For decades, immunology textbooks described the immune system as a two-department operation. Innate immunity was the fast but forgetful security guard, while adaptive immunity was the detective who remembered faces, fingerprints, and suspicious spike proteins. Then researchers discovered that the security guard had been taking notes all along.
This phenomenon, known as trained immunity, describes long-lasting functional changes in innate immune cells after an infection, vaccination, or other biological stimulus. Instead of recognizing one specific pathogen, trained cells can respond more quickly or forcefully to a range of later threats. That possibility attracted enormous attention during the COVID-19 pandemic, particularly when scientists explored whether existing vaccines such as bacillus Calmette-Guérin, better known as BCG, could temporarily strengthen defenses against SARS-CoV-2.
The idea remains scientifically exciting, but it is not a magic force field. Clinical trials have produced mixed results, and researchers have learned that innate immune memory can be protective, ineffective, or even harmful depending on how it is programmed. In other words, the immune system remembersbut, like the rest of us, it does not always remember the right lesson.
What Is Trained Immunity?
Trained immunity is a form of biological memory found within the innate immune system. It can affect monocytes, macrophages, natural killer cells, neutrophils, and the stem and progenitor cells that continually produce new immune cells in bone marrow. After an initial stimulus, these cells may undergo metabolic and epigenetic changes that influence how they react to a later infection.
Epigenetic changes do not rewrite the DNA sequence. Instead, they alter how easily particular genes can be activated. Think of the genome as a massive cookbook. Training does not add new recipes; it leaves certain useful pages open on the counter so the immune system can start cooking sooner.
Metabolic reprogramming also matters. Trained cells may change how they generate and use energy, supporting faster production of inflammatory cytokines, antimicrobial molecules, and other defensive tools. These effects can persist after the original stimulus disappears, sometimes because the programming reaches long-lived bone marrow progenitor cells rather than remaining limited to short-lived circulating monocytes.
How It Differs From Adaptive Immune Memory
Adaptive immunity relies mainly on B cells and T cells. B cells can produce highly specific antibodies, while T cells identify infected cells or coordinate other parts of the immune response. This system can remember distinctive molecular targets for years.
Trained immunity is broader and less precise. It does not necessarily recognize the same pathogen that triggered it. A vaccine or infection may instead prepare innate cells to react more vigorously when they encounter an unrelated virus, bacterium, or fungus. This nonspecific protection is sometimes called a heterologous effect.
The two systems are not competitors. Innate immunity buys time, shapes inflammation, and helps activate adaptive defenses. Adaptive immunity then supplies targeted antibodies and specialized cellular responses. An effective immune reaction is less like a solo performance and more like a band whose drummer, guitarist, and vocalist have finally agreed to play the same song.
Why Trained Immunity Became a COVID-19 Research Priority
When SARS-CoV-2 emerged, population-specific vaccines did not yet exist. Researchers therefore investigated whether licensed vaccines with known nonspecific immune effects could provide temporary protection while COVID-19 vaccines were being developed.
BCG became the leading candidate. Originally designed to protect against tuberculosis, this live attenuated vaccine has been associated with reduced rates of certain unrelated infections in some populations. Controlled human experiments conducted before the pandemic also showed that BCG could epigenetically reprogram monocytes and improve resistance to an attenuated yellow fever vaccine virus, supporting a genuine biological mechanism rather than mere immunological wishful thinking.
The strategy was attractive for several reasons:
- BCG had already been administered to billions of people.
- Its safety profile and manufacturing process were well characterized.
- It could stimulate innate responses without requiring an exact genetic match to SARS-CoV-2.
- It might offer broad protection against multiple respiratory pathogens.
Early ecological studies appeared to show lower COVID-19 mortality in countries with universal BCG vaccination programs. Unfortunately, national comparisons were vulnerable to major confounding factors, including differences in population age, testing capacity, healthcare access, reporting practices, epidemic timing, and public-health restrictions. Country-level correlation is useful for generating a hypothesis, but it cannot prove that an old vaccination scar stopped a new coronavirus.
What Did the BCG Trials Actually Find?
Randomized controlled trials delivered a more cautious answer than the early headlines.
In the large international BRACE trial, healthcare workers were randomly assigned to receive BCG or a saline placebo. BCG vaccination did not reduce the risk of symptomatic COVID-19 during the primary analysis period. The study also lacked enough severe outcomes to determine conclusively whether BCG affected hospitalization or death.
Another controlled trial involving Dutch healthcare workers found no meaningful reduction in SARS-CoV-2 infection, illness duration, or COVID-19 severity, although researchers observed a possible difference in antibody production after infection.
Some smaller studies reported more encouraging signals. The ACTIVATE-2 trial involving older adults with underlying health conditions suggested that BCG might reduce the risk of possible COVID-19, and fewer vaccinated participants required hospitalization. However, the number of clinical events was small, and some endpoints were not confirmed by laboratory testing.
