Persistent neuroimmune alterations in children who are HIV-exposed but uninfected at age 6-7 years: Associations with language development in a South African birth cohort

This study of a South African birth cohort reveals that perinatal HIV exposure in uninfected children is associated with persistent alterations in immune development and neurometabolites, which correlate with reduced language skills and disrupted neuroimmune pathways critical for school readiness.

The Gist

The Big Picture: A Tale of Two Groups of Children

Imagine a large group of children in South Africa growing up in a community where HIV is common. The researchers are looking at two specific groups:

1. The "Unexposed" Group (HU): Children whose mothers did not have HIV.
2. The "Exposed but Uninfected" Group (HEU): Children whose mothers did have HIV, but thanks to medicine, the children were born healthy and do not have the virus themselves.

The Mystery: Even though the HEU children don't have the virus, they often struggle more with learning and school readiness than their peers. The researchers wanted to know: Why? What is happening inside their bodies and brains that makes school harder for them?

The Investigation: A 7-Year Detective Story

The researchers followed these children from before they were born until they were 6 or 7 years old (the age they start school). They looked at three main things:

1. The "Alarm System" (Immune System): They took blood samples to check for "cytokines." Think of these as chemical smoke signals or alarm bells the body rings when there is inflammation or stress.
2. The "Brain Chemistry" (MRS Scan): They used a special MRI machine (like a high-tech camera for brain chemicals) to look at the brain's "fuel" and "building materials."
3. The "School Report Card" (ELOM): They tested the children's language, math, and motor skills to see how ready they were for school.

The Findings: What They Discovered

1. The Mother's "Smoky Kitchen"

During pregnancy, the mothers with HIV had a different "smoke signal" profile than mothers without HIV. Their bodies were constantly ringing a low-level alarm (high levels of a marker called sCD14).

• The Analogy: Imagine the mother's body is a kitchen. In the HIV-positive mothers, the smoke detector was going off constantly, even if there wasn't a fire. This "smoke" (inflammation) crossed the placenta and reached the baby's developing brain, potentially confusing the baby's own construction crew.

2. The Children's "Broken Construction Crew"

When the children grew up to age 6–7, the researchers found that the HEU children's brains looked slightly different chemically:

• Less "Glue" (Choline): In the white matter (the brain's wiring), there was less "choline." Think of choline as the insulation tape on electrical wires. If the insulation is thin or missing, the electrical signals (thoughts) might get lost or slow down.
• Less "Fuel" (Glutamate): In the grey matter (where thinking happens), there was less glutamate. Think of this as gasoline for the brain's engine. Less fuel means the engine runs a bit quieter or weaker.

3. The "Smoking Gun": IL-8

The study found one specific chemical, IL-8, that was a major troublemaker.

• The Analogy: IL-8 is like a hyperactive security guard. In healthy children, this guard calms down as they grow up. But in the HEU children, this guard kept running around shouting "Fire!" even when there was no fire. This constant shouting (inflammation) seemed to mess with the brain's ability to build its wiring (white matter) properly.

4. The School Connection

The results were clear:

• The HEU children scored lower on language tests than the unexposed children.
• In the healthy (unexposed) group, children with better "insulation tape" (choline) in their brains spoke better and were more ready for school.
• Crucially: In the HEU group, this link was broken. Even if they had decent "insulation," it didn't seem to help their language skills as much. It's as if the construction crew was so confused by the constant "smoke signals" that the good insulation didn't work as well.

The Takeaway: Why This Matters

This study tells us that HIV exposure changes the body's immune "software" permanently, even if the child never gets the virus.

• The Metaphor: Imagine two cars. One is a brand-new car (Unexposed). The other is a car that was built in a factory where the air was full of dust (Exposed). Even if the second car is cleaned and runs fine, the engine parts might be slightly worn, and the sensors might be hypersensitive.
• The Result: Because of this "dust" (immune dysregulation), the HEU children's brains develop differently. They have trouble with the "wiring" (white matter) and "fuel" (glutamate) needed for language.

What Should We Do?

The researchers suggest that we can't just treat the virus; we need to treat the inflammation.

• For Doctors: We might need to monitor these children's immune systems more closely, not just for infections, but for signs of this "false alarm" inflammation.
• For Parents: Early intervention is key. Since language is the area most affected, giving these children extra support with reading and talking at home and school could help bridge the gap.

In short: Being exposed to HIV in the womb leaves a "scar" on the immune system that confuses the brain's construction crew. This leads to slower wiring and less fuel for language, making school a bit harder to start. But by understanding this, we can find new ways to help these children thrive.

Technical Summary

1. Problem Statement

Children who are HIV-exposed but uninfected (HEU) face increased risks of neurodevelopmental delays, mortality, and morbidity compared to their HIV-unexposed (HU) peers, despite not contracting the virus. While maternal immune activation (MIA) during pregnancy is hypothesized to disrupt neurogenesis and myelination, the specific neuroimmune pathways linking perinatal HIV exposure to long-term school readiness remain unclear. Previous studies have shown heterogeneous results regarding cytokine profiles and limited data on how these early immune dysregulations persist into school age and correlate with brain neurometabolites and functional outcomes. This study aims to map the longitudinal trajectory of immune markers from pregnancy to age 5, their association with brain neurometabolites at age 6–7, and their impact on school readiness, specifically focusing on the HEU population in a high-burden South African setting.

2. Methodology

Study Design & Cohort:

• Cohort: A sub-study of the Drakenstein Child Health Study (DCHS), a population-based birth cohort in the Western Cape, South Africa.
• Participants: 268 children aged 6–7 years (94 HEU, 174 HU).
• Exclusion Criteria: Congenital anomalies, genetic disorders, confirmed HIV infection in the child, low Apgar scores, neonatal ICU admission, maternal substance use, or MRI contraindications.

