Replication of a study of frequency analysis of the resting awake EEG in mild probabke Alzheimer's disease

Lawrence A. Coben, David Chi, Abraham Z. Snyder, Martha Storandt

Research output: Contribution to journalArticle

104 Scopus citations


In the resting EEG, the percentage power in the delta, theta, alpha, and beta bands and the mean frequency were computed in an occipital-vertex derivation for two samples of subjects. Theoriginal sample (n = 79) and the new sample (n = 43) each contained a mild probable Alzheimer's disease (SDAT) group and a healthy elderly control group. Group medians in both samples were higher in the SDAT than in the healthy subjects for percentage delta and theta, and were lower for percentage alpha and beta and for mean frequency. Percentage theta and mean frequency were consistent across the two samples in showing statistically significant differences between SDAT and healthy groups. The ability of each EEG measure to detect individual subjects with SDAT was assessed. The most effective measure, percentage theta, had only modest sensitivity (about 20%),but this was attained at a specificity of 100%. The accurate detection of an individual at the mildstage requires that the predictive value of a positive test be hight to avoid misclassification of non-SDAT subjects as SDAT. This,in turn, requires a specificity of virtually 100% when the prevalence is low. The low sensotivity puts several constrainss on the usefulness of the EEG. For this reason, when the dementia is at the mild stage the EEG would be a useful detector Af probable Alzheimer's disease only under certain limiting conditions, including high prevalence, high specificity, and a willingness to accept a high rate of falsely negative tests.

Original languageEnglish
Pages (from-to)148-154
Number of pages7
JournalElectroencephalography and Clinical Neurophysiology
Issue number3
StatePublished - Mar 1990


  • Alzheimer's disease
  • EEG
  • Frequency analysis

Fingerprint Dive into the research topics of 'Replication of a study of frequency analysis of the resting awake EEG in mild probabke Alzheimer's disease'. Together they form a unique fingerprint.

  • Cite this