Psychological Medicine, 2002, 32, 259–265. " 2002 Cambridge University Press DOI: 10.1017\S0033291701005001 Printed in the United Kingdom Evidence of an early information processing speed deficit in unipolar major depression " G.TSOURTOS, J.C.THOMPSON C.STOUGH From the Department of Psychiatry, University of Adelaide, South Australia; and Neuropsychology Laboratory and Brain Sciences Institute, Swinburne University, Victoria, Australia ABSTRACT Background. Slowing of the speed of information processing has been reported in geriatric depression, but it is not clear if the impairment is present in younger patients, if motor retardation is responsible, or if antidepressant medications play a role. Method. Twenty unmedicated unipolar depressed inpatients were compared with 19 medicated depressed in-patients and 20 age-, sex- and verbal IQ-matched controls on inspection time (IT), a measure of speed of information processing that does not require a speeded motor response. We also examined the relationship between IT and current mood and length of depressive illness. Results. Unmedicated depressed patients showed slowing of information processing speed when compared to both medicated depressed patients and controls. The latter two groups were not significantly different from each other. Slowing of IT was not associated with current mood, but was negatively correlated with length of illness since first episode. No differences in IT were found between patients receiving medication with anticholinergic effects and patients receiving medication with no anticholinergic effects. Conclusions. The findings indicate that unipolar depression is associated with a slowing of speed of information processing in younger patients who have not received antidepressant medication. This does not appear to be a result of motor slowing. memory impairments, medicated neuropsycho- INTRODUCTION logical impairment in patients with geriatric Patients suffering depression often report the depression. Both motor and cognitive speed subjective experience of a slowing in mental appear to be impaired in depression (Sobin & speed (O’Connor et al. 1990). Cognitive slowing Sackheim, 1997; Calgiuri & Ellwanger, 2000), may contribute to neuropsychological impair- although Elliott et al. (1996) found that middle- ment associated with unipolar major depression aged (mean age 49 years) depressed patients (MD), as well as depression secondary to other were impaired on a measure of cognitive speed illnesses (Brebion et al. 2000; Fann et al. 2001). but not motor speed. For example, Brown et al. (1994) reported that In contrast to these findings, Purcell et al. elderly (& 65 years old) depressed patients (1997) reported that younger patients (mean age showed slower performance on a range of 37 years) were impaired on measures of atten- neuropsychological tests than age-matched con- tional set-shifting and planning, but not cog- trols. Nebes et al. (2000) reported that slowing nitive speed. It was concluded by Purcell et al. of information speed, as well as working (1997) that younger patients with depression do not show the cognitive slowing that is reported " in middle-aged and older patients. A problem Address for correspondence: Professor Con Stough, Neuro- psychology Laboratory, Swinburne University, PO Box 218 Haw- with this interpretation is that the cognitive thorn, Victoria 3122, Australia. speed measure used by both Elliott et al. (1996) 259 260 G. Tsourtos and others and Purcell et al. (1997) was time to respond 1987). Subjects are instructed to take as long as (‘thinking time’) during a planning task (the necessary to respond, and to focus on accuracy. Cambridge Neuropsychological Test Auto- A preliminary study by Tsourtos et al. (1995) mated Battery (CANTAB) Tower of London). reported that a mixed group of psychiatric in- As this measure involves a number of cognitive patients diagnosed with either depression, schizo- operations including processing speed, it is phrenia, mania or anxiety disorder had signifi- unclear if one can conclude from Purcell et al.’s cantly longer ITs than a healthy control group. (1997) study that younger depressed patients do The present study examined IT performance not show cognitive slowing. Tarbuck & Paykel in unipolar depressed in-patients of a similar age (1995) reported that older depressed patients to those in the study by Purcell et al. (1997). It (mean age 69 years) were slower than younger was hypothesized that the depressed patients depressed patients (mean age 41 years) on a would be impaired relative to age-, sex- and IQ- choice reaction time (RT) measure. However, matched control subjects. Many studies of the RT improved to a similar extent in both groups neuropsychological profile of depression have following