A Comparison of the 3 COVID-19 Vaccines

The three main COVID-19 vaccine platforms that saw widespread global use differ fundamentally in how they teach your immune system to recognize the virus. mRNA vaccines (Pfizer-BioNTech and Moderna) deliver genetic instructions wrapped in tiny fat particles, adenoviral vector vaccines (AstraZeneca and Johnson & Johnson) use a harmless modified virus as a delivery vehicle, and protein subunit vaccines (Novavax) present a lab-made copy of the spike protein directly. All three reduce severe illness, hospitalization, and death, but they diverge in efficacy numbers, side-effect profiles, how quickly protection fades, and which rare complications they carry.

How Each Platform Works

The mRNA vaccines from Pfizer-BioNTech (BNT162b2) and Moderna (mRNA-1273) package synthetic messenger RNA inside lipid nanoparticles, essentially tiny spheres made of specialized fats. The mRNA and a key fat called an ionizable cationic lipid sit in the core of the particle, while helper lipids form the outer shell.¹ The two vaccines use different ionizable lipids to get the job done: Moderna uses SM-102, while Pfizer-BioNTech uses ALC-0315, both of which outperform older lipid designs at delivering mRNA into cells.² Once inside your cells, the mRNA instructs them to produce the SARS-CoV-2 spike protein, which your immune system then learns to attack.

Adenoviral vector vaccines take a different route. They use a replication-deficient adenovirus, a common cold virus that has been gutted so it cannot reproduce, to ferry the spike protein’s genetic code into your cells.³ AstraZeneca’s vaccine uses a chimpanzee adenovirus (ChAdOx1), while Johnson & Johnson’s uses a human adenovirus (Ad26). Your cells read the instructions and produce the spike protein, prompting an immune response. The advantage of this platform is its familiarity: adenoviral vectors have been used in vaccines for decades and are stable at normal refrigerator temperatures, making them easier to distribute in resource-limited settings.

Novavax’s NVX-CoV2373 skips the genetic-instruction step entirely. Instead, it delivers pre-made spike proteins grown in insect cells, paired with an adjuvant called Matrix-M. This adjuvant consists of purified saponins from the bark of the Quillaja saponaria tree, combined with cholesterol and phospholipids to form tiny cage-like nanoparticles. Matrix-M activates immune cells at the injection site and in nearby lymph nodes, improving both the strength and breadth of the antibody response.⁴ This protein subunit approach is the most traditional of the three and is the same basic technology behind hepatitis B and whooping cough vaccines.

How They Performed in Clinical Trials

Head-to-head clinical trials comparing all three platforms in the same population were never conducted, so researchers relied on network meta-analyses, statistical methods that indirectly compare vaccines across separate trials. In one such analysis of phase III trials, no statistically significant differences emerged among the vaccine types for preventing symptomatic infection, though there was a trend favoring mRNA vaccines. Among individual vaccines, Pfizer-BioNTech ranked highest, followed by Moderna, with Novavax close behind.⁵ A separate network meta-analysis found a similar pattern: Pfizer-BioNTech and Moderna had the highest probability of efficacy against symptomatic COVID-19, while the adenoviral vector vaccines from AstraZeneca and J&J ranked lowest among those analyzed.⁶

These rankings came with a crucial caveat: each vaccine’s phase III trial was conducted in a different country, at a different point in the pandemic, against different circulating variants, and with different definitions of what counted as a “case.” The mRNA vaccines were tested earlier, when the original strain dominated. AstraZeneca and J&J trials ran later, when more transmissible variants were circulating. Raw efficacy percentages compared across separate trials are a rough guide, not a precise ranking.

Real-World Effectiveness

Once billions of doses were administered, real-world data painted a more detailed picture. A large prospective cohort study following about 18,000 participants over 18 months found that mRNA vaccines had the highest effectiveness: roughly 71% against infection, 84% against severe disease, and 88% against death, outperforming inactivated and viral vector vaccines.⁷ A separate observational study comparing four vaccine types found Pfizer-BioNTech’s effectiveness against asymptomatic infection was about 92%, compared to roughly 84% for AstraZeneca.⁸

The gap between platforms narrowed considerably when you looked at the outcome that matters most: keeping people alive and out of the hospital. All three platforms showed strong protection against severe outcomes. The differences mattered most for preventing mild or asymptomatic infection, where mRNA vaccines maintained a measurable edge.

