Background: Different rates of cognitive progression have been observed among Alzheimer disease (AD) patients. The present study aimed at evaluating whether the rate of cognitive worsening in AD may be predicted by widely available and easy-to-assess factors. Methods: Mild to moderate AD patients were recruited in the ICTUS study. Multinomial logistic regression analysis was performed to measure the association between several sociodemographic and clinical variables and 3 different rates of cognitive decline defined by modifications (after 1 year of follow-up) of the Mini Mental State Examination (MMSE) score: (1) "slow" progression, as indicated by a decrease in the MMSE score ≤1 point; (2) "intermediate" progression, decrease in the MMSE score between 2 and 5 points; and (3) "rapid" progression, decrease in the MMSE score ≥6 points. Results: A total of 1005 patients were considered for the present analyses. Overall, most of the study participants (52%) exhibited a slow cognitive course. Higher ADAS-Cog scores at baseline were significantly associated with both "intermediate" and "rapid" decline. Conversely, increasing age was negatively associated with "rapid" cognitive worsening. Conclusions: A slow progression of cognitive decline is common among AD patients. The influence of age and baseline cognitive impairment should always be carefully considered when designing AD trials and defining study populations.