Prevalence of mild behavioral impairment among a Chinese cohort of elderly with mild cognitive impairment and subjective cognitive impairment

Introduction Mild behavioural impairment (MBI) describes later-life onset of sustained and meaningful neuropsychiatric symptoms (NPS) of any degree of severity before the onset of dementia or even the onset of cognitive symptoms. To our knowledge, there was only one study conducted in China and it only included non-dementia patients but did not differentiate among the cognitive groups. Methods A total of 88 subjects (28 cognitively normal, 15 subjective cognitive impairment, and 45 mild cognitive impairment) were recruited from Memory Clinic from 1st August 2020 through 16th Dec 2021. Inclusion criteria were that subjects should be of Chinese Han ethnicity. MBI was diagnosed according to the International Society to Advance Alzheimer’s Research and Treatment - Alzheimer’s Association (ISTAART-AA) criteria. MBI-checklist (MBI-C) score was either self-rated or informant-rated. NPI severity of 0 was defined as having a score of 0 for all Neuropsychiatric Inventory Questionnaire (NPI-Q) questions. NPI severity of 1 was given when the subject had a maximum score of 1-3. NPI severity of 2 referred to a subject having a score of four or over for at least one question. This cross-sectional study aimed to 1) calculate the prevalence of MBI among SCI/CN and MCI subjects in the Chinese population; 2) compare the performance of MBI-C and NPI-Q scales in MBI case ascertainment. Results Our results showed that the prevalence of MBI was 7% in CN/SCI group and 11.1% in MCI group assessed by MBI-C. The prevalence of NPS (severity >0) was 69% in CN/SCI group and 72.7% in MCI group assessed by NPI-Q. Spearman’s correlation test showed that the correlations of MBI and NPI severity were significant in overall subjects (correlation coefficient=0.3460, p=0.0011) and MCI group (correlation coefficient=0.3710, p=0.0131), and showing a trend of significance in CN/SCI group (correlation coefficient=0.3040, p=0.0503). However, Cohen’s kappa test showed that MBI-C and NPI-Q scale had only slight agreement in the overall and the both cognitive groups (overall: κ=0.0101; CN/SCI: κ=0.0667; MCI: κ=0.0917). Conclusion In conclusion, prevalence of MBI is around 7-11% among CN/SCI and MCI subjects in the Chinese population. MBI-C and NPI-Q were different tools and the former should be relied on to detect MBI. Further studies should validate an optimal cut-off score on MBI-C for each cognitive group in the Chinese population.