To systematically elucidate the constitutional characteristics of AIS patients with hypertension and establish targeted BP management strategies, we developed a stepwise analytical framework grounded in evidence synthesis (Figure 1). First, we characterized TCM constitution distributions in foundational populations (AIS and hypertensive cohorts) to identify baseline traits linked to cerebrovascular pathology. Due to limited direct evidence in AIS with hypertensive population, we expanded analyses to broader ischemic stroke (IS) with hypertensive population. To address the risks of hypotension/hypoperfusion from aggressive BP lowering, we further examined TCM constitution patterns in hypotensive and ischemic cerebrovascular disease with hypoperfusion cohorts, identifying vulnerability traits. By synthesizing these findings and relevant research, we mapped the constitutional profiles in AIS with hypertensive patients and correlated them with clinical outcomes such as recurrence and hypoperfusion risks. This integrative approach enabled us to make targeted hypotheses: considering which TCM constitutions are, respectively, suitable for intensive BP control and conservative strategies.

Studies were eligible for inclusion if they met the following criteria: (1) studies using the standardized TCM Constitution Classification and Determination criteria published by the China Association of Chinese Medicine (CACM) (14); (2) cohort studies, case–control studies, or cross-sectional studies; (3) studies providing sufficient data for calculating TCM constitution distribution in patients with AIS, hypertension, hypotension, AIS with hypertension, IS with hypertension, or ischemic cerebrovascular disease with hypoperfusion. Exclusion criteria were: (1) studies involving populations with other comorbidities; (2) studies analyzing only a single constitution type; (3) duplicate publications or secondary analyses of the same population; (4) reviews, commentaries, case reports, or animal studies; (5) studies with incomplete data reporting or inaccessible full texts.

The outcome measure was the distribution of TCM constitution types in patients with AIS, hypertension, hypotension, AIS with hypertension, IS with hypertension, or ischemic cerebrovascular disease with hypoperfusion. To quantify this outcome, prevalence was used to describe the proportion of each constitution type within the study population, while odds ratios (ORs) were employed to compare the occurrence of specific constitution types between different patient groups. TCM constitution types were classified using the validated TCM Constitution Classification and Determination Criteria (ZYYXH/T157-2009) published by the CACM (14). (Supplementary material 1). This standardized tool defines nine constitution types: Balanced constitution (Type A) and eight unbalanced constitution types (Type B-I), including Qi-deficiency (Type B), Yang-deficiency (Type C), Yin-deficiency (Type D), Phlegm-dampness (Type E), Dampness-heat (Type F), Blood-stasis (Type G), Qi-stagnation (Type H), and Special Diathesis (Type I). Each participant completed the TCM Constitution Classification and Determination Questionnaire (Supplementary material 1), which consists of Likert-scale questions (5-point scoring) corresponding to the nine constitution domains. Raw scores were calculated by summing item responses for each constitution type. Conversion scores were derived using the formula:

A Balanced constitution was defined as having a conversion score of ≥60, with all eight unbalanced constitution types simultaneously scoring <30, while unbalanced constitutions were categorized into three levels: “yes” for conversion scores ≥40, “tendency” for scores ranging from 30 to 39, and “no” for scores <30. The detailed questionnaire items and criteria for constitution classification are presented in Supplementary material 1.

A comprehensive search was conducted across nine electronic databases, including China National Knowledge Infrastructure (CNKI), Wanfang Data, VIP Database, China Biology Medicine (CBM), PubMed, Web of Science, Scopus, the Cochrane Library, and ScienceDirect. The search encompassed articles published in both Chinese and English languages up to January 2025. The included research types comprised cross-sectional study, case–control study, and cohort study. The search strategy employed a combination of Medical Subject Headings (MeSH) terms and free-text keywords, including but not limited to “acute ischemic stroke,” “acute cerebral infarction,” “constitution,” “ischemic stroke,” “cerebral infarction,” “hypertension,” “low blood pressure,” “hypoperfusion,” “traditional Chinese medicine,” “cohort study,” “case–control study,” and “cross-sectional study,” along with their respective Chinese equivalents. Complete search syntax exemplifying the PubMed strategy is detailed in Supplementary Tables 1–6 (Supplementary material 2), with database-specific adaptations implemented accordingly. The reference management process was implemented using EndNote 20 software. Two investigators independently performed a two-stage screening procedure involving sequential evaluation of titles, abstracts, and full-text documents. All stages of the literature search and selection process were subjected to reciprocal verification to determine eligible records. Any discrepancies were adjudicated through consensus discussions with senior reviewers.

