Get your genome out of my risk pool - Matthew Steinglass

http://www.economist.com/blogs/democracyinamerica/2011/06/genomics-and-health-care-inequality

Part of the logic behind insurance is that it's a risk pool; none of us knows when we're gonna go, so we agree to split the costs. But genetic profiling may increasingly give each of us our own set of pre-existing conditions, good or bad. And that may test people's willingness to chip in for the health costs of their fellow-citizens. When "it coulda been me" turns into "nope, it couldn't", we may start seeing...hm, I was about to say "a breakdown in social solidarity", but then I remembered we're talking about America here. How about "even less willingness to do anything for people who aren't as lucky as you are."

In other words, genetic redlining. While benefitting the healthy, whether they are healthy through their own actions or not, this also completely breaks down the ability of insurance to function. No longer will insurance be able to forecast average losses over large groups of people, but to hedge bets against various smaller sub-groups, each with their own risk profile.
On one hand it makes sense to charge a higher premium to a smoker who is obese and never works out, or to lower premiums for a jogger who has low blood pressure. But this is something entirely different. This would mean that the jogger's efforts might be for naught if they are genetically predisposed to cancer. People who are healthy and lucky assume it is through their own control, but often this is not true.
On the other hand, in one sense this is already available. How many times have you answered questions about your health or your parents health, or whether certain diseases run in your family? Granted, this would improve the data that was being gathered, but a lot of this information is already available, though privacy laws wall off much of what would be useful.

I don't see this as much of a threat.