A longitudinal study of adults with well-controlled HIV found no clear association between the strength of their immune response to cytomegalovirus (CMV) and the later development of cognitive impairment.
The study followed 93 adults for an average of 6.5 years. All participants were receiving treatment that had suppressed HIV for at least six months, and none experienced a viral rebound during follow-up. The findings were reported by Chauhan and colleagues in The Journal of Infectious Diseases.
What the study found
At the beginning of the study, 48 participants had normal cognition but later developed impairment. They were compared with 45 matched participants who remained free of impairment. The groups were matched on age, sex, CD4 count category and years of follow-up.
Average CMV IgG levels were similar: 9.57 EU/mL among participants who later developed impairment and 9.37 EU/mL among those who did not. After adjustment, the odds ratio for baseline antibody levels was 1.07, with a 95% confidence interval of 0.61 to 1.92.
CMV-reactive T-cell responses, measured with an ELISpot test, also were not associated with later impairment. The adjusted odds ratio was 1.00, with a 95% confidence interval of 0.98 to 1.01. Baseline ELISpot averages were 91 responding cells per million among 26 people who later developed impairment and 109 among 31 who did not.
Over time, antibody levels fell slightly in both groups, while ELISpot responses increased. Neither change was linked to a change in thinking scores. The researchers said the increase in T-cell responses may have reflected silent CMV reactivations, although CMV viral levels were not measured.
How cognition was assessed
Participants completed cognitive testing every 48 weeks using Trail Making Tests A and B and the Digit Symbol test. These assessments primarily covered processing speed, attention and mental flexibility.
Impairment was defined as scoring at least one standard deviation below the norm on two tests, or at least two standard deviations below the norm on one test, after adjustment for age, sex, race and education.
People who later developed impairment had lower baseline thinking scores than controls, despite not meeting the study’s definition of impairment at that time. Median scores were −0.35 in the later-impairment group and 0.73 among controls.
Smaller immune-cell analysis
A flow-cytometry analysis examined immune-cell types in a much smaller subset, with only 14 to 17 people per group. Some cells thought to help protect against CMV tended to be less common among people who later developed impairment, but the study did not perform significance testing for these differences.
For example, the CMV-responsive share of natural killer cells releasing interferon-gamma averaged 3.36 in the later-impairment group and 6.92 among controls. CMV-responsive regulatory T cells averaged 2.59 and 1.39, respectively.
These findings are exploratory rather than evidence of a confirmed relationship. Stored blood-cell samples that could not be used reduced the size of the cell-based analyses and may have made those subsets unrepresentative.
Why the findings differ from earlier research
Earlier cross-sectional studies had linked stronger CMV immune responses with worse cognition in people with HIV. In the US CHARTER cohort, higher CMV antibody levels correlated with poorer overall cognitive performance among 38 people receiving suppressive treatment. A Danish study of 52 people with HIV receiving stable treatment also reported associations involving CMV-specific CD4 T-cell responses, antibody levels and neurocognitive performance.
The new study differs because it followed participants over time. Its results suggest that the measured CMV immune responses did not predict later impairment in this particular group, rather than proving that CMV has no role in cognitive changes associated with HIV.
What remains uncertain
The main antibody analysis included 93 people, while the ELISpot and flow-cytometry analyses were smaller. A modest association could therefore have gone undetected. The confidence interval for the antibody result was also broad, ranging from about 40% lower odds to nearly twice the odds.
Only three participants were CMV-negative, so the study mainly compared different strengths of immune response among people who had evidence of CMV exposure. It did not compare CMV-infected and CMV-uninfected groups. The cognitive tests also did not measure memory, language and other domains in depth.
Neuropathy was twice as common among participants who later developed impairment, at 35.4% compared with 17.8%. Neuropathy, White race and Hispanic ethnicity were linked to cognitive change in the antibody analysis, but the study was designed around CMV immune responses and does not establish why these factors were associated with change.
Because the participants were mostly middle-aged men with well-controlled HIV in US research cohorts, the findings may not apply in the same way to older adults, women or people whose HIV is less controlled.
Key takeaways
- In 93 adults with treatment-suppressed HIV, CMV antibody levels were not clearly associated with cognitive impairment over an average of 6.5 years.
- Baseline CMV-reactive T-cell responses measured by ELISpot were not associated with later impairment.
- Changes in CMV antibody and T-cell responses were not linked to changes in thinking scores.
- The smaller immune-cell analyses were exploratory and did not include significance testing for reported differences.
What this means
The study adds longitudinal evidence to a debate previously informed mainly by cross-sectional research. For this group of adults with well-controlled HIV, the measured CMV immune responses did not predict later cognitive impairment. The findings do not rule out smaller effects, effects in other populations, or relationships involving cognitive domains that were not measured in depth.
What the evidence cannot tell us
- The main antibody analysis included 93 people, and the ELISpot and flow-cytometry analyses included smaller subsets.
- A modest association may have been missed, and the antibody confidence interval was broad.
- Some stored blood-cell samples were unusable, making cell-based analyses smaller and potentially unrepresentative.
- Flow-cytometry differences were described as small, and no significance testing was performed for them? No; the fact sheet says no significance testing was done; this string should not contain typo. Need correct in final? We can edit.
AI tools were used to assist with the preparation of this article under our editorial standards.
Leave a Reply