The following information was systematically extracted from eligible studies: first author, publication year, study population, sample size, research design type, country of origin, and patient distribution across TCM constitution types. Two investigators independently performed data extraction using a pretested form and cross-checked the results for consistency, with discrepancies resolved through discussions.

Quality assessment was conducted using standardized tools appropriate to each study design: cross-sectional studies were evaluated with the Agency for Healthcare Research and Quality (AHRQ) checklist (15), while cohort and case–control studies were appraised using the Newcastle-Ottawa Scale (NOS) (16). For the AHRQ checklist, each of the 11 items (e.g., clarity of data sources, inclusion/exclusion criteria, handling of confounding factors) was evaluated as “Yes,” “No,” “Unclear,” or “Not Applicable.” Scoring criteria were defined as follows: “Yes” responses were assigned 1 point, while “No,” “Unclear,” and “Not Applicable” received 0 point. For cross-sectional studies, Item 11 about follow-up was deemed “Not Applicable” due to the inherent design of such studies. Consequently, this item was excluded from the total score calculation, resulting in a maximum achievable score of 10. Study quality was classified based on established thresholds in the literature (17) studies fulfilling ≥7 scores were categorized as high quality, those scoring 4–6 as moderate quality, and those with ≤3 points as low quality.

The NOS evaluates studies across eight domains (e.g., cohort selection, comparability, outcome assessment), with a maximum score of 9. Quality classification followed established thresholds: studies scoring ≥7 points were classified as high quality, those with 5–6 points as moderate quality, and those with ≤4 points as low quality (18). Two researchers independently conducted quality assessments, followed by cross-verification of all entries. Any discrepancies were resolved through consensus discussions or, when necessary, adjudication by a third reviewer.

All statistical analyses were performed using Review Manager 5.3 and Stata software (version 14.0). Two types of analyses were conducted: (1) pooled prevalence estimates for single-group proportions, and (2) comparative analyses between groups using ORs. For synthesizing single-group prevalence estimates (e.g., TCM constitution distributions), pooled proportions were calculated using the logit transformation to stabilize variances of proportional data. Results were back-transformed to the original proportion scale via the inverse logistic transformation and reported as proportions (%) with 95% confidence intervals (CIs). Heterogeneity across studies was assessed using Cochran’s Q test (p-value) and the I² statistic. A random-effects model was applied if significant heterogeneity was detected (p < 0.1 or I² ≥ 50%); otherwise, a fixed-effects model was used. Comparative effect analyses (e.g., AIS with hypertension population vs. AIS without hypertension population) employed the inverse variance method for effect size calculation. Heterogeneity assessment and model selection criteria followed the same approach as described for prevalence analyses. All comparative results are expressed as ORs with 95% CIs.

Sensitivity analyses using leave-one-out cross-validation and subgroup analyses stratified by clinically relevant covariates were conducted to investigate potential sources of heterogeneity. When more than 10 studies were included in the analysis, funnel plots, Egger’s test, and Begg’s test would be employed to assess publication bias (19). If publication bias was detected, the trim-and-fill method would be applied to examine its impact (20).

Fifty-four studies were ultimately included for analysis. The literature selection flowcharts are provided in Supplementary Figures 1–6 (Supplementary material 2). The characteristics of the studies are summarized in Table 1. The studies were conducted in China, with publication years spanning from 2009 to 2024, and sample sizes ranging from 9 to 26,579 participants. Research design types included cross-sectional (n = 40), cohort (n = 4), and case–control (n = 10). 28 studies have analyzed the distribution characteristics of TCM constitution in the AIS population. 13 studies focused on individuals with hypertension, three on those with hypotension, two on AIS patients with hypertension, 10 on IS with hypertension, and two on ischemic cerebrovascular disease with hypoperfusion. The distribution characteristics of the nine TCM constitutions for each population were also recorded in detail, respectively.

