Journal of Homeopathy: wheat

According to ScienceDaily, the journal Homeopathy has published a 2-part special issue on the biological models of homeopathy. Being the obsessive quirk that I am, I decided to get ahold of this 2-part special issue of Homeopathy to see what all the fuss is about.

ScienceDaily says that:

The special issue makes an important contribution to this debate, by reviewing laboratory experiments with high dilutions. It includes reviews and new findings in biosystems, ranging from whole animal behavioral, intoxication and inflammation models through diseased and healthy plant models, to test tube experiments using isolated cells, cell cultures or enzymes.

The editor in chief concludes that -

"Throughout its 200 year history claims that homeopathy has 'real' (as opposed to placebo) effects have been hotly contested. Our special issue brings together a wide range of scientific work in biological systems, where there can be no placebo effect, showing that there are now several biological experiments which yield consistently positive results with homeopathic dilutions."

One of the first articles I looked at was an article on the effect of arsenic on wheat seedlings [1]. This article was a review of an experiment on wheat seedlings that had been replicated 17 times among 2 research groups. The experiment was like this: To test potentiated arsenic as a cure for arsenic poisoning, researchers first exposed wheat seedlings to a 1%, 1.2%, or 1.6% solution of arsenic. Then, they had three outcome groups: 1 treated with water, 1 treated with 45X potentized water, and one treated with a 45X potentation of 1% Arsenic. Then, they measured wheat shoot length as an outcome after 7 days.

So, here's what happened with the two research groups. The first lab group conducted the experiment, and said experiment resulted in a 24% increase in shoot growth as compared to the placebo. This was the expected outcome, as homeopathic arsenic is taken to be a remedy for arsenic poisoning. A different research team later replicated the experiment, only to find the opposite – they found that the arsenic group had a reduction in hoot length by -3%. Curious about this discrepancy, they performed the experiments again. They found no statistically significant effect. So, they performed a meta analysis of all of the seed data and found a 3.2% reduction in seed growth and an (apparently nonsignificant) trend in reduction of germination rates.

The authors report that these findings are puzzling, but the fact that there was any effect at all shows that homeopathy does something. Their conclusion is that treatment of arsenic poisoned wheat seeds with homeopathic arsenic leads to statistically significant, yet contrary, effects. Without going into huge amounts of detail, I can say that the authors did a fairly good job of eliminating other factors that might have contributed to their results. Regardless, the meta-analysis certainly did not show that homeopathic arsenic is an effective treatment for arsenic poisoning in wheat.

I think that the experiment producing both positive, negative, and no effects does not point to homeopathy having some kind of effect, but rather it illustrates that science is messy and inexact. A 3.2% difference in wheat shoot growth is barely noticeable. Their mean growth rate was between 29 and 65mm. When a meta analysis reveals a 3.2% difference between treatment groups, we're talking about a difference in wheat shoot measurements of 1-2mm.

Surprisingly, the researchers in this paper talked about effect size. Effect size, simply is a measure of the strength of the relationship between two variables. In scientific experiments, it is often useful to know not only whether an experiment has a statistically significant effect, but also the size of any observed effects. In other words, if we look at a group of 20 people in a treatment group and 20 people in a placebo group, there might be significant differences, even when those differences are essentially meaningless. What is really important is how much these groups are different or how big the differences are. It is always possible to show that there is significant difference between two groups, unless they are 100% Identical. The important part is to what degree groups are different. For this, you need to know the effect size. In this case, the effect size of a 3% difference between their wheat groups amounts to 0.04. Effect size for Cohen's d an effect size of 0.2 to 0.3 might be a "small" effect, around 0.5 a "medium" effect and 0.8 to infinity, a "large" effect. If a small effect is from 0.2-0.3, and the effect size of the experimenter's wheat growth is 0.04, then I would chalk up the differences to experimental noise. This difference in groups is immensely tiny.

I have to wonder why the experimenters chose only 7 days of germination instead of taking measurements at 14 days, and then at 28 days. I don't know if a 1-2mm difference in shoot length among groups will actually translate into any measureable effects of plant size as the plants grow, but since the authors of this paper did not comment on further growth after the 7 day period, one cannot make any claims either way. There are also about a million factors that can influence early seed growth. I admit only rudimentary knowledge of horticulture, but even if seeds are placed side by side in rows in a pan, there could be enough of a difference in lighting, soil temperature, air temperature, and carbon dioxide distribution to account for the small differences seen among groups, even if the groups are dispersed. One thing the experimenters did do was to look for statistically significant differences among shoot growth of 2 pans of seeds where the seeds were placed side-by side and subjected to otherwise identical growing situations. They didn't find any differences. Here is how they arranged the seeds:

The 90 envelopes of each cardboard box were grouped in 9 groups of ten seeds each. The 9 groups of each box were randomly allocated to 3 3 treatment groups with the aid of a computer generated randomization list. The list was prepared newly for each box and each experiment. Thus, in one box, ten seeds of the first group were followed by ten seeds of the second group, and so on (corresponding to the order in which they were planted).

There are other research papers on homeopathy for treatment of arsenic poisoning. One was a study conducted on mice [2] which produced positive results, but it was unblinded. There have been other studies conducted on humans, but I'll just say that the guys over at NESS tore those studies apart. I am unconvinced that this minute difference among plant seedlings will translate into a clinically meaningful effect with regard to using homeopathy as a treatment for disease (also noting that the tiny effect the researchers found in their meta-analysis was contrary to homeopathy, stunting shoot growth rather than increasing it) and the standards for validity of non-homeopathic remedies are much higher. Given all of the other research published on homeopathy which has shown that its effects are equivalent to placebo in humans, I have to wonder why researchers are backtracking into preliminary studies and then saying, "See? High dilutions do something… even if what it does is barely measurable and contrary to the law of similar" when we already have ample evidence that homeopathy is of little clinical use.

1. Lahnstein L, Binder M, Thurneysen T, Frei-Erb M, Betti L, Peruzzi M, Heusser P, Baumgartner S. Isopathic treatment effects of Arsenicum album 45x on wheat seedling growth – further reproduction trials. Homeopathy (2009) 98:198-207
   
2. Mallick P, Chakrabarti (Mallick) J, Bibhas G, Khuda-Bukhsh AR. Ameliorating Effect of Microdoses of a Potentized Homeopathic Drug, Arsencium Album, on Arsenic-Induced Toxicity in Mice. BMC Complementary and Alternative Medicine, (2003)3:7

“The G-spot 'doesn't appear to exist', say researchers”

An article I recently read boldly claims that The G-spot 'doesn't appear to exist', say researchers. I read this with a sigh, as I know from experience how greatly distorted any research findings can get when they are published in mainstream media. Clearly, this was an instance of such distortion. I was curious to see what the actual study said, and went off to find it. You may read it here, if you are curious.

Sadly, it wasn’t very distorted after all.

In fact, the press release from the Department of Twin Research & Genetic Epidemiology was worse than the articles I had read. It presents the following conclusion from their study:

The complete absence of genetic contribution to the G-Spot, an allegedly highly sensitive area in the anterior wall of the vagina which when stimulated produces powerful orgasm, casts serious doubt on its existence, suggests a study by the Department of Twin Research to be published in the Journal of Sexual Medicine.

The investigators carried out this study by recruiting 1804 female volunteers from the TwinsUK registry aged 23-83 years. All completed questionnaires detailing their general sexual behavior and functioning, and a specific question on self-perception of the G- Spot. The researchers found no evidence for a genetic basis. This led to the conclusion that – given that all anatomical and physiological traits studied so far have been shown to be at least partially influenced by genes – the G-Spot does not exist and is more a fiction created by other factors e.g. an individual’s own sexual and relationship satisfaction or self-report is an inadequate way to assess the G-Spot and researchers should in future focus more on ultrasound studies.

The impression I took from the mainstream press articles, and which was reinforced by the institute’s press release, was that the existence of the G-spot was inferred to correspond to study participants’ reports of whether they had one. If this were so—if we could determine anatomy by poll—I expect I could find some people with more spleens than kidneys and more livers than lymph nodes.

---

I took the trouble to read the actual paper (it’s fairly short and quite accessible). The reality turns out not to be quite so bad. The main point—well, let me make an aside here and say that I find it extremely odd that what seemed to be the main point emphasised in the paper was considerably de-emphasised in the press release and consequent mainstream articles, seriously reducing their credibility. Anyway, back to the point:

The main point of the paper is that if the G-spot exists, it is an anatomical structure; if it is an anatomical structure, it is presumably genetically inherited. Even if some women have it while some don’t, we expect to find a strong correlation in twins. Since heterozygotic twins share 50% of their genome, and monozygotic (‘identical’) twins share 100% of their genome, if it’s genetically heritable at all, we should see a correlation in twins, especially monozygotic ones: If one twin has it, the other should (more often than is the case with unrelated people); if one twin does not, the other shouldn’t. Because twins are typically raised in extremely similar environments, even environmental factors should be similar. In particular, monozygotic twins should be more similar to each other than heterozygotic twins for heritable (but not environmental) factors.

