A DNA test that tells a person exactly what to eat sounds like the future of weight loss. The pitch is simple and tempting: send in a saliva sample, learn which foods suit a person’s genes, and follow a diet written for their body alone instead of a generic plan.
A new randomized trial has now put that pitch to a direct test, and the result is not what the sellers would have hoped for. Over 12 months, adults who received diet and exercise advice tailored to their genes lost no more weight than adults who received standard, guideline-based counseling. Both groups lost weight. The gene-based group simply did not lose more.
The study, known as the MyGeneMyDiet Study, was carried out by researchers at the Food and Nutrition Research Institute in the Philippines and published in the journal Genes and Nutrition in 2026. Genetic tailoring was feasible, the researchers wrote, but it provided no clinical advantage over standard counseling.
The promise in one paragraph
The idea behind a gene-based diet is that small differences in DNA change how a person’s body handles food and exercise. Three gene variants sat at the center of this trial. The FTO variant has been linked in earlier research with appetite and body weight. The TCF7L2 variant has been linked with how the body handles fat and blood sugar. The UCP1 variant has been linked with how the body burns energy as heat. The theory is that a person carrying one of these variants should eat or move differently from a person who does not, and that a diet built around those variants should therefore work better than a standard plan.
Genes do influence body weight, and twin and family studies have shown for decades that weight runs partly in families. What the theory skips over is scale. Each of these variants, on its own, shifts weight by a small amount. A diet built on three of them is still built on small effects, and the trial set out to find whether those small effects add up to a real advantage on the bathroom scale. They did not.
Two plans, side by side
The trial was designed so that the only difference between the two groups was the genetic tailoring. Both groups received the same amount of attention, the same counseling schedule, and the same pre- and post-test genetic counseling. The comparison was clean: tailored advice against standard advice, nothing else.
| Gene-based group (MyGeneMyDiet) | Standard group (control) | |
|---|---|---|
| Who it was for | Adults carrying at least one of three risk variants: FTO, TCF7L2, or UCP1 | Same eligibility, same gene testing and counseling |
| Exercise advice | FTO carriers: 30 to 60 minutes of moderate aerobic exercise every day | At least 150 minutes of moderate exercise per week |
| Calorie advice | UCP1 carriers: daily calories cut by 150 kcal below estimated baseline | A daily deficit of about 500 kcal |
| Fat advice | TCF7L2 carriers: dietary fat limited to 15 to 20 percent of daily energy | Fat at 25 to 30 percent of daily energy, with 55 to 65 percent carbohydrate and 10 to 15 percent protein |
| Food model | Advice built from each person’s variant combination | Philippine Nutrition Practice Guidelines and the Pinggang Pinoy food model, with World Health Organization activity guidance |
| Counseling | Online sessions at months 0, 3, 6, and 12 | Online sessions at months 0, 3, 6, and 12 |
| Structure | Active phase for 6 months, then a self-directed phase to month 12 | Active phase for 6 months, then a self-directed phase to month 12 |
The table makes the design easy to see. The gene-based plan was not a different world of eating. It was the same kind of advice — move more, eat less, watch the balance of nutrients — with the numbers adjusted according to three gene variants. The standard plan, meanwhile, was not a careless control. It was built on national nutrition guidelines and a recognized food model, with a clear calorie target and a clear exercise target. That is the right comparison for a test sold to the public.
The trial, month by month
Recruitment ran across Metro Manila and surrounding provinces between December 2021 and November 2022. The researchers enrolled Filipino adults aged 19 to 59 with a body mass index between 25 and 40, the range that covers overweight and obesity. To take part, a person had to carry at least one risk variant among the three genes, so every participant had a genetic result worth tailoring advice to.
Sixty-six people were randomized, and 52 began the intervention: 29 in the gene-based group and 23 in the standard group. The researchers recruited extra people to allow for dropouts. In the first six months, the active phase, both groups received regular online counseling and followed their assigned plan. Forty-nine people completed assessments at month 6.
