Figures
Abstract
Background
Previous research has shown that academic physicians conflicted by funding from the pharmaceutical industry have corrupted evidence based medicine and helped enlarge the market for drugs. Physicians made pharmaceutical-friendly statements, engaged in disease mongering, and signed biased review articles ghost-authored by corporate employees. This paper tested the hypothesis that bias affects review articles regarding rimonabant, an anti-obesity drug that blocks the central cannabinoid receptor.
Methods/Principal Findings
A MEDLINE search was performed for rimonabant review articles, limited to articles authored by USA physicians who served as consultants for the company that manufactures rimonabant. Extracted articles were examined for industry-friendly bias, identified by three methods: analysis with a validated instrument for monitoring bias in continuing medical education (CME); analysis for bias defined as statements that ran contrary to external evidence; and a tally of misrepresentations about the endocannabinoid system. Eight review articles were identified, but only three disclosed authors' financial conflicts of interest, despite easily accessible information to the contrary. The Takhar CME bias instrument demonstrated statistically significant bias in all the review articles. Biased statements that were nearly identical reappeared in the articles, including disease mongering, exaggerating rimonabant's efficacy and safety, lack of criticisms regarding rimonabant clinical trials, and speculations about surrogate markers stated as facts. Distinctive and identical misrepresentations regarding the endocannabinoid system also reappeared in articles by different authors.
Citation: McPartland JM (2009) Obesity, the Endocannabinoid System, and Bias Arising from Pharmaceutical Sponsorship. PLoS ONE 4(3): e5092. https://doi.org/10.1371/journal.pone.0005092
Editor: Christian Gluud, Copenhagen University Hospital, Denmark
Received: October 15, 2008; Accepted: March 1, 2009; Published: March 31, 2009
Copyright: © 2009 McPartland. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The author has no support or funding to report.
Competing interests: The author previously received research and/or salary support from the Cannabinoid Research Institute, a research division of GW Pharmaceuticals (www.gwpharm.com).
Introduction
The epidemic of obesity began, as many modern epidemics do, with a reclassification. In 1998, the number of overweight and obese individuals in the USA swelled instantaneously by 37 million, when a NIH task force redefined overweight as a body mass index (BMI) ≥25 kg/m2, and obesity as BMI ≥30 [1]. The task force was criticized for ignoring studies that disputed BMI as a valid surrogate marker for adiposity, and for circuitously basing its reclassification upon opinions and flawed studies authored by its own members, rather than independent studies [2], [3]. Nearly 90% of the obesity task force members had financial ties to the weight-loss industry, including pharmaceutical companies and weight loss clinics [4]. The Chair of the task force stated that pharmaceutical corporations “have no influence over what I say. … I'm not accepting payment directly. It comes through a company that runs continuing education. Maybe that's a bad thing. But if you did away with this, you would wipe out 80 percent of the medical education programs” [4]. The Chair of the task force was former President of the NAASO (North American Association for the Study of Obesity, also called the Obesity Society). The NAASO is an accredited continuing medical education (CME) provider. The ex-President approved NAASO CME programs as free of commercial bias [5], [6], although the programs were funded by Sanofi-Aventis (the manufacturer of a weight-loss drug), and he received financial support from Sanofi-Aventis [7]. A year after chairing the obesity task force, the ex-President was identified in a lawsuit as the guest author of a ghostwritten review on obesity commissioned by Wyeth-Ayerst regarding long-term, off-label use of “fen-phen” (fenfluramine and phentermine) [8].
In 2004, researchers from the Centers for Disease Control (CDC) reported that obesity caused 400,000 deaths in the year 2000 [9]. Despite the fact that this statistic was unsupportable (and was downsized after a congressional inquiry [10]), Medicare officials promptly announced they would treat obesity as a disease, opening the way for government reimbursement of treatments [11]. A second group of researchers reanalyzed the data used to generate the 400,000 number. After adjusting for confounding factors, obesity-related deaths in 2000 numbered 25,814 – less than 7% of the original estimate [12]. U.S. Surgeon General Richard Carmona subsequently announced, “Obesity is the terror within…the magnitude of the dilemma will dwarf 9–11 or any other terrorist attempt” [13]. Severe obesity is an indisputable health hazard, and its prevalence is rising. But the framing of obesity as a 9–11 terror is an example of disease mongering, which includes the promotion of new diseases, the expansion of illness boundaries, the medicalization of normal physiology, and the expansion of markets for disease treatments [14].
