Research Article

Are All Placebo Effects Equal? Placebo Pills, Sham Acupuncture, Cue
Conditioning and Their Association

--
       X
     * Irving Kirsch,
       Affiliations: Program in Placebo Studies, Beth Israel Deaconess
       Medical Center, Harvard Medical School, Boston, Massachusetts,
       United States of America, School of Psychology, Plymouth
--
       X
     * Ted J. Kaptchuk
       Affiliation: Program in Placebo Studies, Beth Israel Deaconess
       Medical Center, Harvard Medical School, Boston, Massachusetts,
       United States of America
--
Abstract

   Placebo treatments and healing rituals have been used to treat pain
   throughout history. The present within-subject crossover study examines
   the variability in individual responses to placebo treatment with
   verbal suggestion and visual cue conditioning by investigating whether
   responses to different types of placebo treatment, as well as
   conditioning responses, correlate with one another. Secondarily, this
   study also examines whether responses to sham acupuncture correlate
--
   to participate in two sequential experiments. Experiment one is a
   five-session crossover study. In each session, subjects received one of
   four treatments: placebo pills (described as Tylenol), sham
   acupuncture, genuine acupuncture, or no treatment rest control
   condition. Before and after each treatment, paired with a verbal
--
   subjects completed experiment one, and 45 completed experiment two. The
   results showed significantly different effects of genuine acupuncture,
   placebo pill and rest control on pain threshold. There was no
   significant association between placebo pills, sham acupuncture and cue
   conditioning effects, indicating that individuals may respond to unique
   healing rituals in different ways. This outcome suggests that placebo
   response may be a complex behavioral phenomenon that has properties
   that comprise a state, rather than a trait characteristic. This could
   explain the difficulty of detecting a signature for “placebo
   responders.” However, a significant association was found between the
   genuine and sham acupuncture treatments, implying that the non-specific
--

   Citation: Kong J, Spaeth R, Cook A, Kirsch I, Claggett B, et al. (2013)
   Are All Placebo Effects Equal? Placebo Pills, Sham Acupuncture, Cue
   Conditioning and Their Association. PLoS ONE 8(7): e67485.
   doi:10.1371/journal.pone.0067485
--
Introduction

   Placebo treatments and healing rituals have been used since the
   beginning of human history [1], [2]. The systematic study of placebo
   and ritual is still in its infancy [3], [4]. Whether all placebo
   treatments, or medical rituals, have equivalent effects remains
   unknown. This raises the question: Do patients who respond to one
   placebo intervention also tend to respond to other placebo
   interventions?

   In a previous clinical trial [5] of chronic pain patients, we found
   that sham acupuncture reduced pain significantly more over time than
   did placebo pills, while placebo pills offered more short-term benefits
   of improving pain-disturbed sleep over sham acupuncture. Thus it showed
   that not all placebo treatments are equal. However, this clinical trial
   involved multiple, concurrent experimental arms and was not designed to
   answer the question of whether individuals who tend to respond to sham
   acupuncture also tend to respond to placebo pills.

   In another study [6], Colloca and colleagues compared the placebo
   effects of verbal suggestion and conditioning to a control condition
   and found that verbal suggestion alone could not produce significant
--
   effects.

   Elucidating the relationship between different placebo modalities
   paired with verbal suggestion (suggestion-evoked placebo effects) as
   well as understanding their association with conditioning-evoked
   placebo effects will enhance our understanding of the variability
   observed in the placebo response. We are particularly interested in
   whether placebo responses can be characterized by the involvement of
   relatively stable traits or states, depending on particular
   circumstances.
--
   on chronic pain have shown contradictory results and often fail to show
   superiority over sham acupuncture [8]. This ambiguity may be the result
   of acupuncture's sizeable placebo effects as well as large
   inter-individual variability in response to acupuncture treatment [5],
   [8]–[11]. It is well known that some patients respond well to placebo
   and acupuncture treatments while others do not [5], [9]. Thus, as a
   secondary aim, this paper also addresses whether individuals who
--
   In this study, all subjects participated in two experiments
   sequentially. The first experiment was a multi-session crossover study
   [12] designed to test the analgesic effects of placebo Tylenol (pill),
   sham acupuncture, and genuine acupuncture (electroacupuncture) as
   compared to a no treatment (rest control). Subjects also participated
--

