Precision Medication Brings Genetic Insights to Weight problems Remedy

Obesity research has come a long way since the start of the 21^st century. From a disease whose treatment mostly relied upon lifestyle interventions and surgery, obesity now has many available treatment options.

“The current days are very exciting because it’s the first time that we have very potent and safe drugs,” said Ricardo Cohen, a bariatric surgeon at the Hospital Alemão Oswaldo Cruz. Indeed, the introduction of drugs like glucagon-like peptide 1 (GLP1) receptor agonists and other appetite-inhibiting medications have helped people with obesity increase their weight loss more effectively than only changing their diet and exercise habits.^1-3

Similarly, for people who previously would have had to undergo surgery as an initial treatment or follow up after weight regain, Cohen said he can now offer different options. “It’s a very exciting and very rewarding experience to go from where to put a stitch to choose which is the best treatment for our patients,” he said.

While these advances have drastically improved obesity treatment, physicians have found that there is considerable variability among how well different medicines work for different patients. With many factors contributing to obesity, there is likely not a universal solution for the condition.

To improve obesity intervention outcomes, many researchers are taking a more precision medicine-based approach by conducting genome-wide association studies (GWASs) and using machine learning tools to find the best treatment for each patient.

GWASs Reveal Complexity of Obesity

Obesity is a complex disease that causes excess weight gain that leads to a variety of negative health outcomes.⁴ For years, physicians and scientists erroneously viewed obesity as a consequence of poor dietary decisions, such as overeating or choosing predominantly high-caloric foods.

However, research over the last decade points to genetic and environmental factors influencing weight gain, and around the world, obesity rates are increasing.⁵ “It suggests that there isn’t just some sort of sudden societal failure of willpower at play here,” said Deirdre Tobias, an epidemiologist and obesity scientist at the Brigham and Women’s Hospital and Harvard University.

Indeed, environmental changes like the increased availability and energy density of foods along with lifestyles leading to less sleep, less physical activity, and greater stress can promote obesity. Yet at the same time, not all people respond in the same way to these external influences; in fact, GWASs point to several genes influencing these outcomes.⁶ Furthermore, many factors that regulate a person’s energy intake and expenditure are rooted in physiology, making long-term weight loss challenging.⁷

Despite this, Tobias said that obesity prevention strategies have taken a “one-size-fits-all” approach directed at individuals, such as providing education on diet and encouraging more physical activity. These interventions have seen wide variability in their success in helping patients lose weight, especially over time.^5,8-10 Similarly, the long-term success of both anti-obesity medications and surgery varies greatly in helping people maintain weight loss.^10-12

To better understand obesity and begin to address these outstanding intervention questions, researchers dove into the genetics influencing this condition. Over the last decade, scientists have identified several gene variants related to obesity using GWASs. Ruth Loos, a human geneticist at the University of Copenhagen, leads a team that used these studies to find links between different physiological metrics and health outcomes in the context of obesity. One of the most common metrics used in obesity research is body mass index (BMI), or a ratio of a person’s height to their weight, with higher values correlating to greater weight and, thus, obesity.

Using GWASs focused on BMI, Loos’s team and other researchers identified variants in genes involved in brain functions such as signaling eating behaviors and neurodevelopment, “suggesting that, okay maybe obesity is really a disease of the brain where some people feel more hungry or less satiated or feel a need for rewards,” Loos said.

While BMI is a useful tool for studying populations, physicians like Cohen find it less helpful for assessing a specific patient, since many other factors can influence an individual’s health and needs. Additionally, most BMI metrics are based on those from adult white males in Western populations. These often poorly predict obesity risks in different ethnic groups, who can experience obesity complications at lower or higher BMIs than in the white population, and in children.^13-15

Although the metric has led to valuable findings, on its own, Loos said, “Body mass index is a phenotype that’s actually pretty remote from our genome or from our biology.” She pointed out that it does not take into account how fat is deposited around the body, which is important in determining health risk.

To get better insight into the role of fat distribution in obesity, her team also performed GWASs in this context. “These variants that determine fat distribution, they don’t point to the brain. They actually point to the peripheral biologies,” she said. For example, her team identified genes involved in the regulation of adipose cells, leptin production, and insulin sensitivity.^16,17

From these associations, Loos’s team created gene risk scores (GRSs) to inform the likelihood that an individual may develop obesity and the severity of it based on their gene variants.^18,19 However, “The challenge is not finding these variants,” she said, but “the real bottleneck in our research—and that’s not only just in obesity genetics but also in diabetes and cardiovascular disease genetics—is what do these variants mean?”