Cannabinoids, evidence based medicine, and surrogate markers
Two new anti-obesity drugs, rimonabant (Acomplia®, Sanofi-Aventis) and taranabant (Merck), work by a new mechanism: blockade of the cannabinoid 1 (CB1) receptor, an integral part of the endocannabinoid system (ECS) [15]. Obesity leads to excessive endocannabinoid production, which drives CB1 in a feed-forward dysfunction [16]. Endocannabinoids as well as plant cannabinoids in marijuana stimulate appetite, so it makes sense that a CB1 antagonist would suppress appetite. However, endocannabinoids do more than modulate appetite. The ECS plays important roles in neurogenesis, neurodegenerative diseases, mood disorders, pain perception, gut function, immunity, and inflammation [17]. These important roles suggest that ECS blockade might cause adverse effects. However, this type of physiological rationale is not accepted by evidence based medicine (EBM) guidelines [18]. EBM accepts randomized clinical trials (RCTs) as best evidence. Pharmaceutical corporations increasingly recognize the value of RCTs in shaping EBM. They treat RCTs as important resources to be managed, thereby extending their marketing arm into the peer-reviewed medical literature [19]. Pharmaceutical corporations spent US$57.5 billion on marketing alone in 2004. This was substantially greater than US$31.5 billion expended on domestic pharmaceutical research [20].
Four rimonabant-in-obesity (RIO) RCTs, all funded and conducted by Sanofi-Aventis, have been published, although 25 RCTs of rimonabant for the treatment of obesity and diabetes are completed or underway [21]. This 4-to-25 ratio suggests “publication bias,” which arises when pharmaceutical corporations choose not to publish unfavorable studies [22], [23]. Criticisms of the RIO trials included the use of unvalidated or disputed surrogate endpoints [24], [25], favorable claims not supported by trial data [26], overstated treatment efficacy [27], downplayed adverse effects [25], [28], [29], lack of internal validity and external validity or generalizability [30], [31], and failure to disclose financially-conflicted interests [32]. However, these criticisms and other types of narrative reviews are not recognized by EBM; EBM relies upon meta-analyses of RCTs [18]. A meta-analysis of the RIO RCTs concluded that rimonabant was safe and effective [33]. The meta-analysis was funded by Sanofi-Aventis. Industry-funded meta-analyses tend to be less transparent, have more methodological flaws, and make more pro-industry conclusions regarding drugs than do independent meta-analyses [34]. Consistent with this, four independent meta-analyses of the RIO trials have questioned rimonabant's efficacy and potential for adverse effects [26], [35]–[37].
One taranabant clinical trial has been published in the peer-reviewed literature [38]. The study also used unvalidated or disputed surrogate endpoints, made claims not supported by trial data, and downplayed adverse effects. The use of surrogate endpoints instead of clinical end points has come under recent scrutiny. Surrogate markers help get drugs to the market quickly, but they may not correlate with disease outcome. Just because patients with flu have a fever, for example, doesn't mean that treating the fever will clear the infection [39]. Rosiglitazone was fast-tracked through FDA approval because it lowered serum glucose levels, but a meta-analysis showed that rosiglitazone increased the risk of myocardial infarction [40]. According to Steven Nissen, author of the meta-analysis, “Wait long enough, and you're going to find that all surrogates eventually fail due to these off-target effects” [39]. Serum cholesterol serves as a surrogate for cardiovascular disease, and ezetimibe-simvastatin lowers serum cholesterol, but ezetimibe-simvastatin did not slow the development of atherosclerosis in patients [41]. Rimonabant and taranabant RCTs employed several disputed surrogate markers. Even “metabolic syndrome,” a composite surrogate measure, has limited value as a cardiovascular risk marker [42], [43]. Sanofi-Aventis answered these criticisms with the STRADIVARIUS study, which measured rimonabant's effects upon coronary artery atheroma volume [44]. The study was conducted by Steve Nissen, a champion of clinically-relevant outcome measures [39], [40], [45]. Coronary artery atheroma volume, however, is a nonvalidated surrogate endpoint for cardiovascular outcomes [46], [47]. The study showed that rimonabant had no effect on percent atheroma volume, although Nissen and colleagues noted improvements in secondary nonvalidated surrogate endpoints, such as normalized total atheroma volume [44].
ECB, CME, and review articles
In addition to RCTs and meta-analyses, clinicians base rational EBM decisions upon CME presented by fellow physicians. Clinicians must participate in CME to fulfill licensure requirements, making them a “captured audience” for corporate-sponsored messages. Pharmaceutical corporations routinely seed CME with review articles that promote their products, thereby further unraveling EBM [48]. Review articles often contain industry bias [49], [50], especially articles in journal supplements, which are not usually peer reviewed [19]. Journal supplements are quite lucrative to medical journals, because pharmaceutical corporations sponsor them. Corporate employees may ghostwrite review articles, and then influential physicians are recruited to sign the articles [51]. Whereas authorship establishes accountability and responsibility, ghost authorship increases the potential for conflicted manipulation. Documents made public in litigation showed that Wyeth-Ayerst [8], Pfizer [52], [53], and Merck [54] employed corporate authors to ghost author CME review articles.
