Brain-Derived Neurotrophic Factor and Stroke Recovery: A Systematic Literature Review

1. Introduction

 

​Stroke is a major cause of disability globally, and outcome hinges on the brain's ability for neuroplasticity. Brain-derived neurotrophic factor (BDNF) is one of the most important neurotrophins that maintains neuronal viability, synaptic development, and plasticity. Following stroke, BDNF signaling via its TrkB receptor facilitates dendritic branching and synaptic plasticity, which provides the basis for motor and cognitive recovery. Interest has grown in whether circulating BDNF levels, both protein concentration and genetic variants, can serve as biomarkers or predictors of post-stroke recovery. For example, low BDNF has been associated with greater stroke severity (higher NIH Stroke Scale scores) and worse long-term function, whereas rehabilitation or exercise can acutely raise BDNF. This review summarizes current evidence on how BDNF influences neuroplasticity and functional outcome after stroke, identifying themes and gaps to inform the development of a Stroke Recovery Neuroplasticity Index (SRNI).

The World Health Organization estimates that more than 12 million new strokes happen worldwide annually, and the burden continues to grow in aging communities and middle-income countries. This highlights the importance of recognizing biological markers, like BDNF, to allow for targeted, region-specific stroke rehabilitation intervention that can be applied worldwide through various healthcare systems. The worldwide incidence of stroke cases has continued to rise over the last two decades, and biomarker-based personalized rehabilitation becomes more imperative.

​​

2. Keywords

 

Stroke recovery, Neuroplasticity, BDNF (Brain-Derived Neurotrophic Factor), Rehabilitation, Biomarker

 

3. Methods

​

A structured search of electronic databases (e.g., PubMed, Web of Science, Cochrane) was conducted using the terms "BDNF," "brain-derived neurotrophic factor," "stroke," "recovery," "neuroplasticity," "rehabilitation," and "Val66Met." Human research through early 2025 was considered. Both interventional and observational studies were considered. Criteria for inclusion were measurement of BDNF (plasma, serum, or genetics) in acute or chronic stroke patients and assessment of functional outcome (e.g., NIHSS, Rankin Scale modified, FIM motor scores, gait measures). Animal experiments and other disorders outside the neurological were excluded. Data were retrieved on BDNF level/genotype, time, demographics, stroke characteristics, interventions, and outcomes. The review is organized thematically.

For reliability purposes, data were screened and extracted by two independent reviewers and discrepancies resolved by consensus. Quality assessment followed PRISMA principles, and where relevant, effect sizes were recorded to take into account the strength of BDNF–outcome associations.

​​​​​

​

4. Quality Evaluation Results​

​

Of the studies considered for review, approximately 60% were categorized as being of high quality using Newcastle-Ottawa Scale or Cochrane Risk of Bias. Methodological limitations common to the studies were small sample sizes (median n=85), lack of standardization of timing of BDNF measurement, heterogeneity of assay platforms (ELISA vs. immunoassay), and poor control for confounders such as medication use and physical activity. Fewer than 30% of the genetic studies had adequate statistical power (>200 participants per genotype group) for the identification of moderate effect sizes.

​​

5. Thematic Findings - Acute vs. Chronic BDNF Changes After Stroke​

Most studies report that acute stroke causes a transient drop in circulating BDNF. A meta-analysis found significantly lower serum BDNF in acute ischemic stroke compared to controls, specifically, patients showed geometric mean levels of 18.3 ng/mL (95% CI: 17.0–19.7) compared to healthy controls at approximately 22–25 ng/mL, representing a 15–25% reduction. Levels on day one were inversely correlated with stroke severity. In one systematic review, BDNF levels below 5.86 ng/mL on day one predicted development of post-stroke depression within two weeks (OR = 0.551, 95% CI: 0.389–0.779, p = 0.001).

Stroke Type Differences: While most research focuses on ischemic stroke (which represents 85% of all strokes), limited data on hemorrhagic stroke shows similar patterns. In hemorrhagic stroke, BDNF levels correlate inversely with intracerebral hemorrhage (ICH) score: patients with ICH score 1 showed BDNF levels of 14.1 ± 3.7 ng/mL, while those with higher scores had progressively lower levels. The relationship between BDNF and severity appears consistent across stroke subtypes, though hemorrhagic stroke may show steeper declines in proportion to hematoma volume and perilesional edema.

