Serum vitamin D and its association with acute ischemic stroke severity: A systematic review

INTRODUCTION

Stroke represents a major cause of global mortality and disability, and approximately 62% of cases being Acute Ischemic Stroke (AIS).^[1] The clinical outcomes of stroke patients are determined by various prognostic factors, including the severity of neurological deficits at initial presentation, evaluated using the National Institute of Health Stroke Score (NIHSS).^[2]

Vitamin D is acquired from food and synthesized in the skin, then converted in the liver into 25−hydroxyvitamin D [25(OH)D].^[3] In addition to its essential role in calcium metabolism and bone health, vitamin D affects the central nervous system, inflammation, and immune response.^[3]Vitamin D deficiency is quite prevalent worldwide and has been linked to an increased risk of cardiovascular disease and stroke.^[3] At present, approximately 71% of stroke patients are estimated to be vitamin D deficient, which has been associated with clinical outcomes.^[4,5]

Recent studies have increasingly addressed the potential role of vitamin D in AIS, given that low vitamin D levels have been associated with a higher likelihood of stroke and more severe presentations.^[6] Vitamin D may have neuroprotective effects by reducing neuronal damage, controlling blood pressure, lowering inflammation, and improving endothelial function.^[7] Furthermore, vitamin D deficiency in AIS patients may lead to worse clinical outcomes, higher mortality rates, and prolonged hospitalization.^[8]

However, findings from existing studies remain inconsistent. Some studies have found a significant association between reduced vitamin D concentrations and more severe stroke, whereas others have reported no meaningful relationships.^[9,10]Therefore, a systematic review is needed to clarify the relationship between vitamin D levels and AIS severity, and to evaluate whether vitamin D can serve as a predictive or prognostic factor. The findings of this review are expected to inform clinical recommendations on the value of vitamin D screening in patients with AIS.

MATERIALS AND METHODS

This systematic review and meta-analysis were performed in line with the PRISMA 2020 reporting guidelines [Supplementary Table 1], and the protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42025633269).

RESULTS

Study selection

A total of 156 records were identified through three databases, searching and citation tracking. After removal of duplicates, 117 articles remained for titles and abstracts screening. Of these, 93 records were excluded, and 24 full-text articles were assessed for eligibility against predefined criteria. Following full-text review, six studies were excluded, leaving 18 studies eligible for inclusion in the systematic review, of which 14 studies provided sufficient data and were included in the meta-analysis [Figure 1].^[5,7,9-24]

Close

Figure 1: Flowchart for selection of studies for inclusion.

Export to PPT

Association between serum vitamin D levels and NIHSS at admission

Among the 18 included studies, 14 reported an inverse association between serum vitamin D levels and stroke severity measured by the NIHSS at hospital admission [Tables 2 and 3]. Fourteen comparisons involving 2,242 participants contributed to the quantitative synthesis.

