modulating speed-accuracy strategies in major depression

Running Head: Speed-Accuracy in Depression

Modulating Speed-Accuracy Strategies in Major Depression

Antonino Vallesia,b,#, Francesca Canalazc, Matteo Balestrieric, Paolo Brambillac,d,e

aDepartment of Neurosciences: SNPSRR, Universitá degli Studi di Padova, Italy

bCentro di Neuroscienze Cognitive, Universitá degli Studi di Padova, Italy

cDepartment of Experimental & Clinical Medicine (DISM), Inter-University Center for Behavioural

Neurosciences (ICBN), University of Udine, Udine, Italy

dDepartment of Psychiatry & Behavioral Sciences, University of Texas Medical School at Houston,

TX, USA

eIRCCS ‘E. Medea’, Scientific Institute, Udine, Italy

#Corresponding Author’s address:

Antonino Vallesi

Department of Neurosciences: SNPSRR, University of Padova

Via Giustiniani, 5, 35128 Padova

Phone: +39 049 821 4450

Fax: +39 049 821 8988

E-mail: [email protected]

Number of words in the abstract: 148

Number of words in the text: 2638

Number of tables: 2

Number of figures: 1

Number of supplementary material: 0

Speed-Accuracy in Depression – Vallesi et al. 1

mailto:[email protected]

Abstract

Background: Depression is associated with deficits in cognitive flexibility. The role of general

slowing in modulating more specific cognitive deficits is however unclear. Aim: We assessed how

depression affects the capacity to strategically adapt behavior between harsh and prudent response

modalities and how general and specific processes may contribute to performance deficits.

Methods: Patients suffering from major depression and age- and education-matched healthy

controls were asked to randomly stress either speed or accuracy during perceptual decision-making.

Results: Diffusion models showed that patients with depression kept using a less conservative

strategy after a trial with speed vs. accuracy instructions. Additionally, the depression group

showed a slower rate of evidence accumulation as indicated by a generally lower drift rate.

Conclusions: These data demonstrate that less efficient strategic regulation of behavior in

depression is due not only to general slowing, but also to more specific deficits, such as a rigid

dependence on past contextual instructions. Future studies should investigate the neuro-anatomical

basis of this deficit.

Keywords: flexibility, executive functions, depression, speed-accuracy trade off, perceptual

decision-making, diffusion models.


Introduction

Depression is a major psychiatric disorder, which is usually accompanied by deficits in cognitive

functioning, including impairments in cognitive flexibility (Airaksinen et al., 2004; Meiran et al.,

2011; Whitmer and Banich, 2007). In particular, Whitmer and Banich (2007), by using a task-

switching paradigm, showed that depressive rumination is associated with deficits in inhibiting

previous mental sets. Moreover, patients with major depression show significant deficits when

performing the Wisconsin Card Sorting Test (WCST), a test of cognitive flexibility, including

problems in shifting cognitive sets when appropriate (e.g., Franke et al., 1993; Merriam et al.,

1999). Perseverative responses and other deficits in the WCST are predicted by the severity of

depressive symptoms independently of general intellectual abilities (Martin et al., 1991).

Cognitive flexibility deficits are possibly mediated by prefrontal serotonin deficiency (Clarke et

al., 2004). From the functional-anatomical point of view, the dorsolateral prefrontal cortex

(especially on the left hemisphere), which is reliably activated during tasks tapping cognitive

flexibility (Kim et al., 2011; Vallesi, 2012), has been shown to be hypo-metabolic in depression

(Bench et al., 1992; Davidson et al., 2002; Drevets, 2000; Mayberg et al., 1999), although many

fMRI studies have shown that this region may be inefficiently hyper-active during task execution

(see Graham et al., 2013, for a recent meta-analysis). Depression is also accompanied by an

abnormal pattern of activation of medial prefrontal structures such as the anterior cingulate cortex

(Bench et al., 1992; Diener et al., 2012; Drevets et al., 1992; Kennedy et al., 2001), a region

implicated in energization, drive, and in the effortful allocation of cognitive and motor control

(Paus, 2001; Shenhav et al., 2013; Stuss, 2011; Vallesi, 2012).