Everyday Side Effects

The vaccines differed in when and how they made people feel lousy. For the mRNA vaccines, side effects were typically worse after the second dose than the first: more fatigue, headache, muscle aches, and fever. For the adenoviral vector vaccines, the pattern reversed. The J&J single-dose vaccine and AstraZeneca’s first dose tended to produce stronger reactions than subsequent doses. A real-world safety monitoring study found that the incidence of seven common adverse events after the J&J viral vector vaccine was about twice that of the two mRNA vaccines.⁹

A study comparing adverse event reports across all four major vaccine brands found that the percentage of people reporting any side effect ranged from about 53% for Pfizer to roughly 94% for Moderna. Serious adverse events were rare across the board, though AstraZeneca had the highest rate of serious reports at about 0.23%. Pfizer was associated with the lowest overall rate of adverse events.¹⁰ The Novavax protein subunit vaccine generally produced milder reactions than either the mRNA or vector vaccines, which was one reason it appealed to people hesitant about newer technologies.

Rare but Serious Complications

Each platform carried its own rare safety signal, and these platform-specific risks became one of the starkest differences among the three.

Myocarditis, or inflammation of the heart muscle, was linked primarily to the mRNA vaccines. It occurred most often in young men after the second dose, with symptoms typically appearing two to four days later. Most cases involved men with a median age of 20 to 30, with hospital stays averaging two to four days.¹¹ When studies applied four key filters simultaneously, looking specifically at young men after a second mRNA dose, the incidence ranged from about 8 to 39 cases per 100,000. The highest estimates, around 37 to 39 per 100,000, were in boys aged 12 to 17 after a second Pfizer dose. Moderna’s rate in men aged 18 to 24 after the second dose was about 30 per 100,000.¹² The vast majority of these cases resolved fully with standard care, but the signal was real enough to influence dosing recommendations for younger age groups in several countries.

The adenoviral vector vaccines carried a different rare risk: vaccine-induced immune thrombocytopenia and thrombosis, known as VITT. This condition involved unusual blood clots combined with dangerously low platelet counts, typically appearing 5 to 24 days after the first dose. Patients presented with severe thrombocytopenia, sharply elevated D-dimer levels, and antibodies against platelet factor 4, a pattern that signaled abnormal platelet activation.¹³ Both AstraZeneca and J&J vaccines were associated with VITT.¹⁴ While extremely rare, VITT was often severe and sometimes fatal, which ultimately led several countries to restrict or discontinue the adenoviral vector vaccines.

Novavax’s protein subunit vaccine had the mildest rare-event profile of the three platforms, though some cases of myocarditis and pericarditis were reported. Its overall safety data was less extensive simply because far fewer doses were administered worldwide compared to the mRNA and vector vaccines.

How Quickly Protection Fades

All COVID-19 vaccines see their protection drop over time, but the speed and pattern of that decline differed by platform. After primary immunization, neutralizing antibody levels waned at different rates depending on the vaccine type. In one analysis, titers from heterologous (mixed) primary series dropped about 3.4-fold every 90 days, while viral vector vaccine titers dropped about 1.4-fold over the same period.¹⁵ That might sound like viral vector vaccines held up better, but they started from a lower peak, so a slower decline from a lower starting point did not necessarily mean better protection overall.

Booster doses significantly slowed the decline. Modeling of antibody decay after mRNA vaccination suggested that each successive booster reduced the rate at which antibodies fell, maintaining protective levels for longer. Younger adults generated higher peak antibody levels after their initial two doses, but by the third and fourth doses, that age gap largely disappeared.¹⁶

An important finding is that protection against severe disease and death held up far better than protection against infection. Even when neutralizing antibody levels dropped to the lower limit of detection, vaccine effectiveness against severe and fatal outcomes stayed above 75%.¹⁵ This pattern held across vaccine types and suggests that other immune defenses, particularly T cells, continue doing protective work even after antibody levels crater.