Table 2: Data extraction of level Vitamin D and NIHSS at admission

S. No.	References	Level of serum Vitamin D at admission (ng/mL); (Mean/median)	NIHSS at admission (Mean/median)	Vitamin D assay method	Admission timing
1.	Kim et al., 2023[7]	17.1 (12.2−24.2)	NR	CLIA	7 days
2.	Li et al., 20187[9]	19.1 (11.9−25.6)	NR	ELISA	0–24 h
3.	Bayat et al., 2023[10]	AIS=24.3±1.4, control=29.9±1.5	NR	CLIA	0–24 h
4.	Wang et al., 2014[11]	14.3 (10.1–21.5)	8 (5−12)	CLIA	0–24 h
5.	Kamal et al., 2023[12]	AIS = 19.79±10.79; control=28.57±42.84	20.73±9.444	CLIA	Not specified
6.	Miao et al., 2021[13]	25 (9.2)	4.14 (4.18)	NR	NR
7.	Qiu et al., 2017[14]	14.2 (10.3−18.9)	8.0 (4−12)	NR	Within 24 h
8.	Masoud et al., 2023[15]	NR	12.0±4.7	ELISA	Not specified
9.	Zhang et al., 2018[16]	NR	<20=3 (2−7); ≥20=3 (1−5)	CBPA	Within 24 h
10.	Flifel et al., 2020[24]	>30=37.95±7.86; <30=16.77±7.64	>30=3.93±2.13; <30=11.51±10.45	NR	24–48 h
11.	Tu et al., 2014[17]	14.2 (10.2−18.9)	8 (4−12)	CLIA	0–24 h
12.	Safari et al., 2022[18]	AIS=22.40 (18.83−25.86); control=32.07 (20.13−39.12)	12.97 (10.65−15.35)	ELISA	>24 h
13.	Markišic et al., 2017[19]	30.2±9.0	14±7	NR	Not specified
14.	Alharbi et al., 2022[20]	NR	7 (4,12)	CLIA	Within 7 days
15.	Fahmy et al., 2019[5]	AIS=21.260±12.5, control=28.72±11.284	11.375±4.836	ELISA	Not specified
16.	Aggarwal et al. 2022[21]	19.5±9.4	19.0 (11.5, 23.0); 15.0 (9.0, 20.0); 7.5 (5.0, 15.5); 6.0 (4.0, 14.0)	CLIA	Within 7 days
17.	Wei et al., 2018[22]	18 (13–24)	7 (3–14)	CLIA	0–48 h
18.	Simon et al., 2024[23]	NR	NR	CLIA	<24 h

AIS: Acute ischemic stroke, NIHSS: National institutes of health stroke scale, mRS: Modified rankin scale, MRI: Magnetic resonance imaging, DWI: Diffusion weighted imaging, CLIA: Chemiluminescent immunoassay, ELISA: Enzyme-linked immunosorbent assay, CBPA: Competitive binding protein assay, OR: Odds ratio, CI: Confidence interval, NR: Not reported, AIS: Acute ischemic stroke

The random-effects meta-analysis showed a moderate inverse correlation between serum vitamin D levels and NIHSS scores (r=−0.39, 95% CI −0.45 to −0.32; I² = 65%; p < 0.0001), indicating that lower vitamin D concentrations were associated with greater neurological severity at presentation [Figure 2]. Given the presence of substantial between-study heterogeneity, pre-specified subgroup analyses were undertaken according to stroke onset, geographic regions, and vitamin D assay methods.

Close

Figure 2: Forest plots of correlations between serum vitamin D levels and National Institutes of Health Stroke Scale at admission. (a) Subgrouped by stroke onset timing (early <24 h, delayed >24 h or mixed, and unspecified). (b) Subgrouped by geographic region (East Asia vs. other regions). (c) Subgrouped by vitamin D assay method (Chemiluminescent immunoassay-based, Enzyme-linked immunosorbent assay-based, and unspecified), CI: Confidence interval, COR: Correlation coefficient.

Export to PPT

Subgroup analysis by stroke onset

Subgroup analyses by stroke onset, timing of NIHSS assessment relative to stroke onset, were categorized into delayed (>24 h), early (<24 h), and unspecified [Figure 2a]. The inverse association between serum vitamin D levels and initial NIHSS was observed in all subgroups: delayed onset subgroup (r=−0.38, 95% CI −0.44 to −0.31; I² = 18%), early-onset subgroup (r=−0.44, 95% CI −0.54 to −0.34; I² = 78%), and unspecified subgroup (r=−0.30, 95% CI −0.39 to −0.20; I2 = 1%).

Heterogeneity was substantially reduced in the delayed onset and unspecified subgroups, but remained high in the early onset subgroup. Visual inspection of the forest plot suggested that the study of Li et al.,^[9] contributed disproportionately to heterogeneity in this subgroup. After excluding this study in a sensitivity analysis [Supplementary Figure 1], the early onset subgroup remained statistically significant (r=−0.39, 95% CI −0.45 to −0.33) with no residual heterogeneity (I² = 0%).

