Spatial shifts in visual attention in normal ageing and dementia of the Alzheimer type

https://doi.org/10.1016/S0028-3932(02)00057-XGet rights and content

Abstract

Using a spatial-cueing paradigm, we assessed the ability of Alzheimer’s disease patients, age-matched controls and younger participants to use cues to guide attention to the location indicated by the cue. In separate experiments, we attempted to isolate cues that attract attention automatically (exogenous cueing) and those that require the wilful movement of attention (endogenous cues). We found significant cueing effects for all three groups of participants for both types of cue. However, the group with Alzheimer’s disease showed far greater cueing effects when using an exogenous cue, whilst no difference between group’s ability to use the cue was found for the endogenous cue. No differences in cueing were found for either cue type as a function of normal ageing. We further tested whether the differences in cueing found in the group with Alzheimer’s disease was due to a generalised slowing of function. After transforming the data to take account of the overall slowing of all responses in this group, we still found significant differences between this group and the control groups. We conclude that patients with Alzheimer’s disease have an abnormality in automatic, but not controlled visuospatial attention.

Introduction

Whilst dementia of the Alzheimer type (DAT) is often considered an impairment of memory, it is now recognised that there are also disturbances of attentional processes which may be apparent at early stages of the disease [34], [36], [37]. Some changes in attentional processes also appear to accompany the normal ageing process [9], [14]. Thus, we are left to consider whether the impairments in function seen in DAT can be considered extensions of the normal ageing process, or whether they differ in some fundamental manner. In other words, are differences for a particular task or psychological function (if there are any) quantitative or qualitative? In a similar vein, it has often been shown that older participants tend to perform tasks slower than younger controls. It has been suggested, therefore, that many results in the literature can be accounted for by a generalised slowing, rather than an impairment in a specific psychological function [29], [46]. Thus, for any particular task, we should establish whether any function that we claim may be impaired in either normal ageing, or in DAT, is impaired beyond what may be predicted from a generalised slowing of function.

A popular method for assessing attentional selection is the spatial-cueing paradigm introduced by Posner [40]. In this paradigm, the participant responds to targets that can occur at one of two locations on either side of the fixation mark. Before the target appears, one of these locations is cued in some manner so that ‘attention’ may be focused at this location. It is commonly found that participants are better at detecting (as indexed by reaction time or sensitivity measurements) targets at this cued location in comparison to the uncued location. In this paradigm, the eyes are always fixated on the central mark, hence it is the covert movements of attention that are being assessed.

Within this paradigm, there is a distinction made between the types of cue used. First, the participant can wilfully move their attention to a location because they believe that the target is likely to occur at this location. Typically, this is done by a cue occurring at a central location (such as an arrow) that signals to the participants that the target is most likely to occur at the location indicated. Therefore, the participant must understand this cue, and deliberately move their attention to the indicated location. Given the pattern of deficits reported by DAT patients, any failure to use such a cue may not be due to any deficit in the attention system per se, but due to an inability to understand the nature of the cue or to remember the meaning of the cue throughout the experiment. Indeed, the effect of such a cue can be eliminated in ‘normal’ participants if the participant’s resources are elsewhere employed [16]. Such a cue that solely engages the wilful movements of attention is commonly referred to as an ‘endogenous’ cue in the recent research literature [50].

Second, some events (such as a flash of light) seem automatically to attract attention even if there is no a priori reason why the participant should move their attention. This is often operationalised by flashing one of the location boxes (a peripheral cue). To make sure that the cue does not attract any wilful processing the cue is not predictive of the target’s location. Such a cue that automatically attracts attention is termed an exogenous cue [50]. Such a cue is still effective even if the participant’s resources are employed elsewhere [16]. There is evidence that endogenous and exogenous cues have separate and additive effects [42].

Not surprisingly, the spatial-cueing paradigm has been used to assess attentional function in both normal ageing and in DAT. One influential study [33] compared DAT patients with age-matched controls. The participants had to make a judgement about a target letter (was it a consonant or a vowel?) that occurred at one of two locations. Cues that occurred at a central location (an arrow) or at a peripheral location, and at a range of cue to target time intervals, were tested. They found that the normal difference between the cued and uncued location was exaggerated for some conditions in the DAT patients. For the peripheral cue, the greater effect of cue was present only for short (200 ms) intervals between the cue and target, whereas the exaggeration was seen only at the long time interval (2000 ms) for the central cues. The results were interpreted as indicating an abnormality of attention in DAT for both automatic and effortful shifts of attention (mapping on to the exogenous and endogenous distinction drawn above). The cue–target intervals when the abnormality is seen is in line with the known properties of these systems [28]. Exogenous shifts of attention are fast, but short-lived (hence, an abnormality would only be seen at short cue–target intervals), endogenous attention is slow but sustained (hence, an abnormality would only be seen at long cue–target intervals). There are, however, a number of possible problems with this interpretation we believe warrant further experimentation.

