Rapid Responses to:

NEWS: Susan Mayor Meta-analysis shows difference between antidepressants and placebo is only significant in severe depression BMJ 2008; 336: 466 [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Will Depression study serve to act as new MMR study scare?

DAVID G SAMUEL   (27 February 2008)

A Call For Regulation of Media Reporting

Justin Christian Marley   (27 February 2008)

The anti-depressant debate - are we missing the point and are health professionals part of the problem?

Anthony ( Tony) P Steele-Perkins   (27 February 2008)

Depression and cerebral blood flow.

Les O. Simpson   (28 February 2008)

Antidepressant Benefits: Misinferance from ordinal scales?

Peter Tennant, Alan Tennant   (1 March 2008)

Probably biased

Andrew Montgomery   (1 March 2008)

Response Strategy Required

Sian J.A. Harris   (2 March 2008)

ERRATUM - Antidepressant Benefits: Misinferance from ordinal scales?

Peter Tennant, Alan Tennant   (2 March 2008)

quick

benjamin dean   (3 March 2008)

Lets celebrate the Placebo Effect

Alastair E Dobbin   (3 March 2008)

Low dose anti-depressants in non-major depression

jyotirmoy ghosh   (5 March 2008)

Is it time to come back to Hippocrates’ principles in the treatment of depression?

Gianluca Castelnuovo, Elena Faccio, Gianpiero Turchi, Alessandro Salvini, Enrico Molinari   (7 March 2008)

Sauce for the goose

R N Stenning   (8 March 2008)

A boost for placebos?

John Merrick   (11 March 2008)

Lies, damned lies, statistics and the BMJ

Zekria Ibrahimi   (2 April 2008)

Effectiveness of Anti-depressants

Dr Arun Viswanath   (7 April 2008)

