Clock drawing test in mild and moderate dementia of the Alzheimer's type: a comparative and correlation study

(a) To compare two different clock drawing tests (CDTs) in mild and moderate dementia of the Alzheimer's type (DAT); (b) To examine presumed correlation between these CDTs and some demographic, cognitive and activities of daily living (ADL) variables in mild and moderate DAT.

Cross-sectional study. Psychogeriatric outpatient clinic. 49 DAT patients, total; 26-mild, 23-moderate, mean age 77.8 and 80.6, respectively. Evaluations included the Mini-Mental State Examination (MMSE), the Cambridge Cognitive Examination (CAMCOG), the Instrumental Activities of Daily Living Scale (IADL), and a Basic Activities of Daily Living (BADL)-dressing subscale. Severity of dementia was determined with the Clinical Dementia Rating (CDR). Each clock was blindly scored by the same investigator, according to Shulman's and Freedman's methods.

Mild and moderate DAT groups were similar in age, gender and education. Performance on Shulman's clock was similar between groups while moderate DAT subjects performed significantly worse on Freedman's clock compared to mild DAT patients. Both clocks correlated highly in mild and moderate DAT. CDT scores correlated significantly with age and education only in mild DAT. Neither clock correlated with ADLs in either stage of dementia severity. CDTs correlated with the MMSE score, and the CAMCOG score in mild DAT, and only with the CAMCOG score in moderate DAT. These correlations were still significant after controlling for age and education.

Different aspects of cognition and dementia severity are reflected depending on how a clock drawing is scored. Some scoring systems may have greater sensitivity than others in monitoring progression of cognitive deterioration. Correlation between different CDTs and the variables studied (demographic, cognitive, ADLs), when present, is not ubiqitous and changes with the dementia severity.