HIV-associated cortical and subcortical volume reductions, white matter changes, metabolite abnormalities, and regional glucose metabolism that vary relative to HIV clinical factors (for example, viral load, nadir CD4), and HAND severity are evident on magnetic resonance imaging (MRI), magnetic resonance spectroscopy (MRS), and positron emission tomography (PET) [
32-
34]. Although neuroimaging abnormalities are usually most significant in cases of opportunistic brain infection, HIV also has direct and indirect effects on brain structure and function. Historically, research focused on the basal ganglia and cerebral white matter, regions considered to be particularly vulnerable to HIV [
35]. Yet when compared with seronegative controls, people with HIV also show reduced grey matter and cortical thinning [
32-
34,
36], especially in frontal and temporal regions [
37]. A study of asymptomatic individuals with HIV revealed decreased frontal grey matter volumes in the absence of other brain changes [
38], suggesting that cortical atrophy begins in frontal areas. White matter hyperintensities (WMHs) on MRI reflect white matter damage, and studies show WMHs occurring at younger ages in people with HIV than they do in adults aging without HIV [
39,
40]. Case-controlled diffusion tensor imaging (DTI) studies show that HIV is associated with lower white matter integrity globally [
41] and in specific areas such as the corpus callosum, internal capsules, and the frontal and parietal lobes [
42,
43]. There is evidence that the earliest HIV-associated white matter effects occur in the frontal lobes [
44], with more widespread damage occurring as the disease increases in severity [
43,
45]. MRS studies show increased myoinositol (MI), choline (Cho), and total creatine (Cr) in brain disorders with chronic inflammation and glial activation, including HIV [
32,
34,
46,
47]. N-acetylaspartate (NAA), a marker of neuronal integrity that decreases in response to neuronal damage, has been shown to be lower in people with HIV compared with age-matched controls, especially in the basal ganglia and frontal white matter [
46,
48]. Although NAA associations with plasma HIV-RNA show that neuronal injury is related to current viral replication, neuronal injury is also found in virologically suppressed patients and may be attributed to the effects of chronic immune activation and inflammation. PET studies have shown that glucose hypometabolism in the frontal cortex, suggesting deficient functioning, and basal ganglia hypermetabolism occur in HIV [
49]. Increased basal ganglia metabolism may seem counterintuitive, although HIV-infected astrocytes require increased glucose to proliferate [
50], and the basal ganglia are known to be particularly vulnerable to HIV [
35].
The neuroimaging abnormalities that have been observed historically among HIV-infected people are similar to those observed among older adults without HIV. As people reach advanced age, cortical and subcortical volumes gradually decrease [
51]. Additionally, older age in healthy cohorts has consistently been found to be one of the most important independent predictors of greater WMH volume [
52]. WMHs in frontal and parietal regions have been especially associated with older age and greater cognitive dysfunction [
53], and longitudinal studies show greater age-related volumetric decline in anterior versus posterior white matter regions [
54]. Declines in DTI measures of white matter integrity also occur with increased age [
55], with anterior regions showing the greatest changes [
56]. Furthermore, MRS research indicates that there is an age-related decline in NAA and increases in Cho and Cr [
57]. Finally, PET research shows age-related declines in glucose metabolism, beginning with frontal lobe changes [
58].
cART-era research shows that the neuropathology of HIV appears to be changing in that it now involves cortical as well as subcortical structures [
33]. This signifies that HIV may progress to involve processes that bear a greater resemblance to age-related neurodegenerative diseases, such as AD, of which cortical atrophy and ventricular enlargement are hallmarks [
59]. As in HIV, WMHs have been shown to occur in frontal and parietal lobes in AD, and the degree of WMH in parietal lobes and posterior periventricular areas corresponds with level of cognitive impairment [
60]. AD is also associated with widespread DTI abnormalities [
61]. However, cortical changes in AD cases are typically more pronounced than in cases of HIV, and hippocampal atrophy occurs early and ubiquitously in AD whereas the hippocampus is not as vulnerable in HIV [
62,
63]. Also, unlike in HIV, in AD the largest DTI effects are in hippocampal areas. A recent review of MRS abnormalities in AD showed NAA decreases and MI increases similar to HIV. In AD, decreased NAA is generally found in all major lobes of the brain as well as the medial temporal lobe and the posterior cingulate gyrus [
64]. MI increases are also common, and changes in NAA and MI are associated with level of AD neuropathology. PET research in AD shows parietal, temporal, and posterior cingulate glucose metabolism decreases that predict cortical volume loss [
65], with decreases in the frontal cortex as the disease progresses [
66], whereas in HIV frontal hypometabolism is seen early on in the disease.