The prediction follows that patients suffering from major depression will show impairment in

flexible regulation of behavior, especially in tasks recruiting the dorsolateral prefrontal cortex and

the anterior cingulate cortex. Cognitive flexibility is for instance required when trading off speed

and accuracy. This capacity is important in everyday life because it allows us to flexibly adapt to

different and quickly changing environmental and endogenous demands. It has been shown that


switching from speed to accuracy by adopting stricter criteria for decision-making involves the left

dorsolateral prefrontal cortex (Vallesi et al., 2012). Patients suffering from depression are thus

expected to be impaired in this condition. A second prediction concerns possible deficits in the

energization process based on anterior cingulate cortex, which should produce generally slower

responding (Stuss et al., 2005).

Although a neuroimaging study should be set up to directly test the link between the

involvement of these key regions and cognitive flexibility problems in depression, the present study

aimed at finding behavioral evidence for deficits in speed-accuracy strategy regulations in major

depression. A perceptual decision-making task was adopted in which speed and accuracy

instructions were manipulated on a trial-by-trial basis to understand whether depression is

associated with cognitive flexibility deficits in trials that require a switch in response strategy, and

in slow response patterns especially in trial sequences with high time pressure. To gain deep

insights on the possible mechanisms of depression-related deficits in speed-accuracy trade off

regulation, we adopted a diffusion model analysis (e.g., Ratcliff, 1978; Voss and Voss, 2007) of the

performance data (i.e., response times and accuracy), which allowed us to estimate more

informative decisional and non-decisional sub-processes.

Methods

Participants

Twenty patients with a current or previous diagnosis of Major Depression (mean age: 47 years,

range: 23-72; mean education years: 13 years, range: 5-18; 5 males; mean score for the Hamilton

Depression Rating Scale for Depression (HDRS) at the time of testing: 6.8, range: 0-18) and 28

healthy volunteers (mean age: 48 years, range: 23-73; mean education years: 14.6 years, range: 5-

19; 13 males) took part in the experiment. According to t-test analysis, the two groups were

matched for age [t(46)=.16, p=.87] and years of education [t(46)=1.37, p=.18]. All the patients with


depression apart from two were under different anti-depressant treatments at the time of testing,

while all of them had previously been under pharmacological treatment.

Patients with major depression met the Diagnostic and Statistical Manual of Mental Disorders,

fourth edition (DSM-IV) criteria and were recruited from the out-patient Psychiatric Clinic of the

University Hospital of Udine, Italy, as diagnosed with Structured Clinical Interview for Axis I

DSM-IV Disorders (SCID-I). Diagnoses were confirmed by the clinical consensus of two staff

psychiatrists. All the patients in the depression group did not have any other comorbid Axis-I

diagnosis, including history of substance or alcohol abuse. They had been administered various

(mainly serotoninergic) drugs for variable life-time periods. Sixteen patients were treated with

antidepressants: 8 took selective serotonin reuptake inhibitors (SSRI), 1 SSRI and Tricyclic

Antidepressant (TCA), 4 selective norepinephrine reuptake inhibitor (SNRI), 1 TCA, 1 selective

noradrenaline reuptake inhibitor (NARI) and 1 agomelatina.

Table 1 reports the type of antidepressant and dosage, length of disease, and HDRS scores,

number of past episodes, age at depression onset and time in remission. Additionally, Table 1 also

reports scores of an adapted Antidepressant Treatment History Form rating scale (see Sackeim,

2001) on a 4 point-scale (0=no drug, 3= highest drug load). This procedure was slightly adapted

since some antidepressants were not present in the original study, and we assimilated those drugs

with drugs with similar profile among those reported in Sackeim’s (2001) study (e.g., venlafaxina

and duloxetina, escitalopram and citalopram).

The study was approved in advance by the ethics committee: “IRCCS E. Medea Associazione La

Nostra Famiglia”. The authors assert that all procedures contributing to this work comply with the

Helsinki Declaration of 1975, as revised in 2008.

----Insert Table 1 about here----

Materials and task


The basic task was to judge whether the predominant color in the target square was orange or

green by responding with the index and middle fingers of the right hand (keys “B” and “N” of the

laptop keyboard, appropriately covered with orange and green labels, respectively). The association

between prevailing color and response button was counterbalanced between-subjects.