   In reality, the manual acupuncture was sham acupuncture and the Tylenol
   painkiller was a placebo pill. In addition, subjects were explicitly
   informed that the control condition was a baseline control for the
   study and, thus, no treatment effects were expected. To maintain
--
   Next, acupuncture was administered for 2 minutes to introduce subjects
   to the acupuncture experience. As several previous studies [18]–[20]
   have suggested that optimism is associated with placebo effects, all
   subjects were asked to complete the Life Orientation Test (LOT) to
   assess individual differences in generalized optimism versus pessimism.
--
   during each experimental session. The only difference among the
   sessions was the treatment condition: electroacupuncture, sham
   acupuncture, placebo Tylenol, or rest control condition. The order in
   which subjects received the 4 experimental conditions was randomized
   prior to study proceedings using a computerized random number
--
Interventions.

   Placebo Tylenol Pill: Subjects were informed that the goal of the
   session was to test the analgesic effect of a non-opioid analgesia
   pill, acetaminophen (Tylenol by brand), on experimental pain. They were
--
   that Tylenol can produce a general analgesic effect on the whole body,
   including the experimental pain applied to the forearms. After orally
   ingesting the pill (in reality an inert placebo pill), subjects waited
   about half an hour prior to the beginning of the post-treatment pain
   assessment.

   Sham (Placebo) Acupuncture: Sham acupuncture was performed using the
   validated Streitberger sham acupuncture device [12], . A small plastic
   ring was covered by an opaque, thin covering and then placed
   over2non-acupoints after the acupuncturist disinfected the area with
   isopropyl alcohol. A placebo needle, which was visually
   indistinguishable from genuine acupuncture needles, was inserted into
   the center of the ring and held in place by the tape. Because the
   Streitberger placebo needle has a blunt tip and a retractable shaft,
   the needle did not actually penetrate the skin; however the subjects
   felt a sensation similar to that of a pinprick or a scratch. After
--
   visual cues. Please see original publication for more details on
   experimental procedures and fMRI results [36]. The present manuscript
   only focuses on the analysis of the association between the placebo
   effects evoked by suggestion in experiment one and the visual cue
   conditioning effects in experiment two. This analysis has not been
--
   HIGH heat pain stimuli) in the 2 test pain sequences. We explored the
   association between the analgesic effect evoked by different treatments
   in experiment one (genuine acupuncture, sham acupuncture and placebo
   pills compared with rest condition, separately) and visual cue effects
   in experiment two by applying non-parametric Spearman correlations
--
   found a similar result (F = 3.63; df = 3, 137; p = 0.015). Post hoc
   analysis among the 4 experimental conditions showed that both genuine
   acupuncture and placebo pills produced significant post-treatment pain
   threshold increases (+0.79, 95% CI:[+0.25, +1.33], p = 0.004; and
   +0.74,95%CI: [+0.19, +0.1.29], p = 0.008 respectively) relative to rest
--
   rating across different groups.

   EA, electroacupuncture group; PA, placebo acupuncture group; PT,
   placebo Tylenol group; RS, resting control group.
   doi:10.1371/journal.pone.0067485.g002
   thumbnail
--
   across the treatment groups did reveal a significant progression in the
   strength of treatment effects when the treatments were ordered as rest
   control condition<placebo Tylenol = sham acupuncture
   <electroacupuncture. In this analysis, the effect of treatment was
   estimated, with the rest condition coded as 0, each of the two placebo
   treatments coded as 1, and the genuine acupuncture coded as 2. Using
   the same ANCOVA models as mentioned above, the effects of treatment
--
   Because only the pain threshold outcome showed significant differences
   among the experimental conditions, further analyses focused only on
   this outcome. In an attempt to determine whether responses to placebo
   Tylenol were significantly related to each subject's response to sham
   acupuncture, we investigated the relationship between post-treatment
   placebo Tylenol pain threshold and the following predictors,
   controlling for pre-treatment placebo Tylenol pain threshold: pre- and
   post-sham acupuncture pain threshold change, pre- and
   post-electroacupuncture pain threshold change, pre- and post- rest
   control pain threshold change, expectancy ratings for each of the4
   conditions(placebo Tylenol, sham acupuncture, genuine acupuncture, and
   rest control), age, gender, and optimism measured by the LOT scale.

