1 2 3 16

Cranio-sacral therapy is firstly implausible, and secondly it lacks evidence of effectiveness (see for instance here, here, here and here). Yet, some researchers are nevertheless not deterred to test it in clinical trials. While this fact alone might be seen as embarrassing, the study below is a particular and personal embarrassment to me, in fact, I am shocked by it and write these lines with considerable regret.

Why? Bear with me, I will explain later.

The purpose of this trial was to evaluate the effectiveness of osteopathic manipulative treatment and osteopathy in the cranial field in temporomandibular disorders. Forty female subjects with temporomandibular disorders lasting at least three months were included. At enrollment, subjects were randomly assigned into two groups: (1) osteopathic manipulative treatment group (n=20) and (2) osteopathy in the cranial field [craniosacral therapy for you and me] group (n=20). Examinations were performed at baseline (E0) and at the end of the last treatment (E1), and consisted of subjective pain intensity with the Visual Analog Scale, Helkimo Index and SF-36 Health Survey. Subjects had five treatments, once a week. 36 subjects completed the study.

Patients in both groups showed significant reduction in Visual Analog Scale score (osteopathic manipulative treatment group: p = 0.001; osteopathy in the cranial field group: p< 0.001), Helkimo Index (osteopathic manipulative treatment group: p = 0.02; osteopathy in the cranial field group: p = 0.003) and a significant improvement in the SF-36 Health Survey – subscale “Bodily Pain” (osteopathic manipulative treatment group: p = 0.04; osteopathy in the cranial field group: p = 0.007) after five treatments (E1). All subjects (n = 36) also showed significant improvements in the above named parameters after five treatments (E1): Visual Analog Scale score (p< 0.001), Helkimo Index (p< 0.001), SF-36 Health Survey – subscale “Bodily Pain” (p = 0.001). The differences between the two groups were not statistically significant for any of the three endpoints.

The authors concluded that both therapeutic modalities had similar clinical results. The findings of this pilot trial support the use of osteopathic manipulative treatment and osteopathy in the cranial field as an effective treatment modality in patients with temporomandibular disorders. The positive results in both treatment groups should encourage further research on osteopathic manipulative treatment and osteopathy in the cranial field and support the importance of an interdisciplinary collaboration in patients with temporomandibular disorders. Implications for rehabilitation Temporomandibular disorders are the second most prevalent musculoskeletal condition with a negative impact on physical and psychological factors. There are a variety of options to treat temporomandibular disorders. This pilot study demonstrates the reduction of pain, the improvement of temporomandibular joint dysfunction and the positive impact on quality of life after osteopathic manipulative treatment and osteopathy in the cranial field. Our findings support the use of osteopathic manipulative treatment and osteopathy in the cranial field and should encourage further research on osteopathic manipulative treatment and osteopathy in the cranial field in patients with temporomandibular disorders. Rehabilitation experts should consider osteopathic manipulative treatment and osteopathy in the cranial field as a beneficial treatment option for temporomandibular disorders.

This study has so many flaws that I don’t know where to begin. Here are some of the more obvious ones:

  • There is, as already mentioned, no rationale for this study. I can see no reason why craniosacral therapy should work for the condition. Without such a rationale, the study should never even have been conceived.
  • Technically,  this RCTs an equivalence study comparing one therapy against another. As such it needs to be much larger to generate a meaningful result and it also would require a different statistical approach.
  • The authors mislabelled their trial a ‘pilot study’. However, a pilot study “is a preliminary small-scale study that researchers conduct in order to help them decide how best to conduct a large-scale research project. Using a pilot study, a researcher can identify or refine a research question, figure out what methods are best for pursuing it, and estimate how much time and resources will be necessary to complete the larger version, among other things.” It is not normally a study suited for evaluating the effectiveness of a therapy.
  • Any trial that compares one therapy of unknown effectiveness to another of unknown effectiveness is a complete and utter nonsense. Equivalent studies can only ever make sense, if one of the two treatments is of proven effectiveness – think of it as a mathematical equation: one equation with two unknowns is unsolvable.
  • Controlled studies such as RCTs are for comparing the outcomes of two or more groups, and only between-group differences are meaningful results of such trials.
  • The ‘positive results’ which the authors mention in their conclusions are meaningless because they are based on such within-group changes and nobody can know what caused them: the natural history of the condition, regression towards the mean, placebo-effects, or other non-specific effects – take your pick.
  • The conclusions are a bonanza of nonsensical platitudes and misleading claims which do not follow from the data.

As regular readers of this blog will doubtlessly have noticed, I have seen plenty of similarly flawed pseudo-research before – so, why does this paper upset me so much? The reason is personal, I am afraid: even though I do not know any of the authors in person, I know their institution more than well. The study comes from the Department of Physical Medicine and Rehabilitation, Medical University of Vienna, Austria. I was head of this department before I left in 1993 to take up the Exeter post. And I had hoped that, even after 25 years, a bit of the spirit, attitude, knowhow, critical thinking and scientific rigor – all of which I tried so hard to implant in my Viennese department at the time – would have survived.

Perhaps I was wrong.

Difficulties breastfeeding?

Some say that Chinese herbal medicine offers a solution.

This Chinese multi-centre RCT included 588 mothers considering breastfeeding. The intervention group received the Chinese herbal mixture Zengru Gao, while the control group received no therapy. The primary outcomes were the percentages of fully and partially breastfeeding mothers, and a secondary outcome was baby’s daily formula intake.

