Previously I talked about how bad alcohol is as a drug. If a lack of bioavailable fun was a medical condition and we put alcohol through safety testing, it wouldn’t get past its phase 1 clinical trial.
So how do we make its replacement?
A cocktail mimetic
Part of what makes alcohol so unique is the wide range of effects it has. It does a number of things at once, hitting a variety of different systems in your body: GABA, NMDA, dopamine and opioid receptors are all targets. Respectively these give you relaxation/loss of inhibition, disassociation, reward and pleasure. Any replacement will need to mimic these effects.
One option is to create a cocktail of other, safer drugs which bind to the same receptors. Something like:
alprazolam for GABA agonism. This gives alcohol its relaxing and disinhibiting properties.
Euphoric, fast, addictive, bad withdrawal effects once you’re dependent.
memantine or methoxetamine for NMDA antagonism. This is what gives alcohol its dissociating effects at higher doses.
Safeish; methoxetamine is a designer drug for people who love ketamine so much they wished it lasted longer.
tianeptine for opioid agonism. Alcohol doesn’t typically bind opioid receptors directly but it does trigger the release of endogenous opioids, making it blissful.
Also reasonably safe, although with addiction potential. Tianeptine is an atypical antidepressant that was thought to act by increasing the reuptake of serotonin (the opposite effect to SSRIs, go figure). Only later was it discovered that it binds mu-opioid receptors and nicknamed (overdramatically) “gas station heroin”.
nothing for dopamine. Dopamine release from alcohol happens because of the other effects, so this one can be left out.
In practice we actually don’t want to replicate all of alcohol’s effects. The dissociation you get from drinking heavily isn’t particularly nice, and is what causes all those people to end up in A&E from falling over when they’re drunk as a skunk.
Likewise with opioids. They’re addictive and are part of what makes alcohol addictive.
There are better options if we look at our replacement alcohol from this lens – of making a better drug rather than just a clone.
a less-addictive GABAergic drug. This is sort of a contradiction in terms; anything that binds GABA is addictive by definition. But there are some drugs in this class that are interesting to us.
bretazenil. Partial agonist at the benzodiazepine binding site, which gives a flat dose-response past some ceiling dose.
TPA023, TPA023B, MK-0777, L-838,417. These were trialled by Merck as an attempt for a non-sedating anxiolytic.
Etifoxene. Binds one subunit of the GABA receptor, and increases endogenous synthesis of neurosteroids like allopregnanolone which are themselves relaxing and sedating. Prescribed in france but safety record isn’t great.
Alcarelle. This is David Nutt’s attempt to make a better alcohol, which is going for GRAS (generally recognised as safe) approval in the US, but it remains to be seen whether it does anything.
NDMA antagonism: nothing, this is all downsides.
a less-addictive opioid. Also possibly a contradiction in terms, but there are some candidates:
buprenorphine. A partial agonist with a better safety profile; often given to opioid addicts or as a less addictive painkiller.
Mitragynine (kratom). Already used socially in an alcohol-like drink in parts of SE Asia.
Tianeptine, discussed above.
dopamine: nothing, the reinforcing effects are something we’re trying to avoid.
Pharmacokinetics
One thing that’s very hard to replicate is alcohol’s pharmacokinetics, or how it’s eliminated from the body. A bunch of liver enzymes break down foreign substances in the body, and the enzyme kinetics dictates how long a drug is in your body and how it’s eliminated.
Most drugs have first-order kinetics, meaning the rate of elimination varies as a function of how much drug is in your body. This gives you the classic exponential decay – these drugs have a half life, whereby every time the half life passes, half of the drug is removed from your body. (This is complicated of course by the fact that some drugs themselves are broken down into other drugs, which have their own activity and half life which can vary from the parent compound.)
Alcohol’s elimination has zero order kinetics, meaning that a constant amount of it is removed per unit time. This gives the old heuristic that roughly one unit of alcohol is removed per hour (a unit is about half a pint of beer, or a single 25mL shot of 40% spirits). This is because one of the enzymes that breaks down alcohol, alcohol dehydrogenase, is saturated at every low amounts of blood alcohol. You can imagine people trying to get through the checkout at a supermarket – once people start queuing, you just get a constant number of people going through per hour. This is unusual in the drug world, and the substances above don’t match it.
Delivery mechanisms
Alcoholic drinks are sipped, not taken in a pill. The above are all water-soluble so you could add them to a drink, although it would probably taste digusting. But alcohol’s pharmacokinetics also mean that you can dose it quite easily; as you get drunker you sip slower (in theory; in practice as you get more disinhibited your sensibilities around not drinking so much start to fade).
Drugs with first-order kinetics don’t have the same property. Even worse, these drugs have different half-lives, so some of the effects will last longer than others – making dosing them by gradually sipping even more difficult.
Ideally you want a combination with onset within single-digit minutes which then plateaus and is gradually eliminated. You get the buzz, while also being able to dose yourself somewhat accurately.
Benefits
What’s the point of doing this?
Well, our theoretical cocktail of drugs would have a much better safety profile. Alcohol has a notoriously low therapeutic index – the gap between the amount you need to have fun vs the amount you need to die is very small, and this is why people get alcohol poisoning and vomit in their sleep and die. But the TI of our cocktail is only as good as the worst one in it.
It’s also much less carcinogenic, one of the huge causes of the disease burden of even moderate drinking. Once thought to be good for you in small amounts (think a small glass of red wine), alcohol is now considered to unsafe at any dose. We can improve on that enormously.
We have an antidote for some of these drugs. Overimbibe on alcohol and there’s not much you can do except wait out. Take too much benzodiazepine and you can be given flumazenil, a competitive antagonist at the same binding site, already on the crash trolley. Maybe you could even have a pill you take at the end of the night which switches off our mix and lets you get home safely without risking falling over and splitting your head open. Remember kids: drunk walking is just as dangerous (for you) as drunk driving.
No hangovers! Alcohol is metabolised into toxic acetaldyhe and formaldehyde, which is why the next day is so rough, with nausea and a splitting headache and generally feeling incredibly rough. Ours would have none.
And who could forget: alcohol is full of calories! Ethanol has 7kcal per gram, almost as bad as fat. Our cocktail has zero.
In the next instalment: can we skip all this, and design a better drug from scratch as our alcohol replacement? And if we can, how do we get it through clinical trials and into the hands of the public?
I had a professor once. Very short, skinny, Korean man. In his 60s. Somehow he was able to drink like a fish, get totally wasted, and go back to work the next morning like nothing had happened. He said that he limited it to "just once a week" now, but when he was younger he used to do it 2 or 3 times a week.
My guess is that he must have been using something like this: https://www.theguardian.com/world/2025/sep/28/south-korea-hangover-cures-booming-market-traditional-novel-remedies-k-cure But I don't know exactly what, or how it worked, or how he lived. But there is definitely a place for this sort of medicine to "cure" alcohol...
I've taken memantine and n=1 but for me it doesn't feel fun. At 10mg/day it's totally not noticeable, at 20mg/day I feel woozy but not euphoric, like I'm going to keel over all the time.
OTOH 400 mg of oral bioidentical progesterone *does* basically make me feel drunk for a few hours without the hangover!
Problems with prog:
- it's expensive, requires a prescription
- kind of a wildcard hormone and doesn't have consistent effects from person to person. side effects basically. can mess up the menstrual cycle.
- long story but from asking claude looks like there could be issues with GABA-A receptor adaptation if you do it too often.