تاثیر حواس پرتی خنثی و عاطفی منفی بر روی عملکرد حافظه و ارتباط آن با شکایت حافظه در افسردگی اساسی

Abstract Patients with major depression (MDD) often report relevant cognitive problems in everyday life while performance in standardised neuropsychological tests is not severely disturbed. This discrepancy may partly be due to the differences between the demands of everyday life with the presence of emotionally relevant distractors and standardised neuropsychological settings without those distractors. In the present study, we hypothesise that patients with major depression (MDD) show an increased distractibility towards emotionally negative stimuli and that this distractibility is related to complaints of cognitive functioning in everyday life. Thirty MDD patients and 48 healthy participants performed our recently developed learning paradigm with neutrally and negatively valenced distraction as well as without distraction. Both groups also performed a neuropsychological test battery as well as self- and observer ratings of impairments in memory and attention in every day life. In the MDD sample, cognitive impairments were reported by the patients and their relatives but were not found in the neuropsychological tests. We found a trend towards a poorer memory performance with negatively valenced distraction in the MDD sample when compared to the performance of healthy subjects. However, this impairment was not related to the self- and observer ratings. This result may be due to the fact that the distractors were not personally relevant to the subjects whereas everyday life implies such distractors. Further research is needed to explore everyday cognitive functioning of patients with MDD.

1. Introduction Neuropsychological impairment is regarded as a key feature of major depression (MDD) (World Health Organization, 1991 and American Psychiatric Association, 1994). Ebmeier et al. (2006) concluded that these impairments are responsible for much of the MDD morbidity and have important clinical implications including strategies for treatment. Indeed, self-reports and reports by patients' relatives indicate severe disturbances of memory and attention in MDD (Lahr et al., 2007). However, it is surprising that severe impairments are not found in neuropsychological studies using standardised tests. In meta-analyses, moderate deficits with no more than one standard deviation below the means of healthy control subjects were reported — specifically in the areas of executive functions, memory and attention (Christensen et al., 1997 and Veiel, 1997). Furthermore, subjective complaints often do not predict failure in objective tests (Mowla et al., 2008). It is still a matter of debate why the reported severe dysfunctions are not confirmed with neuropsychological tests. It seems possible that patients overestimate objective deficits. An alternate explanation focusses on the differences between neuropsychological settings and the demands of everyday life. In a previous study of our group (Lahr et al., 2007), the patients' self- reported deficits of memory and attention exceeded their relatives' ratings. This finding supports the notion that patients indeed may exaggerate their deficits. However, we also found the ratings of the patients' relatives clearly exceeded the impairments found by neuropsychological test measures. This result argues for the possibility that differences between subjective and objective measures may be explained by the differences between neuropsychological settings and the demands of everyday life. One striking difference between everyday life and neuropsychological settings is the presence of distracting stimuli (Potter and Hartman, 2006). Whereas neuropsychological tests usually have to be presented under laboratory conditions (Lezak, 1995), distracting noise or other conflicting information are usually present in everyday life. Consequently, the success of purposive information processing also depends on the ability to inhibit interfering information. In accordance with this assumption, Potter and Hartman (2006) hypothesised that poorer performance on tests of executive functions, such as inhibition, would be associated with higher levels of memory complaints in everyday life. They reported that decreased response inhibition in depressed patients as assessed by means of the Stroop-paradigm predicted memory complaints, even after controlling for memory performance. The authors concluded that the memory problems reported by depressed patients are possibly related to difficulty inhibiting distraction from irrelevant information in everyday life. With regard to inhibition, impaired performance in the Stroop task was found in MDD patients (Schatzberg et al., 2000 and Harvey et al., 2004). However, many studies did not find a general inhibition problem in depressed patients (Degl'Innocenti et al., 1998, Murphy et al., 1999 and Markela-Lerenc et al., 2006). There is some evidence that cognitive changes in depression are more obvious with affectively meaningful stimuli (Erickson et al., 2005). Lau et al. (2007) concluded that inhibitory dysfunction in MDD patients is most likely valence-specific. In their study, patients with MDD, healthy controls and a clinical control group (non-depressed, anxious patients) were compared with regard to a measure of cognitive inhibition (Prose Distraction Task, PDT), and a measure of motor response inhibition (Stop-Signal Task, SST). Both tasks were modified in order to present emotionally valenced semantic stimuli. No group differences were revealed in the SST but participants with MDD demonstrated impairments on the PDT, and these impairments were most pronounced for negatively valenced adjectives. Similarly, Goeleven et al. (2006) found that depressed patients showed a specific problem inhibiting negative information in a priming task with pictures of sad and happy facial expressions. A specific inhibition problem with negative distraction is in accordance with an attentional bias towards negatively valenced information in MDD patients. This bias may also lead to an enhanced memory for negatively valenced emotional material (Leppanen, 2006). In the present study we hypothesise, first, that the everyday problems reported by MDD patients and their relatives exceed their impairments in neuropsychological tests and that self-reports indicate greater disturbances than their relatives' ratings. Second, we are interested in the investigation of inhibitory functions in MDD patients. We hypothesise that patients with MDD when compared to healthy control subjects show poorer performance in a recently developed learning paradigm (Beblo et al., 2006) when emotionally negative distractors are presented. By contrast, we expect the MDD patients' performance to be unimpaired when learning stimuli are presented with neutral or no distraction. Third, we hypothesised that the problems of MDD patients with inhibiting negatively valenced, irrelevant information is related to self-reported cognitive impairments in every day life.

