Category: 8. Health

Found in everything from protein bars to energy drinks, erythritol has long been considered a safe alternative to sugar. But new research suggests this widely used sweetener may be quietly undermining one of the body’s most crucial protective barriers – with potentially serious consequences for heart health and stroke risk.

A recent study from the University of Colorado suggests erythritol may damage cells in the blood-brain barrier, the brain’s security system that keeps out harmful substances while letting in nutrients. The findings add troubling new detail to previous observational studies that have linked erythritol consumption to increased rates of heart attack and stroke.

In the new study, researchers exposed blood-brain barrier cells to levels of erythritol typically found after drinking a soft drink sweetened with the compound. They saw a chain reaction of cell damage that could make the brain more vulnerable to blood clots – a leading cause of stroke.

Related: Study Reveals How This Artificial Sweetener May Cause Heart Damage

Erythritol triggered what scientists call oxidative stress, flooding cells with harmful, highly reactive molecules known as free radicals, while simultaneously reducing the body’s natural antioxidant defences. This double assault damaged the cells’ ability to function properly, and in some cases killed them outright.

But perhaps more concerning was erythritol’s effect on the blood vessels’ ability to regulate blood flow. Healthy blood vessels act like traffic controllers, widening when organs need more blood – during exercise, for instance – and tightening when less is required.

They achieve this delicate balance through two key molecules: nitric oxide, which relaxes blood vessels, and endothelin-1, which constricts them.

The study found that erythritol disrupted this critical system, reducing nitric oxide production while ramping up endothelin-1. The result would be blood vessels that remain dangerously constricted, potentially starving the brain of oxygen and nutrients. This imbalance is a known warning sign of ischaemic stroke – the type caused by blood clots blocking vessels in the brain.

Even more alarming, erythritol appeared to sabotage the body’s natural defence against blood clots. Normally, when clots form in blood vessels, cells release a “clot buster” called tissue plasminogen activator that dissolves the blockage before it can cause a stroke. But the sweetener blocked this protective mechanism, potentially leaving clots free to wreak havoc.

The laboratory findings align with troubling evidence from human studies. Several large-scale observational studies have found that people who regularly consume erythritol face significantly higher risks of cardiovascular disease, including heart attacks and strokes.

One major study tracking thousands of participants found that those with the highest blood levels of erythritol were roughly twice as likely to experience a major cardiac event.

However, the research does have limitations. The experiments were conducted on isolated cells in laboratory dishes rather than complete blood vessels, which means the cells may not behave exactly as they would in the human body. Scientists acknowledge that more sophisticated testing – using advanced “blood vessel on a chip” systems that better mimic real physiology – will be needed to confirm these effects.

The findings are particularly significant because erythritol occupies a unique position in the sweetener landscape. Unlike artificial sweeteners such as aspartame or sucralose, erythritol is technically a sugar alcohol – a naturally occurring compound that the body produces in small amounts.

This classification helped it avoid inclusion in recent World Health Organization guidelines that discouraged the use of artificial sweeteners for weight control.

Erythritol has also gained popularity among food manufacturers because it behaves more like sugar than other alternatives.

While sucralose is 320 times sweeter than sugar, erythritol provides only about 80% of sugar’s sweetness, making it easier to use in recipes without creating an overpowering taste. It’s now found in thousands of products, especially in many “sugar-free” and “keto-friendly” foods.

Trade-off

Regulatory agencies, including the European Food Standards Agency and the US Food and Drug Administration, have approved erythritol as safe for consumption. But the new research adds to a growing body of evidence suggesting that even “natural” sugar alternatives may carry unexpected health risks.

For consumers, the findings raise difficult questions about the trade-offs involved in sugar substitution. Sweeteners like erythritol can be valuable tools for weight management and diabetes prevention, helping people reduce calories and control blood sugar spikes.

But if regular consumption potentially weakens the brain’s protective barriers and increases cardiovascular risk, the benefits may come at a significant cost.

The research underscores a broader challenge in nutritional science: understanding the long-term effects of relatively new food additives that have become ubiquitous in the modern diet.

While erythritol may help people avoid the immediate harms of excess sugar consumption, its effect on the blood-brain barrier suggests that frequent use could be quietly compromising brain protection over time.

As scientists continue to investigate these concerning links, consumers may want to reconsider their relationship with this seemingly innocent sweetener – and perhaps question whether any sugar substitute additive is truly without risk.

Havovi Chichger, Professor, Biomedical Science, Anglia Ruskin University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Type 1 diabetes, as most textbooks tell it, shows up early, marches in with unmistakable warning signs, and never goes away. A single drop in daily insulin can mean trouble.

That story has shaped treatment for nearly a century, and in many places, it works. Yet most of the data behind that story comes from Europe and North America. What if a different chapter has been unfolding elsewhere all along?

