By analyzing metabolic signals tied to health and diet, the scientists were able to reconstruct details about ancient climates and landscapes, including temperature, soil conditions, rainfall, and vegetation. The results, published in Nature, point to environments that were significantly warmer and wetter than those found in the same regions today.

Studying metabolites — the molecules produced and used in digestion and other chemical processes in the body — can reveal information about disease, nutrition, and environmental exposure. While metabolomics has become a powerful tool in modern medical research, it has rarely been applied to fossils. Instead, most studies of ancient remains rely on DNA, which mainly helps establish genetic relationships rather than day-to-day biology.

“I’ve always had an interest in metabolism, including the metabolic rate of bone, and wanted to know if it would be possible to apply metabolomics to fossils to study early life. It turns out that bone, including fossilized bone, is filled with metabolites,” said Timothy Bromage, professor of molecular pathobiology at NYU College of Dentistry and affiliated professor in NYU’s Department of Anthropology, who led the international research team.

Why Fossil Bones Can Preserve Chemistry

In recent years, scientists discovered that collagen — the protein that provides structure to bones, skin, and connective tissues — can survive in ancient bones, including dinosaur fossils.

“I thought, if collagen is preserved in a fossil bone, then maybe other biomolecules are protected in the bone microenvironment as well,” said Bromage, who directs the Hard Tissue Research Unit at NYU College of Dentistry.

Bone surfaces are porous and filled with tiny blood vessel networks that exchange oxygen and nutrients with the bloodstream. Bromage proposed that during bone growth, metabolites circulating in blood could become trapped inside microscopic spaces within the bone, where they might remain protected for millions of years.

To test this idea, the team used mass spectrometry, a technique that converts molecules into charged particles for identification. Tests on modern mouse bones revealed nearly 2,200 metabolites. The same approach also allowed researchers to detect collagen proteins in some samples.

Testing Fossils From Early Human Landscapes

The researchers then applied this method to fossilized animal bones dating from 1.3 million to 3 million years ago. These samples came from earlier excavations in Tanzania, Malawi, and South Africa, regions known for early human activity.

The fossils belonged to animals with modern relatives still living near those sites today. The team analyzed bones from rodents (mouse, ground squirrel, gerbil) as well as larger animals, including an antelope, a pig, and an elephant. Thousands of metabolites were identified, many of which closely matched those found in living species.

Health Diet and Disease Written in Bone

Many of the detected metabolites reflected normal biological processes, such as the breakdown of amino acids, carbohydrates, vitamins, and minerals. Some chemical markers were linked to estrogen-related genes, indicating that certain fossilized animals were female.

Other molecules revealed signs of illness. In one striking case, a ground squirrel bone from Olduvai Gorge in Tanzania, dated to about 1.8 million years ago, showed evidence of infection by the parasite that causes sleeping sickness in humans. The disease is caused by Trypanosoma brucei and spread by tsetse flies.

“What we discovered in the bone of the squirrel is a metabolite that is unique to the biology of that parasite, which releases the metabolite into the bloodstream of its host. We also saw the squirrel’s metabolomic anti-inflammatory response, presumably due to the parasite,” said Bromage.

Tracing Ancient Diets and Environments

The chemical evidence also revealed what plants the animals consumed. Although plant metabolite databases are far less complete than those for animals, the researchers identified compounds linked to regional plants such as aloe and asparagus.

“What that means is that, in the case of the squirrel, it nibbled on aloe and took those metabolites into its own bloodstream,” explained Bromage. “Because the environmental conditions of aloe are very specific, we now know more about the temperature, rainfall, soil conditions, and tree canopy, essentially reconstructing the squirrel’s environment. We can build a story around each of the animals.”

These reconstructed habitats align with previous geological and ecological research. For example, Olduvai Gorge Bed in Tanzania has been described as freshwater woodland and grassland, while the Upper Bed reflects drier woodlands and marshy areas. Across all studied locations, the fossil evidence consistently points to climates that were wetter and warmer than today.

“Using metabolic analyses to study fossils may enable us to reconstruct the environment of the prehistoric world with a new level of detail, as though we were field ecologists in a natural environment today,” said Bromage.

Research Team and Support

Additional study authors include Bin Hu, Sher Poudel, Sasan Rabieh, and Shoshana Yakar of NYU College of Dentistry; Thomas Neubert, Christopher Lawrence de Jesus, and Hediye Erdjument-Bromage of NYU Grossman School of Medicine; along with collaborators from institutions in France, Germany, Canada, and the United States. The research was supported by The Leakey Foundation, with additional support for the scanning electron microscope provided by the National Institutes of Health (S10 OD023659 and S10 RR027990).

