Title | Multiple imputation of cognitive performance as a repeatedly measured outcome. |
Publication Type | Journal Article |
Year of Publication | 2017 |
Authors | Rawlings AMonica, Sang Y, Sharrett ARichey, Coresh JJ, Griswold M, Kucharska-Newton AMaria, Palta P, Wruck LMiller, Gross ALawrence, Deal JAnne, Power MCarolyn |
Secondary Authors | Bandeen-Roche KJean |
Journal | Eur J Epidemiol |
Volume | 32 |
Issue | 1 |
Pagination | 55-66 |
Date Published | 2017 01 |
ISSN | 1573-7284 |
Keywords | Cognition, Computer Simulation, Data Interpretation, Statistical, Female, Humans, Male, Middle Aged, Models, Statistical, Observational Studies as Topic, Outcome Assessment, Health Care, Prospective Studies, Research Design |
Abstract | Longitudinal studies of cognitive performance are sensitive to dropout, as participants experiencing cognitive deficits are less likely to attend study visits, which may bias estimated associations between exposures of interest and cognitive decline. Multiple imputation is a powerful tool for handling missing data, however its use for missing cognitive outcome measures in longitudinal analyses remains limited. We use multiple imputation by chained equations (MICE) to impute cognitive performance scores of participants who did not attend the 2011-2013 exam of the Atherosclerosis Risk in Communities Study. We examined the validity of imputed scores using observed and simulated data under varying assumptions. We examined differences in the estimated association between diabetes at baseline and 20-year cognitive decline with and without imputed values. Lastly, we discuss how different analytic methods (mixed models and models fit using generalized estimate equations) and choice of for whom to impute result in different estimands. Validation using observed data showed MICE produced unbiased imputations. Simulations showed a substantial reduction in the bias of the 20-year association between diabetes and cognitive decline comparing MICE (3-4 % bias) to analyses of available data only (16-23 % bias) in a construct where missingness was strongly informative but realistic. Associations between diabetes and 20-year cognitive decline were substantially stronger with MICE than in available-case analyses. Our study suggests when informative data are available for non-examined participants, MICE can be an effective tool for imputing cognitive performance and improving assessment of cognitive decline, though careful thought should be given to target imputation population and analytic model chosen, as they may yield different estimands. |
DOI | 10.1007/s10654-016-0197-8 |
Alternate Journal | Eur J Epidemiol |
PubMed ID | 27619926 |
PubMed Central ID | PMC5332286 |
Grant List | HHSN268201100012C / HL / NHLBI NIH HHS / United States HHSN268201100009I / HL / NHLBI NIH HHS / United States U01 HL096812 / HL / NHLBI NIH HHS / United States HHSN268201100010C / HL / NHLBI NIH HHS / United States T32 AG027668 / AG / NIA NIH HHS / United States HHSN268201100008C / HL / NHLBI NIH HHS / United States UL1 TR001079 / TR / NCATS NIH HHS / United States P50 AG005146 / AG / NIA NIH HHS / United States HHSN268201100005G / HL / NHLBI NIH HHS / United States U01 HL096917 / HL / NHLBI NIH HHS / United States HHSN268201100008I / HL / NHLBI NIH HHS / United States HHSN268201100007C / HL / NHLBI NIH HHS / United States HHSN268201100011I / HL / NHLBI NIH HHS / United States HHSN268201100011C / HL / NHLBI NIH HHS / United States U01 HL075572 / HL / NHLBI NIH HHS / United States U01 HL096902 / HL / NHLBI NIH HHS / United States T32 HL007024 / HL / NHLBI NIH HHS / United States HHSN268201100006C / HL / NHLBI NIH HHS / United States HHSN268201100005I / HL / NHLBI NIH HHS / United States U01 HL096814 / HL / NHLBI NIH HHS / United States HHSN268201100009C / HL / NHLBI NIH HHS / United States R01 HL070825 / HL / NHLBI NIH HHS / United States HHSN268201100005C / HL / NHLBI NIH HHS / United States U01 HL096899 / HL / NHLBI NIH HHS / United States HHSN268201100007I / HL / NHLBI NIH HHS / United States T32 HL007055 / HL / NHLBI NIH HHS / United States K99 AG052830 / AG / NIA NIH HHS / United States |