In clinics across sub-Saharan Africa, nurses have long whispered to visiting researchers that some young patients labeled “type 1” seemed to hold on without insulin for weeks, even months.

The hunch was hard to test – supplies are scarce, and lab work can be expensive – but the anecdotes kept piling up. Those whispers have now turned into evidence, and the findings are shaking up what we thought we knew about diabetes.

Surprising African clinic data

Nearly 900 children, teens, and young adults in Cameroon, Uganda, and South Africa joined a project called the Young-Onset Diabetes in Sub-Saharan Africa (YODA) study.

Blood samples, genetic scans, and clinical histories poured in. The results, published in The Lancet Diabetes & Endocrinology, were striking.

About 65 percent of participants lacked the autoimmune fingerprints that define classic type 1 diabetes, and the usual high-risk immune genes were missing as well.

Only 8.5 percent carried more than one diabetes-related autoantibody, a stark contrast to the roughly 90 percent seen within months of diagnosis in Western cohorts.

Redefining the diabetes diagnosis

Instead of the immune system destroying insulin-making cells, this new subtype looks like something else altogether – an insulin-deficient illness that has not yet revealed its trigger.

It sits outside type 2 diabetes, which is tied to insulin resistance, and it doesn’t match malnutrition-related diabetes either. Clinicians had their suspicions; now the data back them up.

“These new research findings confirm our long-standing suspicion. We have always wondered why many young people diagnosed with type 1 diabetes manage to survive without insulin, at least for some time, which would be unusual in typical type 1 diabetes,” said Dr Jean-Claude Katte of the University of Exeter.

Types of diabetes

The YODA team compared African participants with more than 3,000 children and adolescents from the U.S. SEARCH for Diabetes in Youth study.

The novel pattern emerged in a small group of Black American children and was absent in White participants. That split hints at a mix of ancestral roots and environmental factors. It also rules out simple lab error.

Every sample was tested with assays calibrated for African populations, and the same pattern appeared across three countries with different diets, infections, and climates.

“These findings are a wake-up call. They challenge our assumptions about type 1 diabetes and show that the disease may present differently in African children and adolescents,” said Professor Moffat Nyirenda, who directs the MRC/UVRI–LSHTM Uganda Research Unit.

“We urgently need to deepen our investigations into the biological and environmental factors driving this form of diabetes and ensure our diagnostic and treatment approaches are fit for purpose in African settings.”

Dr Katte added, “Our next step is to investigate possible causes – ranging from infections and nutritional factors to environmental toxins. If we can find the cause, we may be able to prevent new cases and find new treatments.”

Professor Eugene Sobngwi of Cameroon’s Ministry of Public Health put it bluntly, saying, “We must invest in context-specific research. If we don’t, we risk misdiagnosing and mistreating millions of people.”

What this means globally

Roughly nine million people worldwide live with type 1 diabetes. If two out of every three young patients in parts of Africa actually have a different disease, the global numbers – and the research priorities that follow – need a hard look.

Mislabeling matters. Classic type 1 patients need lifelong insulin from day one; those with the new subtype may respond to therapies that protect or boost their still-working beta cells.

Catching that difference could spare families the cost and risk of unnecessary high-dose insulin while steering researchers toward fresh prevention strategies.

Diagnosing diabetes type

The study leaves open questions. Autoantibodies can fade, so testing closer to the time of diagnosis could capture more detail.

In addition, teasing apart infections, dietary patterns, and possible toxins will take careful fieldwork and, just as importantly, local funding.

Yet the path is clear: broaden the research lens, update diagnostic manuals, and train healthcare teams to spot red flags – such as a young patient who stays stable on surprisingly little insulin.

The YODA data show that those cases are not rare outliers; they are part of a hidden majority in several African settings.

Diabetes textbooks will need an update, but the bigger story is about equity in science. Decades of work built on narrow datasets can miss whole chapters of human biology. Broadening the map doesn’t just help the regions being studied; it fine-tunes medicine for all.

The full study was published in the journal The Lancet Diabetes & Endocrinology.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–

In the modern world, animals are mostly seen through the eyes of science, conservation, or affection. But go back a few hundred years, and the view was very different.

People didn’t just fear wild animals because of what they could physically do– they feared them for what they were believed to represent. In many cultures, certain animals became part of local myths, tied to bad luck, death, or even the supernatural.

Disclaimer: The Times of India does not claim these beliefs as fact. This article is based on folklore, mythology, and historical narratives from publicly available sources.

Let’s look at a few animals that still carry reputations as omens or bad signs in some parts of the world.

A breakthrough AI model can spot silent structural heart disease from a simple ECG, promising to catch dangerous conditions earlier, streamline patient care, and close the diagnostic gap missed by traditional screening.

Study: Detecting structural heart disease from electrocardiograms using AI. Image Credit: DC Studio / Shutterstock

In a recent study published in the journal Nature, a group of researchers investigated whether an artificial intelligence (AI) electrocardiogram (ECG) model can reliably detect diverse structural heart diseases (SHDs) across various hospitals and care settings, outperforming standard physician review. The model, called EchoNext, was developed as a multitask classifier to address collinearity among different SHD component labels.

