The usefulness of microbiome profiling for geriatric patients with neuropsychiatric conditions: a scoping review

Abstract

Introduction: Neuropsychiatric disorders encompass psychiatric and neurodegenerative diseases. These conditions are particularly challenging to diagnose in the elderly due to their overlapping cognitive, affective, and behavioural symptoms. Recent studies indicate microbiota imbalances exist in both psychiatric and neurodegenerative conditions, suggesting potential biomarkers for diagnosis and prognosis. The objective of this scoping review is to provide a comprehensive overview of how microbiome profiling has been used in research to develop diagnostic and prognostic models for elderly patients with neuropsychiatric conditions. Methods: Following JBI guidelines, a comprehensive search was conducted across four electronic databases (MEDLINE, PsycINFO, SCOPUS, EMBASE) with strategies developed alongside academic librarians. Two independent reviewers screened titles, abstracts, and full texts based on pre-specified inclusion criteria. Data was extracted focusing on participants, study methods, and key findings. Results: Thirty-one studies employing microbiome-based predictive models were identified, primarily targeting Alzheimer’s and Parkinson’s diseases, with limited research on psychiatric conditions which was only found for depression and schizophrenia. Most studies used 16S rRNA gene sequencing and machine learning models. Findings highlight the potential of gut microbiota data to enhance predictions of neuropsychiatric conditions, though limitations included small, non-diverse cohorts and a lack of methodological standardisation. The review highlights the need for larger, longitudinal, multicentere studies to validate models and improve clinical applicability. Discussion: Microbiome-based predictive models for neuropsychiatric conditions in the elderly show promise. Future research should focus on longitudinal studies and expanding profiling methods to validate findings in larger, diverse cohorts.