Well, this turned out not to be the case: Heterozygous twins report that they have G-spots about as often as do monozygous twins, and this is the real point of the paper. It’s not as spectacular as the mainstream news articles, but I’m surprised that they so failed to emphasise this in their own press release. Ah well, such is the hunt for fame, I suppose.

In the conclusion of the real, scientific paper, the authors are of course forced to admit that

A possible explanation for the lack of heritability may be that women differ in their ability to detect their own (true) G-spots.

They, of course, do not believe this to be the case. We may reasonably ask, why not? And how good is your evidence? My thoughts will be very tentative, because I’m not an expert in any related field; but we may at least reason about it.

---

First, I will note that the study’s exclusion criteria were, at times, a bit puzzling.

Women who reported that they were homo- or bisexual were excluded from the study because of the common use of digital stimulation among these women, which may bias the results.

I daresay it may bias them! For example, if the G-spot exists, it’s a specific anatomical location inside the vagina. Because it is postulated to be a very specific location, it may be difficult to stimulate with the penis, which is after all not prehensile and may not be angled so as to optimally stimulate a specific location. This postulated spot could perhaps be more easily located and stimulated with the fingers. Therefore, if it does exist, and if we are restricted to self-reporting as evidence, I would expect to find much stronger evidence for this in a population with common use of digital stimulation. The people I would ask first are the people whose answers they discarded. I would be very curious to see how their data are affected if they include this population. What was their rationale for the exclusion criterion? Was it determined beforehand, or after the data were in? Would it contradict their conclusion? What if this population were considered exclusively?

This looks like a very serious weakness to me, as the exclusion criterion seems to be specifically geared towards reaching a particular conclusion. (I can’t think of anything much more damning I could possibly say about a study.) It’s not the only thing that makes me raise an eyebrow, though (but it is the strongest).

Another thing is that, well, some traits just aren’t very heritable. (This is why we measure heritability; if there weren’t variation in how strongly phenotypic traits are associated with genes, there’d be no need.) I suppose the authors may reasonably expect their readership to be familiar with not just the concept of heritability (as I am), but also what kind of numbers we should expect (as I am not). Is a “close to 0” heritability common, or unusual, or rare, or impossible in variable phenotypic traits? Still, it is possible that heritability of the G-spot—not necessarily its existence, but perhaps its precise location and orientation, or its sensitivity—is relatively low. Is the study still powered to detect it? How does this render it more vulnerable to other confounders?

There are various criticisms leveraged against twin studies in general. Twin studies are potentially wonderful tools because monozygotic twins offer unique opportunities to investigate heritability. (Personally, I think the most interesting ones are of that rarity of rarities, pairs of monozygotic twins raised apart; the surprising similarities they show in a very wide range of behavioural traits is strong evidence of genetic conditioning.) But they are not perfect.

And finally, I make the observation that the institute—the Department of Twin Research & Genetic Epidemiology—maintain a database of twins (an awful lot of them: Some 11,000 people). This is great; it enables them to efficiently perform twin studies. However, studying the same sample over and over again is problematic. If you look at the same N people, examining them for different properties over and over again, you’re bound to find an apparent correlation eventually. Think about it: If you pick 100 names at random from a phone book, you’ll expect about half of them to be male, half female; and about 8–15 of them to be left-handed…but if you examined them for blood pressure, and dietary habits, and sexual preferences, and number of children, and so on for any number of questions, it would be bizarre if they were an average sample in every respect. This is a problem with data mining. Clearly, the department’s database is pretty large, but then they’ve already published over 400 research papers. At what number of papers should we statistically expect to find spurious calculations?

---

All in all, the study was a bit more sensible than mainstream media had me thinking at first, but as research papers go, I found it surprisingly unimpressive. In particular, the exclusion criterion that discarded answers from gay and bisexual women smells very fishy, and I wouldn’t be terribly surprised if it “biased” the results so far as to invalidate their conclusion.

In a general sense, I trust science—I trust the scientific method, and (to a lesser but considerable degree) I trust that scientific consensus will move toward the right answers: Science is often characterised as an asymptotic approach to the truth (we may never know it exactly, but we will get ever closer). However, when considering a single study, one should be cautious. Never trust what the mainstream press says about it at all, whether you like what it says or not—ordinary reporters lack scientific savvy, good science reporters are rare, and after the editors have their say, it’s often dubious whether the scientists behind a finding would agree with anything the press has to say about them except, perhaps, the scientists’ names.

And while the scientific method is excellent, and the scientific consensus is the best approach we have to knowledge, some studies just aren’t worth the paper of the webpages they’re published on. If you want to adjust your opinions according to a single study, read it. Read it critically.

3H1P: ‘sup w/ evolution and Jesus?

 3H1P is a blogging project wherein three heathens (Ziztur, Flimsy and Petter) and one pastor (Keith) answer questions posed by readers of the blog and discuss various issues related to religion, philosophy, science, etc. If you have a question that you'd like to see answered by 3H1P, ask it in the comment box. We promise we'll probably get to it.

I know this question was meant for Pastor Keith, but I decided to go ahead and answer it, too.

Reader Mitch asks:

I've read that most Christians are ok with evolution. But the truth of evolution shows that the story of Adam and Eve in the Garden is a myth, hence there was no "Fall", no "Original Sin". Without "Original Sin" there is no need for "Salvation". Jesus is pointless. So, how can a Christian accept evolution?

I would like to question the initial premise of this question that most Christians are okay with evolution. I'm glad that you prefaced this factoid with "I've read" rather than simply proclaiming it as a fact – it shows that you're likely to know the difference between what people claim and what is actually true.

According to every reputable public opinion polling establishment out there, most American Christians are not okay with evolution. According to the Pew forum, only 6% of evangelical Christians, 31% of mainline Christians, and 25% of Catholics in he US believe that "humans and other living things evolved over time through natural selection" (in 2006). According to the Gallup poll, 24% of people who attend church weekly or more accept evolution – though it does not say that they accept evolution in its entirety. People who are both Christian and accept evolution fall squarely in the minority, and I can imagine that an even tinier portion of Christians both fully accept evolution and fully believe in a literal interpretation of the Garden.Of course, as fellow blogger Marc_Newcomb points out, this says nothing about Christians outside the US. Do most Christians believe in evolution? I'm not entirely sure.

Regardless, the meat of your question is still important. How can a Christian accept evolution? I'm not a Christian, but in my experience most Christians who accept evolution accept a blend of both Christianity and evolution in such a way that the level of cognitive dissonance is not so much that they are winning the Dissonance Olympics with their impressive brain gymnastics.

I would imagine that one can accept in evolution on a continuum from total confidence in evolution to total rejection, with many varieties of partial acceptance in between. We can even plot this on a totally unscientific but still useful graph:

There are a lot of people out there who consider evolution to be true, but that it was guided by the hand of their god. These people are more likely to believe that the Garden story is mythical, but can still believe that we are all sinners by nature and need salvation. They could believe "Adam and Eve" are allegories for the human race, or any number of different things. So I don't think that evolution and Christianity are necessarily in conflict with each other, depending on what parts of which you accept. Obviously, a literalist interpretation of the Bible is in conflict. If you believe that the world was created in 6 literal days, then to accept evolution in addition to the literal creation story is to accept that Billy is both sixty years old and two days old.

Acupuncture affecting regulation of pain

A new acupuncture vs. placebo acupuncture study has been making headlines such as "Chinese acupuncture affects brain's ability to regulate pain, study shows" due to results of a study suggesting that there may be a difference in opioid receptor response in acupuncture vs. placebo acupuncture.

Several large-scale studies [1-3] have been released showing that acupuncture and various forms of placebo acupuncture have clinically insignificant differences in the reduction of pain, proponents of acupuncture are now looking at brain-imaging to explore the mechanisms of acupuncture and placebo acupuncture to determine if acupuncture and placebo acupuncture operate via different mechanisms.

In this study [4], researchers hypothesized that long term acupuncture therapy may result in increased opioid receptor availability and that these effects would not be observed in a placebo acupuncture group. Their subjects consisted of 20 women randomly divided into 2 groups of 10 subjects. One group received traditional acupuncture treatment while another group received non-invasive, placebo acupuncture. Results from PET scans using contrast material were taken during a 90-minute period, during which acupuncture treatment or sham acupuncture treatment was administered during the 45-90 minute timeframe. A period followed in which subjects received 7 acupuncture or sham acupuncture treatments, and then the PET scan procedure was repeated, for a total of 9 treatments. Results indicate acupuncture therapy evoked short-term increases in MOR binding potential, in multiple pain and sensory processing regions including the cingulate (dorsal and subgenual), insula, caudate, thalamus, and amygdala. Acupuncture therapy also evoked long-term increases in MOR binding potential in some of the same structures including the cingulate (dorsal and perigenual), caudate, and amygdala. These short- and long-term effects were absent in the sham group where small reductions were observed, an effect more consistent with previous placebo PET studies.

