Study: lower memory performance was linked to higher HIV self-stigma in men with HIV, mediated by mental quality of life

Internalized HIV self-stigma, the process in which a person living with HIV absorbs society's prejudices about the condition, has been associated with avoidance, social exclusion, psychological distress and poorer engagement with care. The authors noted a recent study estimated nearly two-thirds of men who have sex with men (MSM) with HIV had moderate-to-high HIV self-stigma, and that memory is the most commonly affected cognitive ability in people with HIV (HIV-associated neurocognitive impairment estimated at 27.4% in one cited figure).^[1]

A 2021 cross-sectional study in Psychology Research and Behavior Management evaluated, in MSM living with HIV, how cognitive functioning, health-related quality of life (HRQoL) and HIV self-stigma related, and, the authors state, was the first to test whether HRQoL mediated that link.

What the study tested

The study was a single-center, cross-sectional analysis at the Infectious Diseases Unit of the General University Hospital of Alicante (Spain). Eligible participants had an HIV diagnosis, were aged 18 or older, were men who have sex with men, were on antiretroviral treatment, and gave informed consent. Comorbidities, dementia or other central nervous system disease, mental-health diagnoses, chronic hepatitis, active cancer, diabetes, hypertension, cardiovascular disease, hypothyroidism or malnutrition, were excluded.

From a clinic cohort of 900 patients with HIV, 630 met criteria; after a power calculation (required sample 62), 72 were enrolled and 2 failed to complete the questionnaires, leaving 70 HIV-positive MSM. Mean age was 45.41 years (SD = 12.67); 54 (77.1%) were Spanish and 55 (78.6%) employed. Mean time since diagnosis was 9.85 years (SD = 6.42), and 66 of 70 (95.7%) had a viral load ≤50 copies/mL.^[1]

Cognition was measured with the Online General Cognitive Assessment Battery (CogniFit, Inc.), a roughly 30-minute test of five domains, reasoning, memory, attention, coordination and perception. Per the platform's normative data, the highest score per domain was 800; scores over 400 indicated adequate performance and lower scores low performance, the threshold used to split participants into high- and low-performance groups.

HIV self-stigma used the HIV Internalized Stigma Scale (HIV-ISS), a 10-item self-report scored 10 to 50 (higher = greater stigma). HRQoL used the Spanish Medical Outcomes Study-HIV (MOS-HIV), which yields a Physical Health Summary (PHS) and Mental Health Summary (MHS), each on a 0–100 scale (higher = better). Pearson correlations, a hierarchical regression, and a bootstrapped (10,000-resample) PROCESS mediation tested how memory, PHS and MHS predicted self-stigma.

What it found

Among the five domains, only memory was significantly correlated with HIV self-stigma (r = −0.276, p < 0.05): lower memory, higher stigma. Memory was also the only domain positively correlated with both quality-of-life summaries (PHS r = 0.257, MHS r = 0.291, both p < 0.05); reasoning, attention, coordination and perception showed no significant association with self-stigma. Both HRQoL summaries were strongly negatively related to self-stigma (PHS r = −0.428, MHS r = −0.596, both p < 0.01).^[1]

Split by memory performance, low-memory participants scored significantly higher on internalized stigma: t(67.418) = 2.709, p = 0.009, d = 0.66, 95% CI [1.21, 8.02], a medium-sized difference. Memory was the only domain to produce it; the other four showed no group differences (p > 0.05).^[1]

In a hierarchical regression with all five domains entered, only memory was significant, explaining 5.6% of the variance in self-stigma (β = −0.551, p = 0.005). With PHS and MHS added, memory dropped out and only MHS remained significant (β = −0.674, p < 0.001); the model then explained 31.6% (ΔR² = 0.261, p < 0.001).^[1]

How quality of life mediated the link

A multiple-mediation model entered memory as predictor, PHS and MHS as parallel mediators, and HIV self-stigma as outcome. Memory predicted both PHS (B = 0.017, p = 0.03) and MHS (B = 0.026, p = 0.01), but among the mediators only MHS predicted self-stigma (B = −0.400, p = 0.0001). The indirect effect through MHS was significant (−0.010, SE = 0.005, bias-corrected 95% CI [−0.023, −0.001]); the PHS path was not.^[1]

Critically, once MHS entered the model, the direct memory–stigma relationship was no longer significant (B = −0.006, p = 0.261), consistent with full mediation by the mental component of quality of life. The model was significant, F(3,66) = 13.162, p = 0.00001, predicting 37.4% of the variance in self-stigma. In the standardized path diagram, the total memory→stigma effect (c = −0.275, p < 0.05) fell to a non-significant direct effect (c′ = −0.115) once the MHS path (b = −0.675, p < 0.01) was modeled.^[1]

Limitations, what this does NOT show

The authors stated that the effect of HIV self-stigma on cognition or HRQoL "is likely to be bidirectional, and additional research is necessary to clarify this issue."

No causality. The cross-sectional design "only allowed measuring the association between the variables; it did not enable establishing any causality." The direction between memory, quality of life and self-stigma cannot be settled from these data.

Self-report and small, regional sample. Questionnaires were self-administered, so "some participants may not have been sincere in their responses" (possible desirability bias). The sample was small (N = 70) and drawn from a single Spanish region, which the authors noted "may limit the generalisability of its conclusions."^[1]

Unmeasured confounders. No social HIV determinants, income, social support or isolation, were measured, and the authors acknowledged these "could be possible confounders of the obtained results." This study tested no treatment; it measured cognitive performance and its statistical relationships with stigma and quality of life.

Where this fits in the broader evidence

The findings echo work in other stigmatized populations: the authors cited research in people with schizophrenia in which neurocognitive deficits, executive dysfunction in particular, limited the ability to reject society's stereotypes, reducing "stigma resistance." Memory, one of the domains most related to executive functioning, was the only one linked to self-stigma here. For the full list of peer-reviewed research using CogniFit's technology, see the research index; related pages include the long COVID brain-fog pilot and the brain-training apps scoping review.