Visual stimuli were squares of 100 mm2 presented centrally against a constantly grey

background. Lighter and darker grey pixels randomly dispersed in the square frame (50% each)

were used to form the fixation space during the presentation of cue stimuli. Cues were triplets of

capital letters (VEL, for speed, that is “velocità” in Italian, or ACC for accuracy) appearing on the

top of the grey square frame at the beginning of the trial and disappearing with the target offset.

Orange and green pixels were randomly dispersed in the square in various ratios (44/56, 47/53,

53/47, 56/44) to form target stimuli (adapted from Vallesi et al., 2012). Cues appeared on the top of

the screen 1000 ms before the presentation of the target and disappeared with the target offset. The

target lasted 2000 ms, followed by a blank screen of 500 ms. Response deadline was 2500 ms.

Moreover, participants received a feedback which was different according to whether they fulfilled

the speed-accuracy instructions or not (see below). The feedback was presented at the center of the

screen for 1500 ms.

A first familiarization run with no speed-accuracy instructions or feedback (32 trials) was

performed before the beginning of the real test. Two experimental runs with feedback were

subsequently performed (64 trials each). In each run, the 4 orange/green proportions were presented

pseudo-randomly and equiprobably. In the runs with cue (all but the first familiarization run), the

combination of 2 cue type (accuracy vs. speed) and 2 previous cue type (accuracy vs. speed) factors

was also presented randomly and with the same probability.

In the first familiarization run, participants were asked to simply perform the color estimation

task (i.e., “are there more green or orange pixels”). During the next 2 training runs with feedback,

participants were required to stress either speed or accuracy according to the nature of the cue at the

beginning of the trial. Visual feedback was displayed as a sentence (font: Courier New, black color)


for 1000 ms after each trial, followed by a 500 ms blank screen. The provided feedback could be

one of the following:

- “Sbagliato, attenzione!” (Wrong, be careful! in English): if participants failed to obey the

accuracy instruction by making a mistake.

- “Sii più veloce” (Be quicker!): if RT was < average RT + 1 SD of the familiarization run

during speed trials.

- “Tempo scaduto”! (Time expired!): if no response was collected before deadline.

- “Bene così” (Well done!): in any other case.

During the last run, which was a test run, participants performed the same task as during the

training runs with feedback but now no feedback was provided at the end of every trial.

Data Analysis

The 32 trials from the familiarization run, trials with no response, trials with RTs < 200 ms, and

the first trial of each run were excluded from further analyses. Both RTs and accuracy data were

submitted to a 2x2x2x2 mixed ANOVA, with phase (training with feedback, test), preceding cue

(accuracy, speed) and current cue (accuracy, speed) as the within-subject factors, and group

(Depressed, Controls) as the between-subjects factor. Tukey’s HSD test was used as the post-hoc

test to detect the sources underlying significant effects.

RT and accuracy data were also combined together in an additional analysis based on diffusion

models (Ratcliff, 1978; Voss and Voss, 2007), as this model is specifically suited to analyse

differences in speed-accuracy trade-off. The analysis of two-choice RT data by means of the

diffusion-model assumes that information is continuously accumulated until one of two thresholds

is passed. The analysis is based on the distributions of both correct and erroneous responses. From

these distributions a set of parameters is estimated that allows drawing conclusions about non-

decisional and decisional processes. Specifically, the model parameter t0 represents the duration of

non-decisional processes such as stimulus encoding and response execution. The model parameter


‘v’ (drift rate) indicates the strength of the systematic influence that drifts the decision process from

a starting point (parameter z) to one of two response thresholds. When either response threshold is

reached, a response is initiated. Moreover, the distance between response thresholds is indicated by

the parameter ‘a’. This parameter thus indicates how much information is required before either

response is initiated. This is a measure of conservativeness/liberality. The z/a measure is instead

used as a measure of bias towards one of the two responses (range 0-1), with 0.5 indicating no

decisional bias.