At day 3 and 7 after delivery, significant differences were found in favour of Zengru Gao group on the percentage of full/ partial breastfeeding. At day 7, the percentage of full/ partial breastfeeding of the active group increased to 71.48%/20.70% versus 58.67%/30.26% in the control group, the differences remained significant. No statistically significant differences were detected on primary measures at day. While intake of formula differed between groups at day 1 and 3, this difference did not achieve statistical significance, but this difference was apparent by day 7.

The authors concluded that the Chinese Herbal medicine Zengru Gao enhanced breastfeeding success during one week postpartum. The approach is acceptable to participants and merits further evaluation.

To the naïve observer, this study might look rigorous, but it is a seriously flawed RCT. Here are just some of its most obvious limitations:

  • All we get in the methods section is this explanation: Participants were randomly allocated to the blank control group or the intervention group: Zengru Gao, orally, 30 g a time and 3 times a day. This seems to indicate that the control group got no treatment at all which means there was no blinding nor placebo control. The authors even comment on this point in the discussion section of their paper stating that because we included new mothers who received no treatment as a control group, we were able to prove that the improvement in breastfeeding was not due to the placebo effect. However, this is a totally nonsensical argument.
  • The experimental treatment is not reproducible. The authors state: Zengru Gao, a Chinese herbal formula, which is composed of 8 herbs: Semen Vaccariae, Medulla Tetrapanacis, Radix Rehmanniae Praeparata, Radix Angelicae Sinensis, Radix Paeoniae Alba,Rhizoma Chuanxiong, Herba Leonuri, Radix Trichosanthis. This is not enough information to replicate the study outside China where the mixture is not commercially available.
  • The primary outcome was the percentage of fully, and partially breastfeeding mothers. Breastfeeding was defined as mother’s milk given by direct breast feeding. Full breastfeeding meant that no other types of milk or solids were given. Partially breastfeeding meant that sustained latch with deep rhythmic sucking through the length of the feed, with some pause, on either/ or both breasts. We are not being told how the endpoint was quantified. Presumably women kept diaries. We cannot guess how accurate this process was.
  • As far as I can see, there was no correction for multiple testing for statistical significance. This means that some or all of the significant results might be false-positive.
  • There is insufficient data to show that the herbal mixture is safe for the mothers and the babies. At the very minimum, the researchers should have measured essential safety parameters. This omission is a gross violation of research ethics.
  • Towards the end of the paper, we find the following statement: The authors would like to thank the Research and Development Department of Zhangzhou Pien Tze Huang Pharmaceutical co., Ltd. … The authors declare that they have no competing interests. And the 1st and 3rd authors are “affiliated with” Guangzhou Hipower Pharmaceutical Technology Co., Ltd, Guangzhou, China, i. e. work for the manufacturer of the mixture. This does clearly not make any sense whatsoever.

I have seen too many flawed studies of alternative medicine to be shocked or even surprised by this level of incompetence and nonsense. Yet, I still find it lamentable. But, in my view, the worst is that supposedly peer-reviewed journals such as ‘BMC Complement Altern Med’ publish such overt rubbish.

It would be easy to shrug one’s shoulder and bin the paper. But the effect of such fatally flawed research is too serious for that. In our recent book MORE HARM THAN GOOD? THE MORAL MAZE OF COMPLEMENTARY AND ALTERNATIVE MEDICINE, we discuss that such flawed science amounts to a violation of medical ethics:  CAM journals allocate peer review tasks to a narrow range of CAM enthusiasts who often have been chosen by the authors of the article in question. The raison d’être of CAM journals and CAM researchers is inextricably tied to a belief in CAM, resulting in a self-referential situation which is permissive to the acceptance of weak or flawed reports of clinical effectiveness… Defective research—whether at the design, execution, analysis, or reporting stage—corrupts the repository of reliable medical knowledge. Ultimately, this leads to suboptimal and erroneous treatment decisions…

The authors of this systematic review aimed to summarize the evidence of clinical trials on cupping for athletes. Randomized controlled trials on cupping therapy with no restriction regarding the technique, or co-interventions, were included, if they measured the effects of cupping compared with any other intervention on health and performance outcomes in professionals, semi-professionals, and leisure athletes. Data extraction and risk of bias assessment using the Cochrane Risk of Bias Tool were conducted independently by two pairs of reviewers.

Eleven trials with n = 498 participants from China, the United States, Greece, Iran, and the United Arab Emirates were included, reporting effects on different populations, including soccer, football, and handball players, swimmers, gymnasts, and track and field athletes of both amateur and professional nature. Cupping was applied between 1 and 20 times, in daily or weekly intervals, alone or in combination with, for example, acupuncture. Outcomes varied greatly from symptom intensity, recovery measures, functional measures, serum markers, and experimental outcomes. Cupping was reported as beneficial for perceptions of pain and disability, increased range of motion, and reductions in creatine kinase when compared to mostly untreated control groups. The majority of trials had an unclear or high risk of bias. None of the studies reported safety.

Risk of bias of included trials. “+” indicates low risk of bias, “−” indicates high risk of bias, and “?” indicates unclear risk of bias.

The authors concluded that no explicit recommendation for or against the use of cupping for athletes can be made. More studies are necessary for conclusive judgment on the efficacy and safety of cupping in athletes.