There are several pieces of information regarding this study that were left out of news headlines an abstracts, so I will attempt to summarize them here and then offer my own analysis of the results. However, it should be clear that I am not a neurologist, and thus my knowledge of neurology is somewhat limited.

Subjects were blinded to which treatment group they were in, and they were also asked to guess which treatment group they thought they were assigned to after the first PET scan. There was no significant difference between groups, and thus it can be assumed that the subjects remained adequately blinded, though the study does not mention any blinding of the researchers, making it quite likely that the researchers were unblinded, which could have an effect on the study results.

During the acupuncture treatment, needles were left in during PET scan acupuncture treatment measurement during the 45-90 minute timeframe, while no needles were retained during the sham acupuncture group given that no needle penetration occurred during sham acupuncture. Clearly then, PET scans during that 45-90 minute period involve one group receiving an active treatment (given that needles were in their skin) while the sham group received inactive treatment (given that no needles were present). It seems obvious to me that there will be neurobiological differences between a group of people being measured while needles are inserted into them and a group of people who do not have needles in them, so their results are not surprising. Additionally, even though subjects may have not know which group they were in (they had to have had no prior acupuncture experience to participate), they quite likely were aware of whether or not needles remained in place during the PET scans, especially given the fact that both treatment groups involved placement on the head and ear. This knowledge could provide an explanation for the differences in treatment groups and is not addressed in the study. As such, even though the subjects were ignorant of whether or not they were receiving placebo treatment or not, the same cannot be said of their ignorance of the presence of needles placed in their body during PET scans.

What I find especially interesting about this study is the discussion of opponents of acupuncture in the introduction. The researchers wrote:

“Recent controversy in the field of acupuncture research was generated when several large scale randomized controlled trials in chronic pain patients failed to show superiority of acupuncture over sham acupuncture methods. This has led opponents of acupuncture therapy to suggest that it is no more effective than a placebo intervention.”

I fail to see why one needs to be an opponent of acupuncture therapy to suggest exactly what the large scale randomized controlled trials are suggesting – that acupuncture therapy is not superior to placebo intervention. This sentence seems to indicate potential bias on the part of the researchers in this study. The data from these studies are very clear.

If the clinical results between acupuncture and placebo acupuncture are the same, it seems to me that potential side-effects are far more important than the fact that acupuncture and placebo acupuncture potentially operate via different mechanisms. This difference in mechanisms is irrelevant, or at least far less relevant than clinical results or side effects. It could be argued that different types of placebo acupuncture have different neurochemical mechanisms of action as well, but thus far no study has documented these potential differences. This study is weak in that it only compares two different treatments. If the study had used acupuncture and two different types of placebo acupuncture that had already been established through trials to have similar clinical results, and then shown that the mechanism of action for the placebo acupuncture was the same while the acupuncture group had a different mechanism, then the results would be more convincing.

The researchers indicated that previous studies indicated that the neurobiological response to acupuncture was distinct from pain and sham acupuncture, but one of the articles [5] I read in support of this claim (there were three total, and all appeared to be from the same group of people as evidenced by common authors. I read the newest one.) failed to blind subjects to which treatment they were receiving and thus is poor evidence to support their claim. Instead, this seems to support evidence that there is a different neurobiological response in individuals who know they are receiving acupuncture or a placebo, which is to be expected.

Regardless, it seems fairly obvious to me that measuring neurobiological responses in a PET scan while some subjects have needles inserted during the scan and others do not is measuring a neurobiological response to needles being in the skin versus not in the skin.  Sticking needles in subjects would likely provoke a different neurochemical response in subjects when compared to placebo acupuncture, which involved no needle insertion. So, if you do two different physical things to people, this provokes different neurochemical responses. Didn’t we already know this? At least this study does not argue that acupuncture and placebo acupuncture have different effects. Instead, it argues that acupuncture and placebo acupuncture have different mechanisms. My less than dignified response is, “So what?”


1. Brinkhaus B., Witt CM, Jena S, Linde K, Streng A, Wagenpfeil S, Irnich D, Walther HU, Melchart D, Willich SN. Acupuncture in patients with chronic low back pain: a randomized controlled trial. Arch. Intern. Med. 2006;166:450–457.


2. Linde K, Streng A, Jurgens S, Hoppe A, Brinkhaus B, Witt C, Wagenpfeil S, Pfaffenrath V, Hammes MG, Weidenhammer W, Willich SN, Melchart D. Acupuncture for patients with migraine: a randomized controlled trial. JAMA 2005;293:2118–2125.

3. Melchart D, Streng A, Hoppe A, Brinkhaus B, Witt C,Wagenpfeil S, Pfaffenrath V, Hammes M, Hummelsberger J, Irnich D, Weidenhammer W, Willich SN, Linde K. Acupuncture in patients with tension-type headache: randomized controlled trial. BMJ 2005;331:376–382.

4. Harris RE, Zubieta JK, Scott DJ, Napadow V, Gracely RH, Clauw DJ. Traditional Chinese acupuncture and placebo (sham) acupuncture are differentiated by their effects on μ-opioid receptors (MORs) NeuroImage 2009;47:1077-1085

5. Napadow V, Kettner N, Liu J, Li M, Kwong KK, Vangel M, Makris N, Audette J, Hui KK.  Hypothalamus and amygdala response to acupuncture stimuli in carpal tunnel syndrome. Pain 2007;130: 254–266.

How much money do we spend on CAM?

Finally! A real survey outlining the cost of Supplements, Complimentary and Alternative Medicing (SCAM's) has been published by a government survey.

Now, when people use "Big Pharma" as a scapegoat for illogical arguments about why SCAM's work, we can point out that "Big SCAM's" have their fingers in the healthcare pie too. Read on:

Americans spent $33.9 billion out-of-pocket on complementary and alternative medicine (CAM) over the previous 12 months, according to a 2007 government survey [1]. CAM is a group of diverse medical and health care systems, practices, and products such as herbal supplements, meditation, chiropractic, and acupuncture that are not generally considered to be part of conventional medicine. CAM accounts for approximately 1.5 percent of total health care expenditures ($2.2 trillion [2]) and 11.2 percent of total out-of-pocket expenditures (conventional out-of-pocket: $286.6 billion [2] and CAM out-of-pocket: $33.9 billion[1]) on health care in the United States.

Approximately 38 percent of adults use some form of CAM for health and wellness or to treat a variety of diseases and conditions, according to data from the 2007 National Health Interview Survey (NHIS) [3]. 

Of the $33.9 billion spent on CAM out-of-pocket, an estimated $22.0 billion was spent on self-care costs—CAM products, classes, and materials—with the majority going to the purchase of nonvitamin, nonmineral, natural products ($14.8 billion) such as fish oil, glucosamine and Echinacea. U.S. adults also spent approximately $11.9 billion on an estimated 354.2 million visits to CAM practitioners such as acupuncturists, chiropractors, massage therapists, etc.

To put these figures in context, the $14.8 billion spent on nonvitamin, nonmineral, natural products is equivalent to approximately one-third of total out-of-pocket spending on prescription drugs, and the $11.9 billion spent on CAM practitioner visits is equivalent to approximately one-quarter of total out-of-pocket spending on physician visits.

"These data indicate that the U.S. public makes millions of visits to CAM providers each year and spends billions of dollars for these services, as well as for self-care forms of CAM," said Richard L. Nahin, Ph.D., MPH, acting director of NCCAM's Division of Extramural Research and lead author of the cost of complementary and alternative medicine analysis. "While these expenditures represent just a small fraction of total health care spending in the United States, they constitute a substantial part of out-of-pocket health care costs."

 Here are some important references. I suggest you memorize them:


1. Nahin, RL, Barnes PM, Stussman BJ, and Bloom B. Costs of Complementary and Alternative Medicine (CAM) and Frequency of Visits to CAM Practitioners: United States, 2007. National health statistics reports; no 18. Hyattsville, MD: National Center for Health Statistics. 2009.

2. Office of the Actuary, Centers for Medicare and Medicaid Services, National Health Expenditure Data for 2007. U.S. Department of Health and Human Services. Available at: http://www.cms.hhs.gov/NationalHealthExpendData/02_NationalHealthAccountsHistorical.asp#TopOfPage.

3.  Barnes PM, Bloom B, Nahin RL. Complementary and Alternative Medicine Use Among Adults and Children: United States, 2007. National health statistics reports; no 12. Hyattsville, MD: National Center for Health Statistics. 2008.

College attendance and religiosity

Recently I read here that a study [1] had been conducted which looks at the trends between the study of certain subjects in college and religious observance. The study concluded that very religious high school students are more likely than less religious high school students to attend college.

This may surprise the skeptical world. I’ve heard many times that people with high levels of religiosity tend to be less educated and less intelligent whereas people with low religiosity tend to be more educated and more intelligent. Typically people cite an article published in nature as evidence for this phenomenon [2], if they cite an article at all. So why is this study saying that people who are more religious are more likely to attend college?