Indeed, while GWASs have demonstrated thousands of variants correlated with obesity, most of these studies do not provide insights into how a variant affects gene expression or the encoded protein or what cell types are affected. Another hurdle in answering this question is that many variants lie in non-coding regions of the genome, so researchers must then determine with what gene that particular variant interacts and how it influences that gene or set of genes.^20,21

Loos said that making these connections involves creating gene maps and, though her group isn’t making them, they can translate these variants to their biological function.^22,23 For example, researchers used single-cell profiling and spatial transcriptomics to explore how gene variants impact gene expression across cells and in tissues in obesity.^24,25

However, Loos and her team also observed that, in some individuals, BMI did not appear to correlate with expected obesity comorbidities, like type 2 diabetes or cardiovascular disease. Instead, variants in genes that regulated adipocyte function, inflammation, and insulin influenced metabolic activity separately from an individual’s BMI and pointed to roles in individuals’ fat cell responses to excess energy.^26,27

“Some fat cells just cannot expand and, if there’s too much fat going in, they get inflamed. They almost burst in a way, like they cannot cope, whereas other people have fat cells that really cope well; they expand easily, [and] they remain fairly healthy,” Loos said.

Based on this finding, her team proposed a new subtype of obesity that they termed “metabolically healthy obesity” in which people with high BMIs don’t present with the same degree of negative health outcomes as others with similar high BMIs.²⁷ This group of people, Loos said, may benefit from different intervention strategies.

Studying the differences in these fat cell and metabolic responses and the genes that regulate them could also lead to new obesity treatments and insight into disease mechanisms. Recently, Loos and her team identified more than 200 loci that increased a person’s adiposity but were associated with lower risk of eight cardiometabolic conditions.²⁸ They created a genetic score using these loci to determine a person’s risk for obesity without these comorbidities. Loos said that, with more research, these scores could be helpful in identifying patients with higher risk for negative health outcomes and guiding treatment strategies.

Obesity Phenotypes Help Tailor Treatment Approaches

Another outstanding question in obesity medicine is how genetics influence the response to interventions. Andres Acosta, an obesity biologist at the Mayo Clinic, has seen this frustrating reality firsthand. “I’m in the trenches caring for patients that are suffering from this epidemic. It’s very difficult to actually tell a patient ‘Here’s how you’re going to do with this intervention,’ whatever it is: a diet, medication, device, or surgery,” Acosta said. These uncertainties, he said, leave patients with questions about whether a specific treatment is worth their time and money. In response to this challenge, researchers like Acosta are exploring ways to improve obesity treatment through precision medicine.

It was key to tackle this heterogeneity of response in obesity to really try to bring the right intervention for the right patient.

—Andres Acosta, Mayo Clinic and Phenomix Sciences

For example, he and his group observed differences between people with lower satiation—or signals of fullness—and faster gastric emptying in people with obesity; this led the researchers to propose four different phenotypes: “Hungry Brain,” “Hungry Gut,” “Emotional Hunger,” and “Slow Burn.” Individuals in the Hungry Brain group experience abnormal satiation and often ingest more calories before their brain signals to them that they are full.²⁹ Meanwhile, in the Hungry Gut phenotype, people have faster gastric emptying, leading to them feeling hungry between meals more often.³⁰ People with Emotional Hunger eat in response to their emotions, while the Slow Burn phenotype arises from decreased resting metabolic activity in people that causes weight gain.^31-34

In observational and cross-sectional studies using these phenotypes to inform treatment—either medication or a diet regimen—Acosta and his team showed that such phenotype-guided interventions improved weight loss.^35,36 For example, recommendations for people with the Hungry Brain phenotype included appetite suppressing medication or, in a dietary intervention study, time-restricted eating on a low-calorie diet. In contrast, people with Hungry Gut obesity types were assigned a protein supplement before each of their three daily meals or, in medication trials, received a GLP-1 agonist. Similarly, tailoring treatments for Emotional Hunger and Slow Burn, either with diet or medication, improved patient outcomes.

Although researchers have made great strides in identifying genes and traits that influence obesity, Acosta said that the translation of these findings to the clinic has been slow. After identifying these obesity phenotypes, Acosta cofounded Phenomix Sciences to address this translational issue. “It was key to tackle this heterogeneity of response in obesity to really try to bring the right intervention for the right patient,” he said. By doing this, he hoped that he and his team could develop approaches that would also address the underlying cause of the person’s obesity.

