Intelligence as measured by DNA tests Suppose a person has an IQ of 130, as measured by a standard IQ test, when he is 18 years old. He then goes on to develop severe mental illness, which reduces his IQ to 100. After this, he takes a DNA-intelligence test, like the one offered by GenePlaza, which claims to measure his intelligence from his DNA. Will this test show an IQ closer to 130 or 100? Note that mental illness is governed by a combination of genetic and environmental factors.

The answer will depend on the data that genetic model was developed from. I would suspect that most research on genetic factors of IQ has been performed in healthy, normal adults, so the model will predict IQ from DNA as if the subject were a healthy, normal adult (whether they are or not). It's unlikely that the DNA-IQ model was developed using a large cohort of mentally ill people, so the model may produce inaccurate results when applied to those people. I would expect the test to produce a score closer to 130 (if it is indeed an accurate model, which I will not discuss here).

Even if there are genetic predictors of the mental illness, it's likely not a 1:1 correspondence between DNA features and the illness, so this effect would be diluted when correlating against a third variable like IQ.