Lisa M. Kuhns, Anna L. Hotton, Rob Garofalo, Abigail L. Muldoon, Kaitlyn Jaffe, Alida Bouris, Dexter Voisin, John Schneider
9 min
Medication adherence among youth living with HIV (YLH) is notoriously challenging, often hindered by a range of co-morbid psychosocial factors. While prior research has established that individual factors like depression or substance use contribute to poor health outcomes, this study investigates whether these factors cluster together in a 'syndemic'—an additive, synergistic group of conditions—to further exacerbate the challenges of ART adherence and viral suppression in young populations.
The researchers analyzed cross-sectional data from 212 YLH (ages 16–29) recruited between 2011 and 2014 from two sites in Chicago. The study team measured five specific psychosocial indicators: depressive symptoms, anxiety symptoms, moderate/heavy alcohol use, moderate/heavy marijuana use, and HIV-related stigma. These were aggregated into a 'syndemic index' (ranging from 0 to 5) to assess their combined impact on self-reported ART adherence and objective viral load suppression.
The study identified a high prevalence of psychosocial challenges, with 46% of participants reporting high HIV-related stigma and significant portions experiencing depression, anxiety, and substance use. The analysis revealed a clear linear dose-response relationship: as the number of syndemic conditions increased, both self-reported medication adherence and the likelihood of viral load suppression significantly decreased. This suggests that the cumulative burden of these issues creates a much greater barrier to effective HIV treatment than any single condition alone.
Effective HIV treatment relies on consistent, high-level adherence to ART. This study highlights that standard, one-size-fits-all adherence interventions may be insufficient for YLH who face multiple, compounding barriers. The results suggest that clinical care for this population should integrate mental health, substance use, and social support services directly with biomedical HIV treatment to improve long-term health outcomes and public health efforts aimed at preventing further HIV transmission.
Medication adherence among HIV-infected individuals is critical to limit disease progression and onward transmission. Evidence indicates that among youth living with HIV (YLH), adherence is suboptimal and related to co-morbid psychosocial conditions. Cross-sectional data from 212 YLH, ages 16-29, collected between 2011-2014 in Chicago were analyzed to assess the relationship of multiple psychosocial conditions (e.g., depressive symptoms, anxiety symptoms, moderate/heavy marijuana use, moderate/heavy alcohol use, HIV-related stigma) to ART adherence (i.e., a "syndemic.") Adherence was regressed on an index of increasing numbers of psychosocial conditions, controlling for demographic and treatment factors as well as enrollment site. The mean age of participants was 24, 89% were male, 87% black, and 91% behaviorally infected. Psychosocial conditions were prevalent, including 38% and 34% with high depressive and anxiety symptoms, respectively, 54% and 25% with a moderate/high level of marijuana and alcohol use, respectively, and 46% reporting high HIV-related stigma. In regression analysis, the likelihood of ART adherence decreased with the number of syndemic conditions (linear dose response, p = 0.02) as did the odds of viral load suppression (p = 0.008). Interventions to address these conditions in concert with biomedical treatment as prevention for YLH are needed.
Alex: Were both groups selected because they were already having trouble taking medication?
Sam: Only the north side trial required reported adherence problems. The south side trial recruited recently diagnosed Black young men who reported sex with men and had a supportive person in their network. So this was not a representative sample of all young people with HIV.
Alex: And although the data came from trials, the comparison here wasn’t randomized. It was a snapshot before the interventions.
Sam: That’s the design to keep in mind. Most participants were Black, male, and had acquired HIV behaviorally rather than at birth. The researchers adjusted their models for recruitment site, age, medication count, and gender identity. Adjustment helps address measured differences, but it doesn’t erase selection.
Alex: Now make the index concrete. What counted as a condition, and what would move someone into a higher category?
Sam: The index counted five indicators: high depressive symptoms, high anxiety symptoms, high HIV-related stigma, and moderate or high-risk alcohol and marijuana use. Each indicator contributed equally. For example, high depressive symptoms and moderate or high-risk marijuana use would each add a point.