In the past few years, several CME review articles of rimonabant have been published. Some authors of the review articles also served on the NIH obesity task force, the NAASO board, and coauthored RIO publications. One rimonabant review article was presumably ghostwritten because the authors were listed as “editors,” without identifying a primary author, and a Sanofi-Aventis copyright appeared in small print on the back cover of the supplement [55]. The purpose of this paper was to test the hypothesis that rimonabant review articles expressed a high incidence of bias.
Methods
Review articles were identified through a MEDLINE search using the keywords endocannabinoid AND obesity AND rimonabant, limited to articles published prior to 2007 (when the FDA reviewed rimonabant). To be included in the analysis, a review article had to meet the following criteria:
- ample information (≥2 paragraphs) describing the ECS system and obesity;
- ample information (≥1 paragraph) describing obesity and rimonabant;
- authored by a USA physician who received financial support from Sanofi-Aventis.
Because the study aimed to uncover identical bias by different authors, only one article by each author was analyzed; the earliest published article was analyzed and subsequent publications were excluded. Evidence of real or potential conflicts of interest (i.e., financial support in the form of honoraria, education support, research funding, or identification as a consultant or speaker) was obtained by searching Google using each physician's name combined with Sanofi-Aventis. Biases and misrepresentations were measured by three methods:
- Review articles were analyzed with a validated instrument for monitoring bias in CME [56]. The Takhar instrument consisted of 13 questions (see Table 1), with scores graded on a scale from 1 (strongly disagree this paper displays bias) to 4 (strongly agree this paper displays bias), or N/A (not applicable). The instrument was designed for oral CME presentations, so it was slightly modified (e.g., “author” instead of “speaker”).
- The “RIO bias tally” searched articles for biased statements or inappropriate omissions that originally appeared in the RIO publications (see Table 2), graded on a scale from 0 (no bias evident in review article), 1 (a mix of biased and unbiased statements) to 2 (bias evident), or N/A (no statement made regarding the item).
- Articles were scanned for recurrent themes and for identical misrepresentations about the ECS written by different authors.
Results
Eight review articles were identified that met inclusion criteria [57]–[64]. Only three of eight articles disclosed authors' conflicts of interest with Sanofi-Aventis. Two articles carried the statement that the author had “no conflict of interest,” despite easily accessible information to the contrary—searches with Google revealed all the authors served as consultants, or on speaker bureaus, or received other financial support from Sanofi-Aventis.
The Takhar CME bias instrument demonstrated bias in all the review articles (Table 3), with mean scores ranging from 2.6 (weakly biased) to 3.6 (strongly biased). Collectively, the mean Takhar score was 3.08 (95% CI: 2.78 to 3.39). Ten industry-friendly statements that originally appeared in RIO publications reappeared in the eight review articles (Table 3). Biased statements included disease mongering, speculations regarding surrogate markers stated as facts, lack of acknowledgment of RIO design flaws, and exaggerated statements of rimonabant's efficacy and safety. Mean scores from the Takhar bias instrument (Table 3) correlated with mean scores from the RIO bias tally (Table 3), but fell short of statistical significance (r = 0.50, p = 0.11).
Recurrent visual and textual themes emerged from the eight review articles. Graphics from Sanofi-Aventis promotional materials [55] reappeared in a review article [61]. Articles by different authors in different journals nevertheless used similar stock photos from Getty Images, Inc. (e.g., [60] and [63]). Three unusual yet identical misrepresentations about the ECS appeared in articles by different authors:
- The hypothalamus was named first in descriptions of CB1 expression in the brain. Three of eight articles shared this misrepresentation [57], [58], [63]. Actually CB1 expression is relatively low in the hypothalamus. In human brain, the rank order of CB1 receptor density is: substantia nigra>globus pallidus>hippocampus>cerebral cortex>putamen>caudate>cerebellum>amygdala>thalamus = hypothalamus [65]. This error may arise from the fact that Sanofi-sponsored research has focused upon the hypothalamus (e.g., [66]).
- Adipose tissue was listed amongst tissues with dense CB1 expression. Six of eight articles stated this [57]–[61], [63]. Sanofi-sponsored research has highlighted adipose tissue in rimonabant's “peripheral effects” (e.g., [67]). However, most independent studies have found that CB1 expression is relatively low or undetectable in adipose and adipocyte-rich tissue such as bone marrow [68]–[71]. Adipocyte CB1 expression actually decreases in obese research participants [67], [72].
