Notably missing from the HIV literature is an examination of differential associations by sex or race/ethnicity and APOE’s association with cognition and brain integrity. In a meta analysis of aging research studies, APOE e4 women were found to be at greater risk of AD compared to APOE e4 men but only between the ages of 65 and 75 . Additionally, there are known differential effects of APOE status on AD risk by race. The effect of APOE status on AD risk is significantly attenuated in African Americans/Black people compared to non-Hispanic white people . Therefore, future examination of the relationship between sex, race/ethnicity, APOE status, and other genetic markers of AD risk within PWH is certainly warranted. In summary, considering the HIV literature, the middle-aging literature, and the finding that episodic memory was associated with prefrontal structures rather than medial temporal lobe structures, episodic memory in middle-aged PWH is more likely related to frontally mediated etiologies. This could indicate that memory in middle-aged PWH is associated with HIV disease. Notably, this association was seen in PWH on ART without a detectable viral load, showing that this association is seen even in PWH who are virally suppressed. However, it is of course difficult to differentiate between the effect of HIV itself versus the effect of comorbid conditions, many of which may be increased in PWH due to the downstream effects of HIV and ART, or a combination of the two. The medial temporal lobe was not associated with episodic memory, which overall may indicate that at this age range, preclinical AD is not likely a contributor to memory functioning. However, the middle-aging literature does not provide a good estimate of when, on average, to expect to start detecting differences, even small,vertical grow factory in memory and medial temporal structures in those that are on an AD trajectory; therefore, it is possible that this group is too young to even start detecting any preclinical AD effect.
This is further complicated because the middle-aging literature is demographically different from the CHARTER sample, thus highlighting the need for more diverse aging studies. Additionally, this study did not specifically examine differences in the associations between memory and brain structures by AD risk ; thus, future research should examine memory associations by AD risk, particularly given that APOE status was associated with delayed recall. Relatedly, these findings show that on average this group is not showing associations with memory and the medial temporal lobe and early signs of preclinical AD, but this does not mean that no participants are on an AD trajectory. In fact, given base rates, some of this group will eventually develop AD. First, the multi-level models examining the cross-level interactions between time and medial temporal structures with dichotomous recognition as the outcome did not converge. This analysis would have examined if baseline medial temporal lobe structures are associated with greater likelihood of impaired recognition over time. Given that the models did not converge, this indicates the models were over parameterized and that the model was not supported by the data. This was possibly affected by the modest sample size, with a particularly small group of participants with impaired recognition at baseline. Examining the variability in recognition over time within this study is still meaningful. For example, of the 12 participants that were impaired at baseline, only two remained impaired. Moreover, in those that were not impaired at baseline but were impaired at some point in time, most reverted back to unimpaired at subsequent visits. Only four participants remained impaired in recognition over time, although with limited follow-up. There is not data on why these participants do not have additional follow-up , and thus it is hard to make any definitive conclusion as to if consistently impaired recognition is a risk factor for negative outcomes. However, it would certainly be warranted to examine if consistent recognition impairment is associated with negative outcomes in a larger group of middle-aged and older PWH. For example, this small group of participants that were consistently impaired in recognition memory could represent those that are progressively declining and are on more of an AD trajectory. Moreover, a better understanding of how those that are consistently impaired differ from those that revert to unimpaired recognition would be beneficial.
There are multiple reasons that may explain why recognition impairment status was variable over time. First, HIV-associated neurocognitive impairments are known to fluctuate over time. For example, in the CHARTER study, 17% of the sample improved over time . Therefore, this could simply reflect the heterogeneous and fluctuating course of HAND over time. Second, recognition is sometimes used as an embedded performance validity measure. While all participants were administered a standalone performance validity test at the beginning of the neuropsychological evaluation to verify credible test performance, effort can fluctuate throughout testing. That said, none of the participants at baseline were below the proposed cut-off of ≤5 for HVLT-R recognition , making this explanation less likely. Lastly, this variability over time may be in part due to the psychometric properties of the HVLT-R and the BVMT-R. Recognition for both the BVMT-R and the HVLT-R are skewed with known ceiling effects, meaning that there is limited variability in this variable . Therefore, a one- or two-point difference can result in large differences in the normative score. Moreover, there are known modest interform differences on the HVLT-R recognition . Additionally, while the HVLT-R and BVMT-R test-retest reliability of recognition show adequate test-retest stability coefficients, the test-retest reliability of recognition is less reliable than other test measures such as total learning or delayed recall . Next, longitudinal delayed recall was examined. Most notably, there was little decline in delayed recall over time; the delayed recall T-score decreased by 0.041 per year. Additionally, there was little variability in this slope given that the standard deviation of the slope was 0.678. None of the cross-level interactions between medial temporal lobe structures and years since baseline were significant indicating that medial temporal lobe structures at baseline were not associated with a change in delayed recall. However, given that there was little variability in delayed recall over time, this was not surprising.