The test for subgroup differences was not statistically significant (p = 0.12), indicating that stroke onset, or timing of the NIHSS assessment after the onset of stroke, did not reliably modify the strength of the association, although it appeared to contribute to heterogeneity.

Subgroup analysis by stroke geographic region

Subgroup analyses by geographic region were classified into East Asia vs. other regions, reflecting the predominance of studies conducted in East Asian countries. The stratification showed that the inverse correlation between vitamin D levels and NIHSS was consistent across regions [Figure 2b], both East Asia (r= −0.41, 95% CI −0.53 to −0.30; I² = 84%), and other regions (r=−0.37, 95% CI −0.43 to −0.31; I² = 12%). Although heterogeneity was high in East Asian studies, the test for subgroup differences was not statistically significant (p =0.52), indicating no evidence that geographic region modified the association between vitamin D levels and initial stroke severity. Removal of Li et al.,^[9] reduced heterogeneity without changing the pooled estimate [Supplementary Figure 1].

Subgroup analysis by vitamin D assay method

When stratified by vitamin D assay method, all subgroups demonstrated an inverse correlation [Figure 2c]: Chemiluminescent immunoassay (CLIA)-based assays (r = −0.38, 95% CI −0.42 to −0.33; I² = 10%), Enzyme-Linked Immunosorbent Assay (ELISA)-based assays (r = −0.37, 95% CI −0.58 to −0.16; I² = 85%), and unspecified assays (r = −0.42, 95% CI −0.57 to −0.27; I² = 0%). The test for subgroup differences was not statistically significant (p = 0.86), suggesting that the assay method did not influence the pooled association. As in other subgroup analyses, exclusion of Li et al.,^[9] substantially reduced heterogeneity in the ELISA-based assays subgroup without changing the direction or magnitude of the effect [Supplementary Figure 1].

Overall, this meta-analysis demonstrates a consistent, moderate inverse correlation between serum vitamin D levels and stroke severity at presentation. The association was robust across subgroups defined by stroke onset, geographic region, and assay method. However, the presence of residual heterogeneity, largely driven by a single influential study, indicated that the results should be interpreted with caution.

Funnel plot inspection did not reveal marked asymmetry, suggesting the absence of potential publication bias [Supplementary Figure 2]. Overall, these findings suggest that higher serum vitamin D concentrations are associated with less severe neurological deficit at admission, but between-study heterogeneity limits the certainty of this conclusion.

Association between serum vitamin D levels and functional outcomes

Six studies, including 538 participants, assessed the association between serum vitamin D levels and functional outcomes measured by mRS. All six studies reported an inverse association [Table 3 and Figure 3]. The random effects meta-analysis demonstrated a moderate inverse correlation between vitamin D levels and mRS scores (r= −0.43, 95% CI −0.60 to −0.27; I² = 83%; p < 0.0001), indicating that lower vitamin D levels were associated with worse functional outcomes [Figure 3]. Between-study heterogeneity was substantial; therefore, pre-specified subgroup analyses were conducted.

Table 3: Table of correlation between Vitamin D and NIHSS, mRS, and infarct volume

