LAUSR

Self-reported cognitive difficulties or decline have been frequently associated with an increased risk of developing dementia (Jessen et al., 2014; Reisberg et al., 2010). However, other studies have noted no association between self-reported cognitive difficulties and decline and subsequent dementia (Jorm et al., 1997). There has also been significant debate as to the associations between self-reported cognitive difficulties and depression. Specifically, as to whether depression results in people reporting more cognitive difficulties or if people experience an increase in objective cognitive difficulties due to depression (e.g. Yates et al., 2015; Zlatar et al., 2014). The paper by Hill and colleagues (2019), investigating the temporal bi-directional associations between self-reported memory and depressive symptoms, covers two of the important potential issues in investigating these associations, the differing methods used in measurement of self-reported cognitive difficulties and the temporal association with depressive symptoms. By addressing the associations of self-reported memory with depressive symptoms overtime, they are able to highlight the importance of considering how self-reported cognition is measured as well as its complex association with depression. These are key considerations that need to be addressed if self-reported cognition is going to be a useful tool in predicting future cognitive decline or dementia. Many different approaches have been used to assess self-reported cognitive difficulties, so much so that even the terminology used can be somewhat controversial. The most common cognitive domain considered is memory but even within this single domain, there are wide variations in how self-reported memory is assessed. These methods range from single items to complex multiquestion measures and from asking people to report on their current memory, to compare themselves to others or to reflect on changes over specific time periods (Rabin et al., 2015). These variations may go some way in explaining the disparities in findings. Hill and colleagues address this issue by including three methods of assessing subjectivememory – frequency of memory problems, perceived 1-year decline, and perceived 10-year decline. This allowed them to investigate the differences in methods of assessing self-reported memory in relation to depressive symptoms. Crosssectional studies have noted that the measurement of self-reported cognitive difficulties can influence whether they are associated with mood and with cognitive function. Opdebeeck and colleagues (2019) reported different associations of self-reported cognitive difficulties with objective cognition and mood by style of question asked and by cognitive domain, with more consistent associations of self-reported cognitive difficulties with mood than with cognition. The study by Opdebeeck and colleagues was cross-sectional so while it could highlight the importance of considering the method of assessing self-reported cognitive difficulties, it could not provide any information on the temporal associations or directionality. A key issue in definitively identifying whether selfreported cognitive difficulties are to prove a useful predictor of future risk of dementia is in detangling the complex associations between self-reported cognition and depression. The study by Hill and colleagues set out to address this issue using data from the Einstein Aging Study, a longitudinal cohort study of community-dwelling people aged over 70. The study included 1,163 participants who had no clinical diagnosis of mild cognitive impairment or dementia during the study. There were up to 11 waves of data for each participant, and the sample represented a relatively diverse population. As mentioned above, self-reported memory was assessed in three different ways, while depressive symptoms were assessed with the commonly used 15-item Geriatric Depression Scale (GDS-15; Sheikh & Yesavage, 1986). The total GDS was calculated with 14 questions with the item “Do you feel you have more memory problems than most?” excluded to avoid conflating self-reported memory with depressive symptoms. Multilevel linear modeling adjusted for age, sex, education, race, and income was used to assess the temporal associations between the three measures of self-reported memory and depressive symptoms. There weremedium associations between the three self-report memory items at baseline, indicating that while they are certainly related, these measures likely taped into different aspects of self-reported memory. At baseline, participants who reported higher frequency of memory problems or memory decline in comparison to 1 and 10 years ago reported higher levels of depressive symptoms than those who did not report memory decline (Hill et al., 2019). International Psychogeriatrics (2020), 32:6, 681–683 © International Psychogeriatric Association 2020