ischemic stroke. Diabetes Res. Clin. Pract. 2005;69:293–298. doi: 10.1016/j.diabres.2005.02.001. [DOI] [PubMed] [Google Scholar]
- 3.Sherwani S.I., Khan H.A., Ekhzaimy A., Masood A., Sakharkar M.K. Significance of HbA1c test in diagnosis and prognosis of diabetic patients. Biomark. Insights. 2016;11:95–104. doi: 10.4137/BMI.S38440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kamouchi M., Matsuki T., Hata J., Kuwashiro T., Ago T., Sambongi Y., et al. Prestroke glycemic control is associated with the functional outcome in acute ischemic stroke: the Fukuoka Stroke Registry. Stroke. 2011;42:2788–2794. doi: 10.1161/STROKEAHA.111.617415. [DOI] [PubMed] [Google Scholar]
- 5.Stevens R.J., Coleman R.L., Adler A.I., Stratton I.M., Matthews D.R., Holman R.R. Risk factors for myocardial infarction case fatality and stroke case fatality in type 2 diabetes: UKPDS 66. Diabetes Care. 2004;27:201–207. doi: 10.2337/diacare.27.1.201. [DOI] [PubMed] [Google Scholar]
- 6.ACCORD Study Group, Gerstein H.C., Miller M.E., Genuth S., Ismail-Beigi F., Buse J.B., et al. Long-term effects of intensive glucose lowering on cardiovascular outcomes. New Engl. J. Med. 2011;364:818–828. doi: 10.1056/NEJMoa1006524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Georgakis M.K., Harshfield E.L., Malik R., Franceschini N., Langenberg C., Wareham N.J., et al. Diabetes mellitus, glycemic traits, and cerebrovascular disease: a Mendelian randomization study. Neurology. 2021;96:e1732–e1742. doi: 10.1212/WNL.0000000000011555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Miwa K., Koga M., Nakai M., Yoshimura S., Sasahara Y., Koge J., et al. Etiology and outcome of ischemic stroke in patients with renal impairment including chronic kidney disease: Japan stroke data bank. Neurology. 2022;98:e1738–e1747. doi: 10.1212/WNL.0000000000200153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ikawa F., Ichihara N., Uno M., Shiokawa Y., Toyoda K., Minematsu K., et al. Visualisation of the non-linear correlation between age and poor outcome in patients with aneurysmal subarachnoid haemorrhage. J. Neurol. Neurosurg. Psychiatry. 2021;92:1173–1180. doi: 10.1136/jnnp-2020-325306. [DOI] [PubMed] [Google Scholar]
- 10.Toyoda K., Yoshimura S., Nakai M., Koga M., Sasahara Y., Sonoda K., et al. Twenty-year change in severity and outcome of ischemic and hemorrhagic strokes. JAMA Neurol. 2022;79:61–69. doi: 10.1001/jamaneurol.2021.4346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Adams H.P., Jr., Bendixen B.H., Kappelle L.J., Biller J., Love B.B., Gordon D.L., et al. 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. doi: 10.1161/01.str.24.1.35. [DOI] [PubMed] [Google Scholar]
- 12.Araki E., Goto A., Kondo T., Noda M., Noto H., Origasa H., et al. Japanese clinical practice guideline for diabetes 2019. J. Diabetes Investig. 2020;11:1020–1076. doi: 10.1111/jdi.13306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Miwa K., Nakai M., Yoshimura S., Sasahara Y., Wada S., Koge J., et al. Clinical impact of body mass index on outcomes of ischemic and hemorrhagic strokes. Int. J. Stroke. 2024;19:907–915. doi: 10.1177/17474930241249370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tuttolomondo A., Pinto A., Salemi G., Di Raimondo D., Di Sciacca R., Fernandez P., et al. Diabetic and non-diabetic subjects with ischemic stroke: differences, subtype distribution and outcome. Nutr. Metab. Cardiovasc. Dis. 2008;18:152–157. doi: 10.1016/j.numecd.2007.02.003. [DOI] [PubMed] [Google Scholar]