Time-restricted eating (TRE) is a form of intermittent fasting that limits daily food intake to a window of no more than ten hours, followed by a fasting period of at least 14 hours. The approach has become popular as a simple strategy to support weight management and metabolic health. Animal studies show that TRE can protect rodents from diet-related obesity and metabolic problems. In humans, earlier studies have reported benefits such as improved insulin sensitivity, healthier blood sugar and cholesterol levels, and modest reductions in body weight and body fat. As a result, TRE has been widely viewed as a promising tool for preventing insulin resistance and diabetes.

Conflicting Evidence From Earlier Studies

Despite its popularity, past research on TRE has produced mixed results. Many studies have not been able to determine whether observed health improvements came from shorter eating windows, unintentional calorie reduction, or a combination of both. In addition, most earlier trials did not carefully track calorie intake or control for other factors that could influence metabolic outcomes.

To address these gaps, Prof. Olga Ramich, Head of the Department of Molecular Metabolism and Precision Nutrition at the DIfE and Professor at the Charité — Universitätsmedizin Berlin, designed the ChronoFast trial. The goal was to test whether an eight-hour eating window could improve insulin sensitivity and other metabolic markers when calorie intake was kept constant.

How the ChronoFast Study Was Conducted

The study used a randomized crossover design and included 31 women with overweight or obesity. Each participant followed two different eating schedules for two weeks at a time. One schedule involved early time-restricted eating between 8 a.m. and 4 p.m.(eTRE). The other followed a later schedule from 1 p.m. to 9 p.m. (lTRE). Throughout both phases, participants ate nearly identical meals with the same calorie and nutrient content (isocaloric).

Researchers collected blood samples during four clinic visits and performed oral glucose tolerance tests to assess glucose and fat metabolism. Continuous glucose monitoring tracked blood sugar levels over 24 hours while food intake was recorded in detail. Physical activity was monitored using a motion sensor. In collaboration with Prof. Achim Kramer from the Charité — Universitätsmedizin Berlin, the team also examined changes in the body’s internal clock using isolated blood cells.

Measuring the Body’s Internal Clock

Human biology follows internally generated rhythms that roughly align with the length of a day, which is why they are known as circadian clocks (Latin: circa and dia). These rhythms help regulate nearly every physiological process, including sleep and metabolism. Almost all cells in the body contain their own internal clock, which can be influenced by light, physical activity, and food timing.

To measure individual circadian phases, Prof. Dr. Achim Kramer developed the BodyTime assay. This test requires only a single blood sample and provides an objective snapshot of a person’s internal timing. The ChronoFast study used this method and confirmed that eating schedules can shift internal clocks in humans.

No Metabolic Improvements Found

Despite expectations based on earlier research, the ChronoFast study found no clinically meaningful changes in insulin sensitivity, blood sugar, blood fats, or inflammatory markers after the two-week interventions. “Our results suggest that the health benefits observed in earlier studies were likely due to unintended calorie reduction, rather than the shortened eating period itself,” explains Ramich.

While metabolic measures remained largely unchanged, the timing of meals did affect circadian rhythms. Analysis of blood cells showed that the internal clock shifted by an average of 40 minutes during the late eating schedule compared to the early schedule. Participants following the later eating window also went to bed and woke up later. “The timing of food intake acts as a cue for our biological rhythms — similar to light,” says first author Beeke Peters.

Calories and Individual Timing May Matter Most

The findings highlight the importance of calorie balance in achieving health benefits from intermittent fasting. “Those who want to lose weight or improve their metabolism should pay attention not only to the clock, but also to their energy balance,” Ramich concludes.

Future research will need to explore whether combining time-restricted eating with reduced calorie intake produces stronger benefits. Scientists also aim to better understand how individual factors, including chronotype and genetics, may influence how people respond to different eating schedules.

The clogged pathways are known as “enlarged perivascular spaces,” and the findings suggest they could serve as an early warning signal for Alzheimer’s, the most common form of dementia.

“Since these brain anomalies can be visually identified on routine magnetic resonance imaging (MRI) scans performed to evaluate cognitive decline, identifying them could complement existing methods to detect Alzheimer’s earlier, without having to do and pay for additional tests,” said Associate Professor Nagaendran Kandiah from NTU’s Lee Kong Chian School of Medicine (LKCMedicine), who led the study.

Justin Ong, a fifth-year LKCMedicine student and the study’s first author, emphasized the importance of early detection. He noted that identifying Alzheimer’s sooner gives doctors more time to intervene and potentially slow the progression of symptoms such as memory loss, reduced thinking speed, and mood changes. The research was conducted as part of LKCMedicine’s Scholarly Project module in the School’s Bachelor of Medicine and Bachelor of Surgery programme.

Why Studying Asian Populations Matters

The study stands out because it focuses on Asian populations, an area that has been underrepresented in Alzheimer’s research. Most existing studies have concentrated on Caucasian participants, which may limit how broadly their findings apply.

The NTU team examined nearly 1,000 people in Singapore from different ethnic backgrounds that reflect the country’s population. Participants included individuals with normal cognitive function as well as those experiencing mild thinking difficulties.

