Abstract
Objective:
Patients with severe health anxiety have complex interpersonal relationships with medical providers and others in their social context, often resulting in conflictual interactions with providers and perception of poor medical care. An adequate understanding of the causes and consequences of these interactions is lacking, particularly 1 informed by the experience of the patient. This study used qualitative methods to explore the development and maintenance of health anxiety from the perspective of patients with lived experience of coping with health anxiety and their interactions with the healthcare system.
Method:
We conducted qualitative interviews with 11 primary care patients purposely sampled to describe their experience living with health anxiety, provider interactions, and social and family interactions surrounding health and health anxiety. We extracted themes related to living with health anxiety and interactions with providers and other significant relationships.
Results:
Thematic content analysis revealed 5 themes including 3 causal themes, 1 response theme, and 1 theme reflecting factors that mitigate health anxiety. Causal themes included subthemes reflecting predisposing factors of the self, key stimulus events from patient learning history, and maladaptive social interaction factors. The response theme was comprised of 2 subthemes: logical conclusions and health anxiety symptoms. The mitigating factors theme included subthemes of a trusting care relationship and recognition of disconfirming evidence.
Conclusion:
The themes and constituent subthemes identified in this study largely map onto cognitive-behavioral theory of health anxiety, demonstrating alignment between patient experience and theory. The resulting model also identifies potential points of intervention in the developmental and maintenance process. We provide recommendations to maximize those points of intervention.
Keywords
Introduction
Disorders of health anxiety, including illness anxiety disorder and somatic symptom disorder, are characterized by recurrent intrusive doubt and anxiety about the possibility of having or acquiring an illness and perseverating on the potential seriousness of the illness. 1 Individuals experiencing severe health anxiety commonly engage in behaviors aimed at eliminating the perceived illness threat or otherwise eliminating the anxious distress aroused by these fears. 2 Those behaviors, referred to here as safety behaviors, often consist of frequent visits to physicians and other healthcare providers to receive reassurance of safety and request repeated or additional testing. 3 Health anxious individuals may also engage in reassurance-seeking from other trusted individuals such as friends and family, or in the age of rapid access to information via the Internet, seek reassurance through online forums and health-focused websites (ie, “paging Dr. Google”). 4
The symptoms of health anxiety and efforts aimed at eliminating threat of illness interact with the medical system in profound and complex ways. Physical health symptoms do not always have a precise cause and an ambiguous diagnosis is common in primary care and medical subspecialties.5,6 A lack of certainty is often unacceptable to individuals suffering from health anxiety and becomes another source of potential threat, represented in the psychopathology literature as intolerance of uncertainty. 7 The attempts to eliminate uncertainty regarding health threats necessarily involves the healthcare system, usually through safety behaviors that involve interaction with healthcare professionals, including recitation of information gained through the use of health-related websites, requests for additional testing, and repeated communications with providers. 8
The interaction between healthcare professionals and patients who are confronting their own perceptions of health anxiety can become conflicting at times, with physicians and other healthcare providers often unable to alleviate the worries of a person experiencing health anxiety. 9 Providers may correctly identify that the patient does not have a significant underlying somatic illness contributing to the experienced symptoms, instead identifying that the excessive worry concerning their benign symptoms is contributing to the patient’s suffering. When providers correctly identify the problem as one of health anxiety and communicate this to the patient, there is potential for the patient to interpret this as an indication that their provider is not taking the problem seriously. This in turn can be interpreted by the patient to mean that their physical health complaints are “all in [the patient’s] head,” effectively minimizing patient concerns and assigning blame to the patient’s perspective.9,10 Prior work to understand the health anxious patient perspective supports the observation that these patients perceive their providers as invalidating their complaints and attributing the complaint to anxiety. 11
Physicians in turn endorse difficulty managing patients presenting with vague symptoms or those without a readily identifiable medical cause. 12 In a survey of primary care providers, only 25% of surveyed physicians reporting good or excellent ability in managing somatoform disorders, and only 14% feeling excellent or very good satisfaction in managing the patients experiencing them. Those providers who reported being most comfortable also endorsed being in a position to establish long-term close relationships with their patients. 10 This type of relationship is difficult to maintain in the context of high workload demands and physician burnout that results in less constructive communication styles and relationship formation. 13
Prior research aimed at understanding the development and maintenance of health anxiety has led to advances in the management of this problem, but is primarily focused on quantitative measurement of symptoms and related constructs. There are limited data exploring health anxiety from the patient perspective, particularly in relation to the intersection of health anxiety and the healthcare system. The addition of qualitative methods complements the findings of quantitative analyses, lending credibility to quantitative findings through comparison with the lived experience of patients of interest. 14 In the present study, we conducted a qualitative analysis of the patient perspective and experience with health anxiety as it intersects with primary care and extracted themes from responses to semi-structured patient interviews. The primary aim of this analysis is to characterize the factors that exacerbate health anxiety symptoms within the medical system, as well as factors that may mitigate suffering in this population. A secondary aim is to use this analysis to propose recommendations to guide care of health anxious persons and reduce the stigma these persons experience when interacting with the healthcare system. Finally, we present a model of the development and maintenance of health anxiety informed by patient experience and align this model with existing theory underlying cognitive and behavioral conceptualization of health anxiety.