- Hepatic tissue was listed amongst tissues with dense CB1 expression. Five of eight articles stated this [57], . Sanofi-sponsored researchers claim that hepatocytes contribute to rimonabant's peripheral effects in mice [73]. On the other hand, CB1 expression is sufficiently low that some independent studies failed to identify CB1 in liver at all [68], [70]. Recent studies have detected CB1 in fibrotic liver cells; hepatic CB1 expression in humans may be limited to cirrhotic or other pathological conditions [74].
Discussion
All the authors of rimonabant review articles held academic positions, many at prestigious institutions. They typified “medical opinion leaders” sought by pharmaceutical corporations to sign ghostwritten articles [48], [52], [53], [75]. Seven of eight rimonabant review articles appeared in journal supplements, which are non-peer-reviewed, usually industry-funded publications known to carry a high incidence of bias [19]. The Takhar bias instrument demonstrated bias in all eight articles (Table 3). The mean score was 3.08 (95% CI: 2.78 to 3.39), significantly greater than the 2.5 score that separates unbiased from biased publications [56]. In comparison, the mean score for 17 accredited CME events evaluated by Takhar and colleagues was 1.65 (95% CI: 1.32 to 1.99) [56]. An additional analysis of non-Sanofi-supported review articles would provide compelling data for comparison. But “absence of evidence is not evidence of absence”—to prove lack of Sanofi support would require much more than a Google search.
The RIO bias tally identified ten nearly-identical industry-friendly statements or inappropriate omissions in articles written by different authors (Table 3). These statements originally appeared in RIO publications, which acknowledged editorial assistance by Sanofi-Aventis (e.g., [7]), whereas the review articles did not. Nearly identical illustrations reappeared in several articles, and distinctive factual misrepresentations reappeared in articles by different authors. Replication of passages in a single author's work may indicate only carelessness, but replication of passages in articles by different authors raises the question of whether a common ghost author was involved. One rimonabant review article (not included in the analysis because we found no evidence that Sanofi funded its authors) listed the authors as “editors,” and the primary (ghost-) author was unidentified [55]. Of course, judgment regarding ghostwriting or plagiarism should be withheld until the candidate publications are appraised by an editorial board or ethics committee. The same proviso was made by Errami and Garner [76], who used a computational text-searching algorithm to identify “duplicate publications” and “plagiarism” in abstracts cited by PubMed. A search of their database (http://spore.swmed.edu/dejavu/) using the keyword “rimonabant” revealed seven pairs of duplicate publications (including one review article identified herein). The cases of suspected plagiarism repeatedly engaged in disease mongering, which expands the market for those who sell disease remedies. Disease mongering and “supersizing” of rimonabant's indications have been criticized [43], and satirized by a description of “indolebant,” a fictional CB1 antagonist that treats “extreme laziness” [77].
Rationally choosing the best medication, like other sorts of clinical decision-making, has increasingly relied upon EBM. Thus whoever generates EBM, by funding RCTs, meta-analyses, and CME, may bias clinical decision-making regarding pharmaceuticals [78]. Industry-friendly bias is not unique to rimonabant publications. The discovery process in recent litigation has revealed that many pharmaceutical corporations recruit and train influential physicians for the purpose of manipulating their colleagues. These physicians sign biased ghostwritten articles without disclosing conflicts of interest [48], [52], [53], [75]. For participating in Parke-Davis promotional CME efforts, physicians received honoraria up to US$158,250, in addition to paid travel, lodging, and amenities at luxury resorts [53]. This behavior is not limited to MDs — a pharmaceutical marketing disclosure law in Vermont revealed that the third highest recipient in the state was an osteopathic physician who received $99,843 in 2007 [79]. Despite the fact that osteopathy began as an essentially drug-free school of medicine, the pharmaceutical industry now imparts significant financial leverage over that profession.
Financial conflicts of interest also bias clinical practice guidelines and FDA decisions. An analysis of 44 clinical guidelines revealed 87% of panelists received financial support from pharmaceutical companies, yet only two of the guidelines disclosed panelists' financial conflicts [80]. A larger investigation of over 200 guidelines revealed about 70% of guideline panels being affected [81]. In one guideline panel, every panel member was paid by a drug manufacturer, and that manufacturer's drug was recommended by the panel [81]. Members of FDA drug advisory committees have financial conflicts but rarely recluse themselves from voting, and they tend to vote in favor of corporations that sponsored them [82]. Ten of the 32 FDA panelists that voted in favor of rofecoxib and valdecoxib received fees from the makers of those drugs [83]. Had those financially conflicted panelists been reclused, the FDA would have voted against continued sales of rofecoxib and valdecoxib [83].
