Influence of age and executive functioning on verbal memory of inpatients with depression

Journal of Affective Disorders 68 (2002) 261–271 www.elsevier.com / locate / jad

Research report

Influence of age and executive functioning on verbal memory of inpatients with depression Philippe Fossati

a,b ,

*, Franc¸oise Coyette c , Anne-Marie Ergis a,d , Jean-Franc¸ois Allilaire a

a

´ ` Hospital, 47 Boulevard de l’ Hopital ˆ Department of Psychiatry, Salpetriere , 75651 Paris Cedex 13, France b Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada M6 A 2 E1 c Neuropsychological Rehabilitation Center, Saint-Luc University Clinic, 10 Avenue Hippocrate, 1200 Bruxelles, Belgium d Psychology Department, Charles de Gaulle-Lille, University, 59653 Villeneuve d’ Ascq, France Received 26 April 2000; accepted 24 October 2000

Abstract Background: Despite many studies demonstrating memory and executive impairments in young and old depressed patients, the relationships between age, executive functioning and memory have not been evaluated in depression. The aim of this study was to investigate if older patients were more vulnerable than younger patients to the impact of depression on memory and if the differences between young and old depressed could be related to executive functioning. Methods: Forty-nine inpatients, with unipolar and bipolar depression, ranging in age from 19 to 72 years were compared with 70 controls on a verbal memory task. Age cut-off of 45 years was used as a categorical variable to divide subjects into subgroups. A subset of patients (n 5 41) was also evaluated with the modified version of the Wisconsin Card Sorting Test and separated into a non-dysexecutive group and a group of patients with mild-executive impairment. Results: Depressed patients exhibited memory deficits with a pattern of memory failure — impaired free recall and normal cued recall and recognition — interpreted as a retrieval problem. Both age and executive function influenced memory performance in depression, however neither group 3 age interaction nor age 3 executive status interaction were significant. Multiple regression analysis showed that free recall scores were related to age and psychomotor retardation in depressed patients. Conclusion: Age and executive functioning have different influences on the memory performance of depressed patients. Our findings support an ‘executive memory decline hypothesis’ in young as well as old depressed patients. The memory deficits in depression may be associated with both trait and state factors and raise questions about the long-term cognitive functioning of patients with recurrent affective disorders.  2002 Elsevier Science B.V. All rights reserved. Keywords: Age; Executive functioning; Verbal memory; Inpatients; Depression

*Corresponding author. Tel.: 1 33-1-4216-1233; fax: 1 33-14216-1236. E-mail address: [email protected] (P. Fossati).

1. Introduction Many studies have demonstrated memory impairments in depressed patients (Golinkoff and Sweeney,

0165-0327 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0165-0327( 00 )00362-1

262

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

1989; Bornstein et al., 1991; Ilsley et al., 1995). However, not all depressed patients show memory deficits and there is marked disagreement about the reasons why certain patients develop more severe memory problems than others (Jonhson and Magaro, 1987; Wolfe et al., 1987; Kessing et al., 1996; Tham et al., 1997; Veiel, 1997; Austin et al., 1999). One important area of controversy in this domain is the relationship between the effects of age and the effects of depression on memory. Many clinicians believe that older patients are more vulnerable to the cognitive impact of depression than younger patients. Although some authors have suggested that the effects of age and depression interact to induce a synergistic decline in memory functions (King et al., 1998), others studies have found few differences between young and old depressed and a lack of age 3 depression interaction regarding memory performance (Cavanaugh and Wettstein, 1983; Niederehe, 1986; Niederehe and Yoder, 1989; Lyness et al., 1994). Moreover, a recent meta-analysis have unexpectedly demonstrated that depression had a greater impact on memory in younger than older patients (Burt et al., 1995). These discrepancies in studies may reflect a number of methodological issues including differences in the definition of the severity of depression, selected tasks for studies, methods of testing for age effects, the influence of other cognitive impairments on memory performance and the heterogeneity of normal cognitive aging. Anatomical changes in the aging human brain predominate in the temporal and frontal lobes, and consequently the typical pattern of neuropsychological change in normal aging involves declines in memory and executive functioning (Craik et al., 1990; Daum et al., 1996; Salat et al., 1999; Bryan and Luszcz, 2000). Two main hypotheses about the age-related memory decline in healthy subjects have been formulated: the processing resource hypothesis and the executive decline hypothesis (Bryan et al., 1999). The executive decline hypothesis of cognitive aging comes from a neuropsychological perspective and suggests that the age-related executive decline may be responsible for observed age differences in memory performance (Troyer et al., 1994; Crawford et al., 2000). The executive functions involve a set of cognitive behaviors — including problem solving,

dealing with novelty, selecting strategies and monitoring performance — that control and integrate other cognitive activities such as episodic memory. While impairments on executive function have been extensively demonstrated in young (Fossati et al., 1999) and older depressed patients (Beats et al., 1996; Palmer et al., 1996), to our knowledge the executive decline hypothesis and the relationships between age, executive functioning and memory have not been directly evaluated in depression. The aim of the present study was to test this hypothesis by investigating the influence of age on memory performance of hospitalized patients with depression and the extent to which performance on tests of executive function mediate the relationship between age, depression and memory scores. More specifically this study addressed two specific questions: (1) are the older patients more vulnerable to the cognitive effects of depression than younger patients? (2) Are these eventual differences on memory performances between young and old patients related to executive functioning?

