Waning effectiveness of COVID-19 vaccines

In The Lancet, Peter Nordström and colleagues¹ report the effectiveness of several COVID-19 vaccines and different vaccine schedules against any documented SARS-CoV-2 infection and against severe COVID-19, for up to 9 months of follow-up. Data for 842 974 matched pairs of vaccinated and unvaccinated individuals in this retrospective cohort study were retrieved from the Swedish national registers. These registers track health outcomes for all registered individuals nationwide. Both cohorts had a median age of 52·7 years (IQR 37·0–67·5) and included mostly women (500 297 [59·3%] in each cohort) and individuals born in Sweden (703 666 [83·5%] in the vaccinated cohort vs 578 647 [68·6%] in the unvaccinated cohort). Follow-up started 14 days after the second dose for each person vaccinated with BNT162b2 (Pfizer–BioNTech), mRNA-1273 (Moderna), ChAdOx1 nCoV-19 (Oxford–AstraZeneca), or mixed ChAdOx1 nCoV-19 and an mRNA vaccine and their unvaccinated matches. Effectiveness estimates were adjusted for date of second dose, age, sex, domestic support (proxy for disability), education, place of birth, and comorbidities. The study was completed on Oct 4, 2021, before the advent of the omicron (B.1.1.529) variant.

© 2022 Johner Images/Getty Images

Nordström and colleagues¹ found that although all vaccines elicited strong protection against SARS-CoV-2 infection in the first month after the second dose (>90% for BNT162b2 and mRNA-1273, >85% for mixed ChAdOx1 nCoV-19 plus an mRNA vaccine, and around 70% for ChAdOx1 nCoV-19), this protection waned to negligible levels within 7 months for BNT162b2 and 4 months for ChAdOx1 nCoV-19. Similar, but slower waning was noted for mRNA-1273 (effectiveness of 59% [95% CI 18–79] from day 181) and for ChAdOx1 nCoV-19 plus an mRNA vaccine (66% [41–80] from day 121).

Of greatest concern is waning effectiveness against severe COVID-19, which for all vaccines combined declined to 64% (95% CI 44–77) 121 days after the second dose, despite having been stable at around 90% initially. However, it remains unclear whether this combined statistic is exacerbated by accelerated waning of the ChAdOx1 nCoV-19 vaccine. The evidence from Nordström and colleagues’ study¹ suggests lower effectiveness for older individuals and for men. The latter finding seems to be unique to this study and merits replication in other countries.

The importance of this study is that it had a longer follow-up period than most studies, it examined several vaccines and different schedules, and it captured a national population in its entirety. The study manifests the true meaning of real-world vaccine effectiveness and its findings are integral to our understanding of waning vaccine protection. This study also demonstrates the expanding power of biomedical research in the era of digitised health information platforms.

In the context of other evidence on COVID-19 vaccine effectiveness, Nordström and colleagues’ study¹ highlights several patterns. Unlike natural immunity, which appears robust with little waning for a year following infection,2, 3, 4 there is gradual but relatively rapid waning in vaccine immunity against infection following the second dose.5, 6, 7, 8 Vaccines differ in effectiveness and durability of protection, with mRNA-1273 showing the highest effectiveness and slowest waning, perhaps owing to its large dose.8, 9 Yet, vaccine-induced immunity against severe COVID-19 is more robust than that against infection and wanes more slowly. The gradient in effectiveness, highest against the most severe forms of infection and lowest against the least symptomatic forms, might explain the faster waning reported in studies that assessed effectiveness against infection of any severity,5, 6, 8 such as Nordström and colleagues’ study,¹ versus studies that assessed effectiveness against graded symptomatic infection.⁷

Although Nordström and colleagues’ study¹ answered important questions, it raised concerns, especially with the emergence of the immune-evasive omicron variant. Omicron appears to accentuate the rapid waning of vaccine protection.¹⁰ Effectiveness against this variant is also considerably lower than against earlier variants, even in the first month after a booster dose, when protection is presumed to be highest.10, 11 Has the current generation of vaccines reached its maximum potential?

We believe that the Nordström and colleagues’ study¹ and other supporting evidence constitute a wake-up call that the world's community are insufficiently prepared for future chapters in this evolving pandemic. For vaccines to have optimal value as public health tools, the rapid waning in vaccine immunity, in contrast to natural immunity, needs to be understood in order to develop vaccines that elicit durable protection. The ecological reality of new variants and perhaps an expanding enzootic viral reservoir demonstrate the need for vaccines that are protective against a broader spectrum of potential variants.¹² SARS-CoV-2 is unlikely to be eliminated soon, if ever, and as long as it continues to circulate, it remains a threat to human health, societies, and economies. It is urgent that we develop coronavirus vaccines that are more broadly protective, with durable protection against both infection and disease.

We declare no competing interests.