Will Depression study serve to act as new MMR study scare? 27 February 2008
DAVID G SAMUEL, FINAL YEAR MEDICAL STUDENT CF48 2AS Send response to journal: Re: Will Depression study serve to act as new MMR study scare?	As a final year medical student I read with interest the study published in this weeks PLoS Medicine regarding the efficacy of Anti-depressants – specifically SSRI’s such as “Prozac”. While I agree with their findings that these drugs may not always be effective in treating every case of depression, especially milder forms of the illness, I fear the study will cause a rebound reaction amongst the public and lead to many patients discontinuing their medication without tailoring their doses gradually. This may exacerbate their depressive illness and make their overall health status even worse. I also feel the media hysteria which has followed its publication may serve to act a similar catalyst sparking public fear and outcry to the panic that followed the infamous Wakefield MMR study. I also feel that it undermines the positive effect that Anti-depressants do undoubtedly have on many people’s lives, especially those with more severe depressive illnesses who are suffering biological symptoms such as poor sleep patterns. Most GP’s would argue that SSRI’s do have their place in treating depression and they are effective. They would also undoubtedly support increasing the availability of other treatment modalities which as CBT and other psychological therapies. The government may wish to take this wake up call that further funding is required to increase and improve the availability and breadth of these treatment services available to patients. However, I feel the study has highlighted an important issue that is often overlooked amongst the medical profession. Studies disproving the efficacy of treatments are rarely published as they do not have the “sell” factor. In addition, the majority of successful trials are often funded by large pharmaceutical companies who naturally have a vested interest in seeing positive results. Maybe now is the time for the government to work with major drug companies and create an independent study group which look at the efficacy of drugs. NICE and other organisations should also use this study as a warning that their guidelines and recommendations may not always be based on sound evidence based medicine and studies. I only hope this study is viewed in context, is not used as a tool to scare the general public and does not lead to patients doubting the treatments given to them in good faith by doctors, acting on their best judgement and experience. Anti-depressants undoubtedly are only one component of a multi-disciplinary approach to treatment the major chronic illness and health burden that is Depression but surely years of experience and success cannot be undone by the results of this single trial? David Gwynfor Samuel 5th year medical student - Cardiff University Merthyr Tydfil. CF48 2AS welshsledge@hotmail.com Competing interests: None declared
A Call For Regulation of Media Reporting 27 February 2008
Justin Christian Marley, Specialist Registrar in Psychiatry Mental Health Unit, Rotherham District General Hospital, Moorgate Road, Rotherham, S60 2UD Send response to journal: Re: A Call For Regulation of Media Reporting	Dear Editor, The article purports to show that antidepressants are little better than placebo for the treatment of depression. I have read the media's response to this article with significant concern as in some articles there has been little or no analysis and the paper has been used to criticise both the diagnosis of depression and the efficacy of antidepressants. A closer examination of the paper reveals a number of difficulties with the authors' conclusions. Firstly the studies considered are those for marketing approval and are of a very short duration, indeed barely long enough to produce the expected benefit in depression. Secondly the authors have omitted a large number of other antidepressants from the study. Thirdly, for the selected antidepressants, a number of studies during the period in question, have not been included. Fourthly subsequent studies for the selected antidepressants, some of which are of a longer duration, have been neglected. Finally, and most importantly, the results of this analysis of a small number of antidepressants has been generalised to the entire class. Given the significant morbidity and mortality associated with depression such reporting in the media, is in my opinion irresponsible and may put lives at risk. The time has come to call for regulation of media reporting of sensitive health related issues. Competing interests: The author has received small gifts from pharmaceutical companies.
The anti-depressant debate - are we missing the point and are health professionals part of the problem? 27 February 2008
Anthony ( Tony) P Steele-Perkins, Clinical lead/ HoD Somerset Occupational Health Taunton & somerset NHS FT, TA1 4AS Send response to journal: Re: The anti-depressant debate - are we missing the point and are health professionals part of the problem?	No doubt there will be flurry of correspondence over the efficacy of SSRIs,but could I comment on the wider issue as a long time consultant occupational physician whose major practice is advising those functionally disabled through a loss of mental well-being, and enabling hopefully a return to well-being and work. First, most of my practice in the NHS and public sector/teaching work arena concerns mainly short term coping reactions to life stress events. These can be profoundly disabling in terms of day to day functioning, especially with a mix of personality and "I must and will cope at all costs" despite increasing and widening internal conflicts. They may indeed fulfil the criteria of a mixed mild or even moderate anxiety or depressive episode ( ICD10). My major aim is answer the unspoken fears of "what's happening to me" and am I going mad" by explaining in simple terms how the mind works,using basic psychology, and to re-assure and normalise by understanding physiological/arousal, cognitive and behavioural symptoms. This in itself is often sufficient to start a recovery, together with advice on some basic CBT principles. Obviously standard history taking should re-assure me of missing more major psychiatric illness and risks, and recovery often involves liaison with GPs, specialists, and other health stakeholders. The important point is that most of these cases are adjustment disorders/reactions as defined in ICD10, with accompanying features of anxiety and depression - which is not surprising given the way these are defined. It is concerning that we health professionals may be falsely labelling these folk with the stigma of depression, and all that goes with this diagnosis. Is this surprising? NO in my opinion, as most standard screening questionnaires used in the primary care setting to effect a "diagnosis" as the HADS, GHQ, Beck etc, will come up with a depressive answer. Where is the questionnaire to screen against Adjustment Reaction? Surely, significant mental illness apart, we should be assessing against loss of