Researchers tracked “excess deaths” among adults aged 25 to 44 years and found disparate causes to blame
Studies conducted at the University of Minnesota and Boston University found that mortality rates among young adults have risen substantially since 2010 due to a variety of factors, pointing to a possible “mortality crisis” as they get older.
The researchers used data from the Centers for Disease Control and Prevention (CDC) and US Census Bureau to analyze nearly 3.4 million deaths in the US between 1999 and 2023 among adults aged 25 to 44 years, according to a Boston University press release.
They then used mortality data from 1999 through 2010 to project expected death rates for the later years and compared those projections to the actual post-2010 mortality rate to calculate the number of “excess deaths,” defined as “those [deaths] above what had been projected for a given period.”
“What we didn’t expect is how many different causes of death have really grown for these early adults,” said study lead author Elizabeth Wrigley-Field, PhD, of the University of Minnesota, in a U of M pressrelease. “It’s drug and alcohol deaths, but it’s also car collisions, it’s circulatory and metabolic diseases—causes that are very different from each other. That tells us this isn’t one simple problem to fix, but something broader.”
From that perspective, clinical laboratories could be part of the solution in tracking down these early conditions and steering young patients towards healthier outcomes.
“The rise in opiate deaths has been devastating for Americans in early and middle adulthood,” said sociologist Elizabeth Wrigley-Field, PhD, of the University of Minnesota in a press release. (Photo copyright: University of Minnesota.)
One-Two Punch
In 2019, excess mortality amounted to 41.7 deaths per 100,000 population, nearly 35% higher than expected, the researchers wrote. Then, in 2021 during the pandemic, excess mortality from all causes was nearly three times higher: 116.2 deaths per 100,000 population. In 2023, excess mortality decreased, but only to 79.1 deaths per 100,000 population.
“As a result, early adult mortality was 70.0% higher in 2023 than it would have been had pre-2011 trends continued,” the researchers wrote in Jama Network Open.
Speaking with Healio, Wrigley-Field described a “one-two punch that these age groups have seen: first, rising mortality since 2010; then the pandemic, only partially recovered from.”
Five causes accounted for nearly 75% of the excess deaths in 2023, the researchers found:
Drug poisoning, such as opiate overdoses (31.8%)
Residual natural causes (16%)
Transport-related deaths, such as motor vehicle accidents (14.1%)
Alcohol-related deaths (8.5%)
Homicide (8.2%)
The researchers also found that cardiometabolic conditions accounted for 9.2% of excess deaths. These include metabolic, circulatory and endocrine, and nutritional conditions, they noted.
Study co-author Andrew Stokes, PhD, of Boston University characterized the latter as a red flag, according to the Boston University press release. “Usually, it takes a lifetime to manifest cardiovascular disease and related mortality,” he said.
“Our findings underscore the urgent need for comprehensive policies to address the structural factors driving worsening health among recent generations of young adults,” he said in the U of M press release. “Solutions may include expanding access to nutritious foods, strengthening social services, and increasing regulation of industries that affect public health.”
Policy Measures
In their paper, the researchers suggested that policymakers should pay more attention to underlying causes such as opioid use, alcohol consumption, and traffic safety, as well as “ongoing consequences of the COVID-19 pandemic—which may be expressed in causes of death related to long-term consequences of infection, medical disruption, and social dislocation—and to deleterious health trends that predated it.”
“Individuals might not necessarily be able to reverse those factors, or their consequences, on their own, but public health collectively has been very successful at improving health through policies like tobacco regulation, to name one example,” Wrigley-Field told Healio.
She added that “the cardiometabolic causes of death stand out because these are really a bellwether of population health. These causes tend to be very responsive to the fundamentals of healthy living: healthy food, exercise, sleep, limited exposure to tobacco and air pollution, and limited experience of excessive stress.”
Young adults have also been dealing with the “expansion of industries that affect public health—processed foods and beverages, prescription drugs and OxyContin, alcohol, combined with this creeping effect of the obesity epidemic,” Stokes said.
He added: “These are the ages, 25 to 44, in which behaviors become entrenched and life course risks start to develop. And if we’re seeing this excess mortality in this generation now, it’s also an indication of what may happen to population health as a whole in decades ahead as this generation ages.”
This information can inform physicians and laboratorians about what diagnostic tests to consider for young people showing symptoms, even if their ages traditionally don’t indicate a chronic condition.