2. Methods

2.1. Subjects Forty-nine patients, ranging in age from 19 to 72 years, admitted to the Department of psychiatry at ´ ´ ` hospital for a major depressive the Pitie-Salpetriere episode without psychotic features, gave their consent to participate in this study. Some of the young patients were included in a previous study of the executive functioning in unipolar depression (Fossati et al., 1999). Two psychiatrists made diagnoses using a checklist encompassing DSM-IV criteria for major depression. No structured interview was used in this study. Each subject completed physical and neurological examination, laboratory examinations, including complete blood count, electrolyte and thyroid function tests. In addition, 34 patients (13 in young group and 21 in old group) underwent a brain CT scan. Patients with concurrent Axis II diagnoses or substance abuse were excluded, as were those with brain diseases likely to affect cognition, e.g., dementia, Parkinson’s disease, stroke, head injury. Patients were not tested within 12 months of receiv-

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

ing ECT and were recruited within 1 week of admission. Two patients (66 and 59 years old) had received ECT more than 1 year previously. Forty-four of the 49 patients were on antidepressant medication. All patients were tested within 2 weeks of first receiving or starting antidepressant medication. Twenty-eight of the 49 patients were taking tricyclics with anticholinergic properties. Thirty-two patients were also on benzodiazepines, and 20 of the 49 patients were on neuroleptics used as hypnotic. Severity of depression was assessed with the Montgomery–Asberg depression rating scale (MADRS, Montgomery and Asberg, 1979) and ´ ` the Salpetriere Retardation Rating Scale (SRRS, Widlocher, 1983). Only subjects with scores of 20 or more on the two scales were included in the study. Thirteen patients (8M / 5F) met DSM-IV criteria for bipolar I disorder and 36 patients (8M / 28F) met criteria for unipolar disorder. There were no differences between unipolar patients (UP) and bipolar patients (BP) in years of education, scores on MADRS and ERD scales and mean duration of index episode (t-tests for unequal variances, a 5 0.05). Bipolar patients had experienced significantly more affective episodes than UP (Mann–Whitney U 5 89.5, P , 0.001). The mean duration of illness and the mean number of hospitalizations were higher in BP than in UP (respectively, U 5 136, P , 0.05 and U 5 158, P , 0.05). Despite these differences between UP and BP regarding the course of the illness, BP performed as the same level than UP for all the memory scores (data no reported here and available on request). Therefore, BP and UP were collapsed into a single group for comparisons with controls. Seventy healthy non-depressed subjects from the community and hospital staff, ranging in age from 19 to 74 years, composed the control sample. The size of the sample was selected to match the sex, age, years of education and the overall number of patients in the study and to provide greater stability in statistical comparisons. All controls were screened for presence or history of psychiatric, substance abuse or neurological disorder. The study was performed in two parts. In one part, we separated the group of patients and controls into four subgroups according to an age cut-off of 45 years. The young depressed patients (n 5 28) did not differ from older depressed patients (n 5 21) on the

263

treated / non-treated ratio (respectively, for benzodiazepines Pearson x 2 5 0.03, ns; antidepressant medication x 2 5 0.02, ns). Older patients received less anticholinergic drugs than younger patients ( x 2 5 3.87, df 5 1, P , 0.05). In part two of this study, a subset of patients (n 5 25 in the young group and n 5 16 in the old group; 31 unipolar patients and 10 bipolar patients) were tested with the modified version of the Wisconsin Card Sorting Test (Nelson, 1976) as part of a comprehensive battery evaluating executive function in depression. The proportion of patients with treatment was the same across age groups (anticholinergic drugs, Pearson’s x 2 5 3.27, df 5 1, ns; benzodiazepines, x 2 5 0.45, ns; antidepressant, x 2 5 0.22, ns).

2.2. Neuropsychological evaluation 2.2.1. Memory task: free and cued selective reminding procedure (adapted by Deweer and Van der Linden from Grober et al., 1988) This verbal learning test assesses episodic memory under conditions that control for cognitive processing during encoding. The procedure is described in detail elsewhere (see Grober et al., 1988; Fossati et al., 1999). In short, participants were asked to search a card containing four verbal items, point to and name each item (e.g., dentist) when its category cue (e.g., profession) was given verbally. After the search was completed, the card was removed, and immediately cued recall of those items was tested. Following this trial of cued recall, a next set of four items was presented for the search procedure, continuing until all 16 items were identified and retrieved. The study phase was followed by three tests trials of free recall preceded by 30 s of interference by counting backward. Participants were instructed to recall as many items as they could. After each free recall was exhausted, cued recall of those items that were not retrieved was tested by presenting the category cue. The participants were verbally reminded of any item they failed to retrieve by cued recall. Following the last free and cued recall trials, recognition memory for the items was tested by presenting one word at a time for the subject to decide whether it was on the list.

