Ad watchdog says Nucleus Genomics cannot substantiate embryo IQ and cancer claims

On September 11, 2026, BBB National Programs' National Advertising Division (NAD) issued recommendations to Nucleus Genomics after reviewing advertising for Nucleus IVF+, a service marketed to couples undergoing IVF who want to compare multiple viable embryos across genetic disease risk and predicted physical traits before selecting one for transfer. NAD initiated the inquiry through its own monitoring -- not at a competitor's request -- and examined a slideshow on the Nucleus website and an Instagram video featuring the company's founder. It recommended that Nucleus modify or discontinue five categories of claims: disease risk comparison, overall cancer risk, eye and hair color, height, and intelligence as measured by IQ-point values. Nucleus said it would appeal.

The advertising at issue was specific. The Nucleus IVF+ website slideshow displayed numeric values for height -- plus-or-minus inch differences between embryos -- and IQ-point ranges for cognitive ability. Founder statements in an Instagram video described the ability to select the "smartest" embryo. A separate website slide told viewers they could "see which embryo has the lowest risk for cancer." These claims are the type that draw consumers in because the numbers feel concrete: not a vague statement of lower risk, but a specific embryo ranked first.

NAD reviewed the supporting evidence Nucleus provided, which included published polygenic risk-score research and the company's proprietary performance metrics. The advertising body's central finding was that population-level statistical evidence does not validate accuracy for comparing a small group of five or six specific embryos. On height, NAD acknowledged that adult height is highly heritable, but said the company's measures did not quantify accuracy for identifying the tallest embryo in a given cohort, or validate the inch-level differences the slideshow displayed. On cognitive ability, it reached the same conclusion: evidence that polygenic scores correlate with outcomes in large populations does not establish that the specific IQ-point values shown for a set of embryos are meaningful. On cancer, NAD also found no evidence that Nucleus had established a methodology for combining separate cancer-type scores into a single overall cancer ranking.

One claim Nucleus did not wait to have ruled on. Before the decision was finalized, the company voluntarily and permanently discontinued a website statement promising that parents could see their future child's "health, traits, and potential before pregnancy." NAD said it would treat that exit as equivalent to a discontinuation recommendation. On the five remaining recommendations, Nucleus said it plans to appeal to the National Advertising Review Board (NARB), the appellate body within the BBB National Programs system, saying in a statement that the decision cannot be reconciled with the evidence it submitted. If NARB upholds NAD's recommendations and Nucleus does not comply, the matter can be referred to the Federal Trade Commission.

For anyone considering a consumer whole genome sequencing service rather than IVF embryo screening, this ruling still carries practical weight. Nucleus Genomics also sells a standard consumer WGS kit, and the polygenic risk scores at the center of this ruling are the same general methodology used in the disease-risk sections of most consumer WGS health reports, including those from Nucleus, Sequencing.com, and SelfDecode. A polygenic risk score can place you approximately on a population-level distribution, but it is less reliable as a precise number attached to one specific person. NAD's findings are a useful reference when reading any WGS provider's health-report marketing: when a report shows your exact percentile for type 2 diabetes or breast cancer, that number is a statistical estimate, not a measurement, and the precision its formatting implies is the same kind the advertising watchdog found unsubstantiated here. The reports can surface meaningful risk signals worth discussing with a clinician -- the point is not that the numbers are worthless, but that the confidence their framing often suggests outruns what the underlying models can actually deliver.