functionality, with the outcome aim to restore functionality, and with it emotional resilience and health, work, and well -being? Could I point readers to the Marks "Work and Social Adjustment Scale (WSAS)" and suggest that this is a more appropriate model for use in the primary care/ occupational health setting? Perhaps I should add my own opinion on "anti-depressants". To engage in behavioural therapy, be it group, one to one, or via the latest and promising IT based self help CBT packages, the individual must have a modicum of cognitive performance especially concentration, and the ability to engage and take on board the processes. I have recommended these, only to be told at review "I couldn't take it in doctor". To achieve this in a reasonable timescale noting the employment clock is ticking, and the bio-psycho-social implications, I do believe, based on clinical observation of many cases, there is a place for psychotropic medication, especially starting in a very low dose to minimise side effects. However I also consider that the chemical medical model cannot work on its own, which brings me round to where I started. Competing interests: None declared
Depression and cerebral blood flow. 28 February 2008
Les O. Simpson, retired experimental pathologist Dunedin, New Zealand, 9077 Send response to journal: Re: Depression and cerebral blood flow.	Maybe a benefit will arise from the current debate about the usefulness of antidepressants in depression, if it draws attention to the published studies concerning reduced regional cerebral blood flow in depression which have failed to influence clinical opinion. Perhaps the relevance of such information can be appreciated by reference to the study by Lucey et al. (1) They showed that in patients with panic disorder with agoraphobia, obsessive compulsive disorder and post traumatic stress disorder, there were regional reductions in cerebral blood flow, but in each disorder in different regions of the brain. It is proposed that depression is the dysfunctional state arising from inadequate rates of blood flow to deliver the necessary oxygen and nutrient substrates to sustain normal tissue function in specific regions of the brain. This means that the condition would be reversible when there were adequate rates of blood flow. In 1990, Sackheim et al (2) critically examined what had been written about regional cerebral blood flow in mood disorders. An important contribution was made by Bench et al (3) who reported their findings in a study in which previously scanned patients were rescanned on remission. They concluded, "Thus, recovery from depression is associated with increases in regional cerebral blood flow in the same area in which focal decreases in regional cerebral blood flow are described in the depressed state, in comparison with normal subjects." Similar findings were reported in another paper (4) which reported that the reduced rate of blood flow in the left frontal region which had been observed during depression, returned to normal during remission. The lack of attention given to such findings probably reflects the current antipathy to reports which imply a role for the flow properties of blood (blood rheology). It is possible to interpret such changes in blood flow in terms of the effects of poorly deformable red cells, which reflect change in their environment by a reduction in fluidity of the the cell membrane. Normalisation of the cell environment restores normal levels of deformability. So it is not surprising that depression is a frequent problem in chronic disorders which are known to have altered blood rheology manifested as increased blood viscosity and poorly deformable red cells, such as in diabetes for example. Kamada et al (5) in 1986 reported that sardine oil so increased the fluidity of the membranes of diabetic red cells that they were unable to distinguish such cells from those of non-diabetics. Ten years later Maes et al (6) noted that, " Major depressed subjects had significantly lower C18-3 omega-3 in cholesteryl esters than normal controls. Major depressed subjects showed significantly lower total omega-3 polyunsaturated fatty acids... than minor depressed subjects and healthy controls." A later study of the omega-3 fatty acid content in the diet and in red cell membranes of depressed patients (7) noted that, "Lower red blood cell membrane n-3 polyunsaturated fatty acids are associated with the severity of depression," and concluded, "The findings raise the possibility that depressive symptoms may be relieved by n-3 polyunsaturated fatty acid supplementation." Ten years later there is no indication that the significance of such findings have been recognised and therapy for depression has been based upon antidepressants rather that omega-3 rich fish oil which might correct the primary problem. Possibly, if the current debate leads to a more public recognition of the problems of cerebral blood flow in depression, then maybe those who suffer from depression will explore the potential benefits of taking 2 x 1000mg capsules of fish oil, three times daily. Many studies have used 10 capsules daily, and one study reported that the maximal tolerable dose was twenty grams daily. Because of the need for the enzyme delta-6-desaturase to be functional in order to utilise the plant derived alphalinolenic acid, it is safer to use fish oil as a source of omega-3 fatty acids. An alternative would be to increase the dietary intake of omega-3 fatty acids by including sardines or oily fish in meals on a daily basis. In addition, because regular light exercise has been shown to reduce blood viscosity, an activity such as walking or dancing should be part of the daily programme. A good example was an Australian study involving "pram pushing" which was shown to be beneficial for women with post-partum depression. References. 1. Lucey JV, Costa DC, Adshead G, et al. Brain blood flow in anxiety disorders: OCD, panic disorder with agoraphobia and post traumatic stress disorder. Br J Psychiatry 1997;171:346-50. 2. Sackheim HA, Prohovnik I, Moeller JR, et al. Regional cerebral blood flow in mood disorders. Arch Gen Psychiatry 1990;47:60-70. 3. Bench CJ, Frackowiak RS, Dolan RJ. Changes in regional cerebral blood flow on recovery from depression. Psychol Med 1995;25:247-61. 4. Navarro V, Gasto C, Lomena F, et al. Normalisation of frontal cerebral perfusion in remitted elderly major depression: a 12-month follow-up SPECT study. Neuroimage 2002;16:781-7. 5. Kamada T, Yamashita T, Baba Y, et al. Dietary sardine oil increases erythrocyte membrane fluidity in diabetic patients. Diabetes 1986;35:604- 11. 6. Maes M, Smith R, Christophe A, et al. Fatty acid composition in major depression: decreased omega-3 fractions in cholesteryl esters and increased C20:4 omega-6/C20:5 omega-3 ratio in cholesteryl esters and phospholipids. J Affect Disord 1996;38:35-46. 7. Edwards R, Peet M, Shay J, et al. Omega-3 polyunsaturated fatty acid levels in the diet and in red blood cell membranes of depressed patients. J Affect Disord 1998;48:149-55. Competing interests: None declared
Antidepressant Benefits: Misinferance from ordinal scales? 1 March 2008
