Background: It is generally accepted that Alzheimer's disease (AD) is mainly characterized by memory disorders. Although recent studies also point to an important role of attention deficits early in the disease, this notion has not yet emerged in clinical practice. Our aim was to assess whether attention, quantified by reaction times, can discriminate between patients with mild AD and controls and therefore contribute to clinical diagnosis.

Methods: In a cross-sectional study, 33 patients with mild AD were matched with cognitively healthy elderly controls for age, gender, educational level and depressive mood. Selective attention (SA), alternating attention (AA) and error-rates were measured by a modified reaction time test.

Results: Significant differences between both groups were found for all measures. Logistic regression showed that SA (corrected for individual processing speed) and error-rates could correctly classify subjects with an overall hit ratio of 81%. When attention measures were not corrected for individual processing speed, the overall hit ratio improved to 97%.

Conclusion: SA and AA deteriorate in patients with mild AD and these measures can be used to discriminate between patients and matched controls, independently of depressive mood.