Background

Every minute, another United States (US) patient enters the hospital with symptoms that may mask underlying SHD. Treating SHD already drains the nation of more than 100 billion dollars each year. Yet, an estimated 6.4% of older adults carry clinically significant valvular heart disease (VHD) that has never been diagnosed, in addition to 4.9% already diagnosed, making the total prevalence over 11%.

Early echocardiography saves lives, but ultrasound labs, trained readers, and patient travel costs remain barriers, leaving busy clinicians guessing whom to scan.

Large-scale digital ECG archives and modern AI offer a low-cost alternative: if one ten-second ECG could reliably uncover silent disease, scarce imaging resources could be directed to those who need them most.

Further research is needed to determine whether algorithm-guided screening improves survival and equity. Additionally, the paper discusses potential deployment strategies for such models, including both “gatekeeper” and “safety net” applications, each with unique benefits and trade-offs for clinical practice.

About the study

Investigators assembled 1,245,273 paired ECG-echocardiogram records from 230,318 adults treated between 2008 and 2022 at eight NewYork-Presbyterian (NYP) hospitals, reserving patient-level splits for training, validation, and testing.

SHD was labeled when any guideline defined abnormality was present with left ventricular ejection fraction (LVEF) ≤ 45%, left ventricular wall thickness ≥ 1.3 cm, moderate or worse right ventricular dysfunction, pulmonary artery systolic pressure (PASP) ≥ 45 mm Hg, or tricuspid regurgitation jet velocity ≥ 3.2 m/s as an alternative pulmonary hypertension definition, moderate or worse regurgitation/stenosis of any valve, or a moderate/large pericardial effusion.

The authors note these thresholds are somewhat arbitrary, as different studies and guidelines may use varying cutoffs.

A convolutional neural network named EchoNext ingested the raw 12-lead waveform, along with seven routine ECG parameters and age/sex data. Performance was first measured on a held-out NYP test set, and then on external cohorts from Cedars-Sinai, the Montreal Heart Institute, and the University of California, San Francisco.

Generalization across age, sex, race, ethnicity, and clinical context was assessed. Silent “shadow” deployment ran EchoNext on 84,875 consecutive ECGs from patients without previous echocardiography, storing scores but not influencing care.

Finally, a single-site pilot, Detecting Structural Heart Disease Using Deep Learning on an Electrocardiographic Waveform Array (DISCOVERY), prospectively invited adults with no recent imaging to undergo echocardiography stratified by a predecessor model’s risk score; EchoNext was analyzed post hoc.

Study results

EchoNext, an AI-powered ECG model, excelled in retrospective analysis. Within the eight-hospital NYP test set, it detected composite SHD with an area under the receiver operating characteristic (AUROC) of 85.2% and an area under the precision–recall curve (AUPRC) of 78.5%. Accuracy remained consistent across academic and community campuses and did not falter when training and test sites were exchanged, demonstrating generalization.

External validation at Cedars-Sinai Medical Center, the Montreal Heart Institute (MHI), and the University of California, San Francisco, yielded AUROC values of 78 to 80%, despite higher disease prevalence.

Disease-specific performance: LVEF ≤ 45% achieved AUROC 90.4%, while PASP ≥ 45 millimeters of mercury reached 82.7%. The authors emphasize that AUPRC values for component diseases are highly dependent on the underlying disease prevalence and should not be directly compared across conditions or use cases.

A 150-trace reader study compared EchoNext with thirteen cardiologists. Reviewing wide age, sex, waveform, and ECG intervals, physicians identified SHD correctly in 64% of cases. The AI alone achieved 77% accuracy, and when clinicians were shown the algorithmic risk score, their accuracy increased modestly to 69%, underscoring that the model captured prognostic patterns that were hidden from expert eyes. It is important to note that cardiologists in this assessment had access only to de-identified ECGs and routine parameters, without any clinical context, which is not typical of standard clinical care.

To estimate clinical opportunity at scale, the team silently ran EchoNext on 124,027 ECGs recorded in 2023 from 84,875 adults who had never undergone echocardiography. The model flagged nine percent of traces as high risk. Usual care, nevertheless, left 45% of these individuals without follow-up imaging, suggesting that an estimated 1,998 cases of silent SHD might have been intercepted had the alert been live, based on modelled prevalence and sensitivity scenarios provided in the paper.

Among the 15,094 patients who eventually received echocardiography, EchoNext preserved accuracy (AUROC 83%; AUPRC 81%) and delivered a positive predictive value of 74%, reinforcing its reliability in a contemporary workflow. The paper also provides modelled performance estimates at different prevalence scenarios and sensitivity thresholds, underscoring the practical implications for population-wide screening.