Peter Tennant, Biostatistician School of Clinical Medical Sciences, University of Newcastle upon Tyne, Newcastle upon Tyne. NE1 4LP, Alan Tennant Send response to journal: Re: Antidepressant Benefits: Misinferance from ordinal scales?	It is very rare for the results of medical research to make national headlines and potentially grave if these studies are later shown to be flawed.[1] Thus, if a study is awarded a particularly high profile, for whatever reason, the scientific community (and the media) should make extra effort to ensure that the research is on the soundest scientific footing. While the recent meta-analysis by Kirsch et al.[2] should be praised for reducing the influence of publication bias,[3] it is curious that little has been made of the apparent disregard of one of the fundamental rules associated with the statistical procedure used, that is regression analysis. Most statistical procedures have rules, or assumptions, that govern their proper use. Considerable literature has been developed to help scientists to test these assumptions, and choose the correct procedure for a particular situation. Regression and, thus, meta-regression, both have a clear list of assumptions that underpin their use.[4] One of the most fundamental of these is the assumption of interval scaled data. For those not familiar with this term, an interval scale is what we would commonly associate with continuous variables such as height or weight. The crucial characteristic is that the distance between any two consecutive points is the same wherever you are on the scale (i.e. the difference between 99cm and 100cm is the same as the difference between 149cm and 150 cm = 1cm). This concept, known as equal interval scaling, is a requirement for many mathematical operations such as the calculation of a change score. Of course, not all outcomes are as simple to measure as height and weight. Many concepts in medicine are not directly measurable. Depression, for example, is typically ascertained by a clinical diagnostic interview, or by adding the answers to a series of questionnaire responses (items) such as those that appear in the Hamilton Depression Rating Scale (HRDS) as reported in the meta-regression. Such scales form what is known as ordinal scaled data where it possible to say, for example, that one patient has more depression than another, or that a patient has improved after treatment. However, any further inferences, such as the size of the change, are inappropriate because, unlike for an interval scale, the distance between any two consecutive points on an ordinal scale cannot be said to be the same as the difference between two consecutive points elsewhere on the scale. Indeed, rather than adopting a linear relationship, ordinal scales are sigmoidal with a single point change at the extremes representing a larger change in the underlying trait (in this case depression) than a single point change in the centre.[5] Given this observation, the NICE definition of clinical significance as ‘at least 3 points’ on the HRSD [6] is essentially a non-uniform condition that is easier to achieve among centrally-scoring patients than those with extreme scores (such as the minimally or severely depressed). Today there is a considerable theoretical and empirical literature that demonstrates the limitations of ordinal data, as well as some relatively simple ways to overcome the problem found in articles referring to Item Response Theory (IRT) or Rasch analysis.[7] Among other things, these applications have revealed that the misuse of ordinal scaled data can produce erroneous data and drive inaccurate conclusions.[8,9] Consequently, concerns must be raised over the accuracy of the results of the meta-regression performed by Kirsch et al, given they have undertaken sophisticated mathematical operations on data which do not support such activities. Moreover, it is worth noting that even the calculation of a mean, a standard deviation, and a change score are invalid on ordinal data, given that these all assume equal interval scaling. Rather than perform a meta-regression, we feel it would have been more appropriate to have performed a traditional meta-analysis by comparing, for example, the relative odds of moving to a lower depressive state (an approach that would have utilised the value of the ordinal data, rather than ignored the limitations). Although this may lead to the same conclusion, it is possible that it may not; either way the uncertainty makes it very difficult to have faith in the current analysis when it is taken in its existing form. References: 1 Elliman D, Bedford H. MMR: where are we now? Arch Dis Child 2007;92:1055-1057 2 Kirsch I, Deacon BJ, Huedo-Medina TB et al. Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration. PLoS Medicine 2008;5;e45 3 Mayor S. Meta-analysis shows difference between antidepressants and placebo is only significant in severe depression. BMJ 2008;336:466 4 Field A. Discovering Statistics Using SPSS (Second Edition). London: Sage Publications 2005 5 Wright BD, Linacre JM. Observations are always ordinal; measurements, however, must be interval. Arch Phys Med Rehabil 1989;70:857 -860 6 National Institute for Clinical Excellence Depression: management of depression in primary and secondary care. London: NICE 2004 7 Pallant JF, Tennant A. An introduction to the Rasch measurement model: An example using the Hospital Anxiety and Depression Scale. Br J Clin Psychology 2007;46:1-18 8 Merbitz C,Morris J,Grip JC. Ordinal scales and foundations of misinference. Arch Phys Med Rehabil 1989; 70:308-12 9 Tennant A. Goal attainment scaling: Current methodological challenges. Disab & Rehab 2007;29:1583-1588 Competing interests: None declared
Probably biased 1 March 2008
Andrew Montgomery, locum Auckland Send response to journal: Re: Probably biased	Unfortunately this paper became headline news in the New Zealand's most widely read newspaper. Fortunately this paper mitigated the harm caused by publishing perhaps 100 rapid responses (personal accounts) from people who had been placed on SSRIs for depressive illness. These responses are readily accessible online and should be read by all those who believe that SSRIs are little better than placebo even in more severe depression. These accounts tell the story of many people who have found these drugs to be life saving. Many are based upon several year histories of battling with depression. I have not read the paper by Hull and nor do I intend to. I have treated numerous people with depression successfully with both medication and "counselling". Depression as with many "diseases" is partly inherited. Despite the fact that many physical features (eg good looks, height and body weight) are unequivocally substantially inherited (genetically determined) many people - unfortunately including many doctors - seem to believe that pyschological traits are not similarly genetically prescribed. Competing interests: None declared
Response Strategy Required 2 March 2008
