The present study sought to examine the profile and predictors of neurocognitive impairment and neuropsychiatric disturbances in individuals with AEH. Extending the scant previous neurocognitive research on this topic (e.g., Moore et al., 2011
), our results revealed elevated rates of both neurocognitive impairment and neuropsychiatric distress in the AEH sample that were significantly higher than their matched seronegative counterparts. These data support prior neuroimaging findings suggesting alterations in brain metabolism, blood flow, and functional connectivity in the early phase of HIV infection (e.g., fronto-striatal pathways; Ances et al., 2009
). Of clinical relevance, given the high rates of HIV transmission that have been observed during the AEH period (e.g., Brenner et al., 2007
), these neurocognitive and neuropsychiatric factors may play a significant role in risk-related behaviors during this critical stage of HIV disease. Furthermore, this study provides valuable insight with regard to the clinical factors associated with these neurocognitive and neuropsychiatric disturbances, particularly MA and alcohol use, which are often independently implicated in risky sexual behaviors and HIV transmission, and could also be targeted for intervention.
In the present AEH sample, 61% of individuals were classified as neurocognitively impaired (i.e., a GDS of ≥ 0.5) on an abbreviated battery of tests designed to be sensitive to HAND. This proportion is slightly higher than is typically seen in chronic HIV infection (e.g., 40–50%; Heaton et al., 2011
) and is nearly double that which was observed in Moore and colleagues’ AEH sample (2011). Several factors may be contributing to the high rate of HIV-associated neurocognitive impairment in this study. First, the present sample of individuals with AEH was examined much closer to their estimated date of infection than in the Moore et al. study (i.e., 11 vs. 16 weeks, respectively). Thus, it is possible that we captured the CNS functioning of our cohort during the height of their adverse inflammatory processes and altered brain metabolism, which may have made our AEH sample more vulnerable to neurocognitive impairment. To this end, we observed a significant relationship between neurocognitive impairment and plasma viral load, suggesting that increased viremia at this time point might be directly related to worse neuropsychological outcomes, as consistent with the findings of Marcotte et al (2003)
. The elevated rate of NCI may also simply be due to sampling differences. Another possible explanation is that the neurocognitive battery administered included tests that are particularly sensitive to incident brain injury, particularly HAND (Carey et al., 2004b
). However, battery selection is unlikely to fully explain this finding because, despite a relatively elevated rate of impairment in our seronegative comparison group that was matched on numerous neurocognitive risk factors that are similarly sensitive to such tests (e.g., substance use), we still observed a statistically significant difference between groups and nearly a fourfold increase in impairment in AEH.
The overall pattern of neurocognitive deficits observed in this AEH sample varies from what one would expect in the cART era. In more recently studied chronically infected HIV samples, the most frequently observed neurocognitive deficits are in the domains of learning and executive functions (Heaton et al., 2011
). Indeed, our sample exhibited the highest rates of impairment on a test of verbal list learning (i.e., HVLT-R) relative to seronegative counterparts, as consistent with the post-cART era experience of the HIV epidemic. However, the AEH sample also showed greater deficits on a test of information processing speed (i.e., Trail-Making Test Part A) relative to the comparison group, a finding that is more consistent with pre-cART era neurocognitive deficits (Heaton et al., 2011
). Nevertheless, this finding is also consistent with the results Wang et al. (2011)
report in their AEH cART era cohort. This may be reflective of a more global, diffuse neural process that is consistent with the cascade of neuroinflammatory events in the early days of HIV infection. This pattern of deficits may be critical in explaining the high rates of HIV transmission during the AEH period. For instance, individuals with decreased information processing speed may be unable to adequately process rapidly evolving social situations that may lead to risk behaviors. Similarly, persons with deficits in verbal learning may have difficulty effectively acquiring pro-health messages related to HIV prevention (e.g., public health campaigns to encourage clean needle use) and treatment. Future studies may wish to include other aspects of memory (e.g., prospective memory; Carey et al., 2006
) and executive functions (e.g., planning; Cattie et al., 2012) that may be especially important to the clinical implications of HAND in AEH (e.g., carrying a condom or remembering to bring a clean needle to an event).