Prospective evidence came from the DISCOVERY pilot, which recruited 100 imaging-naive adults. Post hoc EchoNext scoring revealed clear tiers, with previously unrecognized SHD present in 73% of high-risk participants, 28% of moderate-risk participants, and 6% of low-risk participants; moderate to severe left-sided VHD followed a similar gradient.

These results illustrate the model’s capacity to triage scarce echocardiography resources toward those most likely to benefit, while sparing low-risk individuals unnecessary testing. The original trial used a predecessor model (ValveNet) to stratify risk and recruit participants, and the EchoNext model was applied retrospectively to these participants for further analysis.

Conclusions

To summarize, EchoNext demonstrates that an AI-enhanced ECG can detect SHD associated with LVEF reduction, elevated PASP, and significant VHD, with AUROC and AUPRC metrics superior to those of cardiologists. By flagging high-risk patients for timely echocardiography, the algorithm promises to shrink diagnostic delay and the billion-dollar burden of SHD while maintaining equity across sites and demographics. However, the authors caution that AI-based screening may also carry potential risks, including patient anxiety from false positives or bias in clinical adoption, and highlight the need for further study of these aspects.

The public release of code and data encourages independent validation; however, large pragmatic trials must verify that AI-guided ECG screening truly improves survival, quality of life, and healthcare value. Notably, the authors have released a large de-identified dataset and a benchmark AI model (the Columbia mini-model) to support further research and enable transparent comparison of future algorithms.

Newswise — Consumption of soft drinks, supplemented with white sugar, alters the DNA of gut bacteria and affects the host immune system. The good news? These effects are reversible.

The findings by researchers from the Ruth and Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Prof. Naama Geva-Zatorsky, Ph.D. student Noa Gal-Mandelbaum, and other members of the Geva-Zatorsky team were recently published in Nature Communications. Dr. Tamar Ziv and the Smoler Proteomics Center at the Technion assisted with the research.

Gut bacteria are important members of the microbial community within our body, i.e., the microbiome. hese bacteria, which have co-evolved with humans for generations, are so essential to human health in general and to the development of the immune system in particular that we cannot function without them. The human gut is constantly influenced by changes in the environment. To keep up, our gut bacteria must adapt quickly. They do this through a process called functional plasticity, which allows them to change their behavior and functions in response to factors like nearby microbes, our health status, and what we eat.

A previous study by the Geva-Zatorsky lab discovered that one way gut bacteria adapt to environmental changes is through DNA inversions—rapid genetic switches that help them respond and defend themselves. In their current study, the researchers investigated how these DNA inversions occur in response to dietary factors. They found that consuming soft drinks, which contain white sugar, can alter the DNA of gut bacteria and, in turn, impact the host’s immune system. The study focused on Bacteroides thetaiotaomicron – a prominent gut member that takes part in preventing gut inflammation, preserving the gut mucus layer, and protecting the host from pathogen invasion. The researchers studied the effects of consuming different dietary components on the DNA inversion profile of these bacteria, in vitro, in mice, and humans. They discovered that white sugar consumption causes DNA inversions in these bacteria, which led to changes in inflammatory markers of the immune system, including ones in T-cell populations, cytokine secretion, and gut permeability.

The good news is that these effects are reversible; once mice stopped consuming white sugar, the bacterial DNA inversion state reverted, and the immune system’s state returned to normal. This study stresses the importance of studying the complex effects of nutrition on the microbiome and our health state, and the researchers assess that this will allow tailored dietary recommendations to human subjects to improve their immune system’s state and their health in general.The research was supported by the Technion President’s Fund, RTICC – the Rappaport Technion Integrated Cancer Research Center, the Alon Fellowship for Outstanding New Faculty, the Seerave Foundation, CIFAR, and the European Research Council (ERC).

For the full article in Nature Communications

Click here for photos

In the photos:

1. Prof. Naama Geva-Zatorsky (Credit: Studio Lucido)
2. Noa Gal-Mandelbaum
3. The effect of a sweetened beverage on bacterial function – a demonstration of functional plasticity. Created by AI, inspired by an illustration by Tom Blum, a student in the research group.

# # #

Since 1924, the Technion – Israel Institute of Technology has been home to dreamers, inventors, and innovators who make the seemingly impossible possible. Within classrooms and laboratories, Technion students and faculty reach beyond their disciplines to pursue the unconventional. Alumni carry that pioneering spirit to the highest positions in the Israeli government, entrepreneurial startups, and corporations around the world. From technologies that keep Israel safe, to environmentally friendly rocket fuel, to microfluidic chips that detect heart disease and cancer, Technion people help the State of Israel flourish while solving some of humankind’s most intractable problems.