We used the fast-dm free software

(http://www.psychologie.uni-heidelberg.de/ae/meth/fast-dm/index-en.html) to estimate the

parameters of the diffusion model (Spaniol et al., 2006; Voss and Voss, 2007; Voss et al., 2004). A

multi-dimensional optimization method (i.e., Simplex downhill search) was used to optimize the

fitting between predicted and empirical distributions and the Kolmogorov–Smirnov test was used to

statistically assess this fitting. We allowed ‘z’, ‘a’ and ‘v’ to vary with each of the four conditions

(2 preceding cue x 2 current cue), while other parameters of the diffusion model were assumed to

be constant for all conditions. Each of the three varied parameters was treated as the dependent

variable of a 2x2x2 mixed ANOVA, with preceding and current cue type as the within-subject

factors and group (Depression vs. Control) as the between-subjects factor. The effect of group on

the constant parameter t0, which indicates the duration of non-decisional processes, was also

assessed through a two-sample t-test.

For each significant effect involving the group factor, within the depression group we performed

(i) Pearson’s correlations between the relevant dependent variable and the HDRS, and (ii)

Spearman Rank Order correlations with the Antidepressant Treatment History form Rating scores

(cf., Sackeim, 2001), since the latter were distributed on a scale of just 4 points. We only reported

correlations in the text if they reached the significance level (p<.05).

Results


http://www.psychologie.uni-heidelberg.de/ae/meth/fast-dm/index-en.html

All the behavioral data are shown in Table 2.

Accuracy. In the ANOVA concerning the speed-accuracy experimental runs, there was a strong

trend for a cue main effect [F(1, 46)=3.73, p=.059], with participants being more accurate for

accuracy (90%) than for speed cues (88.7%). Depressed patients were on average less accurate than

their controls [86.8% vs. 91.8%; F(1, 46)=10.13, p=.0026], although they were still able to perform

the task with a reasonable accuracy level. Accuracy improved from the training phase with

feedback (87.2%) to the test phase (91.4%) in both groups [F(1, 46)=27.2, p<.0001]. An interaction

between phase and group [F(1, 46)=5.33, p=.025] was better qualified by a 3-way interaction with

the preceding cue [F(1, 46)=7.55, p=.008]. Post-hoc Tukey’s tests revealed that this interaction was

mainly due to differential sequential effects for the two groups in the test phase. Specifically, the

accuracy level in the Depression group depended on the preceding cue type when no feedback was

provided during the test phase, with worse accuracy when the current trial followed speed (88%) vs.

accuracy (91.7%) instructions in the preceding trial (p=.004). There was no modulation by the

preceding cue in the control group (93.4 vs. 92.5%; p=.97). This was further confirmed with

separate ANOVAs for each phase, which revealed a significant preceding cue by group interaction

for the test phase [F(1, 46)=9.58, p=.0033], but not for the training one (p=.55). Correlation

analyses performed on the depression group showed that, for the test phase, the difference between

the two preceding cues (accuracy - speed) in terms of accuracy level was negatively correlated with

the antidepressant rating scale (Spearman R=-.52, p=.019), that is, the higher the antidepressant

load, the smaller the differential carry-over effects from the preceding cue type. No other effect was

significant.

Response Times (RTs). RTs were shorter in the test run than in the training runs with feedback

[887 vs. 974 ms; F(1, 46)=40.57, p<.00001]. They were shorter for preceding speed cues than for

preceding accuracy cues [920 vs. 940 ms; F(1, 46)=13.25, p=.0007]. Moreover, RTs were shorter

for current speed cues than accuracy cues [893 vs. 967 ms; F(1, 46)=31.37, p<.00001], this


difference being more accentuated in the training runs than in the test run [cue x phase interaction:

F(1, 46)=10.46, p=.002]. A preceding x current cue interaction [F(1, 46)=27.3, p=.00001] was

mainly due to the fact that RTs in speed trials preceded by speed trials were shorter than speed trials

preceded by accuracy trials (872 vs. 915 ms, p=0.0002), while the preceding cue did not modulate

RTs for current accuracy trials (966 vs. 969 ms; p=.98). No other effect was significant.

----Insert Table 2 about here----

Diffusion model analysis.

