No, risk reduction is something different and can not be estimated from a case-control study. You can only estimate relative risk under the condition that the disease is rare.

You can say odds reduction yet it would be a new concept.

The factor for which OR is below 1.0 is considered as a protective one , also including 0.5

Thank you dear friend. What is the best interpretation of OR= 0.5 for clinicians? I want clinical meaning but not epidemiological interpretation. 

What do you mean by clinical meaning?

I think clinical meaning/importance depends on the disease and in order to evaluate it we must be doctors.

OR=0.5 in a case control study means that odds of exposure in cases is half the odds of exposure in controls (this is actually exposure odds ratio which is the outcome of a case-control study and might be unimportant/meaningless). 

Since exposure odds ratio is equal to disease odds ratio, when OR=0.5 a more important/meaningful explanation is that exposed have half of the odds that unexposed has. So being exposed is protective against the disease.

An odds ratio (OR) is a measure of association between an exposure and an outcome. The OR represents the odds that an outcome will occur among those with a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure. Accordingly OR=0.05 means that  those with history of exposure had a half chance of getting outcome than those who did not have exposure. Namely this kind of exposure had a protective role and could be a kind of curable treatment etc.

It's best to keep statistical and clinical interpretations at "arm's length".  

An OR of 0.5 is exactly the same as an OR of 2.0 if you had switched the exposure levels.  In other words, you have found a "protective" exposure as it is shown, BUT if you were to re-draw the table showing the 'other' EXP group, or the "non-exposed" at the top, the OR would show 2.0, meaning that the ILL people were TWICE as likely to have been in the NON-EXPOSED group compared to the NON-ILL people. OR you can say: that the NON-ILL subjects were twice as likely to be in the EXPOSED group compared to the ILL subjects.

You must remember to describe the original suspect exposure as a 'protective' exposure variable, BUT REMEMBER THAT THE "PROTECTION" CAN BE STATISTICAL AS WELL AS CLINICAL OR PHARMACEUTICAL. By this I mean that if there were only TWO exposure levels, and all subjects were exposed to only ONE, then IF the relationship between the HARMFUL-EXP and illness were 2.0 (in your case), then the relationship between the illness and the OTHER EXP would automatically be 1/2 or 0.5.   In such a case, the 'protective' exposure would NOT magically be conferring some prophylaxis or remedial function.      

I agree what was written till now, anyway, please, pay attention to the confidence interval before concluding that there is a  protective effect. If the confidence interval contains 1, there is not statistical significant difference between exposed and not exposed!

protective .

An odds ratio of 0.5 would mean that the exposed group has half, or 50%, of the odds of developing disease as the unexposed group.  In other words, the exposure is protective against disease.   However your confidence interval will determine if this is statistically significant or not:  if the CI does not include 1, then it is statistically significant.  If it crosses 1, then it is not significant (although the lack of significance may be because of a small sample size and not because the exposure isn't protective---but you can't tell for sure until you do another study!)

And since you want a clinical interpretation for individual patients, keep in mind that this is an odds ratio for a group of patients---not a certainty for each individual patient.  There were still some patients in the study who had the (protective) exposure but who still got the disease.  

1. Risk can not be said, since it is case control study. Risk can only be estimated from longitudinal study. 

2. You have to report OR with its confidence interval value. If the upper bound of CI is also less than 1, then you can certainly say that exposure is negatively related to the disease (protective). Without CI difficult to comment.

The last point is well taken. An odds ratio is understood (but not often explained) as an estimate of the "true" relative risk.  The latter requires the true incidence rates for each exposure group, and these are rarely available in a "case-control' model. But having clarified that, the O.R. is still a useful descriptive parameter, provided the confidence limits are observed and interpreted correctly. It is particularly vulnerable to small cell sizes of course.   

OR 0.5 depicts that the individual in particular category has 50% lower odds of having an outcome

OR 0.5 shows that compare to a reference group, that category has less that 50% odds of having an outcome

the exposed group has 50% more protected than the unexposed group in developing the outcome.

.