# Thread: Interpreting Odds Ratios... weird question

1. ## Interpreting Odds Ratios... weird question

Hi,

I sort of have a weird question. I have a basic understanding of ORs, and can interpret them, but it's never been clear to me why we interpret values above/below 1 as percents and values 2+/- as x times. Let me be clearer. Let's say we're talking about the association of diabetes with glycemic control.

I might have an OR = 1.2, or an OR =4.0.

In the first OR, I would say diabetics are 20% more likely to achieve glycemic control. In the second OR, I would say 4 times more likely to achieve glycemic control. So, back to my question, why do we say % in the former, but x times in the latter?

2. ## Re: Interpreting Odds Ratios... weird question

Its just to mislead the uneducated reader, since 8% greater odds for outcome sounds better than 1.08 times greater odds. Just kidding, but not totally.

It is just an etiquette thing. Actually you can do whichever you want, but most adhere to what you presented.

3. ## The Following User Says Thank You to hlsmith For This Useful Post:

Ceci18 (02-06-2017)

4. ## Re: Interpreting Odds Ratios... weird question

What seems to me a bit more questionable is the use of the term "likely". Strictly we have a 20% increase in the odds, so I guess the explanation "20% more likely" might be misleading .

5. ## The Following User Says Thank You to rogojel For This Useful Post:

Ceci18 (02-06-2017)

6. ## Re: Interpreting Odds Ratios... weird question

Yeah, I have a boilerplate phrasing I always use, even though some people like to say more likely, I stay away from it many because is it appropriate, maybe/maybe not?

Then you get to the more nuanced statements where you say, males had a 2 times greater ADJUSTED odds of Y than females. People need to use that term more often if it was taken from a multiple regression model, and if multiple different models were used they need to say what was controlled for in particular.

7. ## The Following User Says Thank You to hlsmith For This Useful Post:

Ceci18 (02-06-2017)

8. ## Re: Interpreting Odds Ratios... weird question

Thank you, both, for your answers! I suppose I should be sticking to the "official" interpretation by using "odds of" instead of the likely language. Maybe that would have spared my confusion with the % vs. x. Thanks again and happy holidays!

9. ## Re: Interpreting Odds Ratios... weird question

What I find even more interesting is that with odds ratio/relative risk/hazard ratio of 1.2 one can conclude that first group has 20% increased risk of something, but also that the second group has 16.66% reduced risk of something (because 1/1.2=0.8333). It is just shown reciprocally

10. ## Re: Interpreting Odds Ratios... weird question

I believe invertibility holds with hazard ratios and odds ratios, but not with relative risks.

But, yes it is a nice property!!

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts