Emerging Research Pilots

When a person’s arteries become stiff (called “aortic stiffness”), it can leave them vulnerable to neurocognitive impairment and future organ damage. However, we do not know enough about whether and how aortic stiffness contributes to other comorbidities (illnesses) in people living with perinatal HIV infection (PHIV) or with perinatal HIV-exposure but uninfected (PHEU). 

For this Emerging Research Pilot study, investigators hypothesize that higher rates of metabolic abnormalities and inflammation could make youth with PHEU (ages 9-14) and young adults living with PHIV (ages 22-29) more likely to develop arterial stiffness. Having stiff arteries could put these youth and young adults at risk for altered brain structure and function, which could eventually cause organ damages. Investigators will measure aortic stiffness among youth and young adults in the TERBO BRAIN study by using a measure called carotid-femoral pulse-wave velocity (cfPWV). This involves sending a painless and invisible pulse wave up and down the body to see how long it takes to travel through the arteries. When someone has arterial stiffness, the wave will take longer to travel through the body. Researchers will then compare these youth and young adults to a group of similar individuals from the general pediatric and young adult population.

Investigators will: 

1. Look at differences in arterial stiffness between young adults living with PHIV and youth with PHEU, and then compare them to similar individuals from the general population. 
2. Attempt to understand how and why aortic stiffness happens. 
3. Describe how aorterial stiffness may alter brain structure and function in youth with PHIV or PHEU by analyzing the cfPWV data collected in the TERBO BRAIN study. 

Aortic stiffness is not yet widely recognized as a possible consequence of perinatal HIV and in utero HIV exposure. Yet it is a factor that could be important to understand because it could be a factor that puts a person at higher risk for other comorbidities, possibly resulting in organ and brain damage. This study aims to demonstrate that cfPWV could be a simple and noninvasive tool to identify individuals with aortic stiffness. It could also eventually establish aortic stiffness as a known early risk factor among youth with PHIV or PHEU, helping to prevent them from developing additional illnesses as a result.

In the United States, there are many different dialects of English; some examples include Appalachian English, African American English, General American English, Creole-English, and Spanish-influenced English. These dialects of English have different but equally complex sets of rules, yet some are more stigmatized than others. 

Researchers have found that some children may be misdiagnosed as language impaired if there is a mismatch between the dialect on the test and the child’s dialect. One possible reason for the misdiagnosis could be test biases. Most standard language tests use General American English to evaluate and diagnose children’s language. Yet for children who speak a different dialect or speak English as a second language, these tests could give false and harmful results. One way to correct for this bias is to modify the scoring to be sensitive to markers of weak language within and across all dialects of English. 

As part of their participation in SMARTT, children complete a language test called the Clinical Evaluation of Language Fundamentals (CELF-4) at ages 5 and 9. This Emerging Research Pilot study seeks to determine whether dialect biases exist in the CELF-4, and to develop a better approach to reduce this bias.

The project will test three scoring approaches: 

• Unmodified scoring: Scores the test exactly as the manual recommends. This may over-identify language impairment (some kids without language impairment are falsely diagnosed).  
• Modified scoring: Modifies scoring to allow for any speech that is consistent with the child’s dialect and does not penalize a child for speaking a dialect that differs from General American English. This may under-identify language impairment because it is not sensitive to developmental differences within that dialect (some kids with language impairment are not diagnosed) 
• Strategic scoring: Modifies scoring if it is consistent with a child’s dialect and sensitive to developmental differences within that dialect. This is predicted to reduce bias while also providing an optimal and more balanced diagnostic outcome.

A group of researchers will listen to anonymous recordings of some children in SMARTT completing their language test at age 5 or 9. For each recording, they will identify the dialect the child speaks and then will use strategic scoring to re-score the language test. They will compare the scoring systems to each other to see if bias was present and reduced in a developmentally appropriate way. Children will only be recorded if they and their parent consent to this. 

It is critical to investigate and adjust language scoring as needed to better detect language impairment that may come about after a child is exposed to ARVs in the womb, and to best support children and families across a diversity of linguistic backgrounds. If the study’s hypotheses are correct, strategic scoring can be applied to the CELF-4 within SMARTT. It will reduce bias and increase validity of the effects of perinatal exposure to antiretroviral treatments on language development.