Sian J.A. Harris, GP Heron Practice,John Scott Health Centre,Woodberry Down,Green Lanes,London. N4 2NU Send response to journal: Re: Response Strategy Required	Dear Editor, This meta-analysis has serious, probably fatal, methodological flaws and draws incorrect inferences, as shown by the substantial peer review response (see BMJ rapid responses and PLoS website). Yet, despite this, the message that will be retained by the public and unfortunately many doctors will be that “Prozac is no better than sugar pills”. One only has to have seen the front page of the review section of Thursday’s Guardian newspaper – “What was all that about ? The creation of the Prozac myth ”- to see the potential for significant harm. The aftermath - the measured response from the scientific community –does not make the headlines of course and again the damage will have been done, as clearly seen with the MMR debacle. How many patients will now not come forward or will decline medication and how many GPs will not prescribe where major benefit may have been achieved (e.g. in dysthymic patients) because of this failure? Should we not be developing robust strategies to quickly counter such misrepresentation in the future? Dr Sian Harris GP The Heron Practice John Scott Health centre Woodberry Down Estate London N4 2NU Sian.harris@nhs.net Competing interests: None declared
ERRATUM - Antidepressant Benefits: Misinferance from ordinal scales? 2 March 2008
Peter Tennant, Biostatistician School of Clinical Medical Sciences, University of Newcastle upon Tyne, Newcastle upon Tyne. NE1 4LP, Alan Tennant Send response to journal: Re: ERRATUM - Antidepressant Benefits: Misinferance from ordinal scales?	Where the original response states: "Rather than adopting a linear relationship, ordinal scales are sigmoidal" It should have said, "Rather than adopting a linear relationship, ordinal scales are usually sigmoidal" We apologise for any confusion that this may have caused. Competing interests: None declared
quick 3 March 2008
benjamin dean, sho oxford Send response to journal: Re: quick	I just wanted to point out that the use anecdotal cases of patients who claim that they have benefited from SSRIs is not a convincing comeback to this detailed review. The study shows SSRIs to be as good as placebo, thus they will help some patients! This is not a point in question. There are numerous medical treatments that some patients will insist help them, even though there is no decent scientific evidence behind their claims, this doesn't mean these treatments work. Also the patients who have not been helped by SSRIs won't be writing in to express this sentiment! Competing interests: None declared
Lets celebrate the Placebo Effect 3 March 2008
Alastair E Dobbin, General Practitioner Brunton Place Surgery, 9 Brunton Place EH5 3RN Send response to journal: Re: Lets celebrate the Placebo Effect	I read with dismay the responses to this article. It seems that none of the writers have studied the broad sweep of Irving Kirsch's work, and I would suggest close scrutiny of his earlier papers (Refs 1 and 3). In his earlier papers he examined: (a) 39 studies of depressed patients from 1974 to 1995 (1). The studies included patients with a primary diagnosis of depression, were randomized, and controlled for patients who received no treatment. All the studies compared drugs with placebo. The results of this I suggest your readers study for themselves. Interestingly treatment effects for SSRIs are no more than they were for drugs such as thyroxine, barbiturates and benzodiazepines, all of which have been used in the past to treat depression (yes, hard to believe). And guess what is the best predictor for the effect of any antidepressant drug of any class? Placebo effect, the greater the placebo effect, the greater the drug effect; 'As a proportion of the drug response, the placebo response was constant across different types of medication (75%), and the correlation between placebo effect and drug effect was .90. These data indicate that virtually all of the variation in drug effect size was due to the placebo characteristics of the studies.' See a pattern emerging? And what is the difference between all these active drugs and placebo? Amazingly, side effects! And what happens if you give your placebo side effects? I leave the Cochran collaboration (run by statisticians of the highest calibre with no outside influences and completely free of any drug company sponsorship) to say this 'The more conservative estimates from the present analysis found that differences between antidepressants and active placebos were small.'(2) (b). The second paper by Kirsch and colleagues(3) is on the 10 year FDA data from the USA on the commonest anti-depressants, fluoxetine, paroxetine sertraline, nefazadone, venlafaxine and citalopram. All the studies published and unpublished on these drugs have to be released under FDA regulations after 10 years, which happened in 2002. Kirsch et al looked at all the data from these studies, all of which used the Hamilton Depression Index. They found: duplication of drug by placebo 82%, mean Hamilton Depression Index difference 1.80 points (drug/placebo). For NICE to approve a drug in this class today they expect a treatment effect difference of 3 points, so if these drugs were released today they would not be approved by NICE. Pretty staggering really. And yet to read these responses, most of which look like their authors have not studied the breadth of Kirsch's work, you would think it was the best thing that the studies showing no effect of the treatments were suppressed and the others showing positive effects published (sometimes repeatedly). None of your correspondents have done more than pay lip service to what appears to me as essentially a suppression of the truth by the drug companies in order to increase sales. We are not going to move forward now if we are constantly bemoaning and denying the placebo effect. We cannot expect the public to trust us if we do not recognise the real problem here, the power of the pharmaceutical companies. We need a paradigm shift in our thinking. Now is the time to celebrate the placebo effect and perhaps even harness it, there are ways to do it, what clinicians need is a broader understanding of suggestion and positive visualisation and the incredible healing powers of the mind and spirit. 1. Kirsch I, Sapiristein G (1998) Listening to Prozac but hearing placebo: A meta-analysis of published clinical trials Prevention & Treatment Jun Vol 1(2) 2 Moncrieff J, Wessely S, Hardy R. Active placebos versus antidepressants for depression. Cochrane Database of Systematic Reviews 2001, Issue 2. The full text of the review is available in The Cochrane Library (ISSN 1464-780X). 3.Kirsch, I., Moore, T.J., Scoboria, A., et al (2002) The Emperor's new drugs: An analysis of antidepressant medication data submitted to the US Food and Drug Administration. Prevention & Treatment, 5, Article 23. http://www.journals.apa.org/prevention/volume5/pre0050023a.html Competing interests: Alastair dobbin is a researcher in mental health in primary care Irving Kirsch is a co-applicant for a trial of Positive Mental Training, a 3rd wave CBT intervention for self help based on self hypnosis developed by Alastair Dobbin. Alastair Dobbin has a commercial interest in this intervention (company director, shareholder) and is not an applicant or investigator on the trial Irving Kirsch has no commercial interest in the intervention.