In summary, financial conflicts permeate the system and are by no means limited to corporations referenced in this article, such as Merck, Parke-Davis, Pfizer, Sanofi-Aventis, and Wyeth-Ayerst. On balance, pharmaceutical corporations do good work and aid in humanitarian efforts. For example Sanofi-Aventis provides artemisinin at cost to malaria-endemic countries [84]. Nevertheless, ghost authorship and the corrupting effects of covert financial support must cease. Only three of eight rimonabant review articles disclosed corporate sponsorship; two authors specifically denied conflicts. Lack of disclosure prevents readers from judging the credibility of an author. Medical journals should require stronger author disclosure procedures, and universities should discipline academics who sign ghostwritten articles. This behavior should be regarded as unethical misconduct [85]. More broadly, researchers with conflicts of interest should not be allowed to sit on guideline committees and regulatory boards. Corporate funding of CME programs and review articles should be abolished.
Post script
While this paper was under review, Merck halted taranabant RCTs, and Sanofi-Aventis removed rimonabant from the European market. The FDA rejected rimonabant after data submitted by Sanofi-Aventis revealed adverse effects in RIO trials that went unreported in RIO publications [86], including one death in a rimonabant-treated subject (ruled a suicide by the FDA, [86]) that did not appear in the pertinent publication [7]. Although the risk-benefit ratio of cannabinoid receptor blockade may preclude its use for chronic conditions such as obesity and drug or alcohol dependence, cannabinoid receptor blockade could serve in the treatment of acute endocannabinoid dysregulation, such as hepatic cirrhosis, hemorrhagic or endotoxic shock, cardiac reperfusion injury, and doxorubicin-induced cardiotoxicity [15].
Author Contributions
Conceived and designed the experiments: JMM. Performed the experiments: JMM. Analyzed the data: JMM. Contributed reagents/materials/analysis tools: JMM. Wrote the paper: JMM.
References
- 1. NIH (1998) Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report. NIH Publication No. 98-4083. Available: http://www.nhlbi.nih.gov/guidelines/obesity/ob_gdlns.htm. Accessed Feb 29, 2008.
- 2. Strawbridge WJ, Wallhagen MI, Shema SJ (2000) New NHLBI clinical guidelines for obesity and overweight: will they promote health? Am J Public Health 90: 340–343.
- 3. Prentice AM, Jebb SA (2001) Beyond body mass index. Obes Rev 2: 141–147.
- 4. MacPherson K, Silverman E (1997) Fat's overlap. Newark Star-Ledger [Newspaper] Feb 17, 1997.
- 5. NAASO (2004) Understanding and treating obesity. Obesity On Line. Available: http://www.obesityonline.org/meetings/treating_obesity/index.cfm#Accreditation9620Statement. Accessed Dec 10, 2006 [site no longer online].
- 6. NAASO (2008) Understanding and treating obesity. Obesity On Line. Available: http://www.obesityonline.org/meetings/treating_obesity. Accessed Sept 15, 2008.
- 7. Pi-Sunyer FX, Aronne LJ, Heshmati HM, Devin J, Rosenstock J (2006) Effect of rimonabant, a cannabinoid-1 receptor blocker, on weight and cardiometabolic risk factors in overweight or obese patients: RIO-North America: a randomized controlled trial. JAMA 295: 761–775.
- 8. Birmingham K (1999) Lawsuit reveals academic conflict-of-interest. Nature Medicine 5: 717.
- 9. Mokdad AH, Marks JS, Stroup DF, Gerberding JL (2004) Actual causes of death in the United States, 2000. JAMA 291: 1238–1245.
- 10. Waxman HA (2004) Committee on Oversight and Government Reform: Questions on Estimate of Mortality from Obesity. Available: http://oversight.house.gov/story.asp?ID298. Accessed Feb 21, 2008.
- 11. US-HHS (2004) HHS Announces Revised Medicare Obesity Coverage Policy. News release, July 15, 2004. Available: http://www.hhs.gov/news/press/2004pres/20040715.html. Accessed Feb 21, 2008.
- 12. Flegal KM, Graubard BI, Williamson DF, Gail MH (2005) Excess deaths associated with underweight, overweight, and obesity. JAMA 293: 1861–1867.
- 13. Saguy AC, Almeling R (2008) Fat in the fire? Science, the news media, and the “obesity epidemic” Sociological Form 23: 53–83.