T-Cell Responses Across Platforms

Antibody levels get the most attention, but T cells are a critical second layer of defense, especially for preventing severe disease. A comparison of immune responses in Jordan found that Pfizer recipients had significantly higher T-helper cell responses (around 91%) compared to recipients of non-mRNA vaccines like Sinopharm, Sputnik V, and AstraZeneca (around 84%). However, the pattern flipped for cytotoxic T cells, the type that directly kills infected cells: non-mRNA vaccine recipients showed higher levels (about 74% versus 58% for Pfizer). Total T-cell counts did not differ significantly between groups.¹⁷

This finding hints at something nuanced: different platforms may train slightly different arms of the cellular immune system. The adenoviral vector vaccines appeared to generate relatively more robust killer T-cell responses, which may help explain why they offered solid protection against severe disease even when their antibody-driven protection against mild infection was lower. No single platform clearly “won” on cellular immunity; they each had strengths.

Mix-and-Match Boosting

When booster campaigns began, many countries allowed or encouraged people to receive a different vaccine type for their booster than they received for their primary series. This heterologous boosting strategy proved to be more than a logistical convenience. Studies found that mixing platforms generally produced stronger immune responses than staying with the same vaccine. In an NIH-sponsored trial, homologous boosters raised neutralizing antibody titers by a factor of 4 to 20, while heterologous boosters raised them by a factor of 6 to 73. Additionally, boosting mRNA-primed individuals with the J&J adenoviral vector vaccine substantially increased spike-specific CD8+ T cells, the killer cells that had been lower in the mRNA-only group.¹⁸

The mix-and-match approach also addressed practical problems. Countries that had relied heavily on AstraZeneca for primary doses, then pulled it due to VITT concerns, needed an alternative booster. Switching those recipients to an mRNA booster gave them a strong immune bump and improved cellular and humoral responses without significantly increasing adverse reactions.¹⁹

The Omicron Problem and What Boosters Fixed

Omicron exposed the limits of two-dose immunity more starkly than any previous variant. A study of paired blood samples from healthcare workers found that after two mRNA doses, there was little to no neutralizing capability against Omicron, and neutralizing capacity against all tested variants was essentially gone by eight months. A booster dose changed the picture dramatically, producing a greater than 15-fold increase in neutralization against Omicron in paired pre-and-post-boost samples.²⁰ The booster did not just restore protection; it broadened it, improving both the quantity and quality of antibodies against a range of variants.

Immunocompromised Individuals

For people with weakened immune systems, the differences between vaccine platforms mattered less than the simple fact that all vaccines worked less well. A systematic review found that after one dose, organ transplant recipients were about 16 times less likely to develop protective antibodies compared to healthy controls. After two doses, only about a third of transplant recipients achieved seroconversion. People with blood cancers, autoimmune conditions, and solid tumors fell in between, with seroconversion rates ranging from roughly 62% to 89% after two doses. People with HIV responded essentially the same as healthy controls.²¹

Real-world data from 10 U.S. states confirmed this gap. Among immunocompromised patients who received three mRNA doses, effectiveness against hospitalization was about 81%, compared to 96% in people with healthy immune systems. Transplant recipients had the lowest effectiveness of any immunocompromised subgroup, and similar patterns held for viral vector vaccines.²² A third dose of mRNA vaccine helped convert some non-responders with cancers and autoimmune conditions, though results for transplant recipients remained inconsistent.

Children and Adolescents

Pediatric data accumulated more slowly, but the picture eventually clarified. A meta-analysis of mRNA vaccination in children aged 5 to 11 found that two doses cut the risk of symptomatic infection roughly in half and reduced hospitalizations by about two-thirds. The most striking finding was the reduction in multisystem inflammatory syndrome in children, a dangerous post-infection complication, which dropped by about 95%. Myocarditis in this age group was estimated at about 1.8 per million after the second dose.²³

