Future independent RCTs



I know that Soylent is by no means a pharmaceutical, a drug or anything that might fall into the category of a medical intervention. However, as more people use it, are there plans to conduct clinical, drug-style randomized controlled trials to more accurately observe the risks and benefits of this health “intervention” in people?

After all, RCTs are the best and most fair way we can know whether one intervention works better than placebo (or some other currently available intervention), so why not use this tool to try and understand Soylent better? The devil lies in the detail, and by carefully and independently giving access to pooled data from such trials in a regular and almost automatic fashion, and having more eyes on this data, I believe the potential benefits and risks can be better understood.

I don’t know how many people are currently using this or whether a trial of such a scale is even possible at the moment, but what is important is that the plural of anecdote is not data. Transparency is also vital here, not necessarily in full disclosure of ingredients used, but in making data about these trials available to everyone.

I know that no other food company bothers to do this. Soylent, I feel, should take a more responsible approach, not because it is potentially a more all-encompassing “intervention” (you could exclusively live off fast-food, for example), but because it should set the example for other food manufacturers and be responsible and evidence-based about food just like everything else.

I think it’s a great concept but trials need to be run on this intervention to fully understand benefits and risks.


There are 25000 orders for trials being shipped right now. It’s a user funded trial, Microsoft style. If people keep ordering and survive, the trials will be successful.

But seriously ? It’s food. Enjoy it. It’s not medication.


I don’t want to confuse the word “trial”. By trial I meant a standard clinical trial where you take a group of people, randomize them, give them two things and make sure they don’t know which is which. Obviously these types of trials are difficult to implement but it can be done.

I completely disagree with your second point. Let’s say there’s a background risk of developing a certain type of negative abnormality in the general population, let’s say this risk is 3%. If a trial shows that people using this type of food have a 5% risk, that’s 500 more people being affected (assuming 25000 users). I know it’s not medication, but it’s still an intervention that could have possible harms and benefits, no matter how many biological justifications you make.

The body is complex and I think this product should recognize that and be responsible about it.


I think its a good idea, but I don’t think they have any more responsibility to test it than say McDonald’s testing what happens if all you eat are Big Macs. Every individual has a responsibility to approach their food choices with caution and make adjustments if things aren’t working.

I’m hoping in the absence of a formal study that we get lots of anecdotal evidence from people posting their experiences both good and bad. Blood measurements are a good indicator of health so I think its important to have them until you are confident that its working and stable.


I love the idea of clinical trials, but I’m not holding my breath since they are not RECOMMENDING that people go 100%. So if they allow you 3-4 meals a week that are muggle food, it becomes more and more difficult to see exactly what is doing what.

Alternatively, I’d love to see a study that has some parameter along the lines of replacing 2 meals a day with Soylent and leaving meal three unchanged to observe changes. I think that would give us some great insight without putting Soylent on the hook for a recommendation it has refused to make.


I saw this discussed several months ago, and one big problem loomed: how do you do the randomized study? Usually you give one group the actual medication and the other a placebo and compare results, without the members of the group (or sometimes even the people running the study, at least until the end) knowing who’s in which group. If we had an alternate food source that was indistinguishable from Soylent that we knew was safe…we’d be calling it Soylent. The placebo is the sticky part here, because you have to give people actual nutrition, not a blank pill. If you can come up with a way to get around this, we’d love to hear it; many people would like hard data, but no one has (yet) come up with a way to do that RCT.


The treatment is much more demanding on which side? Even RCT’s do not randomly sample from the population. Recruitment strategies may aim for representative samples, but it’d probably be most efficacious to do paired or stratified recruitment. Every person who is recruited would be someone willing and interested to drink Soylent. Then, have two separate control groups, one eating what they’d normally eat and one eating a dietitian guided diet.

Main problem with these studies is going to be self report. Nutrition studies are always shit because it’s always self-report which have serious reliability issues. There are some ways for better reliability but it’s all generally bandaids over gaping wholes.


McDonalds doesn’t claim you can live on Big Macs. Soylent is supposed to be sufficient in and of itself, consumed on a highly regular basis. I think that is a significant difference.


Its food. Its no different than going to the store and getting a Swansons TV dinner. What is this intervention stuff you keep talking about? It is no such thing. Its food. Food and nothing more.


Companies can claim a lot of things, they usually don’t need to justify them. See: the current POM Wonderful suit. What tells them what they need to prove is what classification they’re sold under. That’s why herbal supplements are often viewed as a scam (because most of them are).

It’s an intervention in that it’s a very different way to ‘eat’. Interventions can be anything, it just means it is what you’re interested in evaluating the effects of.


Good point, it really does come down to the claims. Are they claiming this will cure disease or will work for everyone, or that it works for them and it might work for you too? I see the latter as being a responsible position, and if they start making medical claims then they need to do testing to be morally responsible.


In my opinion, the idea of a double blind study is not applicable to this situation for the reasons already stated. We need to be careful of not latching on to concepts and adhering to them dogmatically like they are some sort of holy grail that proves or disproves theoretical arguments. Were we to do that, we’d be giving credence to the not so ridiculous idea that faith in science is just another form of religion. At some point methodology needs to take a back seat to reason and pragmatism. I don’t wish to insult or offend or “attack the person” of the original poster here. Rather my intent is to suggest a paradigm shift in ones thinking. Yes science and all it’s methods are useful, if not necessary. But to switch off ones intellect in order to blindly follow a formula - that’s not science. That’s religion. Again this is not to attack the person, just an observation.


Thank you for the response.

I agree, I don’t think Soylent has any more responsibility than what any other food company has. Individuals do have a responsibility to control their own food choices. What I would like to see, is companies becoming more aware of real risk and benefit in how they are shaping the diet of the world. I know it’s an idealistic aspiration, but nonetheless one that I feel resonates with Soylent’s focus on evidence and transparency.

The problem with anectdotal evidence is that it is by definition misleading when applied to larger groups of people. When even the best formal studies can be misconstrued systemically or by bad practice, there is little hope in getting real epidemiological information from anecdotes. Of course, it can be important in the initial stages and for individuals to decide to get involved.


Clinical trials, blinding and randomizations can be extremely difficult things to pull off. However, there are no trials that don’t have weaknesses, and these, along with excruciatingly detailed and clever methodology is often what makes up the bulk of clinical trial design. The whole point is that for every intervention there is, in theory, a test that is the most fair possible, and that is what must be strived for in order to learn something new.

I’m not necessarily saying that it’s possible to pull off some sort of RCT on this kind of consumption, but that task should lie with academic institutions outside of this industry to ensure transparency and full disclosure of data. No trial is perfect, but the basic idea is really important here.


So create an excruciatingly detailed and clever methodology for Random Clinically Testing soylent.

The body has been a black box forever, and yet we’re still alive and living longer than before using the knowledge we’ve amassed. Using the current “best practices” to create a food product is common sense.

It just feels to me like we’re ringing the alarmist “but it’s not real food!” bell again.