Effects of gender on response to treatment with rivastigmine in mild cognitive impairment: A post hoc statistical modeling approach

Steven Ferris, Roger Lane, Nikolaos Sfikas, Bengt Winblad, Martin Farlow, Howard H. Feldman

Research output: Contribution to journalArticle

14 Scopus citations


Background: Epidemiologic studies have identified several demographic factors, including gender, that may influence the progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD). Objective: This analysis aimed to develop a sensitive model for detecting treatment benefits in patients with MCI by controlling for factors that predict progression to AD. The study used this statistical modeling to investigate the effect of gender on treatment response in patients with MCI. Methods: This post hoc analysis used data from the InDDEx (Investigation in Delay to Diagnosis of Alzheimer's disease with Exelon) study, a long-term (3- to 4-year), multicenter, randomized, double-blind, placebo-controlled trial of rivastigmine 3- to 12-mg/d capsules in 1018 patients with MCI. Baseline variables that were significantly associated with progression to AD within the InDDEx study population were identified. A Cox proportional hazards multivariate regression model that adjusted for these predictive factors was applied to the overall study population and to the gender subgroups. Results: Of 31 baseline variables analyzed, 13 were found to be significantly associated with progression to AD in the overall population. After adjustment using the Cox proportional hazards regression model, rivastigmine was associated with a significantly lower risk for progression to AD compared with placebo in the overall population (1016 patients; hazard ratio [HR] = 0.747; P = 0.045). This effect of rivastigmine was evident in women (530 patients; HR = 0.676; P = 0.046) but not in men. Conclusions: In these patients with MCI, a significant decrease in the risk for progression to AD was found with rivastigmine over 3 to 4 years. The proportion of women whose disease progressed to AD was significantly lower in the rivastigmine group than the placebo group (P = 0.046). This effect was not found in men. These data suggest that prospectively adjusting for predictive factors could lead to a more accurate estimation of treatment benefits in future studies of MCI.

Original languageEnglish (US)
Pages (from-to)345-355
Number of pages11
JournalGender Medicine
Issue number2
StatePublished - Jul 1 2009


  • Cox proportional hazards regression analysis
  • gender
  • MCI
  • mild cognitive impairment
  • progression to AD

ASJC Scopus subject areas

  • Medicine(all)
  • Gender Studies

Fingerprint Dive into the research topics of 'Effects of gender on response to treatment with rivastigmine in mild cognitive impairment: A post hoc statistical modeling approach'. Together they form a unique fingerprint.

  • Cite this