Methods
Participants
We used purposeful random sampling 15 of adult patients with health anxiety currently in the care of a primary care team consisting of at least 1 primary care clinician (MD, DO, PA, or APRN) and a nursing support team (RNs, LPNs, and Medical Assistants). Participants were receiving their care from 1 of 3 regional primary care clinics that are part of a large Northeast U.S. healthcare system. The qualitative analysis described here was conducted within a larger randomized clinical trial focused on brief psychosocial intervention for health anxiety approved by the Institutional Review Board of the hospital system where the trial was conducted. Thus, all participants were participating in a clinical trial at the time of the interviews; however, all interviews were conducted prior to initiation of the intervention procedure that was the focus of that trial. The sample consisted of 11 patients with a diagnosis of either illness anxiety disorder (n = 10) or somatic symptom disorder (n = 1). Participants completed a diagnostic evaluation using the Health Anxiety Interview 16 adapted for DSM-5 as part of the primary study.
Procedures
The study team administered semi-structured qualitative interviews to a purposefully random sample of patients endorsing severe health anxiety to gain a better understanding of the patient-provider relationship for this subset of the primary care population. Our study team consisted of 3 interviewers, including a clinical psychologist, advanced medical student, and research coordinator with experience in administering qualitative interviews. The study interview (Supplemental 1) was developed by the first author (RB) to inquire about patient perspectives on (1) health and health anxiety as a construct and term, (2) provider interactions surrounding health and health anxiety, and (3) social and family interactions surrounding health and health anxiety. The average duration of the interviews was 30 min.
All interviews were professionally transcribed. Transcriptions were imported into Atlas.ti (Version 8, Scientific Software Development GmbH) for coding and further analysis. 17 Two coders (RB and AB) coded 2 interviews followed by collaborative review and determination of an initial codebook. Subsequent interviews were independently single coded, with coding audits performed by the first author to ensure coding consistency and iterative collaborative refinement to eliminate redundant codes. This occurred at approximately every second transcription coded. Interviews continued until saturation was achieved at which point no additional, non-redundant codes were identified after 2 consecutive interviews and all thematic elements were sufficiently developed to inform the resulting model. 18
Our analytic plan was an inductive thematic content analysis intended to produce a model of how patients’ experience of health anxiety relates to provider interactions in the healthcare setting and social and family interactions outside of the healthcare setting. 19 After initial coding of all interviews and agreed upon saturation of the data, we met again to conduct a final pruning of codes to ensure elimination of redundant or non-meaningful codes that were unrelated to the goals of the analysis. We used an axial coding process to identify themes and constituent subthemes that could inform a model of health anxiety and its presence in the patient-provider and larger social context. 20 We adhered to the SRQR checklist in the process of manuscript preparation. 21
Results
Twelve primary care patients participated in the qualitative interviews. Our inductive analysis yielded 39 codes collapsed into 5 themes comprised of 14 subthemes. These themes are presented in Table 1 with conceptual definitions of themes, subthemes, and representative quotes denoted by a participant number. These participant numbers are non-identifiable codes not related to their study identification number. The relation between themes along with the constituent subthemes is also presented in Figure 1. The following summary of the qualitative results includes the higher order themes comprised of subthemes drawn from the data.
Themes, Subthemes, Definitions, and Quotes Reflecting Health Anxious Patient Experience.
Hypothesized model depicting the axial relations between qualitative themes and constituent subthemes in the development and maintenance of health anxiety. Directional arrows indicate the proposed effects, with stimulus events leading to responses moderated by maladaptive social interaction factors and predisposing factors, and mitigating factors that reduce the intensity of health anxiety responses.
Theme 1—Predisposing Factors
The theme of Predisposing Factors included the subthemes of Negative Orientation Toward the Future, Lack of Medical Knowledge, and Characterological Features. The subtheme of Negative Orientation Toward the Future was conceptualized as a bias toward expecting the worst possible outcomes for health, even when this is unknown. Patients exhibiting Negative Orientation Toward the Future would interpret unknown outcomes in an automatically negative direction (eg, expecting negative outcome from ambiguous events). The Lack of Medical Knowledge subtheme reflects the understandable reality that patients do not have an adequate understanding of the medical causes of their health experience, due to not having formal education or training in medical sciences. The Characterological Features subtheme represents enduring patterns of behavior and cognition, such as interpretation biases and need for certainty related to health anxiety or other aspects of life. These subthemes together comprise a predisposition toward development of pathologic anxiety that, when combined with specific learning events, could result in health anxiety.