--
   regression models. The results (Table 4) show the estimated regression
   coefficients and associated p-values. In the left column, each variable
   is considered one at a time, controlling only for baseline placebo
   Tylenol pain threshold and the session number in which placebo Tylenol
   was received. In the right column, all variables appear together in a
   single multivariate model. In both settings, there is no significant
   association between post-treatment placebo Tylenol pain threshold and
   the sham acupuncture pain threshold change, suggesting the 2 placebo
   modalities are not associated. No other predictors were found to be
   significantly related to placebo Tylenol pain threshold in either
   univariate or multivariate models.
   thumbnail
--
       original image ()

   Table 4. Potential predictors of placebo Tylenol response using pain
   threshold as outcome.
   doi:10.1371/journal.pone.0067485.t004
--
   threshold as well as the session number in which electroacupuncture was
   received: pre- and post- sham acupuncture pain threshold change, pre-
   and post-placebo Tylenol pain threshold change, pre- and post- rest
   control pain threshold change, expectancy ratings for each of the 4
   conditions(genuine acupuncture, sham acupuncture, placebo Tylenol, and
   rest control), age, gender, and optimism measures using the LOT scale.

--
   for each of the 4 experimental conditions are: 4.4±0.3 for the genuine
   acupuncture condition, 4.1±0.4 for the sham acupuncture condition,
   5.3±0.3 for the placebo Tylenol condition, and 0.6±0.2 for the rest
   control condition. The ANOVA showed that there was a significant
   difference in expectancy ratings across the 4 experimental conditions
--
   expectancy rating in all 3 treatment groups were significantly greater
   than the rest control condition (p<0.001 for each). Expectancy ratings
   in the placebo Tylenol condition were also greater than both the sham
   acupuncture condition (p = 0.001) and the genuine acupuncture condition
   (p = 0.003).
--

   In this study, we investigated the analgesic effects produced by
   placebo Tylenol, sham acupuncture, genuine acupuncture and a rest
   control condition, as well as the association between the effects of
   verbal suggestion on evoked placebo treatments (placebo Tylenol and
   sham acupuncture), electroacupuncture, and conditioning cue effects.
   The results showed that genuine acupuncture and placebo pills could
   significantly increase subjects' pain threshold compared with rest
   control condition. Regression analysis showed that there were no
   significant associations between individuals' responses to placebo
   pills, sham acupuncture, electroacupuncture and conditioning cue
   effects; however, subjects' responses to sham acupuncture correlated
--

   To the best of our knowledge, the question of whether individuals are
   likely to respond to different placebo routes of administration in a
   similar or different manner was an explicit topic of research in
   several experiments performed between 1956 and 1965 and each concluded
   that sham injections were more powerful than placebo pills
   [39]–[42].Further supporting this belief, a relatively old
   meta-analysis that included 35 RCTs for treatment of acute migraine
   extracted the placebo responses of trials that utilized an oral placebo
   and compared those responses with placebo responses in trials that used
   an injected placebo. The analysis showed that the injected placebo was
   statistically and clinically superior to the oral placebo [43].

   Our team performed an RCT comparing placebo pills and sham acupuncture
   across 270 patients with chronic arm pain. The results showed that sham
   acupuncture was statistically and clinically superior to placebo pills
   over time, while placebo pills were more beneficial in the short-term
   for improving sleep disturbances due to the pain condition [5]. A
   recent Cochrane meta-analysis of placebo effects [44] revisited this
   question of pill placebos versus device placebos and found that RCTs (n
   = 61) with “physical placebos” (which included sham acupuncture and
   such devices as sham electrotherapy and ultrasound) produced larger
   placebo effects than studies with “pharmacological” pill controls. In a
   separate study, Linde and colleagues [10] re-analyzed this
   meta-analysis [44] to investigate whether effects associated with sham
   acupuncture differed from those of other 'physical placebos'. They
   found pooled standardized mean differences were −0.41 (95% CI(−0.56,
   −0.24)) between sham acupuncture and no treatment, and −0.26 (95% CI
   −0.37, −0.15) between other physical placebos and no treatment,
   implying a larger effect of acupuncture treatment. In another
   meta-analysis from the same group [8], the authors also found that sham
--

   In our present study, we did not find any differences between the sham
   acupuncture treatment and placebo pills in the measurement of pain
   threshold, tolerance, and pain ratings. Additionally, there was no
   significant association between any of the pain measurement changes
   after the 2 modes of placebo treatments (placebo pill and sham
   acupuncture) in both univariate analysis and after adjusting for
   expectancy and optimism, suggesting a lack of consistency in placebo
   response across different modes of placebo treatments. This result is
   consistent with the findings from a recent study, in which Whalley and
   colleagues [45] found that placebo effects across trials were highly
   correlated when placebo creams bore the same name but were not
   significantly correlated when placebo creams had different names.

