Treatment expectations and clinical outcomes following repetitive transcranial magnetic stimulation for treatment-resistant depression.

BACKGROUND: Patient expectations, including both positive (placebo) and negative (nocebo) effects, influence treatment outcomes, yet their impact on acute repetitive transcranial magnetic stimulation (rTMS) for treatment-resistant depression (TRD) is unclear. METHODS: In this single-center retrospective chart review, 208 TRD patients completed the Stanford Expectation of Treatment Scale (SETS) before starting open-label rTMS treatment. Patients were offered two excitatory rTMS protocols (deep TMS or intermittent theta-burst stimulation), which stimulated the left dorsolateral prefrontal cortex. A minimum of 20 once daily treatments were provided, delivered over 4-6 weeks. Primary outcomes were 1) remission, measured by a post-treatment score of <8 on the Hamilton Depression Rating Scale (HAMD-17), and 2) premature discontinuation. The change in HAMD-17 scores over time was used as a secondary outcome. Physicians were blinded to SETS scores. Logistic and linear regression, adjusting for covariates, assessed SETS and HAMD-17 relationships. RESULTS: Of 208 patients, 177 had baseline and covariate data available. The mean positivity bias score (positive expectancy minus negative expectancy subscale averages) was 0.48 ± 2.21, indicating the cohort was neutral regarding the expectations of their treatment on average. Higher positive expectancy scores were significantly associated with greater odds of remission (OR = 1.90, p = 0.003) and greater reduction in HAMD-17 scores (ß = 1.30, p = 0.005) at the end of acute treatment, after adjusting for covariates. Negative expectancy was not associated with decreased odds of remission (p = 0.2) or treatment discontinuation (p = 0.8). CONCLUSIONS: Higher pre-treatment positive expectations were associated with greater remission rates with open-label rTMS in a naturalistic cohort of patients with TRD.

Placebo effects of repetitive transcranial magnetic stimulation on negative symptoms and cognition in patients with schizophrenia spectrum disorders: a systematic review and meta-analysis.

Background: Negative symptoms and cognitive impairments are highly frequent in schizophrenia spectrum disorders (SSD), associated with adverse functional outcomes and quality of life. Repetitive transcranial magnetic stimulation (rTMS) has been considered a promising therapeutic option in SSD. However, placebo effects of rTMS on these symptoms remained unclear. Objective: To investigate placebo effects of rTMS on alleviating negative symptoms and cognitive impairment in patients with SSD and to explore potential moderators. Methods: We systematically searched five electronic databases up to 15 July 2023. Randomized, double-blind, sham-controlled trials investigating effects of rTMS on negative symptoms or cognition in patients with SSD were included. The pooled placebo effect sizes, represented by Hedges' g, were estimated using the random-effects model. Potential moderators were explored through subgroup analysis and meta-regression. Results: Forty-four randomized controlled trials with 961 patients (mean age 37.53 years; 28.1% female) in the sham group were included. Significant low-to-moderate pooled placebo effect sizes were observed for negative symptoms (g=0.44, p<0.001), memory (g=0.31, p=0.010), executive function (g=0.35, p<0.001), working memory (g=0.26, p=0.004), and processing speed (g=0.36, p=0.004). Subgroup analysis indicated that placebo effects were affected by sham stimulation methods, rTMS targeting approaches, and stimulation frequency. Conclusions: Placebo effects of rTMS on negative symptoms and cognition in patients with SSD are significant in a small-to-moderate magnitude, which might be mediated by rTMS parameters. Our findings will provide new insights for practitioners to further optimize and establish standardized rTMS protocols for future RCTs tackling cardinal symptoms in SSD. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/, identifier CRD42023390138.

Boosting treatment outcomes via the patient-practitioner relationship, treatment-beliefs or therapeutic setting. A systematic review with meta-analysis of contextual effects in chronic musculoskeletal pain.