Hospitals in 38 states confirmed patient infections of the dangerous, drug-resistant fungus
Rapidly spreading Candida auris fungus is once again showing up in hospitals throughout the United States, with multiple cases confirmed in Georgia and Florida. Hospital laboratories and pathology departments are encouraged to take advantage of CDC resources to help in the diagnosis of this deadly pathogen.
Candida auris (C. auris) spreads between patients in hospital settings, is resistant to anti-fungal medications, and can cause severe illness, according to the Centers for Disease Control and Prevention (CDC). Tracking data from CDC’s National Notifiable Diseases Surveillance System found 4,514 new clinical cases of C. auris in the US in 2023.
“The number of clinical cases has continued to increase since the first US case was reported in 2016,” said the CDC of past outbreaks of C. auris. “Based on information from a limited number of patients, 30–60% of people with C. auris infections have died. However, many of these people had other serious illnesses that also increased their risk of death.” The fungus has been spreading at a high rate from 2016-2023 with several cases cropping up recently in Georgia.
According to representatives from the Georgia Department of Public Health, “the state has seen over 1,300 cases as of the end of February,” WJCL reported.
The Hill reports a significant recent increase in the spread of the fungus in all but 12 states. Though the number of cases in each state remains small, the overall percentage of increased cases is large and growing.
And a study conducted at Jackson Health System in Miami, Fla., and published in the American Journal of Infection Control, found that “The volumes of clinical cultures with C. auris have rapidly increased, accompanied by an expansion in the sources of infection.”
“If you get infected with this pathogen that’s resistant to any treatment, there’s no treatment we can give you to help combat it. You’re all on your own,” Melissa Nolan, PhD, associate professor of epidemiology and biostatistics at the Arnold School of Public Health, University of South Carolina, told Nexstar. (Photo copyright: University of South Carolina.)
CDC Recommendations
The deadly fungus was first detected in 2016 in US hospitals, and the number of cases in hospital patients has grown every year based on CDC data from 2023. Invasive medical procedures can provide a gateway for C. auris to infect patients, and the immunosuppressed nature of these patients can lead to further complications.
Invasive procedures that could expose a patient to C. auris include the placing of breathing and feeding tubes, and the insertion of vein or urinary catheters.
“We’ve had four people at one time on and off over the past few months, and in years past, it was unusual to have one or even two people with Candida auris in our hospital,” Timothy Connelly, MD, told WJCL about the spread of the fungus at Memorial Health in Savannah, Ga.
Cases have also rapidly increased in Miami according to the Jackson Health System study. The researchers found that, “The volumes of clinical cultures increased every year and infection sources expanded.”
The CDC considers C. auris “an urgent antimicrobial resistance threat” based on the severe risk an infected patient can face. “The rapid rise and geographic spread of cases is concerning and emphasizes the need for continued surveillance, expanded lab capacity, quicker diagnostic tests, and adherence to proven infection prevention and control,” said Meghan Lyman, MD, in a CDC news release.
Fungal Infection is Difficult to Treat and Diagnose
C. auris has been shown to be resistant to antifungal medications, making it an acute threat to ill patients. And since it tends to infect already sick patients, it can be difficult to detect because symptoms of infection can be generic, such as fever or chills.
The fungus is also adept at surviving on hospital surfaces.
“It’s really good at just being, generally speaking, in the environment,” Melissa Nolan, PhD, associate professor of epidemiology and biostatistics at the Arnold School of Public Health, University of South Carolina, told Nexstar. “So, if you have it on a patient’s bed for example, on the railing, and you go to wipe everything down, if in whatever way maybe a couple of pathogens didn’t get cleared, then they’re becoming resistant. And so over time, they can kind of grow and populate in that hospital environment.”
CDC Resources to Help Identify C. auris
C. auris also can be misidentified with other candida species fungi. The CDC recommends identification using a diagnostic device “based on matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF).” The CDC also recommends using supplemental MALDI-TOF databases and molecular methods to help distinguish C. auris from other candida.
Prompt clinical laboratory diagnosis is extremely important to stem outbreaks as they become more frequent in hospital settings. The CDC offers resources for hospital pathology departments to aid in screening and detection.
“I think we need to do a better job of predicting,” Nolan told Nexstar. “Moving forward [we need] more funding to support quality surveillance of these potential infectious strains so that we can know in advance, and we can do a better job of stopping disease spread before it becomes a problem.”
According to the CDC, the fungus typically spreads in hospital settings and is not known to affect healthy people.