Considering the authors’ stated aim, this conclusion seems odd. Surely, they should have concluded that THERE IS NO CONVINCING EVIDENCE FOR THE USE OF CUPPING IN ATHLETES. But this sounds rather negative, and the JCAM does not seem to tolerate negative conclusions, as discussed repeatedly on this blog.

The discussion section of this paper is bar of any noticeable critical input (for those who don’t know: the aim of any systematic review must be to CRITICALLY EVALUATE THE PRIMARY DATA). The authors even go as far as stating that the trials reported in this systematic review found beneficial effects of cupping in athletes when compared to no intervention. I find this surprising and bordering on scientific misconduct. The RCTs were mostly not on cupping but on cupping in combination with some other treatments. More importantly, they were of such deplorable quality that they allow no conclusions about effectiveness. Lastly, they mostly failed to report on adverse effects which, as I have often stated, is a violation of research ethics.

In essence, all this paper proves is that, if you have rubbish trials, you can produce a rubbish review and publish it in a rubbish journal.

The authors of this review aimed to present an overview of the literature on physicochemical research performed on homeopathic preparations with respect to publication quality and methods used. They searched major scientific databases to find relevant publications from its origin to the end of 2015. Publications were assessed using a scoring scheme, the Manuscript Information Score (MIS). Information regarding country of origin of the research and experimental techniques used was extracted.

The authors identified 183 publications. The rate of publication in the field was 2 per year from the 1970s until 2000. Afterward, it increased to over 5.5 publications per year. The quality of publications was seen to increase sharply from 2000 onward: before 2000, only 12 (13%) publications were rated as ‘‘high quality’’ (MIS ‡7.5); 44 (48%) publications were rated as ‘‘high quality’’ after 2000.

Countries with most publications were Germany (n=42, 23%), France (n=29, 16%), India (n=27, 15%), and Italy (n=26, 14%). Techniques most frequently used were electrical impedance (26%), analytical methods (20%), spectroscopy (20%), and nuclear magnetic resonance (19%).

The authors concluded that physicochemical research into homeopathic preparations is increasing both in terms of quantity and quality of the publications.

They also announce that there will be a further paper on the subject: In part 2, we aim to identify the most interesting experimental techniques. With this, we aim to be in a position to generate meaningful hypotheses regarding a possible mode of action of homeopathic preparations.

It might be relevant to ask who the authors are and where they come from. They are Klein SD, Würtenberger S, Wolf U, Baumgartner S, and Tournier A. And their affiliations are:

  • Institute of Complementary Medicine, University of Bern, Switzerland.
  • Scientific & Regulatory Affairs, Hevert-Arzneimittel GmbH & Co. KG, Germany.
  • Society for Cancer Research, Arlesheim, Switzerland.
  • Institute of Integrative Medicine, University of Witten/Herdecke, Germany.
  • Homeopathy Research Institute (HRI), London, United Kingdom.

In other words, they are without exception proponents of homeopathy, some burdened with considerable conflicts of interest in the subject. Personally, I think it unlikely that anything meaningful will ever come of this research. But unsurprisingly, the enthusiasts beg to differ: on facebook, the HRI claimed that this new systematic review is a major step towards developing clear and testable hypotheses regarding the mode of action of homeopathy.

In a previous blog-post I have tried to explain my reservations in some detail; please allow me to repeat them here:

… homeopaths have been keen to find more rational support for their theories. Thus they have developed several ‘sciency’ concepts to explain the mode of action of their highly diluted homeopathic remedies. For instance they postulated that water can form secondary structures that hold some information of the original substance (stock), even if it has long been diluted out of the remedy. Alternatively, they claimed that the shaking of the remedy generates nano-particles or silicone-particles which, in turn, are the cause of the clinical effects.

Today, I want to assume for a minute, that one of these theories is correct – they cannot all be right, of course. Homeopaths regularly show us investigations that seem to support them, even though it only needs a real expert in the particular field of science to cast serious doubt on them. I will nevertheless assume that, after potentisation, the diluent retains information via nano-particles or some other phenomenon. For the purpose of this mind-experiment, I grant homeopaths that, in this respect, they are correct. In other words, let’s for a moment assume that the ‘memory of water’ theory is correct.

As I have been more than generous, I want homeopaths to return the favour and consider what this would really mean: information has been transferred from the stock to the diluent. Does that prove anything? Does it show that homeopathy is valid?

Could the homeopaths who make this assumption be equally generous and answer the following questions, please?

  1. How does a nano-particle of coffee, for instance, affect the sleep centre in the brain to make the patient sleep? Or how does a nano-particle of the Berlin Wall or a duck liver affect anything at all in the human body? The claim that information has been retained by the diluent is no where near to an explanation of a rational mode of action, isn’t it?
  2. Most homeopathic remedies are consumed not as liquids but as ‘globuli’, i. e.  tiny little pills made of lactose. They are prepared by dropping the liquid remedy on to them. The liquid subsequently evaporates. How is it that the information retained in the liquid does not evaporate with the diluent?
  3. The diluent usually is a water-alcohol mixture which inevitably contains impurities. In fact, a liquid C12 remedy most certainly contains dimensions more impurities than stock. These impurities have, of course, also been vigorously shaken, i. e. potentised. How can we explain that their ‘potency’ has not been beefed up at each dilution step? Would this not necessitate a process where only some molecules in the diluent are agitated, while all the rest remain absolutely still? How can we explain this fantastic concept?
  4. Some stock used in homeopathy is insoluble (for instance Berlin Wall). Such stock is not diluted but its concentration in the remedy is initially lowered by a process called ‘trituration’, a process which consists in grinding the source material in another solid material, usually lactose. I have granted you that potentisation works in the way you think. But how is information transferred from one solid material to another?
  5. Everything we drink is based on water containing molecules that have been inadvertently potentised in nature a million times and therefore should have hugely powerful effects on our bodies. How is it that we experience none of these effects each time we drink?