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

264

2.2.2. Executive task: Wisconsin Card Sorting Test (modified version, Nelson, 1976) Subjects were asked to sort 48 cards on the basis of three possible categories (shape, number, color). With this version of the task, subjects were asked to change the sorting principle to another category after six consecutive correct responses. 2.3. Statistical analysis Multivariate analysis of variance (MANOVA) and repeated measures analysis of variance were conducted on neuropsychological variables of verbal learning with a 2 (depressed, controls) 3 2 (older, younger) factorial model. Significant results at P , 0.05 level were followed by F-univariate tests and specific comparisons of means between specific subjects groups using Bonferroni correction for multiple comparisons. In the second part of the study, the role of executive functioning on memory was evaluated by separating the subjects into ‘dysexecutive’ and ‘non-dysexecutive’ groups. Two (young, old) 3 2 (dysexecutive, non-dysexecutive) factorial ANOVAS were performed on the data. When the assumption of abnormality was not met,

the non-parametric Mann–Whitney was used. To control for the role of treatment, analyses of covariance were performed using treatments as dichotomic covariates. The relationships between cognitive and clinical measures were then examined for depressed group by Pearson’s correlation coefficient. Finally, multiple linear regression was used to analyze the predictive value of course of illness, age, executive level and severity of depression on memory performance. The data were analyzed using SPSS 10.0 for Windows NT 4.0.

3. Results

3.1. Influence of age on memory performance To evaluate the role of age on memory performance we divided the patients and controls groups into four subgroups according to the age cut-off of 45 years as previously defined. The demographic and clinical characteristics of patients and controls are summarized in Table 1. Within the depressed groups, a multivariate analysis of variance with age as between subject factor and MADRS, SRRS scores,

Table 1 Demographic and clinical characteristics of patients and controls (values given are means with standard deviation in parentheses) Depressed

Sex ratio b Age (years) Years of education b MADRS score SRRS score Age at onset of first episode Mean duration of illness (years) Mean duration of index episode (months) Number of depressive episodes Number of hospitalizations a b

Statistics depressed a

Control

Younger (n 5 28)

Older (n 5 21)

Younger (n 5 27)

Older (n 5 43)

6M / 22F 32.4 (7.35) 13.1 (2.67)

10M / 11F 57.2 (8.91) 12.5 (3.31)

14M / 13F 29.9 (6.30) 13.7 (2.42)

17M / 26F 57.72 (6.83) 12.0 (1.98)

25.7 (5.65) 23.0 (5.38) 25.3 (6.71)

25.3 (5.11) 23.38 (5.77) 45.7 (13.21)

F

P

0.07 0.05 48.45

ns ns , 0.001

7.5 (8.37)

11.7 (11.5)

2.12

ns

6.5 (5.77)

5.2 (5.69)

0.61

ns

1.9 (0.99)

2.5 (1.69)

2.47

ns

1.6 (1.13)

1.7 (1.18)

0.10

ns

Univariate comparisons between young and old depressed with P adjusted according to the Bonferroni’s method. Comparisons between controls and depressed; male / female ratio, x 2 5 6.13, df 5 3, ns; years of education, F(3,115) 5 2.57, ns.

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

and clinical features as dependent variables showed a significant effect of age (F(7,41) 5 13.04, P , 0.001). This effect was due to the age at first depressive episode (see Table 1), significantly being greater in older than younger patients (F(1,47) 5 48.45, P , 0.001; Bonferroni). Compared to the controls depressed patients showed memory deficits on free recall. However, depressed patients were not more vulnerable than normal controls to the effects of aging. An analysis of variance on the three free recall trials of the memory test with diagnosis and age as between subject factors, and study-test trials as within-subject factor revealed: (1) an effect for diagnosis (F(1,115) 5 16.25, P , 0.001); (2) an effect for age (F(1,115) 5 17.90, P , 0.001); and (3) an effect for trials (F(2,114) 5 89.52, P , 0.001). However, no interaction of diagnosis 3 age was observed (F(1,115) 5 1,84; ns). Likewise, interactions of age 3 trials (F(2,114) 5 2.12, ns), diagnosis 3 trials (F(2,114) 5 0.44, ns) and triple interaction of age 3 diagnosis 3 trials (F(2,114) 5 0.28, ns) were not significant. Pairwise comparisons showed that controls performed better than depressed subjects (P , 0.001; Bonferroni) and young subjects performed better than older subjects (P , 0.001; Bonferroni). Since it is generally known that both benzodiazepine and anticholinergic drugs can impair memory,