First, in the study of Parasuraman et al. [33], the peripheral cue was also predictive of the target’s location, hence it did not isolate the exogenous system. Whilst several studies show that purely endogenous attention is absent at short cue–target intervals [14], [26], [33], others have shown that an endogenous component can modify the effects of an exogenous cue even at very brief cue–target intervals [23], [42], [54]. Therefore, deficits in attention observed at brief cue–target intervals for a peripheral cue cannot be interpreted unambiguously to show a deficit in either (or both) attention system(s).

Second, whilst the results show a greater cueing effect in some circumstances for the DAT patients, does this necessarily indicate a specific problem of attention? As we noted earlier, there is substantial evidence for a generalised slowing of function in DAT, and it is noticeable that reaction times for the DAT patients of Parasuraman et al. [33] (irrespective of any cueing condition) are considerably larger than the controls. Could a generalised slowing of function predict changes in the magnitude of the cueing effects in the covert paradigm in the absence of any specific attentional dysfunction? The answer is yes [47]. If the target appearing at the uncued location takes longer to capture attention (due to the generalised slower processing) then the effect of the cue will be greater.2 Thus, in the presence of overall changes in RT, a greater cueing effect per se does not necessarily indicate any specific dysfunction of attentional systems. However, we note that such an argument cannot predict both conditions where greater cueing effects occurred and conditions where no such effects occurred.

Third, Parasuraman et al. [33] also performed an experiment in which the task of the participant was to merely detect the onset of the target (as opposed to making some judgement about the letter). This paradigm more closely followed the original studies of Posner et al. [41] and many others. Only central cues were assessed. Under these circumstances, no significant differences in cueing were found between the DAT patients and the controls. This might indicate that endogenous cueing is not abnormally affected in DAT. However, as pointed out by Parasuraman et al. [33], such detection tasks may require less attentional resources to be allotted to target processing, and thus, they may be a less sensitive assay of attentional dysfunction. Such a position is supported by results from Brawn and Snowden [4] that show greater effects of cueing when comparing discrimination as opposed to detection tasks. Nevertheless, if attention is affected in DAT, then detection tasks should also be predicted to reveal this change if they are sensitive enough. Our experiments, therefore, included both detection and discrimination tasks.

Since the study of Parasuraman et al. [33], several other studies have also examined the spatial-cueing paradigm and DAT. Oken et al. [32], using a discrimination task, and Buck et al. [5] and Danckert et al. [8], using a detection task, also indicate abnormal attentional processes. However, other studies using only detection tasks failed to support this notion [6], [9]. Again, the distinction between exogenous and endogenous attention was blurred in these studies, and so one cannot attribute deficits (or lack of them) to one system or the other (or both).

Maruff and Currie [21] used a peripheral cue, but varied the validity (the percentage of trials on which the target occurred at the cued location) of this cue from 80 through 50 to only 20%. They found greater cueing effects in DAT patients for both 80 and 50% valid cues. When the cue is only 50% valid, the experiment should isolate the exogenous system and, therefore, the result here indicates an abnormality for DAT patients with this system. For a cue that is 80% valid, both the exogenous and the endogenous systems are working together, so the greater cueing effect found for the DAT patients could be due to the action of the exogenous system alone, or a combination of deficits from both systems. Finally, when the cue is only 20% valid, the two systems are in opposition. Under these circumstances, the control participants showed a reversal of the effect so that they were faster for targets occurring opposite to the cue. This can be interpreted as the endogenous information being more effective than the exogenous information. The DAT patients did not show such a reversal, and one could interpret this as suggesting that the exogenous information was more effective than the endogenous information. Given that the results from the 50% validity condition indicate a stronger exogenous effect for the DAT patients, this could explain the results for the 20% conditions. Once again, this study does not consider the overall differences in RTs when considering these effects. Therefore, a generalised slowing could explain the results for the exogenous alone cue alone, but could not easily explain the overall pattern of results.

