participants with SMC and those with no SMC (Jungwirth et al. 2004 ). Generally,
then, these studiesfind that those with cognitive complaints do not have lower
memory performance than those without such complaints. However, in both
studies, those with SMC had more depressive symptoms. One limitation is that the
memory tests in both studies were quite easy, resulting in a restricted range of
scores (i.e., many subjects could obtain the maximum score on the tests). This
potentially limits the ability tofind a true relationship.
Other studies dofind a relationship between SMC and memory performance.
The Amsterdam Study on the Elderly (AMSTEL), involving 2537 non-depressed
and non-demented participants aged 65–85 years, found that SMC, as assessed by a
general memory rating and 3 specific problems, were associated with worse verbal
memory and overall orientation, after controlling for age, sex, and pre-morbid
intellectual function (Jonker et al. 1996 ). A particularly informative study on the
relation between SMC and memory performance involved 16,964 women (mean
age 75 years) enrolled in the Nurse’s Health Study (Amariglio et al. 2011 ). SMC
were assessed by 7 yes/no questions. SMC, as measured by the sum of endorsed
items, was strongly associated with worse performance on a global cognitive
measure and delayed verbal memory (Amariglio et al. 2011 ). Indeed, for every
additional item endorsed, the risk of cognitive impairment was increased by 20%.
Similarly, in a study of 2389 unimpaired primary care patients, SMC, as assessed
by a single rating of memory decline and 4 questions about everyday memory
function, were associated with worse verbal memory, but only in those with few
depressive symptoms (Jessen et al. 2007 ). These results indicate that SMC are a
predictor of cognitive impairment. This pattern offindings was also evident in a
study of 263 primary care patients, suggesting that SMC relate to objective memory
declines in clinical settings where memory complaints are frequently brought to the
attention of a healthcare provider (Snitz et al. 2008 ).
SMC Assessed by Validated Questionnaires
Comprehensive measures of SMC sometimes relate to non-memory cognitive
functions. For instance, although the initial validation study utilizing the MAC-Q
(Crook et al. 1992 ) found that SMC were related to memory complaints in a sample
of 232 older adults, smaller studies found that SMC were related to verbalfluency
and processing speed, but not verbal memory (Minett et al. 2005 , 2008 ). Finally,
Amariglio et al. ( 2012 ) examined specific SMC subscales with the Everyday
Cognition (E-Cog) scale, a 39-item measure, scored on a Likert scale (1 =“Better or
no change”to 4 =“Consistently much worse”), which was specifically developed to
assess cognitive abilities in older adults (Farias et al. 2008 ). The E-Cog Language
correlated with the ability to name an object depicted on a sheet of paper (e.g.,
escalator) and E-Cog Attention correlated with working memory performance
(Amariglio et al. 2012 ), suggesting specificity to the relationship between SMC and
cognitive function.
13 Subjective Memory Complaints and Objective Memory Performance 285