Long-lasting cognitive impairment in unipolar major depression: a 6-month follow-up study

Abstract

The aim of the study was to investigate cognitive impairment in major depression both acutely and after 6 months. All patients were investigated within a neurocognitive experimental setting at two testing sessions: at inclusion and after 6 months. Automatic and effortful information processing was investigated with a visual search paradigm. Twenty-one patients with recurrent major depression according to DSM-IV and a Hamilton Depression Rating Scale score >18 were included in the study. Healthy subjects, matched for age and gender, were used as a control group. The results showed that the depressed patients performed equal to the control group on trials requiring automatic information processing at both sessions. However, the patients were impaired compared to the control group on trials requiring effortful information processing, also at both sessions. The depressed patients showed no improvement in cognitive performance from test 1 to test 2. The results indicate that the depressed patients had an impaired performance for effortful, but not automatic, visual search performance, and that the impairment remained after 6 months, despite significant improvement in their depression scores.

Introduction

Major depression is associated with cognitive dysfunction during periods of acute illness (Veiel, 1997, Elliot, 1998, Austin et al., 2001). A large number of patients with major depression have recurrent multiple episodes of depression between periods of symptom reduction and nonsymptomatic phases.

However, a major issue is whether cognitive impairment manifested during periods of depression is long lasting or improves during remission and recovery (see Austin et al., 2001, for an overview). There is a growing consensus in the literature that cognitive impairment seen during episodes of illness persists during episodes of remission (Trichard et al., 1995, Beats et al., 1996, Paradiso et al., 1997, Reischies and Neu, 2000), although not all studies have shown this (Calev et al., 1986, Bazin et al., 1994, Elliot et al., 1996).

It has been suggested that cognitive impairment worsens for every episode of depression (Brown et al., 1999, Sweeney et al., 2000) and that cognitive impairment observed in a nonsymptomatic phase is related to number of previous episodes (Kessing, 1998). The risk for a relapse after a first episode is estimated at 50%. For patients having a history of three episodes or more, the estimated relapse risk is increased to 90% (Kasper and Eder, 1994).

Knowledge about cognitive functioning in phases of remission can have implications for treatment, rehabilitation and risk of relapse. There is, however, no evidence that symptom reduction is followed by reduction in cognitive impairment to a similar degree.

Hartlage et al. (1993) have proposed a ‘cognitive effort hypothesis’, suggesting that cognitive deficit in depression is dependent on the difficulty of the task to be performed, with impairment increasing in accord with the cognitive effort the task requires. Demanding tasks, no matter what cognitive function they assess, will have serious detrimental effects in depressed patients. In this context, automatic processing of information is defined as a stimulus-driven process without invoking attention, while effortful processing is an instruction-driven process that requires attention and allocation of cognitive capacity.

This view is supported by other findings, which have shown that depressed patients are more impaired on tasks that require more effortful information processing compared to tasks that require more automatic information processing (Hashler and Zacks, 1979, Weingartner et al., 1981, Cohen et al., 1982, Roy-Byrne et al., 1986, Georgieff et al., 1998, Thomas et al., 1999). However, it is still unsettled whether this also holds true in patients tested longitudinally, during periods of symptom reduction as well as during acute phases of the illness.

The aim of the present study was therefore to investigate if impairment in effortful information processing shown in an acute phase of depression still existed in a period of symptom reduction at a 6-month follow-up.

A visual search paradigm was chosen in order to experimentally manipulate effortful and automatic information processing. In this task, the subject has to detect a target stimulus on a computer screen, where the target is surrounded by distractor stimuli. Visual search time is dependent on how basic stimulus features (i.e. color, orientation, size) are shared by the target and the distractors. When the target shares many features with the surrounding distractors, search for a target becomes more demanding and effortful (Duncan and Humphreys, 1989, Wolfe, 1998), with an increased search and response time. The time it takes to respond from the onset of the stimulus display (reaction time, RT) is presumed to indicate the level of cognitive effort involved. Using a paradigm with different levels of cognitive effort, we predicted that depressed patients would: (a) show increased RTs on trials requiring effortful but not automatic information processing, and compared to a healthy control group, (b) this would prevail at a second test after 6 months in the patient group, despite symptom reduction, and (c) that the healthy control group would perform equally at the two test occasions.