This isn't a particularly good study because it is also more or less a drug study with the drug making a large impact on risk. Paul Jaminet comments here:

"Is the DHA-cancer connection drug-driven?: Via Dennis Mangan, I’m informed of a helpful comment on the Brasky study by commenter Karl at FuturePundit. Key points: Half of the subjects in the study were on the drug finasteride (Propecia, Proscar). The finasteride arm had 78% more high-grade cancers than the placebo arm.

While tissue DHA levels were strongly associated with high-grade cancer, EPA levels weren’t. On the one hand this is consistent with our idea that high tissue DHA drives cancer via the DHA-angiogenesis mechanism we’ve been discussing: only DHA, not EPA, produces high-grade cancer.

However, normally tissue DHA and EPA are obtained from food and levels go up or down together. Few people in the study supplemented DHA alone; so DHA and EPA levels should have been high or lower together and both should have shown the association.

A plausible explanation would be that the drug was raising DHA levels artificially, and triggering high-grade cancers.

Unfortunately the study doesn’t report the level-response relationship between DHA and high-grade cancers separately in the finasteride and placebo arms – only in the two groups combined. I see this as a failure of the reviewers and an embarrassment to the authors.

Taking Propecia for an enlarged prostate may be much more dangerous than eating salmon.

This plausible hypothesis makes me more comfortable maintaining our recommendation to eat a pound of cold-water marine fish per week."

http://perfecthealthdiet.com/?p=3338

By the entire body of literature I can conclude 4 things.

1. Trans fats kill you dead.

2. Too much linoleic acid is a causative factor in all cancers.

3. Inflammation is a causative factor in all cancers, and a certain amount of DHA is necessary for optimal levels of inflammation.

4. Too much DHA IS bad, but whether or not it will cause cancer in healthy people is highly debatable. Fish-consuming populations tend to have very low rates of cancer. What constitutes "too much" is also highly debatable.

So indeed even if it wasn't a bad study it would still be at odds with the literature. Like here for example http://www.ncbi.nlm.nih.gov/pubmed/10584888

Denise Minger picked this study apart in a guest-column written for Mark's Daily Apple. It's worth reading:

http://www.marksdailyapple.com/fish-oil-prostate-cancer/

First things first: Can we really conclude that the men with high DHA levels were eating more fish (or fish oil) than the rest of the gang?

The answer is a resounding “nope” – and this is why the current fish-condemning media coverage is a load of hooey. Although consuming more DHA can definitely boost your serum levels, the reverse isn’t always true: elevated DHA doesn’t automatically mean you’re a sushi fiend or loyal fish-oil guzzler. Guess what else can increase omega-3 fats in your blood? You’ll get a hoot from this one: low-fat diets.

I also talked about this and Denise's response on Latest in Paleo, Episode 14: http://www.latestinpaleo.com/paleo-podcast/2011/5/2/episode-14-bacon-saves-manhood.html

I wonder how much information is available on:

a) The species of oily fish consumed e.g. apex predators like tuna or small fry like sardines/anchovies.

b) The quality of the fish e.g. farmed vs wild caught.

For example, mercury, which is present in greater concentrations the higher up the food chain you go, is also a factor in prostate cancer. Equally, farmed fish are as polluted as their "toxic soup" environments/diets, so I imagine that the hefty dose of PCBs/heavy metals along for the ride with the O3s would cancel out some of the latter's benefits from a cancer-prevention view.

Edit: a quick Google search reveals a rather graphic insight into the exact ingredients of this "toxic soup".

http://www.dietmindspirit.org/2008/01/30/why-you-shouldnt-eat-this-fish-pangas-pangasius-vietnamese-river-cobbler-white-catfish-gray-sole/

I would be much more cautious when you administer DHA to those susceptible to developing prostate cancer (as the study was).

Consider:

1 tbsp of fish oil per day is equivalent to 4.8g of n-3 and about 1.5g of DHA. Do you realize how much fish consumption that is: ~1,800g of fish per day. The Japanese only consume around 60g of fish per day or 150mg of n-3. Not even those Inuits or Eskimos could consume that much fish in a day. I think the math is correct here. Somebody point out if it's not. ~1/3 of DHA per n-3. That's what we're talking about. Even a teaspoon of fish oil is equilvanet to eating 600g of fatty fish from the North Sea.

The fish oil is a very concentrated form of DHA, EPA and other n-3 acids. Presumably, it is necessary if you eat a SAD where n-6 trumps n-3. But a tablespoon or more each day for something that only accrues benefit through this "balancing act"? Maybe the Inuits and Eskimos don't get cancer because they never get to consume 1,800g+ of fish.

That's the point that gets missed. The difference between what works and what doesn't is the dose: At a low dose, fish oil seems to work. At 1 tbsp? I'm not so sure.

Rather than ask my question, I am posting on this thread.

As an older woman eating grain free, sugar free, seed oil free and low carb, do I take 1 tsp. Carlson Cod liver oil to help omega 3s since I can't eat grass fed beef (Kurt Harris), do I eat no more than 1 lb. salmon a week (Paul Jaminet), and what about my friend with bone cancer...no fish oil?