Taken together, the evidence does not establish BCG vaccination as a reliable method for preventing COVID-19. Differences among trials may reflect age, previous BCG exposure, vaccine strain, timing, local transmission, circulating variants, prior SARS-CoV-2 infection, and access to COVID-specific vaccines. The biological effect of trained immunity may be real while its practical protection against one particular disease remains inconsistent.
Why a Sound Mechanism Can Still Produce a Disappointing Trial
Biology rarely behaves like a light switch. A stronger innate response does not automatically prevent infection. SARS-CoV-2 can interfere with interferon signaling, alter monocytes and neutrophils, and provoke damaging inflammation. A broad immune boost may help in the earliest hours of exposure but offer little benefit once the virus has established itself.
Timing is equally important. Trained immunity develops over days or weeks and may fade or change over time. If exposure occurs too soon, training may be incomplete. If it occurs months later, the most useful effects may have weakened. Meanwhile, a highly transmissible variant can overwhelm modest nonspecific protection.
Finally, protection from infection is different from protection from severe disease. A trained immune response might fail to prevent a positive test while still changing viral load, inflammatory balance, symptom duration, or the probability of hospitalization. Trials need enough participants and sufficiently precise endpoints to detect these differences.
The Double-Edged Nature of Innate Immune Memory
Trained immunity is often described as an enhanced response, but “more inflammation” is not always a prize worth winning. Severe COVID-19 can involve both inadequate early antiviral activity and excessive later inflammation. An immune system that arrives late and then flips every table in the restaurant is not providing ideal customer service.
Research has shown that severe SARS-CoV-2 infection can leave persistent epigenetic changes in hematopoietic stem and progenitor cells. Their descendant monocytes may remain unusually responsive for months, with some signatures detected for approximately a year. Interleukin-6 signaling appears to be one pathway involved in creating this long-lasting inflammatory memory.
This finding creates an important distinction between protective trained immunity and maladaptive immune training. Properly directed training could improve early resistance to infection. Poorly regulated training could contribute to persistent inflammation, tissue injury, vascular problems, or post-acute symptoms.
Researchers have proposed that lingering innate immune reprogramming may be one piece of the long COVID puzzle, although long COVID is heterogeneous and cannot be explained by a single mechanism. Other possibilities include viral persistence, autoimmunity, endothelial dysfunction, reactivation of dormant viruses, and abnormal nervous-system signaling.
Can COVID-19 Itself Train the Immune System?
Yes, but the outcome may depend heavily on disease severity, tissue location, and the type of immune cell involved.
Animal research has found that recovery from SARS-CoV-2 infection can reprogram lung macrophages and produce stronger responses to a later influenza infection. In experimental models, antiviral genes remained more accessible, allowing cells to react more efficiently during the second viral challenge.
That result does not mean people should seek infection to “train” their immunity. Natural infection carries unpredictable risks, including hospitalization, vascular complications, organ damage, and long COVID. The scientific goal is to reproduce helpful immune programming without forcing patients to pay the admission price of a potentially dangerous disease.
Vaccination offers a safer way to develop targeted protection. Current COVID-19 vaccines primarily work by stimulating adaptive immunity against SARS-CoV-2, but research suggests they may also influence innate cells. A 2025 study reported persistent epigenetic changes in monocyte-derived macrophages after mRNA vaccination. The clinical meaning of those changes is still being investigated, so they should not be interpreted as proof of broad protection against unrelated infections.
How Trained Immunity Could Improve Future Vaccines
The most valuable legacy of this research may not be repurposing BCG as a stand-alone COVID intervention. It may be learning how to design vaccines that deliberately coordinate innate and adaptive memory.
Smarter Vaccine Adjuvants
Adjuvants are ingredients that improve a vaccine’s immune response. Future adjuvants might be selected not only for their ability to enhance antibodies and T cells but also for their capacity to produce controlled, beneficial innate training.
The ideal adjuvant would create a rapid antiviral state without causing prolonged inflammatory overactivation. Achieving that balance will require detailed knowledge of dose, delivery route, metabolic programming, age-related immune changes, and underlying health conditions.
Mucosal Immune Training
SARS-CoV-2 usually enters through the nose and upper respiratory tract, making mucosal immunity especially important. Scientists are studying nasal vaccines and other approaches intended to establish protection directly where exposure begins.
Research in animal models indicates that immune training can occur in tissue-resident lung macrophages, not only in circulating cells. BCG and respiratory viral exposures have both been shown experimentally to influence lung-resident innate memory. This raises the possibility of vaccines that prepare airway defenses to respond immediately, before viral replication gains momentum.
Combined Vaccine Platforms
Another strategy is to combine a SARS-CoV-2-specific antigen with a component that promotes trained immunity. In animal studies, BCG-adjuvanted COVID-19 vaccine platforms have produced strong immune responses and protection against experimental viral challenge. These results are promising but remain preclinical; success in hamsters or mice does not guarantee equal performance in humans.