OBJECTIVE: To ascertain whether manipulating contextual effects (e.g. interaction with patients, or beliefs about treatments) boosted the outcomes of non-pharmacological and non-surgicaltreatments for chronic primary musculoskeletal pain. DESIGN: Systematic review of randomized controlled trials. DATA SOURCES: We searched for trials in six databases, citation tracking, and clinical trials registers. We included trials that compared treatments with enhanced contextual effects with the same treatments without enhancement in adults with chronic primary musculoskeletal pain. DATA SYNTHESIS: The outcomes of interest were pain intensity, physical functioning, global ratings of improvement, quality of life, depression, anxiety, and sleep. We evaluated risk of bias and certainty of the evidence using Cochrane Risk of Bias tool 2.0 and the GRADE approach, respectively. RESULTS: Of 17637 records, we included 10 trials with 990 participants and identified 5 ongoing trials. The treatments were acupuncture, education, exercise training, and physical therapy. The contextual effects that were improved in the enhanced treatments were patient-practitioner relationship, patient beliefs and characteristics, therapeutic setting/environment, and treatment characteristics. Our analysis showed that improving contextual effects in non-pharmacological and non-surgical treatments may not make much difference on pain intensity (mean difference [MD] : -1.77, 95%-CI: [-8.71; 5.16], k = 7 trials, N = 719 participants, Scale: 0-100, GRADE: Low)) or physical functioning (MD: -0.27, 95%-CI: [-1.02; 0.49], 95%-PI: [-2.04; 1.51], k = 6 , N = 567, Scale: 0-10, GRADE: Low) in the short-term and at later follow-ups. Sensitivity analyses revealed similar findings. CONCLUSION: Whilst evidence gaps exist, per current evidence it may not be possible to achieve meaningful benefit for patients with chronic musculoskeletal pain by manipulating the context of non-pharmacological and non-surgical treatments. TRIAL REGISTRATION: This systematic review was prospectively registered in PROSPERO (registration number: CRD42023391601).

The Effect of Singular Nonverbal Behaviours of Experimenters on Pain Reports.

Introduction: Studies suggest facial expressions of caregivers may be important in placebo effects; however, this has not been systematically tested. This experiment investigated the effects of caregivers' singular positive nonverbal behaviours (NBs) on pain reports. Methods: Fifty-one males and 53 females (total of 104) participants were randomized to four groups that were displayed positive facial expressions, tone of voice, body movement, or neutral NBs of videotaped experimenters. Subjective reports of pain, stress, arousal, and cardiac activity were obtained in a pre-test, a conditioning phase, and at a post-test. Four minutes of heat pain was induced in each test, and a placebo cream was administered before the conditioning and post-test in all groups. Results: There were no differences between the NB groups in the reduced pain. Males had larger reduction in pain in the post-test, and females had lower arousal than the opposite sex. During the conditioning, females had larger reduction in pain ie, unconditioned pain response (UPR). In females, the UPR predicted the reinforced expectation ie, increase in expectations from conditioning to post-test, and fear of minor pain negatively predicted both the UPR and reinforced expectation. Discussion: Singular NBs of caregiver were weak to enhance placebo effects. Females had lower pain during conditioning, and the UPR amplitude in females was associated with positive expectations. Moreover, for females, fear of minor pain weakened the UPR and expectations of cream. Conclusion: No NB of caregivers is more effective in reducing pain. Caregivers' NBs are less effective when displayed individually. Males and females may be different in underlying mechanisms of placebo effects.

Harnessing Placebo Effects for the Treatment of Functional Cognitive Disorder: A Feasibility Pilot Study.

OBJECTIVE: Limited research has directly investigated whether and how placebo effects can be harnessed for the treatment of functional neurological disorder (FND), despite a long-standing and controversial history of interest in this area. METHODS: A small exploratory study was conducted with adults with a cognitive subtype of FND recruited from a single cognitive neurology center in the United States. Participants were given the expectation of receiving cranial stimulation that could benefit their memory symptoms; however, the intervention was sham transcranial magnetic stimulation (placebo). Outcomes included measures of short-term memory testing, subjective memory rating, and state anxiety before and after stimulation. After the study, the true objective and rationale for investigating placebo effects were explained in a scripted debriefing session. Acceptability of the study design and qualitative feedback were collected. Institutional ethics approval and signed consent were obtained. RESULTS: Three patients (female, N=2; male, N=1; average age=57 years) were recruited. Outcome data were analyzed descriptively at the patient level. Trends of improvement in subjective memory rating, but not objective cognitive test scores, and decreases in state anxiety were observed. After the debriefing session, all patients found the study design to be acceptable (ratings of 70%, 90%, and 100%), and two of the three patients believed that withholding mechanistic information about the intervention was needed to leverage placebo effects as treatment. CONCLUSIONS: In the first study to prospectively investigate the feasibility of harnessing placebo effects for the treatment of FND, promising preliminary findings were obtained, and methods and resources for use in larger future studies are offered.