Now, homeopaths, let me propose a deal.

If you can answer these questions satisfactorily, I will no longer doubt your memory of water theory. If you cannot do this, I think you ought to admit that all your ‘sciency’ theories about the mode of action of highly diluted homeopathic remedies are really quite silly – more silly even than Hahnemann’s idea of a ‘spirit-like’ effect.





Traditional Chinese Medicine (TCM) is popular, not least because it is heavily marketed and thus often perceived as natural and safe. But is this assumption true?

This study analysed  liver tests before and following treatment with herbal Traditional Chinese Medicine (TCM) in order to evaluate the risk of liver injury. Patients with normal values of alanine aminotransferase (ALT) as a diagnostic marker for ruling out pre-existing liver disease were enrolled in a prospective study of a safety program carried out at the First German Hospital of TCM from 1994 to 2015. All patients received herbal products, and their ALT values were reassessed 1-3 d prior to discharge. To evaluate causality for suspected TCM herbs, the Roussel Uclaf Causality Assessment Method (RUCAM) was used.

The report presents data of 21470 patients. ALT ranged from 1 × to < 5 × upper limit normal (ULN) in 844 patients (3.93%) and suggested mild or moderate liver adaptive abnormalities. A total of 26 patients (0.12%) experienced higher ALT values of ≥ 5 × ULN (300.0 ± 172.9 U/L, mean ± SD). Causality for TCM herbs was estimated to be probable in 8/26 patients, possible in 16/26, and excluded in 2/26 cases.

Compared with the large TCM study cohort, patients in the liver injury study cohort were older and contained a higher percentage of women, whereas the duration of the hospital stay was similar in both cohorts. The TCM herbs were rarely applied mostly as mixtures consisting of several herbs adding up to 35 different drugs during the patients’ four-week stay. The daily dosage was 95 ± 30 g and thus slightly higher than in the TCM study cohort. Among the many herbal TCM used by the 26 patients in the liver injury cohort, Bupleuri radix and Scuterllariae radix were the two TCM herbs most frequently implicated in liver injury, with variable RUCAM-based causality gradings. Most of the patients received one to six TCM drugs that were associated with potential liver injury as evidenced from the scientific literature, e.g., one patient (case 8) received six potentially hepatotoxic herbal TCM drugs during their hospital stay.

The authors concluded that in 26 (0.12%) of 21470 patients treated with herbal TCM, liver injury with ALT values of ≥ 5 × ULN was found, which normalized shortly following treatment cessation, also substantiating causality.

In the discussion section of the paper, the authors comment that the use of TCM is widely considered less risky as compared with synthetic drugs, although data on direct comparisons are not available in support of this view. Populations using herbal TCM, drugs, either alone, or combined experience more drug-induced liver injury (DILI) than herb-induced liver injury (HILI), possibly due to a higher use of drugs. Valid data of incidence and prevalence of HILI caused by TCM herbs are lacking, and respective data cannot be derived from the present study.

This study is most valuable, in my view. Its strength is clearly the huge sample size. Top marks for the authors for publishing it!

Having said that, we need to take the incidence figures with a pinch of salt, I think. In reality they could be much higher because:

  • other settings will not be as tightly supervised as the unusual hospital setting;
  • in most other situations the quality of the Chinese herbs might be less controlled;
  • there could be adulteration;
  • there could be contamination.

The ‘elephant in the room’ obviously is the inevitable question about benefit. Like any other treatment, TCM cannot be judged on the basis of its risk but must be evaluated according to its risk/benefit balance. I realise that this was not the subject of the present study, but it is nevertheless crucial: do the benefits of TCM outweigh its risks?

I am not aware that this is the case (but more than willing to consider any sound evidence readers might supply). More importantly, I am not aware of good evidence to show that, for any condition, TCM would be superior in terms of risk/benefit balance than conventional options. This is not a trivial issue: clinicians have the ethical obligation to apply the best (the one with the most positive risk/benefit balance) treatment to their patients.

If I am right, then TCM should not be used in therapeutic routine in or outside hospitals.

If I am right, the ‘First German Hospital of TCM‘ should close asap; it would be violating fundamental ethical principles.

If I am right, the debate about the risks of TCM is almost irrelevant because we simply should not use it.

Or did I misunderstand something here?

What do you think?


The aim of this three-armed, parallel, randomized exploratory study was to determine, if two types of acupuncture (auricular acupuncture [AA] and traditional Chinese acupuncture [TCA]) were feasible and more effective than usual care (UC) alone for TBI–related headache. The subjects were previously deployed Service members (18–69 years old) with mild-to-moderate TBI and headaches. The interventions explored were UC alone or with the addition of AA or TCA. The primary outcome was the Headache Impact Test (HIT). Secondary outcomes were the Numerical Rating Scale (NRS), Pittsburgh Sleep Quality Index, Post-Traumatic Stress Checklist, Symptom Checklist-90-R, Medical Outcome Study Quality of Life (QoL), Beck Depression Inventory, State-Trait Anxiety Inventory, the Automated Neuropsychological Assessment Metrics, and expectancy of outcome and acupuncture efficacy.

