According to the growing clinical interest in early indicators of dementia,
numerous studies have examined the association between subjective memory c
omplaints and cognitive performance in old age. Their results are contradic
tory. In this paper, studies carried out over the last 10 years are compare
d with regard to the study design and the assessment instruments used. The
results are discussed with particular reference to the diagnostic validity
of subjective memory complaints. The majority of case-control studies and c
ross-sectional studies of non-representative samples could not demonstrate
an association between subjective memory complaints and cognitive performan
ce. Most field studies of larger representative population samples, however
, have come to the opposite conclusion. A consistent assessment of these st
atistically significant associations against the background of diagnostic v
alidity showed that memory complaints cannot be taken as a clear clinical i
ndicator for cognitive impairment. Subjective memory complaints may reflect
depressive disorders and a multitude of other processes, of which an objec
tive impairment of cognitive performance is just one aspect. As a consequen
ce, an inclusion of subjective memory complaints as a diagnostic criterion
for the diagnosis of "mild cognitive disorder" according to ICD-10 is not j
ustified.