Two hundred and thirty-eight. That is the total number of slots across every active treatment trial for pediatric Long COVID on earth.
Three trials. One gut barrier drug (48 children). One immunomodulator (up to 160, ages 6–25). One supplement (30 children, open-label). That is the entire pipeline for a condition affecting an estimated 5.8 million U.S. children.
The adult pipeline has over twenty concurrent trials — baricitinib, LDN, semaglutide, IVIG, stellate ganglion block, daratumumab — with thousands of participants. I have covered them across a dozen posts. This post is about the children those trials were never built for.
What the Schools See
Two independent datasets now document what Long COVID does to children's education.
Reeder et al. analyzed 1,976 children from the NIH RECOVER pediatric cohort — 406 school-age (6–11) and 1,570 adolescents (12–17). Every school metric they measured was worse in children with Long COVID:
| Outcome | School-Age (6–11) | Adolescent (12–17) |
|---|---|---|
| Worsened grades | 18% vs 7% (aRR 2.18) | 29% vs 11% (aRR 2.39) |
| Concentration difficulty | 38% vs 14% (aRR 2.52) | 37% vs 11% (aRR 3.26) |
| Social isolation | 28% vs 9% (aRR 2.77) | 43% vs 21% (aRR 1.95) |
| Seeking IEP | 35% vs 22% (RR 1.30) | 27% vs 15% (aRR 1.66) |
Separately, Ford et al. examined 11,057 children aged 5–17 using the National Health Interview Survey. Children with Long COVID were 2.5 times more likely to miss 18 or more school days per year. Memory impairment: 18.3% vs 8.6%. Concentration problems: 14.3% vs 7.7%.
One study is a prospective clinical cohort. The other is a nationally representative survey. They converge on the same finding. A child’s grades dropping is not a measurement artifact. A child missing thirty days of school is not a construct validity problem. These are among the cleanest outcome data in the entire Long COVID literature — functional consequences measured through institutional records, not self-report symptom scales.
What the Clinics See
Miller et al. studied 214 children at the University Hospitals Pediatric COVID Recovery Clinic. Their central finding: Long COVID symptom burden in children — sleep disturbance, anxiety, depression, fatigue — exceeds national norms and exceeds scores reported by pediatric patients with chronic pain, cancer, sickle cell disease, and autism.
The most common interventions the clinic could offer: diet changes (81.8%), pacing (65.9%), sleep hygiene (61.2%). These are management strategies for a condition with no approved therapy.
“Many patients had been dismissed by multiple providers previously as having only anxiety and school avoidance.”
— David W. Miller, Medical Director, Pediatric Integrative Medicine, UH Connor Whole Health
Who Gets Hit Hardest
Rhee et al. studied 4,584 children across 52 RECOVER sites. Food insecurity doubled Long COVID risk (aOR 2.39). Discrimination and low social support doubled it again (aOR 2.17). One protective factor: food security, even amid economic hardship.
In Post #40, I documented how adult Long COVID measurement tools show differential sensitivity across racial groups — the same instrument measuring different things in different populations. The pediatric problem is prior to that. The measurement tools were never built for children in the first place. The treatment infrastructure was not adapted. It was absent.
The Known Path
The biology connecting infection to school failure runs through the gut.
Yonker et al. demonstrated in a Phase 2a trial (n=12, double-blind, randomized) that larazotide — a zonulin blocker that seals the intestinal barrier — accelerates spike protein clearance in children with MIS-C. Spike levels correlated with IFN-γ (P=0.004) and IL-6 (P<0.0001). The mechanism: gut barrier fails, spike translocates to blood, systemic immune activation follows.
This maps onto the pathways I traced in Post #31 (gut barrier and peroxisome repair) and Post #38 (gut-derived extracellular vesicles reprogram macrophages). In Post #18, I documented how COVID rewires children’s metabolism before symptoms appear — which means the treatment window may be earlier and more consequential than in adults.
A larazotide Long COVID trial in children is now enrolling. Forty-eight seats.
The Pipeline
| Trial | Drug | Mechanism | N | Ages | Status |
|---|---|---|---|---|---|
| Yonker (MGH) | Larazotide | Zonulin / gut barrier | 48 | Pediatric | Enrolling |
| RECOVER-TLC | Low-dose naltrexone | Immunomodulation | 160 | 6–25 | Summer 2026 |
| Calgary | Taurine | Antioxidant / metabolic | 30 | Pediatric | Open-label |
| Total slots that include children | 238 | ||||
For comparison: adults currently have active trials for baricitinib (550 patients), abrocitinib, upadacitinib plus pirfenidone (348 patients), IVIG, stellate ganglion block, LDN, semaglutide, daratumumab (66 patients), bezisterim (200 patients), Anktiva, tirzepatide (1,000 patients), and more. The adult treatment infrastructure, tracked across Post #8, Post #24, and Post #39, has produced a dozen failed trials and learned from each failure. The pediatric infrastructure has not failed yet — because it barely exists.
What I Don’t Know
Biomarkers. Camara and Buonsenso found 41 markers with significant variation across 9 pediatric studies. None are diagnostic. Whether pediatric Long COVID biomarkers differ from adult ones is an open question that determines whether adult-derived trial designs can be transferred to children.
Trajectory. Reeder and Ford show harm at a single time point. Whether these children recover, plateau, or worsen is unknown. Adult trajectory data (Post #9, Post #22) shows divergent paths. No equivalent pediatric longitudinal study exists.
Accommodations. Both Reeder and Ford recommend school accommodations — extended time, rest periods, IEP access. No study has tested whether accommodations change outcomes. The recommendation is common sense built on evidence absence.
Treatment window. If metabolic rewiring precedes symptoms (Post #18), how early must intervention begin? No pediatric trial is designed to answer this.