I can find articles about relative risk but I’m wondering what the NNT is if you get one group to exercise and another remain sedentary regarding primary prevention for CVD, secondary prevention, prevention of diabetes, prevention of complications of existing diabetes, etc.
relative risk reduction seems to run about 30-40%, at least for heart disease. I don’t know about diabetes.
Maybe rephrase with less jargon? I have never heard of “NNT”.
Guessing he means Number Needed to Treat, based on some quick googling. iljitsch’s comment is still valid; be careful of technology-specific abbreviations; there’s lots of professions and nationalities in SDMB.
I figured if someone didn’t know what NNT meant they wouldn’t know the answer anyway.
Yeah, but you also gotta figure that if someone doesn’t know what NNT means (I didn’t) then they also wouldn’t know that they don’t know the answer, and they’d be badgering you to explain.
Or they might be interested in expanding their knowledge base a little bit by reading explainers and the answers. Fighting ignorance for the masses and all that.
And if you explain the acronym in the OP, you avoid having the thread sidelined by a conversation about what the acronym means.
I’ve lost interest in the thread anyway. Lets discuss ice cream. I am partial to rocky road, but much of that is because I have coprophagia and the peanuts and marshmallows excite me sexually. What were we talking about again, NIN?
Dental care?
This site - http://www.thennt.com - was covered on Newshour last night. Nothing specific to your question but I get the feeling the site is actively expanding, so you might check it later.
Yup. But the numbers can be all over the place. I’ve heard switching to the Mediterranean diet resulted in a relative risk reduction of having a first heart attack by about 30% (so 30% fewer heart attacks vs controls). I think exercise is about the same for relative risk. That sounds good, but looking at the NNT it is something like 50-350 high risk people (people over 55 with multiple risk factors) who need to be treated for 5 years to prevent one heart attack or stroke. On a societal level of millions of people there is benefit, but most individuals will not benefit.
I can’t find info on exercise and NNT for diseases of old age, but I can find info on diet. And I’m guessing/wondering if the NNT is the same for both since I believe relative risk reduction is about the same for both.
Also I’m wondering if diet and exercise are more effective for diabetes gauging by NNT than they are for cardiovascular disease.
The diabetes bit is a quick google search for this:
Not easily finding the others though. Sorry. Not sure if there is some way to translate RR numbers into NNT ones. I think you also need the incidence, i.e. the absolute risk, and then you can calculate it. Maybe someone with more stats expertise will chime in.
Or just enjoy your ice cream!
Thing about exercise though is that the intervention impacts many different measures, not just first heart attack and stroke. The NNT to impact one of those measures in a significantly positive way is likely much much smaller since most people will end up with one of those many items.
Thanks for that, a NNT of 61 per year for mild exercise to prevent death with type II diabetes is much better than I was expecting.
Okay here are some tools we can use to make the conversions. The nomogram they reference is here (pdf). What you need in each case is the absolute risk rate (the risk of events to the control population, referred to there as the ARR or CER). Then if you have the odds reduction (OR) or relative risk (RR) you can calculate the number needed to treat (NNT).
For those who wonder why this matters: a decision to utilize a costly intervention (be that cost money, or time, or risks of adverse side effect) is better informed by this number than relative risk reduction alone. This can be seen by the following extreme illustration - Something that completely prevents a very rare event if given to every person will have a 100% reduction in relative risk but the number needed to treat will be huge, so the costs/risks of it have to be very low to justify it.
So taking the example of the Mediterranean diet for prevention of a first cardiovascular event (and only that outcome) - the data shows that the absolute risk of myocardial infarction, stroke, and death from cardiovascular causes over 5 years went from 4.4 to 3.4% on a Mediterranean diet inclusive of nuts. That’s an absolute risk reduction of 1% and thus a NNT of 100. That’s pretty good. The limit is that roughly 95 out of 100 won’t have a CV event in the next 5 years in the control group, hence if an intervention was 100% effective the NNT to prevent an event in the next five years (and only the next five years) would be 20.
Exercise is felt to be the same numbers but I am not easily finding any prospective controlled trials looking at exercise alone to harvest the numbers out of. Most are observational. Even the secondary prevention studies are not sufficiently powered to look at death rates as end points - they generally use proxy measures.
But if you find the actual studies Wesely, that’s at least how you do it.
Bumping to share this NYT article about NNTs. It links to that nnt clearinghouse mentioned by jasg.