For adolescents, mRNA vaccines showed efficacy between roughly 91% and 100% against symptomatic infection in randomized trials, with adverse events overwhelmingly mild to moderate: injection-site pain, headache, fatigue, and fever.²⁴ Novavax’s protein subunit vaccine was tested in adolescents as well, showing about 80% efficacy against symptomatic COVID-19, with adolescents generating higher antibody levels than young adults. Side effects were largely mild and transient, and no adverse events led to study discontinuation.²⁵

Vaccination During Pregnancy

A key question for pregnant individuals was whether vaccine-generated antibodies would cross the placenta and protect newborns. Research confirmed that mRNA vaccination during pregnancy produces antibodies that transfer to the fetus. In one evaluation, anti-SARS-CoV-2 IgG was detectable in about 95% of maternal blood samples at delivery and roughly 88% of cord blood samples. Importantly, neither the spike protein nor vaccine mRNA was detectable in cord blood or placental tissue, indicating the vaccine’s genetic payload does not reach the fetus at readily detectable levels, only the protective antibodies do.²⁶ The duration and strength of that passive protection in newborns remain active areas of research.

Hybrid Immunity

By the time most of the world had been both vaccinated and infected at least once, researchers found that the combination, called hybrid immunity, outperformed either infection or vaccination alone. In a study tracking infections during the Alpha, Beta, and Delta waves, hybrid immunity provided consistently higher protection against both infection and severe COVID-19 than either natural infection or vaccination by itself, regardless of which mRNA vaccine was used.²⁷

The antibody numbers tell the story vividly. In people with hybrid immunity, neutralizing titers were 8 to 23 times higher than in those vaccinated alone, depending on the variant tested. The biggest jumps were against variants like Beta and Omicron BA.1, the very strains most likely to dodge vaccine-only immunity. The quality of those antibodies also improved: they neutralized a broader range of variants more efficiently.²⁸ During the Omicron wave, hybrid immunity provided stronger protection against subsequent infection and similar protection against hospitalization and death compared to two vaccine doses alone.²⁹

Why None of Them Stopped Transmission

All three vaccine platforms were designed and tested primarily for their ability to prevent disease, not for their ability to prevent the virus from entering or replicating in your nose and throat. Injected vaccines are good at generating systemic immunity, meaning antibodies in your blood and immune memory cells throughout your body. They are not as effective at generating mucosal immunity, the secretory IgA antibodies and tissue-resident memory cells in the respiratory tract that would stop the virus right at the point of entry.³⁰

This is why breakthrough infections occurred even in fully vaccinated people, and why vaccinated individuals could still spread the virus. Injected vaccines do not robustly or consistently stimulate the mucosal immune response considered essential for blocking infection and onward transmission. Mucosal vaccines, delivered as nasal sprays or inhaled drops, have shown promise in animal models by stimulating secretory IgA and respiratory-tract memory cells, significantly reducing viral replication and transmission.³¹ Several mucosal COVID-19 vaccines have been approved in countries like China and India, though none have been widely adopted in the West. This remains one of the most important frontiers in COVID-19 vaccine development.

Manufacturing and Why Supply Varied So Much

The three platforms differ enormously in how easy they are to manufacture at scale, which had direct consequences for global access. mRNA vaccines are produced in a cell-free process, essentially mixing synthetic RNA with lipid components in a controlled reaction. This is fast and highly concentrated. Modeling of vaccine production found the mRNA platform to be roughly 200 to 300 times more productive per unit of bioreactor volume per year than the adenoviral vector platform, which relies on growing live cells in bioreactors and carries more biological variability and a higher failure rate.³²

However, the mRNA vaccines required ultra-cold storage, at least initially. Pfizer-BioNTech’s vaccine needed minus 70°C for long-term storage in its original formulation, while Moderna’s was stable at normal freezer temperatures. Adenoviral vector vaccines and Novavax could be stored in a standard refrigerator. For much of the world, particularly sub-Saharan Africa, South Asia, and rural communities without cold-chain infrastructure, AstraZeneca and later Novavax were the only practical options. The manufacturing speed advantage of mRNA meant little if the doses could not be stored and transported to the people who needed them. This logistical gap shaped the global vaccine rollout at least as much as any efficacy comparison did.