Artificial intelligence tools for radiology, clinical laboratory, and pathology diagnostics continue to advance and improve
Researchers in Germany have developed a fully automated, artificial intelligence (AI) tool that improves the diagnosis of prostate cancer. Developed by mediaire, a company that creates AI-based tools for radiologists, the software reduces clinical workloads and could be beneficial in counteracting issues associated with variability in magnetic resonance imaging (MRI) reporting. This is another example of AI’s growth in the clinical diagnostic industry, including clinical laboratory and pathology medicine.
The software, called mdprostate, has received the mandatory certification mark (CE or European Conformity) for products sold within the European Economic Area (EEA). It is now commercially available in those countries and was recently incorporated into the picture archiving and communications system (PACS) of some healthcare organizations and applied to a group of patients who had undergone a multiparametric prostate MRI (mpMRI).
The goal was to compare the overall performance of mdprostate against radiologists who executed the initial interpretations of the mpMRIs, according to Health Imaging.
“Mdprostate is intended to support radiologists by automating time-consuming processes and improving the objectivity of diagnosis through data quantification,” said Tonia Michaely, chief of staff at mediaire, in a news release.
“By providing objective assessments and standardizing lesion detection and classification, AI has the potential to augment radiologists’ performance throughout the PCa [prostate cancer] diagnostic pathway,” Nadine Bayerl, Dr. med., a radiologist with the Institute of Radiology at University Hospital Erlangen and corresponding author of the mediaire study, told Health Imaging. (Photo copyright: University Hospital Erlangen.)
Scoring Cancer Risk
To perform the comparison, a team of researchers applied the AI tool to 123 prostate MRI exams followed by systematic and targeted biopsies. The software was instructed to automatically segment the prostrate, calculate prostate volume, and classify lesions per the Prostate Imaging Reporting and Data System (PI-RADS).
PI-RADS, according to the America College of Radiology, is a reporting method that indicates how likely a lesion is to be clinically significant cancer on a score of one to five:
PI-RADS 1: very low (clinically significant cancer is highly unlikely to be present).
PI-RADS 2: low (clinically significant cancer is unlikely to be present).
PI-RADS 3: intermediate (the presence of clinically significant cancer is equivocal).
PI-RADS 4: high (clinically significant cancer is likely to be present).
PI-RADS 5: very high (clinically significant cancer is highly likely to be present).
For PI-RADS scores greater than two, mdprostate generated 100% sensitivity and dismissed all cancers for lesions that were below that threshold. For PI-RADS scores of four or higher, the AI tool yielded 85.5% sensitivity and specificity of 63.2% for clinically significant cancers.
Deep Learning in Diagnostic Pathway
“In practical terms, these results indicate that when a case falls below the PI-RADS ≥ 2 cutoff, clinicians can rule out malignancy with a high degree of confidence,” the authors explained in the European Journal of Radiology. “This capability is particularly valuable in clinical decision-making, as it allows for the safe avoidance of unnecessary biopsies or further invasive procedures in these patients.”
“Recent advances in deep learning algorithms, facilitated by larger labeled datasets, improved computing hardware, and refined training techniques, have led to several studies highlighting the diagnostic value of deep learning algorithms in prostate imaging,” radiologist Nadine Bayerl, Dr. med., Institute of Radiology at University Hospital Erlangen and corresponding author of the study, told Health Imaging.
The software “demonstrated high diagnostic performance in identifying and grading prostate lesions, with results comparable to those reported in meta-analyses of expert readers using PI-RADS,” the researchers noted in their published study.
“Its ability to standardize evaluations and potentially reduce variability underscores its potential as a valuable adjunct in the prostate cancer diagnostic pathway. The high accuracy of mdprostate, particularly in ruling out prostate cancers, highlights its clinical utility by reducing workload and enhancing patient outcomes,” they concluded.
AI in Clinical Laboratories and Pathology
Dark Daily has frequently covered AI’s expanding role in clinical laboratory testing and pathology diagnostics. At the recent Executive War College, a dozen sessions explored its growth in the industry. During one session, Sam Terese, CEO and president at Alverno Laboratories said, “AI is allowing us to drive our business. It is really resonating that we need to use AI in the future.”
Members who could not attend the 2025 Executive War College can order audio recordings of these valuable sessions by clicking here.