   We did not find any significant sham acupuncture effects compared to
   the rest control. Additionally, analgesic effects across all treatment
   modes were, in general, mild. This is consistent with previous findings
   that indicate that placebo effects evoked by verbal suggestion alone
   are weak in healthy subjects compared to patients [6], [46], [47],
   particularly considering that acupuncture treatment is a novel
   treatment method in the United States [15].The fact that sham
   acupuncture had significantly lower expectancy ratings than placebo
   Tylenol further supports this view.

--
   two are necessarily the same. We found that although the explicit
   conscious expectancy for genuine acupuncture is lower than the
   expectancy for placebo pills, the analgesic effect (as measured by pain
   threshold and tolerance) produced by genuine electroacupuncture trends
   is higher than the effect produced by placebo pills. This dissociation
   suggests that the specific effect of acupuncture may significantly
   contribute to the analgesic effect of electroacupuncture, which is
   consistent with previous neuroimaging studies that found different
   brain activation patterns associated with placebo and acupuncture
   analgesia [16], [17].

--
   alone, and conditioning at the level of N1 and N2–P2 components of CO2
   laser-evoked potentials (LEPs) and subjective pain reports. They found
   that both verbal suggestions (placebo cream) and conditioning modified
   the N2–P2 complex. Verbal suggestion in combination with the
   application of placebo cream could not produce subjective pain rating
   differences as compared with control condition, while conditioning with
   placebo cream produced more robust reductions of LEP amplitudes and
   subjective pain rating decreases. Our results are consistent with their
   findings of weak placebo effects with verbal suggestion alone and
   robust effects using a conditioning paradigm. Although we cannot
   perform a direct comparison, in our results the p-value of the
   conditioned visual cue effect was more significant compared to the
   placebo effect observed with verbal suggestion alone. More importantly,
   our results extend their findings showing that there is no association
   between the placebo analgesia observed with verbal suggestion and the
   effects of visual cue conditioning, suggesting that each may be
   associated with different mechanisms.
--
   Although our study is based on an experimental model in healthy
   volunteers, it may have implications for clinical care and research.
   Until now it has been understoodthat aggregate placebo responses in
   RCTs and experiments are variable [50]. Our study may suggest a new
   dimension of variability in placebo effects: responses may differ
   within individual subjects according to route of administration (pills
   vs. sham acupuncture) as well as environmental cues and learning
   processes (e.g., verbal suggestion and conditioning). This finding
   implies that placebo responses may not be dependent on stable
   individual traits but rather are more a characteristic of the state
   circumstances of individuals or combination of both trait and state.
   This result may help explain the difficulty of detecting reliable and
   consistent placebo responders [51]. Furthermore, the fact that people
   have unique responses to different routes of administration suggests
   that, for some, pills may work better than injections or vice versa.
   Moreover, the significant correlation between genuine and sham
   acupuncture on subjective outcomes suggests that elucidating whether
   acupuncture is more than a placebo effect will be difficult.

   The potential limitations of the study include the following: 1) the
--
   should adjust for any residual effect from a previous treatment as well
   as other potential confounding factors, such as hyperalgesia due to
   repeated pain application; 2) The placebo effect of sham acupuncture
   was not statistically superior to no-treatment control. It is possible
   that had we had a larger placebo effect, we might have found a
   correlation between pill and sham device; 3) The study was performed on
   healthy subjects in a relatively small sample size and thus it is
--

   In summary, in this crossover study, we found that individuals respond
   differently to different types of verbally suggested placebo treatments
   and cue conditioning. In addition, we found a significant association
   between genuine and sham acupuncture treatments, which implies that
--
References