The authors first rightfully point out that there is a pressure in the United States toward being religious, yet despite this pressure (especially from families), religiosity is remaining more or less steady and even swinging downward as the years go by. One common culprit blamed on this is college, given that college tends to be the first time people are separated from their families for an extended period of time.

The goal of the study was to look at how contents of college curriculum affect student values and to distinguish these effects from patterns of selection based on already-held values. The study hypothesizes that college students are confronted (in varying degrees) to three streams of thought in college that have certain negative attitudes toward religion, and that these streams of thought may have an effect on religiosity. Those streams of thought are:

Science – consisting of a commitment to truth, the scientific method and open-mindedness toward evidence. Natural science fields have a strong scientist content.

Developmentalism – consisting of a commitment to freedom and progress. Economics and business have a strong developmentalist content.

Postmodernism – consisting of a commitment to relativism of truth and morality and the idea that truth and morality are determined by those who are most powerful. The humanities and social sciences have a strong postmodernist content.

The article delves much deeper into exactly what these three streams of thought are and how they come into conflict with religiosity (in a refreshingly impartial manner), but for the purposes of this article I will leave them simply defined.  If on accepts that different majors are tied to different streams of thought, it is possible to test whether any of these three streams of thought contribute to reduced religiosity by looking at initial choices of major and changes of major over time in concordance with any changes in religiosity. They specifically examined the changes in religiosity from high school and into college, using a sample size of literally thousands of students in Michigan from high school and through college.

Here are some of their findings:

-compared to business majors, social sciences and humanities have a statistically significant negative effect on both attendance of religious services and the rating of the importance of religion.
-education majors were more religious than other majors and their religiosity increased over time.
-religiosity increases over time for business majors
-religious attendance decreased for students who are undecided about college major.
-religious attendance decreased for respondents who did not go to college.
-students in science and engineering have less trust in god than people who have not gone to college or business majors (as measured by asking the respondents, “if we just leave things to god, they will turn out for the best [disagree, mostly disagree, neither agree nor disagree, mostly agree, agree]”).

-students who were social science, humanities or engineering tended to think religious organizations should have less of an effect on society, whereas subjects who had not gone to college thought religious organizations should have more influence on society.

The study was fairly well-designed, so I was quite surprised to find that students who were more religious were more likely to attend college, given that it seems to contradict other studies (such as the one reported on in Michael Shermer’s How We Believe: Science, Skepticism and the Search for God [3]) which indicate that atheism increases with education level.  Religiosity was rated on a 1-4 scale (“how often to you attend religious services? [1=never, 2=rarely 3=once or twice a month 4=about once a week or more]”), and each point on the scale corresponded with a 14% increases in the likelihood of going to college. A change in the rated importance of religion by one point amounted to an 8% increase in the likelihood of going to college.

I think that a weakness in the study lies in the fact that they did not take into account the different denominations of college attendees, and rather used “religious attendance” and “importance of religion” as their measures of religiosity. As such, the study ignored the positive or negative impacts of particular denominations or religious schools of thought, especially Sectarianism (the belief that religious rewards will be given exclusively to the adherents of a particular faith) and fundamentalism (finding value in sacred texts, especially the belief in the inerrancy of biblical texts).  Studies have shown that sectarianism and fundamentalism in particular has a negative impact on educational attainment [4], and especially the educational attainment of women [5]. Sectarian and fundamentalist individuals are also more likely to choose a religious college.

One other interesting thing to note is that the reported amount of time people spend at religious services does not seem to correlate with actual religious service attendance – studies show the actual number is about half of what people report – suggesting that many people overestimate (or lie about?) how often they attend [6]. The best this study can say with regard to religious service attendance is that the people who likely overestimate how often they attend are more likely to go to college.

1. Kimball MS, Mitchell CM, Thornton AD, Young-Demarco LC. Empirics on the Origins of Preferences: The Case of College Major and religiosity. NBER (2009) Working Paper No. 15182

2. Larson EJ, Witham L. Leading scientists still reject god. Nature 1998:394;313

3. Shermer, M. How We Believe: Science, Skepticism, and the Search for God. New York: William H Freeman. 1999:76–79. ISBN 071673561X.

4. Sherkat DE. Religion and higher education: the good, the bad, and the ugly. 2007: Online at http://religion.ssrc.org/reforum/Sherkat.pdf

5. Sherkat, DE, Darnell A. The Effect of Parents' Fundamentalism on Children's
Educational Attainment: Examining Differences by Gender and Children's Fundamentalism Journal for the Scientific Study of Religion 1999:38;23-35.

6. Hadaway CK, Marler L, Chaves M. what the polls don’t show: a closer look at U.S. church attendance. American Sociological Review 1993:58;741-752

Chinese Herbal Medicine for Endometriosis

Media outlets are reporting that Chinese herbs may relieve symptoms of endometriosis, using a Cochrane review of 2 research articles. Headlines read “Chinese herbs show early promise for endometriosis”  and  “Chinese Herbs May Relieve Endometriosis Symptoms, Review Finds”  .

Endometriosis is a medical condition in which some of the endometrial cells (typically found in the uterus under the fluctuating influence of female hormones) are found outside of the uterine cavity. Symptoms include many nonspecific complaints such as pelvic pain, infertility, nausea, unusual menstruation, chronic fatigue, mood swings, back pain, ovarian cysts, constipation, urinary tract infections, diarrhea, anemia, etc. Appropriate diagnosis is by laparoscopic biopsy – a doctor will use a laparoscopic instrument to remove suspected extrauterine endometrial cells and examine them.  Treatments vary and can include hormonal treatments or surgery to remove the cells. In China, treatment of this disorder with Chinese herbal medicine (CHM) is routine.

Both of the media reports linked to above say that the Cochrane reviewers found some evidence that CHM has comparable benefits to conventional drug therapy after laparoscopic surgery for people with endometriosis but that the review has limitations. The primary author of the study is quoted as saying “"I think the positive message is that Chinese herbal medicine may offer equivalent benefits to conventional medicine but with fewer side effects.”

I found the Cochrane review [1] and noted that reviewers collected 110 studies for review and graded them based on methodological criteria. They dropped all but two of the 110 studies due to excluding trials with poor methodology, unconfirmed randomization procedures or ones diagnosing endometriosis without an appropriate laparoscopic biopsy.

Did the two retained articles feature research with superior methodology? First, let’s look at what those two articles were, and what the author concluded from this review:

The first article [2] had two treatment arms: women treated with CHM orally (2x/day) and via enema (1x/day) after laparoscopic surgery versus women treated with gestrinone (2x/wk) after laparoscopic surgery for 3 months. The results showed no difference between rates of symptom relief or pregnancy in either group.

The second article [3] had three treatment arms: women treated with CHM orally(2x/day), women who treated with CHM orally and via enema(1x/day), and women treated with danazol(1x/day) for 3 months. These women did not undergo laparoscopic surgery, but instead were only biopsied for diagnostic purposes.  Women obtained greater symptomatic relief with oral and oral plus enema CHM versus danazol, oral plus enema CHM shower a greater reduction in dysmenorrhoea pain scores than danazol and shrinkage of adenexal masses. There were no differences for other factors (lumbrosacral pain, rectal discomfort, vaginal nodules).

The author concluded that post-surgical administration of CHM may have comparable benefits to gestrinone but with fewer side effects, that oral CHM may be better for treatment than danazol and may be more effective at relieving dysmenorrheal and shrinking adnexal masses when used with a CHM enema.

So, what are our weaknesses?

1.    No placebo control: There was no arm of the first study which looked at women receiving laparoscopic surgery alone without CHM or danazol, and no arm of the second study which looked at women receiving no treatment or a placebo pill treatment.

2.    Poor blinding: I should not have to point out that if you enroll in a study that has a pill treatment arm and an enema treatment arm, it is impossible for the participants to be blinded to which treatment group they are in. And enema, as you probably know, is a procedure in which liquids are forced into the rectum through the anus. It might be possible to blind participants to whether or not they are getting CHM versus the other medications, but I bet most people can tell the difference between a Chinese medicine pill and the other pills in the study. The researchers were also not blinded as to which treatment group women were in, though the paper indicates the assessors were blinded to which treatment group the women were in.

3.    Inadequate comparison treatments: Danazol is no longer commonly used as a treatment for endometriosis, and gestrinone is not available in the USA. These studies would have been much more robust had they compared it to typical drug treatments for endometriosis. In the world of conventional treatments for endometriosis, these two drugs can hardly be called conventional.

4.    Poor outcome measures: In both of the studies, a clinical outcome of “no effect” was recorded if there were no change in symptoms or if the symptoms became worse. Recording worsening symptoms as “no effect” biases the data toward a positive outcome.


I think that the most appropriate take home message or finding of the study is this: the massive stockpile of clinical trials that explore CHM for treating endometriosis have serious methodological shortcomings. 