. Results 3.1. Sample characteristics and clinical data Mean age of the patients was 37.7 years (range: 18–69, S.D. = 12.9), basic school education was 11.3 years (range: 9–13, S.D. = 1.5) on average. Fifteen out of the 30 patients were women. With regard to these demographic variables the healthy subjects were similar to the patients (mean age: 36.8, range: 21–62, S.D. = 11.0; mean education: 11.4 years, range: 10–13, S.D. = 1.3; 25 out of 48 women). Groups did not differ with regard to intelligence as indicated by means of the LPS-4 (Horn, 1983). Results (Table 3) indicate an average level of intelligence (slightly above 100) in both groups. An average level of intelligence corresponds to the subject's education level. Fifteen patients also showed anxiety disorders (two patients showed two different anxiety disorders) — agoraphobia/panic disorder (n = 7), posttraumatic stress disorder (n = 5), specific phobias (n = 2), generalised anxiety disorder (n = 2) and other anxiety/somatisation symptoms (n = 1). With regard to Axis II co-morbidity, eight patients showed personality disorders such as avoidant personality disorder (n = 5), obsessive–compulsive personality disorder (n = 2) and narcissistic personality disorder (n = 1). The mean BDI-score in the patients was 22 (S.D. = 7.7), indicating a moderate level of depression. The control group did not show any signs of depressive symptoms (all scored below 10 in the BDI). Nineteen patients received antidepressive medication (selective serotonine reuptake inhibitors: n = 15, tricyclics: n = 2; others: n = 2) and seven patients received neuroleptic medication. One patient received a mood stabiliser (Lamictal). 3.2. Inhibition With regard to the learning performance with neutral and emotional distraction (based on a modified version (Beblo et al., 2006) of the Auditory Verbal Learning Test (Rey, 1964 and Spreen and Strauss, 1998) as described above), we found no main effect due to group. Patients and healthy subjects showed similar overall performance (Table 1). Data show a main effect due to condition (F(2,75) = 49, P < 0.001). Post hoc analyses indicate that the performance in the standard condition was superior to the performance in the condition with neutral distraction (t(77) = 6.9, P < 0.001). The performance in the condition with neutral distraction seemed to be superior to the condition with negative distraction (t(77) = 2.0, P = 0.045); however, after alpha correction this result failed to be significant. Not surprisingly, there was also a main effect due to the learning trials (F(2,75) = 567, P < 0.001) with the best performance in trial 3 (trial 3 > trial 2: t(77) = 15.9, P < 0.001) and the lowest score in trial 1 (trial 2 > trial 1: t(77) = 28.3, P < 0.001). In addition, there was a condition × trial interaction (F(4,73) = 3.8, P = 0.008). Performance differed between the conditions for all three trials (trial 1: F(2,76) = 51, P < 0.001; trial 2: F(2,76) = 33, P < 0.001; trial 3: F(2,76) = 25, P < 0.001) with F-values indicating the differences to be most prominent in the first trial and at least prominent in the last. With regard to the group × condition interaction we found a trend (F(2,75) = 2,4, P = 0.097) indicating a disproportionally inferior performance by the depressed patients in the condition with negative distraction ( Fig. 1). For explorative purposes, we performed post hoc analyses and found a trend that the depressed patients performed worse only under negative distraction (t>(76) = 2,2, P = 0.025, n.s. after alpha correction). Table 1. Learning performance with neutral and emotional distraction (number of correctly recalled targets) based on