265

we repeated the previous analysis after adjusting for benzodiazepine as a covariate. Results were unchanged with main effects for trials (F(2,113) 5 33.0, P , 0.001), diagnosis (F(1,114) 5 7.12, P , 0.01) and age (F(1,114) 5 17.72, P , 0.002). Again, the interaction of age 3 group was not significant. Within young and old patients, those received anticholinergic drugs did not differ from those who did not receive these anticholinergic medications (F(1, 45) 5 1.96, ns). Patients recalled fewer verbal items than normal controls on trials 1, 2, and 3. Follow-up univariatetests for each trial are summarized in Table 2. Likewise, older subjects recalled fewer items than controls on each trial (Table 2). Compared to other groups, patients and older subjects showed normal improvement across the three trials. Patients and controls recalled most of the remaining items by cued recall, resulting in absence of difference between groups according to age or diagnosis (see Table 2). Finally a 2 (depressed, controls) 3 2 (young, old) analysis of variance was conducted on the consistent free recall index (the percentage of items consistently recalled across the three free recall trials). This index was not calculated in eight old controls due to loss of data. There was a main effect of group (F(1,107) 5 7.89; P , 0.01), age (F(1,107) 5 12.88; P , 0.001) and no significant

Table 2 Subjects’ performance on the free and cued selective reminding task according to age and diagnosis (mean (standard deviation))a

Free recall 1 Free recall 2 Free recall 3 Total recall 1 Total recall 2 Total recall 3 SFR STR CFRb CTR Index CFR% Index SSC

Group 1 D, , 45 years (n 5 28)

Group 2 D, . 45 years (n 5 21)

Group 3 C, , 45 years (n 5 27)

Group 4 C, . 45 years (n 5 43)

9.9 (2.09) 12.0 (1.97) 12.4 (1.99) 15.4 (0.83) 15.6 (0.73) 15.85 (0.44) 34.4 (5.23) 46.9 (1.58) 6.2 (2.47) 15 (1.44) 52.6 (16.92) 93.6 (7.38)

8 (2.58) 9.6 (2.37) 10.9 (2.03) 14.9 (1.24) 15.5 (0.74) 15.6 (0.66) 28.6 (5.86) 46.1 (2.30) 3.4 (2.08) 14.5 (1.66) 33.9 (16.26) 91.4 (10.85)

10.9 12.7 13.5 15.1 15.7 15.8 37.2 46.7 7.2 15.0 56.7 90.8

10.0 (2.18) 11.4 (2.51) 12.7 (2.23) 15.2 (1.14) 15.5 (0.95) 15.8 (0.39) 34.2 (5.97) 46.7 (2.17) 6.20 (3.12) 14.97 (1.83) 49.8 (20.73) 92.2 (11.39)

(2.17) (1.72) (1.62) (1.27) (0.65) (0.36) (4.68) (1.71) (3.06) (1.41) (18.40) (10.68)

Statistics: groups

Statistics: age

F

P

F

P

13.07 8.50 14.56 U * 5 1653.0 U 5 1653.5 U 5 1588.5 16.25 U 5 1582.5 12.09 U 5 1462.5 7.89 U 5 1712.0

, 0.001 , 0.005 , 0.001 ns ns ns , 0.001 ns , 0.002 ns , 0.007 ns

11.33 18.78 9.51 U 5 1685.0 U 5 1659.5 U 5 1703.0 17.90 U 5 1719.5 13.23 U 5 1659.5 12.88 U 5 1717.5

, 0.001 , 0.001 , 0.005 ns ns ns , 0.001 ns , 0.001 ns , 0.001 ns

a D, depressed; C, controls; U, Mann–Whitney test; SFR, sum of free recall; STR, sum of total recall; CFR, consistence of free recall; CTR, consistence of total recall; Index CFR, consistence of free recall %; SSC, sensitivity to semantic cueing. b Controls . 45 years, n 5 35.

266

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

interaction of group 3 age (F(1,107) 5 2.71, ns). Patients and older subjects did not retrieve the same items across the trials. Recognition was intact in patients and controls and showed a ceiling effect (Table 2). Likewise, patients and controls (young and old) did not differ for the number of intrusive errors. These results suggest a pattern of memory failure in depression — impaired free recall and normal cued recall and recognition — which appears to mainly reflect a retrieval deficit. Depressed patients in this study were not more vulnerable than normal controls to the effects of age as revealed by the non-significant interaction age 3 group.

3.2. Relationships between memory scores and clinical variables In order to check for an association between cognitive measures and clinical variables we performed correlational analyses in depressed patients. The retardation score, age and the number of previous depressive episodes were significantly associated with the free recall score on trial 1 (respectively, r 5 2 0.28, P , 0.05; r 5 2 0.29, P , 0.002; r 5 2 0.47, P , 0.05) and only a weak correlation was found between this free recall score on trial 1 and the number of hospitalizations (r 5 2 0.27, P 5 0.055). A multiple regression analysis was conducted to evaluate the relative contribution of course illness, age and severity of index depressive episode to memory performance. Age, SRRS score, number of previous depressive episodes and number of hospitalizations were entered into the regression model in a stepwise manner. A model including two variables, age and SRRS score, yielded significant contributions to the prediction of free recall 1 scores (F(2,46) 5 5.59, P , 0.008). The model explained about 20% of the variance (R 2 5 0.19). Incremental change in variance attributed to the addition of the SRRS to the equation was equal to 8.2%.