The author’s main conclusion (and the conclusion parroted by the press), that CHM may work to alleviate symptoms of endometriosis, seems spurious in light of this. Additionally, it appears that researchers used a specific mixture of herbs (Nei Yi) in the two studies, which raises the question: why the author did not title his paper “Nei Yi for endometriosis”?  Perhaps he wanted his readers to focus on the fact that this was a Chinese herbal medicine versus a “conventional” medicine.

Lastly, it is worth noting that although the authors of the review state that there is no conflict of interest in the publication of this review, the primary author is an acupuncture and Chinese medicine practitioner at a center for Chinese medicine in the UK.

References:

[1] Flower A, Liu JP, Chen S, Lewith G, Little P. Chinese herbal medicine for endometriosis. Cochrane Database of Systematic Reviews 2009, Issue 3. Art. No.: CD006568. DOI: 10.1002/14651858.CD006568.pub2.

[2] Wu SZ,Chen XL,Chen WZ, Li SY.Clinical analysis of the treatment of endometriosis using Nei Yi pills and Nei Yi enema. Journal of Liaoning University of TCM 2006;8(7):5–6.

[3] Wu SZ, Chen XL, Chen WZ. Clinical observation of Nei Yi pills combined with Nei Yi enema in the treatment of endometriosis. Chinese Archives of TCM 2006;24(3):431–3.

Coincidences

Brief IM log snippet:

(10:20:50 AM) David: http://www.youtube.com/watch?v=obIGsb-IZMo&feature=user
(10:22:11 AM) Petter: I paused my music to check that out.
(10:22:28 AM) Petter: In a rather amusing coincidence, the song I paused was a song by Tom Lehrer called "New Math".
(10:22:40 AM) David: wow, what are the odds? ;)
(10:23:10 AM) Petter: Probably inversely proportional to the frequency with which such remarkable coincidences happen to me.

Other peculiar coincidences of my life:

• 
  At a lake, swimming with a friend; we see a column of smoke rising from nearby and make jokes along the lines of I guess I shouldn't have planted that car bomb (yeah, yeah; we were about fourteen). Heading home, we pass the source of the smoke; the burning wreck of a car.
  
• 
  The weirdest one: Playing a role playing game with a friend (same friend, incidentally). The setting is a sci fi one; the game…kind of superhero-ish, a bit like Batman with a lot more guns. Having broken up some crime or other, the hero fires a few rounds into the air to attract the attention of the authorities as he makes his sortie. You accidentally shoot out a streetlight, I declare. Annoyed, he says something to the effect of Oh, come on!, then falls silent as the light bulb in my desk lamp cracks.
• 
  This is the only time in my life I've seen a light bulb do that; I guess thermal expansion and contraction had slowly worked microscopic cracks to the point where most of the glass bulb fell off and shattered on the desk, leaving the screw-in part in the socket, adorned with some jagged pieces of glass.
  

Remarkable coincidences like these seem to tempt some people to commit superstition. Me, I like to wonder the following: What are the odds that anyone should go through life without running into at least a few remarkable coincidences? After all, every single day, every one of us experiences many hundred events (using the term loosely), for a grand total of millions of events over a lifetime. On average, therefore, a person should expect to run into at least a couple of coincidences that are literally one in a million rare; coincidences that are one in a thousand, meanwhile, are a dime a dozen—you'll go through thousands of them. Or, in one of my favoite self-coined phrases, Every single day, on Earth, there are half a dozen people having a one-in-a-billion kind of day.

For a more thorough (but very accessible) discussion of this sort of thing, read Richard Dawkins's Unweaving the Rainbow, in particular chapter 7, Unweaving the Uncanny.

Homeopathy and the curse of the scientific method Pt2

This is part 2 of my analysis of a double-blind placebo controlled study of the efficacy of a homeopathic anxiety product produced by Homeopet, and the response to that study in the form of a journal article. Part one can be found here, with more info on Homeopet here. 

Typically, when one starts to speak of research, one starts to hear of phrase like "statistical significance" and learns that when group X is statistically significantly different from group Y, that they are different in some kind of measurable way.

In the case of a homeopathic drug treatment such as the one in Overall (2009), one of the first things we will want to know is if there is a significant difference between the homeopathic 'treatment' group and the placebo group.

A lot of published research on homeopathy will report "significant" differences, but will not state whether or not these differences are large enough to warrant the claim that we have detected a phenomena. As Overall puts it in her article, "A rare statistical association may have no effect on the biology of the system, but could misdirect incautious readers to believe that this was not the case". Thus, to really say something meaningful about the differences in treatment groups, one has to say something about effect size.

Effect size, simply is a measure of the strength of the relationship between two variables. In scientific experiments, it is often useful to know not only whether an experiment has a statistically significant effect, but also the size of any observed effects. In other words, if we look at a group of 20 people in a treatment group and 20 people in a placebo group, there might be significant differences, even when those differences are essentially meaningless. What is really important is how much these groups are different or how big the differences are. It is always possible to show that there is significant difference between two groups, unless they are 100% Identical. The important part is to what degree groups are different. For this, you need to know the effect size. To shamelessly quote the Wikipedia entry on effect size:

Presentation of effect size and confidence interval is highly recommended in biological journals. Biologists should ultimately be interested in biological importance, which can be assessed using the magnitude of an effect, not statistical significance. Combined use of an effect size and its confidence interval enables someone to assess the relationship within data more effectively than the use of p values, regardless of statistical significance. Also, routine presentation of effect size will encourage researchers to view their results in the context of previous research and will facilitate incorporating results into future meta-analysis. However, issues surrounding publication bias towards statistically significant results, coupled with inadequate statistical power will lead to an overestimation of effect sizes, consequently affecting meta-analyses and power-analyses.

If we establish that for a given treatment and placebo group, there is both a significant difference and a large effect size, then we must propose a testable mechanism to explain this difference. Of course, there is the rub - CAM studies rarely contain testable mechanisms of action. How many times have you heard that an alternative medicine "baffles scientists" or "can't be explained by science"? You can bet the effects can be explained by science - we will just say what alt-med practitioners don't want to hear - that their methodology is flawed or their effect is entirely placebo.  They also rarely contain information about effect size (even though, once one has the data in place, the calculations can be done in a matter of minutes.)

If alt-med such as homeopathy wants to be considered alongside science, it has to play the science game and the results of studies need to be held to the same standards as science-based medicine. To do that, anyone with a vested interest in the outcome needs to be excluded from scientific experimentation (such as pet owners who believe homeopathy works, practicing homeopaths, or in-house researchers employed or funded by the company promoting the treatment, such as is the case with Cracknall (2008). Additionally, the methodology of studies must be sound and transparent, using such methods as blinding, placebo arms and discussion and calculation of effect size. Last, there must me a mechanism - a mechanism of action os one of those things that separates evidence-based medicine from science-based medicine.

There is no conceivable mechanism for the action of homeopathic preparations. As a reminder, homeopathic 'remedies' work by serial dilution to the point in which there may not even be one molecule of the original 'active ingredient' in the remedy. This is not the same as herbal medications, which actually have some sort of substance beyond a carrier substance such as water or alcohol.


Here, I must quote Overall verbatim, because the author says something absolutely concise:

The ‘herbal’ and ‘natural’ types of interventions are more appropriately part of CAM. Virtually all commercially produced pharmaceuticals have their source and/or intellectual roots in active compounds isolated from plants. It is ironic that ‘herbal’ sources themselves are often considered ‘safe’, but that commercially produced derivatives – the compounds that are forced to undergo toxicity testing – are not.

 Herbal medications at least have some sort of potential mechanism, as they actually have ingredients in them. Homeopathy, on the other hand, relies on water molecules "remembering" the substances that were shaken into them. This defies reason, testing and observation and is essentially magic. Of your mechanism relies on magic to work, you have no real mechanism.

Refs:
Overall, K., Dunham, A., Homeopathy and the curse of the scientific method. The Veterinary Journal (2009) 180: 141-148
Online abstract

Cracknell, N., Mills, D. A double-blind placebo-controlled study into the efficacy of a homeopathic remedy for fear of firework noises in the dog (Canis familiaris) The Veterinary Journal (2008) 177:80-88

“Why Most Published Research Findings are False”

“Orac” over at Respectful Insolence has a writing style that’s fairly prone to offend—definitely pugnacious, and very fond of side swipes at those he dislikes (primarily alternative medicine quacks)—and I don’t blame him for his distaste, which in fact I share, but it does sometimes make his essays a bit harder to slog through. (He also has an inordinate fondness for beginning sentences with Indeed. This is one area where I can tentatively claim superiority: I can also be pugnacious and come off as offensive, but while I am no less prone than Orac to complicated sentence structure, I’ve never been accused of any such repetitive verbal tic.)

