T N van den Kommer1, H C Comijs, M G Dik, C Jonker, D J H Deeg. 1. Longitudinal Aging Study Amsterdam and EMGO Institute, VU University Medical Center, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands. tn.vandenkommer@vumc.nl
Abstract
OBJECTIVE: To develop two classification models for use in primary care to aid early identification of persons at risk for persistent cognitive decline. METHODS: Data were used from the Longitudinal Aging Study Amsterdam (LASA), an ongoing populationbased study. The study sample consisted of 2,021 non-demented men and women aged 58-88 years. Data on relevant predictors of persistent cognitive decline were collected at baseline. RESULTS: The incidence of persistent cognitive decline after three years of follow-up was 4.0 %. In the first model, in which predictors already known or otherwise easily assessed (first set) were included, age was the strongest predictor of persistent cognitive decline, with an increased risk for persons > 75. In addition, having memory problems, low education, and a MMSE score of < or = 24, resulted in a predictive value for persistent cognitive decline of 43.5 %. In the second classification model, in addition to the first set, predictors requiring additional measurement (e.g. markers determined in blood) were included in the analyses. Age was again the strongest predictor of persistent cognitive decline. In persons > 75 years, having a low total cholesterol level (< 5.0 mmol/L) and a MMSE score of < or = 24 resulted in a predictive value of 30.0 %. CONCLUSIONS: Both models lead to a substantial increase of the predictive value for persistent cognitive decline, that is from 4.0 % to 43.5 % and 30.0 %, and may identify to a large extent a different subsample of persons who are at risk for persistent cognitive decline. The developed classification trees could be useful for case-finding of persons at risk for future persistent cognitive decline who are therefore at risk for dementia, in a feasible and cost-effective manner.
OBJECTIVE: To develop two classification models for use in primary care to aid early identification of persons at risk for persistent cognitive decline. METHODS: Data were used from the Longitudinal Aging Study Amsterdam (LASA), an ongoing populationbased study. The study sample consisted of 2,021 non-demented men and women aged 58-88 years. Data on relevant predictors of persistent cognitive decline were collected at baseline. RESULTS: The incidence of persistent cognitive decline after three years of follow-up was 4.0 %. In the first model, in which predictors already known or otherwise easily assessed (first set) were included, age was the strongest predictor of persistent cognitive decline, with an increased risk for persons > 75. In addition, having memory problems, low education, and a MMSE score of < or = 24, resulted in a predictive value for persistent cognitive decline of 43.5 %. In the second classification model, in addition to the first set, predictors requiring additional measurement (e.g. markers determined in blood) were included in the analyses. Age was again the strongest predictor of persistent cognitive decline. In persons > 75 years, having a low total cholesterol level (< 5.0 mmol/L) and a MMSE score of < or = 24 resulted in a predictive value of 30.0 %. CONCLUSIONS: Both models lead to a substantial increase of the predictive value for persistent cognitive decline, that is from 4.0 % to 43.5 % and 30.0 %, and may identify to a large extent a different subsample of persons who are at risk for persistent cognitive decline. The developed classification trees could be useful for case-finding of persons at risk for future persistent cognitive decline who are therefore at risk for dementia, in a feasible and cost-effective manner.
Authors: Serge Gauthier; Barry Reisberg; Michael Zaudig; Ronald C Petersen; Karen Ritchie; Karl Broich; Sylvie Belleville; Henry Brodaty; David Bennett; Howard Chertkow; Jeffrey L Cummings; Mony de Leon; Howard Feldman; Mary Ganguli; Harald Hampel; Philip Scheltens; Mary C Tierney; Peter Whitehouse; Bengt Winblad Journal: Lancet Date: 2006-04-15 Impact factor: 79.321
Authors: Tessa N van den Kommer; Daniel E Bontempo; Hannie C Comijs; Scott M Hofer; Miranda G Dik; Andrea M Piccinin; Cees Jonker; Dorly J H Deeg; Boo Johansson Journal: Dement Geriatr Cogn Disord Date: 2009 Impact factor: 2.959
Authors: Tessa N van den Kommer; Dorly J H Deeg; Wiesje M van der Flier; Hannie C Comijs Journal: J Gerontol B Psychol Sci Soc Sci Date: 2018-04-16 Impact factor: 4.077
Authors: Tomiko Yoneda; Jonathan Rush; Eileen K Graham; Anne Ingeborg Berg; Hannie Comijs; Mindy Katz; Richard B Lipton; Boo Johansson; Daniel K Mroczek; Andrea M Piccinin Journal: J Gerontol B Psychol Sci Soc Sci Date: 2020-01-14 Impact factor: 4.077
Authors: Elżbieta Kuźma; Maya Soni; Thomas J Littlejohns; Janice M Ranson; Natasja M van Schoor; Dorly J H Deeg; Hannie Comijs; Paulo H M Chaves; Bryan R Kestenbaum; Lewis H Kuller; Oscar L Lopez; James T Becker; Kenneth M Langa; William E Henley; Iain A Lang; Obioha C Ukoumunne; David J Llewellyn Journal: J Alzheimers Dis Date: 2016 Impact factor: 4.472
Authors: Carlijn M Maasakkers; Jurgen A H R Claassen; Paul A Gardiner; Marcel G M Olde Rikkert; Darren M Lipnicki; Nikolaos Scarmeas; Efthimios Dardiotis; Mary Yannakoulia; Kaarin J Anstey; Nicolas Cherbuin; Mary N Haan; Shuzo Kumagai; Kenji Narazaki; Tao Chen; Tze Pin Ng; Qi Gao; Ma S Z Nyunt; John D Crawford; Nicole A Kochan; Steve R Makkar; Perminder S Sachdev; Dick H J Thijssen; René J F Melis Journal: Sports Med Date: 2020-02 Impact factor: 11.136