Contextual effects: how to, and how not to, quantify them.

The importance of contextual effects and their roles in clinical care controversial. A Cochrane review published in 2010 concluded that placebo interventions lack important clinical effects overall, but that placebo interventions can influence patient-reported outcomes such as pain and nausea. However, systematic reviews published after 2010 estimated greater contextual effects than the Cochrane review, which stems from the inappropriate methods employed to quantify contextual effects. The effects of medical interventions (i.e., the total treatment effect) can be divided into three components: specific, contextual, and non-specific. We propose that the most effective method for quantifying the magnitude of contextual effects is to calculate the difference in outcome measures between a group treated with placebo and a non-treated control group. Here, we show that other methods, such as solely using the placebo control arm or calculation of a 'proportional contextual effect,' are limited and should not be applied. The aim of this study is to provide clear guidance on best practices for estimating contextual effects in clinical research.

Can Pharmacological Conditioning as an Add-On Treatment Optimize Standard Pharmacological Treatment in Patients with Recent-Onset Rheumatoid Arthritis? A Proof-of-Principle Randomized Clinical Trial.

Medication regimens using conditioning via variable reinforcement have shown similar or improved therapeutic effects as full pharmacological treatment, but evidence in patient populations is scarce. This proof-of-principle double-blind randomized clinical trial examined whether treatment effects in recent-onset rheumatoid arthritis (RA) can be optimized through pharmacological conditioning. After four months of standardized treatment (n = 46), patients in clinical remission (n = 19) were randomized to the Control group (C), continuing standardized treatment (n = 8), or the Pharmacological Conditioning (PC) group, receiving variable treatment according to conditioning principles (n = 11). After eight months, treatment was tapered and discontinued linearly (C) or variably (PC). Standard treatment led to large improvements in disease activity and HRQoL in both groups. The groups did not differ in the percentage of drug-free clinical remission obtained after conditioning or continued standard treatment. The PC group did show a larger decrease in self-reported disease activity (Cohen's d = 0.9) and a smaller increase in TNF-α levels (Cohen's d = 0.7) than the C group. During all phases, more differences between groups were found for the patients who followed protocol than for the intention-to-treat sample. Although the results are not conclusive, pharmacological conditioning may have some advantages in terms of disease progression and stability, especially during the conditioning phase, compared with standard clinical treatment. The effects may be particularly beneficial for patients who show a good initial response to increased medication dosages.

Expectancy for Adderall influences subjective mood and drug effects regardless of concurrent caffeine ingestion: A randomized controlled trial.

RATIONALE: Nonmedical prescription stimulant use (NPS; use without a prescription or in ways other than prescribed) is common among college students. Despite the potential for negative consequences, students continue engaging in NPS for cognitive enhancement purposes, which may be maintained by expectancy and placebo effects. OBJECTIVES: This study examined if a placebo administered under the guise of Adderall influenced subjective mood/drug effects and cognitive performance. Furthermore, this study examined if concurrent caffeine ingestion incrementally enhanced Adderall-related placebo effects. METHODS: Undergraduate students with features that put them at elevated risk for NPS (N = 121) completed measures of mood and drug effects and cognitive assessments on two separate laboratory visits in this parallel randomized controlled trial. Visit 1 was a baseline control visit, on which no drug was expected or received. On visit 2, subjects were randomized to: (1) expect/receive no drug (control); (2) expect Adderall/receive placebo; or (3) expect Adderall/receive 200 mg caffeine. RESULTS: There were several significant condition × visit interactions for subjective effects, including amphetamine effects, energy and efficiency effects, and feeling high. In most cases, participants who expected Adderall reported greater positive subjective effects on visit 2 compared to controls; however, there were generally not incremental enhancements for those ingesting caffeine compared to placebo. There were no significant effects for any cognitive tests. CONCLUSIONS: Expectation for prescription stimulant effects influenced subjective outcomes in a sample of high-risk college students. These findings may inform expectancy challenge interventions to reduce NPS. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03648684.