Alex: Those labels could hide important differences. Were they clinical diagnoses, or thresholds on questionnaires?
Sam: They were questionnaire thresholds, not a set of confirmed diagnoses. Depression and anxiety used standard symptom cutoffs. Substance-use measures assessed use and associated problems. High stigma meant scoring above the sample’s average on a youth-focused scale, not crossing a clinical diagnostic threshold.
Alex: So the count is easy to interpret, but it treats unlike concerns as equivalent. How did they measure medication-taking?
Sam: Participants rated the percentage of each HIV medicine they had taken during the previous month. The researchers averaged those ratings across their antiretroviral therapy, the medicines used to treat HIV. They also obtained viral load from medical records: the amount of HIV measured in blood.
Alex: A blood measure gives another perspective, but it isn’t identical to someone’s reported doses. How was it used?
Sam: Viral suppression meant having a viral load below the study’s specified threshold. Among participants with those records available, fifty-nine percent were suppressed. The authors used suppression as a secondary outcome, because viral load data were missing for some participants.
Alex: What is the strongest result to hold onto? I’d rather hear one comparison than every category.
Sam: In the adjusted model, participants with four indexed conditions reported adherence about twenty-two percentage points lower than participants with none. That’s a difference on the reported medication-taking scale, not a relative increase in missed-dose risk. The confidence interval excluded no difference, although it was wide.
Alex: Did adherence keep falling at every higher count? Otherwise, a smooth dose-response description could overstate the pattern.
Sam: It did not fall at every step. The estimate became more negative through the category we just discussed, then weakened in the category with every indexed condition. That highest category’s confidence interval included no difference. Still, the overall statistical test supported a declining linear trend.
Alex: So there’s evidence for a cumulative pattern, but not a clean staircase. What happened when they checked viral suppression?
Sam: Higher condition counts were also associated with lower odds of suppression. That trend was statistically significant after the same adjustments. My reading is that this strengthens the finding beyond self-report alone, but the biological analysis used fewer participants because of missing records.
Alex: Did each component individually relate to adherence? If not, the count might conceal which concerns matter most.
Sam: Stigma, anxiety, and depressive symptoms were negatively correlated with adherence. Alcohol and marijuana risk were not individually significantly correlated with adherence in that analysis. The count’s association therefore doesn’t establish equal contributions from its components, or identify the best treatment target.
Alex: The main causal objection remains: these measurements were collected together. Poor adherence and psychosocial distress could be connected without the count explaining the direction.
Sam: The snapshot cannot establish which came first, or whether changing the conditions would improve adherence. Generalization is limited too, given the urban, predominantly Black male sample and selective recruitment. The study also omitted childhood sexual abuse and physical victimization, which other syndemic research has included.
Alex: With those limits, what do the authors actually propose for care, rather than claim they’ve proved?
Sam: They propose screening and addressing mental health, substance use, and stigma alongside medication support. Their discussion also cites the north side intervention’s follow-up: its benefit diminished over time among participants with combined depressive symptoms and marijuana risk. That motivates testing more intensive, multicomponent support; it doesn’t establish its effectiveness.
Alex: Who should read the full paper, and where should they begin?
Sam: Researchers designing youth HIV adherence interventions should read it, along with clinicians studying psychosocial support. Start with the measures: the thresholds determine what this count means. Then read the adjusted regression results, including the weakening estimate at the highest count, and the discussion of intervention durability.
Alex: And for listeners outside that area, what is the line to carry into a conversation with colleagues?
Sam: In this Chicago sample, more psychosocial difficulties went with poorer medication-taking and lower odds of viral suppression. The practical lesson is a question for intervention research, not a proven prescription: can addressing clustered concerns make adherence support more durable? Remember that medication-taking difficulties may come bundled with other difficulties.
Alex: Thanks for listening.