Construct validity of Hopkins Verbal Learning Test—Revised component process measures in an HIV-1 sample

Abstract

Executive dyscontrol of episodic verbal learning and memory secondary to prefrontostriatal circuit neuropathophysiology is a common feature of HIV-1 infection. Prior research indicates that standard clinical learning and recall indexes from Hopkins Verbal Learning Test—Revised (HVLT-R) are among the most sensitive indicators of HIV-associated neurocognitive disorders. Emerging data support the validity of qualitative component process measures derived from the HVLT-R (e.g., Semantic Clustering); however, no prior studies have examined these particular indices of performance in an HIV-1-infected population. In the present study, we examined the construct validity of HVLT-R component process indices in a sample of 42 persons with HIV-1 infection and 29 demographically similar seronegative comparison participants. The HIV-1 sample performed significantly below the seronegative group on Total and Delayed Recall, Semantic Clustering, and the Retrieval Index. No between-group differences were observed on Serial Clustering, Pair Frequency, Learning, Repetitions, Semantic False Positive Recognition Errors, or the Recognition Discrimination Index. In addition, the HVLT-R component process measures demonstrated evidence of convergent and divergent validity with standard clinical tests in the HIV-1 sample. Findings support the construct validity of HVLT-R component process measures and are commensurate with prior literature indicating that HIV-1 disease is associated with deficient executive control of encoding and retrieval within verbal episodic memory.