- 15.Larsson S.C., Scott R.A., Traylor M., Langenberg C.C., Hindy G., Melander O., et al. Type 2 diabetes, glucose, insulin, BMI, and ischemic stroke subtypes: Mendelian randomization study. Neurology. 2017;89:454–460. doi: 10.1212/WNL.0000000000004173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tsai C.F., Anderson N., Thomas B., Sudlow C.L.M. Risk factors for ischemic stroke and its subtypes in Chinese vs. Caucasians: systematic review and meta-analysis. Int. J. Stroke. 2015;10:485–493. doi: 10.1111/ijs.12508. [DOI] [PubMed] [Google Scholar]
- 17.Aquil N., Begum I., Ahmed A., Vohra E.A., Soomro B.A. Risk factors in various subtypes of ischemic stroke according to TOAST criteria. J. Coll. Physicians Surg. Pak. 2011;21:280–283. [PubMed] [Google Scholar]
- 18.Al Kasab S., Nguyen T.N., Derdeyn C.P., Yaghi S., Amin-Hanjani S., Kicielinski K., et al. Emergent large vessel occlusion due to intracranial stenosis: identification, management, challenges, and future directions. Stroke. 2024;55:355–365. doi: 10.1161/STROKEAHA.123.043635. [DOI] [PubMed] [Google Scholar]
- 19.Banerjee C., Chimowitz M.I. Stroke caused by atherosclerosis of the major intracranial arteries. Circ. Res. 2017;120:502–513. doi: 10.1161/CIRCRESAHA.116.308441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Carson A.P., Howard G., Burke G.L., Shea S., Levitan E.B., Muntner P. Ethnic differences in hypertension incidence among middle-aged and older adults. Hypertension. 2011;57:1101–1107. doi: 10.1161/HYPERTENSIONAHA.110.168005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Aronson S.M. Intracranial vascular lesions in patients with diabetes mellitus. J. Neuropathol. Exp. Neurol. 1973;32:183–196. doi: 10.1097/00005072-197304000-00001. [DOI] [PubMed] [Google Scholar]
- 22.Iwase M., Yamamoto M., Yoshinari M., Ibayashi S., Fujishima M. Stroke topography in diabetic and nondiabetic patients by magnetic resonance imaging. Diabetes Res. Clin. Pract. 1998;42:109–116. doi: 10.1016/s0168-8227(98)00103-x. [DOI] [PubMed] [Google Scholar]
- 23.Kameyama M., Fushimi H., Udaka F. Diabetes mellitus and cerebral vascular disease. Diabetes Res. Clin. Pract. 1994;24(suppl):S205–S208. doi: 10.1016/0168-8227(94)90250-x. [DOI] [PubMed] [Google Scholar]
- 24.Beausang-Linder M., Bill A. Cerebral circulation in acute arterial hypertension--protective effects of sympathetic nervous activity. Acta Physiol. Scand. 1981;111:193–199. doi: 10.1111/j.1748-1716.1981.tb06724.x. [DOI] [PubMed] [Google Scholar]
- 25.Gao Y., Jiang L., Wang H., Yu C., Wang W., Liu S., et al. Association between elevated hemoglobin A1c levels and the outcomes of patients with small-artery occlusion: a hospital-based study. PLoS One. 2016;11 doi: 10.1371/journal.pone.0160223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jeong J., Park J.K., Koh Y.H., Park J.M., Bae H.J., Yun S.M. Association of HbA1c with functional outcome by ischemic stroke subtypes and age. Front. Neurol. 2023;14 doi: 10.3389/fneur.2023.1247693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fang Y., Zhang S., Wu B., Liu M. Hyperglycaemia in acute lacunar stroke: a Chinese hospital-based study. Diab. Vasc. Dis. Res. 2013;10:216–221. doi: 10.1177/1479164112459663. [DOI] [PubMed] [Google Scholar]
- 28.Kuo I.C., Lin H.Y.H., Niu S.W., Lee J.J., Chiu Y.W., Hung C.C., et al. Anemia modifies the prognostic value of glycated hemoglobin in patients with diabetic chronic kidney disease. PLoS One. 2018;13 doi: 10.1371/journal.pone.0199378. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix A. Japan Stroke Data Bank investigators.
Appendix B. Supplemental materials.