"57% of participants reported a health improvement." A figure like this, taken from a study published in 2026, would make a great headline for any DNA nutrition company, ours included. Before trusting it, a few questions are worth asking. Here are seven, each illustrated with a real study from the past year.
1. Was there a comparison group?
The study behind that 57%, published in BMC Nutrition in April 2026 by a company that sells DNA-based nutrition plans, compared customers who said they had changed their habits with customers who had not: 57% versus 12% reported an improvement. But there was no group that changed its habits without genetic advice. So the result shows that changing habits helps, not that the genetics made the difference. Only a randomized trial with a control group can separate the two.
2. Who answered?
Out of 47,299 eligible customers, 874 answered the survey, under 2%. People who are happy with a service are more likely to reply, which tends to make results look better. In the same study 77% of respondents were white women, so the findings may not apply to everyone.
3. How was the outcome measured?
A reported improvement or a self-reported weight is weaker evidence than a measured weight or a blood test. In the vitamin D trial published in July 2026, the outcome was the blood level of 25(OH)D measured after 12 months: much harder to argue with.
4. Who wrote and funded it?
All the authors of the 57% study worked for, or had a financial interest in, the company whose service was studied, and they declared it, which is good practice. A conflict of interest does not make a result false, but it means you should look for independent confirmation. The same rule applies to us: we sell reports, so judge our claims by the independent studies we cite.
5. How big is the effect, and how precise?
A September 2026 meta-analysis found that genetics-based advice led to 0.52 kg more weight loss than standard advice. Statistically significant, yes, but the confidence interval runs from about 1 kg down to almost nothing. Significant means unlikely to be pure chance; it does not mean large.
6. Did the gene change the outcome, or just the advice?
A randomized trial in slow caffeine metabolizers gave everyone the same advice and told only some of them their CYP1A2 genotype. Both groups cut their caffeine by more than half, and knowing the gene added nothing measurable. To show that genetic information helps, a study has to isolate it.
7. Who was studied?
The vitamin D trial included people over 70, mostly white, in northwest England. Its results are solid for that group but may differ for a 30-year-old or for other populations. Always check whether the participants look like you.
A quick checklist
- Randomized, with a control group?
- Enough participants, and few lost along the way?
- Measured outcomes rather than impressions?
- Independent authors, or declared interests and independent replication?
- An effect size with its confidence interval, not just a percentage?
- Participants similar to you?
No single study ticks every box. What counts is the overall weight of evidence, which is why our reports list the scientific references behind each finding, so you can check them yourself. See the reports.
This article is educational and is not medical advice.