As discussed in the introduction, worse baseline medial temporal lobe structures, particularly the hippocampus and entorhinal cortex, have been associated with an increased risk of future AD, MCI, and decline in cognition in older adults without HIV . This relationship is less understood in middle age. One study by Gorbach et al. found that hippocampal atrophy was associated with a decline in episodic memory in adults over the age of 65 but not in middle-aged adults between the ages of 55 to 60. As highlighted above, it is possible that the cohort from the current study is too young to expect to see associations between medial temporal lobe structures and longitudinal memory. Importantly, the current study only examined cross-sectional structural MRI; therefore, we cannot assume that smaller or thinner medial temporal lobe structures are indicative of atrophy. Additionally, this study does not have an HIV-negative comparison group and did not use normatively-adjusted morphometric values , so it is unclear if participants in this cohort deviate from average, although accelerated brain atrophy has been demonstrated in PWH previously . Therefore, research examining changes in the medial temporal lobe and how that change relates to episodic memory, particularly recognition memory,vertical grow indoor in persons with and without HIV over the age of 65 is needed. This research may help to better understand if medial temporal lobe structures are associated with the risk of an AD trajectory and if these associations differ by HIV-serostatus. While there may be some individuals in this group that are experiencing objective decline, on average, in this group of middle-aged PWH we did not observe a decline in delayed recall T-scores over time. These T-scores are age-corrected, so the raw scores on the tests may be declining but they are not declining at a rate greater than what would be expected for age. Additionally, these T-scores also account for practice effects, which if unaccounted for can mask decline, although the best method of practice-effect correction is still debated . Similar results showing stable cognition over time were found in a study by Saloner et al. in a larger sample of CHARTER participants aged 50 and over. This study employed growth mixture modeling, and none of the three latent classes demonstrated a decline in global T-score over time. However, other studies of PWH over the age of 50 have observed a greater than expected effect of aging on episodic memory and a recent systematic review found accelerated neurocognitive aging in 75% of longitudinal studies in PWH . Some researchers have questioned if accelerated aging could be due to a neurodegenerative cause such as AD given the high prevalence of risk factors for AD in PWH such as chronic inflammation, increased cardiometabolic comorbidities, and lower brain reserve . While emerging studies have demonstrated some possible ways to disentangle HAND and aMCI , it remains unclear if PWH are at increased risk of AD or if a neurodegenerative etiology could, at least in part, account for someof the observed accelerated aging.
For example, Milanini et al. showed a low frequency of amyloid positivity, measured via PET imaging, among virally suppressed PWH over the age of 60, and the rates of amyloid positivity were similar to published rates among an age-matched seronegative sample. However, a recent study among Medicare enrollees did find a higher prevalence of AD and related disorders among PWH . In summary, this aim showed that recognition was variable over time. While amnestic decline could not specifically be tested given that recognition models did not converge, these analyses indicated that within this group, medial temporal lobe integrity was not associated with a decline in delayed recall over time. Additionally, delayed recall only marginally declined over time , thus adding to the mixed literature examining episodic memory in middle-aged and older PWH. Overall, this study did not detect clear signs of preclinical AD in this group, as delayed recall did not change over time and baseline measures of medial temporal lobe integrity were not associated with memory over time as seen in HIV-negative older adults. However, it is not clear if these associations would be expected in a middle-aged cohort of PWH due to a lack of literature on this topic in middle-aged adults. Therefore, it would be beneficial to re-examine this analysis in an older cohort of PWH.The last aim of this study was to examine if the medial temporal lobe mediates a relationship between peripheral inflammation and memory. It was hypothesized that medial temporal lobe structures would mediate a relationship between peripheral inflammation and episodic memory. Five peripheral biomarkers of inflammation were examined , and these biomarkers were chosen given that they have been associated with cognition in AD and HIV. In this mediation model, the association between peripheral biomarkers of inflammation and medial temporal lobe structures was also explored and the relationship between medial temporal lobe structures and memory was also reported, although this second relationship was already explored in aim 1. First, the mediation models examining recognition indicated poor model fit. Therefore, the relationship between the five plasma biomarkers of inflammation and recognition was examined instead. Greater levels of plasma CRP were associated with lower odds of having impaired recognition. None of the other plasma biomarkers of inflammation were associated with recognition impairment. These findings are generally not in line with the HAND , middle-aging , or older adult literature . Aging and HIV studies have found that a greater concentration of these plasma biomarkers of inflammation are associated with greater risk of HAND, worse memory, and an increased risk of future development of MCI or AD. However, many of these studies only find weak associations, and these studies do not examine recognition memory. The current study had a very small sample of PWH with impaired recognition; thus, it is possible that the CRP finding is spurious, and this finding should not be over-interpreted. Therefore, analyses should be reexamined in a larger, more generalizable sample. Next, a single-mediator model was used to examine if medial temporal lobe structures mediate the relationship between plasma biomarkers of inflammation and delayed recall.