How long do immune protection and antibody levels last after mRNA?

The short version: antibody levels after mRNA vaccination drop sharply within months, but memory B cells persist longer. How long protection lasts depends on what you measure and against which variant.

The evidence here is mostly about antibody decay, not about real-world protection against severe disease over years. And most of it was funded or conducted by the manufacturers or their academic partners. The key studies track antibody levels, not hard outcomes like hospitalization or death. Antibodies are a surrogate endpoint. A rise in titers does not prove protection, and a drop does not prove vulnerability.

Here is what the numbers show. After two doses of BNT162b2, spike IgG peaked around 3,800 BAU/mL and fell below a 50% protection threshold in about 7 days [3]. A booster pushed the peak to about 8,500 BAU/mL and extended time above the 50% threshold to about 99 days [3]. But those thresholds were calibrated against the original strain and BA.1 Omicron. Against later variants, the real protection is shorter. A 2025 study in Malaysia found that after two doses of Comirnaty, neutralizing inhibition was still about 91% at 24 weeks, but that was against the original strain, not current variants [2].

The durability is very different if you have had a prior infection. Hybrid immunity (infection plus vaccination) extended the antibody half-life to about 448 days under a simple model, and to 810 days under a two-phase model [4]. That is roughly double the half-life of vaccine-only immunity [4]. The booster-only group in another study had an antibody half-life of about 117 days, while the hybrid group had about 242 days [3].

Memory B cells, which can produce new antibodies if the virus returns, appear to persist longer. A 12-month study found that mRNA vaccines, especially mRNA-1273, induced more durable memory B cell responses than natural infection [1]. But memory B cells are not the same as circulating antibodies. If the virus mutates enough, the memory cells may not recognize it.

The evidence does not include long-term follow-up beyond 12 months for most studies, and no study in this retrieval compared vaccinated to unvaccinated over years for all-cause mortality. The waning data is based on antibody levels, not on infection or death rates. And the studies that do exist were mostly funded by the manufacturers or conducted by their affiliates.

My call: antibody levels wane substantially within 3-6 months after mRNA vaccination, and protection against infection drops with them. Memory B cells persist longer but may not cover new variants. The evidence for long-term protection against severe disease is thin, and the most durable immunity comes from hybrid immunity, not from vaccination alone. Confidence: moderate for antibody decay data, low for long-term clinical protection.