- 14. Moynihan R, Henry D (2006) The fight against disease mongering: generating knowledge for action. PLoS Med 3: e191.
- 15. McPartland JM (2008) The endocannabinoid system: an osteopathic perspective. J Am Osteopath Assoc 108: 586–600.
- 16. Matias I, Di Marzo V (2007) Endocannabinoids and the control of energy balance. Trends Endocrinol Metab 18: 27–37.
- 17. Pacher P, Batkai S, Kunos G (2006) The endocannabinoid system as an emerging target of pharmacotherapy. Pharmacol Rev 58: 389–462.
- 18. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS (1996) Evidence based medicine: what it is and what it isn't. BMJ 312: 71–72.
- 19. Smith R (2005) Medical journals are an extension of the marketing arm of pharmaceutical companies. PLoS Med 2: e138.
- 20. Gagnon MA, Lexchin J (2008) The cost of pushing pills: a new estimate of pharmaceutical promotion expenditures in the United States. PLoS Med 5: e1.
- 21. NIH (2007) National Institutes of Health Clinical Trials Registry. Available at: http://www.clinicaltrials.gov. Accessed Jun 9, 2007.
- 22. Bekelman JE, Li Y, Gross CP (2003) Scope and impact of financial conflicts of interest in biomedical research: a systematic review. JAMA 289: 454–465.
- 23. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R (2008) Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med 358: 252–260.
- 24. Padwal R, Majumdar S (2006) Metabolic risk factors, drugs, and obesity. N Engl J Med 354: 974–975.
- 25. Lambert ML, Roberfroid D, Vlayen J (2007) Rimonabant in obese patients with type 2 diabetes. Lancet 369: 553–555.
- 26. Anonyme (2006) Rimonabant: new drug. Obesity: loss of a few kilos, many questions. Prescrire Int 15: 123–126.
- 27. Roberfroid D (2007) Rimonabant in obese patients with type 2 diabetes. Lancet 369: 553–555.
- 28. Gadde KM (2006) Effect of rimonabant on weight and cardiometabolic risk factors. JAMA 296: 649.
- 29. Simons-Morton DG, Obarzanek E, Cutler JA (2006) Obesity research–limitations of methods, measurements, and medications. JAMA 295: 826–828.
- 30. Cleland SJ, Sattar N (2006) Does rimonabant pull its weight for type 2 diabetes? Lancet 368: 1632–1634.
- 31. Yanovski SZ (2005) Pharmacotherapy for obesity–promise and uncertainty. N Engl J Med 353: 2187–2189.
- 32. Editor (2006) Correction: failure to disclose financial interest. JAMA 295: 1252.
- 33. Van Gaal L, Pi-Sunyer X, Despres JP, McCarthy C, Scheen A (2008) Efficacy and safety of rimonabant for improvement of multiple cardiometabolic risk factors in overweight/obese patients: pooled 1-year data from the Rimonabant in Obesity (RIO) program. Diabetes Care 31: Suppl 2S229–240.
- 34. Jorgensen AW, Hilden J, Gotzsche PC (2006) Cochrane reviews compared with industry supported meta-analyses and other meta-analyses of the same drugs: systematic review. BMJ 333: 782.
- 35. Curioni C, Andre C (2006) Rimonabant for overweight or obesity. Cochrane Database Syst Rev CD006162.
- 36. Rucker D, Padwal R, Li SK, Curioni C, Lau DC (2007) Long term pharmacotherapy for obesity and overweight: updated meta-analysis. BMJ 335: 1194–1199.
- 37. Christensen R, Kristensen PK, Bartels EM, Bliddal H, Astrup A (2007) Efficacy and safety of the weight-loss drug rimonabant: a meta-analysis of randomised trials. Lancet 370: 1706–1713.
- 38. Addy C, Wright H, Van Laere K, Gantz I, Erondu N, et al. (2008) The acyclic CB1R inverse agonist taranabant mediates weight loss by increasing energy expenditure and decreasing caloric intake. Cell Metab 7: 68–78.
- 39. Ledford H (2008) Drug markers questioned. Nature 452: 510–511.
- 40. Nissen SE, Wolski K (2007) Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 356: 2457–2471.
- 41. Kastelein JJ, Akdim F, Stroes ES, Zwinderman AH, Bots ML, et al. (2008) Simvastatin with or without ezetimibe in familial hypercholesterolemia. N Engl J Med 358: 1431–1443.
- 42. Kahn R, Buse J, Ferrannini E, Stern M (2005) The metabolic syndrome: time for a critical appraisal: joint statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 28: 2289–2304.
