Alcohol, bad science, and sophistry

In an article in the Guardian today, Professor Nutt made a series of statements about alcohol that are deeply depressing, and not for the reasons one might perhaps think. They were as follows, with the reasons for despair after each:

"Alcohol is a toxin that kills cells such as microorganisms, which is why we use it to preserve food and sterilise skin, needles etc. Alcohol kills humans too. A dose only four times as high as the amount that would make blood levels exceed drink-driving limits in the UK can kill. The toxicity of alcohol is worsened because in order for it to be cleared from the body it has to be metabolised to acetaldehyde, an even more toxic substance. Any food or drink contaminated with the amount of acetaldehyde that a unit of alcohol produces would be immediately banned as having an unacceptable health risk."

Where to begin with the tawdry idiocy of this? Perhaps simply replacing 'alcohol' with sodium chloride (or table salt as the less pretentious apparently refer to it) is as good a place as any, giving us - "Sodium chloride is a toxin that kills cells such as microorganisms, which is why we use it to preserve food and steralise skin, needles etc [not so much the second, but salt water is a good, if nippy, way to steralise a wound]. Sodium chloride kills humans too." But only if you give them it in doses that are vastly higher than that which one would normally consume in or with one's food. Alcohol is a toxin and acetaldehyde is nasty, but at the doses that one normally consumes alcohol our body is quite clearly able to process them. What Prof. Nutt has done here is analogous to the famous case in which people were convinced to try and ban dihydrogen monoxide (H2O) because of the dangers it posed. That he is trying to pass off the equivalent of such a prank as serious comment is tragically disingenuous. 

Like all the best lies, each sentence has an element of truth to them (or, rather, they are all true, but do not apply to the context that he is putting them forward in) - if one was not slightly educated in science then one would probably read them and accept them as entirely reasonable. Somebody who has been given the gift of an intellect and education that has taken them to the top of academia should not be abusing the ignorance (used in an entirely non-pejorative sense) of others in this way.

"Although most people do not become addicted to alcohol on their first drink, a small proportion do. As a clinical psychiatrist who has worked with alcoholics for more than 30 years, I have seen many people who have experienced a strong liking of alcohol from their very first exposure and then gone on to become addicted to it. We cannot at present predict who these people will be, so any exposure to alcohol runs the risk of producing addiction in some users."

A supposed scientist putting forward an anecdote as evidence, putting him in the esteemed company of creationists, militant anti-abortionists, homeopaths, et al. Not only that, but his anecdote and surrounding statements are what can only be described as bullshit (cf. Frankfurt, 1986). 

Firstly, nobody and nothing (with the usual caveats about freak outliers) has become addicted to anything upon their first exposure. They may have become dependent upon their first exposure, but they did not become addicted. Again, most people may not know the difference, but as a pharmacologist Prof. Nutt most certainly does. Glossing over that, yes, a great many of people who liked alcohol the first time they tried it went on to become addicted. But a great many that liked it didn't go on to be addicted. And a great many who hated it the first time will have gone on to be addicted, and some who liked it a little, although perhaps it could have done with a little more ice, will have gone on to be addicted (and some not). None of this is relevant to anything, other than to a description of all the possible combinations of liking something the first time (or not) and getting addicted to it (or not). It is rhetoric, nothing more.

Does any of this matter though if the outcome for the individual is the same? Well, yes. Politicians and journalists may bend and abuse words to fit their needs and blind people into doing what they want them to be doing, but we should - nay, must - be able to expect better from a member of the academy. Being a scientist and an academic is to a small part of a great cooperative entity striving towards truth. When one member of this effort sullies this enterprise by resorting to lies and half-truths in this way it belittles all the other members and moves us all as a species that little bit further back into the darkness.  

"The supposed cardiovascular benefits of a low level of alcohol intake in some middle-aged men cannot be taken as proof that alcohol is beneficial. To do that one would need a randomised trial where part of this group drink no alcohol, others drink in small amounts and others more heavily. Until this experiment has been done we don't have proof that alcohol has health benefits. A recent example of where an epidemiological association was found not to be true when tested properly was hormone replacement therapy. Population observations suggested that HRT was beneficial for post-menopausal women, but when controlled trials were conducted it was found to cause more harm than good."

More misrepresentation here. The study that I assume that he is referring to here (Rossouw, 2002) did find an increased risk of breast cancer, heart disease, stroke, and pulmonary embolism in subjects who were more than five years post-menopausal; but it also found a reduced risk of colon cancer and hip fracture in this group, as well as a reduced risk of heart disease for those subjects who were less then five years post-menopausal. So, as with most science, the results were complicated and do not in their entirety support the claim that he makes. If one is to go around cherry-picking results that suit then one could equally just chose the results from this trial that refer to heart disease in women that are less than five years post-menopausal and say that HRT definitely does more good than harm. 

Almost as depressing is his willful misrepresentation of the scientific method (I can't bring myself to believe that he might genuinely understand it). Certainly it is true that a (double-blind) randomised controlled trial would be the best method of establishing whether alcohol has health benefits or not, but to claim that we have no proof prior to this is just silly. The trials that have been done so far tell us something - not everything, but something - within what will have been carefully delineated set of limitations. They do not give us proof (in a strong sense of the word), but nor would the randomised trial that he says is necessary before we can say anything. All of which brings us into the domain of questions about truth and science, an area that I suspect he is not seeking to take his argument into (especially given that many if not all of the trials that he will say show that dangers of alcohol will not meet the artificial standards of proof that he has set out here).  

