Mental Suggestion
If ever there was a good example of why health journalists should seek to gain a proper understanding of the material they are writing about it is this:
Study finds vitamins boost mental health
“People with mental illness made “remarkable” improvements by taking a daily dose of nutritional supplements rather than conventional medicines, a trial has found.
“The work by a Canterbury University clinical psychologist has shown the potential that consumption of the right micronutrients, such as vitamins, minerals and amino acids, could have for helping a range of mental health problems.”
Cue hordes of depressed people rushing off to the vitamin store to cure their depression…
There is only one problem with the above opening lines. It gives entirely the wrong impression. Of course the journalist who wrote this can claim that this impression is erroneous. After all, he carefully used weasel words such as “potential”, “could have” and “range”. But there is no doubt that the opening line supports the heading. And the heading is simply claptrap.
Firstly, this is a tiny trial of 14 people, so it is hardly a definitive study. Compare that to the literally tens of thousands of people involved in trails of medications such as Prozac.
Secondly, the mental illnesses studied here were ADHD and Severe Mood Dysregulation (similar to but less severe than bipolar disorder). Both of these are usually paediatric diagnoses although this trial was with adults. Presumably, they all therefore had long-standing diagnoses. Both of these problems are unusual in that have been shown in other studies to be strongly related to diet. It is therefore dangerous to drawn any conclusions from this study about other mental illnesses. I was particularly concerned to read this:
““Most of the individuals were in a moderate to severe depressed state at the start of the trial,” Dr Rucklidge said..
““At the end of the eight weeks, the mean score on the depression measure fell in the normal non-depressed range, which is a fairly remarkable change in such a short time, especially as many had not experienced such improvements with other conventional treatments.”
This is exactly the sort of statement that should be avoided. It gives the impression that vitamin therapy is an effective form of treatment for depression and yet there is no evidence for this beyond Dr. Rucklidge’s tiny study. It is well within the bounds of possibility that both the placebo effect and the simple act of engaging in continuing assessment of these people lead to their improvement. Yet a mental health patient will read this and take themselves off their medication to “try vitamins”. They will not ask their GP first (trust me on this), they will simply do it.
Guess who will be cleaning up the resultant mess. Not Dr. Rucklidge.
When will doctors stop promoting inconclusive, often unpublished studies and journalists learn to ignore self-serving clinicians?
Related posts:
- Vitamin Victims? Here is a good illustration why newspapers desperately need proper...
- Mental Problems Nick Smith has reiterated that ACC has become another branch...
- Attractive Option Healtheries is recalling a batch of children’s vitamins after metal...
- Accidental Rape? The battle between counsellors treating sexual abuse victims and ACC...
- Chronic Fatigue Syndrome Chronic Fatigue Syndrome sufferers the world over will be very...
Jan 22 10 8:26 am
It was apparently an open label trial ( victims know what they are getting ) – not sure if placebo effect would apply, but certainly extra TLC would be difficult to account for. Is she a medical doctor, or just a Ph.D. in psychology?.
Jan 24 10 7:15 pm
Open label makes placebo effect more likely as people know what they are getting. There was no placebo control in this trial at all. Rucklidge is a Ph.D, not a medical doctor.
Jan 22 10 9:34 am
I’ve corresponded with a few health researchers whose studies have hit the press, with the press making bold causality claims unjustified by the underlying study. One answer I’ve gotten is that the journos are following the press releases, and that the university’s (not Canterbury in this case) press agent told them to sex up the causality claims in the press releases or there’d be zero press interest.
This is disgusting.
The press exists to sell papers. Most readers don’t care about reports of studies that have all the caveats – they want bold claims or nothing. So the papers are only interested in studies that can be presented as conclusive. Some researchers (and more importantly their employers) care a lot about positive press coverage for their work. And so stuff gets tarted up.
I’m increasingly of the view that it’s not that journalists don’t understand the difference between causality and correlation but rather that they specifically don’t care because their readers don’t care. I’d wondered for a while whether it would be worthwhile putting together some kind of one-day summer thing for journos: what you need to know to understand social science research (basic stats, causality/correlation). But the readers wouldn’t care, so why should they?
Eric Crampton´s last blog ..Reasons not to live in Australia
Jan 24 10 7:16 pm
Cynical, Eric – but distressingly true.
Jan 22 10 3:02 pm
“When will doctors stop promoting inconclusive, unpublished studies and journalists learn to ignore self-serving clinicians?”
When they have an understanding of the processes and rigour behind good scientific research, which, regrettably for many of them, will probably be never.
Jan 24 10 4:52 pm
Hi, I’m Julia Rucklidge, Associate Professor of Psychology, the researcher behind the micronutrient study you blogged about.
You write my study is “unpublished”. Actually, it is published in the Journal of Attention Disorders. http://jad.sagepub.com/cgi/content/abstract/1087054709356173v1. We have also published on this topic in CNS Spectrums, Science, Expert Review in Neurotherapeutics, and Journal of Anxiety Disorders. I am happy to provide you with all these references.
Of course other researchers have been publishing on this topic for decades. There are RCTs that show benefit of micronutrients in the treatment of symptoms associated with Alzheimer’s (Remington et al., 2008), behavioural problems in children (Schoenthaler and Bier, 2000) and offending behaviour in incarcerated populations (Gesch et al., 2002). Multivitamin/mineral supplementation has also been shown to improve nonverbal IQ in normal children by up to 8.3IQ compared with placebo (Benton and Cook, 1991).
You write there is” no evidence” that vitamin therapy is an effective form of treatment for depression beyond my tiny study. Actually, vitamin and mineral therapy has been undertaken for more than 100 years, as documented in this review of the literature: Kaplan, B. J., Crawford, S. G., Field, C. J., & Simpson, J. S. (2007). Vitamins, minerals, and mood. Psychological Bulletin, 133 (5), 747-760.
You also should know that if a “tiny” study finds an effect, then we know that the effect is large. It is far more worrisome when researchers have to sample thousands, in order to detect the effect.
When will medical bloggers read the peer-reviewed scientific literature before writing blogs?
Jan 25 10 1:58 am
Thank you for replying to my post. I had already realised your study had been published but by that time Projectman had already commented on that part and I could not properly alter it to reflect that. I did, however read through your study. I am also well aware that vitamins and micronutrients are important for childhood development and that dietary changes can seriously effect our moods. However all this is a far cry from actually treating depression with diet alone. And no where in the Kaplan article does it assert that micronutrients should be considered as a treatment for depression based on the literature available. I do concede that the matter is worth studying further.
May I respectfully suggest that you read a decent book on statistics. Your assertion that “if a “tiny” study finds an effect, then we know that the effect is large” is complete nonsense mathematically. Consider throwing ten coins in the air – it is feasible that all 10 may come down tails. This is not a “large effect”. The chances of this happening randomly are 1:1024 – not very large odds. However, if I took a thousand coins and threw them, the chances of me getting a thousand tails are astronomically small. The result of a thousand coin tosses is far more likely to show that the probability of a tail is a mere 50%.
I’m afraid that, statistically speaking, your study is little better than anecdotal evidence. This is not to suggest that it is not valuable. It is just to point out that you cannot draw the conclusions you appear to suggest in the Herald article. All we can say from your study is “That’s interesting – we should do a randomised trial with more people”. And that is more than enough to justify your work.