• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

Virta Health Two Year Results

ziggy_w said:
For those who are interested, here is an article reporting Virta's two-year results.

Still look very good, but maybe quite as good as the one-year results.

https://www.biorxiv.org/content/biorxiv/early/2018/11/28/476275.full.pdf
Click to expand...
Thanks for the link! The interesting thing for me is that although the study cohort put back on some weight and had increased HBA1c even though they remained in ketosis,, their insulin resistance seemed not to have worsened. Also the 'placebo' (treated with the standard care package) did not seem to show any improvements in the many biomarkers. I know it is biased and non randomized but am glad that studies like this are being done to show what results can be achieved without medications.
 
I havnt read the full data as yet, but one thing that always puzzles me about the virta studies is how patients were allocated to the treatmnet as usual and study groups - has anyone got a link showing this?
My experience with scientific studies comparing an intervention with “ treatment as usual” is that people are randomly assigned, with 50% of participants in each group. The numbers in the Virta data indicates about 75% of participants in the study group and 25% in the treatment as usual group, which also makes drop out rates prior to data reporting even more important. There were some also apparant differences in starting baseline measures - things like starting waist circumferance, ethnicity, gender balance etc between the two groups.
 
Last edited:

Hi @NicoleC1971,

I agree. The way I see it the results are still very strong. The drop-out rate is a bit higher now, but maybe partially due to people just not getting the required lab tests. I am also not so much worried about the non-randomization (this might also be because I do research in another area where randomiziation trials are often not feasible and therefore uncommon).

In the social sciences if you want to look at causation, you need to show three things: Correlation (diet and changes in bg are related), temporal precedence (the change in diet becomes before the change in bg), and elimination of plausible alternative explanations. Randomization is intended to deal with the third factor. Since we don't have randomization, the question is: Are there are any other differences between the two groups that might have caused the effect on blood glucose, insulin resistance, blood pressure, blood lipids, etc.?
 
Last edited:
People were allowed to choose the arm of the study they were in. Remember they are trying to show the likely outcome if health insurance companies offer virta as an option. Given the high cost of drugs in the USA, they have shown that within a few years offering virta--health as a option will save money.
 
We all fast every night. If skipping the odd meal or two was bad for our health, we would have died out millions of years ago. In my opinion

Anyway it works for me once a month and has a noticeable impact on getting my blood glucose even lower, and not just in the immediate term, but through the whole month or so. Personally I’m an IF advocate, although not necessarily for weight loss
 

Thanks for that info @ringi
Can you tell us where you read it?

The control group/different arm of study has always bothered me a bit.
I mean, here you are, with a protocol that you are convinced will profoundly help people, and you put a group of people into the study arm which won't get those benefits.
What an ethical dilemma.
However, if people are offered the choice and self-select then the problem is resolved.

I wonder how they cope with people wanting to switch arms mid study, because if I had chosen the orthodox treatment option, then I would be wanting to switch, based on the findings so far.
 

Yes, but Dr Phinney has strong views against fasting due to lean mass loss. He feels it is not worth it...but I see that as a limitation of the Virta's protocol.
 
We don't all respond to fasting in the same way (as shown by many posts on this forum).

Plus, if fasting was incorporated in this Virta study, then it would be introducing another massive variable and would, frankly, completely disrupt the results and be one more criticism for detractors to use against the protocol.
 
kokhongw said:
I believe one of the limitation of Virta's program is that unfortunately, they do not believe fasting has any long term benefits..
https://blog.virtahealth.com/science-of-intermittent-fasting/
Click to expand...
That will partly be down to Dr Phinney's dislike of it as a theraputic treatment. Which is a shame but he has kind of made so many statements on his belief that is not good that he needs to undergo a Noakes like conversion for him to back it.

Note to self.. read all the thread before contributing..
 
I think for me, the simple ethos regarding fasting is eat when you’re hungry and stop when you’re not. If I get hungry when I’m fasting then I’ll break the fast and eat.

It strikes me that nature knows when I need to feed, and cares little for meal times prescribed by society.
 

Yes...so the question is whether we are looking for an optimal protocol ( a combo of known approach LC/Keto/IF/Extend fast) for the individual or out to prove that only one approach (well formulated keto) is needed...

My view is that long term keto only has its limitation...although it is already remarkably effective. Some of us really need that extra glucose/fats clearance made possible with IF/Extended fast.
 
Going by the conclusion posted here I don’t read that as overly negative on fasting.

It talks about “potential”, “can” not will, “full day” fasts more than “twice a week”. Not overnight, 16/8, weekly or monthly fasts. I can’t see anything wrong about the last two points either.

Even the first talks about practice outpacing published data which is an unclear choice of wording at best imo leading my interpretation to they don’t know what current practice is doing as it’s outside previous research parameters.

Perhaps the full report words it differently
 
The relatively modest HbA1c reduction of 0.9% means there's still much room for improvement in the protocol.

 
All else aside for a moment, I’d personally classify a 0.9% absolute HbA1c reduction as pretty significant. That’s the difference between 31-41mmol/mol. If I’m doing my maths correctly
 
There was a lot of discussion about their study on twitter when the first year results come out, they also did a few Q&A sessions on Facebook. If I recall correctly, their service is half the price for the 2nd year, with much less intensive surport. They claim that they are now getting better results on drug cost savings then the study group.

Recently they have taken on a few large employers who "self insure" for health care and have started a payment system when they only get paid if the pt get a greatly improved A1c.

What is not clear until you know it, is that they excluded very few people from the study and set it up to show the "return on investment" of health insurance company going with them, rather then to prove the science.
 
You must log in or register to reply here.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn More.…