to the editor: We were interested to read the useful review of the dementia work-up by Drs. Adelman and Daly in American Family Physician.¹ However, the authors did not mention executive function testing, which helps greatly in differentiating the various causes of dementia.²

Executive function is defined by the Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV) as the cognitive process that organizes simple ideas, behaviors, and affects into complex actions, the one best action for the environmental cue, and the right step for the goal.³ The more commonly employed Mini-Mental State Examination (MMSE) serves as an adequate screen for Alzheimer’s disease but misses the subcortical dementias that do not exhibit early language deficits.

Subcortical dementias include, but are not limited to, the motor neurologic deficit dementias such as Parkinson’s disease, Huntington’s disease, and normal-pressure hydrocephalus. They distinguish themselves by their early motor deficits and late appearance of language deficits. The characteristic memory loss is subtler than the cortical dementias (such as Alzheimer’s disease) and involves forgetfulness rather than amnesia.

The Executive Interview (EXIT25) is a 25-item bedside examination of executive function that takes 15 minutes for a trained lay examiner to administer. The higher the score, the greater the executive dysfunction. Each item is scored zero for a perfect answer, 2 for an incorrect answer, and 1 if the patient needs prompting to obtain a correct answer. For example, if a subject were to get totally incorrect answers on the 25 items, he or she would obtain a score of 50. Executive function worsens with age, so the cutoff for normal older persons is 10 to 15 out of 50 points, whereas the cutoff for young adults is 5 out of 50 points. EXIT25 scores correlate with competency, behavior problems,⁴ diabetes control, and level of institutional care.⁵ Scores in the high 20s correlate with the nursing home level of supervision required for patients’ cognitive deficits.

Tests of executive function (e.g., EXIT25) identify early, subtler subcortical dementias such as the vascular dementias and the aforementioned neurologic disorders. Other type I or subcortical dementias (e.g., apathetic depression, irreversible sequelae of vitamin B₁₂ dementia, hypothyroidism, late return to baseline from severe medical illness, or polypharmacy) all present with a similar clinical pattern.⁴ We estimate that as many as 40 percent of dementias may go undiagnosed with a work-up relying primarily on the MMSE.

Tests of frontal-subcortical system disease are supported by histopathologic and positron emission tomography and single-photon emission computed tomography studies that reveal early destruction of the frontal-subcortical pathways in subcortical dementia.⁶

In patients with memory problems that respond to cueing and intact language skills, but with motor deficits, the appearance of executive dysfunction suggests subcortical dementia. The EXIT25 screening instrument helps delineate the dementia subtype but also aids in determining the setting of care. We have found executive function testing to be an invaluable tool in our approach to the dementia work-up.

in reply: We read the letter from Dr. Gershman and colleagues with interest. They propose that a scale to measure executive function may help differentiate the various types of dementias, especially the subcortical dementias that “do not exhibit early language deficits.” The Mini-Mental State Examination (MMSE) can be a useful screening test in patients with dementia; however, it is not highly sensitive, especially in persons with early disease and those with high levels of premorbid cognitive functioning. Because language impairment is usually a late phenomenon in patients with Alzheimer’s disease, the questions on the MMSE related to language are among the least helpful for identifying mild and moderate disease. In addition, language is preserved early in patients with subcortical dementias.

For our article on initial evaluation of dementia,¹ we examined mental status examinations that had test characteristics (i.e., specificity, sensitivity, and predictive value) available. Our review of readily available literature does not indicate that the Executive Interview (EXIT25) has been well evaluated in these respects. EXIT25 does appear to uncover impairment in patients with normal MMSE scores.^2,3 There is insufficient data to recommend the EXIT25 as a screening tool for dementia. Also, the EXIT25 requires training and takes up to 15 minutes to administer, reducing its utility in a primary care office setting.

Neurocognitive testing is useful in patients with suspected dementia who score well on the MMSE or who have high levels of premorbid cognitive functioning. This type of testing is not always readily available and is not covered by most insurance plans. These tests also take two or more hours to administer.

The EXIT25 potentially may be a useful instrument to augment established screening tests for patients with possible dementing illnesses. For patients who score in the “normal” range on the MMSE, the EXIT25 may be sensitive enough to identify patients with dementia and may obviate the need for comprehensive neuropsychological testing.