Tropicamide eyedrops cannot be used for reliable diagnosis of Alzheimer's disease

To evaluate the mydriatic effect of tropicamide eyedrops as a diagnostic test for Alzheimer's disease.

In a double-blind, placebo-controlled study, we assessed pupillary responses in 22 normal control subjects, 23 patients with probable Alzheimer's disease, 4 patients with isolated memory difficulty, and 6 patients with non-Alzheimer's dementia. Three separate studies were performed, the second and third on a subset of the original group. With use of infrared binocular pupillography, after 5 minutes of dark adaptation, we averaged pupil size during a 1-minute interval for baseline determinations. We then instilled 0.01% tropicamide into one eye. In the first two studies, we averaged pupil size for a 1-minute period at 5-minute intervals for 30 minutes, followed by a pupil light reflex test. In the third study, we measured pupil size every 5 minutes for 45 minutes and omitted the light reflex test.

No significant difference was noted in pupil dilatation between normal subjects and patients with Alzheimer's disease and between patients with non-Alzheimer's dementias and the Alzheimer's disease group in all three studies. Furthermore, on reperformance of the test in the same patients, more than 50% changed from a group above or below 13% pupil dilatation (a cutoff reported to distinguish Alzheimer's disease from normal control subjects) to the opposite group.

Results of this study indicate that pupil measurement after instillation of tropicamide cannot be used as a reliable diagnostic test for Alzheimer's disease. Moreover, test-retest reliability with use of dilute tropicamide eyedrops is questionable.