The American Technion Society supports visionary education and world-changing impact through the Technion – Israel Institute of Technology. Based in New York City, we represent thousands of US donors, alumni and stakeholders who invest in the Technion’s growth and innovation to advance critical research and technologies that serve the State of Israel and the global good. Since 1940, our nationwide supporter network has funded new Technion scholarships, research, labs, and facilities that have helped deliver world-changing contributions and extend Technion education to campuses in three countries.

Newswise — Children’s Hospital Los Angeles was first in the nation to offer a clinical liquid biopsy assay for pediatric solid tumors. Launched in late 2022 by CHLA’s Center for Personalized Medicine, the test is the only one of its kind using whole genome sequencing to detect copy number changes in these tumors—including brain tumors.

Katrina O’Halloran, MD, MS—a pediatric oncologist in the Neuro-Oncology Program in the Cancer and Blood Disease Institute at CHLA—was an investigator on the initial pilot study and is leading research into the team’s clinical use of this groundbreaking test for patients with brain tumors. Dr. O’Halloran presented the team’s latest data at the Children’s Oncology Group meeting in April, as well as at the 8th Biennial Pediatric Neuro-Oncology Conference in May.

She shares how the neuro-oncology team is using this liquid biopsy assay today—and how it could change the paradigm for future patient care.

How does the test work?

A liquid biopsy enables us to detect tumor DNA or other tumor markers in a minimally invasive way. Our test uses a sample of cerebrospinal fluid (CSF), which we typically obtain through a lumbar puncture.

CHLA’s liquid biopsy platform is also available for other pediatric solid tumors using a blood sample, and for retinoblastoma using the aqueous humor fluid in the eye. But for brain tumors, we use CSF because the blood-brain barrier may prevent tumor fragments from crossing into the blood.

After we collect the sample, the Center for Personalized Medicine performs low-pass whole genome sequencing. This gives us a bird’s eye view of the genome and any copy number changes—gains or losses of specific chromosomes—which help determine the presence of tumor DNA.

In addition, the lab now offers a new, targeted sequencing panel for liquid biopsy applications. Rather than looking across the whole genome, this panel zooms in on specific cancer-causing genes to detect DNA sequence-level alterations.

What types of brain tumors are you testing?

We have been routinely using it in patients with embryonal tumors, especially those with medulloblastoma. Embryonal tumors have more of the copy number changes that you see with the low-pass whole genome sequencing.

Our new targeted sequencing panel test is helpful for other types of tumors, such as diffuse midline gliomas, which tend to have single-gene changes that are disease-causing.

How is the team using liquid biopsy in clinical care?

We typically use it at diagnosis and at the end of therapy to help assess the patient’s response to treatment. For some very high-risk patients we’re performing it more often, including in surveillance. We want to see if we can detect relapse at a much earlier stage—before it shows up on MRI.

Just a few weeks ago, we also diagnosed our first patient based solely on liquid biopsy. That child had a rare central nervous system tumor called diffuse leptomeningeal glioneuronal tumor (DLGNT). We had a suspicion this was the case, but this tumor is quite challenging to biopsy in a traditional way. Because DLGNT has a specific genetic alteration, we were able to diagnose it with a liquid biopsy—sparing the child an invasive brain surgery.

Are the results influencing treatment decisions?

In some cases, yes. That said, the liquid biopsy test is one factor, in combination with many other factors, that we use in our shared decision-making with families.

For example, we had one patient who had a metastatic tumor at diagnosis and because of side effects was not able to receive all planned therapy. The patient then had a positive liquid biopsy at the end of treatment. Because of those risk factors, along with some concerning MRI findings, we decided to continue treatment with a different regimen. Once the child had several negative liquid biopsies, we stopped therapy.

It has now been over a year, and the patient has continued to test negative and is back at school and doing well. Would we have made the same decisions without the liquid biopsy? Maybe not. The liquid biopsy particularly helped us know when to stop that additional therapy.

What does your data show so far?

To date we have sent 81 liquid biopsies from 35 patients, ranging in age from less than 1 year to 20 years. Out of those, 20 liquid biopsies have found additional DNA alterations when compared with the tissue biopsy. In other words, it’s possible the liquid biopsy may provide a more comprehensive assessment. A tissue biopsy only reflects the tumor in one location, and we know that tumors are not uniform.

We’ve also found that most patients with persistently positive liquid biopsies ultimately experienced progression or relapse, indicating that the test seems to be predictive of recurrence.

In addition, in 16 cases of serial liquid biopsies, we saw changes over time, including new alterations consistent with tumor evolution. This is exciting because we know tumors evolve, and this is part of how they become resistant to therapy. Liquid biopsy allows us to watch this genomic evolution in almost real time.

Can physicians from other institutions order this liquid biopsy?

Yes. It is CAP-certified and CLIA-certified, and we generate a formal pathology report that we share with the outside provider and patient. It typically takes two to three weeks to get results, which is similar to pathology results taken from actual tumor biopsies.

How do you think this test will impact patient care in the future?

