Consistent with research in other populations demonstrating the value of intra-individual RT variability as an indicator of neurocognitive status (Bruce et al., 2010
; Dixon et al., 2007
; MacDonald et al., 2003
; Strauss et al., 2007
), results of this study of HIV-positive individuals demonstrated that elevated RT variability was associated with poorer overall cognitive ability. Higher RT variability in this study was also associated with indicators of recent immunological status (i.e., recent CD4 count and viral load), peak immunological dysfunction (i.e., nadir CD4 count and highest viral load), and, consistent with preliminary indications from our previous study (Levine et al., 2008
), poorer adherence to antiretroviral medication regimens. This latter finding is of particular interest, as medication adherence represents an important indicator of functional ability in this population. Recent findings suggest that medication adherence and cognitive dysfunction may be reciprocally related in HIV/AIDS (Ettenhofer, Foley, Castellon, & Hinkin, 2010
). However, it was not possible to evaluate the degree to which the current findings might represent the potential for RT variability to predict individuals' ability to adhere, versus any possible protective effect of antiretroviral adherence on RT variability.
In comparison with RT variability, mean RT latency was associated with a relatively narrower set of indicators, including recent immunological status and global cognitive ability, but not medication adherence, lowest CD4 count, or highest viral load. Follow-up analyses demonstrated that RT variability accounted for additional variance in global cognition, medication adherence, and peak immunological dysfunction beyond that explained by RT latency. Moreover, the relationship between RT latency and global cognition was no longer significant when RT variability was included in the model. Beyond demonstrating its incremental value, this finding suggests that RT variability may also be more fundamentally related to global cognition than RT latency in this population.
Among indicators of cognitive ability, only learning/memory was associated with medication adherence, and no significant associations between cognitive ability and immunological status were found in this sample. Considering our modest sample size, these findings should not be interpreted as evidence for the absence of relationships between the variables in question. For example, significant relationships between medication adherence and executive function have been demonstrated in previous research (Hinkin et al., 2002
). Instead, the current pattern of results is best interpreted as preliminary evidence that RT variability may be more sensitive to poor medication adherence and recent immunological dysfunction than indicators of processing speed, attention, executive functions, and motor functioning, and may be more sensitive to poor medication adherence and peak immunocompromise than RT latency.
The potential significance of this latter finding is highlighted further by recent research suggesting that peak immunosupression may be more strongly related to neurocognitive decline than day-to-day fluctuation in immunological status (Cysique, Maruff, & Brew, 2006b
). Neither neuroanatomical data nor a suitable HIV-negative control groups were available in this study to determine whether RT variability was specifically related to HIV-related frontal/subcortical disturbance. However, our participants' above-average mean estimated premorbid verbal IQ suggests that the poorer-than-average mean performance demonstrated on neurocognitive tasks and indicators of RT latency and variability probably represent true weaknesses in this population. Therefore, in combination with the previously documented relationships between RT variability and neuropathology in aging and mixed neurological samples (Bunce et al., 2007
; Stuss et al., 2003
; Walhovd & Fjell, 2007
), preliminary findings from this study suggest that RT variability might also provide enhanced sensitivity to important aspects of HIV-associated neurocognitive decline.
Future research on this topic may be strengthened by improved measurement of neurological, behavioral, and immunological status. For example, additional measures of neurological status (e.g., neuroimaging) and behavioral functioning (e.g., driving capacity) would provide valuable points of reference for the neuroanatomical and functional significance of elevated RT variability in HIV/AIDS. Likewise, relationships between RT variables and “most recent” immune indicators in this study may have contracted or otherwise changed during the time frame between the collection of information about immune status and the administration of RT measures. Therefore, a more systematic approach to measurement of immune status may be useful in examining the links between RT variability and immunocompromise in HIV/AIDS.
Additionally, sample size limited the degree to which statistical comparisons could be made between the different aspects of RT variability assessed in this study. Comprehensive examination of different forms of RT variability in future investigations may provide additional information regarding the utility of these novel measures and possible mechanisms of impairment in HIV/AIDS. Finally, in future research, it will be important to compare RT variability and other neurocognitive indicators between HIV-positive and HIV-negative groups in order to more directly verify the extent of elevations in RT variability that may be present.
In conclusion, this study provides a window into the possible relevance of RT variability as an indicator of neurocognitive burden in HIV/AIDS patients, and the potential value of RT variability measures as an adjunct to other methods of neuropsychological assessment. Although neuropsychologists typically focus on mean level of performance rather than consistency of performance, data from this study suggest that by doing so we may be overlooking important sources of information. It is possible that, like the proverbial ‘canary in the coal mine’, increased RT variability may prove to be a harbinger of incipient declines in level of performance. However, the neurobiological mechanisms of RT variability in HIV/AIDS remain unclear. Additional research will be needed to evaluate the degree to which the sources of RT variability in HIV/AIDS may parallel those suggested by research in other populations, and to specify elements of this variability that may be most fundamentally related to various aspects of neurological, immunological, and behavioral functioning.