Developing Intelligence

Aging is associated with some slow but measurable forms of cognitive decline, but there is debate over the type of cognitive changes taking place. A recent study by Rush, Barch & Braver uses a series of interesting tasks to clarify the nature of this cognitive decline. The results seem to show that changes in “context processing” – the ability to internally represent environmental cues to control thought and action – but not inhibition or processing speed underlie aging-related decrements in cognitive function. The work has implications for our understanding of and interventions for this change across the lifespan.

Previous work with the same AX-CPT task had always been interpreted to reflect “context processing” but due to the nature of this task, such decreases in performance might have also reflected failures to inhibit or cancel responses, or general slowing in processing speed, both of which are often cited as reasons for cognitive decline in old age.

To address these issues, Rush et al gave 51 younger (19 years of age on average) and 56 older (74 years of age on average) adults the AX-CPT task (to index context processing), a series of inhibition tasks (Stroop, no-go, stop-signal and a garden path paradigm), and a simple reaction time task to index processing speed.

(The details of these tasks will not be described here; I’ll just explain the interesting findings, and provide relevant methodology information where necessary.)

The Results:

Younger adults actually performed worse on certain trials (“AY” trials to be exact) of the AX-CPT task than older adults, replicating previous work on aging with the AX-CPT task. Theoretically, AY trials are special because they tend to “trip up” those subjects who are better able to represent context: the “A” context typically predicts a target response, and so subjects who better represent that context may be more likely to false alarm, or slow down, when they subsequently get a low-probability “Y” stimulus. In contrast, older and younger adults did not differ in accuracy to other trial types (“BX” trials), indicating this was not merely a general trend in terms of worse performance between younger & older adults.

After statistically controlling for differences in processing speed between the age groups, there were still age-related differences in “AY” trial errors and “BX” trial reaction times – indicating that these age related differences in context processing cannot be reduced to processing speed differences.

Age effects were also found on two of the inhibitory tasks, including the task requiring inhibition of an automatic reading response (Stroop), and the one involving inhibition of a prepotent manual response (Stop Signal). [No age effects were found on a go-nogo task nor another other task involving garden path sentences.]

Age-related change in context processing (as indexed by AY trials in AX-CPT) was larger than the age-related change on the Stroop task, but not significantly larger than the age-related change on the Stop Signal task. However, individual performance on these tasks was not correlated – indicating that both may have relationships to aging and yet measure dissociable constructs. Based on the lack of consistent correlations among the inhibition tasks, the authors conclude that inhibition is itself a multifactorial construct.

Similarly, based on the fact that AX-CPT showed numerically larger relationships with aging than other tasks (at least in terms of AY trial performance), the authors conclude that context processing is a powerful explanatory construct for understanding age-related decline in cognitive functioning.

The take-home message is that cognitive decline is a multi-faceted phenomenon, entailing decreases in both total processing speed, context-processing, and also some capacity that is required by the Stop Signal task (putatively referred to as inhibition).


  1. #1 Matt
    November 7, 2007

    Nice summary.It seems another facet in cognitive decline is metacognition. Older adults seem to experience a decline in the ability to accurately monitor and spontaneously control cognition during complex tasks.

  2. #2 Dan
    November 18, 2007

    I automatically translate this, in my mind, to “old people are out of touch with reality, but the rest of their cognition basically works fine”. Interesting stuff.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.