A tuberculosis vaccine rarely used in the United States reached the central nervous system of older adults in a small study, providing researchers with direct human evidence for a mechanism behind one of the strangest patterns in dementia research.
Marc Weinberg of Massachusetts General Hospital and colleagues ran two related one-year, open-label trials in 23 adults aged 55 and older, 11 with biomarker evidence of Alzheimer's-related pathology and 12 without. Using lumbar puncture to sample cerebrospinal fluid, they showed that the Bacille Calmette-Guerin vaccine reaches the brain and alters immune activity there. The work was published in July in Communications Medicine.
Weinberg drew the boundary himself. What we didn't do is prove that BCG prevents Alzheimer's, he told CIDRAP News, adding that the team did not show it treats the disease either.
That distinction is the whole story for households. Three unrelated vaccines now show a similar statistical association with lower dementia rates, and the research has moved a step closer to explaining why without moving any closer to a recommendation.
Spinal Fluid Evidence from 23 Older Adults
BCG prevents tuberculosis and is not routinely given in the United States outside a few uses, including treatment for some bladder cancers. Scientists at Mass General Brigham studying those off-label applications noticed that people receiving BCG appeared to develop Alzheimer's and other dementias less often.
The new work did not test whether that observation holds up. It tested something narrower and more mechanical, namely whether the vaccine can reach the organ where the effect would have to occur.
Participants received two intradermal BCG doses a month apart, then gave blood and spinal fluid samples over the following year. The team found durable immune reprogramming in cerebrospinal fluid cells. Shifts in amyloid beta levels, lower in spinal fluid and higher in blood, appeared only in participants without Alzheimer's-related changes at the outset, not in those who already had them.
Twenty-three participants is a very small sample, and the design was open-label with no placebo group. What it establishes is biological plausibility, not benefit.
Higher-Dose Flu Shots and a Sex Difference
Separate work from Paul Schulz and colleagues at UTHealth Houston has been building a parallel case around influenza vaccination for about seven years, after a biomedical informatics team flagged an unexpected signal in their databases.
Their first analysis, published in the Journal of Alzheimer's Disease, compared adults aged 65 and older who received a flu vaccine with those who did not. Three years of vaccination was associated with roughly a 20 percent lower risk of Alzheimer's, and six years with about a 40 percent lower risk, persisting for at least eight years.
The team then asked a sharper question. Comparing people who got a standard-dose flu vaccine with those who got a high-dose version removes much of the healthy vaccinee problem, since both groups are people who show up for shots. In that comparison published in Neurology, first-authored by Avram Samuel Bukhbinder, the high dose outperformed the standard dose by about 15 percent.
The researchers also reported that the apparent anti-dementia effect lasted roughly two years, far longer than the four to six months of influenza protection, and that women appeared to benefit more from the high dose than men. Women develop Alzheimer's about twice as often as men, which makes that finding worth following.
Three Unrelated Vaccines, One Recurring Signal
MedicalDaily previously reported on the shingles vaccine study linking it to lower dementia risk, which found a 24 percent lower relative risk across more than 500,000 older adults over four years. That work was led by Daniel Harris of the University of Delaware using nursing home records and Medicare claims, and was published in Annals of Internal Medicine. It was funded by GlaxoSmithKline, which makes the shingles vaccine, a disclosure readers should weigh alongside the finding.
What has changed since that report is the shape of the evidence rather than its strength. A shingles vaccine, an influenza vaccine and a tuberculosis vaccine target completely different pathogens, yet all three now carry a similar association.
Researchers have offered mechanisms that differ by vaccine. For shingles, the varicella zoster virus lives in nerve tissue, can replicate in cerebral arteries, and raises stroke risk for up to a year after an episode, and stroke is an established cause of vascular dementia. For influenza, the leading hypotheses involve the immune system either clearing amyloid plaque more efficiently or reacting to it less aggressively.
Glen Finney, who directs the Memory and Cognition Program at Geisinger Health, has framed the common thread as immune retraining and has raised the question of whether the effect belongs to vaccines broadly or to a specific subset. That remains unanswered.
Association Is Not Prevention
Every study described here is observational except the small BCG trials, and none of them demonstrates that a vaccine prevents dementia. Researchers involved in this work say randomized trials comparing vaccines against placebo are the missing piece, as CIDRAP News reported.
Those trials would be slow and difficult. Because a person can receive a recombinant shingles vaccine from age 50 and cognitive changes may take many years to surface, follow-up would need to run for a long time, with all the attrition that implies.
The healthy vaccinee problem also has not disappeared. People who keep up with vaccines tend to be healthier, better connected to care, and more likely to have other protective habits, and statistical adjustment only goes so far. The high-dose versus standard-dose comparison addresses this more convincingly than most designs, but it does not eliminate it.
For a household deciding on fall shots, the practical guidance has not changed. Influenza and shingles vaccination are recommended for older adults on the strength of what those vaccines were designed to do, which is prevent serious infection, hospitalization, and death. BCG is not available for this purpose, and no one should seek it out on the basis of a 23-person study.
A reasonable conversation with a clinician covers eligibility for the recombinant shingles vaccine, whether a high-dose or adjuvanted flu vaccine is appropriate at a given age, and what is covered under Medicare Part B or Part D. What is not reasonable is treating any of this as a dementia prevention plan, or delaying an appointment while waiting for trials that may take a decade to report.
Key Questions Answered
What is genuinely new here? A small trial showed the BCG tuberculosis vaccine reaches the human central nervous system and changes immune activity there, published in July. Separate flu vaccine research compared high-dose and standard-dose recipients.
Do vaccines prevent dementia? No study has shown that. The research describes statistical associations in observational data, and the researchers themselves say randomized trials are needed.
Should someone ask for a BCG vaccine? No. BCG is not used in the United States for this purpose, and the study involved 23 people and tested biology rather than benefit.
Why do three different vaccines show a similar pattern? The leading idea is that vaccination retrains the immune system in ways that may reduce brain inflammation, but the mechanism differs by vaccine, and none is confirmed.
Does this change flu or shingles vaccine advice? No. Both remain recommended for older adults based on infection prevention, which is a separate and far better established benefit.
What is the healthy vaccinee problem? People who get vaccinated tend to be healthier overall, which can make vaccines look more protective than they are. Comparing vaccine doses rather than vaccinated and unvaccinated groups reduces it.
Did women benefit more? In the high-dose flu comparison, researchers reported a larger apparent benefit in women. That finding comes from observational data and has not been confirmed in a trial.