Section snippets

Cost and benefit analysis

In many studies of covert attention, the effects of the valid and invalid cues are compared to a neutral cue, allowing calculation of the benefits of the valid cue (neutral–valid RTs) and the cost of the invalid cue (neutral–invalid RTs). However, the interpretation of such an analysis is difficult. Indeed, some authors specifically recommend including no such neutral condition ([17], p. 41). Instead, recent studies simply calculate ‘validity’ effects by comparing the effects of the valid and

Participants

Three groups of individuals participated in the study: one in which individuals had a diagnosis of mild to moderate probable Alzheimer’s disease, an age-matched healthy control group, and a young control group. The Alzheimer’s (DAT) group was recruited from the Memory Clinic at Llandough Hospital, Cardiff and originally consisted of 19 individuals. However, seven of these people were unable to perform the tasks reliably enough for their data to be included in the analysis. The DAT group,

Results

Because not all participants were able to perform all the tasks, it was not possible to perform a direct comparison of detection and discrimination tasks and the statistical analysis is constructed accordingly.

General discussion

Our study aimed to examine separately attention shifting due to exogenous cues (automatic shifts) and to endogenous cues (wilful shifts) in DAT patients and in the normal ageing process. We found no evidence that DAT patients differ from age-matched controls for endogenous cues; however, we found a greater effect of an exogenous cue. This pattern of results was consistent across the two tasks used. Although we confirmed the commonly reported slowing of all reaction times with age, we found no

Acknowledgements

This project was funded by an Alzheimer’s Society Research Fellowship to Dr. A. Tales.

References (56)

  • G. Rizzolatti et al.

    Reorienting attention across the horizontal and vertical meridians: evidence in favour of a premotor theory of attention

    Neuropsychologia

    (1987)
  • D.L. Robinson et al.

    Visuospatial attention: effects of age, gender, and spatial references

    Neuropsychologia

    (1990)
  • G. Tassinari et al.

    Covert orienting to non-informative cues: reaction time studies

    Behavioural Brain Research

    (1995)
  • M.L. Voytko

    Cognitive changes during normal aging in monkeys assessed with an automated test apparatus

    Neurobiology of Aging

    (1993)
  • M.G. Baxter et al.

    Spatial orienting of attention in adult and aged rhesus monkeys

    Behavioral Neuroscience

    (1996)
  • P.T. Brawn et al.

    Can one pay attention to a particular colour?

    Perception and Psychophysics

    (1999)
  • P.T. Brawn et al.

    Attention to overlapping objects: detection vs. discrimination of luminance changes

    Journal of Experimental Psychology: Human Perception and Performance

    (2000)
  • B.H. Buck et al.

    Spatial- and object-based attentional deficits in Alzheimer’s disease. Relationship to HMPSO-SPECT measures of parietal perfusion

    Brain

    (1997)
  • P. Caffarra et al.

    Orientation of visual attention in Alzheimer’s disease: its implication in favor of the interhemispheric balance

    Neuropscyhiatry and Neuropsychology and Behavioral Neurology

    (1997)
  • Cohen J. Statistical power analysis for the behavioural sciences. 2nd ed. Hillsdale, NJ: Erlbaum,...
  • J. Danckert et al.

    Inhibitory processes in covert orienting in patients with Alzheimer’s disease

    Neuropsychology

    (1998)
  • M.E. Faust et al.

    Inhibition of return and visuospatial attention in healthy adults and individuals with dementia of the Alzheimer type

    Neuropsychology

    (1997)
  • W.A. Fletcher et al.

    Saccadic eye movement dysfunction in Alzheimer’s disease

    Annals of Neurology

    (1986)
  • C.L. Folk et al.

    Aging and shifts of visual spatial attention

    Psychology and Aging

    (1992)
  • F. Folstein et al.

    Mini-mental state, a practical method for grading the cognitive state of patients for the clinician

    Journal of Psychiatric Research

    (1975)
  • P.M. Greenwood et al.

    Controlling the focus of spatial attention during visual search: effects of advancing aging and Alzheimer disease

    Neuropsychology

    (1997)
  • A.A. Hartley et al.

    Age differences and similarities in the effects of cues and prompts

    Journal of Experimental Psychology: Human Perception and Performance

    (1990)
  • Jonides J. Voluntary versus automatic control over the mind’s eye’s movement. In: Long J, Baddeleys A, editors....
  • Cited by (0)

    1

    Present address: Department of Care of the Elderly, University of Bristol, Clinical Research Centre and Memory Disorders Clinic, The BRACE Centre, Blackberry Hill Hospital, Manor Road, Bristol BS16 2EW, UK.

    View full text