Research presented this week at the Alzheimer’s Association International Conference in London suggests that the use of blood tests for Alzheimer’s could eventually transform the way the disease is diagnosed and even predict who might develop Alzheimer’s and when.

One study presented on Tuesday showed that, when primary care doctors had access to blood test results, their accuracy in diagnosing whether a patient’s cognitive symptoms were caused by Alzheimer’s improved considerably.

Another study, presented on Wednesday and published in the journal JAMA, suggested that the blood tests could help predict an individual’s risk of developing Alzheimer’s-related cognitive impairment over two, five and 10 years’ time.

The findings signal that these tests — which detect forms of two proteins, amyloid and tau, that are biological markers of Alzheimer’s — have growing promise to help patients be diagnosed more easily and to inform healthy people, who do not have any symptoms of cognitive impairment, of their future risk.

The goal is to transform Alzheimer’s into a disease that can be treated at its earliest stages, and to ultimately find ways to delay and even prevent its devastating memory and thinking problems.

However, while experts are excited about the potential of the tests, they say questions remain about how best to use them, and some have concerns about their accuracy.

Diagnostic Help for Primary Care Doctors

Two blood tests are currently approved in the United States to help doctors diagnose Alzheimer’s in people with cognitive problems. Different variations of the tests exist, but the one found to be most accurate is typically referred to as the p-tau217 test, named for the type of tau it measures.

Because the tests are much cheaper and easier to administer than a PET scan or a lumbar puncture (the current gold standards for diagnosing Alzheimer’s), they could help more people receive a diagnosis and treatment.

The study presented at the Alzheimer’s Association conference on Tuesday, which has not yet been published in a peer-reviewed journal, provided new data on the accuracy of the tests when used in a primary care setting.

In the study, primary care doctors evaluated patients who had cognitive impairment and were asked to determine whether the patients’ symptoms were caused by Alzheimer’s, both before and after they received the patients’ blood test results. The patients were then evaluated by neurologists and received a PET scan or lumbar puncture to confirm the diagnosis.

When the primary care doctors had access to the blood test results, the accuracy of their diagnosis improved from 62 percent to 88 percent. The physicians also reported that they would change their treatment plans — for instance, starting or stopping medications, referring a patient to a memory clinic or discharging them — in more than 60 percent of the cases, based on the blood test results. The accuracy of the blood tests themselves was around 90 percent.

Before the tests existed, primary care doctors were faced with “almost an impossible task” of trying to determine whether a patient’s symptoms were caused by Alzheimer’s using very limited tools, said Dr. Sebastian Palmqvist, an associate professor of neurology at Lund University in Sweden, who led the research. Now they have “an actual test that could help them do the job.”

The potential benefit of the tests may be greatest for people who do not have access to a specialty memory clinic.

“Not everybody has a PET scan available; not everybody can do spinal taps out in rural Minnesota, rural America,” said Dr. Ronald Petersen, a professor of neurology at the Mayo Clinic in Rochester, Minn. In such places, he added, a blood test is “much more viable.”

Dr. Palmqvist said he thinks that, if a doctor does a thorough evaluation in addition to the blood test, in most cases they can trust the results to make a diagnosis without ordering a PET scan to confirm.

But other experts currently think the blood tests alone should not be used to diagnose a person with Alzheimer’s. In his clinic, Dr. Petersen said he sees “instances where there’s inconsistencies. The blood test is positive and the PET scan is negative, or the PET scan is positive and the blood test negative.” As a result, he corroborates the blood test results to make a diagnosis, particularly if he is going to recommend that a patient try one of the new drugs designed to treat Alzheimer’s.

The issue of inaccuracy is particularly pressing in primary care settings, where a broad swath of patients may be screened, leading to potentially higher false positive rates. False negative results can also occur with the tests, though they are less common.

There may be several reasons the tests might return a false positive. A notable one is that other chronic health conditions can skew the tests’ results, particularly kidney disease, which affects roughly one-third of Americans 65 and older.

Other data presented at the conference, also currently unpublished, showed similar levels of test accuracy in a real-world setting, with false positive results occurring with the p-tau217 test, specifically, about 10 percent of the time. But without confirmation from a PET scan or lumbar puncture, that could still result in “a lot of people that are walking around with a positive test who likely may not all have Alzheimer’s disease,” said Michelle Mielke, a professor of epidemiology and prevention at Wake Forest University School of Medicine who led the work.