3.3. Influence of executive functioning and age on memory performance In the part two of this study, we evaluated a subset of patients (25 young and 16 older patients) with the modified version of the Wisconsin Card Sorting Test

(MCST). Clinical features of the patients are summarized in Table 3. Young and old patients were matched for the level of executive function assessed by problem solving capacity (Table 3). As there is no standardized definition of a dysexecutive syndrome (Rabbit, 1997), the median of the number of categories completed on the MCST was used to classify subjects into ‘dysexecutive’ and ‘non-dysexecutive’ groups. All the patients in the non-dysexecutive groups completed six categories on the MCST (Table 4). There is no normative data with the MCST for young subjects under 45 years (see De Zubicaray and Ashton, 1996). Lineweaver et al. (1999) however, recently provided age and education corrections for MCST scores for subjects above 45 years. In regard to these normative data, 37.5% of the old dysexecutive patients had mild executive impairment and 25% had scores in the low-average range. 100% of non-dysexecutive old subjects had scores in and above the average range. Therefore, in the group as a whole the dysexecutive patients had only a mild executive impairment. As expected the mild-dysexecutive patients exhibited memory impairments on the free recall trials compared to the non-dysexecutive subjects. The ANOVA on the three free recall trials with age and executive status as between subject factors and study-test trials (3) as within-subject factor thus revealed effects for age (F(1,37) 5 6.13, P , 0.02) and dysexecutive status (F(1,37) 5 3.60, P 5 0.06). Again the interaction of age 3 executive status was not significant (F(1,37) 5 1.22, ns) suggesting that the differences on memory performance between old and young patients cannot be related to executive dysfunction. The follow-up univariate tests confirmed the effects of age and executive levels on memory performance (Table 4). Results were unchanged when treatments were used as covariates (respectively, ANCOVA with benzodiazepine as covariate, group effect F(1,36) 5 5.61, P , 0.03, and executive effect F(1,36) 5 3.75, P 5 0.06; ANCOVA with antidepressant as covariate, group effect, F(1,36) 5 6.73, and executive effect, F(1,36) 5 3.16, P 5 0.08). Pearson correlations between age, executive, memory and clinical scores appear in Table 5. As seen, age, SRRS and executive scores had different influences on the memory performances of depressed patients. The number of achieved categories on the MCST was significantly

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

267

Table 3 Demographic and clinical features of unipolar patients evaluated with the Modified Card Sorting Test (values given are means with standard deviation in parentheses)

Sex ratio Age (years) Clinical features: b Years of education MADRS score SRRS score Age at onset of first episode Mean duration of illness (years) Mean duration of index episode (months) Number of depressive episodes Number of hospitalizations MCST: b,c Categories Total errors (TE) Perseverative errors (PE) PE / TE (%) Verbal span: b Verbal forward Verbal backward

Young depressed (n 5 25)

Old depressed (n 5 16)

4 M / 21 F 32.6 (7.30)

7M / 9F 54.8 (8.32)

13.2 25.4 23.2 25.2

13.3 26.3 24.2 44.4

Statistics Fa

P

(3.42) (4.70) (4.97) (11.90)

10.05 0.01 0.27 0.34 44.23

, 0.001 ns ns ns , 0.001

7.1 (6.71)

10.2 (10.09)

1.46

ns

7.0 (5.86)

6.1 (6.07)

0.22

ns

1.9 (0.93)

2.6 (1.82)

2.38

ns

1.5 (0.96)

1.6 (1.20)

0.09

ns

1.45

ns

5.0 (1.70) 8.2 (8.34) 2.3 (3.46)

5.1 (0.98) 9.3 (6.70) 2.5 (2.33)

17.8 (19.98)

19.2 (14.99) 3.18

ns

6.1 (1.26) 4.6 (1.21)

6.6 (1.07) 4.2 (0.77)

(2.71) (5.48) (5.56) (6.55)

a

One-way analysis of variance, unless otherwise specified. Multivariate analysis of variance. c MCST, Modified version of the Wisconsin Card Sorting Test. b

related to free recall 2,3 whereas SRRS was mainly correlated with the first trial of free recall. Age had a weak correlation with all free recall trials.

4. Discussion In this study, depressed patients exhibit memory impairment compared to normal controls. Considering the relationships between depression and age we find parallel negative effects of age on depressed and controls’ performances of free recall. These results suggest that the older patients are not more vulnerable than younger patients to the impact of depression on memory.