    1. 1. Shapiro AK, Shapiro E (1997) The Powerful Placebo: From Ancient
       Priest to Modern Physician. Baltimore: The Johns Hopkins University
       Press.
    2. 2. Kaptchuk TJ (2011) Placebo studies and ritual theory: a
       comparative analysis of Navajo, acupuncture and biomedical healing.
       Philos Trans R Soc Lond B Biol Sci 366: 1849–1858. doi:
--
          + Google Scholar
    3. 3. Brody H, Miller FG (2011) Lessons from recent research about the
       placebo effect – from art to science. Jama 306: 2612–2613. doi:
       10.1001/jama.2011.1850
          + View Article
--
          + Google Scholar
    4. 4. Finniss DG, Kaptchuk TJ, Miller F, Benedetti F (2010)
       Biological, clinical, and ethical advances of placebo effects.
       Lancet 375: 686–695. doi: 10.1016/S0140-6736(09)61706-2
          + View Article
--
    5. 5. Kaptchuk TJ, Stason WB, Davis RB, Legedza AR, Schnyer RN, et al.
       (2006) Sham device v inert pill: randomised controlled trial of two
       placebo treatments. Bmj 332: 391–397. doi:
       10.1136/bmj.38726.603310.55
          + View Article
--
    6. 6. Colloca L, Tinazzi M, Recchia S, Le Pera D, Fiaschi A, et al.
       (2008) Learning potentiates neurophysiological and behavioral
       placebo analgesic responses. Pain 139: 306–314. doi:
       10.1016/j.pain.2008.04.021
          + View Article
--
          + Google Scholar
    9. 9. Kaptchuk TJ, Goldman P, Stone DA, Stason WB (2000) Do medical
       devices have enhanced placebo effects? J Clin Epidemiol 53:
       786–792. doi: 10.1016/S0895-4356(00)00206-7
          + View Article
--
          + Google Scholar
   10. 10. Linde K, Niemann K, Meissner K (2010) Are sham acupuncture
       interventions more effective than (other) placebos? A re-analysis
       of data from the Cochrane review on placebo effects. Forsch
       Komplementmed 17: 259–264. doi: 10.1159/000320374
          + View Article
--
          + Google Scholar
   15. 15. Kong J, Gollub RL, Rosman IS, Webb JM, Vangel MG, et al. (2006)
       Brain activity associated with expectancy-enhanced placebo
       analgesia as measured by functional magnetic resonance imaging. J
       Neurosci 26: 381–388. doi: 10.1523/JNEUROSCI.3556-05.2006
--
   16. 16. Kong J, Kaptachuk TJ, Polich G, Kirsch IV, Angel M, et al..
       (2009) Expectancy and treatment interactions: A dissociation
       between acupuncture analgesia and expectancy evoked placebo
       analgesia. Neuroimage 45: 940–949 PMID: 19159691
   17. 17. Kong J, Kaptchuk TJ, Polich G, Kirsch I, Vangel M, et al..
--
       47: 1066–1076 PMID: 19501656.
   18. 18. Geers AL, Helfer SG, Kosbab K, Weiland PE, Landry SJ (2005)
       Reconsidering the role of personality in placebo effects:
       dispositional optimism, situational expectations, and the placebo
       response. J Psychosom Res 58: 121–127. doi:
       10.1016/j.jpsychores.2004.08.011
--
          + Google Scholar
   19. 19. Geers AL, Wellman JA, Fowler SL, Helfer SG, France CR (2010)
       Dispositional optimism predicts placebo analgesia. J Pain 11:
       1165–1171. doi: 10.1016/j.jpain.2010.02.014
          + View Article
--
          + Google Scholar
   20. 20. Morton DL, Watson A, El-Deredy W, Jones AK (2009)
       Reproducibility of placebo analgesia: Effect of dispositional
       optimism. Pain 146: 194–198. doi: 10.1016/j.pain.2009.07.026
          + View Article
--
          + Google Scholar
   21. 21. Voudouris NJ, Peck CL, Coleman G (1990) The role of
       conditioning and verbal expectancy in the placebo response. Pain
       43: 121–128. doi: 10.1016/0304-3959(90)90057-K
          + View Article
--
          + Google Scholar
   22. 22. Montgomery GH, Kirsch I (1997) Classical conditioning and the
       placebo effect. Pain 72: 107–113. doi:
       10.1016/S0304-3959(97)00016-X
          + View Article
--
   23. 23. Price DD, Milling LS, Kirsch I, Duff A, Montgomery GH, et al.