recovery, indicating that both age and included medicated and unmedicated patients depression may affect information processing within the same group (Austin et al. 1992; speed, but these variables do not interact to Tarbuck & Paykel, 1996; Purcell et al. 1997). In produce cognitive slowing in older depressed order to examine if medication has an effect on but not younger depressed. The aim of the IT in depression, medicated and unmedicated present study was therefore to examine if speed subjects were grouped separately in the present of information processing was slowed in young, study. As our previous research has indicated unipolar depressed individuals. that anticholinergic drugs can impair IT Most measures of information processing (Thompson et al. 2000; Waterham et al. 2002), speed rely on reaction item (RT) as the de- we compared patients receiving antidepressants pendent variable. Measures of RT can often be with anticholinergic effects to those receiving confounded by changes to motor speed. While antidepressants with minimal anticholinergic many methodologies allow for the separation of actions. Depressive symptoms and history have movement time (MT) and decision time (DT) also been reported by some studies to be related from RT, DT still measures the speed of both to cognitive impairment in some studies (Austin the perception and encoding of a stimulus, and et al. 1992) but not others (Purcell et al. 1997; the initiation of a motor action. The DT\MT Schatzberg et al. 2000), thus the relationship paradigm is also constrained by subjects being between level of depression, depressive history able to adopt varying speed-accuracy trade-off and IT were examined. strategies, as accuracy can be increased at the expense of response time. Unlike RT procedures METHOD the inspection time (IT) procedure is widely regarded as a measure of the speed of early Subjects stages of information processing that is not Twenty unmedicated and 19 medicated sensitive to motor speed, speed-accuracy trade- depressed in-patients from a psychiatric ward in offs or other cognitive strategies (Nettelbeck, a general hospital in Adelaide, South Australia 1987; Deary & Stough, 1996). Avoiding tasks in who were clinically diagnosed with (MD) accord- which strategies can improve performance is ing to the DSM-III-R criteria, together with 20 crucial in assessing cognition in depression, as healthy controls, match for age, sex and IQ par- depressed individuals are often impaired in the ticipated. Patients diagnosed with any history of deployment of effective cognitive strategies substance abuse, neurological injury, or con- (Channon & Green, 1999). IT is a measure current psychiatric disorder were excluded. An defined as the minimum duration of stimulus initial informal interview with the control presentation required for near perfect response subjects was used to establish any evidence of on a two-choice visual discrimination task. The substance abuse, neurological injury, or family stimulus duration is controlled by superimposing history of psychiatric illness. The vocabulary a backward mask over the stimulus which subscale from the Weschler Adult Intelligence prevents extended iconic sampling (Nettelbeck, Scale – Revised (WAIS-R; Weschler, 1987) was Information processing in depression 261 item scale (standardized scores range between Table 1. Means (..) of the demographic 25–100) of depression experienced in the past, variables and a visual analogue scale (VAS, scores range Age Sex Vocabulary between 0–10) measuring the extent of depres- sion currently experienced. Additional infor- Unmedicated (N l 20) 39n4 (13n6) 12 F 39n6 (13n3) mation was gathered about the patient’s length Medicated (N l 19) 36n1 (12n8) 15 F 39n5 (13n0) Controls (N l 20) 35n8 (13n7) 14 F 40n3 (11n4) of illness both current episode (weeks) and from initial onset (number of weeks since first episode). The type, dosage (mg\day) and length Table 2. Medication of depressed patients in the of medication administered to the medicated non-cholinergic and anticholinergic medication group was also retrospectively recorded from groups hospital drug charts after drug administration. Subject No. Medication Dose\day mg Inspection time Non-cholinergic An IBM compatible PC with a 14 inch monitor 1 Fluoxetine 20 was used to display the monochrome visual IT 2 Moclobemide 150 task with an accompanying 12 12 cm two 3 Fluoxetine 20 i 4 Moclobemide 500 response choice panel. The two buttons were 5 Moclobemide 150 17 mm in diameter and spaced 107 mm apart. 6 Fluoxetine 20 7 Tranylcypromine 70 To measure IT, a small central circular cue 8 Sertraline 50 appeared prior to the stimulus for 500 ms. The 9 Fluoxetine 20 stimulus was composed of two vertical lines, one 10 Lithium 500 29 mm in length, the other,