Genetic tests and nutrigenomics — what your DNA can and cannot tell you

Today you can order a home DNA test for the price of a nice dinner. Within days you know how you process coffee, whether you carry a “bad” MTHFR variant and what your risk of Alzheimer’s disease is. Most of our DNA is the same in all people, but small differences in individual letters of the genetic code – variants – influence how the body handles food, medicines and stress. The study of how genes and diet affect each other is called nutrigenomics. You do need to know, however, what a test can really tell you and what is already marketing.

Consumer versus clinical test

Most tests you order online do not read the whole genome. A genotyping chip checks hundreds of thousands of selected positions in the DNA. That is enough for ancestry and common variants, but often not for the rare mutations behind serious inherited diseases. A clinical genetic test, by contrast, is ordered by a doctor or geneticist, performed by an accredited laboratory, and the result always comes with professional interpretation and genetic counselling.

A clinical laboratory in the USA reviewed 49 samples in which the raw data of a consumer test showed a serious variant. As many as 40 % of these findings were false positives, and some variants labelled “increased risk” were classified by the laboratory as benign.

MTHFR: the most overrated gene on the internet

The MTHFR gene encodes an enzyme involved in processing folic acid and in methylation – a process important, for example, in breaking down homocysteine. Common variants of this gene slightly slow the enzyme down and, especially with low folate intake, can raise homocysteine.

Professional guidelines are cautious. The American College of Medical Genetics and Genomics (ACMG) states that meta-analyses have not confirmed an association between MTHFR variants and the risk of venous thrombosis, nor between mildly elevated homocysteine and coronary heart disease. According to the ACMG, MTHFR testing has minimal clinical utility and should not be part of routine evaluation. It is more practical to measure homocysteine, folate and vitamin B12 in your blood. These show what is happening in your body right now.

APOE4: a gene that deserves respect

The APOE gene controls the transport of fats and cholesterol and exists in three main variants: ε2, ε3 and ε4. A large meta-analysis of almost 6,000 patients with Alzheimer’s disease and more than 8,600 controls showed that in people of European descent one copy of ε4 increases the odds of the disease about 3.2-fold and two copies about 14.9-fold compared with the most common combination, ε3/ε3.

A 2024 study in Nature Medicine went even further. In people with two copies of APOE4 it found changes typical of Alzheimer’s disease almost always. By age 65 nearly all had abnormal amyloid levels in their cerebrospinal fluid, 75 % had a positive amyloid brain scan, and symptoms began on average at 65. The authors therefore consider it a distinct genetic form of the disease.

At the same time, even APOE4 is not a verdict. In the Finnish FINGER study, a two-year programme of diet, exercise, memory training and blood pressure control improved cognitive results, and in ε4 carriers the benefit was at least as large as in others. If you are a carrier, pay particular attention to sleep, exercise, blood pressure, blood sugar and alcohol.

Polygenic scores: many small votes

Most common diseases are caused not by one gene but by thousands of variants with small effects. A polygenic score adds them up into a single number. A paper in Nature Genetics showed that such a score can identify 8 % of people with more than three times the risk of coronary artery disease. That is twenty times more people than carriers of rare mutations with comparable risk.

A score, however, gives a probability relative to the population, not certainty. It works best in people whose data it was developed on, so far mainly people of European descent. And it knows nothing about how you live: a person with a high score and good habits can be better off than a person with a low score and bad ones.

• What a test can tell you – carrier status for some important variants (for example APOE), a tendency towards certain traits, an approximate polygenic risk and ancestry.
• What it cannot tell you – an exact diagnosis, the date of a disease, the ideal diet or which supplements you need.

Nutrigenomics: a diet based on your genes?

The idea is tempting: the same food affects different people differently. Coffee is an example: the CYP1A2 gene influences how quickly the liver breaks down caffeine, and thus whether an afternoon espresso ruins your sleep. We see it all the time in the community: two people eat the same lunch, one is full of energy, the other falls asleep.

Controlled studies, however, are more sobering. The European Food4Me study from seven countries followed 1,269 people for six months. Personalised advice improved diet more than general advice – people ate less red meat, salt and saturated fat. Adding genetic information, however, did not increase the effectiveness of the advice. Personalisation works; genes add little to it so far.

Protecting your genetic data

You can change a password, but not your DNA. Genetic data also reveal information about your family. The European GDPR classifies them as a special category of personal data, but a lot depends on the consent you click when ordering. Even a large company can face a data breach or bankruptcy in which the database changes owner.

• Read the terms – find out whether the company shares data with partners or uses it for research; you can refuse consent to research.
• Minimise your footprint – a separate e-mail address and as little personal information in your profile as possible work well.
• Download and delete – after downloading your raw data, ask for your sample to be destroyed and your account deleted.
• Do not upload data just anywhere – free interpretation websites often lack clear rules.

Genes are not destiny: epigenetics

In the winter of 1944 – 1945, occupied Netherlands suffered a famine. People conceived at that time had, six decades later, different DNA methylation of the IGF2 gene than their same-sex siblings who had not experienced hunger in that period. The famine did not rewrite the DNA; it changed the marks that determine how it is read. This layer is called epigenetics, and it is influenced by diet, exercise, sleep and stress.

Even lifespan is not written in the genes as firmly as people say. Earlier estimates of its heritability ranged from 15 – 30 %. An analysis of the family trees of hundreds of millions of people showed that part of this similarity is better explained by people choosing partners with a similar lifestyle, and that the true heritability of lifespan is well below 10 %. Genes deal the cards. How you play them is largely up to you.

Conclusion

A genetic test is a good servant but a bad master. It can alert you to weak spots, such as APOE4, and motivate you to take care of yourself earlier. But it will not give you a diagnosis or a ready-made meal plan. Have serious findings confirmed clinically and protect your data. And above all: what you eat, how you move, how you sleep and how you handle stress changes how your genes are expressed every day.