Decisional Bias ‘z/a’. The decisional bias was higher (i.e., closer to the correct response

criterion) for current speed vs. accuracy instructions [F(1, 46)=4.36, p=.042]. There was also a

preceding cue by group interaction [F(1, 46)=5.28, p=.026], which suggested that the two groups

modulated the decisional bias differently depending on the preceding cue. However, no comparison

survived the Tukey’s HSD post-hoc test (all ps>.21). No other effect was significant.

Distance between response criteria ‘a’. This parameter, which measures the conservativeness of

the adopted response criteria, was generally higher in the training runs with feedback than in the test

run [F(1, 46)=13.31, p=.0007]. The ‘a’ value was also higher for preceding accuracy cues than for

speed ones [F(1, 46)=10.46, p=.0023]. Moreover, this value was higher for current accuracy cues

than for speed ones [F(1, 46)=21.78, p=.00003]. A preceding cue x group interaction [F(1, 46)=4.3,

p=.038] indicated that only the patients with depression modulated the parameter ‘a’ as a function

of preceding cue (p<.006), being it lower for preceding speed than for preceding accuracy trials (see

Figure 1 and table 2), while the control group did not show any modulation of this parameter

(p=.81).

----Insert Figure 1 about here----


Drift rate ‘v’. There was a phase main effect [F(1, 46)=25.9, p=.00001], which showed that this

parameter was higher for the test run than for the training runs, and a group main effect [F(1,

46)=10.49, p=.0022], with the drift rate being generally higher in controls than in the depressed

patients. Correlation analyses showed that the average drift rate in the depressed patient group was

negatively correlated with the HDRS (Pearson’s r=-.52, p=.018), that is, the higher the depression

level as measured with the Hamilton scale, the lower the drift rate. No other effect was significant

for the ANOVA.

Non-decisional processes ‘t0’. There was no group difference in the constant parameter t0

(p=-.72).

Discussion

The present study investigated how depression affects the flexible regulation of cognitive

strategies. In particular, the capacity to switch between speed and accuracy in perceptual decision-

making according to the task demands was investigated. Patients with depression had difficulties in

overcoming the instructions presented in the previous trial, as indicated by the presence of

sequential effects in the percentage of errors, especially in the test phase with no feedback. In

particular, they were 4% less accurate when they had already traded off accuracy for speed in the

previous trial (i.e., preceding speed cues), independently of the current instructions. This sequential

effect was negatively correlated with a comprehensive measure of antidepressant treatment load,

suggesting that the dependence on the previous trial instructions ameliorated with a higher

medication load. The control group did not show these sequential effects.

A more complete picture was provided by the diffusion model analysis, which showed that this

deficit in flexibility was mainly present in the depressed group when considering the distance

between response criteria (parameter ‘a’). This distance, which indicates the degree of


conservativeness, was smaller when the previous cue instructed a quick response, independently of

the current cue. These effects suggest less flexibility in adapting the response strategy according to

the external demands in depression, especially when the previous trial instruction was “speed”. This

would be especially problematic when the current instruction cues a switch towards a more accurate

strategy, a condition which previous neuroimaging evidence has shown to recruit the left

dorsolateral prefrontal cortex (Vallesi et al., 2012). It is worth noting that this prefrontal region

shows hypo-metabolism in depression (Davidson et al., 2002; Dougherty and Rauch, 1997; Heller

and Nitschke, 1998). No correlation between depressive symptoms or antidepressant level and this

measure was found, suggesting that problems with response criterion flexible adjustments could be

a stable characteristic of depression.

Another important finding was a generally slower drift rate (i.e., systematic velocity of

transmission of information) in patients suffering from depression than in controls, which is

compatible with many depressive symptoms such as general psychomotor slowness and loss of

drive (Mayberg, 1994; White et al., 1997). However, this finding suggests a more specific slowness

in perceptual evidence accumulation than just general slowing, since there was no difference

between the group with depression and their controls for the duration of non-decisional processes

(t0). The negative correlation with the Hamilton depression rating scale demonstrated that this

deficit was specifically linked with the severity of the depression disease, while a comprehensive

measure of the amount of drugs taken did not show any role for this factor.

