Valkee's claim of a "really peer-reviewed" study:
Did a placebo-controlled imaging study really show that the human brain is activated by light?


Background:

Functional Magnetic Resonance Imaging (fMRI) can explore functional connections between brain areas.80 Resting state fMRI scans are used to find differences in brain function, for example between sick and healthy people. The Valkee group has published some articles about fMRI findings in Seasonal Affective Disorder, not mentioning the earlight device.81 82

These laboratory studies are generally not able to confirm any treatment effects. They can find candidate structures responsible for functional differences between individual brains. They are not testing a hypothesis, like "light is activating the brain". The scan produces data of unclear content, which is then screened for patterns. Patterns will always be found in a large pool of data if the search is extensive enough. You can declare success for any arbitrary thing you found, if you don't tell what you are looking for before the search starts.

Valkee used scanners, lab devices and patients from legitimate research to produce the following results. It announced this study, like others before, as a peer-reviewed proof for the earlight's effects, as an important step for the company. 14


How they did it:

Valkee did scan 51 healthy persons' brains at the resting state. No before-after changes in individual brains, no reactions to light were tested, according to Valkee for technical reasons. Instead, the "as is" state was scanned in groups: In December, the brains of 10 participants underwent imaging with earlight twice the strength of the commercial device. In May, the brains of 14 other persons were scanned, without earlight present. There were sham and earlight sessions in February, for 27 persons. 9 persons underwent the scan with earlight switched on and off.

50 scans were analyzed and searched for differences. 26 scans were done without earlight. 24 other persons had been scanned, partially at other occasions, with earlight. Therefore the groups differed in date of scan, and also in some baseline parameters of unclear relevance, before the analysis.

No significant differences were found in the 9 persons getting earlight on/off. The data for both volunteer groups was computed until one difference showed up, in a brain function cluster they called "lateral visual network IC". Valkee infered from this probably accidental finding, "that brain tissue is inherently light-sensitive". A previous study found 11 such differences between SAD patients and healthy persons.81

An article full of unnecessary technical terms was produced by the Valkee group, nearly unreadable for outside persons.


What does it mean:

Valkee scanned two groups of people at different time of year and with/without earlight. A reaction to the earlight was not detected. A small difference between the groups was possibly seen, in one of diverse tested brain functions. This difference could have been there already before, it is impossible to tell if there was an effect of the earlight.

In plain text - It is pointless, if one recrutes two different groups of people, and then scans the groups until they differ in any point. In Valkee's case, differences between the tested groups were even pre-existing.


How was it published:

Like all other earlight studies, this one was not published in a peer-reviewed journal. It was presented as a PDF file on a commercial website posing as a scientific publication.83 This company has been notorious for using names of respected researchers without their consent, and publishing copies of real articles, to appear like a scientific publisher.84 85 It charges salty fees for storing articles which are not accepted in scientific journals.86 A real quality check would undermine this business model. Thus, article authors name their own "reviewers" - the exact opposite of peer-review.87 92

The "World Journal of Neuroscience" WJNS is not indexed in the scientific databases like MedLine/PubMed, researchers cannot find the articles.93 Clients can request their own journal.88 This seems to be the case here: the first WJNS number consists of one single article.89 The Journal of the Finnish Medical Association classifies such articles as junk publications, which do not have any scientific merit.94 It suggested using the list on www.scholarlyoa.com, where "Scientific Research Publishing" is listed as not trustworthy.95

What was the impact:

Valkee announced this piece directly after their bust in Finland.14 They produced no finnish press release, given the low credibility they had now. An english press release about this "success" paper appeared on Valkee's website,90 to convince those who did not buy earlier fakes.91

Conclusion:
A junk study, published as a fake "peer-review" article in a questionable journal. It has no scientific meaning nor impact, and was just done for Valkee's marketing.

See also: The real placebo-controlled trial showed that the earlight does not work better than placebo.


References & Sources:

14. Juuso Nissilä: Interview w/ Radio Mega, 13.3.2012. MP3 on file.

80. Wikipedia: Functional magnetic resonance imaging. Accessed Sep 3, 2013.

81. Abou Elseoud A , Nissilä J , Liettu A , et al: Altered resting-state activity in seasonal affective disorder. Hum Brain Mapp. 2012 Sep 15. doi: 10.1002/hbm.22164. [Epub ahead of print]

82. Abou Elseoud A , Littow H , Remes J , et al : Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity. Front Syst Neurosci. 2011 Jun 3;5:37. doi: 10.3389/fnsys.2011.00037.

83. Scientific Research Publishing website, accessed Sep 3, 2013.

84. Sanderson K: "Two new journals copy the old". Nature 463, 148 (2010).

85. Wikipedia: Scientific Research Publishing. Accessed Sep 3, 2013.

86. World Journal of Neuroscience > Article processing fees. Accessed Sep 3, 2013.

87. World Journal of Neuroscience > Guidelines > Submission. Accessed Sep 3, 2013.

88. Scientific Research Publishing Website: "Call for new journal proposals". Accessed Sep 3, 2013.

89. World Journal of Journal of Neuroscience: Vol. 1, Number 1, May 2011. Accessed Sep 3, 2013.

90. Valkee authors: "Publishing[sic!] in the peer-reviewed World Journal of Neuroscience: fMRI study finally confirms that brain tissue is responsive to light". Press release, May 29, 2012. available at http://www2.valkee.com/uk/about-us/press/

91. Anderson G: Valkee Study Published In Peer Reviewed Journal? Arctic Startup website, February 14, 2012. Accessed Sep 3, 2013.

92. Wikipedia: Peer review. Accessed Sep 6, 2013.

93. National Library of Medicine catalog entry: World Journal of Neuroscience. Accessed Sep 3, 2013.

94. Lautala T: Open access vaatii tutkijalta valppautta. Suomen Lääkärilehti 34/2013, p. 2026-2028 (subscribers only, Sep 3 2013)

95. Beall J: List of (...) predatory publishers. Website Scholarlyoa.com, accessed Sep 3, 2013.


Main page

The Valkee Story

Valkee's research

Valkee Quotes

FAQ

about this site

Contact

The GLOOM Blog.