With the help of his company, Acosta’s group explored how factors like sex, age, and weight affect variability in satiation, which promotes the feeling of fullness to end a meal. After seeing sex and weight emerge as top non-genetic predictors of the number of calories needed for an individual to experience satiation, the team examined genetic influences of satiation and developed a test for this metric based off of their identified genes and their associated risk of altered satiation.³⁷ To develop this GRS, they used machine learning to study 41 genes previously identified to be involved in obesity, specifically an axis connecting the gut, brain, and adipose tissue. They narrowed their investigation down to 10 genes that predicted patient responses to two anti-obesity medications.

“We were able to show that your biology matters, and if you build a genetic surrogate against your biology, you can identify whether you’re going to respond to two different medications and not in an insignificant way,” Acosta said. Recently, his team presented preliminary findings showing similar predictive results with another GLP-1 agonist.³⁸ “That’s fascinating because now we can take the biology, the trait, the genetic biomarker, or surrogate biomarker and show that you can actually predict response to obesity medications, really bringing precision medicine to the forefront of obesity,” he said.

Acosta’s team and Phenomix Sciences also shared data at a conference in 2025 that demonstrated that this test works in ethnically diverse populations.³⁹ Separately, they showed that they could use the same test to predict individuals who will respond to a GLP-1 medication after bariatric surgery.⁴⁰ Additionally, they showed that the same GRS could predict which individuals were most at risk for obesity-related disorders, like type 2 diabetes in a large cohort of individuals.⁴¹ “[It’s] very interesting to see how this type of gene risk score can also help us to predict who will develop more obesity comorbidities compared to others,” Acosta said.

Obesity Biomarkers Could Improve Precision Medicine Approaches

With the help of genetic studies and clinical trials, precision medicine for obesity is taking shape. “There, I think, is a lot of optimization of treatment allocation when you can match people’s biology to whatever intervention works best for them,” Tobias said.

It’s a very exciting and very rewarding experience to go from where to put a stitch to choose which is the best treatment for our patients

—Ricardo Cohen, Hospital Alemão Oswaldo Cruz

Indeed, these proposed obesity phenotypes also help Cohen, whose patients are normally interested in surgery given his specialty, approach treatment options. “Phenotyping obesity helps immensely [in how we] speak with the patients and show them how we will intervene in them with another drug, surgery plus drug, [or] surgery alone,” he said.

Cohen said that precision medicine for obesity should mirror its use in cancer treatment. “When we see a person with obesity in front of us at the office, we need to think exactly as the oncologist would,” he said. This includes considering all of the characteristics of a specific patient, such as their risk factors and goals as well as their obesity phenotype and working with that person to achieve those goals.

But, Cohen added, “[Precision medicine for obesity is] still in its infancy.” As many of these obesity types are new to the field, not all physicians are aware of them. Additionally, Cohen said that providers don’t have precise markers to determine the type of obesity that a person has. “So, we depend on the clinical judgment, which is good as well, but less precise than the number,” he said.

Whereas chronic diseases like diabetes can use a threshold of a circulating factor, hemoglobin A1c, “We don’t have this in obesity,” he said. “We have different phenotypes. We have clusters of metabolic disease. We have clusters of osteoarthritis [and] impaired daily activities. We have sleep apnea. We have raised blood pressure. We still don’t have a marker.”

He added that a biomarker could improve diagnosis of a person’s obesity phenotype, as opposed to the current method of relying on clinical judgement alone, and be helpful to screen people to improve prevention.

Meanwhile, in addition to continuing to explore the efficacy of their gene test, Acosta said that, for a number of people with obesity, “we just don’t know what is their underlying biology or what’s driving their obesity, clearly telling us that a lot more research is needed.” This is another area that he is interested in exploring.

In her role as an epidemiologist, Tobias is also interested in research exploring anecdotes of patients taking GLP-1s who no longer have the same food behaviors despite being in their same environment. “I do think that could be useful because we could leverage it again to understand more about how to prevent that in the first place. Maybe there are certain triggers about the environment we’re in that can be modified. So, we don’t even have to be on a drug because now we’ve removed whatever that was, and now we can prevent obesity to begin with,” she said.

However, she added that most proposed tests are not reliable enough yet to be used in the clinic. “Although this is a compelling direction for tailoring weight loss to the individual, research has a long way to go.”