ABSTRACT
Introduction: The Prospective and Retrospective Memory Questionnaire (PRMQ) is one of the most commonly used scales to assess both retrospective memory (RM) and prospective memory (PM) complaints. This study aimed to: 1/replicate the previous results concerning the PRMQ latent structure in a French version and 2/provide its psychometric properties in a normal and clinical population.
Method: This observational study included 488 participants divided into five subgroups. A sample of 168 healthy participants (no memory consultation sought), served as controls. Patients were recruited in a memory clinic: 98 “functional” patients (subjective memory complaints but no memory impairment), 83 amnestic-Mild Cognitive Impairment (a-MCI), 82 non-amnestic-MCI (na-MCI) and 57 Alzheimer Disease (AD) patients. Structure, validity, consistency, reliabilitiy and reproducibility of the PRMQ were calculated. Novelty, Area Under the Receiver-Operating Characteristics (AUROC) curve, was used to determine the optimal cut-off, to distinguish “functional” patients from control participants.
Results: The optimal fit model of the French PMRQ was not a tri but a bi-partite model, with a RM and a PM subscale. The convergent validity showed significant correlation with cognitive difficulties (r = .82 and .78, respectively), anxiety (r = .44 and .48, respectively) and depression (r = .23) scales. Cronbach’s alpha was good (α = .79 and .88), as well as the reproducibility (r = .71 and .80). The interaction [Subgroups of participants x PMRQ Subscales] was significant [F(4, 483) = 11.46; p < .001]. The power discrimination was adequate (AUROC = .71 and .74) for detecting “functional” patients compared with controls, in particular for the PM subscale (sensitivity 66.6%, specificity 77.4%).
Conclusions: The PMRQ, with minor changes, was validated in its French form with satisfactory psychometric qualities. This self-rating tool appears useful for identifying significant memory complaints in a normal population and may also be helpful in discriminating between functional/na-MCI and a-MCI/AD patients.
Acknowledgments
The authors are grateful to Mary Harries for her assistance in English, to Jean-Dominique Mouchard and Pr. Marc Ychou for their trust. They also thank Drs. Pierre Senesse, Amélie Darlix, Vanessa Guillaumon, Jean-Pierre Bleuse and Pr. Grégory Ninot for their support.
Disclosure statement
No potential conflict of interest was reported by the authors.