However, those foibles aside, he has written some very good stuff (he’s on my list of blogs I ready daily for a reason), and this article, summarising and explaining the work of a John Ioannidis, was very interesting indeed. The claim it looks at is a very interesting and puzzling one: Given a set of published clinical studies reporting positive outcomes, all with a confidence interval of 95%, we should expect more than 5% to give wrong results; and, furthermore, studies of phenomena with low prior probability are more likely to give false positives than studies where the prior probabilities are high. He has often cited this result as a reason why we should be even more skeptical of trials of quackery like homeopathy than the confidence intervals and study powers suggest, but I have to confess I never quite understood it.

I would suggest that you go read the article (or this take, referenced therein), but at the risk of being silly in summarising what is essentially a summary to begin with…here’s the issue, along with some prefatory matter for the non-statisticians:

A Type I error is a false positive: We seem to see an effect where there is no effect, simply due to random chance. This sort of thing does happen. Knowing how dice work, I may hypothesise that if you throw a pair of dice, you are not likely to throw two sixes, but one time out of every 36 (¹/₆×¹/₆), you will. I can confidently predict that you won’t roll double sixes twice in a row, but about one time in 1,296, you will. Any time we perform any experiment, we may get this sort of effect, so a statistical test, such as a medical trial, has a confidence level, where a confidence level of 95% means there’s a 5% chance of a Type I error.

There’s also a Type II error, or false negative, where the hypothesis is true but the results just aren’t borne out on this occasion. To the best of my knowledge, there is no equivalent of the confidence level for Type II errors.

This latter observation is a bit problematic, and leads into what Ioannidis observed:

Suppose there are 1000 possible hypotheses to be tested. There are an infinite number of false hypotheses about the world and only a finite number of true hypotheses so we should expect that most hypotheses are false. Let us assume that of every 1000 hypotheses 200 are true and 800 false.

It is inevitable in a statistical study that some false hypotheses are accepted as true. In fact, standard statistical practice [i.e. using a confidence level of 95%] guarantees that at least 5% of false hypotheses are accepted as true. Thus, out of the 800 false hypotheses 40 will be accepted as "true," i.e. statistically significant

It is also inevitable in a statistical study that we will fail to accept some true hypotheses (Yes, I do know that a proper statistician would say "fail to reject the null when the null is in fact false," but that is ugly). It's hard to say what the probability is of not finding evidence for a true hypothesis because it depends on a variety of factors such as the sample size but let's say that of every 200 true hypotheses we will correctly identify 120 or 60%. Putting this together we find that of every 160 (120+40) hypotheses for which there is statistically significant evidence only 120 will in fact be true or a rate of 75% true.

Did you see that magic? Our confidence interval was 95%, no statistics were abused, no mistakes were made (beyond the ones falling into that 5% gap, which we accounted for), and yet we were only 75% correct.

The root of the problem is, of course, the ubiquitous problem of publication bias: Researchers like to publish, and people like to read so journals like to print, positive outcome studies rather than negative ones, because a journal detailing a long list of ideas that turned out to be wrong isn’t very exciting. The problem is, obviously, that published studies are therefore biased in favour of positive outcomes. (If not, all 800 studies of false hypotheses would have been published and the problem would disappear.)

Definition time again: A prior probability is essentially a plausibility measure before we run an experiment. Plausibility sounds very vague and subjective, but can be pretty concrete. If I know that it rains on (say) 50% of all winter days in Vancouver, I can get up in the morning and assign a prior probability of 50% to the hypothesis that it’s raining. (I can then run experiments, e.g. by looking out a window, and modify my assessment based on new evidence to come up with a posterior probability.)

Now we can go on to look at why Orac is so fond of holding hypotheses with low prior probabilities to higher standards. It’s pretty simple, really: Recall that the reason why we ended up with so many false positives above—the reason why false positives were such a large proportion of the published results—is because there were more false hypotheses than true hypotheses. The more conservative we are in generating hypotheses, the less outrageous we make them, the more likely we are to be correct, and the fewer false hypotheses we will have (in relation to true hypotheses). Put slightly differently, we’re more likely to be right in medical diagnoses if we go by current evidence and practice than if we make wild guesses.

Now we see that modalities with very low prior probability, such as ones with no plausible mechanism, should be regarded as more suspect. Recall that above, we started out with 800 false hypotheses (out of 1000 total hypotheses), ended up accepting 5% = 40 of them, and that

It's hard to say what the probability is of not finding evidence for a true hypothesis because it depends on a variety of factors such as the sample size but let's say that of every 200 true hypotheses we will correctly identify 120 or 60%. Putting this together we find that of every 160 (120+40) hypotheses for which there is statistically significant evidence only 120 will in fact be true or a rate of 75% true.

That is, the proportion of true hypotheses to false hypotheses affects the accuracy of our answer. This is very easy to see—let’s suppose that only half of the hypotheses were false; now we accept 5% of 500, that is 25 false studies, and keeping the same proportions,

…Let's say that of every 200 500 true hypotheses we will correctly identify 120 300 or 60%. Putting this together we find that of every 160 (120+40) 325 (300+25) hypotheses for which there is statistically significant evidence only 120 300 will in fact be true or a rate of 75% 92% true.

We’re still short of that 95% measure, but we’re way better than the original 75%, simply by making more plausible guesses (within each study, we were still equally likely to make either Type I or Type II errors). The less plausible an idea is, the higher the proportion of false hypotheses will be out of all the hypotheses the idea generates: A true/false ratio. Wild or vague ideas (homeopathy, reiki, …) are very likely to generate false hypotheses along with any true ones they might conceivably generate. More conventional ideas will tend to generate a higher proportion of true hypotheses—if we know from long experience that Aspirin relieves pain, it’s very likely that a similar drug does likewise.

This is not to say that no wild ideas are ever right. Of course they sometimes are (though of course they usually aren’t). What it does mean is that not only should we be skeptical and demand evidence for them, there are sound statistical reasons to set the bar of evidence even higher for implausible than for plausible modalities.

It is also a good argument for the move away from strict EBM (evidence-based medicine) to SBM (science-based medicine) where things like prior probability are taken into account. Accepting 95% double-blind trials at face value isn’t good enough.

Petter on Statistics and Medicine

Petter is such a good blogger, I am inclined to make him a contributor.

Once again, my most awesome commenter Petter Häggholm who wrote this guest post and this Ray a Day post and blogs over here (or by RSS; you can use tags to narrow it down to essay-style posts, or only posts related to skepticism or religion), has extended his EPIC WIN BRAINMEAT to my blog. Enjoy!

During my coffee break, I read an article in Scientific American Mind called Knowing Your Chances (available online). I think it is an outstanding article, and you should read it. The most evocative part may have been a simple example:

Consider a woman who has just received a positive result from a mammogram and asks her doctor: Do I have breast cancer for sure, or what are the chances that I have the disease? In a 2007 continuing education course for gynecologists, Gigerenzer asked 160 of these practitioners to answer that question given the following information about women in the region:

• The probability that a woman has breast cancer (prevalence) is 1 percent.
• If a woman has breast cancer, the probability that she tests positive (sensitivity) is 90 percent.
• If a woman does not have breast cancer, the probability that she nonetheless tests positive (false-positive rate) is 9 percent.

What is the best answer to the patient’s query?

1. The probability that she has breast cancer is about 81 percent.
2. Out of 10 women with a positive mammogram, about nine have breast cancer.
3. Out of 10 women with a positive mammogram, about one has breast cancer.
4. The probability that she has breast cancer is about 1 percent.

Before you read on, take a brief moment to think about it, but also note your gut feeling. Done? Let’s continue:

Gynecologists could derive the answer from the statistics above, or they could simply recall what they should have known anyhow. In either case, the best answer is C; only about one out of every 10 women who test positive in screening actually has breast cancer. The other nine are falsely alarmed. Prior to training, most (60 percent) of the gynecologists answered 90 percent or 81 percent, thus grossly overestimating the probability of cancer. Only 21 percent of physicians picked the best answer—one out of 10.
Doctors would more easily be able to derive the correct probabilities if the statistics surrounding the test were presented as natural frequencies. For example:

• Ten out of every 1,000 women have breast ­cancer.
• Of these 10 women with breast cancer, nine test positive.
• Of the 990 women without cancer, about 89 nonetheless test positive.

Thus, 98 women test positive, but only nine of those actually have the disease. After learning to translate conditional probabilities into natural frequencies, 87 percent of the gynecologists understood that one in 10 is the best answer.

I’m happy to say that I did get it right on the first try, but I strongly agree witht the authors’ opinion that it is not intuitive when the statistics are cited as probabilities rather than natural frequencies. The reason I got it right is because I’ve done a bit of math and a wee bit of stats, I enjoy reading some blogs that talk about medical statistics, I know some of the not-quite-obvious ground rules of probabilities; I know what Type I and Type II errors are (even if I occasionally mix them up)…

…And, perhaps crucially, I’ve spent time thinking about false positives in medical testing before. When I get my periodic routine screenings for STIs (I’ve never had symptoms or tested positive for any, I’m glad to say, but I feel a responsible person should get tested anyway!), I’ve asked myself the hypothetical question What if it did show positive for, say, HIV? What are the odds that I would actually have it? (It turns out that if you’re a heterosexual male, and if you test positive for HIV, there’s about a 50% chance that you don’t have it! You should play it safe, but get re-tested and don’t panic. Some people commit suicide when they get positive test results, even though they’re as likely as not to be healthy.)