Finally, implying that there is only evidence for health benefits for alcohol intake in middle aged men and in the context of cardiovascular disease is unashamedly dishonest. The conditions for which alcohol has been suggested to have a protective effect range from stroke (to the common cold (Takkouche, 2002). This is not to say that these results are not complex and controversial, nor that alcohol consumption does not cause serious health problems - both of these things are obviously true, it is just that the implicit claim that Prof. Nutt makes quite obviously is not. 

For all other diseases associated with alcohol there is no evidence of any benefit of low alcohol intake – the risks of accidents, cancer, ulcers etc rise inexorably with intake.

A final weaselly sentence. No, there is no evidence that low doses of alcohol are of benefit in liver cirrhosis - but no-one in their right mind would suggest that there was (although that doesn't prove anything - has an experiment been done?). It takes a fairly close reading of the sentence to see what he is doing here though. By definition diseases associated with alcohol are not helped by alcohol consumption, but that says nothing about those diseases that are not associated with alcohol. As to the accidents etc, a whole series of questions arise if one stops to think about what he is trying to say. Does that risk of accidents increase with intake in an acute sense or accumulitively over ones life? The first is quite obvious, and isn't saying much at all, whilst the latter is patently absurd. What does he mean by ulcers? Duodenal? But as every biology undergrad knows, such ulcers are to an extremely high degree caused by H. pylori and not diet. Again this is writing that one would expect from a politician or a tabloid journalist - half truths, obfuscation, and the exploitation of lay ignorance.

This article, then, is both sad and dangerous. It is sad because it is evidence of somebody who should know better sullying the whole enterprise of science for measly rhetorical advantage. It is dangerous for this reason too. People are not stupid, and some at least will see through the deceptions. When they do, and when they see Prof. Nutt producing them from his position as an expert, they will grow further distrustful of all those who make any claims from a position of expertise. In doing so we move further towards a world where learning and work are utterly devalued, and where people are entirely unwilling to accept guidance from those who have knowledge, relying instead on instinct or the word of those who can talk the loudest or the smoothest. Such a world would drive one to drink.


References
Frankfurt, H. On Bullshit. Raritan Quarterly Review 1986; 6(2)

Rossouw JE, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trial., JAMA. 2002; 288(3):321-33.

Emberson, JR.; Bennett, DA. Effect of alcohol on risk of coronary heart disease and stroke: causality, bias, or a bit of both?. Vasc Health Risk Manag. 2006; 2(3): 239–49 

Takkouche B, et al. Intake of Wine, Beer, and Spirits and the Risk of Clinical Common Cold. Am. J. Epidemiol. 2002; 155(9): 853-858 

Eccentric does not equal being a dick

Much is made in an article in Wired about the 'eccentricity' of Amy Bishop, the University of Alabama in Huntsville scientist who shot and killed three of her colleagues. Reading about her behaviour, however, makes one rather wonder about the definition of eccentric that is being used. 

Academia is, on the whole, tolerant to those that deviate somewhat from the norm, and a good thing too (although it is, by all accounts, far less tolerant than it once was, and getting worse). With its entire purpose being the advancement of knowledge, discriminating against those who are not content to accept that which is currently held as normal would be fatal. Thus I had an anatomy lecturer who would flit through the streets of Glasgow, academic's cloak streaming behind him; or the astronomy lecturer who could not countenance the use of the phrase "the naked eye", it being in his opinion "too rude". Different, certainly, but also two of the best teachers that I had. 

The manner in which Amy Bishop is described as habitually acting in could not be further from that of these benign and brilliant characters. Rather than having innocent quirks that, if anything, improved their teaching, she is described as being arrogant and abusive to a degree that interfered with her ability to do her job. 

Being obnoxious most definitely does not equal being eccentric, nor would anyone in academia ever be inclined to confuse the two and tolerate the abuse. This is reflected in the fact that she was not kept on in a number of positions, and was not awarded tenure in Alabama. 

What is more telling from the point of view of why she was tolerated and not flagged up as a potential risk  is the description given in the article of the response of the Ipswich community to her behaviour (not that I particularly believe that we should be going around flagging people who aren't very nice as being in some way a risk). By the other residents' accounts, she treated them in an equally obnoxious manner as she did her academic colleagues and students, yet they too tolerated this behaviour for five years. They may have held a spontaneous party upon the departure of the Bishop family, but there seems to be little evidence of them coming together to confront the bully in the midst of their suburban harmony. (Bringing to mind the points made about suburban living made by, for example, Eric Fromm  and Richard Sennett.)


Thus, this sad case does not represent an example of academia tolerating an eccentric, and should in no way be seen in that light. Instead what it represents is an apparent society-wide tolerance of bullies and bullying, or at least an unwillingness of society to work together to confront and ameliorate such individuals - either being a far more disturbing and telling conclusion than those being drawn by the majority of commentators.