We assessed the quality of 40 cross-sectional studies using the AHRQ measurement tool (Figure 2). The results showed that 11 studies had a total score of 7 to 10 points and were considered to be of high quality. 28 studies had a total score of 4 to 6 points and were regarded as of moderate quality, and one study had a total score of 3 points and was considered to be of low quality. The majority of score deductions stemmed from insufficient reporting of consecutive enrollment status in non-population designs, unverified blinding procedures for subjective assessments, undocumented quality control measures for primary outcomes, or incomplete documentation of participant response rates and data completeness. The detailed scores for items of each study are presented in Supplementary Table 7 (Supplementary material 2).

The NOS scale was used to assess the quality of four cohort studies and 10 case–control studies (Figure 3). 10 studies scored 7–9 points (classified as high-quality), while 4 studies scored 5–6 points (categorized as moderate-quality). The primary reasons for score deductions across most studies included inadequate follow-up duration, failure to describe follow-up completeness, and insufficient documentation of non-exposed group recruitment sources. The detailed scores for 8 items of each study are presented in Supplementary Table 8 (Supplementary material 2).

Table 2 presents the meta-analysis results of TCM constitution distribution, whereas Table 3 compares TCM constitution types between the studied populations and other populations. The forest plots for the AIS population are presented in Figures 4, 5, with corresponding subgroup analysis forest plots provided in Figure 6. The forest plots for other meta-analyses and their subgroup analyses are presented in Supplementary Figures 7–19 (Supplementary material 2).

Sensitivity analyses were performed by sequentially excluding each included study. In the meta-analysis of the Special Diathesis constitution among IS with hypertension population, the exclusion of the study by Wu et al. (72) substantially reduced heterogeneity (I² decreased from 53% [p = 0.08] to 0% [p = 0.78]). However, the pooled estimate was not statistically distinct (pre-exclusion: 2% [0.01, 0.04] vs. post-exclusion: 1% [0.01, 0.02]), confirming the low distribution of Special Diathesis constitution in this population. For the remaining five population subgroups, sequential exclusion of individual studies demonstrated no material alteration in heterogeneity indices or effect sizes, indicating robust primary findings.

We assessed publication bias for analyses incorporating more than 10 studies, including meta-analyses of TCM constitution distributions in the AIS population, hypertensive population, and comparative analyses of hypertensive versus non-hypertensive population. Trim-and-fill analysis was performed where publication bias was detected. The overall assessment indicated that while some analyses exhibited publication bias, the original conclusions remained robust after trim-and-fill correction. Funnel plots are presented in Figures 7–9, with Egger’s test, Begg’s test, and trim-and-fill results detailed in Table 4.

For the AIS population, funnel plots for Yin-deficiency, Phlegm-dampness, and Qi-deficiency constitutions were symmetrical (Figure 7), with both Egger’s and Begg’s tests yielding p > 0.05, indicating no significant publication bias. However, funnel plots for Blood-stasis, Balanced, Damp-heat, Qi-stagnation, Yang-deficiency, and Special Diathesis constitutions demonstrated slight asymmetry, with either Egger’s or Begg’s test showing p ≤ 0.05, suggesting potential publication bias. Trim-and-fill analysis revealed no studies requiring addition or removal, confirming the robustness of original findings and indicating that publication bias did not substantively impact conclusions.

In the meta-analysis of TCM constitution distribution among hypertensive populations, funnel plots for all nine constitutions appeared generally symmetrical (Figure 8). With the exception of the Phlegm-dampness constitution which demonstrated significant asymmetry on Egger’s test (p = 0.005), both Egger’s and Begg’s tests for the remaining eight constitutions yielded non-significant results (p > 0.05), indicating no substantial evidence of publication bias. Trim-and-fill analysis for Phlegm-dampness constitution identified two missing studies. After imputation, the pooled prevalence estimate was 28% [0.14, 0.58]—closely aligned with the original effect size in Table 2—thereby reinforcing that Phlegm-dampness is among the most prevalent TCM constitutions in hypertension.

