The association of perceived stress and verbal memory is greater in HIV-infected versus HIV-uninfected women

Abstract

In contrast to findings from cohorts comprised primarily of HIV-infected men, verbal memory deficits are the largest cognitive deficit found in HIV-infected women from the Women's Interagency HIV Study (WIHS), and this deficit is not explained by depressive symptoms or substance abuse. HIV-infected women may be at greater risk for verbal memory deficits due to a higher prevalence of cognitive risk factors such as high psychosocial stress and lower socioeconomic status. Here, we investigate the association between perceived stress using the Perceived Stress Scale (PSS-10) and verbal memory performance using the Hopkins Verbal Learning Test (HVLT) in 1009 HIV-infected and 496 at-risk HIV-uninfected WIHS participants. Participants completed a comprehensive neuropsychological test battery which yielded seven cognitive domain scores, including a primary outcome of verbal memory. HIV infection was not associated with a higher prevalence of high perceived stress (i.e., PSS-10 score in the top tertile) but was associated with worse performance on verbal learning (p < 0.01) and memory (p < 0.001), as well as attention (p = 0.02). Regardless of HIV status, high stress was associated with poorer performance in those cognitive domains (p's < 0.05) as well as processing speed (p = 0.01) and executive function (p < 0.01). A significant HIV by stress interaction was found only for the verbal memory domain (p = 0.02); among HIV-infected women only, high stress was associated with lower performance (p's < 0.001). That association was driven by the delayed verbal memory measure in particular. These findings suggest that high levels of perceived stress contribute to the deficits in verbal memory observed in WIHS women.

Fig. 1

Perceived stress (lower vs. higher) is associated with the verbal memory domain and delayed recall only in the context of HIV. Note. *p<0.001. There was a significant interaction between perceived stress and HIV status on the verbal memory domain (p=0.03) and specifically on delayed recall (p=0.02). Among HIV-infected women, women with higher perceived stress performed worse than women with lower perceived stress on the verbal memory domain and on delayed recall (p’s< 0.001). Conversely, among HIV-uninfected women, women with higher and lower perceived stress performed similarly on the verbal memory domain and on delayed recall (p’s>0.76). The model is adjusted for marijuana use; crack, cocaine, and/or heroin use; smoking; hazardous alcohol use; antidepressants; HCV; income; study site; and number of exposures to the Hopkins Verbal Learning Test

Fig. 2

The interactive association between perceived stress (lower vs. higher) and viral load on the verbal memory domain in HIV-infected women. Note. **p<0.01. There was a significant interaction between perceived stress and viral load on the verbal memory domain (p=0.04). Among HIV-infected women with viral loads ≥10,000 cp/ml, those women with higher perceived stress performed significantly worse than women with lower perceived stress (B=−4.54, SE=1.61, p < 0.01). HIV-infected women with higher and lower perceived stress performed similarly if their viral loads were <10,000 cp/ml (B=0.23, SE=1.13, p=0.84) or undetectable (B=−1.17, SE=0.93, p=0.21). The model is adjusted for marijuana use; crack, cocaine, and/or heroin use; smoking; hazardous alcohol use; antidepressants; HCV; income; study site; number of exposures to the Hopkins Verbal Learning Test; recent CD4 count and viral load; nadir CD4 count; and cART use and adherence