Statistics Question . . .

[jerseyrugger76]

If one of you has easy (i.e. free internet based from home) access to a resource like JSTOR via your university, could you please run a search on the odds of heart attack for an AAS user vs that of a control group? I found next to nothing persuasive or up to snuff scientifically from a simple google search.

I'm stuck at the office and thus cannot run over to Princeton's library myself today, and I think it's a question more than a few people are curious about though no one has said anything.

I've heard of more than a few guys in terrific shape dropping dead of sudden, massive heart attacks in their 30's and they had a few key traits in common.

Note, this is an inquiry into statistical correlation, not causation, so if you don't know the difference, don't bother me.

 

[PICK3]

If one of you has easy (i.e. free internet based from home) access to a resources like JSTOR via your university, could you please run a search on the odds of heart attack for an AAS user vs that of a control group? I found next to nothing persuasive or up to snuff scientifically from a simple google search.

I'm stuck at the office and thus cannot run over to Princeton's library myself today, and I think it's a question more than a few people are curious about though no one has said anything.

I've heard of more than a few guys in terrific shape dropping dead of sudden, massive heart attacks in their 30's and they had a few key traits in common.

Note, this is an inquiry into statistical correlation, not causation, so if you don't know the difference, don't bother me.

You think they've actually complied those stats?

 

[jerseyrugger76]

You think they've actually complied those stats?

Preliminary research has been done somewhere. I'll bet half my k on it.

It may not be a double-blind, perfectly controlled lab study like the kind they do in big pharma for a drug their submitting for FDA approval, but someone has run the numbers and published them in a credible academic journal. The question is too obvious not to have been asked and answered by someone already.

 

[PICK3]

Preliminary research has been done somewhere. I'll be half my k on it.

bet

 

[jerseyrugger76]

bet

What typo?

 

[Tatyana]

If one of you has easy (i.e. free internet based from home) access to a resource like JSTOR via your university, could you please run a search on the odds of heart attack for an AAS user vs that of a control group? I found next to nothing persuasive or up to snuff scientifically from a simple google search.

I'm stuck at the office and thus cannot run over to Princeton's library myself today, and I think it's a question more than a few people are curious about though no one has said anything.

I've heard of more than a few guys in terrific shape dropping dead of sudden, massive heart attacks in their 30's and they had a few key traits in common.

Note, this is an inquiry into statistical correlation, not causation, so if you don't know the difference, don't bother me.

You know the issues with correlation studies I assume.

I could do a search tomorrow, I have datastar and ovid/athens with quite a few journals the NHS subscribes to.

 

[jerseyrugger76]

You know the issues with correlation studies I assume.

I could do a search tomorrow, I have datastar and ovid/athens with quite a few journals the NHS subscribes to.

Many thanks! I knew you'd know where to look!

 

[Raptorex]

has nothing to do with correlation, need to estimate the rates in the exposed and compare to rates in the un-exposed, take the ratio (after controlling for appropriate co-variates which are numerous). Best estimator would be a rate-ratio estimator, also look for an odd-ratio in your search.

Best of luck.

 

[chris302001]

These arent the statistics you were asking for but articles related to the same topics that I looked up a while back. They're worth a read.

 

[jerseyrugger76]

has nothing to do with correlation, need to estimate the rates in the exposed and compare to rates in the un-exposed, take the ratio (after controlling for appropriate co-variates which are numerous). Best estimator would be a rate-ratio estimator, also look for an odd-ratio in your search.

Best of luck.

My point in emphasizing correlation was to fend off anyone who would get weird and defensive about the question being raised at all.

And a multi-variate logistic model would answer the question just fine Mr. Fancy Pants.

 

[LoneTree]

If one of you has easy (i.e. free internet based from home) access to a resource like JSTOR via your university, could you please run a search on the odds of heart attack for an AAS user vs that of a control group? I found next to nothing persuasive or up to snuff scientifically from a simple google search.

I'm stuck at the office and thus cannot run over to Princeton's library myself today, and I think it's a question more than a few people are curious about though no one has said anything.

I've heard of more than a few guys in terrific shape dropping dead of sudden, massive heart attacks in their 30's and they had a few key traits in common.

Note, this is an inquiry into statistical correlation, not causation, so if you don't know the difference, don't bother me.

You may be able to find some small case-control studies.
Long term use of AAS, especially combined with anti-estrogens, will adversely effect cholestrol profile, but probably heart attacks in 30's will only happen in people who have other predisposing factors (smoking, hypertension, diabetes, cholestrol problems, family history).
You can improve your odds by controling other predisposing factors.
There can not possibly be a randomized controlled trial.

 

[Raptorex]

My point in emphasizing correlation was to fend off anyone who would get weird and defensive about the question being raised at all.

And a multi-variate logistic model would answer the question just fine Mr. Fancy Pants.

not really, it would just provide you with an estimate of the relative risk (ie. odds-ratio) with very large confidence intervals given the small sample. Personally, I doubt the data your looking for is available. But best of luck.

 

[Tatyana]

LOLOLOLOLOL

Just look at you two stats whizzes.

I am using pretty basic stuff compared to you two.

I have heard of a few studies, but they are short term (1 year) and the steroid cycles were not standardised.

 

[chris302001]

LOLOLOLOLOL

Just look at you two stats whizzes.

I am using pretty basic stuff compared to you two.

I have heard of a few studies, but they are short term (1 year) and the steroid cycles were not standardised.

Same here, short term, or purely anecdotal.

Damn that rule about not experimenting on humans.....

 

[FISHTALES]

like thirty percent of statistics are made up, twenty seven percent of the time.

 

[ceo]

like thirty percent of statistics are made up, twenty seven percent of the time.