Then came the part that makes this trial more honest than most diet studies. From month 6 to month 12, the counseling stopped and participants lived under self-directed conditions, with outcome assessments at months 9 and 12. By month 12, 27 people completed follow-up: 15 in the gene-based group and 12 in the standard group. Sensitivity analyses confirmed the main findings despite the dropouts.
The timeline matters because it answers the objection that a gene-based diet might need more time to show its edge. Twelve months is a long time in weight-loss research. It did not pull ahead at month 6, and it had not pulled ahead by month 12.
The numbers that matter
The headline numbers are strikingly similar. The gene-based group started at an average weight of 76.7 kilograms and finished the year at 71.7 kilograms, a reduction of 6.5 percent. The standard group started at 76.6 kilograms and also finished at 71.7 kilograms, a reduction of 6.4 percent. At month 6, the adjusted difference between the groups was 0.36 kilograms in favor of the gene-based group, with a confidence interval running from 1.77 kilograms in its favor to 1.04 kilograms against it. That range includes zero, which means the difference is not statistically significant.
The pattern repeated across every measure. Body mass index differed between the groups by 0.11 at month 6, with a confidence interval from 0.51 lower to 0.73 higher, again not significant. Waist circumference differed by 0.27 centimeters, with a confidence interval from 2.23 centimeters lower to 1.69 centimeters higher, again not significant. Body fat percentage differed by 0.92 percentage points, with a confidence interval from 0.86 lower to 1.05 higher, again not significant. Blood markers, including glycated hemoglobin and blood lipids, showed no sustained difference between the groups either. The full account of the trial, including the researchers’ own reading of these figures, is set out in the Bioengineer report on the gene-based diet trial, which is the source for the figures quoted in this article.
Within the gene-based group alone, reported daily intake fell by month 12: energy down by 461.4 kilocalories, protein down by 12.1 grams, carbohydrates down by 46.4 grams, and fat down by 22.0 grams. But the comparison between the groups on dietary intake was not statistically significant, and the control group also reduced its intake. A change inside one group is not the same as an advantage over the other group, and the trial’s conclusion rests on the comparison.
Why both groups lost the same weight
The most useful question the trial raises is not why the gene-based plan failed, but why the standard plan did so well. The answer is that standard weight-loss counseling, done properly, is already a strong intervention. A daily deficit of about 500 kilocalories, regular moderate exercise, and a clear structure for meals will produce weight loss in most adults who follow it, and the control group followed it well enough to lose 6.4 percent of body weight over a year.
That sets a high bar for any new approach. A gene-based plan does not need to beat doing nothing. It needs to beat a competent standard plan, and the gap it must open is measured in kilograms, not in good intentions. The three gene variants in this trial carry modest individual effects, and tailoring advice to them changed the numbers on the plan — 150 fewer kilocalories here, a tighter fat range there — without changing the outcome on the scale.
There is a second, quieter reason the two groups look alike. Adherence was low in both arms, with fewer than half of participants consistently meeting their targets. A plan only works when it is followed, and a genetically tailored plan that is followed half the time will not beat a standard plan that is also followed half the time.
The calorie targets in this trial, whether the 150-kilocalorie cut for UCP1 carriers or the 500-kilocalorie deficit in the standard group, are both set against total daily energy expenditure, explained in plain terms in what TDEE is. A guide to TDEE versus BMR explains how that total differs from the smaller basal figure.
What this means for DNA diet tests
The trial lands in the middle of a growing market. Direct-to-consumer DNA tests for weight loss are sold online and in pharmacies, promising to match a person’s diet to their genes. The marketing often leans on the language of precision medicine, as if a genetic report were a prescription.