S. No.	References	Correlation between vitamin D and NIHSS	Logistic/linear regression between vitamin D and NIHSS	Correlation between Vitamin D and mRS	Correlation between Vitamin D and infarct volume
1.	Kim et al., 2023[7]	−0.402	OR: 6.02 (95% CI, 2.34–15.50) p<0.001	Discharge=−0.274	NR
2.	Li et al., 2018[9]	−0.601	OR: 0.886 (95% CI, 0.805–0.932)	NR	MRI with DWI r=−0.417; p<0.001
3.	Bayat et al., 2023[10]	−0.41	β=−0.381 (95% CI, −1.322–−0.415), p=0.000	NR	NR
4.	Wang et al., 2014[11]	−0.389	NR	NR	MRI with DWI r=−0.355, p<0.0001
5.	Kamal et al., 2023[12]	−0.2459	NR	Discharge=−0.6427	NR
6.	Miao et al., 2021[13]	NR	β=−0.092 (95% CI, −0.029–−0.004), p=0.011	NR	NR
7.	Qiu et al., 2017[14]	NR	OR: 1.62 (95% CI, 1.20−1.89) p< 0.0001	NR	NR
8.	Masoud et al., 2023[15]	−0.22	NR	At 90 days=−0.48
9.	Zhang et al., 2018[16]	NR	≥75=1 50−75=1.11 (0.63−1.96) <50=1.21 (0.93−1.58) ≥50=1 <50=1.38 (0.91−2.07) p=0.070	NR	NR
10.	Flifel et al., 2020[24]	−0.376	NR	At admission=−0.232	NR
11.	Tu et al., 2014[17]	−0.363	NR	NR	NR
12.	Safari et al., 2022[18]	−0.306	NR	NR	MRI with DWI r=− 0.147, p=0.030
13.	Markišic et al., 2017[19]	−0.465	NR	After 3 month=0.049 After 6 month=−0.231	ASPECT score r=0.221, p=0.149
14.	Alharbi et al., 2022[20]	NR	β=−0.04, SE: 0.01, p=0.003	NR	NR
15.	Fahmy et al., 2019[5]	−0.286	Insufficiencyβ=1.059, 95% CI=−1.035–8.036, SE: 0.532, p=0.043, OR:2.884 Deficiency β=2.623, 95% CI=−2.988– 63.553,SE: 0.780, p=0.001, OR: 13.780	After 3 month=−0.641	NR
16.	Aggarwal et al. 2022[21]	−0.457	NR	NR	NR
17.	Wei et al., 2018[22]	−0.305	NR	NR	MRI r=–0.179, p=0.012
18.	Simon et al., 2024[23]	−0.4081	β=-0.3994, p<0.001	NR	NR

AIS: Acute ischemic stroke, NIHSS: National institutes of health stroke scale, mRS: Modified rankin scale, MRI: Magnetic resonance imaging, DWI: Diffusion weighted imaging, ASPECT: Alberta stroke program early computed tomography score, OR: Odds ratio, CI: Confidence interval, NR: Not reported, SE: Standard error

Close

Figure 3: Forest plots of correlations between serum vitamin D levels and functional outcome measured by the modified Rankin Scale. (a) Subgrouped by stroke onset timing (delayed >24 h or mixed; unspecified). (b) Subgrouped by geographic region (East Asia vs. other regions). (c) Subgrouped by vitamin D assay method (Chemiluminescent immunoassay-based, Enzyme-linked immunosorbent assay-based, and unspecified), CI: Confidence interval, COR: Correlation coefficient, CLIA: Chemiluminescent immunoassay, ELISA: Enzyme-linked immunosorbent assay.

Export to PPT

Subgroup analysis by stroke onset

When stratified by stroke onset, the inverse association was observed in both available subgroups [Figure 3a]: delayed onset subgroup (r= −0.26, 95% CI −0.38 to −0.15; I² = 0%) and unspecified subgroup (r= −0.53, 95% CI −0.68 to −0.38; I2 = 68%). The test for subgroup differences was statistically significant (p < 0.01), suggesting that stroke onset may modify the strength of the association. However, this result should be interpreted with caution because no study contributed to the early onset subgroup, and substantial heterogeneity remained in the unspecified subgroup.

Subgroup analysis by stroke geographic region

Only one study was conducted in East Asia, with the remaining studies classified as other regions [Figure 3b]. There was no statistically significant evidence of differences between subgroups (p = 0.06), and this stratification did not explain the observed heterogeneity. Therefore, there was insufficient evidence to determine whether the association differed by geographic region.

