Introduction: Semantic verbal fluency (SVF) tasks are widely used because of their simplicity, brevity and diagnostic accuracy (DA). However, they have the disadvantage of being greatly influenced by socioeducational variables. Our objective is to assess the possible influence of these variables on the fluency test "names of persons" (SVF-np).
Methods: Cross-sectional study in 226 neurology patients classified in dementia (DSM-IV criteria), cognitive impairment without dementia (GENCyD-SEN criteria) and non-cognitive impairment who responded to a SVF-np test and to the classical test of verbal fluency "names of animals" (SVF-an), evaluating the DA of their results for both dementia and cognitive impairment by calculating and comparing the area under the ROC curve (aROC). In addition, the influence of the socioeducational variables and cognitive state on the results was assessed through multiple linear regression analysis.
Results: There were no significant differences between the DA of the SVF-np and SVF-an for dementia (0.88 +/- 0.02 [aROC +/- SD] vs. 0.90 +/- 0.02, respectively) or for cognitive impairment (0.88 +/- 0.02 vs. 0.87 +/- 0.02). The scores of the SVF-an task were associated with age, sex, gender, educational level and cognitive status of the subject. On the contrary, those of the SVF-np task only depended on the cognitive status.
Discussion: The SVF-np task has the same advantages and DA as the SVF-an, but it has the additional advantage of not being influenced by socioeducational variables and not requiring score adjustments or corrections. This independence makes it especially appropriate for use in multicultural populations and those persons with low educational level.