Placebo Hypoalgesia and Nocebo Hyperalgesia Induced by Observational Learning May Be Difficult to Disentangle in a Laboratory Setting.

Observational learning (OBL) (seeing pain/pain treatment in others) can evoke placebo hypoalgesia and nocebo hyperalgesia. Data that compare these effects and illuminates the role of expectations and empathy are scarce. Healthy participants (n = 105) were randomized to: 1) placebo OBL, 2) nocebo OBL, or 3) no-observation control group. OBL consisted of a model simulating pain relief or increase after a sham ointment was applied to one arm. Pain was evoked with thermal stimuli on both arms (ointment, contralateral) at baseline and postobservation. Expectations, pain ratings, and physiological data (eg, skin conductance level) were collected. A 3 × 2 × 2 (Group × Arm × Phase) mixed analyses of variance revealed a 3-way interaction that confirmed that OBL modulates pain: F(2, 93) = 6.08, P = .003, ηp2 = .12. Significant baseline-to-post-observation pain increases were shown in the nocebo OBL group, with a bigger increase for the arm with ointment (both P ≤ .007). In the placebo OBL group, pain was higher for the contralateral relative to the ointment arm (P < .001). Baseline-to-post-observation pain increase was significant for the contralateral arm (P < .001). Expectation mediated these effects. Skin conductance level decreased over time during ointment trials in the nocebo OBL group, suggesting reduced physiological arousal. The findings illustrate that OBL modulates pain through expectations. In the placebo OBL group, the pain did not decrease for the ointment but increased for the contralateral stimuli, which may reflect nocebo learning. Experimental OBL paradigms typically examine relative differences between ointment and contralateral cues. This can complicate disentangling placebo hypoalgesia and nocebo hyperalgesia in laboratory settings. Implications for existing theories are discussed. PERSPECTIVE: Data that systematically compare placebo hypoalgesia and nocebo hyperalgesia induced by OBL are scarce. The current work illustrates that these effects may be more difficult to disentangle than previously assumed, which could have implications for existing theories on OBL and placebo effects and their translation to clinical practice.

Systematic manipulation of experimenters' non-verbal behaviors for the investigation of pain reports and placebo effects.

Objective: Non-verbal behaviors (NBs) of caregivers affect pain reports and placebo effects. However, little experimental research has systematically examined the caregivers' NBs. This study protocol and preparatory study report a systematic manipulation of experimenters' NBs to investigate pain report and placebo effects. Methods: We propose an experiment in which videotaped experimenters (VEs) conduct a pain stimulation and a placebo treatment study. The VEs express one positively enhanced NB and keep the other NBs neutral. Participants will be randomized to either the positive facial expressions (+FE), tone of voice (+TV), body movement (+BM), or neutral NBs (i.e., neutral condition; NC) of the VEs. As a preparatory study for proof of concept, two groups of NB coders from Norway and the USA separately rated the degree of NBs (eye contact, body postures and movements, and tone of voice), and impressions of dominance and being in charge, positivity, and expressivity from each NB video. The NB videos had construct validity and reliability. The +BM and +FE were rated as more dominant and in charge than the +TV and the NC. The +FE and +BM were rated as the most positive and expressive NBs, respectively. Expected results: +FE will have the largest placebo effects on pain and stress levels. However, transmitting the NBs to patients by VEs is challenging. Moreover, controlling for the effects of research assistants present in the testing room is challenging. Discussion: We propose that caregivers' NBs affect pain reports and placebo effects. Moreover, different NBs elicit different impressions, and a better understanding of the role of caregiver NBs requires more rigorous investigations. Lastly, aiming to investigate the caregiver NBs, the varying degrees of micro-NBs and their effects on the formation of impressions should be considered.

The Synchrony-Prosociality Link Cannot Be Explained Away as Expectancy Effect: Response to Atwood et al. ().

Moving in time to others, as is often observed in dance, music, sports and much of children's play cross-culturally, is thought to make people feel and act more prosocially towards each other. In a recent paper, Atwood et al. (2022) argued that the inferential validity of this link found between synchronous behaviour and prosociality might be mainly due to "expectancy effects generated by a combination of (1) experimenter expectancy, leading to experimenter bias; and (2) participant expectancy (i.e., placebo effects)". Here, we counter these arguments with (1) examples of studies devoid of experimenter expectancy effects that nevertheless demonstrate a positive link between synchrony and prosociality, and (2) insights from the developmental literature that address participant expectancy by showing how expectations formed through lived experiences of synchronous interactions do not necessarily threaten inferential validity. In conclusion, there is already sufficient good-quality evidence showing the positive effects of synchronous behaviours on prosociality beyond what can be explained by experimenter or participant expectation effects.