Low dose anti-depressants in non-major depression 5 March 2008
jyotirmoy ghosh, Private practice Durgapur (w.bengal), India: 0343 2536067 Send response to journal: Re: Low dose anti-depressants in non-major depression	Instead of presenting mental problems, Indian patients suffering from anxiety and endogenous depression present with overlapping physical symptoms of functional diseases e.g, non-ulcer dyspepsia, irritable bowel syndrome, gastro-oesophageal reflux disease, migraine etc. However they readily admit to mental problems when asked. Study of a sample of 100 patients showed: (For nil symptom score was 0, off & on- 1, most days-2 and everyday-3; maximum possible total score for each of the symptoms was 300 for 100 patients.). Mental symptoms scores- Depressed 227, Tense -230, Sleep disturbance-203 Follow-up of similar cases treated with common anti-depressants showed the following:- 1. Amitryptiline- dosage- 25 to 50 mgm/day. No. of pts- 83, av. age-35yrs. with systemic diseases- 44 (53%); av. duration of symptoms- 45.8 months; av.duration of intake of drug- 4.7months Result (pts’ own assessment)- Good- 45.9%, satisfactory- 38.5%, no response - 15.7%; adverse effects- 6%. 2. Dothiepen- dosage- 50 to 75 mgm/day. No. of pts.- 51; Av. Age-46.6yrs. with systemic diseases- 31 (61%); av. duration of symptoms-66.3 months; av. duration of intake of drug- 4.9 months Results (pts’own assessment)- Good- 57%, Satisfactory- 31%; no response- 12%. Adv effects – 4%. 3. Sertraline- dosage- 50mgm/day. No. of pts.- 20; av. age 51.6yrs; with systemic diseases- 15(75%); av. duration of symptoms- 11.3m; av. duration of intake of drug- 7.55m. Result (pts’ own assessment)- Good- 30%, satisfactory- 60%, no response- 10%; adv. Effects- nil. Conclusion: Indian patients with anxiety and depression do not complain of mental problems when they come to doctor. Instead they present with overlapping symptoms of NUD, IBS, GORD, Migrain-tension headache, non -articular rheumatism etc. All these functional diseases as well as their mental symptoms respond significantly well when they are treated long term with low dose anti-depressants with little side-effects. Competing interests: None declared
Is it time to come back to Hippocrates’ principles in the treatment of depression? 7 March 2008
Gianluca Castelnuovo, PhD, PsyD, Professor of Psychology and Researcher in Clinical Psychology Istituto Auxologico Italiano IRCCS, Psychology Research Laboratory, 28824, Verbania, Elena Faccio, Gianpiero Turchi, Alessandro Salvini, Enrico Molinari Send response to journal: Re: Is it time to come back to Hippocrates’ principles in the treatment of depression?	The results of the recent meta-analysis of clinical trials about antidepressants (Kirsch et al., 2008; Mayor, 2008) add to the never ending debate in the scientific and clinical community about the choice of the best treatment for depression. The importance of recognizing many factors (psychological variables, family and social context, living habits, etc…) in the etiology and treatment of many disorders could be historically represented in one of the best of Hippocrates’ pronouncement, translated in Latin as “Similia similibus curantur” or, in English, “Like is cured by like.” These Hippocrates’ considerations have been “forgotten” by part of modern medicine, especially in the mental health field. Coming back to Hippocrates’ suggestions, we have to consider the pathology as the product of different factors and the relative treatments have to be built using the “Like is cured by like” principle. Remaining in the mental health field, if a problem is generated at a cultural, social, interpersonal, family, etc… level, perhaps the best treatment is to provide care at the same level. In fact there is a lot of possibilities to treat a psychological or psychopathological problem: working with the single patient, the couple, the family, the social context, the school, etc… but one of the leading trends of official medicine is an individual and biological (pharmacological) response. If a problem is generated at biological, molecular, neurological level (such in a severe depression), perhaps the best choice for treatment is to consider this level, but if a person shows the problem only in some particular and selected situations and contexts (such as at work but not in the family, alone but not with people, etc…), is the bio-pharmacological approach the best treatment that respects the “Similia similibus curantur” principle? Is the bio-pharmacological reductionism (Ahn, Tewari, Poon, & Phillips, 2006; Longino, 1998) the best approach to solve psychopathological problems, such as in mild depression? Moreover it is important to underline that bio-psychopharmacological reductionism could increase some negative effects (Hammad, Laughren, & Racoosin, 2006): in a recent study about children and adolescents, the risk of suicide attempts was 1.52 times higher after antidepressant drug treatment compared with no antidepressant drug treatment (Olfson, Marcus, & Shaffer, 2006). But the most worrying aspect of the use of drugs in mental health field is that from a longer-term perspective, the brains of teenagers are still developing, and the effects of drug use and abuse may be harmful in ways that are not yet understood (Friedman, 2006). Many criticisms about using and following to use pharmacological treatments have been expressed by the same psychiatrists (Moncrieff, 2006; Moncrieff & Cohen, 2006). For example antidepressants create, and not cure, abnormal brain states: “antidepressants are assumed to work on the specific neurobiology of depressive disorders according to a “disease-centered” model of drug action. However, little evidence supports this idea. An alternative, “drug-centered,” model suggests that psychotropic drugs create abnormal states that may coincidentally relieve symptoms. Drug induced effects of antidepressants vary widely according to their chemical class—from sedation and cognitive impairment to mild stimulation and occasionally frank agitation. Results of clinical trials may be explained by drug-induced effects and placebo amplification. No evidence shows that antidepressants or any other drugs produce long-term elevation of mood or other effects that are particularly useful in treating depression” (p. 961, Moncrieff & Cohen, 2006). Moreover the research on maintenance drug treatment is flawed: “If withdrawal induced adverse effects could be effectively managed, the success of drug discontinuation might be much greater than usually assumed and might outweigh the disadvantages of continued treatment” (p. 517, Moncrieff, 2006). Is it time to come back to Hippocrates’ principles in the treatment of depression? References Ahn, A. C., Tewari, M., Poon, C. S., & Phillips, R. S. (2006). The limits of reductionism in medicine: could systems biology offer an alternative? PLoS Med, 3(6), e208. Friedman, R. A. (2006). The changing face of teenage drug abuse-¬the trend toward prescription drugs. N Engl J Med, 354(14), 1448-1450. Hammad, T. A., Laughren, T., & Racoosin, J. (2006). Suicidality in pediatric patients treated with antidepressant drugs. Arch Gen Psychiatry, 63(3), 332-339. Kirsch, I., Deacon, B. J., Huedo-Medina, T. B., Scoboria, A., Moore, T. J., & Johnson, B. T. (2008). Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLoS Med, 5(2), e45. Longino, C. F., Jr. (1998). The limits of scientific medicine: paradigm strain and social policy. J Health Soc Policy, 9(4), 101-116. Mayor, S. (2008). Meta-analysis shows difference between antidepressants and placebo is only significant in severe depression. BMJ, 336(7642), 466. Moncrieff, J. (2006). Why is it so difficult to stop psychiatric drug treatment? It may be nothing to do with the original problem. Med Hypotheses, 67(3), 517-523. Moncrieff, J., & Cohen, D. (2006). Do antidepressants cure or create abnormal brain states? PLoS Med, 3(7), e240. Olfson, M., Marcus, S. C., & Shaffer, D. (2006). Antidepressant drug therapy and suicide in severely depressed children and adults: A case-control study. Arch Gen Psychiatry, 63(8), 865-872. Competing interests: None declared
Sauce for the goose 8 March 2008
R N Stenning, Chiropractor Devon Ex13 5DZ Send response to journal: Re: Sauce for the goose	It is interesting noting the responses that you have had and how quickly those who would likely dismiss complimentary medicine on the grounds of lack of evidence are so quick to cite patient testimonials as the true gauge of a treatment's worth when it suits them. How many clinicians purely follow the published evidence and ignore what they know 'works" - I would suggest not many successful ones. I doubt that there is a single study in which someone cannot pick out some methodological flaw. The trouble is we measure what we can measure and not necessarily what is important. Is it not time to put evidence-base into perspective, as a useful and essential aid to practice but not the be all and end all? Competing interests: None declared
A boost for placebos? 11 March 2008
John Merrick, Retired GP doing some locums Street, Somerset Send response to journal: Re: A boost for placebos?	What will come as no surprise to GP's is not that antidepressants are only slightly better than placebo, but that placebos are remarkably effective. In addition, all patients in the trials will have received the therapeutic effect of a consultation with a doctor and time spent completing depression scales. Unfortunately, treating patients with placebos is now frowned upon - although some would say that this may be main benefit of homeopathy etc. I remember hearing stories of GP's, in the days of fewer effective drugs,having different coloured aspirins in their dispensaries - if one colour didn't work, another was tried. Some confirmation of system of using different colours was published in a trial I think by Huskisson EC (regrettably I could not find the reference), in which he found that different coloured tablets worked better for different complaints. Perhaps we should bring back placebos! Competing interests: None declared
Lies, damned lies, statistics and the BMJ 2 April 2008
Zekria Ibrahimi, Psychiatric Patient Coombs Library UB1 3EU Send response to journal: Re: Lies, damned lies, statistics and the BMJ	The U turn is never the most pretty of manoeuvres. The 180 degree volte face by the BMJ between the 1 and 8 March, 2008, was far from graceful and elevated. On 1 March, the BMJ hiccupped forth an article on antidepressants revolving round the recent Kirsch study. Irving Kirsch indicated that the 'new generation' antidepressants were not much better than placebo except in severe depression.(1). These 'novel' SSRI (Selective Serotonin Reuptake Inhibitors) drugs like Prozac and Seroxat have been as much unfortunate magnets for clinical controversy as they have also seemed to attract blatantly bloated profits for Eli Lilley, GSK, and a pack of companies touting them as part of the all- American lifestyle. The older anti- depressants- the MAOIs (Monoamine Oxidase Inhibitors) and the Tricyclics- induced veritable landslides of side effects. The MAOIs could engender deadly hypertensive problems simply when taken alongside everyday foods ranging from bananas to broadbeans. The Trycyclics like amitriptyline were anticholinergic and thus precipitated nasty autonomic implications. But- lo!- the Selective Serotonin Reuptake Inhibitors were nigh! The archetypal ancestor of them all, Prozac, was alleged to affect 'only' the neurotransmitter, Serotonin; and thus a fake promise could be loudly presented to many a gullible patient- that the SSRIs would soon appear to have fewer side effects, that they were the portal to a psychopharmacological paradise- which, despite the boasts, they actually were not. They were wrongly proclaimed by drug companies (replete with quite dizzy PR spin) to be at the peak of probity and purity. However, with the SSRIs, it was not a matter of reduced, but, rather, different, side effects: motor problems mirroring those for anti- psychotics, from akathisia to tardive dyskinesia; and personality complications, including suicide ideation; and so on... And , according to the Kirsch study, it is not only that the SSRIs are packed with brutal side effects of their own, but also that they are not much better than placebo for many labouring under the crisis that is depression. On 8 March, the BMJ, in an editorial, provided a less than convincing riposte to Kirsch's suggestions. Turner and Rosenthal, who produced it, claimed to have done similar research to Kirsch and to have found the same results in terms of 'effect size'. Turner and Rosenthal admitted the sad fact that there had been 'selective publication' and that 'antidepressant drugs' are much less effective than is apparent from journal articles.' (3) Where the 8 March BMJ editorial was fundamentally unprincipled was in its decision to shift from a medical to a statistical framework. Kirsch himself has admitted that there is a dichotomy between 'clinical' and 'statistical' significance. (4) It was that dichotomy which the 8 March article sought- unfairly- to exploit. Statistics are numbers, not people. Statistics are complicated and they often have a cold- hearted ambience. The obscure and far from easy mathematics of the Null Hypothesis and the Confidence Interval can rapidly be misused to bamboozle the public. Turner and Rosenthal unrepentantly declared that, though their own results replicated those of Kirsch, 'our interpretations of them were quite different... we conclude that each drug was superior to placebo.' (5) In the 8 March editorial, effect size and statistical significance are distinguished and separated. Statistical significance is meant to be 'all or nothing', 'black or white', whereas 'effect size' has a gradualist 'shades of grey' approach. But it is not unreservedly right to suggest that 'tests of significance... tell us nothing about the size of the effect.' They tell us that the difference is large enough not to be random. If we want to test for an even larger difference, we can change P from 0.05 to 0.01 or 0.001, etcetera. The unsettling truth is that, if the NICE guidelines for effect size are 'arbitrary' and 'relative', so are the levels at which P can be set. The 0.05 P- figure is absolutely arbitrary, is plucked from the air, is not as set in stone as the Mosaic tablets. Ststistical procedures can always be 'arbitrary' and 'relative'; they deal with probabilities and 'interpretations', not fixed certainties. Indeed, the 8 March editorial, though not as its original intention, rotates eventually round the protean character of statistics, the worrying reality that all statistical data- and procedures- can be 'arbitrary' and 'relative'. Here is mathematics, not medicine, and very confusing and difficult mathematics too. I could never cope without my statistical gospel- 'Medical Statistics, A Guide to Data Anbalysis and Critical Appraisal', which is a BMJ book. (6) So how does the BMJ itself officially regard 'effect size', denounced as only half adequate by one of its own editorials? 