Research could lead to new biomarkers that detect Alzheimer’s much earlier than existing tests and help scientists understand why some people with the disease do not develop dementia
Key biomarkers for detecting the progression of Alzheimer’s disease have typically been based on amyloid-beta (Aβ) plaques. But these plaques show up after the disease has well-progressed and aren’t suited to early detection of the disease.
Now, researchers at the University of Pittsburgh School of Medicine (Pitt) have developed a cerebrospinal fluid (CFS) test that detects changes in tau protein prior to the formation of neurofibrillary tangles (NFTs) that proceed Aβ plaques.
With further research, Pitt’s test could lead to new clinical laboratory biomarkers that help detect the disease earlier and with more accuracy.
“The clumping of tau protein into well-ordered structures, referred to by pathologists as neurofibrillary tangles, is a more defining event for Alzheimer’s disease as it is more strongly associated with the cognitive changes,” as compared to amyloid-beta pathology, according to a Pitt news release.
The researchers showed that their CSF biomarker test worked independent of discovery of brain amyloid deposits and “correlates with severity of cognitive decline” to enable “early-stage disease diagnosis and intervention,” reported Genetic Engineering and Biotechnology News.
“Our test identifies very early stages of tau tangle formation—up to a decade before any tau clumps can show up on a brain scan,” said Thomas Karikari, PhD (above), senior author and assistant professor of psychiatry at Pitt, in a news release. (Photo copyright: University of Pittsburgh.)
Combining Biomarkers May Lead to Better Alzheimer’s Knowledge
The new biomarkers may also work with existing markers that detect amyloid-beta pathology. This could give researchers and healthcare providers a better understanding of the early stages of Alzheimer’s in specific patients.
“Amyloid-beta is a kindling, and tau is a matchstick,” said Thomas Karikari, PhD, senior author and assistant professor of psychiatry at Pitt. Karikari previously researched amyloid-beta.
“A large percentage of people who have brain amyloid-beta deposits will never develop dementia. But once the tau tangles light up on a brain scan, it may be too late to put out the fire, and their cognitive health can quickly deteriorate. Early detection of tangle-prone tau could identify the individuals who are likely to develop Alzheimer’s-associated cognitive decline and could be helped with new generation therapies,” he added.
“P-tau-217 and p-tau-181 are fantastic biomarkers. However, in the early days after we developed these markers, we wondered why they were much more reflective of amyloid pathology than tau pathology,” Karikari told MedPage Today.
“That’s what inspired this work. We believe that methods combining, say, p-tau-217 and p-tau-262 or 356, would provide more complete information on combined early-stage amyloid and tau pathologies in Alzheimer’s disease,” he noted.
Developing the Alzheimer’s Biomarker Test
Karikari and colleagues turned to biochemistry and molecular biology to develop their new test.
Specifically, they emphasized “building blocks of NFTs including oligomers and protomers” which they called “soluble tau assemblies,” Medical News Today explained.
According to the Pitt news release, using autopsied brain tissue, the researchers found:
A core region of the tau protein where NFTs form.
111 amino acids in the region.
New “phosphorylation sites of p-tau-262 and p-tau-356 can inform the status of early-stage tau aggregation that, with an appropriate intervention, could potentially be reversed.”
In other words, p-tau-262 and p-tau-356 “could predict future NFT production, making them potential biomarkers for early disease,” Medical News Today noted.
“Together, our findings inform about the status of early-stage tau aggregation, reveal aggregation-relevant phosphorylation epitopes in tau, and offer a diagnostic biomarker and targeted therapeutic opportunities for Alzheimer’s disease,” the authors wrote in Nature Medicine.
More Research Planned Before Clinical Lab Use
About seven million Americans are affected by Alzheimer’s, according to the Alzheimer’s Association, which expects that number to grow to 13 million by 2050. A cure for the disease does not exist.
More research is needed before the Pitt researchers’ new CSF assay can be used by clinical laboratories. Karikari said the next step is developing blood assays for the biomarkers, MedPage Today reported.
At least a dozen sessions at the 2025 Executive War College explored artificial intelligence use in clinical labs
Although not explicitly stated, it was clear at the 2025 Executive War College on Diagnostics, Clinical Laboratory, and Pathology Management conference that artificial intelligence must be a path forward for labs to control costs in an unfavorable economic environment.
Even though the technology is largely unproven in clinical lab settings, the potential of artificial intelligence (AI) in labs is enough reason for laboratory leaders to explore it further.