Still, while my gut told me the answer was not A (wherein I did better than most of the gynecologists), I had to think about it for a minute to figure out which was the proper answer. People need to be educated on this stuff. Meanwhile, if you haven’t had the benefit of statistical education, keep this one thing in mind: The obvious answer is not always correct, so if you’re unsure, ask someone who can do the maths. And, sadly, even your doctor may not know. I actually find it rather sad that as after learning to translate conditional probabilities into natural frequencies, 87 percent of the gynecologists understood that one in 10 is the best answer, this means that even after simplification, more than 1 in 10 gynecologists didn’t get it. Your doctor can spot the symptoms and order the right tests, but you may need a mathematically inclined friend to actually calculate the risks.

acupuncture preventing nausea

One of my favorite websites is Science Daily, a science news service that publishes only science-related news stories – typically ones that are attached to actual research.

They have an “alternative medicine” section, and the other day I was browsing around and this headline caught my eye:

Wrist Acupuncture Or Acupressure Prevents Nausea From Anesthesia, Review Finds

A “review”, in scientific literature, is a study in which researchers look at a bunch of peer-reviewed studies (in which direct research was performed) on a particular topic or of a particular type, and analyze all of the studies to determine if the studies show a particular trend. In this case, a systemic review of acupuncture research on preventing post-anesthetic nausea and vomiting was performed.

Researchers claim that their review of the studies showed that wrist acupuncture or acupressure (of what is called the P6 region of the wrist) is a suitable alternative to anti-nausea medications. They reviewed studies that compared acupuncture/pressure to placebo acupuncture/pressure and anti-nausea meds.

Here is what they said of the studies:

The Cochrane reviewers found that compared to sham [placebo] treatment, stimulation of the P6 acupoint can significantly reduce the risk of nausea and vomiting after surgery, with few side effects. Lee said that “for 100 people, of whom 80 would vomit or feel sick after surgery if given sham treatment, about 25 people would benefit from P6 stimulation and 75 would not.”

I find the convoluted wording of this sentence particularly telling. What they are saying is that 80% of people felt sick/vomited after being given sham treatment, and 75% of people felt sick/vomited after being given appropriate P6 stimulation. 25 out of 100 sounds like a lot of people to benefit from P6 stimulation, but not when compared to 20 out of 100 for placebo stimulation.

Here is the next paragraph:

When compared to the use of antiemetic medications, however, the reviewers found no reliable evidence that showed any difference in whether acupoint stimulation is more effective in preventing postoperative nausea and vomiting. They concluded that it can be used as a “suitable alternative or addition to” the medications.

Lee said that in her homeland of Hong Kong, P6 stimulation methods are “not used widely, if at all.

When I read this last sentence, I though maybe there might actually be something to this acupuncture thing. If the research studies were performed in Hong Kong, where your average individual in the hospital will not know where the pressure point on the wrist for reducing nausea is located, then they won’t know the difference between placebo acupoint stimulation and real acupoint stimulation. That would be a great way to control for biases, as it would be hard to blind people to placebo acupoint stimulation in America, since lots of people know about acupuncture. In fact, surveys have concluded that about 10% of the general US population has tried acupuncture. In the hospital, those numbers are abound to be greater, since ill people are much more likely to try acupuncture than healthy people who have no need for alternative medicine.

But the research studies were not conducted in Hong Kong - they were conducted in America. In other words, it is safe to say that for a good number of these patients, the patients were not blinded to whether or not they were receiving placebo acupoint therapy or real therapy, because they know the difference, being familiar with it. Certainly, there are people who have tried acupoint therapy who couldn’t tell the difference, but I bet there are a substantial number of people out there who have not tried acupoint therapy yet know where the P6 wrist pressure point is – like me. As such, I think there are weaknesses in this review, especially given the extremely modest differences of 80% of people experiencing nausea with placebo vs 75% of people experiencing nausea with the real deal.

The Day they Kicked God out of the Schools - Rebuttal

The yesterday I posted a video which my friend had forwarded to me. Here is a transcription of the narrative of the video, and my (long...) rebuttal.

Dear God,

Why didn’t you save the schoolchildren at: Moses Lake, Washington; Bethyel Alaska; Pearl Mississippi; West Paducah, Kentucky; Stamp, Arkansas; Jonesboro, Arkansas; Edinboro, Pennsylvania; Fayetteville, Tennessee; Springfield, Oregon; Richmond, Virginia; Littleton, Colorado; Taber, Alberta Canada; Conyers, Georgia, Dinning, New Mexico; Fort Gibson, Oklahoma, Santee, California; El Cajon, California; Blacksburg Virginia.

Sincerely, Concerned Student

Reply: Dear concerned student,

I am not allowed in schools.

Sincerely, God

This video implies that these unsaved students were victims of school shootings, given the graphics of the video, which shows a map of the U.S. with the states (and Canadian provinces) in which the alleged nonsavings took place.

I've linked to some of the incidents mentioned in the video with quick references to the character of the individuals who were responsible for the shootings. I'll just do the first few, but you'll see a theme:

Moses Lake, Washington; mentally ill, had delusional and godlike thoughts, tormented by bullies.
Bethyel Alaska; sexually abused, depressed, suicidal, bullied
Pearl Mississippi; bullied, satanist
West Paducah, Kentucky; schizophrenic, bullied
Stamp, Arkansas; bullied

The fact of the matter is that mentally ill individuals attend schools, and sometimes mentally ill individuals snap and kill people. The fact of the matter is that kids bully other kids - even christian kids who pray and go to church bully other kids - Flimsy tells me that anyone who wasn't Christian at his high school would have been bullied mercilessly. At my high school, no one particularly cared what your religion was.

Nowhere are any of these incidents due to a kid's atheism.

I know this list of school shootings is long and impressive-sounding, especially with the sniper scope graphics, but let's look at this in a more realistic context. According to the CDC, there were 116 deaths from homicide at school between 1999-2006. Let's compare that to something else really unlikely, such as getting struck by lightning: there are about 100 deaths and 500 injuries due to lightning strike in the US every year.

So, what is the (I assume Christian) god's role in this? The narrator's god's snarky reply is that he simply isn't allowed in school. So, he's apparently not going to protect students because he's not allowed. Apparently the Christian god is not powerful enough to protect children in more secular situations. Apparently, not bothering to save children is some kind of punishment for not forcing children to pray to him.

But, it's not that the Christian god is "not allowed" in public schools. Public school children have not had their rights to practice their religion taken away, so it is only that the Christian god is not mandatory in schools. So apparently, the Christian god will not protect children unless the school mandates a daily prayer to him.

I have to wonder - why doesn't the Christian god save all of the people in church shootings? I am pretty sure he's mandatory there.

Fun! Let's move on, we're barely a quarter into the video:

” How did this get started? I think it started when Madalyn Murray O’hair complained, she didn’t want any prayer in our schools. And we said, “Ok”. Then someone said, “You better not read the Bible in school. The Bible says, “Thou shalt not kill, “Thou shalt not steal, and, Love your neighbor as yourself”…And we said, “Ok”. ”

No, that's not what happened. In 1963 in Abington Township School District v. Schempp the US Supreme Court declared that school sponsored Bible reading in public schools was unconstitutional.  The Abington case began when Edward Schempp, a Unitarian and a resident of Abington Township, Pennsylvania, filed suit against the Abington School District in the Federal District Court for the Eastern District of Pennsylvania to prohibit the enforcement of a Pennsylvania state law that required his children, specifically Ellory Schempp, to hear and sometimes read portions of the Bible as part of their public school education. That law (24 Pa. Stat. 15-1516, as amended, Pub. Law 1928) required that "[a]t least ten verses from the Holy Bible [be] read, without comment, at the opening of each public school on each school day." Schempp specifically contended that the statute violated his and his family's rights under the First and Fourteenth Amendments.[2]

Like four other states, Pennsylvania law included a statute compelling school districts to perform Bible readings in the mornings before class. Twenty-five states had laws allowing "optional" Bible reading, with the remainder having no laws supporting or rejecting Bible reading. In eleven of those states with laws supportive of Bible reading or state-sponsored prayer, courts[clarification needed] had declared them unconstitutional (Boston, 1993, p. 101).

Yes, you can still pray in school.Yes, you can still read the Bible in school.

On top of saying thou shalt not kill or steal and thou shalt love they neighbor and all of that happy stuff, it also says plenty of evil, hateful, disgusting things, which I shall not bother to point out here because you can find hateful or bizarre things in the Bible simply by opening it up to just about any random page of your choosing. I opened one of my bibles to Ezekiel 4:12 and found the Bible commanding people to eat barley cakes cooked over a fire which uses human excrement as fuel - um.. okay?