The absence of group 3 age interaction on verbal memory task is consistent with previous studies (Niederehe, 1986; Niederehe and Yoder, 1989; King et al., 1993; Lyness et al., 1994; Tarbuck and Paykel, 1995) and is at odds with other studies (King et al., 1991, 1998). Several possible reasons may account for these divergent findings, such as for instance the severity and the type of memory task used in various studies (Lyness et al., 1994; King et al., 1998). Although the SSRS and MADRS scores were not particular high, young and young–old patients examined in this study required hospitalizations for their major depressive episode and all had a chronic history of recurrent affective episodes. Our results confirm that in-patient status might reduce the differ-

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

268

Table 4 Performance of unipolar depressed patients on memory task according to age and executive status (values given are means with standard deviation in parentheses) Depressed young Dys.ex (n 5 9) MCST: Categories Total errors (TE) Perseverative errors (PE) PE / TE (%) Free and cued selective Free recall 1 Free recall 2 Free recall 3 Sum of free recall Consistent free recall Index CFR % a

Depressed old

Non-dys.ex (n 5 16)

Dys.ex (n 5 8)

Statistics: age Non-dys.ex (n 5 8)

3.2 16.22

6 3.68

4.37 14.75

6 4

5.44 29.86

0.56 11.09

4.37 29.06

0.75 9.37

reminding task: 10.1 (2.61) 11.4 (2.65) 11.7 (2.77) 33.3 (7.56) 5.8 (3.4) 48.4 (23.9)

9.6 12.3 12.8 34.8 6.5 55.4

7.1 (3.83) 8.8 (2.85) 10.6 (2.26) 26.6 (7.83) 3.1 (2.85) 30.2 (21.8)

(1.92) (1.66) (1.40) (4.13) (1.9) (12.7)

9.0 11.2 12.0 32.2 4.2 38.8

(1.51) (1.66) (1.92) (3.37) (1.48) (10.59)

F

a

4.96 6.62 2.34 6.13 9.90 9.55

Statistics: executive

P

Fa

P

, 0.05 , 0.03 ns , 0.02 , 0.005 , 0.005

0.73 5.27 3.48 3.60 1.24 1.91

ns , 0.03 5 0.07 5 0.06 ns ns

Univariate comparisons with P adjusted accorded to the Bonferroni’s method.

Table 5 Pearson correlations among age, clinical variables, executive scores and memory tasks in patients (n 5 41) evaluated with the MCST

MCST Ca.a Free recall 1 Free recall 2 Free recall 3 Index CFR% b

Age

Hospitalizations

SRRS

MCST Ca a

2 0.05 2 0.27** 2 0.26** 2 0.27** 2 0.34*

0.16 2 0.33* 2 0.02 2 0.10 2 0.12

0.06 2 0.36* 2 0.09 2 0.11 2 0.16

0.06 0.31* 0.31* 0.19

a

MCST Ca., categories achieved. Index CFR%, consistence of free recall %. * P , 0.05; **P 5 0.08. b

ence between young and older depressed especially regarding memory functions (see Tarbuck and Paykel, 1995; Kindermann and Brown, 1997). We have tested the hypothesis that executive difficulties might influence memory problems in depressed patients and contribute to age-related difference in memory performance. As expected patients with mild executive impairment showed significant lower scores that non-dysexecutive patients on free recall tasks. However, again there was no significant interaction of age and executive functioning on these memory scores. Although we used only one task for assessing executive functioning in our patients, our preliminary results suggest that the

differences between old and young patients could not be related to a more detrimental effect of executive dysfunction on memory performance in older patients. These results rather indicate that executive problems may affect the cognitive skills necessary for strategic encoding and retrieval processes in young as well as old depressed patients (Moscovitch, 1992). This executive memory decline hypothesis in depression is consistent with several neuroimaging studies revealing that specific regions of the prefrontal cortex are involved in mood regulation, memory processes and the neuropsychopathology of depression (see Grady, 1999; Mayberg et al., 1999). Our results must be interpreted in the context of

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

several methodological limitations. First, we did not evaluate the patients with a large comprehensive neuropsychological battery. Regarding the great heterogeneity of cognitive aging, our results thus cannot be generalized to other cognitive domains (King et al., 1993). Likewise, we cannot exclude the possibility that the worse free recall performance of the dysexecutive patients reflects differences in general cognitive functioning rather than a specific causal link between executive deficits and retrieval. Second, most of the depressed patients were receiving psychoactive drugs at the time of testing and these medications may have influenced the results. Although we addressed this issue by analyzing the data with pharmacological treatments as covariates, we cannot definitely rule out the possibility that treatments alter memory performance in depressed. Clearly, further studies should be conducted with patients free of treatment to confirm the results reported here. Finally it should be noted that subjects included in this study were selected within a restricted age range. Although our data show a memory decline in young–old normal controls and patients (see Lavigne and Finley, 1990), an interaction of group 3 age may have emerged if we had tested subjects over 70 years of age. Memory deficits of depressed patients were mainly related to the psychomotor retardation score and the number of hospitalizations in the depressed patients. Several authors have emphasized the role of cognitive and motor slowing in depression at the clinical and cognitive levels (White et al., 1997; Dantchev and Widlocher, 1998). Psychomotor retardation might induce deficits in free recall by decreasing cognitive initiation and generation of cue during retrieval. Numerous previous works have linked memory dysfunction in depression with the number of previous episodes and hospitalizations (Bieliauskas, 1993; Tham et al., 1997; Kessing, 1998). The present findings suggest that both state and trait factors contribute to memory impairments in young and older depressed. One possible explanation for these correlational findings is that the actual severity of depression and the executive level influence the controlled and strategic component of memory tasks, whereas trait factors (i.e., history of affective illness) influence non-strategic processing. Although the

269

cross-sectional nature of this study precludes any conclusive interpretation, our results are consistent with the idea that patients with recurrent depression experience a chronic cognitive decline that may be the clinical correlate of reported anatomical changes, such as hippocampal atrophy (Sheline et al., 1996; Basso and Bornstein, 1999). While additional longitudinal studies are needed, the recurrence of depressive episodes and hospitalizations may therefore reflect this progressive cerebral dysfunction.