Overall, HIV-associated neurocognitive impairment in the AEH sample was most strongly associated with high-risk MA use. Independently, MA dependence is associated with damage to dopaminergic frontostriatal circuitry and often results in mild-to-moderate neurocognitive deficits in verbal learning and information processing speed (e.g., Scott et al., 2007
; Woods et al., 2005b
). In the context of AEH, additive neuroinflammatory processes effects of MA and HIV combined may exacerbate this damage and thereby produce the profile observed. Such additive damage at the neural level has been observed in the context of chronic HIV infection and is hypothesized to be a result of a combination of increased dopaminergic and glutaminergic transmission (MA) as well as cytokines, inflammatory responses, and other neurotoxins that stimulate glutaminergic cells (HIV), which may result in apoptosis, vascular pathology, and damage to both grey and white matter (Chana, et al., 2006
). Based on the cross-sectional and present-focused nature of our data, it is unknown whether the assessed MA use existed prior to HIV infection (and was the method by which the individual became HIV-infected) or whether MA use at this assessment was reactionary to knowledge of HIV status, and therefore more acute (cf. chronic MA use). However, our groups were matched on MA use, which helps to suggest that the NP impairment in the AEH sample was not solely due to MA, but instead a combination of risk factors for some participants. Unfortunately, we are not able to rule out the possibly confounding effects of polysubstance (e.g., cocaine, opioids) use or examine specific MA use parameters, as the DAST did not include information regarding onset, duration, and frequency of use. Nevertheless, the specificity of the observed finding to MA is at least partially supported by the absence of significant associations between HAND and alcohol use in this AEH cohort.
In addition to providing insight regarding the neurocognitive profile of AEH, the neuropsychiatric features of this cohort may yield important, and potentially clinically useful, insights into this stage of HIV disease. Notably, findings from two previous AEH cohorts reported levels of depressive symptomatology that were mildly elevated relative to the general population (i.e., Atkinson et al., 2009
) or seronegative adults (i.e., Moore et al., 2011
). In contrast, the present study compared an AEH cohort to seronegative comparison participants on a global, multifactorial measure of neuropsychiatric distress (i.e., POMS). Analyses of the POMS Total Mood Disturbance summary score revealed significantly higher rates of distress among AEH individuals compared to the HIV- group and was roughly comparable to levels that have been observed in chronically infected HIV+ individuals (e.g., Cattie et al., in press
). Moreover, using the normative cut point, over 50% of AEH individuals were experiencing clinically elevated affective distress compared to less than 20% of the seronegative comparison participants. In the same way that lifetime mood disorder onset typically precedes HIV seropositivity notification (Atkinson et al., 2009
), elevated neuropsychiatric symptoms may also have preceded these individuals’ awareness of their HIV status. Nevertheless, such an elevation may be of great importance during the AEH period because they may impact the individual’s response to learning of his or her HIV seropositive status, such as a possible exacerbation of affective symptoms or an increase in maladaptive behaviors (e.g., substance use). Although there is some evidence to suggest that, on average, AEH individuals typically utilize adaptive coping behavior (Atkinson et al., 2009
), it is not clear whether these findings generalize to all individuals with AEH, or whether these adaptive coping strategies would facilitate these individuals following current best-practice recommendations (e.g., initiating cART despite potential side effects). Furthermore, given the evidence of elevated rates of mood disorders and mood symptoms in chronic HIV infection, it may be the case that these adaptive coping strategies are not sufficient to prevent a long duration of symptoms or possible recurrence of symptoms, suggesting that formal psychotherapeutic intervention in the AEH period may be beneficial for both short-term and long-term psychiatric outcomes.
One advantage of the POMS is that it includes a multifactorial assessment of different aspects of neuropsychiatric distress. In our study, the AEH group reported significantly greater elevations than the comparison subjects on the Confusion/Bewilderment and Tension/Anxiety subscales. A trend-level elevation in the Depression/Dejection subscale of the POMS was observed in AEH, which was confirmed by a follow-up analysis of continuous normative z-scores. This evidence of depressive mood converges with prior data on MDD in AEH from Moore et al. (2011)
, who showed higher lifetime rates of MDD in AEH. However, we did not observe group differences in clinical elevations of the Anger/Hostility, Fatigue/Inertia, Anger/Hostility. This pattern of findings may suggest that self-reported distress in this cohort may be specific to certain clusters of affective symptoms, rather than global affective complaints.
