The effect of age, gender and education in cognitively normal elderly on the strategy of two different category verbal fluency tests

Abstract

Background: Clustering and switching strategy had not been extensively studied in cognitively normal elderly in China. Few studies compared the two semantic verbal fluency tests according to different defined categories. Methods: We recruited cognitively normal subjects and administered neuropsychological tests to them including mini-mental status examination (MMSE), auditory verbal learning test (AVLT), trail making test (TMT), complex figure test (CFT), semantic verbal fluency test (VFT), five point test (FPT), etc. We adopted ‘‘supermarket’’ and ‘‘animal’’ as the two defined categories. Two components_clustering (generating words within subcategories) and switching (shifting between subcategories) were examined. Results: Three hundred and seventy four subjects were enrolled which comprise of 156 men and 218 women with the age as 66.26 6 7.16, education year as 12.72 6 2.79. For the supermarket VFT, the mean score of total correct score, subcategory, cluster size, subcategory switch are 20.196 5.83, 4.04 6 1.27, 4.80 6 2.48, 5.94 6 2.91, respectively. With regard to the animal VFT, the scores are 16.79 6 4.56, 2.94 6 1.03, 4.29 6 2.48, 4.64 6 2.49. The zodiac animal score is 6.37 6 2.41. Co-variance analysis showed that: subcategory score was the main significant difference between men and women (supermarket VFT, men, 4.13 6 1.30, women, 3.97 6 1.25, p 1⁄4 0.035; animal VFT, men 3.11 6 1.07, women, 2.82 6 0.99, p 1⁄4 0.038). The cluster size decreased significantly in supermarket VFT with aging (p 1⁄4 0.001), but not in animal VFT. Education level correlated significantly with cluster size (educational year 12, 4.58 6 2.25, educational year >12, 5.11 6 2,73, p 1⁄4 0.007 for supermarket VFT, 3.90 6 2.13, 4.81 6 2.81, p 1⁄4 0.001 for animal VFT, respectively). Conclusions: Women generate fewer subcategories in verbal fluency test than men. Older groups or poor educated population produce fewer words due to smaller cluster size. These results support a dual processing model in verbal fluency and different strategy in each demographic subgroup.