It’s going to help us better tailor care to each patient. My hope is that liquid biopsy monitoring could eventually allow us to de-escalate therapy and reduce toxicities for children who are responding well—and intensify therapy for those who are not responding.

Liquid biopsy findings could also help us understand more about how these tumors evolve to evade treatment. This may lead to new therapies that could also be tailored to each patient. If a tumor begins to develop X changes, we give Y therapy.

That is years down the road. But for the first time, we can monitor the genetic evolution of these tumors in close to real time, in a minimally invasive way. It’s already helping us diagnose patients and assess their response to therapy. That is immensely exciting, and it points to a future where treatment can truly be personalized to improve outcomes for each child.

A groundbreaking study reveals how a single gene variant shapes the immune landscape of the brain and blood, unlocking new opportunities for early detection and intervention across Alzheimer’s, Parkinson’s, and other neurodegenerative disorders.

Study: APOE ε4 carriers share immune-related proteomic changes across neurodegenerative diseases. Image Credit: Andrii Vodolazhskyi / Shutterstock

In a recent study published in the journal Nature Medicine, researchers identified a conserved immune signature associated with the apolipoprotein E ε4 variant (APOE ε4) across cerebrospinal fluid (CSF), brain, and plasma, regardless of the presence of neurodegenerative disease.

The APOE ε4 gene is the most significant genetic risk factor for late-onset Alzheimer’s disease (AD). However, mounting evidence suggests APOE ε4 carriage may play a role in other neurodegenerative diseases. APOE ε4 has been linked to a lower age of onset and a higher risk of Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS), and frontotemporal dementia (FTD). APOE ε4 has also been linked to poor cognition and faster cognitive decline in PD, elevating PD dementia (PDD) risk.

Notwithstanding APOE ε4’s deleterious impact, little is known about its biological mechanisms and whether and how it changes across neurodegenerative diseases. The paper frames this within the evolutionary concept of “antagonistic pleiotropy,” suggesting that the gene’s pro-inflammatory effects may be beneficial against infectious diseases in youth but become harmful with age. Previously, the authors reported that APOE ε4 carriers shared a pro-inflammatory proteomic signature in the CSF regardless of cognitive status in mild cognitive impairment and AD. Whether this applies to other neurodegenerative diseases remained unknown.

The study and findings

The present study explored systemic proteomic changes in APOE ε4 carriers with neurodegenerative diseases. First, SomaScan proteomic data from the Global Neurodegeneration Proteomics Consortium (GNPC) dataset, which reflects real-world clinical heterogeneity, were utilized for the CSF proteome profiling of individuals with AD, PD, or control (non-impaired) status. Using mutual information, 229 CSF proteins associated with APOE ε4 were identified in non-impaired controls.

Classification and regression trees (CART) models showed that these 229 proteins could differentiate APOE ε4 carriers from non-carriers across PD and AD. A functional enrichment analysis of CSF APOE ε4 proteins indicated significant enrichment for apoptosis, viral processes, protein folding and phosphorylation, RNA/DNA processes, cellular processes, and rhythmicity.

APOE ε4 carriers across different neurodegenerative diseases share a common systemic proteomic change reflective of pro-inflammatory immune dysregulation. The figure is created with BioRender.com.

Further analysis for immune-specific processes revealed APOE ε4 enrichment in various infection-related pathways, including herpes, influenza A, hepatitis, measles, and Epstein-Barr virus (EBV). Significant enrichment was also observed for B-cell, T-cell, and inflammatory signaling cascades. Next, immune cell subtype enrichment analysis revealed the most APOE ε4 enrichment in intermediate and non-classical monocytes among innate immune cells.

Among adaptive immune cells, memory cluster of differentiation 8 (CD8) T cells, regulatory T (Treg) cells, and memory CD4 T cells were the most enriched. Besides, γδ T cells and natural killer (NK) cells showed APOE ε4 enrichment. In the liver, a cell-type-specific enrichment analysis revealed the most APOE ε4 enrichment in Kupffer cells and hepatocytes.

Next, the researchers examined whether APOE ε4 CSF proteome changes were reflected in the plasma and used the GNPC dataset for plasma proteome profiling of AD, PDD, FTD, PD, ALS, and non-impaired controls. Fifty-eight plasma proteins associated with the APOE genotype were identified in non-impaired controls. CART modeling revealed that these 58 proteins could strongly differentiate between APOE ε4 carriers and non-carriers across neurodegenerative diseases, and this signature was found to be consistent across different sexes and racial groups.

APOE ε4 plasma processes showed significant enrichment in biological processes, including protein processes, cellular processes, viral processes, DNA/RNA processes, and apoptosis. Viral processes were the most significantly enriched in both plasma and CSF. This was supported by similar enrichments in infection and immune pathways, including hepatitis and EBV. Intermediate and non-classical monocytes and basophils were enriched for APOE ε4 proteins among innate immune cells.

