ofonorow wrote:And can you (or anyone) explain why all these studies were published on average 5 years after the study ended? (I can think of a few reasons...)
You're looking at trials involving thousands of patients whose records have to be collected, reviewed (sometimes by independent bodies if the protocol requires), and then the statistical tests have to be done and checked, and finally the paper has to be written. These are also often multicenter trials, so you can imagine the logistics involved - the investigators might be working on more than one project simultaneously, not to mention their other academic obligations. Furthermore, even once the manuscript is finished, it first has to be sent in for peer review, and modifications/clarifications might have to be made based on the results of the peer review. Even once the paper is approved for publication it may be a few months till the paper is actually published in print.
Naturally the largest study is the Heart Protection Study, and depending on the weighting, this study may have undue influence. (I still maintain that the HPS positive results includes the statin + vitamin group, and not just the statin group. The paper can be read this way, and since they haven't released the data, my interpretation is impossible to refute
Actually, since they included all the 20,000 HPS participants, this data has to include about 4000 who were both on vitamins and statins.
There was no benefit shown with vitamins in the HPS study, i.e. the group on vitamins and statins did not do any better than the one on statins alone. I recall posting the link in another thread.
One wonders how many of the other trials played the same game, adding vitamins to statins?
[/color]
You can go on the CTT website where they've listed the 14 studies used in the trial, and you can look up the design of each trial:
http://www.ctsu.ox.ac.uk/projects/ctt/i ... reference1Taking all years together, the overall incidence of major coronary events was reduced by about one quarter per mmol/L reduction in LDL cholesterol among participants with a previous history of MI or other CHD, as well as among those without any pre-existing CHD (figure 5). But since the absolute risk of events was higher among participants with pre-existing CHD, this reduction of about a quarter per mmol/L LDL cholesterol reduction translated into 30 (95% CI 24–37) fewer such participants having major coronary events per 1000 during an average of 5 years, compared with 18 (14–23) fewer among participants who did not have pre-existing CHD (figure 6 and webfigure 1ii).
So 30 fewer (per 1000) over 5 years is what, 6 per 1000 per year? or 3 per 1000 fewer for those who did not have pre-existing CHD. This is less than a 1% difference in absolute terms between the two groups, correct?
You're looking at major coronary events, not all major vascular events (if you take that into consideration, the risk reductions are greater). Figure 6 shows the summary in a format that's easier to understand. Like you say, if you're looking at major coronary events alone, there were 30 fewer events per 1,000 people after 5 years amongst those with previous CHD (i.e. secondary prevention) vs. 18 fewer events per 1,000 people amongst those without previous CHD (primary prevention). I'm not sure why you are comparing the primary prevention group to the secondary prevention group, unless you want to show that the benefits are greater for those who already have CHD. When comparing to placebo, the absolute risk reduction for the secondary prevention group is around 3% compared to <2% in the primary prevention group - when looking at coronary events alone. When looking at all major vascular events, the absolute risk reductions are roughly 5% and 2%, respectively.
Bear in mind, this is just looking at a 1 mmol/L reduction in LDL-C. If you look at figure 3 (or even the
webfigures if you have access to them), you'll see that the risk reduction is proportional to further reductions in LDL-C, based on the meta-regression. Also note that this is looking at all participants - the benefits vary depending on patient characteristics, i.e. greater benefit in men than in women, greater benefit in patients under 65 compared to those over, etc. As I've said before, the benefit of statin use varies from patient to patient. Certainly, someone who is already at low risk of cardiovascular disease will not derive a clinically meaningful benefit from statin use. On the other hand, there is strong evidence of clinical benefit in patients who are at higher risk.
For someone who has had a previous MI, taking statins for five years results in an absolute risk benefit of 5% - i.e, 1 in every 20 people in a similar situation will manage to avoid having another heart attack by being on the drug. To some, those odds are pretty meagre. But to others, a one in twenty chance of avoiding another heart attack or having to undergo a bypass surgery is something they'd be more than willing to take. If I had those odds, I probably would take it so long as I can afford and tolerate the drugs. But someone else might not think that the cost-benefit ratio is not high enough. Patients need to be able to make an informed decision based on the available evidence.
So we are basing the idea that statins reduce mortality on a less than 1/2 of 1% difference in a population known not to be advised to take vitamin C.
I beg to differ on the latter point; it's simply speculation to say that the study population were being advised not to take vitamin C. But even if they were, the control group would also have had the same recommendations. It does not change the fact that the treatment group benefited.
Quite a small difference between placebo and statin intake, if you ask me, and probably the reason these studies took so long to publish. They are not very good news for the makers of statins.
I guess it's a matter of opinion. We can look at the absolute mortality risk reduction over five years (which is quite low to begin with), or we can look at the reduction in major vascular events. I think the latter is an important endpoint as far as quality of life is concerned.
Studies take long to publish for a variety of reasons, mostly logistical. Most of the studies in this meta-analysis were probably very good news for statin makers (and you only have to look at statin sales over the past decade). The recent JUPITER trial for Crestor was deemed to be good news for AstraZeneca even though the absolute risk reduction was only something like 1% (or perhaps less, I don't recall).
Obviously, statins are not magic bullets. But used in conjunction with other interventions and lifestyle modifications, they can provide a clinically meaningful benefit.
Now, they seem to focus on lower cholesterol, rather than statin use per se. It is hard to discern how this might impact these results, but we know that vitamin C can lead to lower cholesterol levels. So for example, anyone who participated in these studies, taking a high amount of vitamin C on their own - in either group, would apparently drive this analysis toward lower mortality w/r lower cholesterol.
You're implying that by chance, in a population of thousands of patients, those taking vitamin C supplements somehow aggregated in the treatment group. That's literally incredible.
Yes, this seems to be how they achieved this result. This paper is not focused on statins, it is focused on absolute cholesterol reduction. Yes, it is probably true that the statin group's cholesterol was lower, but how do we know for sure? This is my first reading, but I cannot find a statement that these results are on the basis of who took statins, and who didn't, rather on cholesterol reduction.
The analyses are based on the treatment group vs. the control group.
I don't think they've included a table summarizing the differences in cholesterol from each study, but you can find these by looking up the individual studies. For example, from the AFCAPS/TexCAPs study:
Link.
Anyway, if vitamin C lowers cholesterol then perhaps it does prevent cardiovascular disease at sufficient dosages.