Acknowledgements The research reported in this paper was conducted in partial fulfillment of the requirements for the PH.D. degree by the first author at the University of Paris VI. Dr. Fossati is supported by AP-HP and Fondation Singer-Polignac. We wish to thank Dr. Mayberg for rereading the manuscript. The authors would like to thank anonymous reviewers for their helpful comments on an earlier draft of the manuscript and the individuals who participated in this research.

References Austin, M.P., Mitchell, P., Wilhem, K., Parker, G., Hickie, I., Brodaty, H., Chan, J., Eyers, K., Milic, M., Hadzi-Pavlovic, D., 1999. Cognitive function in depression: a distinct pattern of frontal impairment in melancholia? Psychol. Med. 29, 73–85. Basso, M.R., Bornstein, R.A., 1999. Relative memory deficits in recurrent versus first-episode major depression on a word-list learning task. Neuropsychology 13, 557–563. Beats, B.C., Sahakian, B.J., Levy, R., 1996. Cognitive performance in tests sensitive to frontal lobe dysfunction in the elderly depressed. Psychol. Med. 26, 591–603. Bieliauskas, L.A., 1993. Depressed or not Depressed? That is the question. J. Clin. Exp. Neuropsychol. 15, 116–134. Bornstein, R.A., Baker, G.B., Douglass, A.B., 1991. Depression and memory in major depressive disorder. J. Neuropsy. Clin. Neurosci. 3, 78–80. Bryan, J., Luszcz, M.A., 2000. Measurement of executive function: considerations for detecting adult age differences. J. Clin. Exp. Neuropsychol. 22, 40–55. Bryan, J., Luszcz, M.A., Pointer, S., 1999. Executive function and processing resources as predictors of adult age differences in the implementation of encoding strategies. Aging Neuropsychol. Cogn. 6, 273–287.

270

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271

Burt, D.B., Zembar, M.J., Niederehe, G., 1995. Depression and memory impairment: a metaanalysis of the association, its pattern and specificity. Psychol. Bull. 117, 285–305. Cavanaugh, S.V.A., Wettstein, R.M., 1983. The relationship between severity of depression, cognitive dysfunction and age in medical inpatients. Am. J. Psychiatry 140, 495–496. Craik, F.I.M., Morris, L.W., Morris, R.G., Loewen, E.R., 1990. Relations between source amnesia and frontal lobe functioning in older adults. Psychol. Aging 5, 148–151. Crawford, J.R., Bryan, J., Luszcz, M.A., Obonsawin, M.C., Stewart, L., 2000. The executive decline hypothesis of cognitive aging: do executive deficits qualify as differential deficits and do they mediate age-related memory decline? Aging Neuropsychol. Cogn. 7, 9–31. Dantchev, N., Widlocher, D.J., 1998. The measurement of retardation in depression. J. Clin. Psychiatry 59 (Suppl. 14), 19–25. ¨ Daum, I., Graber, S., Schugens, M.M., Mayes, A.R., 1996. Memory dysfunction of the frontal type in normal ageing. Neuroreport 7, 2625–2628. De Zubicaray, G., Ashton, R., 1996. Nelson’s (1976) modified card sorting test: a review. Clin. Neuropsychol. 10, 245–254. Fossati, P., Amar, G., Raoux, N., Ergis, A.M., Allilaire, J.F., 1999. Executive functioning and verbal memory in young patients with unipolar depression and schizophrenia. Psychiatry Res. 89, 171–187. Golinkoff, M., Sweeney, J.A., 1989. Cognitive impairments in depression. J. Affect. Disord. 17, 105–112. Grady, C.L., 1999. Neuroimaging and activation of the frontal lobes. In: Miller, B.L., Cummings, J.L. (Eds.), The Human Frontal Lobes. Functions and Disorders. Guilford Press, New York, pp. 196–230. Grober, E., Buschke, H., Crystal, H., Bang, S., Dresner, R., 1988. Screening for dementia by memory testing. Neurology 38, 900–903. Ilsley, J.E., Moffoot, A.P.R., O’Carroll, R.E., 1995. An analysis of memory dysfunction in major depression. J. Affect. Disord. 35, 1–9. Jonhson, M.H., Magaro, P.A., 1987. Effects of mood and severity on memory processes in depression and mania. Psychol. Bull. 101, 28–40. Kessing, L.V., 1998. Cognitive impairment in the euthymic phase of affective disorder. Psychol. Med. 28, 1027–1038. Kessing, L.V., Dam, H., Jorgensen, O.S., Bolwig, T.G., 1996. Cognitive impairment in affective disorders. Relation to illness characteristics. Nordic J. Psychiatry 50, 305–316. Kindermann, S.S., Brown, G.G., 1997. Depression and memory in the elderly: a meta-analysis. J. Clin. Exp. Neuropsychol. 19, 625–642. King, D.A., Caine, E.D., Conwell, Y., Cox, C., 1991. The neuropsychology of depression in the elderly: a comparative study of normal aging and Alzheimer’s disease. J. Neuropsychiatry Clin. Neurosci. 3, 163–168. King, D.A., Caine, E.D., Cox, C., 1993. Influence of depression and age on selected cognitive functions. Clin. Neuropsychol. 7, 443–453. King, D.A., Cox, C., Lyness, J.M., Conwell, Y., Caine, E.D., 1998.