I recall very distinctly being taught not to murder or steal and to follow the golden rule about a million times, both by my school and my parents. I do not need the Bible to learn these rules. Seriously, you really can teach children appropriate morality without ever referencing the Bible. Lack of mandatory Bible readings in schools has nothing to do with mentally ill students killing other students.

” Dr. Benjamin Spock said that we shouldn’t spank our children when they misbehave, because their personalities would be warped, and we might damage their self-esteem. And we said, “An expert must know what he’s talking about, so we won’t spank them anymore”. ”

Many people other than Spock have determined that there are better, more effective ways of punishing children other than inflicting pain. Pain is not the only type of punishment that works for any animal, including children.There are also reams of peer-reviewed journal articles on this subject, in which actual research has been conducted, not just speculations. Spanking has been linked to abuse, child aggression and anxiety, and increased risk of sexual problems. If one actually bothers to look at the way children learn and spend five minutes consulting reputable literature, one will find that punishments are good, but spanking is not as effective a punishment as other types of punishment. One should absolutely punish their children for misbehaving, but use effective punishments.

Besides, um, spanking is still legal in US schools. Spanking is, however, NOT legal in any European school and spanking is illegal even in the home in many European countries. Funny, the rates of atheism in Europe are higher, spanking is illegal - and the rates of crime are lower. Related specifically to the school shootings, I can't say for sure but I can bet that every one of those children were spanked or abused in some way, either by their parents or someone else.

” Then someone said, “Teachers and principals better not discipline our children when they misbehave”. And the school administration said, “No faculty member in this school better touch a student when they misbehave, because we don’t want any bad publicity, and we surely don’t want to be sued”. And we accepted their reasoning. ”

Wrong. Children are absolutely disciplined when they misbehave. There are ways to discipline a child without inflicting physical pain. Lack of spanking has nothing to do with mentally ill students killing other students.

” Then someone said, “Let’s let our daughters have abortions if they want, and they won’t even have to tell their parents”. And we said, “That’s a grand idea”. Then, some wise school board member said, “since boys will be boys, and their going to do it anyway, let’s give our sons all the condoms they want, so they can have all the fun they desire, and we won’t have to tell their parents they got them at school”. And we said, “That’s another great idea”. ”

It being legal for minors to have abortions is not the same as "letting them have abortions if they want" and abortion is a healthcare issue - minors are entitled to the same rights to healthcare as adults. Only six states have no parental consent laws, and the rest of them do - surprise, there is no rash of school shootings in those particular states. Oh and abortion was legalized in 1973, and crime started dropping off in 1990 - all of those legally aborted babies would have been 17 years old or younger. It has been suggested that a possible hypothesis for the sharp downturn in crime is due to the legalization of abortion, and at least in this case we have an obvious correlation. See the Wiki article on this hypothesis here.

Besides, which, abortions have nothing to do with mentally ill students killing other students.

Okay, condoms. If one takes five minute out of their busy day being hateful and bigoted to actually look at studies done on sexual activity among high school students, one will find that the sexual activity levels are the exact same whether condoms are given out freely or not. Sexual activity is also the same whether we teach our kids abstinence-only or comprehensive sex education. The difference is that kids taught comprehensive sex education are more likely to use condoms. Oh yeah, and condoms have nothing to do with mentally ill students killing other students.

” Then, some of our top elected officials said, “It doesn’t matter what we do in private, as long as we do our jobs”. And we said, “It doesn’t matter what anybody, including the president does in private, as long as we have jobs, and the economy is good”. And someone else took that appreciation a step further, and published pictures of nude children, and then stepped further, by making them available on the internet. And we said, “Everyone’s entitled to free speech”. ”

I don't think that's what they said. I think this refers to an individual's right to privacy. The point was not that only the economy matters, but that people have the right to engage in legal activities while not working without getting fired. Child pornography is illegal, and it should be. If someone claims that child pornography is "free speech", they are either deranged or trying to build up a pathetic straw-man argument.  This has nothing to do with mentally ill students killing other students. Tired of me repeating that yet?

” And the entertainment industry said, “Let’s make TV shows and movies that promote profanity, violence and illicit sex; and let’s record music that encourages rape, drugs murder, suicide and Satanic themes”. And we said, “It’s just entertainment, and it has no adverse effect, and nobody takes it seriously anyway, so go right ahead”. ”

...Okay, you know what? You might actually have something possibly slightly true here. FINALLY! Studies have had conflicting findings on whether or not depictions of violence lead to violence. I've got to ask though - why did this narration leave out the news media reporting industry? This is the only place in which real violence is depicted.

Here is the problem though... Crime has dropped off significantly since 1993:

So if violence on TV is to blame, one would think that this drop off in crime is due to a drop off in depictions of violence. If anything, depictions of violence have gone up, and certainly more violent video games have been released. Oh yeah and.. atheism is on the rise.

” Now we’re asking ourselves why our children have no conscience; why they don’t know right from wrong; and why it doesn’t bother them to kill strangers, classmates or even themselves. Undoubtedly, if we thought about it long and hard enough, we could figure it out. Surely, it has a lot to do with: ” We reap what we sow “.

Your premise fails miserably to the point of being laughable, because we're actually not living in more violent times than back in the "good old days". We're actually enjoying very low crime rates right now. So, rather than your premise supporting that all of the things mentioned in this narration are a cause of ills of society, actual data shows the opposite. If we do reap what we sow, right now we're reaping reduced crime rates!

Children have a perfectly functional conscience, they do know right from wrong, and it does bother them to kill strangers, classmates and themselves. A few mentally ill students killing other students is not a good way to judge the rest of the children in the US, nor is it a good way to judge the crime climate in the US. There will always be mentally ill people in society, and we're not living in more violent times. Still don't believe me? Here, ask Stephen Pinker:



It's time to be realistic. God didn't save these kids - because there is no god.  The crators of this video are using it as a propaganda tool to promote their particular brand of religiosity, and I find it despicable that they would ignore real data in favor of sensationalism while disrespecting the young victims of violent crimes to furthur their own agenda. The creators of this video should be ashamed.

Ray a Day Guest Post: Alien

Here is a guest Ray a Day by The Alien, who is my longtime friend (we've known each other like 10 years):

I am sure I am adding to the beating of a dead, decimated, now barely visible as once a living being horse here with my post, however that does seem to be the point of the Ray A Day blog entries, so I will also enjoy adding to the sense-ful beating. Ray says:

Yet God, in His justice and holiness, cannot let our sins remain unpunished. More than that, because of His love for us, He hates it that sin separates us from Him. He knows even better than we how desperately we need saving, and the only One powerful enough to save us is the One Who created us in the first place. God, therefore, did the only thing He could do that would satisfy both His justice and His love for us. In the Person of Jesus Christ, he took our sins upon Himself, thereby paying our debt and offering us a chance at salvation and a restored relationship with Him.

The Alien holds her head in her hands as she reads this, amazed at the fact that a rather educated-seeming man could write something that contradicts itself not once, but twice, in just one paragraph. It starts with idea that God is All. There is Nothing Without God - note that starts the bible itself. It's the first...words. "In The Beginning God Created the Heavens and the Earth."

If God created everything, and everything is created by God, then nothing can be not created by God. The snake, the devil, all was created by him...and what the Alien fails to understand how the Devil cannot be his creation or become something different than his creation - if nothing else exists that is not of God's creation. I will attempt to render this in mathematical format as I see it. Let's represent God as one. We'll represent the Devil as 0 as it is seen as different, outside of God.

1=1 (there is nothing else)

1+0=1 (Even with the addition of something new, everything still equals God.)

We'll take more numbers here. Thinking of everything God has created, we can assume that they were all a part of him at some point as everything was God to start. The number is huge, so I will make a more comprehensible number here, let's say 321.

321 = 1^321  (All parts of God)

321+0=321 (All parts of God plus devil...don't do anything)

No matter what we do mathematically here (in basic math anyway), the devil does not, cannot, hurt anything. It cannot separate or divide something which is part of a larger something. How is it, Mr. Comfort, if 0 cannot separate 321 parts of God, how can the devil? He cannot, of course, because in the normal world (not the mathematical one), he does not exist! He cannot exist as away from God when it is impossible by the words of your own bible, to be separate from him! No matter what we do, feel, think, or say, we are saying everything that can only exist because we are a part of Him!

Of course, the next part just reads even more insensibly using my same thoughts above. By the words of the Bible and by Comfort here, apparently a Part of God came down to reside in another Part of God to help cleanse other Parts of God from something that does not exist.( I shall use the |, which means Restriction.) Mathematically, 1+1=2 | 0

So us Parts of God needed another Part of God to pay our debt against a nonexistance that does nothing to us in order to guarantee our restoration of a relationship with God.

The Alien wonders if any Christians actually think of the mathematics behind their God and Devil's existance -  because no matter what way I crunch the numbers, even if God exists, God is highly more likely to exist than the Devil is, and Jesus suddenly seems...unnecessary.