Send to

Choose Destination
AIDS. 2015 Mar 13;29(5):547-57. doi: 10.1097/QAD.0000000000000573.

Multivariate normative comparison, a novel method for more reliably detecting cognitive impairment in HIV infection.

Author information

aDepartment of Radiology, Academic Medical Center bDepartment of Neurology, Academic Medical Center cDepartment of Global Health, Academic Medical Center, and Amsterdam Institute for Global Health and Development dDepartment of Internal Medicine, Division of Infectious Diseases, Center for Infection and Immunity Amsterdam (CINIMA), Academic Medical Center eInfectious Diseases Research, Public Health Service fDepartment of Neurology, Onze Lieve Vrouwe Gasthuis gHIV Monitoring Foundation hDepartment of Psychology, University of Amsterdam, Amsterdam, The Netherlands. *Both Tanja Su and Judith Schouten shared first authorship.



The objective of this study is to assess whether multivariate normative comparison (MNC) improves detection of HIV-1-associated neurocognitive disorder (HAND) as compared with Frascati and Gisslén criteria.


One-hundred and three HIV-1-infected men with suppressed viremia on combination antiretroviral therapy (cART) for at least 12 months and 74 HIV-uninfected male controls (comparable regarding age, ethnicity, sexual orientation, premorbid intelligence and educational level), aged at least 45 years, underwent neuropsychological assessment covering six cognitive domains (fluency, attention, information processing speed, executive function, memory, and motor function). Frascati and Gisslén criteria were applied to detect HAND. Next, MNC was performed to compare the cognitive scores of each HIV-positive individual against the cognitive scores of the control group.


HIV-infected men showed significantly worse performance on the cognitive domains of attention, information processing speed and executive function compared with HIV-uninfected controls. HAND by Frascati criteria was highly prevalent in HIV-infected [48%; 95% confidence interval (95% CI) 38-58] but nearly equally so in HIV-uninfected men (36%; 95% CI 26-48), confirming the low specificity of this method. Applying Gisslén criteria, HAND-prevalence was reduced to 5% (95% CI 1-9) in HIV-infected men and to 1% (95% CI 1-3) among HIV-uninfected controls, indicating better specificity but reduced sensitivity. MNC identified cognitive impairment in 17% (95% CI 10-24) of HIV-infected men and in 5% (95% CI 0-10) of the control group (P = 0.02, one-tailed), showing an optimal balance between sensitivity and specificity.


Prevalence of cognitive impairment in HIV-1-infected men with suppressed viremia on cART estimated by MNC was much higher than that estimated by Gisslén criteria, while the false positive rate was greatly reduced compared with the Frascati criteria.



[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wolters Kluwer Icon for Universiteit van Amsterdam Library, Netherlands
Loading ...
Support Center