Subgroup analysis by vitamin D assay method

When stratified by vitamin D assay method, inverse correlations were observed in all subgroups [Figure 3c]: CLIA-based assays subgroup (r= −0.46, 95% CI −0.82 to −0.09; I² = 93%), ELISA-based assays subgroup (r= −0.58, 95% CI −0.73 to −0.42; I² = 49%), and unspecified subgroup (r= −0.24, 95% CI −0.41 to −0.06; I² = 0%) [Figure 3c]. The test for subgroup differences was statistically significant (p = 0.02), indicating that assay methods may influence the estimated strength of the association. Heterogeneity remained very high in the CLIA subgroup, reflecting marked variability between the two contributing studies.

Overall, the pooled evidence suggests an association between lower vitamin D concentrations and worse post-stroke functional status, as measured by mRS. However, the substantial heterogeneity, limited number of studies, and instability of subgroup findings indicate that the results warrant cautious interpretation.

Association between serum vitamin D levels and infarct volume

Four studies, including 710 participants, examined the association between serum vitamin D levels and infarct volume measured by diffusion-weighted MRI (DWI). All four studies reported an inverse association. The random effects meta-analysis showed a statistically significant inverse correlation between serum vitamin D levels and infarct volume (r=−0.29, CI−0.41 to −0.16; I² = 73%; p < 0.001), as shown in Figure 4. It indicates that lower vitamin D concentrations were associated with larger infarct volumes. Between-study heterogeneity was substantial, and subgroup analyses according to the pre-specified criteria were subsequently conducted.

Close

Figure 4: Forest plots of correlations between serum vitamin D levels and infarct volume measured by diffusion-weighted magnetic resonance imaging. (a) Subgrouped by stroke onset timing (early <24 h vs. delayed >24 h or mixed). (b) Subgrouped by geographic region (East Asia vs. other regions). (c) Subgrouped by vitamin D assay method (Enzyme-linked immunosorbent assay-based vs. Chemiluminescent immunoassay-based), CI: Confidence interval, COR: Correlation coefficient, CLIA: Chemiluminescent immunoassay, ELISA: Enzyme-linked immunosorbent assay.

Export to PPT

Subgroup analysis by stroke onset

When stratified according to stroke onset, three studies contributed to the early onset subgroup (r = −0.32, 95% CI −0.46 to −0.17; I² = 74%), while only one study contributed to the delayed onset subgroup (r = −0.18, 95% CI −0.34 to −0.02) [Figure 4a]. The test for subgroup differences was not statistically significant (p = 0.21), indicating no evidence that stroke onset timing modified the association. However, this finding should be interpreted with caution because the delayed-onset subgroup contained only a single study.

Subgroup analysis by geographic area

Three studies were conducted in East Asia, yielding a pooled correlation of r = −0.33 (95% CI −0.46 to−0.20; I² = 67%), while one study was conducted in other regions (r = −0.15, 95% CI −0.31–0.01) [Figure 4b]. No significant subgroup effects were detected (p = 0.09), and the limited number of studies outside East Asia limits the ability to draw any firm conclusions about regional effects.

Subgroup analysis by vitamin D assay method

When stratified by vitamin D assay method, one study used an ELISA-based assay (r = −0.42, 95% CI −0.52–−0.31), and three studies used CLIA-based assays (r = −0.24 (95% CI −0.37–−0.10; I² = 57%) [Figure 4c]. The test for subgroup differences was statistically significant (p = 0.03), suggesting that the assay method may influence the estimated strength of the association. However, this result should be interpreted cautiously because the ELISA-based assay subgroup was represented by only one study.

Overall, this meta-analysis suggests that lower serum vitamin D levels are associated with larger cerebral infarct volumes on MRI. However, a small number of studies, substantial heterogeneity, and instability of subgroup analyses limit confidence in the precision of the effect estimates.