“AI is allowing us to drive our business,” said Sam Terese, CEO and president at Alverno Laboratories, during a general session at the Executive War College. “It is really resonating that we need to use AI in the future.”
Clinical laboratory leaders should constantly ask themselves whether there is an AI solution to a problem, advised Sam Terese (above), CEO and president at Alverno Laboratories. Terese spoke at the 2025 Executive War College. (Photo copyright: LabX.)
‘Not a Lot of Trust’ in AI from Patients
Terese acknowledged that patients do not yet feel comfortable with the technology. “When you throw AI and healthcare together, from the public’s perspective, there’s not a lot of trust,” he said.
That said, Alverno is committed to increased use of AI in clinical labs in 2025, including for:
Terese urged laboratory owners and executives to not underestimate how quickly AI adoption could spread within the clinical lab industry. Digital pathology took half a century to evolve into its current state, but “AI took five years. The timeline is moving rapidly,” he observed.
Don’t Move Blindly Forward with AI, Experts Warn
At least a dozen sessions at this week’s Executive War College addressed an aspect of AI in labs.
Lab leaders must advise their staff to use AI with systems or processes that can tolerate mistakes because AI will get things wrong, Cecchini added.
“I treat AI like an eager intern where you have to check everything it does,” he said.
Presenter Ankit Ranjan, PhD, founder of AI company Sample Healthcare, agreed with that sentiment. He suggested that clinical laboratories should consider AI as a copilot until its algorithms can prove to lab staff that conclusions or predictions are accurate. The long game for AI in labs is not to cut a few staff from the budget but instead act as a revenue driver.
“Inserting AI into end-to-end processes is what really addresses problems,” Ranjan said.
Watch for much more coverage about the state of AI in clinical laboratories in upcoming issues of The Dark Report. If you’re not a subscriber, it’s a great time to take a free trial of our business intelligence briefings.
For the past 14 years, healthcare spending as a percentage of US gross domestic product has stayed flat—17.2% in 2010 compared to 17.8% in 2024—according to numbers presented by Robert Michel, editor-in-chief of Dark Daily and founder of the Executive War College.
“This is not auspicious for either the vendor side of the clinical laboratory business or providers,” Michel told attendees during the conference’s opening session.
“Clinical laboratories all must watch for opportunities to earn revenue through new business models,” said Robert Michel (above), editor-in-chief of Dark Daily. Michel spoke during a general session at the 2025 Executive War College on Diagnostics, Clinical Laboratory, and Pathology Management in New Orleans. (Photo copyright: LabX.)
Spending Blockades Will Push New Business Models for Labs
Michel ran through a series of other, similarly grim statistics that put hard numbers on trends that many laboratory executives and owners already suspected:
Half of Americans have less than $500 in a savings account, according to a January 2025 survey by GOBankingRates.com. Those people face tough financial decisions, including whether to postpone healthcare testing and treatment, Michel said.
Growth in Medicare spending by beneficiary generally stopped in 2010 and remained flat at around $12,500 per beneficiary as of 2023.
“Social Security and Medicaid are going to go broke sooner rather than later,” Michel predicted. “Congress has seen this problem and not reacted.”
Such financial challenges will force the need for new clinical laboratory business models. Among the key areas that will inspire these models are diagnostic data and technology, said Sam Terese, CEO and president at Alverno Laboratories, during his general session at the Executive War College.
“It comes back to using data to predict disease,” Terese explained. “If you can prevent someone from getting seriously ill, you will lower the cost of care.”
Terese pointed out the need to effectively use artificial intelligence (AI) to digest the massive amount of clinical data that labs sit on.
Another New Clinical Laboratory Business Model: Closing Care Gaps
Diagnostic laboratories should also be in the business of identifying care gaps among their patients and consumers. One subset to consider is diabetic and prediabetic people.
“Can the lab identify an A1C patient who should have come in to see their doctor based on the test result?” Michel asked. That type of approach raises the value of a lab test beyond just the result it produced, he added.
During another general session, Sonora Quest Laboratories showed how it determines risk stratification for colorectal cancer by using an algorithm that considers a patient’s age, gender, and minimum of two complete blood count test results to flag at-risk people.
“We’re able to get information to physicians to close that care gap,” said Jen Umscheid, senior director of quality, innovation, and performance excellence at Sonora Quest.
The Executive War College continues through Thursday, with an expected attendance of just over 1,000 delegates, speakers, and vendor representatives. Friday’s Dark Daily will explore how AI topics played out among curious attendees.
