Abstracts

Rationale: Fluency tests are an important element of neuropsychological testing. We use matched verbal and nonverbal tests to contrast left versus right epileptic foci; each lasts 9 min and requires written responses. Our tests yield good correlation with lateralized brain dysfunction, but we wished to shorten administration time without loss in test quality. Shorter verbal fluency tasks exist but use oral responses and thus make a poor match for figural tasks, which must be written. Further, the phonemic verbal fluency tasks either employ different letters for francophones and anglophones, or if the same letters are used the results differ because of differences in letter frequencies between languages. This study aimed to identify stimuli for written 1-min verbal fluency tests (one phonemic, one semantic) suitable for both languages, which is important for our bilingual site and for many others, and to determine the feasibility of reducing the time allowed for our figural fluency test to 1 min per condition. Methods: To find letters of equal frequency in English and French we performed dictionary word counts and selected two (M and D) that were highly similar in the two languages. We chose three semantic categories expected to contain items with which people of all socioeconomic levels should be familiar: grocery, bathroom, hospital. We tested 40 healthy subjects, half per language and matched for age and education, on 1-min written fluency for each of the five verbal stimuli and the two conditions (free and 4-line) of Design Fluency. Results: We found no significant differences between groups in number of words produced for either letter, but scores were most similar for M. The groups also performed similarly on the semantic categories; highest scores were for grocery and the least number of ambiguous responses were for bathroom. All these scores were slightly lower than means reported for 1-min fluencies tested orally. The shortened Design Fluency test yielded higher scores than would be expected as the appropriate fraction of the full test. Scores were slightly higher for the 4-line than the free condition, but there was no difference between language groups on either condition. Conclusions: The high output in the shortened Design Fluency test relative to the full test suggests that a greater proportion of responses is generated early in the test and output slows over time. We know that perseverative responses increase over time in the full test, in keeping with the notion that output is better at the beginning. The higher scores on the 4-line than on the free condition is true also of the full test, again suggesting that the short version yields results like those of the 9-min test. Thus we expect that the 2-min Design Fluency (1 min per condition) will yield meaningful data, and we are collecting patient data to verify that it also still elicits a deficit in the presence of right frontal lobe abnormality. Based on these results it is reasonable to assess fluency adequately, and similarly in anglophones and francophones, using the letter M and the bathroom category as verbal tasks, and the shortened Design Fluency test.