DISCUSSION

Our systematic review of 18 studies found a consistently inverse relationship between serum 25-hydroxyvitamin D levels and the initial severity of AIS. In a pooled meta-analysis of 14 studies, the combined correlation coefficient was r=−0.39 (95% CI: −0.45 to −0.32; p < 0.0001), indicating that lower vitamin D level is significantly associated with more severe stroke at admission. This finding aligns with previous work: for example, Turetsky et al. ^[25] reported that each 10-ng/mL drop in 25-hydroxyvitamin D doubled the risk of a poor 90-day outcome and was associated with larger infarct volumes. Similarly, Kamal et al. ^[12] found that more severe vitamin D deficiency correlated with NIHSS scores and worse outcomes. These results suggest that vitamin D status may be an important predictor of initial stroke severity and short-term prognosis.

Multiple patient factors likely contribute to this association. Age is a major determinant; 13 out of 18 included studies enrolled older adults (> 60 years old), in whom vitamin D levels tend to be lower due to reduced cutaneous synthesis as well as renal activation, and stroke recovery is generally poorer.^[26] Gender-related factors also contribute to vitamin D status, with postmenopausal women frequently exhibiting lower levels due to variations in cutaneous synthesis, body composition, and behavioral factors.^[27] In this review, seven studies had more than 50% women, which could influence the overall vitamin D distribution. Notably, 12 of the included studies explicitly excluded participants on vitamin D supplements; this helps isolate endogenous vitamin D status but also means that vitamin D supplementation practices, either dietary or sunlight exposure, differed across populations.^[27]

Medical comorbidities and lifestyle factors are important confounders. Conditions such as hypertension, diabetes, heart disease, and dyslipidemia both elevate stroke risk and often co-occur with low vitamin D. For instance, Jiang et al.^[28] reported a Chinese cohort that lower 25(OH)D was associated with dyslipidemia, inversely correlated with LDL and triglycerides, and positively with HDL. Vitamin D deficiency itself can promote vascular risk factors through inflammation or endothelial dysfunction.^[29,30] Lifestyle factors, such as poor diet, inactivity, smoking, and heavy alcohol use, further deplete vitamin D and worsen vascular health, potentially intensifying stroke severity.^[31]

Geography and ethnicity also modulate vitamin D status. Darker skin pigmentation and cultural dress can reduce ultraviolet absorption, and regions at higher latitudes receive less ultraviolet-B radiation. Consistently, our review focuses on the Asian and African population, where vitamin D deficiency is highly prevalent. For example, Nimitphong and Holick^[32] reported that over 70% of Indians, including children, pregnant women, and adults, are vitamin D deficient despite the abundant sun. A meta-analysis of 44,717 South East Asian adults showed that the overall prevalence of vitamin D deficiency was 68%, with a weighted mean level of vitamin D of 19.15 ng/mL.^[33] Such an endemic deficiency in the study populations may partly explain why low vitamin D levels emerged as a marker of severe stroke. These regional differences underscore that our findings are most directly applicable to similar high-deficiency settings, and caution is needed in extrapolating to populations with different sun exposure or supplementation practices.

Beyond initial severity, vitamin D appears to be related to downstream outcomes. A subset of studies (6 studies) examined functional recovery by the mRS. Generally, higher vitamin D levels were linked to better mRS scores, indicating less disability. For instance, Kamal et al. ^[12] similarly reported that a more severe deficiency was associated with increased mortality and poor functional outcome. Differences in measurement timing (at admission, discharge, 3 months, 6 months) are likely to contribute to some variability, but overall, the trend is consistent that adequate vitamin D may support recovery. In practical terms, stroke with very low vitamin D tended to have higher rates of poor outcome (mRS >2) and mortality.^[12,25]

Similarly, smaller infarct size has been observed with higher vitamin D. Several studies using MRI-DWI or CT-ASPECTS reported a weak but significant negative correlation between serum 25(OH)D and infarct volume.^[25,28] In other words, patients with low vitamin D on admission often had larger areas of brain ischemia. Safari et al.^[18] even found a positive correlation between 25(OH)D and ASPECTS score (higher ASPECTS indicating smaller infarcts).^[18] These imaging findings reinforce that vitamin D deficiency may influence the biological extent of ischemic damage.

