Comparison of alternate and original forms of the Montreal Cognitive Assessment (MoCA): an Italian normative study

The Montreal Cognitive Assessment (MoCA) is a screening test widely used in clinical practice and suited for detection of Mild Cognitive Impairment. Alternate forms of the MoCA were developed to avoid "learning effect" in serial assessments, and the present study aimed at investigating inter-form parallelism and at providing normative values for the Italian versions of MoCAs 2 and 3.

Three separate convenience samples were recruited: the first (n = 78) completed three alternate MoCA versions for ascertaining inter-form parallelism; the second (n = 302) and the third (n = 413) samples were administered MoCA 2 or 3 to compute normative data.

A three-step procedure complemented by confirmatory factor analysis and a mixed factorial ANOVA suggested that the three MoCA versions are not strictly parallel. Multiple linear regression analysis revealed that age and education significantly influenced MoCA 2 and 3 total scores. No significant effect of sex was found. From the derived linear equation, correction grids for MoCA 2 and 3 raw scores were built and equivalent scores computed. Inferential cutoff for adjusted scores, estimated using a non-parametric technique, were 17.49 for MoCA 2 and 18.34 for MoCA 3. Correlation analysis showed strong correlations of MoCA 2 (r = 0.69, p < .001) and MoCA 3 (r = 0.61, p < .001) adjusted total scores with MMSE adjusted scores.

The three MoCA forms are not strictly parallel. Specifically developed normative data must be adopted for using MoCA in serial cognitive assessments for clinical and research studies.