With regard to anxiety, mild levels of anxiety have been observed in AEH groups previously (Atkinson et al., 2009
), even when measured weeks to months after notification of HIV-seropositive status. In the current cohort, the rate of clinically elevated anxiety symptoms exceeded that of depression symptoms, suggesting that anxiety was the strongest contributor to the overall elevation in total mood disturbance. Unlike mood disorders, for which evidence suggests that onset of these psychiatric conditions preceded these individuals’ awareness of their HIV seropositive status (i.e., notification of positive test results; Atkinson et al, 2009
), it is possible that increases in anxiety symptoms may be a consequence of AEH status. Findings from a cross-sectional investigation such as the present study cannot definitively address this question, but due to the persistence of new HIV transmission this is a potentially important issue to address. Specifically, newly infected individuals may benefit from interventions targeting anxiety symptoms in the AEH period. Anxiety disorders are also common in chronically-infected individuals (Atkinson & Grant, 1994
), and it is possible that a similar pattern may be observed with anxiety symptoms as with depression symptoms, whereby individuals with clinically significant symptoms in the early stages of infection may be more likely to exhibit symptoms in clinical range up to one year later (Perry et al., 1993
Further examination of the total mood disturbance score predictors revealed only high-risk alcohol use to be associated with elevated rates of overall affective distress in the present AEH cohort. A high prevalence of lifetime alcohol use disorders has previously been reported among AEH individuals, with onset largely preceding the notification of HIV serostatus similar to mood disorders (Atkinson et al., 2009
). Therefore, it is possible that this association is simply due to pre-existing problems with alcohol use that continue into the AEH period. Alternatively, the demonstrated anxiolytic properties of alcohol use (e.g., Steele, Southwick, & Pagano, 1986
) suggest that risky levels of drinking may be more directly related to the elevated anxiety symptoms observed among AEH individuals, regardless of whether those increased anxiety symptoms preceded HIV serostatus notification or were a consequence of AEH status. These findings also might also suggest that the AEH individuals in the present cohort may not have been using primarily adaptive coping strategies (cf. Atkinson et al., 2009
) and instead initiated or continued high-risk drinking in response to seropositive status awareness. This relationship is ecologically relevant given that risky alcohol use during the AEH period increase risk of HIV transmission (e.g., Raj et al., 2009
Interestingly, although the rates of both neurocognitive impairment and affective distress are high in AEH, there is likely discordance between the individuals who experience these respective effects, as well as in the relationships between the predictors and their respective outcomes. Indeed, we observed elevations in both the POMS Confusion/Bewilderment scores and objective neurocognitive impairment; however, performance and complaints did not relate to one another, which has been a consistent finding across the chronic HIV infection literature (e.g., Rourke et al., 1999
). Due to the cross-sectional nature of the present study, we cannot address the timing and intersection of these problems with our data, but several considerations emerge. For example, one possibility is that risky alcohol use and illicit MA use may be antecedents to elevated affective distress and neurocognitive impairment, respectively, irrespective of AEH status. Alternatively, AEH may precipitate both affective distress and neurocognitive impairment via neurobiological mechanisms (e.g., dopaminergic disruption), which may subsequently result in AEH individuals abusing alcohol and/or MA. Another possibility is that both risky alcohol use and/or MA use were pre-existing conditions in these AEH individuals before they were notified of their HIV seropositive status, and their use of these substances persisted into their AEH period. Although our data do not directly support this notion, it is also possible that MA use in the AEH context likely negatively impacts neurocognitive performance, whereas elevated anxiety in the AEH period may lead to and increase in risky alcohol use. Future prospective investigations may address these possibilities with longitudinal data.
Regardless of the timing of onset and the directional nature of these relationships, it is clear that individuals in the AEH period may benefit from early detection and remediation of affective distress, alcohol and substance use, and neurocognitive impairment. It is possible that presence of these factors, alone or in combination, may represent risk factors for later consequences, such as continued psychiatric and substance use problems, as well as further neurocognitive decline. Individuals with such constellations of risk factors may also be at increased risk for declines in everyday functioning (Weber et al., 2012). As such, early intervention may reduce these risks, thereby improving long-term heath outcomes. Furthermore, intervening on these factors with remediation and psychoeducation at this time point may also significantly curb risk behaviors that allow for HIV transmission and continue the epidemic.