The influence of psychological traits and prior experience on treatment expectations.

BACKGROUND: Placebo and nocebo responses are modulated by the treatment expectations of participants and patients. However, interindividual differences predicting treatment expectations and placebo responses are unclear. In this large-scale pooled analysis, we aim to investigate the influence of psychological traits and prior experiences on treatment expectations. METHODS: This paper analyses data from six different placebo studies (total n = 748). In all studies, participants' sociodemographic information, treatment expectations and prior treatment experiences and traits relating to stress, somatization, depression and anxiety, the Big Five and behavioral inhibition and approach tendencies were assessed using the same established questionnaires. Correlation coefficients and structural equation models were calculated to investigate the relationship between trait variables and expectations. RESULTS: We found small positive correlations between side effect expectations and improvement expectations (r = 0.187), perceived stress (r = 0.154), somatization (r = 0.115), agitation (r = 0.108), anhedonia (r = 0.118), and dysthymia (r = 0.118). In the structural equation model previous experiences emerged as the strongest predictors of improvement (ß = 0.32, p = .005), worsening (ß = -0.24, p = .005) and side effect expectations (ß = 0.47, p = .005). Traits related to positive affect (ß = - 0.09; p = .007) and negative affect (ß = 0.04; p = .014) were associated with side effect expectations. DISCUSSION: This study is the first large analysis to investigate the relationship between traits, prior experiences and treatment expectations. Exploratory analyses indicate that experiences of symptom improvement are associated with improvement and worsening expectations, while previous negative experiences are only related to side effect expectations. Additionally, a proneness to experience negative affect may be a predictor for side effect expectation and thus mediate the occurrence of nocebo responses.

Joint evaluation of placebo and treatment effects in cluster randomized trials by causal inference models.

The term placebo effect refers to the psychobiological effect of a patient's knowledge or belief of being treated. A placebo effect is patient-driven, which makes it fundamentally different from the usual treatment effect resulting from external actions. In modern clinical research, the presence of a placebo effect is often treated as a nuisance issue, something to be "adjusted away" in estimating a treatment effect of primary interest. However, from a patient-centered perspective, we believe that a possible placebo produces substantial improvements in patient-centered outcomes. Understanding placebo effects is therefore an important part of patient-centered outcomes research. The available methods for estimating placebo effects are designed for individually randomized trials and are not directly applicable to cluster randomized trials (CRTs). There are several challenges in estimating placebo effects in CRTs. A major challenge is the possible presence of interference within clusters, in the sense that a subject's outcome may depend on the beliefs subjects in the same cluster about treatment assignment (mentality) and therefore possible correlation in outcome and mentality among subjects exists in the same cluster. In this article, we extend the previously developed causal inference framework to also encompass CRTs, using the G-Computation and inverse probability weighting (IPW) approaches. We also develop methodologies and further extend the G-Computation and IPW approaches to handle missingness for jointly evaluating placebo effect and treatment-specific effect, specifically in the context of CRTs. The proposed methods are demonstrated in simulation studies and a cluster randomized trial on effect of fermented dairy drink.

What are side effects?

Side effects are ubiquitous in medicine and they often play a role in treatment decisions for patients and clinicians alike. Philosophers and health researchers often use side effects to illustrate issues with contemporary medical research and practice. However, technical definitions of 'side effect' differ among health authorities. Thus, determining the side effects of an intervention can differ depending on whose definition we assume. Here I review some of the common definitions of side effect and highlight their issues. In response, I offer an account of side effects as jointly (i) unintended and (ii) effects due to the causal capacities or invariances of an intervention. I discuss (i) by examining the intentions or reasons behind therapeutic interventions, and I discuss (ii) by appealing to a manipulationist model of causation. The analysis here highlights that side effects are conceptually distinct from related outcomes like adverse events, adverse drug reactions, and placebo effects. The analysis also allows for reflection on the utility of 'side effect' as a technical term in medical research and practice.

