Bilingualism effect for delaying dementia onset: a Bayesian meta-analysis.

Abstract

Evidence suggests that lifelong practice of two or more languages may contribute to cognitive reserve and protect against dementia. Most retrospective studies comparing bilinguals' and monolinguals' age at dementia onset report a delay for bilinguals. However, some studies suggest that such protective effects may be specific to certain types of dementia. Meanwhile, several prospective studies which investigated whether bilingualism reduces the incidence rate of dementia produced null results. Therefore, we conducted the first Bayesian meta-analysis to examine the relationship between bilingualism and age at onset of dementia and bilingualism and the incidence rate of dementia respectively. The results from 18 age at onset studies revealed that bilinguals were, on average, 3.45 years older than monolinguals at symptom onset (95% credible interval [CrI]: [2.8, 4.1]). Moreover, the result was only slightly influenced by years of education, although the effect of education on dementia remains uncertain. The results from six incidence rate studies did not provide robust evidence for bilingualism's preventive effect. Therefore, this meta-analysis demonstrated that bilingualism has potential effects in delaying the onset of dementia but not preventing it. Moreover, subgroup analyses revealed a relatively consistent pattern of delayed symptom onset among bilinguals, though the estimated delays differed across dementia types. However, due to the limited number of studies available for each dementia subtype, uncertainty remains regarding whether bilingualism induces differential effects across dementia types. This highlights the need for further research specifically targeting non-Alzheimer's dementias to better clarify these potential differences.