Fluency in multiple sclerosis: which measure is best?

Abstract

Tests of verbal fluency provide brief and sensitive measures of the deficits in rapidly retrieving overlearned information common in multiple sclerosis (MS). Production of words that begin with the letters F, A, and S is the verbal fluency measure most often used with patients who are fluent in English. However, because of frequency of words beginning with certain letters varies from one language to another, it is unlikely that any fixed set of letters will be appropriate for multicenter trials that involve patients who are fluent in different languages. A possible alternative involves using semantic fluency categories that contain such a large number of exemplars that no fluent speaker of any language could exhaust the category in the allotted response time. To examine the potential usefulness of semantic fluency measures, 203 MS patients and 87 healthy controls generated words that begin with F, A, or S or were exemplars of the categories animals and parts of the body. Receiver operating characteristic (ROC) curve analyses indicated that sensitivities and specificities for the three fluency measures in discriminating patients from controls were quite similar, especially if patients with global cognitive impairment were excluded.