Mean HIT scores decreased in the AA and TCA groups but increased slightly in the UC-only group from baseline to week 6 [AA, −10.2% (−6.4 points); TCA, −4.6% (−2.9 points); UC, +0.8% (+0.6 points)]. Both acupuncture groups had sizable decreases in NRS (Pain Best), compared to UC (TCA versus UC: P = 0.0008, d = 1.70; AA versus UC: P = 0.0127, d = 1.6). No statistically significant results were found for any other secondary outcome measures.

The authors concluded that both AA and TCA improved headache-related QoL more than UC did in Service members with TBI.

The stated aim of this study (to determine whether AA or TCA both with UC are more effective than UC alone) does not make sense and should therefore never have passed ethics review, in my view. The RCT followed a design which essentially is the much-lamented ‘A+B versus B’ protocol (except that a further groups ‘C+B’ was added). The nature of such designs is that there is no control for placebo effects, the extra time and attention, etc. Therefore, such studies cannot fail but generate positive results, even if the tested intervention is a placebo. In such trials, it is impossible to attribute any outcome to the experimental treatment. This means that the positive results are known before the first patient has been enrolled; hence they are an unethical waste of resources which can only serve one purpose: to mislead us. It also means that the conclusions drawn above are not correct.

An alternative and in my view more accurate conclusion would be this one: both AA and TCA had probably no effect; the improved headache-related QoL was due to the additional attention and expectation in the two experimental groups and is unrelated to the interventions tested in this study.

In our new book, MORE HARM THAN GOOD, we discuss that such trials are deceptive to the point of being unethical. Considering the prominence and experience of Wayne Jonas, the 1st author of this paper, such obvious transgression is more than a little disappointing – I would argue that is amounts to overt scientific misconduct.


Take for instance this tweet I got yesterday:

F SThomas‏ @spenthomf

You go too far @EdzardErnst. In fact I was consulted about a child who hadn’t grown after an accident. She responded well to homoeopathy and grew. How much are you being paid for your attempts to deny people’s health choices?

The tweet refers to my last post where I exposed homeopathic child abuse. Having thought about Thomas’ tweet, I must say that I find it too to be abusive – even abusive on 4 different levels.

  1. First, the tweet is obviously a personal attack suggesting that I am bribed into doing what I do. I have stated it many times, and I do so again: I receive no payment from anyone for my work. How then do I survive? I have a pension and savings (not that this is anyone’s business).
  2. Second, it is abusive because it claims that children who suffer from a pathological growth retardation can benefit from homeopathy. There is no evidence for that at all, and making false claims of this nature is unethical and, in this case, even abusive.
  3. Third, if Thomas really did make the observation she suggests in her tweet and is convinced that her homeopathic treatment was the cause of the child’s improvement, she has an ethical duty to do something more about it than just shooting off a flippant tweet. She could, for instance, run a clinical trial to find out whether her observation was correct. I admit this might be beyond her means. So alternatively, she could write up the case in full detail and publish it for all of us to scrutinise her findings. This is the very minimum a responsible clinician ought to do when she comes across a novel and potentially important result. Anything else is my view unethical and hinders progress.

I do, of course, sympathise with lay people who fail to fully understand the concept of causality. But surely, healthcare professionals who pride themselves of taking charge of patients ought to have some comprehension of it. They should know that clinical improvements after a treatment is not necessarily the same as clinical improvement because of the treatment. Is it really too much to ask of them to know the criteria for causality? There is plenty of easy-reading on the subject; even Wikipedia has a good article on it:

In 1965, the English statistician Sir Austin Bradford Hill proposed a set of nine criteria to provide epidemiologic evidence of a causal relationship between a presumed cause and an observed effect. (For example, he demonstrated the connection between cigarette smoking and lung cancer.) The list of the criteria is as follows:

  1. Strength (effect size): A small association does not mean that there is not a causal effect, though the larger the association, the more likely that it is causal.
  2. Consistency (reproducibility): Consistent findings observed by different persons in different places with different samples strengthens the likelihood of an effect.
  3. Specificity: Causation is likely if there is a very specific population at a specific site and disease with no other likely explanation. The more specific an association between a factor and an effect is, the bigger the probability of a causal relationship.
  4. Temporality: The effect has to occur after the cause (and if there is an expected delay between the cause and expected effect, then the effect must occur after that delay).
  5. Biological gradient: Greater exposure should generally lead to greater incidence of the effect. However, in some cases, the mere presence of the factor can trigger the effect. In other cases, an inverse proportion is observed: greater exposure leads to lower incidence.
  6. Plausibility: A plausible mechanism between cause and effect is helpful (but Hill noted that knowledge of the mechanism is limited by current knowledge).
  7. Coherence: Coherence between epidemiological and laboratory findings increases the likelihood of an effect. However, Hill noted that “… lack of such [laboratory] evidence cannot nullify the epidemiological effect on associations”.
  8. Experiment: “Occasionally it is possible to appeal to experimental evidence”.
  9. Analogy: The effect of similar factors may be considered.

