Valkee Ltd seems to have completely abandoned its SAD claims, now the device is called “HumanCharger” and works – clinically tested! – on jet lag. Marketing bubbles re-used: They only changed some words in their PR stuff, the rest is the same as for SAD. The same unproven claims.
Needless to say, these things were never tested. Valkee logic: Because it is approved for SAD, and those are SAD symptoms, it works on these. Therefore, it does the same for jet lag users!
The whole jet lag campaign is based on a study, which appeared already on April 1st in the journal Aerospace Medicine and Human Performance. Valkee says that this proves a jet lag effect. Its SAD studies are all on the Valkee website, this one remained behind the paywall. Typical Valkee users and marketing folks don’t need the facts and are better off without it. To pay 30 Dollars for a Valkee paper? Simply believe!
Update 12.6.17:40: The study is now available from Valkee HumanCharger, direct download from the site. It is not linked yet from any of their sites, and not found neither by google nor google scholar. It was added in the meantime to a press release. The leak for this post was several weaks ago.
Now one of Valkee’s little helpers did not get the memo, and leaked the document into the public domain. I can mirror it here without doing something illegal:
The full text of Valkee’s jet lag trial (4,6 Mb)
I do not have the spare time to comment, but the main outcome is, that the device did not work on practically all things measured. No significant difference at the end of the predefined treatment period for 10 of at least eleven tested scores. Only one subscale of the POMS brought a significance, and this can be explained with multiple testing.
The funny point: It was the fatigue subscale. Exactly the same single result, as the homeopathic “remedy” No-Jet-Lag got in its own company trial.
The difference: The homeopathic treatment is to be taken orally, during the flight. The Valkee device must be used for 6 (six!) days after the flight, 4 (four!) times a day to get to the same result. 24 treatment sessions.
Valkee is nearly as good as the leading homeopathic treatment.
***
Addition: The analysis was done on 52 patients, but the article speaks of 55 (Figure 3). What happened to the rest? Drop-outs happened, but are not reported. The whole statistics are invalidated. Incredible that this got through peer review.
It says on the article: “Up to four doses of TBL per day significantly reduced overall posttravel jet lag symptoms as well as subjective feelings of fatigue, inertia, and forgetfulness. On average, subjects in the TBL group showed a greater rate of overall recovery based on their VAS scores than subjects in the placebo group.”
You say: “Only one subscale of the POMS brought a significance”.
Can you explain?
Yes.
The predefined treatment period was 6 days post-travel. On day 6, only the POMS fatigue subscale had a significant difference. It is inconsequential, if there was at some point in between a difference, as a sufficient study design always has its pre-defined protocol for outcomes. Otherwise, it would be an exploratory approach (“ok, let’s look when there will be a difference”). That’s then not confirming, but hypothesis-generating at best.
Valkee knows that very well and tries to blur the facts in the article, e.g. by making up the “+”-mark for non-significant results in the other subscales (figures 2x), or by claiming a significant difference in the text – without telling that it was not the final outcome.
The “recovery analysis” is an awesome piece: They define a VAS below 2 as recovery (why? why not 1 or 3?) and then present the fancy column graphic (fig. 3). If one counts the numbers, it is done on 52, not 55 patients. That means, it is not an intention-to-treat (ITT) analysis, or they deliberately conceal the results for 3 persons. Adding the worst-case scenario (2 more fails in the TBL, and 1 more success in the placebo group) eliminates the TBL-placebo difference.
At least, they should have reported the dropouts. So, it is just a smoke screen.