help with interpreting article stats

[goldfish]

---

Members don't see this ad.

Dunno if this is the right forum...
Can some1 explain the "per percent increase" part to me?

"Insulin resistance was the strongest predictor of having
nonalcoholic fatty liver disease, with an odds ratio (OR)
of 15 [95% confidence interval (CI): 3.0 to 70] per percent
increase in insulin resistance."

Does that mean for every 1% increase in IR theres a OR of 15? thats kinda weird..

 

[tupac_don]

Dunno if this is the right forum...
Can some1 explain the "per percent increase" part to me?

"Insulin resistance was the strongest predictor of having
nonalcoholic fatty liver disease, with an odds ratio (OR)
of 15 [95% confidence interval (CI): 3.0 to 70] per percent
increase in insulin resistance."

Does that mean for every 1% increase in IR theres a OR of 15? thats kinda weird..

Yea that's how I take it to mean. Say if you had insulin resistance 80 and I had 81% I would be at statistically more risk than you, is it clinically significant?! That you need to decide, but certainly they are saying that increase in insulin resistance, is no good for your liver.

 

[goldfish]

i made a direct quote of the OR=blah blah.. per percent increase.. into my report and my supervisor (endo consultant) said it made no sense... i'm confused.

 

[aragonh]

based on the or and the 95% CI, the varience of the study was very high.

That is usually due to small number of subjects in the study.

I would be careful in quoting this article. Look for other references that support or deny the claim.

Remember, a lot of junk gets published out there.