The role of positive information provision in open-label placebo effects.

Open label placebos (OLPs) appear generally efficacious among clinical samples, but the empirical evidence regarding their use in non-clinical and sub-clinical samples, as well as when administered independent of a convincing rationale, is mixed. Healthy participants (N = 102) were randomised to either a 6-day course of OLP pills with information provision (OLP-plus: N = 35), without information provision (OLP-only: N = 35), or no-treatment control group (N = 32). OLP pills were described as enhancing physical (symptoms and sleep) and psychological (positive and negative emotional) well-being. Well-being was assessed at baseline and on Day 6. Expectancies and adherence were measured. OLP administration interacted with baseline well-being. The OLP-plus group demonstrated increased well-being on all outcomes other than positive emotions, but only when they reported decreased baseline well-being. OLP-only and control groups did not differ. The OLP-plus group demonstrated elevated expectancies, that mediated the OLP effect on physical symptoms relative to control, but only when well-being was lower than average at baseline (i.e. moderated-mediation). Results demonstrate the importance of information provided with OLPs. The moderating effect of baseline outcomes may reconcile inconsistent results regarding clinical and non-clinical samples. Accounting for baseline symptoms in non-clinical and sub-clinical samples is likely to enhance our understanding of when OLPs are effective.

Targeting neural correlates of placebo effects.

Harnessing the placebo effects would prompt critical ramifications for research and clinical practice. Noninvasive brain stimulation (NIBS) techniques, such as transcranial magnetic stimulation and multifocal transcranial electric stimulation, could manipulate the placebo response by modulating the activity and excitability of its neural correlates. To identify potential stimulation targets, we conducted a meta-analysis to investigate placebo-associated regions in healthy volunteers, including studies with emotional components and painful stimuli. Using biophysical modeling, we identified NIBS solutions to manipulate placebo effects by targeting either a single key region or multiple connected areas. Moving to a network-oriented approach, we then ran a quantitative network mapping analysis on the functional connectivity profile of clusters emerging from the meta-analysis. As a result, we suggest a multielectrode optimized montage engaging the connectivity patterns of placebo-associated functional brain networks. These NIBS solutions hope to provide a starting point to actively control, modulate or enhance placebo effects in future clinical studies and cognitive enhancement studies.

Expectations: How and when do they contribute to placebo analgesia?

In placebo research, expectations are highlighted as one of the most influential subjective factors. While some studies have shown a relationship between expectations and pain relief, others have not. However, little is known about how methods of assessment of expectations may affect these conclusions. One of the fundamental considerations is that participants in placebo trials rate their expectations when prompted to rate them on scales in advance, but are less likely to report their prior expectations, when asked to report their experience retroactively in an unprompted manner, often expressing, for example, prior hope or wishes of recovery. This article presents previously unpublished data to elucidate and explore the concepts highlighted by individuals in a placebo analgesia trial when assessed in a prompted and unprompted manner. The data corroborates the role of expectations involved in placebo effects, particularly in placebo analgesia. Thus, the question may be a matter of how and when expectations contribute to placebo effects, rather than if.

Can contagious itch be affected by positive and negative suggestions?

Contagious itch can be evoked by observing people scratching. Verbal suggestions about to-be-received itch can influence itch intensity, as shown by placebo research, but it is unknown whether this extends to contagious itch. The current study aimed to replicate prior findings that listening to scratching and rubbing sounds elicits contagious itch, and to investigate whether suggestions can modulate this process. Healthy participants (n = 140) received positive or negative suggestions about itch in response to the sounds (aimed to decrease or increase expected itch, respectively), or no specific suggestions as a control. Participants listened to a number of audio fragments with scratching and rubbing sounds. The amount of expected itch as well as itch sensation after each audio fragment were measured by self-report. Suggestions had no effect on the expected itch. Both rubbing and scratching sounds significantly elicited itch in all groups. Scratching sounds induced more itch than rubbing sounds exclusively in the control group. These findings indicate that short suggestions might be not effective enough to modify the expectations of people regarding contagious itch. Furthermore, suggestions modulate contagious itch to some degree, but not in the hypothesized direction. Potential similarities and differences in the neurobiological mechanisms of contagious itch and nocebo effects are discussed.