Research has shown that dementia does not affect all ethnic groups in the same way, making region specific studies essential.

“For example, among Caucasians with dementia, past studies show that the prevalence of a major risk gene, apolipoprotein E4, linked to Alzheimer’s is around 50 to 60 percent. But among Singapore dementia patients, it is less than 20 percent,” said Assoc Prof Kandiah, who is also Director of the Dementia Research Centre (Singapore) in LKCMedicine. Because of these differences, findings in one population may not directly apply to another.

How the Brain Clears Toxic Waste

Inside the brain, blood vessels are surrounded by small channels called perivascular spaces. These spaces help drain toxic waste products, including beta amyloid and tau proteins, which are found in high levels in people with Alzheimer’s disease.

When the brain’s waste removal system becomes less efficient, these spaces can enlarge and become visible on MRI scans. Until now, it was unclear whether this change was directly linked to dementia, particularly Alzheimer’s disease.

To answer this question, the NTU researchers compared enlarged perivascular spaces with multiple established indicators of Alzheimer’s. They also examined how these clogged drainage pathways relate to well known disease markers such as beta amyloid buildup and damage to the brain’s white matter, the network of nerve fibers that connects different brain regions.

Comparing Healthy Brains and Early Cognitive Decline

The study included nearly 350 participants with normal thinking abilities, including memory, reasoning, decision making, and focus. The remaining participants showed signs of early cognitive decline, including mild cognitive impairment, a condition that often precedes dementia.

Previous research has shown that people with mild cognitive impairment face a higher risk of developing Alzheimer’s disease or vascular dementia, which is caused by reduced blood flow to the brain.

After analyzing MRI scans, the researchers found that participants with mild cognitive impairment were more likely to have enlarged perivascular spaces than those with no cognitive problems.

Blood Markers Strengthen the Link

In addition to brain scans, the scientists measured seven Alzheimer’s related biochemicals in participants’ blood, including beta amyloid and tau proteins. Elevated levels of these substances are considered warning signs of Alzheimer’s disease.

Enlarged perivascular spaces were linked to four of the seven biochemical measurements. This suggests that people with clogged brain drains are more likely to have increased amyloid plaques, tau tangles, and damage to brain cells, placing them at greater risk of developing Alzheimer’s.

The researchers also looked at white matter damage, a widely used indicator of Alzheimer’s, and found it was associated with six of the seven blood markers. However, further analysis revealed something unexpected.

Among participants with mild cognitive impairment, the connection between Alzheimer’s related biochemicals and enlarged perivascular spaces was stronger than the connection with white matter damage. This finding points to clogged brain drainage as a particularly early signal of Alzheimer’s disease.

Implications for Diagnosis and Treatment

These insights may help doctors make more informed decisions about early treatment strategies, potentially slowing disease progression before lasting brain damage occurs.

“The findings carry substantial clinical implications,” said Assoc Prof Kandiah. “Although white matter damage is more widely used in clinical practice to evaluate for dementia, as it is easily recognised on MRI scans, our results suggest that enlarged perivascular spaces may hold unique value in detecting early signs of Alzheimer’s disease.”

Dr. Rachel Cheong Chin Yee, a Senior Consultant and Deputy Head at Khoo Teck Puat Hospital’s Department of Geriatric Medicine, said the study highlights the role of small blood vessel changes in Alzheimer’s development.

“These findings are significant because they suggest that brain scans showing enlarged perivascular spaces could potentially help identify people at higher risk of Alzheimer’s disease, even before symptoms appear,” said Dr. Cheong, who was not involved in the research.

Rethinking Brain Vessel Disease and Alzheimer’s

Dr. Chong Yao Feng, a Consultant at the National University Hospital’s Division of Neurology who was also not involved in the study, noted that cerebrovascular diseases and Alzheimer’s disease have traditionally been viewed as separate conditions.

“The study’s findings are intriguing as they demonstrate that both diseases do interact in a synergistic manner,” said Dr. Chong, who is also a Clinical Assistant Professor at the National University of Singapore’s Yong Loo Lin School of Medicine.

As a result, doctors reviewing MRI scans should be cautious about assuming cognitive symptoms are caused only by blood vessel problems when markers such as enlarged perivascular spaces are present. These features may also signal a higher risk of Alzheimer’s disease.

“Doctors will then have to use their clinical judgement of the patient’s scan and symptoms, as well as discuss with the patient, to determine if more checks are needed to confirm whether a patient has Alzheimer’s disease or not,” said Dr. Chong.

What Comes Next

The NTU research team plans to track participants over time to determine how many eventually develop Alzheimer’s dementia. This follow up will help confirm whether enlarged perivascular spaces can reliably predict progression to dementia.

If future studies in other populations reach similar conclusions, identifying clogged brain drains on MRI scans could become a routine tool for detecting Alzheimer’s risk much earlier than is currently possible.