Essentially, this is a first study with a clear-cut TSM protocol that was done properly (blindly and with a reasonably large number of participants).

Abstract here:
http://www.ncbi.nlm.nih.gov/pubmed/23237314

The main result is here:
http://imageshack.us/a/img442/1614/tsmstudy.png

Basically, it works but it does not seem to be the miracle cure it is frequently made to be around here. Two other things notable: 1) the placebo effect is bigger than the drug's effect and 2) there is a very significantly higher drop out rate for nalmefene users.

Anecdotally, I've been on TSM for >2 years with very good but not spectacular results - about 3-fold reduction in units at the peak of efficacy, about one year on. Unfortunately, I've slowly but steadily regressed and eventually started to drink heavily again all the while taking naltrexone religiously. I am off TSM for the past 6 months but consider trying to do it again and to see what happens.

Hopefully, they follow up so that they can go well beyond 6 months. Theoretically, the Nalmefene users would continue the downward trend, and the placebo would tend to return to baseline. But, I'd agree that TSM would seem to have a high placebo effect as one is taking a pill and charting one's drinking in the hope that it goes down. This continual effort to reduce further and further downward would be powerful in and of itself. What I've experienced, though, can't possibly be a placebo effect.

Why can't it? Of course it can. If placebo can occasionally cure cancer, it certainly could occasionally do wonders to substance dependence.

My $0.02: TSM or no TSM, there is an overwhelming evidence now that naltrexone/nalmefene reduces urges and/or binge drinking. "Reduces" is a key word. There is definitely a big variability as to how well people respond to naltrexone and almost certainly most of it is genetic in nature (but very complex). Anecdotally, after almost three years of observation here, my feeling (unsubstantiated by real numbers) is that it is people like you, relatively moderate drinkers, who stand best chances of of being helped by naltrexone in the most dramatic way. Few papers also suggest that women may be less likely to benefit. It could be self-selection bias of course, but it seems to me that yes, many more self-proclaimed "cured" are men.

Well, I suppose it's theoretically possible that it's placebo. I was basing that statement on the intense, undeniable physical feeling I get when I take Naltrexone (which, again, theoretically could be placebo) and the notable difference it makes in the way I experience alcohol when I drink it. But, scientifically speaking, I could never prove that it's not a placebo effect, so I'll grant you that.

RESULTS:

Patients taking placebo (n = 289) and patients taking nalmefene (n = 290) were included in the efficacy analyses. At Month 6, there was a SIGNIFICANT effect of nalmefene compared with placebo in reducing the number of heavy drinking days (-2.3 days [95% confidence interval:-3.8 to-.8]; p = .0021) and total alcohol consumption (-11.0 g/day [95% confidence interval:-16.8 to-5.1]; p = .0003). Improvements in Clinical Global Impression and liver enzymes were larger in the nalmefene group compared with placebo at Week 24. Adverse events (most mild or moderate) and dropouts due to adverse events were more common with nalmefene than placebo. The number of patients with serious adverse events was similar in the two groups.

Hi, Nutella, and thank you for posting this.

I'm curious as to why you say t"he placebo effect is bigger than the drug's effect" Both the Results and the Conclusions say the opposite... ?

It doesn't work for everyone is my observation.

It was a miracle cure for me.

On top of the above I have a strong feeling certain 'traits' in certain people make them more likely to succeed - it's not a med that works without individual input.

@CuriousOne:

Re: Placebo effect is larger than the Nal's effect.

Look at the Figure 1 of the paper. At the end point of the study, placebo induced reduction of alcohol consumption of ~ 35 g/day (about two American drink-units) while in Nalterexone group the number is roughly 50 g/day (about 3 units).

All I meant to say is that 2 > (3-2)

You're reading the chart wrong. That's a reduction in drinking, so a 3 drink reduction is a larger effect than a 2 drink reduction

Hey Nutella, thanks for bringing this paper to everyone's attention! I love getting to the science of this whole thing. As much as I believe in naltrexone therapy, there is more to it, and I agree that the promises of TSM are overblown.

Anyway, for anyone reading, I'm a professional scientist. It's part of my job to critique studies like this. This is actually a pretty good paper. Unfortunately Nutella has kinda misunderstood the main finding, so let's go through it



This is mostly correct. In fact the study was even set up so that the patients take the drug 1-2 hours before drinking. The main difference between this study and TSM (or whatever version of TSM that we all employ) is that ALL study participants, including those on placebo were given "motivational and adherence-enhancing intervention to support them in changing their behavior and to enhance adherence to treatment." This kind of counseling could explain a large portion of the placebo effect seen.



This is simply incorrect. The graph reflects net change in drinking, so a lower line means less drinking. Nal was MORE effective than placebo at helping control drinking, as stated in the abstract.




The difference is very significant statistically. "During the main treatment period, 91 (31%) of the placebo-treated patients and 160 (53%) of the nalmefene-treated patients dropped out of the study." So there is a clearly higher dropout rate in the nal group, but a really high dropout rate in the placebo group too. I only mention this so that people don't think that everyone on nal is doomed to stop complying.

Compliance is something that I am starting to get VERY interested in, so I earnestly thank you for bringing this paper to my attention. Many people around here think or state that they are following the system, but have frequent cheat days, or "want to see what it's like to drink without nal again," or have times that they just "can't handle the side effects." In my reading and my experience, 90% compliance does not give you anywhere near 90% of the effect. Letting you brain associate the endorphin rush with the sauce, even if it is only 5% of the time, seems to have a hugely deleterious effect on people's progress.