39 (0.08) <0.0001 0.21 (0.10) 0.0294  D11   21.71 (2.75) <0.0001 20.17 (3.39) <0.0001  D12   0.18 (0.07) 0.0070 0.04 (0.10) 0.6984  D22   0.01 (0.00) 0.0002 0.01 (0.00) 0.0073  Residual variance   5.67 (0.33) <.0001 5.43 (0.44) <0.0001 AD Alzheimer’s disease, D11 and D22 variance of subject-specific intercepts and Lenvatinib cost slopes, respectively, D12 covariance between subject-specific intercepts and slopes, FDur duration of follow-up, GDS Geriatric Depression Scale, MMSE Mini-Mental State Examination, MoCA Montreal Cognitive Assessment, SD standard deviation aIncluded as time-varying variable ‘Years of education’ was the only confounder with significance on the MMSE, as well as the MoCA scores. Based on MMSE,

pure AD patients seemed to be less cognitively impaired at baseline (2.36, p = 0.023), but this difference was not significant in the multivariable analysis after adjusting for years of education (1.48, p = 0.156). There was a slight decrease in MMSE scores over time (−0.04, p = 0.007), and the decrease over time was similar for NVP-BGJ398 nmr both diagnosis groups (−0.03, p = 0.246). The annual estimated mean reduction of MMSE score was less than 1 for both the pure AD (0.84) and the mixed AD (0.48) groups. Similar trends were observed based on the MoCA scores, with annual estimated mean reduction of 0.72 and 0.48 for pure AD and mixed AD groups, respectively

(Table 3). For both MMSE and MoCA scores, the variance of the patient-specific intercept was ‘large’ (>20), indicating that the severity of cognitive impairment at baseline varied substantially from patient to patient. This was expected in data obtained from clinical practice, unlike randomized controlled trial data. The small variances of the patient-specific slopes indicated that the reduction

in cognition over time was similar from patient to patient, and the reduction in cognition did not depend on the severity of cognitive impairment at baseline, as indicated by the small covariances between the patient-specific G protein-coupled receptor kinase intercepts and slopes. These trends were similar for the base, univariable, and multivariable models. 4 Discussion In our study of a clinical cohort of patients with AD, we found that cognitive enhancers are effective in slowing the rate of cognitive decline in both patients with pure AD and those with mixed AD. Importantly, there was a trend to greater cognitive benefit, characterized by a slower rate of cognitive decline in patients with mixed AD than in those with pure AD. The results remain significant even after adjusting for years of education and inherent variability in the severity of cognitive decline between patients. Both the MMSE and MoCA demonstrated a trend towards cognitive benefit for patients with mixed AD when treated with cognitive enhancers. MMSE and MoCA were both validated for screening and monitoring of AD, with the MoCA found to be a better cognitive tool than MMSE [31].