Regarding comparative analyses between hypertensive and non-hypertensive populations, funnel plots demonstrated symmetry for Phlegm-dampness, Blood-stasis, Damp-heat, Qi-stagnation, Yang-deficiency, and Special Diathesis constitutions (Figure 9), with non-significant Egger’s and Begg’s tests (p > 0.05), indicating no publication bias. Although Yin-deficiency (Egger’s test: p = 0.043; Begg’s test: p = 0.272) and Qi-deficiency (Egger’s test: p = 0.030; Begg’s test: p = 0.217) constitutions showed minor asymmetry, trim-and-fill analysis required no study imputation, confirming minimal bias impact. Balanced constitution exhibited significant funnel plot asymmetry (Egger’s test: p = 0.012; Begg’s test: p < 0.0001). Trim-and-fill imputation added four studies, yielding a corrected effect size of 0.28 [0.15, 0.54] (p < 0.0001). This result remained consistent with the original conclusion that Balanced constitution prevalence is significantly lower in the hypertensive population, affirming its role as a protective factor against hypertension.

Approximately 75% of AIS patients present with elevated BP, mediated by neurological injury, autonomic dysfunction, and pre-existing hypertension (3, 4). In addition, acute disruption of cerebral autoregulation after AIS impairs cerebrovascular adaptability to BP fluctuations, exacerbating ischemic tissue vulnerability to both hypoperfusion and hyperperfusion (77). This creates a therapeutic paradox: aggressive early BP reduction may jeopardize penumbral perfusion (78, 79), while sustained hypertension elevates hemorrhagic transformation and reperfusion injury risks, particularly in thrombolysis recipients (80, 81). Large-scale trials (e.g., ENCHANTED, and BEST-II) comparing intensive versus standard BP strategies, have yielded inconclusive results (9–12) due to marked patient heterogeneity. Beyond CA impairment, BP regulation in AIS involves complex interactions among cerebral vasculature pathology, systemic hemodynamics, and endocrine responses, rendering conventional demographic-based stratification inadequate for capturing physiological and pathological variability. This underscores the necessity to transcend numerical BP targets by integrating individualized physiological profiling. Emerging frameworks, including TCM constitutional theory, may enhance precision in predicting antihypertensive responses and tailoring therapeutic strategies.

We therefore systematically investigated TCM constitutional distributions across relevant populations of AIS with hypertension to determine whether these patterns correlate with disease characteristics and therapeutic heterogeneity. Hypertensive and AIS populations share common constitutional patterns, with Phlegm-dampness, Yin-deficiency, Qi-deficiency, and Blood-stasis constitutions being the most prominent. In AIS/IS with hypertensive individuals, the overlap of constitutional profiles, dominated by Phlegm-dampness, Yin-deficiency, Qi-deficiency, and Blood-stasis constitutions, further underscores the shared pathophysiological mechanisms linking hypertension and AIS. These consistent patterns emphasize the fundamental role of the four TCM constitutions in AIS with hypertension, further large-scale studies are warranted to verify these findings. In contrast, the hypotensive population and those with ischemic cerebrovascular hypoperfusion display a unique constitutional distribution. While the Phlegm-dampness, Qi-deficiency, and Yin-deficiency constitutions remain prominent, the Yin-deficiency and Qi-deficiency constitutions are significantly more prevalent compared to normotensive individuals. This distinct constitutional divergence from hypertension and AIS profiles suggests a pathophysiological propensity toward hemodynamic instability, particularly manifesting as hypotension or ischemic cerebrovascular hypoperfusion in individuals with Yin-deficiency and Qi-deficiency constitutions.