Quantitative and qualitative differences in the verbal learning performance of elderly depressives and healthy controls. J. Int. Neuropsychol. Soc. 4, 115–126. Lavigne, V.D., Finley, G.E., 1990. Memory in middle-aged adults. Educ. Gerontol. 16, 447–461. Lineweaver, T.T., Bondi, M.W., Thomas, R.G., Salmon, D.P., 1999. A normative study of Nelson’s (1976) modified version of the Wisconsin Card Sorting Test in healthy older adults. Clin. Neuropsychol. 13, 328–347. Lyness, S.A., Eaton, E.M., Schneider, L.S., 1994. Cognitive performance in older and middle aged depressed outpatients and controls. J. Gerontol. Psychol. Sci. 49, P129–P136. Mayberg, H.S., Liotti, M., Brannan, S.K., McGinnis, Mahurin, R.K., Jerabek, P.A., Silva, J.A., Tekell, J.L., Martin, C.C., Lancaster, J.L., Fox, P.T., 1999. Reciprocal limbic-cortical function and negative mood: converging PET findings in depression and normal sadness. Am. J. Psychiatry 156, 675– 682. Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to be sensitive to change. Br. J. Psychiatry 134, 382–389. Moscovitch, M., 1992. Memory and working-with-memory: a component process model based on modules and central systems. J. Cogn. Neurosci. 4, 257–267. Nelson, H.E., 1976. A modified card sorting test sensitive to frontal lobe defects. Cortex 12, 313–324. Niederehe, G., 1986. Depression and memory impairment in the aged. In: Poon, L.W. (Ed.), Clinical Memory Assessment. APA, Washington, DC, pp. 226–237. Niederehe, G., Yoder, C., 1989. Metamemory perceptions in depressions of young and older adults. J. Nerv. Ment. Dis. 177, 4–14. Palmer, B.W., Boone, K.B., Lesser, I.M., Wohl, M.A., Berman, N., Miller, B.L., 1996. Neuropsychological deficits among older depressed patients with predominantly psychological or vegetative symptoms. J. Affect. Disord. 41, 17–24. Rabbit, P., 1997. Introduction: methodologies and models in the study of executive function. In: Rabbit, P. (Ed.), Methodology of Frontal and Executive Function. Psychology Press, Hove, East Sussex, pp. 1–38. Salat, D.H., Kaye, J.A., Janowsky, J.S., 1999. Prefrontal gray and white matter volumes in healthy aging and Alzheimer disease. Arch. Neurol. 56, 33–344. Sheline, Y.I., Wang, P.W., Gado, M.H., Csernansky, J.G., Vannier, M.W., 1996. Hippocampal atrophy in recurrent major depression. Proc. Natl. Acad. Sci. USA 93, 3908–3913. Tarbuck, A.F., Paykel, E.S., 1995. Effects of major depression on the cognitive function of younger and older subjects. Psychol. Med. 25, 285–296. ´ A.A., Johnson, L., Olsson, Tham, A., Engelbrektson, K., Mathe, E., Aberg-Wistedt, A., 1997. Impaired neuropsychological performance in euthymic patients with recurring mood disorders. J. Clin. Psychiatry 58, 26–29. Troyer, A.K., Graves, R.E., Munro Cullum, C., 1994. Executive functioning as a mediator of the relationship between age and episodic memory in healthy aging. Aging Cogn. 1, 45–53.

P. Fossati et al. / Journal of Affective Disorders 68 (2002) 261 – 271 Veiel, H.O.F., 1997. A preliminary profile of neuropsychological deficits associated with major depression. J. Clin. Exp. Neuropsychol. 19, 587–603. White, D.A., Myerson, J., Hale, S., 1997. How cognitive is psychomotor slowing in depression? Evidence from a metaanalysis. Aging Neuropsychol. Cogn. 4, 166–174. ¨ Widlocher, D., 1983. Psychomotor retardation: clinical, theoretical

271

and psychometric aspects. Psychiatr. Clin. North Am. 6, 27– 40. Wolfe, J., Granholm, E., Butters, N., Saunders, E., Janowski, D., 1987. Verbal memory deficits associated with major affective disorders: a comparison of unipolar and bipolar patients. J. Affect. Disord. 13, 83–92.