       (1999) An analysis of factors that contribute to the magnitude of
       placebo analgesia in an experimental paradigm. Pain 83: 147–156.
       doi: 10.1016/S0304-3959(99)00081-0
          + View Article
--
          + Google Scholar
   24. 24. De Pascalis V, Chiaradia C, Carotenuto E (2002) The
       contribution of suggestibility and expectation to placebo analgesia
       phenomenon in an experimental setting. Pain 96: 393–402. doi:
       10.1016/S0304-3959(01)00485-7
--
          + Google Scholar
   25. 25. Wager TD, Rilling JK, Smith EE, Sokolik A, Casey KL, et al.
       (2004) Placebo-induced changes in FMRI in the anticipation and
       experience of pain. Science 303: 1162–1167. doi:
       10.1126/science.1093065
--
       13354–13362 PMCID: PMC2649754.
   27. 27. Bingel U, Lorenz J, Schoell E, Weiller C, Buchel C (2006)
       Mechanisms of placebo analgesia: rACC recruitment of a subcortical
       antinociceptive network. Pain 120: 8–15. doi:
       10.1016/j.pain.2005.08.027
--
   29. 29. Kleinhenz J, Streitberger K, Windeler J, Gussbacher A, Mavridis
       G, et al. (1999) Randomised clinical trial comparing the effects of
       acupuncture and a newly designed placebo needle in rotator cuff
       tendinitis. Pain 83: 235–241. doi: 10.1016/S0304-3959(99)00107-4
          + View Article
          + PubMed/NCBI
          + Google Scholar
   30. 30. Streitberger K, Kleinhenz J (1998) Introducing a placebo needle
       into acupuncture research. Lancet 352: 364–365. doi:
       10.1016/S0140-6736(97)10471-8
--
          + PubMed/NCBI
          + Google Scholar
   31. 31. White P, Lewith G, Hopwood V, Prescott P (2003) The placebo
       needle, is it a valid and convincing placebo for use in acupuncture
       trials? A randomised, single-blind, cross-over pilot trial. Pain
       106: 401–409. doi: 10.1016/j.pain.2003.08.013
--
          + Google Scholar
   32. 32. Wasan AD, Kong J, Pham LD, Kaptchuk TJ, Edwards R, et al.
       (2010) The Impact of Placebo, Psychopathology, and Expectations on
       the Response to Acupuncture Needling in Patients With Chronic Low
       Back Pain. J Pain 11: 555–563. doi: 10.1016/j.jpain.2009.09.013
--
          + PubMed/NCBI
          + Google Scholar
   40. 40. Traut EF, Passarelli EW (1957) Placebos in the treatment of
       rheumatoid arthritis and other rheumatic conditions. Ann Rheum Dis
       16: 18–22. doi: 10.1136/ard.16.1.18
--
          + PubMed/NCBI
          + Google Scholar
   43. 43. de Craen AJ, Tijssen JG, de Gans J, Kleijnen J (2000) Placebo
       effect in the acute treatment of migraine: subcutaneous placebos
       are better than oral placebos. J Neurol 247: 183–188. doi:
       10.1007/s004150050560
          + View Article
          + PubMed/NCBI
          + Google Scholar
   44. 44. Hrobjartsson A, Gotzsche PC (2010) Placebo interventions for
       all clinical conditions. Cochrane Database Syst Rev: CD003974.
   45. 45. Whalley B, Hyland ME, Kirsch I (2008) Consistency of the
       placebo effect. J Psychosom Res 64: 537–541. doi:
       10.1016/j.jpsychores.2007.11.007
          + View Article
--
          + Google Scholar
   47. 47. Charron J, Rainville P, Marchand S (2006) Direct comparison of
       placebo effects on clinical and experimental pain. Clin J Pain 22:
       204–211. doi: 10.1097/01.ajp.0000161526.25374.e5
          + View Article
--
          + PubMed/NCBI
          + Google Scholar
   50. 50. Hrobjartsson A, Gotzsche PC (2001) Is the placebo powerless? An
       analysis of clinical trials comparing placebo with no treatment. N
       Engl J Med 344: 1594–1602. doi: 10.1056/NEJM200105243442106
          + View Article
--
          + Google Scholar
   51. 51. Kaptchuk TJ, Kelley JM, Deykin A, Wayne PM, Lasagna LC, et al.
       (2008) Do “placebo responders” exist? Contemp Clin Trials 29:
       587–595. doi: 10.1016/j.cct.2008.02.002
          + View Article