- 43. Banzi R, Moja L, Moschetti I, Liberati A, Gensini GF, et al. (2008) Rimonabant for overweight and “metabolic syndrome”: the attempt to supersize disease and risk by pharmaceutical marketing. Intern Emerg Med 3: 53–56.
- 44. Nissen SE, Nicholls SJ, Wolski K, Rodes-Cabau J, Cannon CP, et al. (2008) Effect of rimonabant on progression of atherosclerosis in patients with abdominal obesity and coronary artery disease: the STRADIVARIUS randomized controlled trial. JAMA 299: 1547–1560.
- 45. Mukherjee D, Nissen SE, Topol EJ (2001) Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA 286: 954–959.
- 46. Dora JM, Scheffel RS (2008) Rimonabant and progression of atherosclerosis in obese persons. JAMA 300: 280–281.
- 47. Rumsfeld JS, Nallamothu BK (2008) The hope and fear of rimonabant. JAMA 299: 1601–1602.
- 48. Fugh-Berman A (2005) The corporate coauthor. J Gen Intern Med 20: 546–548.
- 49. Angell M, Kassirer JP (1996) Editorials and conflicts of interest. N Engl J Med 335: 1055–1056.
- 50. Stelfox HT, Chua G, O'Rourke K, Detsky AS (1998) Conflict of interest in the debate over calcium-channel antagonists. N Engl J Med 338: 101–106.
- 51. Tierney WM, Gerrity MS (2005) Scientific discourse, corporate ghostwriting, journal policy, and public trust. J Gen Intern Med 20: 550–551.
- 52. Healy D, Cattell D (2003) Interface between authorship, industry and science in the domain of therapeutics. Br J Psychiatry 183: 22–27.
- 53. Steinman MA, Bero LA, Chren MM, Landefeld CS (2006) Narrative review: the promotion of gabapentin: an analysis of internal industry documents. Ann Intern Med 145: 284–293.
- 54. Ross JS, Hill KP, Egilman DS, Krumholz HM (2008) Guest authorship and ghostwriting in publications related to rofecoxib: a case study of industry documents from rofecoxib litigation. JAMA 299: 1800–1812.
- 55. Plutzky J, Woods SC (2006) The endocannabinoid system and the regulatin of energy metabolism. DOC News Sept 2006: (Suppl)1–28.
- 56. Takhar J, Dixon D, Donahue J, Marlow B, Campbell C, et al. (2007) Developing an instrument to measure bias in CME. J Contin Educ Health Prof 27: 118–123.
- 57. Cannon CP (2005) The endocannabinoid system: a new approach to control cardiovascular disease. Clin Cornerstone 7: 17–26.
- 58. Davis SN (2006) Contemporary strategies for managing cardiometabolic risk factors. Journal of Managed Care Pharmacy 12: (1Suppl)S4–S9.
- 59. Jensen MD (2006) Potential role of new therapies in modifying cardiovascular risk in overweight patients with metabolic risk factors. Obesity (Silver Spring) 14: Suppl 3143S–149S.
- 60. Henry RR (2006) Management by risk modification: the role of CB1-receptor blockade. J Amer Acad Physician Assist Suppl 2006 Nov 8–11.
- 61. Hollander P (2007) Endocannabinoid blockade for improving glycemic control and lipids in patients with type 2 diabetes mellitus. Am J Med 120: (2Suppl1)S18–28.
- 62. Aronne LJ (2007) Therapeutic options for modifying cardiometabolic risk factors. Am J Med 120: S26–34.
- 63. Lillo JL (2007) The endocannabinoid system as a novel approach for managing obesity. J Am Osteopath Assoc 107: (Suppl 2)S12–S20.
- 64. Watson KE (2007) Novel therapies for cardiometabolic risk reduction and implications for clinical practice. Rev Cardiovasc Med 8: Suppl 4S37–42.
- 65. McPartland JM, Glass M, Pertwee RG (2007) Meta-analysis of cannabinoid ligand binding affinity and cannabinoid receptor distribution: interspecies differences. British Journal of Pharmacology 152: 583–589.
- 66. Bensaid M, Gary-Bobo M, Esclangon A, Maffrand JP, Le Fur G, et al. (2003) The cannabinoid CB1 receptor antagonist SR141716 increases Acrp30 mRNA expression in adipose tissue of obese fa/fa rats and in cultured adipocyte cells. Mol Pharmacol 63: 908–914.
- 67. Engeli S, Bohnke J, Feldpausch M, Gorzelniak K, Janke J, et al. (2005) Activation of the peripheral endocannabinoid system in human obesity. Diabetes 54: 2838–2843.