There are plausible biological mechanisms for these observations. Vitamin D plays multiple roles in vascular and neurological health. It improves endothelial function and arterial compliance, partly by upregulating endothelial nitric oxide synthase and reducing inflammation.^[30,34]Al Mheid et al.,^[34] showed that vitamin D insufficiency is associated with increased arterial stiffness and endothelial dysfunction.^[34] Vitamin D also suppresses the renin– angiotensin system, which can lower blood pressure and reduce vascular resistance.^[30,35] Through these pathways, adequate vitamin D may enhance collateral blood flow in the ischemic penumbra: Terpolilli et al.,^[26] demonstrated that boosting nitric oxide (the same mediator enhanced by vitamin D) improves penumbral perfusion in stroke models.

Moreover, vitamin D has neuroprotective and anti-inflammatory effects. It is a neurosteroid involved in brain development and plasticity.^[27] In experimental stroke, vitamin D has been linked to attenuating ischemia-induced blood–brain barrier disruption and reducing pro-inflammatory cytokine release.^[27,31] Yin and Agrawal^[29] noted that vitamin D regulates immune responses by inhibiting pro-inflammatory cell proliferation and cytokine production. Surdu et al.,^[30] also reported that vitamin D deficiency accelerates atherosclerosis and is strongly associated with ischemic stroke and poorer post-stroke survival. In aggregate, these data support the notion that low vitamin D can exacerbate ischemic injury, via more inflammation, oxidative stress, vascular dysfunction, and blunt recovery, whereas normal levels exert vascular and neural stabilization.

The findings on severity complement prior meta-analyses of vitamin D and stroke risk. Zhou et al.,^[36] found a higher risk of ischemic stroke among individuals with lower vitamin D concentrations, with a pooled relative risk of 1.62. Consistently, Sun et al.,^[37] reported an ~1.5-fold increased stroke incidence or adverse outcomes in deficient patients, with similar estimates in analyses restricted to AIS. Our results extend this evidence that not only is low vitamin D a risk factor for having a stroke, it also predicts worse severity and outcome among those who suffer one. This reinforces the possibility that vitamin D status could be considered in stroke risk stratification and management.

We acknowledge several limitations. First, many included studies had relatively small sample sizes, limiting statistical power and precision of effect estimates. Second, there was moderate heterogeneity among studies, reflecting differences in study design, stroke assessment methods, and population characteristics. Third, all studies were conducted in Asia or Northern Africa, where vitamin D deficiency is common, so results may not generalize to other regions or ethnic groups. Finally, as an observational review, our findings are subject to unmeasured confounding, such as nutrition and unrecorded comorbidities, and cannot prove causality. Despite these limits, our synthesis suggests a robust association between vitamin D deficiency and stroke severity. Future research should aim to confirm these findings in larger, more diverse cohorts, clarify dose–response relationships, and test whether correcting vitamin D deficiency can improve stroke outcomes through randomized trials.

CONCLUSION

In this systematic review and meta-analysis, lower serum vitamin D levels were consistently associated with more severe AISs at presentation. Patients with vitamin D deficiency tended to have larger infarct volumes and worse functional outcomes. These results suggest that vitamin D status may be a valuable prognostic indicator in AIS. However, given the moderate heterogeneity and observational nature of the data, our conclusions should be interpreted with caution. Overall, the evidence supports a role for vitamin D in modulating stroke severity and recovery, but further well-designed studies are needed to establish causality and determine whether vitamin D supplementation can improve clinical outcomes in stroke patients.