And this brings me to my 4th and last level of abuse in relation to the above tweet and most other claims of this nature: being ill-informed and stupid while insisting to make a nonsensical point is, in my view, offensive – so much so that it can reach the level of abuse.

Can conventional therapy (CT) be combined with herbal therapy (CT + H) in the management of Alzheimer’s disease (AD) to the benefit of patients? This was the question investigated by Chinese researchers in a recent retrospective cohort study funded by grants from China Ministry of Education, National Natural Science Foundation of China, Beijing Municipal Science and Technology Commission, and Beijing Municipal Commission of Health and Family Planning.

In total, 344 outpatients diagnosed as probable dementia due to AD were collected, who had received either CT + H or CT alone. The GRAPE formula was prescribed for AD patients after every visit according to TCM theory. It consisted mainly (what does ‘mainly’ mean as a description of a trial intervention?) of Ren shen (Panax ginseng, 10 g/d), Di huang (Rehmannia glutinosa, 30 g/d), Cang pu (Acorus tatarinowii, 10 g/d), Yuan zhi (Polygala tenuifolia, 10 g/d), Yin yanghuo (Epimedium brevicornu, 10 g/d), Shan zhuyu (Cornus officinalis, 10 g/d), Rou congrong (Cistanche deserticola, 10 g/d), Yu jin (Curcuma aromatica, 10 g/d), Dan shen (Salvia miltiorrhiza, 10 g/d), Dang gui (Angelica sinensis, 10 g/d), Tian ma (Gastrodia elata, 10 g/d), and Huang lian (Coptis chinensis, 10 g/d), supplied by Beijing Tcmages Pharmaceutical Co., LTD. Daily dose was taken twice and dissolved in 150 ml hot water each time. Cognitive function was quantified by the mini-mental state examination (MMSE) every 3 months for 24 months.

The results show that most of the patients were initially diagnosed with mild (MMSE = 21-26, n = 177) and moderate (MMSE = 10-20, n = 137) dementia. At 18 months, CT+ H patients scored on average 1.76 (P = 0.002) better than CT patients, and at 24 months, patients scored on average 2.52 (P < 0.001) better. At 24 months, the patients with improved cognitive function (△MMSE ≥ 0) in CT + H was more than CT alone (33.33% vs 7.69%, P = 0.020). Interestingly, patients with mild AD received the most robust benefit from CT + H therapy. The deterioration of the cognitive function was largely prevented at 24 months (ΔMMSE = -0.06), a significant improvement from CT alone (ΔMMSE = -2.66, P = 0.005).


The authors concluded that, compared to CT alone, CT + H significantly benefited AD patients. A symptomatic effect of CT + H was more pronounced with time. Cognitive decline was substantially decelerated in patients with moderate severity, while the cognitive function was largely stabilized in patients with mild severity over two years. These results imply that Chinese herbal medicines may provide an alternative and additive treatment for AD.

Conclusions like these render me speechless – well, almost speechless. This was nothing more than a retrospective chart analysis. It is not possible to draw causal conclusions from such data.


Because of a whole host of reasons. Most crucially, the CT+H patients were almost certainly a different and therefore non-comparable population to the CT patients. This flaw is so elementary that I need to ask, who are the reviewers letting such utter nonsense pass, and which journal would publish such rubbish? In fact, I can be used for teaching students why randomisation is essential, if we aim to find out about cause and effect.

Ahhh, it’s the ! I think the funders, editors, reviewers, and authors of this paper should all go and hide in shame.

The title of this post is a statement recently made in an article by Mike Adams in ‘Alternative Medicine News’:

The cancer industry goes to great lengths to deny patients access to any information that they might use to prevent, treat or cure cancer without requiring expensive (and highly toxic) medical interventions. That’s what makes the BMJ documentation of this curcumin cancer cure so astonishing: In years past, the BMJ never would have even tolerated the publishing of such a scientific assessment. So what changed? In truth, the evidence of natural cures for cancer is now so overwhelming that even the BMJ cannot remain in a state of denial without appearing to be hopelessly out of touch with scientific reality.

The story is based on one single patient who apparently was cured of cancer using curcumin (turmeric). The case was also recently (3/1/18) featured on BBC’s ‘YOU AND YOURS’ ( in a similarly uncritical way: no expert was asked to provide an evidence-based assessment and bring some reason into the discussion. Even the DAILY FAIL reported about the story, and predictably, critical assessment had to make way for sensationalism.

So what?

We hear about such nonsense almost every day!

True, but this case is different; it is based on a publication in the highly-respected BMJ (well, actually, it was the ‘BMJ CASE REPORTS’ and not the BMJ, as reported). Here is the article:


A woman aged 57 years was initially diagnosed with monoclonal gammopathy of undetermined significance (MGUS) in 2007 following an incidental finding of M-protein (18 g/L) during investigation for hypertension.

Within 15 months, the patient had rapidly progressed to ISS stage 3 myeloma with M-protein 49 g/L, urinary protein 1.3 g/24-hour, Bence-Jones protein 1.0 g/24-hour, Hb 9.7 g/dL and increasing back pain. She initially declined antimyeloma treatment but 6 months later, following vertebral collapse at T5 and T12, started cyclophosphamide, thalidomide and dexamethasone (CTD) treatment. However, after a week, the patient was admitted with idiosyncratic syndrome including hyponatraemia, a fall in albumin and worsening of blood counts. She received red cell transfusion and her electrolyte abnormalities were carefully corrected.