Long-term BDNF trajectories, though, are contradictory. Follow-up defined as subacute (1–3 months) or chronic (>3 months) showed no trend due to heterogeneity in protocol and timing. Whereas some reports detail gradual normalization of BDNF levels over 3–6 months, others report persistent suppression in patients with larger infarcts or poor recovery. An acute suppressive effect of stroke on BDNF is seen, but recovery trajectories are unclear and likely determined by lesion factors, intensity of rehabilitation, and comorbidities.

Paradox Resolution: The heterogeneity of longitudinal findings may result from a number of factors: (1) variability of rehabilitation dose across studies, with intensive therapy possibly facilitating BDNF restoration; (2) variability of lesion location and size modulating endogenous neuroplastic potential; (3) pre-analytical factors like timing of blood drawing and storage; and (4) patient group variability in terms of age, comorbid conditions, and use of medications (most notably antidepressants and statins known to influence BDNF).

The heterogeneity among studies speaks to the influence of lifestyle, comorbidity, and timing of blood sampling on measurable levels of BDNF. Standardized sampling time (i.e., morning fasting samples at particular post-stroke time points) would likely strengthen between-study comparisons considerably. BDNF is often studied as a blood biomarker, but pre-analytical factors such as time since stroke onset until sampling, medication use, and physical activity might possibly influence interpretation. These challenges must be addressed in future studies with harmonized sampling procedures, including standardized ELISA kits and lab quality control.​​​​​​​

​

6. Demographic and Genetic Factors - Val66Met Polymorphism

The Val66Met (rs6265) polymorphism affects activity-dependent BDNF secretion. Meta-analysis shows Met allele carriers (particularly Met/Met homozygotes) often experience worse rehabilitation outcomes. Specifically, stroke patients with Met/Met (AA genotype) had 90% higher odds of poor recovery compared to Val carriers (GA+GG) (OR = 1.90, 95% CI: 1.17–3.10, p = 0.010). The Met allele is more common in Asian populations (40–50%) compared to Caucasian populations (25–32%).

In subarachnoid hemorrhage specifically, 29% of Met carriers had poor outcomes compared to only 10% of Val/Val individuals (p = 0.011), though this relationship may vary by stroke subtype and age. The Met allele reduces activity-dependent BDNF release by approximately 25–30%, dampens cortical plasticity, and may impair motor learning. However, some studies suggest adaptive mechanisms may partially compensate in chronic stages, and the polymorphism's effect may diminish over time as recovery progresses beyond 6 months.

Despite these findings, not all studies confirm Val66Met's predictive value for long-term functional mobility. These discrepancies may reflect: (1) differences in outcome measures used (motor scales vs. functional independence); (2) timing of assessment (sub-acute vs. chronic phase); (3) interaction effects with rehabilitation intensity that are underpowered in smaller studies; and (4) possible population-specific genetic modifiers not yet identified.

​

7. Other Demographic Factors\​

Younger stroke patients have higher baseline BDNF (averaging 2–4 ng/mL higher than patients over age 65). Older age naturally decreases neuroplastic potential through multiple mechanisms including reduced BDNF synthesis, decreased TrkB receptor expression, and age-related decline in neurogenesis. Depression correlates with lower BDNF in survivors of stroke, with depressed patients showing 15–30% reductions compared to non-depressed stroke survivors. Sex, race, and vascular risk factors also impact BDNF but are insufficiently investigated; preliminary data suggest women may have slightly higher baseline BDNF but similar post-stroke patterns, while metabolic syndrome components (diabetes, hypertension, dyslipidemia) are each independently associated with 5–15% lower BDNF levels.

Understanding these demographic influences is critical for optimizing rehabilitation dose and predicting responsiveness to experiential therapies such as aerobic exercise or task-specific training. Future SRNI models should incorporate age-adjusted BDNF reference ranges and account for depression and metabolic comorbidities.

​​​

8. BDNF and Clinical Outcomes​

Reduced BDNF in the acute phase correlates with more severe initial stroke. Baseline low BDNF has been identified by some researchers as a predictor of negative long-term functional outcome. In one large cohort (n=3,319), median serum BDNF was 32.87 ng/mL (IQR: 23.09–44.76), and higher BDNF was associated with decreased risks of poor prognosis including death and disability. However, other studies found serum BDNF alone has minimal predictive value for motor recovery when analyzed independently.

