cognitive decline on a 3-word recall measure 5 years later, after adjusting for
baseline cognitive performance and depressive symptoms (Wang et al. 2004 ). In a
study of 169 older adults (average age 75), SMC, assessed by a single question
about the presence of memory problems, did not predict verbal memory 1 year later
in participants that were cognitively normal at baseline. However, for cognitively
impaired participants, SMC at baseline were associated with greater decline in
verbal recall at follow-up. In regression models adjusting for age and education in
cognitively impaired participants, SMC were associated with a fourfold increase in
risk of cognitive decline (Schofield et al.1997a). Similarly, in a study of 507
community-dwelling older adults (≥70 years), SMC and objective cognitive
function were assessed at baseline and 3.5 years later. SMC included global cog-
nitive complaints (affirmative responses to two questions—“Have you experienced
a decline in memory?” and“Have you experienced a decline in thinking and
reasoning?”) and subjective memory decline (Likert ratings on 4 questions about
memory problems in specific areas of everyday life). Objective cognitive perfor-
mance was measured by a global screening tool, a test of processing speed, and 4
brief memory tests (3-word recall, name and address recall, face recognition, and
figure reproduction). Both types of SMC did not predict change on cognitive tests
over the follow-up period after controlling for depression and anxiety (Jorm et al.
1997 ). However, when those data were later reanalyzed using structural equation
modeling (Jorm et al. 2001 ), which enables the detection of small effects, SMC did
predict scores 1 year later. In an extension of the Jorm et al. ( 1997 ) study that
entailed a third evaluation 4 years later, SMC were not associated with current
performance, but predicted future memory performance (Jorm et al. 2001 ).
There is also some evidence that SMC may predict decline in cognitive domains
other than memory. In a study of 786 participants in the Longitudinal Aging Study
Amsterdam, SMC, assessed by a single question about the presence of memory
problems, were associated with decline in delayed recall and processing speed, but
not immediate memory, after controlling for age, sex, education, and depression
(Dik et al. 2001 ).
SMC Assessed by Validated Questionnaires
SMC, as assessed by the frequency of forgetting and mnemonics usage, were not
related to change in objective verbal memory over 5 years in the large ACTIVE
trial (Advanced Cognitive Training for Independent and Vital Elderly). This study
involved 1401 older adults between the ages of 65 and 94 years who were in the
memory training and control (no training) groups (Parisi et al. 2011 ). In contrast,
other studies using validated measures of SMC reveal an association between SMC
and cognitive decline over time. In the Epidemiology of Vascular Aging
(EVA) study of 555 older adults aged 59–71 years, SMC, as assessed by total score
on the Cognitive Difficulties Scale, were associated with decline on global cogni-
tion,fine motor speed and dexterity and processing speed 2 years later, independent
13 Subjective Memory Complaints and Objective Memory Performance 289