This trial does not prove that every such test is useless. It tested three variants, one decision framework, and one population, and its authors are careful about that scope. What it does is remove the easiest claim a seller can make: that gene-based advice, as a general approach, beats standard advice. In a year-long randomized trial, it did not. Two earlier lines of evidence point the same way. The DIETFITS trial, which randomized more than 600 people to low-fat or low-carbohydrate diets, found that genetic variants did not predict who lost more weight on which diet. A review of eight trials found only about half a kilogram more weight loss on average, with the lower end of the range close to zero and no significant difference in body fat.
It means the current generation of gene-based diet advice is built on variants whose individual effects are too small to steer a diet better than standard guidance already does. A person considering a DNA diet test should ask what the test changes in practice. If the answer is a modest adjustment to calories or fat that a dietitian would have recommended anyway, the test is charging for information that does not change the plan.
There is also a practical question that buyers rarely ask: what happens after the report arrives. In this trial, participants received counseling at four points across the year, and adherence still stayed below half. The hard part of weight loss is sticking to a plan, and a genetic label does not make that easier.
The limits, stated plainly
Every trial has limits, and this one has several that the researchers state openly.
First, the size. Fifty-two people began the intervention and only 27 finished the year, so small differences could have been missed.
Second, the population. The participants were Filipino adults with overweight or obesity, recruited in one metropolitan area, who volunteered for a structured trial with regular counseling. The results may not extend to other populations.
Third, the genes. Only three variants were used, and each carries a modest individual effect. A different set of variants might behave differently; the trial tests one version of gene-based advice, not every version.
Fourth, the measurement. Diet and activity were self-reported, and self-reports are unreliable in both directions. The pandemic disrupted the trial, and baseline differences between the groups were adjusted for in the analysis.
Fifth, the comparison itself. The standard group received a strong, guideline-based plan, and that is the comparison the gene-based plan lost.
In short: over 12 months, with the same counseling on both sides, advice tailored to three gene variants produced no advantage over standard weight-loss counseling. Both groups lost a meaningful amount of weight. The genes did not decide who lost more.
Frequently asked questions
Did the gene-based diet fail completely?
No, and that distinction matters. People in the gene-based group lost an average of 6.5 percent of their body weight over 12 months. The plan worked. It simply did not work better than the standard plan, which produced a 6.4 percent loss. The trial shows that gene-based advice adds no advantage, not that dieting itself fails.
What were the three genes in the trial?
The trial used three variants: FTO rs9939609 A, linked with appetite and body weight; TCF7L2 rs7903146 T, linked with fat handling and blood sugar; and UCP1 rs1800592 G, linked with burning energy as heat. Each variant changed one part of the advice: exercise for FTO carriers, a calorie cut for UCP1 carriers, and a fat limit for TCF7L2 carriers. Carriers of more than one variant received combined guidance.
How many people were in the study, and how long did it last?
Sixty-six people were randomized and 52 began the intervention, with 29 in the gene-based group and 23 in the standard group. The trial ran for 12 months, with an active counseling phase for the first 6 months and a self-directed phase for the second 6 months. Forty-nine people completed assessments at month 6, and 27 completed follow-up at month 12.
Does this mean DNA diet tests are a waste of money?
Not necessarily, but it weakens the central claim behind them. This trial found no advantage for gene-based advice over standard counseling, and a review of eight trials found only about half a kilogram of extra weight loss on average, with no significant difference in body fat. A test may still help if it identifies a specific, well-evidenced genetic condition.
What should a person do instead of buying a DNA diet test?
The evidence in this trial points to the basics: a clear daily calorie target, regular moderate exercise, and support to keep going. A step-by-step guide to how to calculate TDEE shows how to work out daily energy needs, which is the number both plans in this trial were built around. Taking that number into a conversation with a clinician or dietitian is a better use of time than a genetic report that changes little in practice.
The personalization that matters most may not come from a gene test. It comes from a plan a person can follow, support that keeps them going, and a calorie target built on actual daily energy use. Readers who want that starting number can work it out with the TDEE calculator, and take the result, along with this study, into a conversation with a clinician or dietitian.