The majority of included studies demonstrated moderate-to-high methodological quality as evaluated by standardized tools (AHRQ for cross-sectional studies and NOS for cohort/case–control studies). Subgroup analyses demonstrated that study design constituted a key source of heterogeneity; nevertheless, most subgroups consistently revealed that Phlegm-dampness, Blood-stasis, Qi-deficiency, and Yin-deficiency constitutions persisted as the predominant types across design categories. Sensitivity analyses employing leave-one-out cross-validation confirmed the robustness of primary findings. Publication bias assessments via funnel plots, Egger’s test, and Begg’s test indicated generally symmetrical distributions, with minimal evidence of bias for most constitutions. Where publication bias was detected (e.g., for Phlegm-dampness constitution in hypertensive population), trim-and-fill imputation demonstrated a negligible impact on the original effect sizes, reinforcing the robustness of the primary conclusions. Collectively, these analytical approaches substantiated the reliability and consistency of the identified TCM constitutional patterns underlying the proposed framework.

Extensive research has confirmed that distinct TCM constitutional types are associated with unique pathophysiological characteristics and prognostic implications, offering a framework for individualized BP management in AIS with hypertension. Patients with the Phlegm-dampness constitution often present with metabolic abnormalities such as hyperlipidemia and obesity, predisposing them to atherosclerosis (82–84) and elevating risks of BP irregularities and stroke. Mechanistic studies indicated intestinal microbiota imbalance (characterized by elevated Firmicutes/Bacteroidetes ratio), chronic inflammation, and metabolic dysregulation in these patients (85–88). These processes promoted vascular endothelial dysfunction (89), contributing to vascular-related diseases like hypertension and AIS. Regarding Blood-stasis constitution, several clinical studies revealed altered hemorheological parameters and heightened thrombogenic potential (37, 90), resulting in microcirculatory dysfunction and elevated vascular stenosis risk. Modern pathological studies further demonstrated that these abnormalities directly promote atherosclerosis, thrombosis, and endothelial dysfunction (91, 92). From a TCM perspective, both constitutions involve impeded circulation of Qi, blood, and fluids, leading to pathogenic accumulation within vessels. Consequently, elevated BP in these individuals may impose excessive mechanical stress on already compromised vasculature, significantly increasing the risk of acute vascular events, including plaque rupture, thrombosis, and recurrent stroke (93, 94). Intensive BP control may be warranted for patients with these constitutions to optimize cerebral hemodynamics and mitigate their inherently elevated stroke risk associated with vascular vulnerability.

Conversely, Qi-deficiency and Yin-deficiency constitutions pose unique challenges for BP regulation due to compromised autonomic function and diminished compensatory capacity. This renders patients highly susceptible to BP fluctuations and hypoperfusion events (38, 74, 76, 95). TCM pathology attributes Qi deficiency to the depletion of “vital energy” (Qi), manifesting as impaired functional activity that undermines blood circulation regulation. Yin deficiency, characterized by insufficient fluid nourishment, compromises vascular tone and perfusion pressure. Modern studies linked Qi-deficiency to cellular energy metabolism disorders (notably mitochondrial dysfunction) (96, 97), which impair adenosine triphosphate (ATP)-dependent vascular autoregulation and repair mechanisms. Yin deficiency was associated with chronic inflammation and oxidative stress (98, 99), exacerbating endothelial dysfunction and reducing vascular elasticity and reserve (100). These impairments weaken compensatory responses to BP fluctuations, increasing vulnerability to hypoperfusion, especially in critical organs like the brain. Consequently, a conservative BP management strategy may be more appropriate for them to avoid iatrogenic hypoperfusion while managing risks from sustained hypertension.

Taken together, these findings support a hypothesis for TCM constitution-targeted BP management in AIS patients. Patients with Phlegm-dampness or Blood-stasis constitutions, characterized by vascular stenosis, thrombosis propensity, and metabolic dysregulation, likely benefit from intensive BP control to reduce stroke recurrence risk (101). In contrast, those with Qi-deficiency or Yin-deficiency constitutions exhibit hemodynamic fragility and diminished vascular reserve, suggesting that more conservative BP management strategies could potentially be appropriate to help reduce the risk of hypoperfusion complications.