Although there was evidence of a response to CTD (M-protein 34 g/L), bortezomib and dexamethasone treatment was initiated as an alternative, but this was discontinued after three cycles due to progressive disease (M-protein 49 g/L). The patient was then treated with lenalidomide and dexamethasone with the aim of reducing disease burden prior to high-dose therapy and autologous stem cell transplantation. Treatment was frequently interrupted and dose adjusted to account for neutropenia and despite a minor response after six cycles (starting M-protein 47 g/L, finishing M-protein 34 g/L), in October 2009, she proceeded with stem cell mobilisation. However, neither cyclophosphamide nor plerixafor/GCSF priming were successful. A bone marrow biopsy revealed 50% myeloma cells and a course of CTD was restarted with cautious titration of thalidomide.

The patient achieved a partial response with CTD retreatment over the course of 17 cycles (M-protein 13 g/L) with no further episodes of idiosyncratic syndrome. However, attempts to harvest stem cells in February 2011 and again there months later, both failed. By then, her M-protein had risen to 24 g/L and the patient was too neutropenic to be considered for a clinical trial.

At this point, the patient began a daily regime of oral curcumin complexed with bioperine (to aid absorption), as a single dose of 8 g each evening on an empty stomach. A few months later, she also embarked on a once-weekly course of hyperbaric oxygen therapy (90 min at 2 ATA) which she has maintained ever since. Her paraprotein levels gradually declined to a nadir of 13 g/L, her blood counts steadily improved and there was no evidence of further progressive lytic bone disease.

Outcome and follow-up

The patient continues to take oral curcumin 8 g daily without further antimyeloma treatment. Over the last 60 months, her myeloma has remained stable with minimal fluctuation in paraprotein level, her blood counts lie within the normal range and she has maintained good quality of life throughout this period. Repeat bone imaging in 2014 identified multiple lucencies <1 cm in the right hip and degenerative changes in both hips, but these were attributed to osteoarthritis rather than the myeloma. Recent cytogenetic analysis revealed she had no abnormal cytogenetics by fluorescent in situ hybridisation.


A small but significant number of myeloma patients consume dietary supplements in conjunction with conventional treatment primarily to help cope with the side effects of treatment, manage symptoms and enhance general well-being. Few, if any, use dietary supplementation as an alternative to standard antimyeloma therapy. Here, we describe a case in which curcumin has maintained long-term disease control in a multiply-relapsed myeloma patient. To the best of our knowledge, this is the first report in which curcumin has demonstrated an objective response in progressive disease in the absence of conventional treatment.

Curcumin is a polyphenol derived from the perennial herb Curcuma longa (turmeric) and has, for centuries, been used as a traditional Indian medicine. Several reports published over the two decades have claimed various health benefits of curcumin and this has led to its increasing popularity as a dietary supplement to prevent or treat a number of different diseases.

The biological activity of curcumin is indeed remarkable. It is a highly pleiotropic molecule which possesses natural antioxidant, anti-inflammatory, antiseptic and analgesic properties. More recently, it has demonstrated antiproliferative effects in a wide variety of tumour cells including myeloma cells and exerts its antiproliferative effects through multiple cellular targets that regulate cell growth and survival.

In vitro, curcumin prevents myeloma cell proliferation through inhibition of IL-6-induced STAT-3 phosphorylation and through modulation of the expression of NF-kB-associated proteins such as IkB〈,Bcl-2, Bcl-xL, cyclin D1 and IL-6 and apoptosis-related molecules including p53 and Bax. In other studies, curcumin was shown to circumvent resistance to dexamethasone, doxorubicin and melphalan as well as potentiate the effects of bortezomib, thalidomide and lenalidomide. Furthermore, curcumin-induced cell death was not influenced by myeloma molecular heterogeneity.

The antimyeloma effects of curcumin in the clinical setting however are less clear. Only one phase I/II study has evaluated curcumin treatment in myeloma patients. These patients were either asymptomatic, relapsed or had plateau phase disease. Treatment with curcumin downregulated the expression of NFkB, COX-2 and STAT3 in peripheral blood mononuclear cells, but no objective responses were observed in any subgroup of patients. This may be as a result of small sample size in this study, follow-up was limited to 3 months and clinical responses may have been observed with longer follow-up. However, downregulation of NFkB, COX-2 and STAT3 expression may not correlate with the clinical activity of curcumin and there may be further mechanisms of action that remain unclear, possibly through the modulation of another target. We would not be able to identify any patient-specific mechanisms of activity in this case study, as the patient has been taking curcumin for some time now and baseline bone marrow or peripheral blood samples are not available. However, in the setting of a clinical trial, it may be possible to use next-generation sequencing to help identify a mutation that may be a potential target for curcumin.

Another study examined its effects in preventing the progression of MGUS and smouldering myeloma to myeloma. The results showed that curcumin exerted a trace of biological activity with modest decreases in free light chain and paraprotein levels and a reduction in a marker of bone resorption with curcumin treatment, suggesting the therapeutic potential of curcumin in MGUS and smouldering myeloma. However, more studies are needed to address this further.