General slowing could be a confounding factor when assessing specific cognitive deficits in

depression (White et al., 1997). The present diffusion model approach was therefore useful to

isolate different component processes and separately characterize their contribution to impaired

performance of the task at hand observed in the depression group. In this respect, diffusion models

are superior to standard analyses of raw performance measures, revealing their effectiveness in the

evaluation of both specific and general cognitive deficits in depression.


Although the significant results clearly demonstrate that the present study did not suffer from a

general lack of power, a possible limitation is represented by the small sample size for the patients

with depression. Therefore, replications with future studies including bigger sample sizes and other

tasks with SAT manipulation are warranted to corroborate the present findings and make them more

generalizable.

Most of the patients with depression tested in the present study were under (mainly

serotoninergic) drugs and therefore it is difficult to disentangle the effect of drug from that of

depression per se in driving the observed effects. Notwithstanding the difficulty of recruiting drug-

free patients with clinically relevant depression, a future study conducted on patients at their first

diagnosis could be helpful to better characterize the role of depression in speed-accuracy switching.

In conclusion, the present study shows a depression-related deficit in flexible regulation of

speed-accuracy during decision-making, which may be related to frontal lobe dysfunction. Future

studies should more directly investigate the neuro-anatomical substrate of this deficit.


ReferencesAiraksinen, E., Larsson, M., Lundberg, I., Forsell, Y., 2004. Cognitive functions in depressive

disorders: evidence from a population-based study. Psychol. Med. 34, 83–91.Bench, C.J., Friston, K.J., Brown, R.G., Scott, L.C., Frackowiak, R.S., Dolan, R.J., 1992. The

anatomy of melancholia - focal abnormalities of cerebral blood flow in major depression. Psychol. Med. 22, 607–615.

Clarke, H.F., Dalley, J.W., Crofts, H.S., Robbins, T.W., Roberts, A.C., 2004. Cognitive Inflexibility After Prefrontal Serotonin Depletion. Science 304, 878–880. doi:10.1126/science.1094987

Davidson, R.J., Pizzagalli, D., Nitschke, J.B., Putnam, K., 2002. Depression: perspectives from affective neuroscience. Annu. Rev. Psychol. 53, 545–574. doi:10.1146/annurev.psych.53.100901.135148

Diener, C., Kuehner, C., Brusniak, W., Ubl, B., Wessa, M., Flor, H., 2012. A meta-analysis of neurofunctional imaging studies of emotion and cognition in major depression. NeuroImage 61, 677–685. doi:10.1016/j.neuroimage.2012.04.005

Dougherty, D., Rauch, S.L., 1997. Neuroimaging and neurobiological models of depression. Harv. Rev. Psychiatry 5, 138–159.

Drevets, W.C., 2000. Functional anatomical abnormalities in limbic and prefrontal cortical structures in major depression. Prog. Brain Res. 126, 413–431. doi:10.1016/S0079-6123(00)26027-5

Drevets, W.C., Videen, T.O., Price, J.L., Preskorn, S.H., Carmichael, S.T., Raichle, M.E., 1992. A functional anatomical study of unipolar depression. J. Neurosci. 12, 3628–3641.

Franke, P., Maier, W., Hardt, J., Frieboes, R., Lichtermann, D., & Hain, C. 1993. Assessment of Frontal Lobe Functioning in Schizophrenia and Unipolar Major Depression. Psychopathology 26(2), 76–84. doi:10.1159/000284803

Graham, J., Salimi-Khorshidi, G., Hagan, C., Walsh, N., Goodyer, I., Lennox, B., & Suckling, J. 2013. Meta-analytic evidence for neuroimaging models of depression: state or trait? Journal of Affective Disorders, 151(2), 423–431. doi:10.1016/j.jad.2013.07.002

Heller, W., Nitschke, J.B., 1998. The puzzle of regional brain activity in depression and anxiety: The importance of subtypes and comorbidity. Cognition and Emotion, 12 (3), 421–444. doi: 10.1080/026999398379664

Kennedy, S.H., Evans, K.R., Krüger, S., Mayberg, H.S., Meyer, J.H., McCann, S., Arifuzzman, A.I., Houle, S., Vaccarino, F.J., 2001. Changes in regional brain glucose metabolism measured with positron emission tomography after paroxetine treatment of major depression. Am. J. Psychiatry 158, 899–905.

