Diagnosis of Major Depressive Disorder (MDD) is currently symptom-based and no externally validated tests are available for routine use to confirm clinical diagnoses. Eye movement abnormalities in schizophrenia (SCZ) and bipolar disorder (BPAD) have been consistently reported and their potential as a biological trait marker highlighted. Only a limited amount of research has been conducted in MDD. Eye movement performance of MDD patients (n = 99; F:M = 55:44; Mdn age = 48) was investigated using picture free-viewing, smooth pursuit and fixation stability tasks and recorded using a non-invasive EyeLink1000 infra-red eye tracker. Performance was compared with identical measures from SCZ, BPAD, Primary Care depression (DEP) and control participants. Analysis was conducted using analyses of variance and machine learning using Probabilistic Neural Networks (PNN). We discovered a unique MDD specific eye movement phenotype, which differentiated patients with MDD from other diagnostic groups with remarkable accuracy. MDDs generated a markedly poor smooth pursuit performance, characterised by small signal-to-noise ratio, small tracking gain and large positional error. Patients also exhibited a slow average saccade velocity during free-viewing and pursuit, and poor fixation maintenance on a centralised target. A PNN classifier delineated MDD from controls with exceptional statistical sensitivity (100%) and specificity (99%), independent of state or demographics. MDD was delineated from SCZ and BPAD in all models with above 89% sensitivity and 95% specificity. MDD and DEP patients were delineated with remarkable statistical sensitivity (90%) and specificity (98%). This emerging evidence suggests possible subtypes consistent with clinical features. Testretest reliability was high for a majority of performance measures; however some measures were less robust. Brief neuropsychology assessment advocated the role of frontal lobes in oculomotor behaviour. This preliminary evidence argues for a specific MDD oculomotor dysfunction and represents potential for a diagnostically applicable biological trait marker.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:690602 |
| Date | January 2016 |
| Creators | Nouzová, Eva |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=230162 |
Page generated in 0.002 seconds