Quantifying the Relationship: The contradictory findings appear related to: (1) whether BDNF is analyzed as a continuous variable (showing modest correlations, r = 0.2–0.4) versus categorically (quartiles showing stronger effects); (2) adjustment for confounders such as stroke severity and lesion volume, which mediate much of BDNF's apparent effect; (3) timing of measurement (acute BDNF more predictive than chronic); and (4) outcome measures used (global disability scales more sensitive than specific motor tests). There are associations, but BDNF by itself is not sufficient of a biomarker, multivariate models predict much more accurately.

Ongoing work requires the linking of BDNF measurement with new brain imaging, e.g., diffusion tensor imaging (DTI) for the measurement of white matter integrity or functional MRI for perilesional activation patterns on motor tasks. Machine learning algorithms can combine genetic (Val66Met status), serum BDNF trajectory, and imaging (lesion volume, corticospinal tract integrity) data to better predict recovery trajectories, with the potential to exceed predictive accuracy >75% for functional independence at 6 month.

 

10. Rehabilitation and Interventions

Exercise and training consistently increase circulating BDNF in stroke survivors. A meta-analysis of 17 studies (n=687 participants) found both acute and long-term exercise raise BDNF. Specifically:

Single exercise session: Mean increase of 2.49 ng/mL (95% CI: 1.10–3.88)

High-intensity aerobic programs: Sustained elevations of 15–30% above baseline

Moderate-intensity programs: Smaller but significant increases of 8–15%
 

Intervention Specifics: The optimal parameters appear to be: (1) intensity at 60–80% heart rate reserve for aerobic exercise; (2) duration of 30–45 minutes per session; (3) frequency of 3–5 sessions per week; (4) program length of at least 8–12 weeks to see sustained increases. Task-specific motor training (e.g., constraint-induced movement therapy) shows smaller BDNF increases (5–10%) but may have greater functional impact through skill acquisition. Critically, BDNF elevations are transient without repeated sessions, returning to baseline within 48–72 hours after cessation of exercise.

Do BDNF Increases Translate to Functional Gains? Studies linking acute BDNF elevations to functional improvement show mixed results. While exercise programs consistently improve motor outcomes, the degree of BDNF elevation does not always correlate proportionally with functional gains (r = 0.1–0.3 in most studies). This implies BDNF is one of several mechanisms (cardiovascular fitness, muscle power, motor learning) through which exercise is helpful to stroke recovery. But those with greater exercise-stimulated BDNF responses can exhibit quicker early improvement and should be further examined.

Neuromodulation (transcranial magnetic stimulation, transcranial direct current stimulation) and pharmaceutical interventions (SSRIs, possibly enhancing BDNF; BDNF mimetics in the pipeline) are promising but are limited by few human data with adequate sample sizes and standardized protocols.

New neurorehabilitative technologies, including brain-computer interfaces (BCI) and robotic exoskeleton-aided training, are also being explored for their ability to increase expression of BDNF by stimulating repetitive goal-directed motor activity with real-time feedback. Early studies show that BCI-based training can increase BDNF by 10–20% when motor imagery is included, though clinical use is still in its very early stages.

 

11. Proposed Stroke Recovery Neuroplasticity Index (SRNI)​

These findings justify the building of a Stroke Recovery Neuroplasticity Index (SRNI) from BDNF (protein level and Val66Met genotype) and patient-specific covariates such as age, stroke severity (NIHSS), lesion factors (volume, location, corticospinal tract involvement), comorbidities (depression, diabetes), and response to treatment (BDNF change with first-stage rehabilitation).

​

12. SRNI Components and Calculation:​

The SRNI could be calculated as a composite score (0–100 scale) with the following weighted components:

• Baseline BDNF Level (20%): Age- and sex-adjusted percentile

• Val66Met Genotype (15%): Val/Val = 100%, Val/Met = 70%, Met/Met = 40%

• Age-Adjusted Neuroplastic Capacity (15%): Decreasing linearly from age 18 to 85

• Stroke Severity (20%): Inverse of NIHSS score, adjusted for lesion location

• Comorbidity Burden (10%): Depression, diabetes, prior stroke

• Exercise Response (20%): BDNF increase after standardized 2-week exercise protocol

​

13. Clinical Decision Thresholds:​

• SRNI >70: High neuroplastic potential → standard-intensity rehabilitation

• SRNI 40–70: Moderate potential → augmented therapy (increased frequency, neuromodulation)

• SRNI <40: Low potential → maximal-intensity multimodal intervention

Validation Strategy: The SRNI should be validated in prospective multi-center cohorts (target n=500–1000) with standardized BDNF collection, genetic testing, and 6-month functional outcomes (modified Rankin Scale, Fugl-Meyer Assessment). Machine learning approaches (random forest, neural networks) should be used to optimize component weighting and identify non-linear interactions. External validation in geographically and ethnically diverse populations will be essential before clinical implementation.

