icon-folder.gif   Conference Reports for NATAP  
Cognitive Impairment in Symptomatic HIV over age 60
  "We examine the demographic and cognitive testing profile of a group of symptomatic individuals and explore factors that may predict objective impairment.....71 individuals (56%) met criteria for MND [Mild Neurocognitive Disorder]."
Testing Finds No Impairment in One Third of Older HIV Group With Cognitive Complaints
8th International Workshop on HIV and Aging, October 2-3, 2017, New York
Mark Mascolini
Despite reporting cognitive symptoms, almost one third of a 60-and-older HIV group in San Francisco had normal results on a hefty battery of neuropsychological tests [1]. Another 15% of participants had test-detected impairment, but non-HIV causes could not be ruled out.
University of California, San Francisco (UCSF) researchers who conducted this study noted that cognitive impairment--though usually mild--remains prevalent in groups with HIV well controlled in blood. Cognitive complaints are particularly frequent, as one would expect, in older people with HIV [2]. To get a better understanding of neurocognitive impairment in older HIV-positive people, the investigators analyzed complaints and neuropsychological test scores in people 60 or older enrolling in the HIV Elders Study. This 5-year study tests mindfulness-based stress reduction to relieve cognitive symptoms, stress, anxiety, and depression.
For this analysis the UCSF team enrolled people 60 or older with undetectable HIV RNA in plasma for at least 6 months on a stable antiretroviral regimen. All participants reported cognitive symptoms, none had used cocaine or methamphetamine for 6 months, and none already tried mindfulness therapy. Participants completed a 10-minute 28-item self-report survey to assess memory, language, and executive function. They also had a 90-minute formal cognitive assessment of seven neuropsychological domains. A consensus panel diagnosed participants as having mild neurocognitive disorder or normal cognition.
Among 127 people participating, median age stood at 65 (range 60 to 83), 94% were men, 75% men who have sex with men, and 84% Caucasian. The group had a median CD4 count of 585 and a median 15.7 years of education.
The researchers diagnosed mild neurocognitive disorder in 71 people (56%), while 37 (29%) tested within normal limits for their age and education. The other 19 people (15%) had some cognitive impairment, but the researchers could not attribute these problems to HIV alone because of confounding conditions such as substance use, major depression, and non-HIV neurologic conditions. In the whole group, more self-reported cognitive conditions did correlate with lower overall neuropsychological test results (r = -0.262, P = 0.006).
The 71 people with mild neurocognitive disorder did not differ significantly from the 37 within normal limits in age, gender, education, current CD4 count, nadir CD4 count, or HIV duration. The group with cognitive disorder had a lower proportion of Caucasians than the normal group (82% versus 97%, P < 0.05), but the researchers cautioned that this difference may reflect less than ideal testing norms for all races.
Analysis of results in individual neuropsychological test domains showed a significant difference between the mild impairment group and the normal group, in the expected direction, for each testing domain. Motor performance emerged as a particular weakness in both impaired and normal participants, while memory remained relatively preserved in both groups.
The UCSF team concluded that neuropsychological testing showed no objective impairment in about one third of older HIV-positive people reporting cognitive symptoms. They noted that symptom characteristics proved poor predictors of impairment, while symptom burden "may have some utility" in predicting neurocognitive problems.
1. Javandel S, Milanini B, Daniels J, Hellmuth J, Allen E, Valcour V. Cognitive impairment in symptomatic HIV over age 60. 8th International Workshop on HIV and Aging. October 2-3, 2017. New York. Abstract 8.
2. Goodkin K, Wilkie FL, Concha M, et al. Aging and neuro-AIDS conditions and the changing spectrum of HIV-1-associated morbidity and mortality. J Clin Epidemiol. 2001;54 Suppl 1:S35-43.
Cognitive Impairment in Symptomatic HIV over age 60


Ultimately, this analysis showed that of the group of symptomatic participants we screened, about one third did not show objective impairment on testing and an additional 15% had another condition that precluded us from applying frascati criteria for HAND. It seems that the characteristics of the symptoms reported are poor predictors of impairment, but the burden, or number of symptoms, may be more useful. I want to note that there are certainly some limitations in applying these results. First, it's unclear if the PAOF is really the optimal tool for assessing symptoms. We know that ADLs and IADLs are flawed, but there may be other more appropriate assessments. Second, participants in this study were paid for their participation and were willing to participate in a study. Also, it's not clear how this picture would change in a younger population. As far as next steps, we would also like to explore other possible predictors from the information we have collected. We have extensive data on geriatric syndromes, depression and loneliness in this group that may help paint a clearer picture. 


