Zhejing Ding
Background and rationale. Stroke remains the second-leading cause of death and the leading cause of acquired adult disability worldwide. Despite the success of reperfusion therapies (intravenous thrombolysis and mechanical thrombectomy), no disease-modifying neuroprotective agent has successfully translated from preclinical efficacy to clinical approval. The pharmacological pipeline has historically been dominated by single-target agents—calcium channel blockers, free-radical scavengers, glutamate receptor antagonists, anti-inflammatory biologics—each addressing only one node of a multi-factorial injury cascade. The ketogenic diet (KD), originally introduced in 1921 for refractory pediatric epilepsy, has re-emerged as a paradigmatic alternative because it simultaneously engages multiple convergent neuroprotective pathways via elevation of circulating β-hydroxybutyrate (BHB). Purpose. This scoping-style review pursues four integrated aims: (i) To characterize the bibliometric landscape of KD–stroke research between 2014 and 2026 and to project the field's trajectory through 2030 using a quadratic regression forecast model. (ii) To synthesize contemporary mechanistic literature on KD/BHB in stroke and stratify each of eight convergent neuroprotective axes using an adapted GRADE evidence framework. (iii) To evaluate differential KD efficacy across five stroke subtypes (focal ischemia with reperfusion, intracerebral hemorrhage, ischemia–reperfusion injury and thrombectomy, post-stroke cognitive impairment, neonatal hypoxic-ischemic encephalopathy, and post-cardiac-arrest brain injury). (iv) To propose a four-phase, time-window-stratified translational roadmap aligning KD formulation (exogenous ketone ester, modified Atkins diet, classical KD, MCT-based diet) with the dominant pathophysiological targets of each stroke phase. Methods. A systematic literature search was conducted across PubMed, Embase, and Web of Science for English-language publications indexed between January 2014 and June 2026. The query combined KD-/ketone-related terms with stroke-related terms (full search string archived in the OSF project). Annual publication counts were extracted from PubMed and verified against Web of Science; a quadratic regression model (publications = 0.32 × year² − 1280 × year + 1.28 × 10⁶, R² = 0.94) was fitted with 95% prediction intervals projected through 2030. Each of the eight neuroprotective axes was graded using an adapted GRADE framework (A: human RCTs; B: high-quality preclinical; C: limited preclinical; D: isolated mechanistic). Stroke-subtype efficacy was rated on a 0–5 semi-quantitative scale combining preclinical effect-size consistency, biological plausibility, and human translational signal. Expected outcomes. The review is expected to yield five principal deliverables: (1) A quantitative bibliometric forecast projecting 18–25 KD–stroke publications per year through 2030 (95% prediction interval). (2) A mechanistic integration mapping eight convergent neuroprotective axes engaged by KD-derived BHB, with GRADE evidence rating for each axis and explicit temporal window of dominant action. (3) A stroke-subtype-stratified efficacy matrix identifying focal ischemia with reperfusion as the highest-evidence subtype (GRADE B) and intracerebral hemorrhage / neonatal HIE / post-cardiac-arrest injury as lower-evidence subtypes requiring further investigation. (4) A four-phase translational roadmap (hyperacute 0–6 h → acute 6–72 h → subacute 3–14 days → chronic weeks–months) specifying formulation, target engagement, and trial design recommendations. (5) A critical discussion of open questions, including KD formulation heterogeneity, adherence barriers, drug–diet interactions, cross-ancestry generalizability, and the need for a validated biomarker of cerebral ketolysis. Significance. This review consolidates a fragmented preclinical and early-clinical literature into a unified mechanistic, quantitative, and translational framework. It is expected to inform (a) the design of phase II/III randomized trials of KD formulations in stroke (with readouts maturing 2027–2029 in time to inform 2030 practice guidelines); (b) the selection of biologically grounded surrogate biomarkers; (c) cross-disciplinary dialogue among stroke neurology, metabolic therapy, and neurocritical care communities. Design classification. Scoping review with embedded bibliometric forecast and GRADE-adapted mechanistic synthesis. Reporting follows PRISMA-ScR. Registration type: retrospective.