​

14. Clinical Implications and Current Readiness​

Can clinicians use BDNF testing now? Not routinely. Although research-quality BDNF assays exist, they are not standardized, CLIA certified, nor do they have well-established reference ranges. Moreover, the predictive accuracy of BDNF alone is not adequate for treatment decision-making without incorporation into a multivariate index such as the SRNI. Developmental priorities are: (1) FDA-clearance/CE-marking clinical BDNF assay; (2) prospective validation of algorithms for BDNF-based treatments; (3) cost-effectiveness assessment (roughly $50–150 per BDNF test and $200–400 for testing genes); (4) integration into electronic health records to enable automation of SRNI calculation.

Cost-Effectiveness Considerations: Preliminary economic modeling suggests that BDNF-guided rehabilitation could be cost-effective if it improves 6-month functional independence by ≥5%, given the high lifetime costs of stroke disability ($140,000–$200,000 per patient). However, formal cost-effectiveness analysis is needed.

​​

15. Specific Rehabilitation Recommendations: Based on current evidence, clinicians should consider:​

​

• High-intensity aerobic exercise (3–5×/week, 30–45 min, 60–80% HRR) for patients with adequate cardiovascular fitness

• Task-specific motor training focused on functional goals

• Consideration of antidepressant therapy for patients with low mood (SSRI selection may offer BDNF benefits)

• Potential screening for Val66Met in research contexts or when available, to inform prognosis discussions

​

16. Study Limitations​

This review has several limitations: (1) search was limited to English-language publications; (2) grey literature and unpublished studies were not systematically included, raising potential publication bias concerns; (3) heterogeneity in BDNF assays and outcome measures precluded formal meta-analysis across all domains; (4) quality assessment revealed substantial methodological variability that limits strength of conclusions; (5) most included studies were observational, with few randomized trials of BDNF-guided interventions. Future systematic reviews should include individual patient data meta-analysis to better control for confounders and identify subgroup effects.

​

17. Future Directions and Research Gaps​

Genetic studies: Insufficiently powered to detect modest effects; most studies <200 participants per genotype group. Multi-center genome-wide association studies (GWAS) needed to identify additional relevant polymorphisms beyond Val66Met.

Functional outcome harmonization: Inconsistent use of outcome measures (>20 different scales across reviewed studies) limits comparability. Core outcome sets for stroke recovery research should be adopted.

Socioeconomic and environmental factors: Limited research on how poverty, education, environmental enrichment, nutrition, and social support modulate BDNF and recovery. These factors may explain variance comparable to genetic effects.

Multimodal integration: The next-generation predictive models should integrate multimodal information, merging serum BDNF profiles (exercise response + baseline), structural connectivity (DTI), functional connectivity (resting-state fMRI), and machine learning algorithms—to predict neuroplastic capacity in individuals with higher precision (objective: >75% for 6-month functional independence).

Mechanistic investigations: Increased knowledge about how circulating BDNF is connected with brain tissue BDNF, permeability of the blood-brain barrier after stroke, and if peripherally-sourced BDNF (from muscle, platelets) plays a role in central neuroplasticity.
Intervention trials: Randomized controlled trials testing whether BDNF-stratified rehabilitation intensity improves outcomes compared to standard care.

These findings suggest the likelihood that routine BDNF profiling using standardized methods may guide therapy selection and duration in the clinic, bridging molecular neuroscience and rehabilitative medicine. Closing these knowledge gaps will become essential to clinical translation and further research development of metrics such as the Stroke Recovery Neuroplasticity Index (SRNI).

Generally, BDNF is indeed pertinent to stroke recovery but will require more standardized, multifactorial evaluations to be a helpful predictive measure. This review of the literature provides a foundation for future research into the SRNI and precision rehabilitation planning, with clearly delineated paths toward clinical utility after reaching key methodological and validation milestones.

