Personally I believe (but only IIRC, I still haven't had the time for a full review) the epidemiologic data hardly shows benefits from MK-4*, including the oft-(mis-)quoted Rotterdam study. The best cardiovascular epidemiologic, mechanistic and animal data is on MK-7. The best interventional CVD data is on mega doses of K1 and all other great studies were done with super-mega doses of MK-4 (osteoporosis, HCC, etc).
*mechanistically speaking this only makes sense: content in food is minimal and pharmacokinetics are almost one magnitude worse than MK-7...
Since there is such limited data, I think it would be more prudent to use K2 in the amounts that we are capable of getting from food sources. Since K2 is in butter, eggs, liver (all mk-4) AND bacterial fermentation (mk-7 or other), both forms seem reasonable (perhaps a low-dose combination is best).
That said, it does seem like a lot of the vitamin K2 craze was based on either epidemiological studies (which are suspect since most people in the cohorts had very little K2) or high-dose interventions (human or rat). If the epidemiological studies are dismissible, there may be little data supporting long-term low-dose supplementation.
However, if you accept the premise that the diets we evolved eating are less likely to kill us than the ones we invented recently, I suspect that early humans likely would have had higher levels of k2 than most do today, supporting notion that long-term low-dose supplementation may be healthy. Combining that reasoning with the good data from short-term interventions (at high doses) and mechanistic studies, produces a decent rationale for long-term low-dose supplementation.
Regarding long-term HIGH-dose supplementation, I just don't see why a healthy person would risk having so much more of a substance than we are capable of getting from food without better data.
Edited by Jay, 21 July 2009 - 09:50 PM.