This study tries to answer the question, "is the same true for high intelligence?" The two general theories are the Continuity Hypothesis and Discontinuity Hypothesis. As its name suggests, the Continuity Hypothesis predicts that the high end of the intelligence distribution is continuous; extremely intelligence individuals don't violate the intelligence distribution the way mentally challenged people do. The Discontinuity Hypothesis predicts the opposite.
By analyzing the genes of twins and other close family members, the researchers found strong evidence of the Continuity Hypothesis.
Intelligence is selected for in humans. Therefore any individual gene that significantly boosts intelligence should be expected to have already spread through the population. As a result we should not expect to find any rare genes that make people super-smart. So super-smart people get there with a combination of different genes, each of which contributes very little.
The bottom end of the scale is the opposite story. Evolution says that individual genes that hurt intelligence should be selected against, and are therefore expected to be rare. (Mutation says that they should not be non-existent, but they should be rare.) Therefore there is no surprise in finding rare individual genes that significantly hurt intelligence.
This pattern is not unique to intelligence. It is predicted for any trait that has actively been selected for by evolution over a long period. The top of the range should look continuous. The bottom of the range tends to be dominated by deleterious point mutations.
The same can be said with drugs -- any chemical that your body could have reasonably produced on its own should be expected to reduce your reproductive fitness. So brain drugs should nearly all have side-effects.