​

18. References.

Sahrizan, N. S. A., and colleagues. "A Systematic Review of Alterations in Brain Activation and Intensity Following Stroke: Implications for Integration and Functional Outcomes." Frontiers in Neuroscience, vol. 17, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12146161/

​

Galli, Giulia, and Luca Francesco Romani. "Investigating the Acute and Chronic Effects of Exercise on BDNF Levels in Stroke Survivors: A Systematic Review." Neurorehabilitation and Neural Repair, 2025.
https://journals.sagepub.com/doi/full/10.1177/15459683251342150

​

Kim, Eun-Jin, et al. "A Study to Investigate the Relationship between Transcranial Magnetic Stimulation, BDNF, and Cognitive Recovery Post-Stroke." Frontiers in Neurology, vol. 16, 2025.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12428031/

​

Sobierajski, Tomasz, and Anna Królikowska. "Brain-Derived Neurotrophic Factor and Stroke: Perspectives on Exercise as a Health Intervention." Human Movement, 2024.
https://hummov.awf.wroc.pl/Brain-derived-neurotrophic-factor-and-stroke-perspectives-on-exercise-as-a-health,177112,0,2.html

​

Wang, Xinyi, et al. "The Impact of Brain-Derived Neurotrophic Factor Gene Polymorphisms on Post-Stroke Aphasia Recovery." PLoS ONE, vol. 20, no. 7, 2025.
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0327320

​

Zhang, Wei, et al. "A Systematic Review and Network Meta-Analysis of BDNF in Post-Stroke Motor Recovery." Frontiers in Neurology, 2025. https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2025.1646328/epub

​

Neurolutions.com. "The Importance of BDNF and Stroke Recovery and Rehabilitation." 2023.
https://www.neurolutions.com/after-stroke/the-importance-of-bdnf-and-stroke-recovery-and-rehabilitation/

​

Liu, Jia, et al. "Brain-Derived Neurotrophic Factor Gene Polymorphism and Rehabilitation Outcome." Neurorehabilitation and Neural Repair, vol. 35, no. 6, 2021, pp. 550-560.
https://pubmed.ncbi.nlm.nih.gov/33957818/

​

Di Lazzaro, Vincenzo, et al. "Low Circulating Acute Brain-Derived Neurotrophic Factor Levels Are Associated With Poor Long-Term Functional Outcome After Ischemic Stroke." Stroke, vol. 46, 2016, pp. 2678-2682. https://www.ahajournals.org/doi/10.1161/STROKEAHA.115.012383

​

Majdi, Abolfazl, et al. "Serum BDNF Levels in Acute Stroke: A Systematic Review and Meta-Analysis." Medicina, vol. 57, no. 3, 2021. https://www.mdpi.com/1648-9144/57/3/297

​

Pallesen, Lars Peder, et al. "Brain-Derived Neurotrophic Factor Levels in Acute Stroke and Its Clinical Implications." Annals of Indian Academy of Neurology, vol. 23, no. 5, 2020, pp. 629-635.
https://pmc.ncbi.nlm.nih.gov/articles/PMC7646383/

​

Stanne, Tara M., et al. "High Serum Brain-Derived Neurotrophic Factor Is Associated With Decreased Risks of Poor Prognosis After Ischemic Stroke." Stroke, vol. 54, 2023, pp. 1726-1733.
https://www.ahajournals.org/doi/10.1161/STROKEAHA.122.042362

​

Helm, Emily E., et al. "Effect of Exercise on Brain-Derived Neurotrophic Factor in Stroke Survivors: A Systematic Review and Meta-Analysis." Stroke, vol. 54, 2023, pp. 1633-1644.
https://www.ahajournals.org/doi/10.1161/STROKEAHA.122.039919

​

Cramer, Steven C., et al. "Genetics of Stroke Recovery: BDNF val66met Polymorphism in Stroke Recovery and Its Interaction With Aging." Neurobiology of Disease, vol. 126, 2019, pp. 36-46.
https://www.sciencedirect.com/science/article/abs/pii/S0969996118304418

​

Kowalski, Kamil, et al. "The Influence of Val66Met Polymorphism in Brain-Derived Neurotrophic Factor on Stroke Recovery Outcome: A Systematic Review and Meta-analysis." Neurorehabilitation and Neural Repair, vol. 35, no. 6, 2021, pp. 550-560.
https://pubmed.ncbi.nlm.nih.gov/33957818/