NK cells, γδ T cells, memory CD8 T cells, and naïve CD8 T cells were enriched for APOE ε4 proteins among adaptive immune cells. In the liver, T cells and Kupffer cells were mainly enriched, while hepatocytes showed little enrichment. These data indicate that genotype-specific proteomic changes detected in the CSF were also reflected in the plasma in APOE ε4 carriers and non-carriers.

Further, the team explored whether peripheral proteomic changes were mirrored in the brains of APOE ε4 carriers and non-carriers. To this end, they utilized the proteomic data of the dorsolateral prefrontal cortex (dlPFC) from individuals with AD, FTD, PD, PDD, ALS, and non-impaired individuals, as part of the Accelerating Medicines Partnership for AD (AMP-AD) UPenn Proteomics Study. Using mutual information, 248 APOE ε4 proteins were identified in the dlPFC.

Functional enrichment analyses showed that the three main biological processes (viral processes, apoptosis, and protein folding) identified in the plasma and CSF were also significantly enriched in the dlPFC of APOE ε4 carriers across neurodegenerative diseases. Four of the most significantly enriched immune pathways in the plasma and CSF were also identified in APOE ε4 carriers disease-independently: hepatitis B, EBV, Escherichia coli infection, and viral carcinogenesis. Crucially, the study found that this brain signature was independent of the hallmark neurodegenerative pathologies, including amyloid-β, tau, TDP-43, and α-synuclein, reinforcing that the APOE ε4 effect is a fundamental vulnerability rather than a direct result of disease-specific protein aggregation.

An Unexpected Finding Challenges a Key Biomarker

In a notable and “unexpected finding,” the study revealed that plasma neurofilament light (NEFL) levels, a widely used biomarker of neurodegeneration, were consistently lower in APOE ε4 carriers. The authors note this contradicts some prior research and “raises important questions about the reliability of NEFL as a stand-alone biomarker,” suggesting its levels may be influenced by APOE ε4-related metabolic factors or clearance across the blood-brain barrier.

Finally, a correlation network analysis was performed to investigate the relationship between APOE ε4 proteins and clinical, demographic, and lifestyle factors in PDD, PD, AD, and non-impaired APOE ε4 carriers. They identified disease-specific relationships with APOE in neurodegenerative diseases. APOE was associated with race and sex in AD, diabetes in PD, age, hypertension, and body mass index in non-impaired controls, and chronic obstructive pulmonary disease and resting heart rate in PDD. The authors caution, however, that because the study is cross-sectional, they “cannot infer causality,” as these clinical factors could be a consequence of the neurodegenerative disease rather than a cause.

Conclusions

The findings showed that APOE ε4 carriers share a unique proteomic signature across the plasma, CSF, and brain, regardless of neurodegenerative disease. This signature was associated with enrichment for circulating immune cells and pro-inflammatory immune dysregulation. The authors propose a specific mechanism for this, suggesting that hyperactive peripheral immune cells may interact with and disrupt the blood-brain barrier (BBB), thereby driving neuroinflammation. However, proteins within this signature were uniquely correlated with clinical, demographic, and lifestyle factors in a neurodegenerative disease-specific manner.

This suggests that APOE ε4 confers a systemic biological vulnerability that is essential but insufficient for neurodegeneration, underscoring the need to account for gene-environment interactions. The authors also acknowledge limitations, including the “absence of validated biomarkers” to confirm all clinical diagnoses and the “absence of direct measures of routine inflammatory markers such as C-reactive protein,” which should be addressed in future studies.

Overall, the results reframe APOE ε4 as a pleiotropic immune modulator, rather than an AD-specific risk gene, providing a foundation for early intervention strategies and precision biomarker development across neurodegenerative diseases. The study concludes by calling for a conceptual shift in the field, moving from simply identifying genetic risk loci “toward functional characterization of established variants,” and establishing a roadmap for future research into these complex interactions.

A Food and Drug Administration panel on Monday discussing the use of antidepressants during pregnancy largely amounted to misinformation or facts taken out of context, according to several psychiatrists who tuned into the meeting.

The panel had promised to feature diverse viewpoints about antidepressants and pregnancy. But nearly all of the 10 panelists bucked medical consensus on the drugs’ safety and emphasized what they said were risks of taking the drugs while pregnant — such as causing autism, miscarriages or birth defects. In some cases, they claimed that antidepressants do not work at all and depression goes away on its own.

Three of the 10 panelists were from outside the U.S. Another runs a clinic to help people taper off of psychiatric drugs.

“They were really rousing concerns about safety that are not evidence-based or established, and not at all balanced with concerns about the risks of untreated depression,” said Dr. Joseph Goldberg, a clinical professor of psychiatry at the Icahn School of Medicine at Mount Sinai in New York City.

“I’m disappointed that the FDA brought people in from outside of the United States when there’s so many experts here in the United States who truly know this [medical] literature inside and out,” added Dr. Jennifer Payne, director of the Reproductive Psychiatry Research Program at the University of Virginia.

