Journal of Biology and Nature

Plant-derived preparations occupy a prominent and expanding place in the self-management of memory complaints and age-related cognitive decline, yet the relationship between their pharmacological promise and their demonstrated clinical value remains poorly resolved. This critical narrative review examines the evidence for medicinal plants as interventions for cognitive health in adults, encompassing healthy ageing, subjective cognitive complaint, mild cognitive impairment and established dementia. Literature was identified through structured searching of biomedical and bibliographic indexes together with citation tracking and interrogation of institutional sources, and appraised against criteria appropriate to botanical intervention research rather than to conventional single-molecule pharmacology. The synthesis is organised around five interlocking problems: the conceptual instability of cognitive-enhancement terminology; the evidentiary asymmetry between mechanistic and clinical findings; species-level heterogeneity in trial quality and effect direction; the methodological features that generate apparently contradictory results; and the product-level determinants that make many trials functionally non-replicable. Evidence is strongest, though still modest in magnitude, for standardised extracts evaluated in symptomatic populations over sustained periods, and weakest for acute studies in cognitively intact adults where ceiling effects and outcome multiplicity dominate. Large long-term prevention trials have consistently failed to reproduce the benefits suggested by shorter symptomatic trials, an asymmetry that is better explained by differences in population, exposure and outcome sensitivity than by simple publication bias. Chemical characterisation of the tested material is frequently absent or inadequate, which converts nominally identical interventions into pharmacologically distinct entities and undermines pooled estimates. Botanical authentication failures, undeclared synthetic adulterants and limited systemic and central bioavailability further separate the material studied in trials from the products consumers obtain. Priorities for the field include mandatory extract characterisation, biomarker-anchored dose selection, adequately powered trials in defined symptomatic populations, and harmonised cognitive outcome sets. Present evidence supports continued rigorous investigation of a small number of well-characterised extracts, but does not support general claims of cognitive benefit for medicinal plants as a category.