a modified version (Beblo et al., 2006) of the Auditory Verbal Learning Test (Rey, 1964 and Spreen and Strauss, 1998) in MDD patients and control subjects (CON). Standard Condition Distraction Condition, Neutral Valence Distraction Condition, Negative Valence MDD (n = 30) CON (n = 48) MDD (n = 30) CON (n = 48) MDD (n = 30) CON (n = 48) M S.D. M S.D. M S.D. M S.D. M S.D. M S.D. 1st trial 7.3 1.8 7.7 2.1 5.5 2.3 5.6 2.5 4.7 1.8 5.4 2.3 2nd trial 10.5 2.2 11.3 2.2 9.8 3.0 9.6 2.2 8.0 2.5 9.4 2.5 3rd trial 12.3 1.8 13.1 1.7 11.0 2.1 11.7 2.5 10.5 2.5 11.8 2.7 Sum trials 1–3 30.1 5.2 32.0 5.3 26.3 6.1 26.9 6.5 23.2 6.0 26.6 6.7 M: Mean; S.D.: standard deviation. Main effect for condition (F(2,75) = 49, P < 0.001), post hoct-tests: standard > neutral (t(77) = 6.9, P < 0.001), neutral > negative (t(77) = 2.0, P = 0.045, n.s. after alpha correction). Main effect for trial (F(2,75) = 567, P < 0.001), post hoct-tests: trial 3 > trial 2 (t(77) = 15.9, P < 0.001), trial 2 > trial 1 (t(77) = 28.3, P < 0.001). No main effect for group. Interaction condition × trial (F(4,73) = 3.8, P = 0.008), post hoc analysis revealed differences between the conditions for all three trials (trial 1: F(2,76) = 51, P < 0.001; trial 2: F(2,76) = 33, P < 0.001; trial 3: F(2,76) = 25, P < 0.001), F-values indicating the differences to be most prominent in the first trial. Interaction group × condition (F(2,75) = 2,4, P = 0.097), explorative post hoc analysis revealed a trend for a decreased performance of the depressed patients only under negative distraction (t(76) = 2,2, P = 0.025, n.s. after alpha correction). No further effects. Table options Learning performance with neutral and emotional distraction (mean sum scores: ... Fig. 1. Learning performance with neutral and emotional distraction (mean sum scores: trials 1–3 and 95% confidence intervals) based on a modified version (Beblo et al., 2006) of the Auditory Verbal Learning Test (Rey, 1964 and Spreen and Strauss, 1998) in MDD patients and control subjects. Figure options Regarding working memory performance, we compared the digit and block spans forward with the condition of (emotionally neutral) distraction (the Digit and Block Suppression Test, Table 2). There were no main effects for group (MDD patients vs. control subjects) or modality (verbal vs. spatial). The performance of patients and control subjects in the verbal and spatial modality were similar. We did not compare the conditions with and without distraction because the raw scores of the spans were not comparable. However, previous studies indicate that spans with distraction are clearly shorter than spans without distraction (Beblo et al., 2004). Furthermore, there were no interactions between the variables (group, condition and modality). Table 2. Working memory performance (number of correctly recalled targets) in MDD patients and control subjects (CON). Verbal modality Spatial modality Forward DST Forward BST MDD (n = 30) 7,1 (1,5) 10,4 (3,7) 8,1 (2,4) 9,4 (3,3) CON (n = 48) 7,5 (1,9) 10,5 (3,3) 8,1 (1,6) 9,6 (3,5) DST: Digit Suppression Test, BST: Block Suppression Test. No main effects for modality (verbal vs spatial) or group (MDD vs CON), no interaction. Table options 3.3. Further neuropsychological assessment In addition to inhibition, we also investigated memory, attention, executive functions and visuo–spatial abilities in the study samples. We did not find any significant group differences with regard to these functions (Table 3). Effect sizes also indicate no large group effects (e > 0.6, Bortz, 2005). The maximum effect size was 0.5 in verbal learning, semantic fluency and visual scanning/selective attention. Some further