A spokesperson for the FDA said the claim that the panel was one-sided was “insulting to the independent scientists, clinicians, and researchers who dedicate their expertise to these panels.” The spokesperson added that “[FDA] Commissioner Makary has an interest in ensuring policies reflect the latest gold standard science and protect public health.”

Health and Human Services Secretary Robert F. Kennedy Jr. has called for a probe into the risks of antidepressants. His “Make America Healthy Again” report in May claimed there were “potentially major long-term repercussions” associated with use of the drugs in childhood.

The panel discussion focused on a class of antidepressants known as selective serotonin reuptake inhibitors, or SSRIs, which includes Lexapro, Prozac and Zoloft. The medications increase levels of serotonin, a chemical messenger in the brain that can improve mood.

Psychiatrists often advise women who are taking SSRIs to continue doing so during pregnancy, since the risks of untreated depression tend to outweigh the potential risks of the medication to mother and child. However, the decision is a personal one.

For ethical reasons, there are no randomized-control trials of SSRIs in pregnant women, meaning that data on potential risks mostly comes from observational studies and drug registries.

Some babies born to mothers taking SSRIs may develop symptoms such as jitteriness, irritability or difficulty eating or sleeping that resolve quickly — what’s known as “neonatal adaptation syndrome.” The symptoms may be caused by medication in the baby’s system or withdrawal from it.

Certain studies have also found a slightly elevated risk of miscarriage associated with antidepressant use in pregnancy, though others have found no association. However, there’s no convincing evidence to suggest that SSRIs are linked to autism or birth defects.

“Well-controlled studies continue to not find an association,” Payne said.

Women with a history of depression are also at an increased risk of symptoms recurring during pregnancy, and depression can come with its own risks — including thoughts of self-harm or low birth weight.

“The best thing a pregnant individual could do for herself and her baby is to get the treatment that they need,” said Dr. Nancy Byatt, a perinatal psychiatrist at UMass Chan Medical School, who was not part of the panel.

In addition to raising concerns about side effects, several panelists questioned the efficacy of antidepressants or suggested that data favoring the use of SSRIs was manipulated by the pharmaceutical industry.

“It’s been said that SSRIs help people who are severely depressed. They don’t,” said David Healy, an FDA panelist and fellow at the Royal College of Psychiatrists in the United Kingdom.

Goldberg and other psychiatrists said that’s simply untrue.

“You can say the moon landing was faked. Conspiracy theories abound in our world. But there is not a doubt about whether SSRIs work,” said Goldberg, a past president of the American Society of Clinical Psychopharmacology who has previously consulted for pharmaceutical companies. He said he was invited to join the FDA panel, but declined because the language of the invitation suggested it would not be a fair discussion.

Healy, the panelist, also claimed that some people “recover spontaneously” from depression.

Another panelist, psychologist Roger McFillin — who hosts a podcast that challenges conventional mental health advice — suggested that depression was not an illness, but rather a product of “women just naturally experiencing their emotions more intensely.” He added, without evidence, that many women feel coerced into taking antidepressants.

Psychiatrists who weren’t part of the discussion broadly decried those characterizations.

“I have never, ever, ever, ever heard of a third party pushing a prescription in pregnancy,” Goldberg said. However, he said he is aware of some obstetrician-gynecologists who aren’t as familiar with SSRIs wrongly telling pregnant patients to stop taking the medications.

The American College of Obstetricians and Gynecologists said in a statement on Monday that SSRIs can be lifesaving for some pregnant individuals.

“Today’s FDA panel on SSRIs and pregnancy was alarmingly unbalanced and did not adequately acknowledge the harms of untreated perinatal mood disorders in pregnancy,” the organization said. “On a panel of 10 experts, only one spoke to the importance of SSRIs in pregnancy as a critical tool, among others, in preventing the potentially devastating effects of anxiety and depression when left untreated during pregnancy.”

One panelist, Dr. Kay Roussos-Ross, a psychiatrist and an OB-GYN at the University of Florida College of Medicine, was the most vocal advocate for the use of SSRIs, often pushing back against others’ assessments of the risks. Psychiatrists who weren’t on the panel praised her science-backed commentary.

“All of us can find a study that agrees with exactly what we think,” Roussos-Ross told the panel. “But we need to look at the data very objectively.”

Several psychiatrists who weren’t on the panel said they worry the discussion could lead to regulatory action that makes it harder to access SSRIs. For example, panelist Dr. Adam Urato, chief of maternal-fetal medicine at MetroWest Medical Center in Framingham, Massachusetts, called for stronger warnings on SSRI labels.

“I am just very nervous that some of the discourse around the risks of antidepressants could lead to even more barriers for people seeking care,” said Dr. Lindsay Lebin, an assistant professor of psychiatry at the University of Colorado Anschutz Medical Campus, who was not part of the FDA discussion.