'Effect sizes are important because they can be used to describe the magnitude of the difference between two groups in either experimental or observational study designs.' (7) And the degrees of effect size stated in the BMJ book are the same as those in the NICE guidelines. 'An effect size of 0.2 is considered small, 0.5 is considered medium, and 0.8 is considered large.' (8) Strangely, there is no mention in the BMJ book of these effect size levels being 'arbitrary' or 'relative'. Indeed, 'the effect size and mean difference... give a meaningful interpretation to the P values.' (9) The BMJ book and the BMJ editorial do agree that P values concern mere statistical significance, and effect size seems to allow for medical insight into the data; but there is no mention in the BMJ book of 'd- juice' and other sad fanciful obscurations adumbrated by Turner and Rosenthal. This is a mathematical dilemma, and we must thus, with our own limited comprehension, ask and beg the mathematicians to intervene- who will probably provide us with yet more of a fog of vertiginous puzzles, and not any neat limpid answer. Statistics- in fact, the whole of mathematics- is capable of being even more opaque and uncertain than theology. To, with tired terror, retreat from the mathematics, and head, or limp, back- bewildered and wounded- to the grim battlefield of medicine. What has happened with the anti- depressants unfortunately reflects what has occurred with the anti- psychotics. The typical neuroleptics, the former treatment, had a variety of vicious side effects. Then the atypical anti- psychotics were declared to be a 'miracle', a 'revolution'. But they have alas turned out to be no miraculous revolution. They possess a cascade of different side effects of their own- the 'metabolic syndrome' (10). And they may have no greater efficacy than the previous typical neuroleptics. (11) One conclusion must be that the entire statistical approach to psychiatry is not transparent enough. Drug companies, and even the FDA (Food and Drug Administration), are being insufficiently open about trials. Clinicians simply input the data into some SPSS program, maybe without understanding the first principles in mathematical terms, and without providing enough objectivity in the ultimate medical implications. Kirsch approves of the concept of effect size, but Turner and Rosenthal criticize it. Perhaps the clash between effect size and statistical significance is an echo of the dispute between the seemingly more dynamic Bayesian creed of number- crunching and the more fixed and static conventional frequentist dogma. (12) There is danger ahead when statistics and medicine are not able to ride comfortably in tandem. I do not belong to the priesthood of doctors. I am a Section 3 psychiatric patient who feels betrayed by this debilitating pseudo- metaphysical BMJ debate over psychotropic drugs. REFERENCES: (1) BMJ. 1 March. 2008. Susan Mayor. 336: pg. 466. (2) Talking Back to Prozac: Peter Breggin. Prozac Backlash: Joseph Glenmullen. (3) BMJ. 8 March. Editorial. E. H. Turner and R. Rosenthal. 2008. 336: pg. 516. (4) BMJ. 1 March. 2008. 336: pg. 466. From: Kirsch et al. PLoS Medicine 2008;5:e45. (5) BMJ. 8 March. 2008. 336: pg. 516. (6) Medical Statistics A Guide to Data Analysis and Critical Appraisal. J. Peat and B. Barton. Blackwell and BMJ. (7) Ibid. pg. 53. (8) Ibid. pg. 54. (9) Ibid. pg. 73. (10) Metabolic abnormalities associated with second generation antipsychotics: fact or fiction? Development of guidelines for screening and monitoring. Marc de Hert et al. International Clinical Psychopharmacology. Volume 21 Supplement 2 March 2006 S10- S17 (11) The Cost of Drugs for Schizophrenia. American Journal of Psychiatry. R. Freedman et al. 163: 12. December 2006. (12) BMJ. 24 October 1998. Statistical notes. Bayesians and frequentists. J. M. Bland, D. G. Altman. Competing interests: None declared
Effectiveness of Anti-depressants 7 April 2008
Dr Arun Viswanath, Speciality Registrar Sheffield Care Trust Send response to journal: Re: Effectiveness of Anti-depressants	Firstly, it was a thoughtful approach on the part of the Royal College of Psychiatrists in advising patients to consult their doctor before stopping antidepressants, which possibly can cause discontinuation symptoms, relapse of illness and even suicidal ideations. There is strong scientific evidence to support the involvement of serotonin and depression. Biological models and animal studies have clearly demonstrated this. The authors of the paper (2) studied four heterogeneous second generation antidepressants (fluoxetine, paroxetine, nefazodone, venlafaxine), not all the FDA database, limiting its conclusions. The duration of the study is also small. A recent study showed response (a 50 % reduction in QIDS-SR16 scores ) to citalopram in MDD took on average 6 weeks and remission ( a score of 7 or less on the HAMD or 5 or less on the QIDS-SR16 ) took on average 7 weeks to achieve.(1) We also understand that in these trials about half of the patients did not "complete the treatment". And this percentage was slightly higher in the treatment than the placebo groups. It is very sound practice in randomized clinical trials to compare treatment and control groups on the "intention to treat principle". And in studying antidepressants, since still being around after six weeks and still swallowing the pills can be considered in itself a response. There is also a possibility of the group studied may be treatment resistant. This paper does raise an interesting debate. Independent studies funded by health services rather than drug companies are more reliable. Before making any decision, pragmatic studies should be looked at the earliest as large amount data is already available. 1.Trivedi MH . Evaluation of outcomes with citalopram using measurement based care in STAR*D : implications for clinical practice . Am Journal of Psychiatry 2006:163:28-40 2. Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration : Irving Kirsch, Brett J. Deacon, Tania B. Huedo-Medina, Alan Scoboria, Thomas J. Moore, Blair T. Johnson Competing interests: None declared Editorial note On 28 April 2008, at the author's request, competing interests were changed to "None declared" and the second author's name was removed.