Multilingualism

Research summary

• Multilingualism has one of the more geographically consistent evidence bases in dementia research, though two recent meta-analyses converge on an important nuance: it appears to delay when symptoms appear, not whether dementia develops at all.
• The foundational Canadian study of 184 dementia patients found bilinguals showed symptoms 4 years later than monolinguals despite equivalent disease severity, and a larger Indian study of 648 patients -- notable because it included illiterate participants, ruling out education as the sole explanation -- found a 4.5-year delay that held across Alzheimer's, frontotemporal, and vascular dementia subtypes.
• A systematic review and meta-analysis found bilinguals experienced Alzheimer's symptoms 4.7 years later and were diagnosed with dementia 3.3 years later than monolinguals, but found no significant reduction in the overall odds of developing dementia (odds ratio 0.89, not statistically significant).
• A more recent Bayesian meta-analysis of 18 studies confirmed a 3.45-year average delay in symptom onset but similarly found no robust evidence bilingualism reduces dementia incidence itself.
• Follow-up work has extended the delay effect to specific Alzheimer's variants (a 5-year delay in logopenic variant primary progressive aphasia) and shown that only "active" bilingualism -- continued regular use of both languages, not just past exposure -- predicts delayed onset of mild cognitive impairment.
• Because nearly all of this evidence comes from retrospective comparisons of patients already diagnosed at memory clinics, and effect sizes vary substantially by dementia subtype and study population, it's best read as a reasonably solid case for delaying symptom onset rather than a proven way to prevent dementia outright.