Data Collection:

• Immune Markers: Peripheral blood serum samples were collected from mothers during pregnancy (26–28 weeks) and from children at 6 weeks, 2, 3, and 5 years.
  • Assays: A 13-plex Luminex panel (cytokines: GM-CSF, IFN-γ, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12p70, IL-13, TNFα) and ELISA for sCD14, sCD163, NGAL, MMP-9, and YKL-40.
• Neuroimaging (MRS): Magnetic Resonance Spectroscopy was performed at 6–7 years using a Siemens 3T Skyra scanner.
  • Regions of Interest (ROIs): Midline parietal grey matter and left parietal white matter.
  • Metabolites: Glutamate (neuronal function), Myo-inositol (glial activation/inflammation), Total Choline (membrane turnover/myelination), and N-acetyl-aspartate (neuronal integrity).
  • Processing: LCModel software was used for quantification; data were creatine-referenced ratios and absolute concentrations.
• Neurodevelopmental Assessment: School readiness was assessed using the culturally validated Early Learning Outcome Measure (ELOM), covering language, literacy, motor skills, and cognitive domains.

Statistical Analysis:

• Trajectory Modeling: Linear mixed-effects models (LMMs) were used to characterize immune marker trajectories from 6 weeks to 5 years.
• Associations: Linear regression models with robust standard errors tested associations between:
  1. Immune markers and neurometabolite ratios.
  2. Neurometabolites and ELOM scores.
  3. Immune markers and ELOM scores.
• Adjustments: Models were adjusted for child age, sex, and voxel tissue composition. Multiple comparisons were corrected using the Benjamini–Hochberg (BH) false discovery rate. Mediation analyses were conducted to test if immune/neurometabolite factors mediated the link between maternal HIV and ELOM scores.

3. Key Contributions

• Longitudinal Neuroimmune Mapping: This is one of the few studies to track immune trajectories from pregnancy through school entry in HEU children, linking them directly to in vivo brain chemistry.
• Developmental Shift in Drivers: The study identifies a critical developmental shift where associations between immune markers and brain metabolism transition from being driven by maternal markers in early life to child markers by school entry.
• Specific Biomarker Identification: It highlights IL-8 as a consistent correlate across developmental stages and identifies specific divergent immune trajectories (e.g., sCD14, GM-CSF, IL-5, YKL-40) in HEU children.
• Functional Correlates: It links specific neurometabolite deficits (glutamate and choline) to functional language deficits in HEU children, providing a biological mechanism for observed neurodevelopmental delays.

4. Key Results

Immune Profiles:

• Mothers: HIV-positive mothers had higher sCD14 (monocyte activation) but lower GM-CSF, NGAL, and MMP-9 compared to HIV-negative mothers.
• Children (Trajectories): HEU children exhibited distinct immune trajectories compared to HU peers:
  • Higher baseline sCD14 with a divergent non-linear trajectory.
  • Declining IL-5 (vs. increasing in HU).
  • Lower baseline GM-CSF and IL-10.
  • Non-linear increase in YKL-40 (vs. decline in HU).
• IL-8: Emerged as a consistent immune correlate across stages, showing differential associations with neurometabolites based on HIV exposure status.

Neurometabolite Differences (Age 6–7):

• HEU vs. HU: HEU children had significantly lower glutamate/creatine ratios in midline parietal grey matter and lower total choline/creatine ratios in left parietal white matter.
• Trends: HEU children showed a trend toward higher absolute myo-inositol concentrations in white matter (suggesting persistent glial activation).

Associations with Neurometabolites:

• Shift in Drivers: In early life, maternal markers predicted child brain chemistry; by school entry, child serum markers (specifically IL-8 trajectories) were the primary drivers.
• HEU Specifics: In HEU children, IL-8 trajectories were negatively associated with grey matter choline ratios.
• HU Specifics: In HU children, declining trajectories of IL-8, TNFα, and IL-4 were associated with lower myo-inositol ratios, suggesting a typical maturation of neuroimmune homeostasis.

School Readiness (ELOM):

• Language Deficits: HEU children scored significantly lower on language and literacy domains compared to HU peers.
• Neurometabolite-Outcome Link: In HU children, higher left parietal white matter choline ratios were positively associated with language and overall school readiness. This association was absent in HEU children, suggesting a decoupling of white matter maturation from functional language outcomes in the HEU group.
• Mediation: No single marker or neurometabolite significantly mediated the relationship between maternal HIV and ELOM scores, implying a complex, multi-pathway mechanism.

5. Significance and Implications

• Persistent Neuroimmune Dysregulation: The study provides evidence that perinatal HIV exposure induces persistent alterations in the child's immune system and brain metabolism that extend into school age, even in the absence of active infection.
• Mechanism of Language Delay: The findings suggest that the language deficits observed in HEU children may be driven by disrupted white matter maturation (indicated by low choline) and altered excitatory neurotransmission (low glutamate), potentially exacerbated by chronic, dysregulated peripheral inflammation (e.g., IL-8).
• Clinical and Policy Relevance:
  • Biomarkers: IL-8 and sCD14 are identified as potential prognostic markers for neurodevelopmental risk in HEU children.
  • Intervention: The results underscore the need for optimizing antenatal immune health and monitoring postnatal immune mediators to identify children at risk for language delays.
  • Future Research: Highlights the necessity of integrating cytokine network analysis with advanced neuroimaging to clarify pathways linking immune dysregulation to white matter maturation and to evaluate the specific impacts of antiretroviral exposures.

Limitations: The study was limited by sample size reduction due to strict quality control for MRS data (particularly for absolute concentrations in white matter), which reduced statistical power for some subgroup analyses. Additionally, the single-voxel MRS approach limits the ability to assess network-wide heterogeneity.