test parameters showed medium effect sizes as well (Go/Nogo, tracking, flexibility), whereas the patients' performances in the other tests were almost equal to the control group. Table 3. Neuropsychological data and data of the self and observer ratings of memory and attentional performance in MDD patients and healthy control subjects (t-tests and effect sizes). MDD patients (n = 301) M (S.D.) Controls (n = 481) M (S.D.) T P e Test performance Logical Memory I (WMS-R) 29.4 (7.3) 32.9 (6.3) 2.2 ns 0.5 Logical Memory II (WMS-R) 25.4 (8.1) 28.6 (7.4) 1.7 ns 0.4 Complex Figure Test (copy) 34.1 (1.9) 34.1 (1.9) 0 ns 0 Complex Figure Test (recall) 19.1 (7.6) 18.6 (7.0) 0.2 ns −0.1 Trail Making Test (A, sec) 30.6 (10.2) 26.6 (9.0) 1.7 ns 0.4 Trail Making Test (B, s) 73.4 (28.5) 62.0 (26.3) 1.7 ns 0.4 TAP — Alertness (s) 248 (23.1) 246 (30.6) 0.2 ns 0 TAP — Divided Attention (s) 691 (93.5) 685 (74.1) 0.3 ns 0.1 TAP — Go/Nogo (s) 445 (70.4) 416 (82.8) 1.4 ns 0.4 FAIR (sum score) 165 (51) 190 (47) 2.2 ns 0.5 FAS — Test (verbal fluency, and lexical) 36.5 (10.6) 37 (10.6) 0.2 ns 0 Animals (verbal fluency, and semantic) 23.3 (5.5) 26 (5.9) 1.9 ns 0.5 LPS-4 (intelligence) 26.4 (4.8)2 27.5 (5.1)2 0.9 ns 0.2 Self and observer ratings EMQ (self, n = 20/38)) 93 (38.3) 51 (15.2) 4.7 < 0.001⁎ 1.9 FEDA (self, n = 21/ 40)) 78 (16.4) 119 (9.7) 10.5 < 0.001⁎ 3.3 FEDA (observer, n = 13/37) 95 (21.3) 121 (10.8) 4.4 < 0.001⁎ 1.7 1 Exeptions are indicated. 2 Correspond to an IQ slightly above 100. ⁎ significant (P < 0.017). WMS-R: Wechsler Memory Scale, revised, TAP: Test-Battery of Attentional Performance Testbatterie zur Aufmersamkeitspruefung), FAIR: Frankfurt Attention Inventory (Frankfurter Aufmerksamkeits-Inventar), EMQ: Everyday Memory Questionnaire (higher score indicating poorer performance), FEDA: Questionnaire of Experienced Attention (Fragebogen zur erlebten Aufmerksamkeit; lower score indicating poorer performance), LPS: Leistungspruefsystem. Table options 3.4. Self- and observer ratings of memory and attention performance There were highly significant group differences between MDD patients and control subjects (P < 0.001) in the self- and observer ratings of memory and attention performance ( Table 3) which indicated impairments in MDD patients. Effect sizes (1.7–3.3) were large and outperformed the effect sizes we found with regard to neuropsychological testing. Please note that the FEDA was administered for self- and observer rating. Whereas both ratings were almost equal in the control group, patients' self- and observer ratings differed significantly (t(11) = 3.2, P = 0.007), indicating a more pronounced deficit with self-rating. 3.5. Prediction of the self- and observer ratings of memory and attention performance It was not possible to predict the depressive patients' self- and observer-based ratings of everyday cognitive functioning (criteria: EMQ [R2(3, 16) = 0.114, P = 0.57], FEDA self-rating [R2(3, 17) = 0.192, P = 0.29], FEDA observer-rating [R2(3, 9) = 0.358, P = 0.25]) by their performance in the learning paradigm (predictors: learning without distraction, learning with neutral distraction, learning with negative distraction [sum scores]). In addition, post hoc analyses did not reveal any significant interrelation between self- and observer-based ratings on the one hand, and parameters of the learning performance on the other (including learning without distraction, learning with neutral distraction, learning with negative distraction, difference score 1: learning without distraction minus learning with neutral distraction, difference score 2: learning without distraction minus learning with negative distraction).