There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Pei Minyue |
| Institution: | Peking University |
| Department: | Research Center of Clinical Epidemiology |
| Country: | |
| Proposed Analysis: | Most studies about prevalence of dementia was conducted via a two-phase cross-sectional study. Screening by scales in Phase I, and high-risk participants would take systemic neuropsychological tests in Phase II. In those studies, though the total non-response rates were around 10%, the non-response were mostly happened in Phase II. The non-responder in Phase II were more likely to diagnose with dementia. It was necessary to impute missing data in Phase II or we would underestimate the prevalence in dementia. We went through serval papers in studies in China, nearly all the studies did not impute missing data when calculating the prevalence. Thus, the prevalence had been underestimate. To prove our view, we conducted a research on Chinese Veteran Clinical Research (CVCR) Platform. When we did not impute the missing data the prevalence of dementia in CVCR was 8.5%. And we imputed missing data by different methods, such as multiple imputation, hot deck and regression, the prevalence was between 10-16%, which is higher after imputation and showed large variation. Now, we want to see whether this phenomenon will appear in other population. Whether the prevalence will be higher after imputation in other population and whether they will show variation or consistency among different imputation methods. We had read an article about missing data named “Practical Strategies for Extreme Missing Data Imputation in Dementia Diagnosis” which used ADNI database for external validation. Is it possible that we could use ADNI database to prove our point of view? We need demographic information (sex, age, education level), scales score (MMSE, MoCA, and ADL) and diagnose on baseline visit. Looking forward to your reply. Thanks a lot. p.s.: About CVCR: https://pubmed.ncbi.nlm.nih.gov/24451949/ Our previous work on data imputation: https://pubmed.ncbi.nlm.nih.gov/28755569/ Practical Strategies for Extreme Missing Data Imputation in Dementia Diagnosis https://pubmed.ncbi.nlm.nih.gov/34288882/ |
| Additional Investigators |
