Ascending Digits Task as a Measure of Executive Function in Geriatric Depression

The sample included 258 subjects, equally divided between the nondepressed comparison (N=129) and depressed (N=129) cohorts. When matched against depressives, comparison subjects were comprised of a higher percentage of women (73% versus 59%; χ²=6.26, p=0.012), were older (76.2 versus 73.4 years; t= [256], df=[3.25], p=[0.0013]), and reported more years of education (15.8 versus 14.2; t=[256], df=[4.28], p=[<0.0001]). Both groups were largely Caucasian (84% versus 90%; χ²=2.17, p=0.14). Comparison subjects scored higher on the MMSE than patients (28.72 versus 27.51; t=[205], df=[4.63], p=[<0.0001]). On the ADT, comparison subjects scored 1.75 units higher than patients (9.52 versus 7.75; t=[245], df=[5.74], p=[<0.0001]). On average, depressed participants had an average MADRS score of 22.33 (SD=7.77), a lifetime total of 5.6 (SD=10.3) depressive episodes, and were 25.2 (SD=29.1) months into the current episode.

In a multivariate regression analysis controlling for age, sex, race, education and MMSE score, depressed subjects scored significantly lower on the ADT relative to comparison subjects (Table 1).

In a separate regression analysis restricted to the depressed cohort, the MADRS score was not significantly associated with patient performance on the ADT (p=0.58).

We have demonstrated that older depressed patients perform worse on the ADT compared with elderly comparison subjects. We also found that although a diagnosis of depression was associated with a significantly lower ascending digits score, the latter was apparently not influenced by depression severity when we examined the depressed group separately.

The ADT involves manipulation of information that is thought to be carried out by the central executive system of working memory. In Baddeley’s model of working memory, the central executive system controls attentional resources to two slave systems, the phonological loop, concerned with speech-based information, and the visuospatial sketchpad, which is involved with visual images.⁴ DSF is thought to involve the phonological loop, while DSB is thought to involve the central executive to some degree.^8,9 Because these more commonly used digit span tests do not involve a more complex manipulation of information, the ADT may be better to assess executive function.

There are several potential limitations that need to be addressed. First, it is possible that worse performance on the ADT is due to impairment in attention or motivation. When cognitive demands pass a threshold, depressed individuals may shut off their attentional or motivational mechanisms earlier than comparison subjects. While we did not examine attention or motivation in this study, previous studies have demonstrated that performance on the DSF task is no different between depressed individuals and comparison subjects, thus indicating that attention or motivation is not impaired in depression during this task.^8,10,11

In addition, depression encompasses a variety of symptoms. It is possible that performance on the ADT is impaired in depressed individuals with a certain subset of depressive symptoms, for example those that complain of diminished ability to concentrate. In a post-hoc analysis, we examined the relationship of the MADRS concentration item and ADT score and found no significant association (Pearson r= −0.042, p=0.60). We have also shown that performance on the ADT was not dependent upon the severity of depression as measured by MADRS scores. This is consistent with previous findings that performance on two working memory tasks carried out simultaneously was not related to severity of depression but to the presence of depression.¹²

Additionally, we did not examine the total length of time in which a subject was depressed, nor did we examine the number of recurrences each patient had. There is some evidence that subjects who have recurrent depression perform worse on memory tests than those in depression for the first time.¹³ This, and the length of time a person has been depressed, may contribute to cognitive dysfunction through anatomical changes that accumulate. For example, the hippocampus¹⁴ and the amygdala¹⁵ have been shown to be reduced in size in depressed individuals matched against comparison subjects.

Executive dysfunction in older depressives may be due to degenerative, anatomical frontal lobe changes. Alternatively, the anatomical integrity associated with normal executive function may be maintained, but frontal lobe systems may be either unable to be activated or are inhibited through an unknown mechanism. While the pathophysiology of frontal lobe dysfunction is not known, it may have features of both a degenerative process and a disconnection syndrome. For example, remission from depression has not been shown to improve performance in certain memory tasks.¹² On the other hand, a “Depression–Executive Dysfunction Syndrome” has been described that supports the notion of frontal pathway disruption.³

In conclusion, our study demonstrates that depressed individuals perform worse on the ADT matched against comparison subjects. We speculate that this difference may be due to a deficit in central executive function, as the task involves manipulation of information held in working memory. As studies have demonstrated frontal lobe dysfunction in depressed individuals that correlate with deficits in executive functioning, the ADT may be a useful tool to explore frontal dysfunction in depressed patients. Functional neuroimaging studies using this task may shed further light into the exact nature of executive dysfunction associated with depression.

This study was supported by NIMH grants P50 MH-60451, R01 MH-54846, and K24 MH-70027.