- 68. Galiegue S, Mary S, Marchand J, Dussossoy D, Carriere D, et al. (1995) Expression of central and peripheral cannabinoid receptors in human immune tissues and leukocyte subpopulations. Eur J Biochem 232: 54–61.
- 69. Lynn AB, Herkenham M (1994) Localization of cannabinoid receptors and nonsaturable high-density cannabinoid binding sites in peripheral tissues of the rat: implications for receptor-mediated immune modulation by cannabinoids. J Pharmacol Exp Ther 268: 1612–1623.
- 70. Ryberg E, Vu HK, Larsson N, Groblewski T, Hjorth S, et al. (2005) Identification and characterisation of a novel splice variant of the human CB1 receptor. FEBS Lett 579: 259–264.
- 71. Nieri P, Greco R, Adinolfi B, Breschi MC, Martinotti E, et al. (2003) CB1- and CB2-cannabinoid receptor-independent lipolysis induced by WIN 55,212-2 in male rat adipocytes. Naunyn Schmiedebergs Arch Pharmacol 368: 352–359.
- 72. Matias I, Gonthier MP, Orlando P, Martiadis V, De Petrocellis L, et al. (2006) Regulation, function, and dysregulation of endocannabinoids in models of adipose and beta-pancreatic cells and in obesity and hyperglycemia. J Clin Endocrinol Metab 91: 3171–3180.
- 73. Osei-Hyiaman D, DePetrillo M, Pacher P, Liu J, Radaeva S, et al. (2005) Endocannabinoid activation at hepatic CB1 receptors stimulates fatty acid synthesis and contributes to diet-induced obesity. J Clin Invest 115: 1298–1305.
- 74. Teixeira-Clerc F, Julien B, Grenard P, Tran Van Nhieu J, Deveaux V, et al. (2006) CB1 cannabinoid receptor antagonism: a new strategy for the treatment of liver fibrosis. Nat Med 12: 671–676.
- 75. Gotzsche PC, Hrobjartsson A, Johansen HK, Haahr MT, Altman DG, et al. (2007) Ghost authorship in industry-initiated randomised trials. PLoS Med 4: e19.
- 76. Errami M, Garner H (2008) A tale of two citations. Nature 451: 397–399.
- 77. Moynihan R (2006) Scientists find new disease: motivational deficiency disorder. BMJ 332: 745–746.
- 78. McPartland JM (2008) Embrace evidence…with both eyes open. J Am Osteopath Assoc 108: 190–191.
- 79. State_of_Vermont (2008) Pharmaceutical Marketing Disclosures. Report of Vermont Attorney General William H. Sorrell. Montpelier, VT, http://www.atg.state.vt.us/upload/1215544954_2008_Pharmaceutical_Marketing_Disclosures_Report.pdf: State of Vermont.
- 80. Choudhry NK, Stelfox HT, Detsky AS (2002) Relationships between authors of clinical practice guidelines and the pharmaceutical industry. JAMA 287: 612–617.
- 81. Taylor R, Giles J (2005) Cash interests taint drug advice. Nature 437: 1070–1071.
- 82. Lurie P, Almeida CM, Stine N, Stine AR, Wolfe SM (2006) Financial conflict of interest disclosure and voting patterns at Food and Drug Administration Drug Advisory Committee meetings. JAMA 295: 1921–1928.
- 83. Steinbrook R (2005) Financial conflicts of interest and the Food and Drug Administration's Advisory Committees. N Engl J Med 353: 116–118.
- 84. Bosman A, Mendis KN (2007) A major transition in malaria treatment: the adoption and deployment of artemisinin-based combination therapies. Am J Trop Med Hyg 77: 193–197.
- 85. Moffatt B, Elliott C (2007) Ghost marketing: pharmaceutical companies and ghostwritten journal articles. Perspect Biol Med 50: 18–31.
- 86. FDA (2007) Food & Drug Administration Briefing Document: New Drug Application 21-888. Zimulti (rimonabant). Available at: www.fda.gov/ohrms/dockets/ac/07/briefing/2007-4306b1-fda-backgrounder.pdf. Accessed July 30, 2007.
- 87. Sanofi-Aventis (2006) Package insert: ACOMPLIA 20 mg film-coated tablets. Mens Care Website. Available at: www.menscare.co.uk/acomplia/acomplia.pdf. Accessed Apr 29, 2008.
- 88. EMA (2007) Summary of Product Characteristics: ACOMPLIA 20 mg film-coated tablets. European Medicines Agency. Available at: www.emea.europa.eu/humandocs/PDFs/EPAR/acomplia/acomplia-updated_pi_july07.pdf. Accessed Apr 29, 2008.
- 89. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB (2003) Years of life lost due to obesity. JAMA 289: 187–193.