A false positive is concerning, Dr. Mielke said, because there could be other factors affecting a person’s cognition, like medications, sleep disturbances or substance use, that might not be addressed if the problem is wrongly attributed to Alzheimer’s.

Predicting the Risk of Alzheimer’s

The study presented and published on Wednesday found that, when blood tests measured higher levels of p-tau217, that reading predicted a greater likelihood of developing Alzheimer’s-related cognitive impairment, and that the tests could estimate that likelihood up to 10 years out.

The lead author, Rachel F. Buckley, an associate professor of neurology at Mass General Brigham and Harvard Medical School, said that a strength of the study was its ability to estimate an individual’s risk. For example, she said, the blood test could help predict “If I’m a 75-year-old man and I have this amount of p-tau217 in my blood plasma, what is my percentage level of risk of cognitive impairment over five years?”

Currently, experts do not recommend the blood tests for people who do not have cognitive impairment, except to help screen them for eligibility to participate in clinical trials of potential Alzheimer’s therapies. That is partly because there are still questions about the accuracy of making such predictions in cognitively healthy people. The researchers and other experts added that predicting risk should also include other factors, among them age, genetics and other health conditions.

Crucially, there are not yet any treatment options available for people without symptoms, said Maria Carrillo, the chief science officer of the Alzheimer’s Association, which has issued blood test guidelines and a diagnostic checklist for primary care doctors.

But, she and others said, that could be about to change. In the next six months to two years, experts expect the results of two clinical trials testing whether the anti-amyloid drugs approved for symptomatic patients can also delay or prevent the onset of cognitive impairment in asymptomatic people.

“If indeed the promise of these treatments in asymptomatic individuals becomes reality,” Dr. Carrillo said, then blood tests to help predict Alzheimer’s will be in widespread demand.

“If that hypothesis is confirmed in these clinical trials, then that’s really going to change most of our practice,” said Dr. Adam Boxer, a professor of memory and aging at the University of California, San Francisco. “We may need to start screening people before the onset of symptoms in the primary care setting for Alzheimer’s disease risk, and if they seem to be at very high risk, then we may want to intervene with an anti-amyloid drug.”

The new study, which involved nearly 2,700 cognitively healthy older adults from six different research cohorts, predicted that having very high p-tau217 levels would be associated with a 38 percent risk of developing cognitive impairment in the next five years and a 78 percent risk in 10 years.

For people with the next-highest level of p-tau217, the risk was 24 percent over five years and 62 percent over 10. People with only slightly elevated levels had a risk of 15 percent over five years and 45 percent over 10. There was little difference in the two-year risk for people with varying levels of p-tau217, ranging from 1 to 4 percent.

“This is an important study,” said Dr. Eric Reiman, the chief executive officer emeritus at Banner Alzheimer’s Institute in Phoenix, who is co-leading one of the trials testing anti-amyloid drugs for asymptomatic people. Dr. Reiman, a founder of a company that makes a p-tau217 antibody used in some of the Alzheimer’s blood tests, noted that the tests correlated tau levels with measures of amyloid plaques, and that evidence was what “predicts a person’s short, intermediate and long-term risk of progressing to cognitive impairment.”

Dr. Reisa Sperling, a professor of neurology at Mass General Brigham and Harvard Medical School and the study’s senior author, said that the goal was for Alzheimer’s blood tests to be good enough that they could be used in the same way doctors use cholesterol and blood pressure measures to predict cardiovascular risks, and blood sugar tests to predict diabetes risks.

The idea is to “detect disease before symptoms and provide as much information about what we know and don’t know, so that people can make the right decision about seeking treatment,” said Dr. Sperling, who is a leader of the other trial of Alzheimer’s drugs for asymptomatic people.

“This is our first step to understanding that that is indeed going to be possible for Alzheimer’s,” Dr. Carrillo added. She said that the new studies show “what the future could look like, a future that is actually not that far out.”

But until there is a treatment for preventing or delaying the onset of Alzheimer’s, the experts reiterated that they would not recommend that asymptomatic people take these tests. While some might want the information and feel empowered making lifestyle changes like improved diet or exercise, for others, the prospect of developing Alzheimer’s might be devastating.

“You can imagine how people may react to this differently,” Dr. Petersen said.