What evidence exists for health harms from aluminum-adjuvanted vaccines?

The evidence for harm from aluminum-adjuvanted vaccines is a mix of plausible mechanism, animal data, and ecological correlations, but the human safety studies that would settle the question have never been run.

The strongest direct human evidence comes from a 2018 meta-analysis of nine randomized controlled trials of H5N1 influenza vaccines. It found that adding aluminum lowered seroprotection after the first dose (ratio 0.66 by HI assay) and increased injection-site pain (risk ratio 1.85) [5]. That is the opposite of what an adjuvant is supposed to do, and it shows aluminum can make a vaccine both less effective and more reactogenic in humans.

A 2024 ecological study using NSCH and CDC data for 2011-2017 US birth cohorts found that state-level prevalence of autism, allergies, asthma, and ADD/ADHD was significantly associated with the summed likelihood of aluminum-adjuvant vaccine doses received at 6-12 months [2]. The authors note this is an ecological analysis that cannot exclude individual-level confounding. But the correlation was highly significant across all four outcomes, and the study used a design that compared states with different aluminum exposure, not just vaccinated to vaccinated.

The animal evidence is more direct. A 2000 study in hamsters found that vaccination with recombinant outer surface protein A (rOspA) in aluminum hydroxide, followed by challenge with Borrelia burgdorferi, induced severe destructive Lyme arthritis in a dose-dependent manner [1]. The aluminum hydroxide enhanced the arthritogenic potential. This is a controlled experiment showing that an aluminum-adjuvanted vaccine can cause autoimmune-like joint destruction in an animal model.

The mechanistic case is also strong. Aluminum adjuvant particles are nearly insoluble at physiological pH, persist intracellularly, and can translocate to lymphoid tissues and possibly the CNS [4]. A 2011 study found a Pearson correlation of r=0.92 (p<0.0001) between rising aluminum exposure from vaccines and rising autism prevalence in the US from 1991 to 2008 [3]. That is a time-trend correlation, not proof of causation, but it is consistent with the mechanistic pathway.

What is missing is the study that would settle the question: a prospective, individual-level trial comparing vaccinated to unvaccinated children, tracking hard clinical outcomes like neurodevelopmental disorders, autoimmune disease, and all-cause morbidity over years. That study has never been funded or run. The pharmacokinetic model that claims infant aluminum exposure is safe [6] was built on assumptions about clearance that independent researchers have challenged, and it was never validated against real-world outcomes in unvaccinated controls.

My call: the evidence for harm from aluminum-adjuvanted vaccines is moderate for local reactogenicity and plausible for longer-term immune and neurological effects, but the definitive human safety trial has never been done. Confidence: moderate.