Martin Lindstrom, I think you could be dangerously wrong!

There is an oped in the NYTimes today about cigarette labels. The author, Martin Lindstrom, talks about graphic labels on cigarette containers and their effectiveness for encouraging people to quit smoking. He concludes that they don’t work. Guess why? Functional imaging of course…
We’ve been down this road before, extraordinary claims based on an fMRI experiment. Mr. Lindstrom says the following about his fMRI experiment:

A brain-imaging experiment I conducted in 2006 explains why antismoking scare tactics have been so futile. I examined people’s brain activity as they reacted to cigarette warning labels by using functional magnetic resonance imaging, a scanning technique that can show how much oxygen and glucose a particular area of the brain uses while it works, allowing us to observe which specific regions are active at any given time.

We tested 32 people (from Britain, China, Germany, Japan and the United States), some of whom were social smokers and some of whom were two-pack-a-day addicts. Most of these subjects reported that cigarette warning labels reduced their craving for a cigarette, but their brains told us a different story.

Each subject lay in the scanner for about an hour while we projected on a small screen a series of cigarette package labels from various countries — including statements like “smoking kills” and “smoking causes fatal lung cancers.” We found that the warnings prompted no blood flow to the amygdala, the part of the brain that registers alarm, or to the part of the cortex that would be involved in any effort to register disapproval.

To the contrary, the warning labels backfired: they stimulated the nucleus accumbens, sometimes called the “craving spot,” which lights up on f.M.R.I. whenever a person craves something, whether it’s alcohol, drugs, tobacco or gambling.

Now, to his credit, he does say that more work is needed:

Further investigation is needed, but our study has already revealed an unintended consequence of antismoking health warnings. They appear to work mainly as a marketing tool to keep smokers smoking.

I think that I would like to question his primary claim which is that an increased BOLD signal (or “lighting up” as he calls it) in the nucleus accumbens is somehow indicative of engagement of the reward pathway in these people. This increased BOLD signal is almost always taken as an indication that excitatory transmission is happening. In this case, they are probably thinking dopamine release in the nucleus accumbens. This idea comes from the fact that excitatory transmission requires energy so there is more blood flow and more oxygen consumption happening in the area. What some of these researchers neglect to tell you is that inhibition also requires a great deal of energy consumption. Why you ask? Well the answer is simple. When GABA-A receptors are activated (at least in the adult CNS) chloride flows into the cell. In order to keep the chloride gradient intact (and inhibition going) the neuron has to get rid of this chloride load. To do this, it pumps out chloride, largely through a potassium, chloride transporter called KCC2. Now this transporter is electroneutral; however, in order for it to work, our old friend the sodium-potassium ATPase has to bump up its activity. This requires a good deal of energy, energy expenditure that would likely contribute to a BOLD response. Wanna know more, check out this great review in Neuron that is freely available on PMC.

So, the NYTimes editorial says that the graphic warnings are causing reward. I say not so fast, they may actually be inhibiting the accumbens via an increase in GABAergic activity. Either of us could be right but in the absence of pharmacological data, I will say the jury is out. Like so many other fMRI-based studies, this one is certainly interesting but it is completely open to interpretation and mine is quite the opposite of the author’s in this case. Finally, what about the evidence that these graphic labels actually work? I think we should not ignore these findings all because of one fMRI experiment. Moreover, the limp cigarette Canadian label is totally awesome!

UPDATE: To be clear, the mechanism that I cite (Na+-K+-ATPase) is only one of many mechanisms that may lead to a BOLD response via an increase in inhibitory neurotransmission. I find these lines from the Neuron review particularly intriguing:

Here, it is interesting to note that observations in normal human subjects using PET have shown that administration of the specific GABAA receptor agonist, 4, 5, 6, 7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP; an agonist known to induce a tonic GABA conductance) leads to an increase in glucose metabolism, even though clinical and electroencephalographic monitoring showed a sedative effect and sleepiness after drug administration (Peyron et al., 1994a). In temporal lobe epilepsy patients, the THIP-induced increase of glucose metabolism was highest in the electrically hypoactive focus (Peyron et al., 1994b).

9 responses to “Martin Lindstrom, I think you could be dangerously wrong!