Whether such effects are observed in patients with active disease remains to be seen. The fact that our patient, who had advanced stage disease and was effectively salvaged while exclusively on curcumin, suggests a potential antimyeloma effect of curcumin. She continues to take daily curcumin and remains in a very satisfactory condition with good quality of life. This case provides further evidence of the potential benefit for curcumin in myeloma. We would recommend further evaluation of curcumin in myeloma patients in the context of a clinical trial.


What should we make of this?

I think that much of the reporting around the story was grossly irresponsible. It is simply not possible to conclude that curcumin was the cause of the remission. It could be due to a whole host of other factors. And a case report is just an anecdote; it never can prove anything and can only be used to stimulate further research.

I fully agree with the authors of the case report: curcumin seems worthy of further investigation. But recommending it to patients for self-medication is vastly premature and quite simply dangerous, unethical and naïve bordering on stupid.

And, of course, the above-cited drivel of Mike Adams is just beyond the pale – the evidence for ‘alternative cancer cures‘ is very, very far from ‘overwhelming’; and the ‘cancer industry’ is doing what they can to determine whether turmeric or any other natural remedy can be used to treat cancer and other diseases.

If they are ever successful, the Adams of this world will shout ‘EXPLOITATION!!!’

If their endeavours are not successful, they will complain ‘CONSPIRACY!!!’

The question whether chiropractic is a truly valuable option for people suffering from back pain has been addressed repeatedly on this blog. My answer was usually negative, but proponents of chiropractic tended to argue that I am biased. Therefore I find it constructive to see what an organisation that hardly can be accused of bias says on this topic. An article by ‘SHOW ME THE EVIDENCE’ has recently provided a comprehensive overview of treatments for back pain. This is what they wrote about chiropractic:


Spinal manipulation, the cranking and tweaking on offer when you visit a traditional chiropractor, is among the most popular approaches to back pain. Practitioners lay their hands on the patient and move their joints to or beyond their range of motion — a technique that’s often accompanied by a pop or crack.

There is some evidence the approach can help people with chronic back pain — but not any more than over-the-counter painkillers or exercise, and you need to take precautions when seeking out a chiropractor.

First, a quick look at the evidence. There are two recent Cochrane reviews on spinal manipulation for low back pain: one focused on people with acute (again, episodic/short duration) pain and the other on chronic pain. The 2011 review on chronic low back pain found that spinal manipulation had small, short-term effects on reducing pain and improving the patient’s functional status — but this effect was about the same as other common therapies for chronic low back pain, such as exercise. That review was published in 2011; UpToDate reviewed the randomized trials that have come out since — and also found that spinal manipulation delivered modest, short-term benefits for chronic back pain sufferers.

The Cochrane review on acute pain found that spinal manipulation worked no better than placebo. So people with a short episode of back pain should probably not bother seeing a chiropractor.

“Based on the evidence,” University of Amsterdam assistant professor Sidney Rubinstein, who is the lead author on the Cochrane reviews, told me, “it would appear [spinal manipulation] works as well as other accepted conservative therapies for chronic low back pain, such as non-prescription medication or exercise, but less well for patients with acute low back pain.”

As a chiropractor himself, he had some advice for patients: They should avoid chiropractors who routinely make X-rays or do advanced diagnostics for low back pain because this adds nothing to the clinical picture, particularly in the case of nonspecific low back pain. Patients should also beware chiropractors who put them on extended programs of care.

“Patients who respond to chiropractic care traditionally respond rather quickly,” he said. “My advice is those patients who have not responded to a short course of chiropractic care or manipulation should consider another type of therapy.”

While the risks of serious side effects from spinal manipulation for back pain are rare — about one in 10 million — the risks associated with chiropractic therapy for neck pain tend to be slightly higher: 1.46 strokes for every million neck adjustments.

The issue is the vertebral artery, which travels from the neck down through the vertebrae. Manipulating the neck can put patients at a higher risk of arterial problems, including stroke or vertebral artery dissection, or the tearing of the vertebral artery (though Rubinstein noted that people in the initial stages of stroke or dissection may also seek out care for their symptoms, such as neck pain, which makes it difficult to untangle how many of health emergencies are brought on by the adjustments).


This all seems fairly reasonable to me – except for the following not entirely unimportant points:

  • I am not sure where the evidence about risks of spinal manipulation comes from. In my view, it is not entirely correct: as there is no effective post-marketing surveillance, we cannot possibly name the incidence figures.
  • Neck manipulations are clearly more risky than manipulations lower down. But this does not necessarily mean that back patients are safer than those with neck pain. Chiropractors view the spine as a whole organ and will regularly manipulate the neck (if they sense ‘subluxations’ in this area), even if the patient comes with low back pain.
  • There are also indirect risks with consulting a chiropractor; for instance, they often give incompetent advice about healthcare. This can include discouraging immunisations or treating serious diseases, such as asthma, colic etc., with chiropractic.
  • I think the article should point out that exercise is not just as effective (or as ineffective) as chiropractic, but it is much safer and less expensive.
  • What Rubinstein says about responders is debatable, in my view. In particular, most chiropractors will convince their patients to continue treatment, even if they do not ‘respond’. And ‘responding’ might be simply the natural history of the condition and therefore totally unrelated to the therapy.

The bottom line: Chiropractic is not the best treatment for back pain!

1 2 3 16
Recent Comments

Note that comments can be edited for up to five minutes after they are first submitted.

Click here for a comprehensive list of recent comments.