AGEADJUSTED NORMATIVE DATA AND DISCRIMINATIVE VALIDITY OF COGNITIVE TESTS IN THE UKRAINIAN ADULT PATIENTS WITH MAJOR DEPRESSIVE DISORDER

Abstract

Purpose of the study. We aimed: 1) to compare cognitive functioning in patients with major depressive disorder (MDD) and healthy controls (HC) in the Ukrainian adult population by the results of neuropsychological assessment, that included Perceived Deficit Questionnaire (PDQ5), Rey Auditory Verbal Learning Test (RAVLT), Trail Making Test Part B (TMTB), Digit Symbol Substitution Test (DSST); 2) to obtain age adjusted normative data of RAVLT, TMTB, and DSST tests; 3) to explore the diagnostic utility of PDQ5, RAVLT, TMTB, and DSST tests to separate patients with MDD from HC; 4) to provide cutoff scores of the PDQ5, RAVLT, TMTB, and DSST tests, stratified by age, that discriminate MDD patients from HC, based on the sensitivity (Se) and specificity (Sp) of the obtained scores.
 Materials and methods. 130 MDD medication free patients (according to DSM5) and 70 HC were enrolled in the study. Psychopathological (by MontgomeryAsberg Depression Rating Scale (MADRS) and Clinical Global Impression Severity (CGIS)) and neuropsychological (by PDQ5, RAVLT, TMTB, DSST) parameters were analyzed in all subjects. To assess betweengroup differences parametric and nonparametric tests were used (Ttest, MannWhitney test, chisquared test). Areas under the curve (AUC) of receiver operating characteristic (ROC) were calculated to determine if the results of PDQ5, RAVLT, TMTB, and DSST tests` performance could discriminate MDD patients from HC. Cutoff scores, which separated MDD patients from HC with empirical optimal Se and Sp, were derived from the ROC curves. The statistical threshold was set at p < 0.05.
 Results. Surveyed groups were comparable in age, gender, and level of education. Besides the expected statistical difference in MDD patients and HC on MADRS and CGIS scores, sufficient distinction in neuropsychological test performance was found between the comparison groups. MDD participants were significantly worse (p < 0,0001) in subjective (PDQ5) as well as objective cognitive functioning (RAVLT subtests, DSST, TMTB scores). Significant differences between MDD and HC groups, established during objective cognitive testing, were specific to each age group, despite the general trend of deterioration of cognitive performance with age. ROC analysis was used to examine the utility of PDQ5, RAVLT, TMTB, and DSST tests to discriminate MDD patients from HC. AUCROCs showed that all cognitive measures included in this study adequately differentiated between the performance of HC and MDD patients. We also provided cutoff scores for five age groups in discriminating MDD patients from HC, based on the Se and Sp of the prescribed scores. The age ranges for each group were as follows: Group 1 – 18–24 years; Group 2 – 25–34 years; Group 3 – 35–44 years;
 Group 4 – 45–54 years; Group 5 – 55–65 years. For PDQ5 cutoff scores were: in the whole sample > 3,5 points (Se 90%, Sp 91%); Group 1 > 3,5 points (Se 100%, Sp 83 %); Group 2 > 3,5 points (Se 93%, Sp 89%); Group 3 > 2,5 points (Se 89%, Sp 83%); Group 4 > 2,5 points (Se 100%, Sp 84%); Group 5 > 3,0 points (Se 90%, Sp 100%). For immediate recall of the RAVLT cutoff scores were: in the whole sample < 56,5 words (Se 85%, Sp 82%); Group 1 < 57 words (Se 100%, Sp 73%); Group 2 < 59,5 words (Se 85%, Sp 70%); Group 3 < 59,5 words (Se 91%, Sp 83%); Group 4 < 57,5 words (Se 86%, Sp 74%); Group 5 < 53,5 words (Se 94%, Sp 80%). For proactive interference of the RAVLT cutoff scores were: in the whole sample < 6,5 words (Se 66%, Sp 72%); Group 2 < 7,5 words (Se 83%, Sp 63%); Group 3 < 6,5 words (Se 70%, Sp 75%); Group 4 < 6,5 words (Se 72%, Sp 74%); an unsatisfactory quality of the models for groups 1 and 5 did not allow to determine the cutoff scores for these age groups. For retroactive interference of the RAVLT cutoff scores were: in the whole sample < 13,5 words (Se 86%, Sp 76%); Group 2 < 13,5 words (Se 85%, Sp 89%); Group 3 < 13,5 words (Se 82%, Sp 92%); Group 4 < 13.5 words (Se 82%, Sp 74%); Group 5 < 12,5 words (Se 94%, Sp 80%); Group 1 had an unsatisfactory quality of the model. For TMTB cutoff scores were: in the whole sample > 63 s (Se 70%, Sp 68%); Group 1 > 61 s (Se 91%, Sp 64%); Group 2 > 58,5 s (Se 73%, Sp 60%); Group 3 > 58,0 s (Se 83%, Sp 83%); Group 5 > 71,5 s (Se 90%, Sp 80%); Group 4 had an unsatisfactory quality of the model. For DSST cutoff scores were: in the whole sample < 58.5 points (Se 74%, Sp 63%); Group 2 < 59,5 points (Se 71%, Sp 67%); Group 3 < 60,5 points (Se 78%, Sp 83%); Group 4 < 53,5 points (Se 68%, Sp 72%); groups 1 and 5 had an unsatisfactory quality of the model. < 6,5 слів (Se 70%, Sp 75%).
 Conclusions. Patients with an active episode of MDD demonstrate as subjective as objective cognitive impairments as compared to HC. Cognitive dysfunctioninthe Ukrainiancohortof MDDpatients is characterized by mild impairments in working memory; moderate impairments in alternating attention; and moderate impairments in executive functioning. PDQ5 and neuropsychological tests, such as RAVLT (subtests for immediate recall, proactive and retroactive interference), TMTB, and DSST show from excellent to good diagnostic value for separating patients with MDD from HC. PDQ5, RAVLT, TMTB, and DSST and obtained ageadjusted cutoffs of those tests could be used by clinicians in everyday practice as a method to secure a more valid assessment of cognitive function in MDD patients.