Personalized Immune Training
Not everyone responds to immune training in the same way. Age, genetics, previous infections, metabolic health, medications, microbiome composition, and baseline inflammation may all affect the outcome.
Older adults may benefit from stronger early innate activation because immunologic responses often weaken with age. However, they may also have chronic inflammatory conditions that make excessive activation risky. Personalized approaches could eventually use biomarkers to identify who is likely to benefit, which stimulus should be used, and how long the effect might last.
What Trained Immunity Cannot Replace
Trained immunity should not be presented as an alternative to COVID-specific vaccination, antiviral treatment, cleaner indoor air, testing, or other established preventive measures.
As of the 2025–2026 U.S. season, CDC guidance recommends individual-based decision-making about updated COVID-19 vaccination, with particular emphasis on older adults and people at increased risk of severe disease. Updated vaccines are designed to improve protection against circulating strains, while trained-immunity interventions such as BCG remain investigational for COVID-19 prevention.
People should not seek additional BCG vaccination specifically for COVID-19 unless they are participating in an appropriately supervised clinical study or have another approved medical indication. BCG is a live vaccine and is not suitable for everyone, particularly certain immunocompromised individuals.
Practical Experiences and Lessons From Trained-Immunity Research
The pandemic provided researchers with a crash course in separating an exciting immunological concept from a dependable public-health intervention. One major lesson came from the early country comparisons involving BCG. The patterns looked persuasive on maps, especially when nations with long-standing BCG programs appeared to have fewer COVID-19 deaths. Yet those maps could not adequately account for epidemic timing, population age, healthcare capacity, testing, reporting differences, prior exposure, or public-health policies. The experience reinforced a basic research rule: colorful correlations are excellent at starting conversations and terrible at ending them.
Clinical-trial teams then faced a different challenge. COVID-19 conditions changed rapidly while the studies were underway. Participants gained access to COVID-specific vaccines, new variants emerged, prior infections accumulated, and community transmission rose or fell. A trial designed during one pandemic phase could be answering a slightly different question by the time the last participant completed follow-up. Researchers learned that trained-immunity trials need flexible designs, careful documentation of prior exposure, laboratory-confirmed endpoints, and enough participants to evaluate severe outcomes rather than relying only on reported symptoms.
Healthcare workers participating in BCG studies also illustrated the difference between biological activity and clinically useful protection. Laboratory samples could reveal changes in cytokine production, antibodies, or monocyte behavior even when infection rates were similar between study groups. That experience showed why immune measurements and real-world outcomes must be evaluated together. A biomarker can confirm that something happened inside the body without proving that the change prevented illness.
Another practical lesson involved timing. Immune training is not instantaneous, and its duration is not unlimited. A vaccine given today may create one pattern of innate activity after two weeks and another after six months. Researchers therefore began paying closer attention to the “when,” not merely the “what.” Future studies will likely require repeated immune profiling to identify when protection is strongest, when it fades, and when potentially harmful inflammation appears.
The pandemic also changed how scientists think about recovery. Severe COVID-19 was initially treated mainly as an acute event: survive the infection, clear the virus, and move on. Persistent epigenetic changes in bone marrow progenitors suggested that the immune system might continue operating under altered instructions long after a negative test. This experience encouraged closer collaboration among infectious-disease specialists, immunologists, cardiologists, neurologists, and long COVID researchers.
Perhaps the most useful experience was learning to respect immune balance. Early in the pandemic, “boosting immunity” sounded universally desirable. Later evidence made that phrase look rather clumsy. Patients need effective early antiviral defenses, but they also need mechanisms that shut inflammation down once the threat is controlled. Future trained-immunity therapies will have to be closer to precision coaching than indiscriminate cheerleading.
Conclusion: A Promising Tool, Not a COVID Shortcut
Trained immunity has changed the traditional understanding of immune memory. Innate cells and their bone marrow precursors can retain functional impressions of previous encounters, allowing later responses to become faster, stronger, or simply different.
In the fight against COVID-19, that discovery generated credible hypotheses, valuable clinical trials, and new ideas for vaccine development. BCG demonstrated the biological ability to train innate defenses, but randomized trials did not establish it as a consistent COVID-19 preventive tool. Meanwhile, studies of SARS-CoV-2 infection revealed that immune training has a darker side: severe disease can leave persistent inflammatory programs that may affect health after the acute infection ends.
The next chapter will focus on control. Scientists need to determine which signals generate useful antiviral readiness, which create chronic inflammation, where trained cells should be located, how long their memory should last, and which patients are most likely to benefit. The goal is not simply to make immunity louder. It is to make immunity earlier, smarter, and better behaved.
Note: This article is intended for educational purposes and does not replace medical advice. BCG vaccination and other trained-immunity interventions are not established substitutes for authorized COVID-19 vaccines, prescribed antiviral treatment, or individualized guidance from a qualified healthcare professional.