1
PubMed Central. mRNA-lipid nanoparticle COVID-19 vaccines: Structure and stability
2
Nature Reviews Materials. Lipid nanoparticles for mRNA delivery
3
PubMed Central. Review of COVID-19 viral vector-based vaccines and COVID-19 variants
4
PubMed Central. The Matrix-M™ adjuvant: A critical component of vaccines for the 21st century
5
PubMed Central. Comparative efficacy and safety of COVID-19 vaccines in phase III trials: a network meta-analysis
6
Scientific Reports. Comparing the clinical efficacy of COVID-19 vaccines: a systematic review and network meta-analysis
7
Journal of Medical & Health Sciences Review. COMPARATIVE EFFECTIVENESS OF mRNA, INACTIVATED, AND VIRAL-VECTOR VACCINES IN PREVENTING INFECTIOUS DISEASES
8
PubMed Central. Real-World Effectiveness of Four Types of COVID-19 Vaccines
9
PubMed Central. The Safety of mRNA-1273, BNT162b2 and JNJ-78436735 COVID-19 Vaccines: Safety Monitoring for Adverse Events Using Real-World Data
10
PubMed Central. Description of Frequencies of Reported Adverse Events Following Immunization Among Four Different COVID-19 Vaccine Brands
11
BMJ. Incidence, risk factors, natural history, and hypothesised mechanisms of myocarditis and pericarditis following covid-19 vaccination: living evidence syntheses and review
12
PubMed Central. COVID ‐19 vaccine induced myocarditis in young males: A systematic review
13
PubMed Central. To clot or not to clot? Ad is the question-Insights on mechanisms related to vaccine-induced thrombotic thrombocytopenia
14
PubMed Central. Comparison of vaccine-induced immune thrombocytopenia and thrombosis cases following two adenovirus-vectored COVID-19 vaccines
15
PubMed Central. Comparative duration of neutralizing responses and protections of COVID-19 vaccination and correlates of protection
16
PubMed. Modeling of anti-spike IgG and neutralizing antibody waning after anti-SARS-CoV-2 mRNA vaccination
17
PubMed Central. Comparison of T cells mediated immunity and side effects of mRNA vaccine and conventional COVID-19 vaccines administrated in Jordan
18
PubMed. Homologous and Heterologous Covid-19 Booster Vaccinations
19
PubMed Central. COVID-19 vaccines mix-and-match: The concept, the efficacy and the doubts
20
PubMed Central. Comparison of total and neutralizing SARS-CoV-2 spike antibodies against omicron and other variants in paired samples after two or three doses of mRNA vaccine
21
BMJ. Efficacy of covid-19 vaccines in immunocompromised patients: systematic review and meta-analysis
22
PubMed Central. Effectiveness of COVID-19 vaccines at preventing emergency department or urgent care encounters and hospitalizations among immunocompromised adults
23
PubMed Central. Assessment of Efficacy and Safety of mRNA COVID-19 Vaccines in Children Aged 5 to 11 Years A Systematic Review and Meta-analysis
24
PubMed Central. Safety and efficacy of COVID-19 vaccines in children and adolescents: A systematic review of randomized controlled trials
25
JAMA Network Open. Safety, Immunogenicity, and Efficacy of the NVX-CoV2373 COVID-19 Vaccine in Adolescents: A Randomized Clinical Trial
26
PubMed Central. Transplacental Transfer of Maternal Antibody against SARS-CoV-2 and Its Influencing Factors: A Review
27
PubMed Central. Effects of previous infection, vaccination, and hybrid immunity against symptomatic Alpha, Beta, and Delta SARS-CoV-2 infections: an observational study
28
PubMed Central. An extended interval between vaccination and infection enhances hybrid immunity against SARS-CoV-2 variants
29
PubMed Central. Comparing hybrid and regular COVID-19 vaccine-induced immunity against the Omicron epidemic
30
PubMed Central. SARS-CoV-2 Vaccines: The Advantage of Mucosal Vaccine Delivery and Local Immunity
31
Vaccine. A review of currently licensed mucosal COVID-19 vaccines
32
PubMed Central. Pandemic-response adenoviral vector and RNA vaccine manufacturing