Kim, C., Johnson, N.F., Cilles, S.E., Gold, B.T., 2011. Common and distinct mechanisms of cognitive flexibility in prefrontal cortex. J. Neurosci. 31, 4771–4779. doi:10.1523/JNEUROSCI.5923-10.2011

Martin, D. J., Oren, Z., & Boone, K. 1991. Major depressives’ and dysthmics’ performance on the Wisconsin Card Sorting Test. Journal of Clinical Psychology 47(5), 684–690.

Mayberg, H.S., 1994. Frontal lobe dysfunction in secondary depression. J. Neuropsychiatry Clin. Neurosci. 6, 428–442.

Mayberg, H.S., Liotti, M., Brannan, S.K., McGinnis, S., 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 (5), 675–682. doi:

Meiran, N., Diamond, G.M., Toder, D., Nemets, B., 2011. Cognitive rigidity in unipolar depression and obsessive compulsive disorder: examination of task switching, Stroop, working memory updating and post-conflict adaptation. Psychiatry Res. 185, 149–156. doi:10.1016/j.psychres.2010.04.044


Merriam, E.P., Thase, M.E., Haas, G.L., Keshavan, M.S., & Sweeney, J.A. 1999. Prefrontal Cortical Dysfunction in Depression Determined by Wisconsin Card Sorting Test Performance. American Journal of Psychiatry 156(5), 780–782.

Paus, T., 2001. Primate anterior cingulate cortex: where motor control, drive and cognition interface. Nat. Rev. Neurosci. 2, 417–424. doi:10.1038/35077500

Ratcliff, R., 1978. A theory of memory retrieval. Psychological Review 85, 59–108.Sackeim, H.A., 2001. The definition and meaning of treatment-resistant depression. The Journal of

Clinical Psychiatry 62 (Suppl 16), 10–17.Shenhav, A., Botvinick, M.M., Cohen, J.D., 2013. The expected value of control: an integrative

theory of anterior cingulate cortex function. Neuron 79, 217–240. doi:10.1016/j.neuron.2013.07.007

Spaniol, J., Madden, D.J., Voss, A., 2006. A diffusion model analysis of adult age differences in episodic and semantic long-term memory retrieval. J. Exp. Psychol. Learn. Mem. Cogn. 32, 101–117. doi:10.1037/0278-7393.32.1.101

Stuss, D.T., 2011. Functions of the frontal lobes: relation to executive functions. J. Int. Neuropsychol. Soc. 17, 759–765. doi:10.1017/S1355617711000695

Stuss, D.T., Alexander, M.P., Shallice, T., Picton, T.W., Binns, M.A., Macdonald, R., Borowiec, A., Katz, D.I., 2005. Multiple frontal systems controlling response speed. Neuropsychologia 43, 396–417. doi:10.1016/j.neuropsychologia.2004.06.010

Vallesi, A., 2012. Organization of Executive Functions: Hemispheric Asymmetries. Journal of Cognitive Psychology 24, 367–386.

Vallesi, A., McIntosh, A.R., Crescentini, C., Stuss, D.T., 2012. fMRI investigation of speed-accuracy strategy switching. Hum. Brain Mapp. 33, 1677–1688. doi:10.1002/hbm.21312

Voss, A., Rothermund, K., Voss, J., 2004. Interpreting the parameters of the diffusion model: an empirical validation. Mem. Cognit. 32, 1206–1220.

Voss, A., Voss, J., 2007. Fast-dm: a free program for efficient diffusion model analysis. Behav. Res. Methods 39, 767–775.

White, D.A., Myerson, J., Hale, S., 1997. How Cognitive is Psychomotor Slowing in Depression? Evidence from a Meta-Analysis. Aging, Neuropsychology and Cognition 4, 166–174.

Whitmer, A.J., Banich, M.T., 2007. Inhibition versus switching deficits in different forms of rumination. Psychol. Sci. 18, 546–553. doi:10.1111/j.1467-9280.2007.01936.x


Tables

Table 1. The table shows, for each patient with major depression, arbitrarily numbered from 1 to 20,

the type of antidepressant and dosages currently used, the length of disease, HDRS scores at the

time of testing, number of past depressive episodes, age at illness onset, time in remission and

scores on an antidepressant treatment rating scale.