I work with doctor Victor Valcour on his studies focused on aging and HIV. As an HIV program set within a dementia research center, we have a unique opportunity to compare HAND with other neurodegenerative diseases. Our location in San Francisco also offers us the chance to work with a large geriatric population, many of whom have been living with HIV for 30 years or more


This study is testing mindfulness based stress reduction as an intervention to address these cognitive symptoms, as well as symptoms of stress, anxiety and depression. It is a five year study that is funded by the NINR. We started screening participants in the spring of 2015 with the goal of enrolling 180 individuals and are are currently a little over halfway enrolled. When interested individuals contact us, the first step is a phone interview. 


We also assess for cognitive symptoms using the patients assessment of own functioning or PAOF. We offer people the choice to do the PAOF over the phone or through an online survey. It takes about 10 minutes to complete and asks them to rate how frequently they experience things like forgetting someone's name or forgetting to pay a bill. The symptoms fall under three domains - memory, language and communication and higher level, or executive functioning. Anyone who describes at least one of these issues occurring more than just once in a while is eligible to come for in-person screening. I should also note that there are a couple of questions that inquire about motor functioning, but they were not included in this analysis
Anyone who passes the primary screen is invite for an in-person evaluation and this evaluation is extensive. They spend six hours with us and undergo cognitive testing, functional testing, a geriatric syndromes assessment and a neurological exam. We also run clinical labs to confirm suppression and look for other treatable causes.


Our cognitive testing battery lasts 90 minutes and explores seven cognitive domains. These are the core components of the battery - as you can see, these are standard tests and are used throughout our center. The neuropsychologist who administers the tests also has the option to add or remove individual tests to ensure that we're getting a clear picture of the cognitive profile. 
After this visit, every case is reviewed at conference, attended by two physicians, two neuropsychologists and other research staff. Guided by Fracati criteria, we determine if they are objectively impaired in at least two domains and if they meet criteria for MND. We also look for other medical conditions that could cause the impairment. Anyone who fits criteria and is eligible for MRI undergoes brain imaging and is then enrolled in an 8-week mindfulness class. After completing the class, they are followed for one year.


In a few years I may be able to come back and talk to you about whether mindfulness is having any impact on symptoms, but for now, we are focused on further exploring data from the screening process. 


As of April, we had completed in-person screening for 127 individuals. Among that group of people who have endorsed symptoms on the PAOF, about 30% actually performed within expectations for their age and level of education. I should note that during conference, we do our best to take into account each person's baseline cognitive functioning, but there are times that we feel that our tests are just not sensitive enough to pick up change in people with very high baselines. Of the remaining 70%, we saw 19 individuals who were objectively impaired, but had another condition that made HAND diagnosis impossible. In other words, we could not attribute their impairment exclusively to HIV. Some confounds we've seen have included major depression, stroke or motor vehicle accidents that resulted in persistent cognitive deficits. These individuals with confounds were excluded from our analyses.


First, we looked at demographic variables to see if any of these factors could predict impairment. Remember, everyone represented here is virologically suppressed on stable cART.  We found that there was no significant difference between the impaired and unimpaired groups in age (p=0.138), education (p=0.159) or current CD4 cell count (p=0.702). There does seem to be a significant difference between the two groups in terms of ethnicity, but this may just be an indication that our norms may not be ideal for all races. We repeated our analyses excluding non-white participants and did not see any change in the results. 
We wondered if it would be possible to predict impairment based on the type of symptoms that people endorsed. And the answer seems to be - not really.  Here we compared the two groups to see which type of symptoms they were reporting on the PAOF. We did not find a significant difference in any of the three domains that are covered. Unsurprisingly, we see that for both groups, symptoms in the memory domain seem to be most frequently reported. But overall, it looks like characterizing the type of symptoms is not a good way to predict how people will perform on testing


⋅We also asked if the number of symptoms they endorsed on the PAOF (meaning "how many things did they say happened more than just once in a while") might predict how they would do on testing. Here there does seem to be a trend indicating that people with a higher symptom burden tend to have a lower overall neuropsych performance.
⋅Pearson correlation: r=-0.2624 p=0.0061 - significant
⋅Excluded those with confounds


We also looked at the cognitive testing profile of the two groups. As expected, they differed significantly in each domain assessed. We noted that even in the unimpaired group, motor performance seems to be a weakness, while memory is relatively preserved. 
⋅Repeated excluding non-Caucasian: in case there are questions