Clustering and switching during a semantic verbal fluency test contribute to differential diagnosis of cognitive impairment

The verbal fluency test (VFT) can be dissociated into clustering (generating words within subcategories) and switching (shifting between clusters), which may be valuable in differential diagnosis. In the current study, we investigated the validity of VFT in the differential diagnosis of Alzheimer's disease (AD, n = 65), vascular dementia (VaD, n = 65), mild cognitive impairment (MCI, n = 92), and vascular cognitive impairment without dementia (VCIND, n = 76) relative to cognitively normal senior controls (NC, n = 374). We found that in the NC group, the total correct score was significantly correlated with age and education; males generated more subcategories; cluster size increased with education, and subcategory and switching decreased with age. A significantly progressive advantage was observed in VFT scores in the sequence NC > MCI/VCIND > AD/VaD, and this significantly discriminated dementia patients from the other groups. AD patients performed better in all four VFT scores than VaD patients. Subcategory and switching scores significantly distinguished AD from VaD patients (AD > VaD; mean difference, 0.50 for subcategory, P < 0.05; 0.71 for switching, P < 0.05). MCI patients scored higher than VCIND patients, but the difference did not reach statistical significance. These results suggest that semantic VFT is useful for the detection of MCI and VCIND, and in the differential diagnosis of cognitive impairment.