Have you ever heard the old story about a person searching for lost keys under a streetlamp? When asked why there, the answer was simple: “That’s where the light is.”

Alzheimer’s disease (AD) researchers may have been doing something similar for decades. They have been searching for drugs to remove amyloid from the brains of patients with dementia. This sticky protein plaque, a prominent feature of AD, was believed to be responsible for neuronal destruction, memory loss and disability.

Billions of dollars were spent chasing drugs to remove amyloid plaque from the brain. The results have been deeply disappointing. Many studies flamed out completely. Other drugs slightly slowed cognitive decline but did not restore memory or function.

A new concept about Alzheimer’s disease is emerging. Some researchers believe that the buildup of amyloid plaque occurs in response to an infectious assault on the brain. In other words, amyloid is a natural defense mechanism to an attack from a pathogen.

The most intriguing clue isn’t a high-tech brain scan or a breakthrough drug; it’s the shingles vaccine. Three independent populationwide studies now suggest that this humble shot may modestly reduce dementia risk, forcing medicine to ask whether it’s time to look beyond the lamp post.

These natural experiments involve an old shingles vaccine, Zostavax. Doctors recommend it to reduce the chance of a painful attack of shingles, which could occur to anyone who had chickenpox as a child.

The latest research reinforces the herpes virus theory of Alzheimer’s disease.

There have been two prior studies of this vaccine. The first was set in Wales and the other examined data from Australia. In that research, scientists took advantage of “natural experiments.” In these countries, health care policymakers established arbitrary eligibility dates for people to receive the vaccine. People only slightly older did not differ from those who were eligible in other important respects, but they could not be vaccinated. Because the vaccines were provided, most people who were eligible chose vaccination.

As a result, the protective effects of the vaccine were clear. Those who had been vaccinated were 20% less likely to be diagnosed with dementia over the next several years (Nature, April 2, 2025). That is a relative risk. The absolute difference was 3.5%. In Australia, people eligible for the shingles vaccine were 1.8% less likely to develop dementia (JAMA, April 23, 2025).

The most recent study comes from Canada (Lancet Neurology, February 2026). In Ontario, people born on or after Jan. 1, 1946, were eligible for the shingles vaccine. Electronic health records from 1990 to 2022 demonstrated an absolute difference in dementia diagnoses of 2 percentage points between those eligible for the shingles shot and those who missed it by a few weeks or months. Elsewhere in Canada, where there was no shingles vaccination program, there is no clear difference in risk of dementia by birth date.

The investigators write: “In conclusion, this study provides strong evidence of a protective effect of herpes zoster vaccination on incident dementia.”

The three natural experiments reinforce the idea that amyloid plaque buildup may be a reaction to infection. Hopefully, future research will reveal even more effective ways to delay, prevent or treat dementia.

In their column, Joe and Teresa Graedon answer letters from readers. Write to them in care of King Features, 300 W. 57th Street, 41st Floor, New York, NY 10019, or email them via their website: www.PeoplesPharmacy.com. Their newest book is “Top Screwups Doctors Make and How to Avoid Them.”