Аннотация
BACKGROUND: Preclinical studies show that experimental tooth extraction accelerates hippocampal degeneration, but the human neuroanatomical evidence has never been synthesised. AIM: To map and critically appraise structural MRI studies on associations between masticatory function and hippocampal/medial-temporal morphology in later life. METHODS: A PRISMA-ScR-registered search (PubMed, Embase, Web of Science, Scopus, CINAHL; 1995-15 June 2025) targeted peer-reviewed studies of adults ≥ 60 years reporting quantitative masticatory indicators and in vivo structural neuroimaging outcomes including hippocampal or closely related medial-temporal metrics. Dual reviewers applied a PCC eligibility matrix, extracted data across five domains and appraised methodological quality using an eight-item JBI adaptation tailored to neuroimaging. RESULTS: Eight studies published 2016-2024 (Japan = 4; Taiwan = 3; Sweden = 1) were included. Sample sizes ranged from 15 to 919 participants (median = 80). Six studies quantified dentition/tooth number, three applied objective chewing-performance tests and one used an occlusal-support composite (Eichner class). Five of six hippocampal/medial-temporal analyses reported significant inverse associations between poorer mastication and smaller hippocampal or parahippocampal measures, while one showed a concordant but non-significant trend. Dose-response gradients were evident in four datasets. In one population-based cohort, each 10-tooth decrement corresponded to approximately 2.2% lower bilateral hippocampal volume. In a study of cognitively impaired older adults, greater missing-tooth count was associated with smaller left hippocampal and parahippocampal GMV, whereas no corresponding association was detected in cognitively normal controls. CONCLUSIONS: Current structural MRI evidence in adults ≥ 60 years describes predominantly inverse associations between masticatory compromise and hippocampal/medial-temporal integrity; causality cannot be inferred. Longitudinal population-based neuroimaging using standardised chewing metrics, hippocampal subfield segmentation and rigorous adjustment for vascular, nutritional and inflammatory covariates is required to determine temporality and to test whether oral rehabilitation is associated with slower medial-temporal decline.