Structural neuroimaging biomarkers for obsessive-compulsive disorder in the ENIGMA-OCD consortium: medication matters.
Bruin WB., Taylor L., Thomas RM., Shock JP., Zhutovsky P., Abe Y., Alonso P., Ameis SH., Anticevic A., Arnold PD., Assogna F., Benedetti F., Beucke JC., Boedhoe PSW., Bollettini I., Bose A., Brem S., Brennan BP., Buitelaar JK., Calvo R., Cheng Y., Cho KIK., Dallaspezia S., Denys D., Ely BA., Feusner JD., Fitzgerald KD., Fouche J-P., Fridgeirsson EA., Gruner P., Gürsel DA., Hauser TU., Hirano Y., Hoexter MQ., Hu H., Huyser C., Ivanov I., James A., Jaspers-Fayer F., Kathmann N., Kaufmann C., Koch K., Kuno M., Kvale G., Kwon JS., Liu Y., Lochner C., Lázaro L., Marques P., Marsh R., Martínez-Zalacaín I., Mataix-Cols D., Menchón JM., Minuzzi L., Moreira PS., Morer A., Morgado P., Nakagawa A., Nakamae T., Nakao T., Narayanaswamy JC., Nurmi EL., O'Neill J., Pariente JC., Perriello C., Piacentini J., Piras F., Piras F., Reddy YCJ., Rus-Oswald OG., Sakai Y., Sato JR., Schmaal L., Shimizu E., Simpson HB., Soreni N., Soriano-Mas C., Spalletta G., Stern ER., Stevens MC., Stewart SE., Szeszko PR., Tolin DF., Venkatasubramanian G., Wang Z., Yun J-Y., van Rooij D., ENIGMA-OCD Working Group None., Thompson PM., van den Heuvel OA., Stein DJ., van Wingen GA.
No diagnostic biomarkers are available for obsessive-compulsive disorder (OCD). Here, we aimed to identify magnetic resonance imaging (MRI) biomarkers for OCD, using 46 data sets with 2304 OCD patients and 2068 healthy controls from the ENIGMA consortium. We performed machine learning analysis of regional measures of cortical thickness, surface area and subcortical volume and tested classification performance using cross-validation. Classification performance for OCD vs. controls using the complete sample with different classifiers and cross-validation strategies was poor. When models were validated on data from other sites, model performance did not exceed chance-level. In contrast, fair classification performance was achieved when patients were grouped according to their medication status. These results indicate that medication use is associated with substantial differences in brain anatomy that are widely distributed, and indicate that clinical heterogeneity contributes to the poor performance of structural MRI as a disease marker.