Patient number

Type and dosages of antidepressants

Length of disease in months

HDRS scores

N° past depressive episodes

Age at illness onset

Time in remission in months

Antidepressant Treatment

History Form Rating Scale

1 Amitriptilina 10 mg/die 17 8 1 49 10 1

2Fluoxetina 20 mg;

amitriptilina 20 mg/die

96 8 1 42 80 3

3 Paroxetina 20 mg/die 12 18 0 22 0 3

4 Duloxetina 60 mg/die 21 1 2 41 12 2

5 None 24 8 1 28 18 0

6 Duloxetina 60 mg/die 456 6 3 14 444 2

7 Paroxetina 20 mg/die 12 0 0 40 4 3

8 Venlafaxina 75 mg/die 42 5 1 68 33 2

9 None 72 2 1 41 61 0

10 Duloxetina 60/mg die 96 0 2 39 83 2

11 Sertralina 50 mg/die 48 9 1 27 34 2

12 Citalopram 20 mg/die 17 8 1 49 9 3

13 None 6 0 2 24 1 0

14 Agomelatina 25 mg/die 72 13 1 44 66 1

15 None 89 6 1 36 77 0

16 Escitalopram 10 mg/die 26 5 1 38 15 2

17 Bupropione 300 mg/die 720 11 6 28 677 3

18 Escitalopram 10 mg/die 84 13 2 56 68 2

19 Citalopram 20 mg/die 36 8 1 40 22 3

20 Fluoxetina 20 mg/die 276 5 5 68 48 3


Table 2. Behavioral data shown according to group and task condition. For each variable, mean and standard error of the mean (S.E.M.) are shown. ‘Acc’ and ‘Spd’ stand for Accuracy and Speed current cues, respectively. ‘Prec Acc’ and ‘Prec Spd’ indicate accuracy and speed instructions in the preceding trial, respectively.

Depression Group Control GroupPrec Accuracy Prec Speed Prec Accuracy Prec SpeedAcc Spd Acc Spd Acc Spd Acc Spd

Accuracy (%) Training Phase Mean 83.8 83.0 84.0 84.5 90.7 90.6 91.3 89.9S.E.M. 2.1 2.9 2.8 2.1 1.1 1.1 0.9 1.2

Test Phase Mean 94.0 89.5 89.2 86.8 93.1 91.9 93.8 93.1

S.E.M. 1.5 2.0 2.1 2.1 1.0 1.0 1.2 0.8

RTs (ms) Training Phase Mean 1051 972 1046 922 999 924 995 880

S.E.M. 30 48 32 41 39 32 39 33Test Phase Mean 918 893 932 860 897 870 902 825

S.E.M. 37 43 34 40 34 34 37 36

Bias 'z/a' Training Phase Mean 0.58 0.57 0.52 0.56 0.52 0.54 0.52 0.57

S.E.M. 0.02 0.02 0.02 0.03 0.02 0.02 0.02 0.02Test Phase Mean 0.58 0.58 0.54 0.56 0.56 0.58 0.57 0.61

S.E.M. 0.02 0.03 0.03 0.04 0.02 0.02 0.02 0.02

Distance 'a' Training Phase Mean 1.69 1.53 1.68 1.45 1.63 1.48 1.65 1.45


S.E.M. 0.07 0.07 0.06 0.07 0.05 0.05 0.05 0.05

Drift rate 'v' Training Phase Mean 0.96 1.10 1.23 1.30 1.72 1.73 1.75 1.70


S.E.M. 0.20 0.22 0.20 0.27 0.17 0.14 0.15 0.14


Figure Caption

Figure 1. Mean values for the diffusion model parameter ‘a’ (and standard errors of the mean),

which indicates the distance between response criteria (smaller values denote more

conservativeness) are shown according to task conditions and groups. The asterisk indicates

significant Tukey’s HSD test at p<.01.


modulating speed-accuracy strategies in major depression

Documents