TY - JOUR
T1 - An empirical comparison of meta- and mega-analysis with data from the ENIGMA Obsessive-Compulsive Disorder working group
AU - ENIGMA OCD Working Group
AU - Boedhoe, Premika S. W.
AU - Heymans, Martijn W.
AU - Schmaal, Lianne
AU - Abe, Yoshinari
AU - Alonso, Pino
AU - Ameis, Stephanie H.
AU - Anticevic, Alan
AU - Arnold, Paul D.
AU - Batistuzzo, Marcelo C.
AU - Benedetti, Francesco
AU - Beucke, Jan C.
AU - Bollettini, Irene
AU - Bose, Anushree
AU - Brem, Silvia
AU - Calvo, Anna
AU - Calvo, Rosa
AU - Cheng, Yuqi
AU - Cho, Kang lk K.
AU - Ciullo, Valentina
AU - Dallaspezia, Sara
AU - Denys, Damiaan
AU - Feusner, Jamie D.
AU - Fitzgerald, Kate D.
AU - Fouches, Jean-Paul
AU - Fridgeirsson, Egill A.
AU - Gruner, Patricia
AU - Henna, Gregory L.
AU - Hibar, Derrek P.
AU - Hoexter, Marcelo Q.
AU - Hu, Hao
AU - Huyser, Chaim
AU - Jahanshad, Neda
AU - James, Anthony
AU - Kathmann, Norbert
AU - Kaufmann, Christian
AU - Koch, Kathrin
AU - Kwon, Jun Soo
AU - Lazaro, Luisa
AU - Lochner, Christine
AU - Marsh, Rachel
AU - Martinez-Zalacain, Ignacio
AU - Mataix-Cols, David
AU - Menchon, Jose M.
AU - Minuzzi, Luciano
AU - Morer, Astrid
AU - Nakamae, Takashi
AU - Nakao, Tomohiro
AU - Narayanaswamy, Janardhanan C.
AU - van den Heuvel, Odile A.
AU - Twisk, Jos W. R.
AU - ENIGMA-OCD Working-Group
AU - Cho, Kang Ik K
AU - Fouche, Jean-Paul
AU - Hanna, Gregory L
AU - Nishida, Seiji
AU - van Wingen, Guido A
PY - 2019/1/8
Y1 - 2019/1/8
N2 - Objective: Brain imaging communities focusing on different diseases have increasingly started to collaborate and to pool data to perform well-powered meta- and mega-analyses. Some methodologists claim that a one-stage individual-participant data (IPD) mega-analysis can be superior to a two-stage aggregated data meta-analysis, since more detailed computations can be performed in a mega-analysis. Before definitive conclusions regarding the performance of either method can be drawn, it is necessary to critically evaluate the methodology of, and results obtained by, meta- and mega-analyses. Methods: Here, we compare the inverse variance weighted random-effect meta-analysis model with a multiple linear regression mega-analysis model, as well as with a linear mixed-effects random-intercept mega-analysis model, using data from 38 cohorts including 3,665 participants of the ENIGMA-OCD consortium. We assessed the effect sizes and standard errors, and the fit of the models, to evaluate the performance of the different methods. Results: The mega-analytical models showed lower standard errors and narrower confidence intervals than the meta-analysis. Similar standard errors and confidence intervals were found for the linear regression and linear mixed-effects random-intercept models. Moreover, the linear mixed-effects random-intercept models showed better fit indices compared to linear regression mega-analytical models. Conclusions: Our findings indicate that results obtained by meta- and mega-analysis differ, in favor of the latter. In multi-center studies with a moderate amount of variation between cohorts, a linear mixed-effects random-intercept mega-analytical framework appears to be the better approach to investigate structural neuroimaging data.
AB - Objective: Brain imaging communities focusing on different diseases have increasingly started to collaborate and to pool data to perform well-powered meta- and mega-analyses. Some methodologists claim that a one-stage individual-participant data (IPD) mega-analysis can be superior to a two-stage aggregated data meta-analysis, since more detailed computations can be performed in a mega-analysis. Before definitive conclusions regarding the performance of either method can be drawn, it is necessary to critically evaluate the methodology of, and results obtained by, meta- and mega-analyses. Methods: Here, we compare the inverse variance weighted random-effect meta-analysis model with a multiple linear regression mega-analysis model, as well as with a linear mixed-effects random-intercept mega-analysis model, using data from 38 cohorts including 3,665 participants of the ENIGMA-OCD consortium. We assessed the effect sizes and standard errors, and the fit of the models, to evaluate the performance of the different methods. Results: The mega-analytical models showed lower standard errors and narrower confidence intervals than the meta-analysis. Similar standard errors and confidence intervals were found for the linear regression and linear mixed-effects random-intercept models. Moreover, the linear mixed-effects random-intercept models showed better fit indices compared to linear regression mega-analytical models. Conclusions: Our findings indicate that results obtained by meta- and mega-analysis differ, in favor of the latter. In multi-center studies with a moderate amount of variation between cohorts, a linear mixed-effects random-intercept mega-analytical framework appears to be the better approach to investigate structural neuroimaging data.
KW - IPD meta-analysis
KW - MRI
KW - linear mixed-effect models
KW - mega-analysis
KW - neuroimaging
U2 - https://doi.org/10.3389/fninf.2018.00102
DO - https://doi.org/10.3389/fninf.2018.00102
M3 - Article
C2 - 30670959
SN - 1662-5196
VL - 12
JO - Frontiers in neuroinformatics
JF - Frontiers in neuroinformatics
M1 - 102
ER -