  1. I don’t know enough about cognitive neuroscience to know if this makes any sense, but could it still be activation of the craving center and be useful in getting people to not smoke?
    Handwaving pop-neuroscience talks about coupling certain neural pathways- an easy to understand example is coupling “this smell and this memory”. Could the brain be linking “this craving and this feeling/thought” (the feeling/thought presumably being negative)? Or is that completely ruled out from the lack of amygdala and cortext involvement?

    Did they do the control experiments in non-smokers?
    And they must have done the control of having subjects in the MRI machine with random sentences and pictures of cigarettes, right?
    Because otherwise, it’s not the warning labels that are advertising for the companies, but the researchers!!!

  2. Fuck me! Now they’re using the fMRI fake-ass phrenology shit to actually make decisions about public health policy!?!?!? AIEEEEE!

  3. Becca,
    I have no idea what the controls are but it doesn’t really matter. As CPP notes, they are using too little information to urge a public health policy decision. My point was that even if the result had adequate controls, it still tells you nothing about whether the people were “craving” or not. “Lighting up” the nucleus accumbens does not have to mean activation in the sense that the author assumes. In fact, the subjects reported the opposite. Why not just believe what they said rather reaching some conclusion about the fMRI data that is contrary to their words.

    That review should be manditory reading for all fMRIers, as far as I am concerned.

  4. Oh Juniorprof- I’m totally with you on the “science and policy – their doing it wrong!” verdict. I was just curious about how this stuff worked.

  5. wow, that’s a singleminded interpretation in that article. he didn’t subject his study to much analytical rigor, did he?

    i do wonder whether simply presenting smoker subjects with the idea of a cigarette is enough to cause a craving. (ex-smoker husband does not like to talk about cigarettes, it makes him want one.)

    seems to me proper controls could have helped to make that clearer.

    i don’t have time to read the article you linked right now, but i will come back to it because it looks interesting.

  6. I did not bother to read said article or original research paper. That said, fMRI is still data. How it is interpreted is a whole ‘nother matter. I agree that it is not necessarily activating a “reward” signal but nevertheless, there are changes in the nucleus accumbens when the images are presented. For whatever that’s worth.

    Another interpretation of the above might be that indeed, any kind of association with cigarettes, even the mention of the word or something associated with it (however “graphic”) still elicits a drug-anticipatory response. This is somewhat along the lines of what the op-ed guy was saying, I guess.

    Second, could it be that graphic cigarette labels do not work simply because to a smoker, they cannot override the actual reward sensation/association of cigarette smoking? For example, cortical and cognitive networks vs. classical mesolimbic reward pathways. I know so many people who don’t give a crap if smoking KILLS, they just like smoking. They should know better (medical doctors, for example) and yet they still chain smoke. I am waiting for Drugmonkey to chime in here.

  7. Could the brain be linking “this craving and this feeling/thought” (the feeling/thought presumably being negative)? Or is that completely ruled out from the lack of amygdala and cortext involvement?

    Becca, in terms of interpreting fMRI data, I always take it exactly at face value. If an area “lights up” there is an increase in blood flow and oxygen consumption vs the control condition (generally just a baseline, sitting in the machine doing whatever you do while sitting in a machine). If a brain area doesn’t “light up” it may or may not be active. There is not enough information gathered from your standard fMRI to know one way or another. So, their conclusions about amygdala, etc., don’t sound right to me. Absence of evidence will never be evidence of absence.

  8. These pieces really annoy me. There’s a certain type of science piece (particularly using fMRI) that is of such a bad quality that it would never even reach the NY Times science section, but the op-ed pages seem to love to publish them.

    This op-ed is neo-phrenology plain and simple. Any piece that bases an argument on “The amygdala is the part of the brain that registers alarm” is garbage. Even if the craving /nucleus accumbens theory was right, you show anything related to an addition, it will induce craving. The question is whether there is more craving from a box of cigarettes with a picture of a camel or a box of cigarettes with a skull and crossbones and a poison warning.

    At the most basic level, I thought that the rates of smoking ARE decreasing around the US and even around much of the world. The piece starts with the nice slight of hand by using raw population numbers instead of proportions of populations.

  9. Hi there, people from Liverpool John Moores University. Feel free to join in the discussion, or fill me in on what you are discussing about this post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s