Theme 2—Stimulus Events
The theme of Stimulus Events included the subthemes of Negative Experiences with Health, Negative Experiences with Physicians, and Medicine Not Meeting Needs. Negative Experiences with Health includes patients’ experiences in which their health was threatened by a diagnosed medical illness (eg, prior cancer diagnosis, shingles) from which they have fully recovered. Negative Experiences with Physicians similarly includes experiences highlighting problematic interactions with physicians and other medical providers that created a negative perception of care. Examples include physicians appearing unconcerned about a symptom or making a perceived mistake, or appearing sarcastic in clinical encounters that led the participant to see them as a source of threat to health. Medicine Not Meeting Needs was conceptualized as the patients’ perception that medicine does not adequately address health needs as a function of structural deficits in the larger medical system. For example, multiple participants alluded to providers seeming overly busy and rushed, leading to a lack of attention to detail in the encounters, among other perceived issues in the medical system. Together this theme comprises the potent learning events that interact with predisposing factors to constitute a specific psychological vulnerability to health anxiety. 22
Theme 3—Maladaptive Social Interaction Factors
The Maladaptive Social Interaction Factors are those features of social behavior that tend to worsen interpersonal interactions with people in the health anxious patient’s social environment and contribute to a cyclical process of worsening health anxiety. These include Lack of Social Understanding, Failed Health Stewardship, Unreasonable Expectations, and Learned Mistrust. Lack of Social Understanding was conceptualized as difficulty expressing distress to others on the part of the patient, but also difficult of others to understand or empathize with the patient’s distress, as well as difficulty effectively expressing that distress to others. Failed Health Stewardship includes experiences in which participants believed that they were “falling through the cracks” and being failed by their providers or others who were not listening or taking complaints sufficiently seriously. This also encompassed instances where participants noted that they feared being thought of as a “hypochondriac” because of the concern that this would lead to more dismissal of concerning symptoms. Unreasonable Expectations is a conflict between the patient’s expectation of a certain level of care from the medical system and providers and the reality of what the system can provide. This results in a social interaction with providers in which the patient leaves the encounter thinking the provider has failed when the reality of the environment does not allow for the expected care. Finally, Learned Mistrust was a consequence of repeated instances of problematic encounters with providers resulting in a belief that the medical system could not be trusted. The Maladaptive Social Interaction Factors necessarily has some overlap with Stimulus Events, as it is these events that contribute to the interpersonal relationship challenges reported by the participants, but focuses on the relational aspects of the encounter.
Theme 4—Responses
The Responses theme includes the overt and covert actions taken by patients to address perceived health threats or to otherwise cope with health fears. This theme is comprised of Logical Conclusions and Health Anxiety Symptoms. Logical Conclusions is a predominantly covert, cognitive behavior, while Symptoms of Health Anxiety include covert and overt behaviors (eg, checking and reassurance seeking). Logical Conclusions is characterized by the use of a rational process to try to understand the cause of symptoms that leads to a misinterpretation of a sensation. For example,1 participant noted that the fact they previously had a confirmed symptom of illness meant that the next similar unexplained medical symptom would surely result in another confirmed medical illness. The theme of Health Anxiety Symptoms was identified as containing those statements demonstrating safety behaviors (eg, checking and reassurance-seeking), worry process, or catastrophic misinterpretation of benign symptoms that is emblematic of health anxiety.
Theme 5—Mitigating Factors
In contrast to the previous 4 themes, the Mitigating Factors theme included subthemes reflecting participants’ suggestions for how health anxiety could be reduced, primarily focused on interpersonal behaviors of others. This theme was comprised of the themes Communication and Trust Building and Providing Counter Evidence. Communication and Trust Building included experiences participants reported wherein a provider took additional time to listen to their patients’ complaints or engaged in a thoughtful conceptualization of their patients’ symptoms. Providing Counter Evidence included instances in which the participants had identified countervailing evidence that their interpretations of sensations or symptoms were irrational or expressed insight that the fear might simply be a function of health anxiety.
Summary of Model Structure
The resulting model demonstrating the relation of themes operates such that Stimulus Events, comprised of negative experiences with personal health, negative interactions with providers, and the experience of medicine not meeting the patient’s needs function as the learning history that fuels health anxiety. These background aversive experiences are subsequently strengthened by broader social interactions (eg, a lack of understanding from peers and family) and intraindividual predisposing factors. The culmination of these themes contributes to the health anxiety response. As indicated by the participants, providers are in a position to mitigate those responses through opportunities to provide evidence countering the worst fears of patients combined with forming the trusting relationship and strong communication skills of high-quality patient-provider relationships.
Discussion
The principal aim of this study was to utilize the lived experience of patients endorsing severe health anxiety to inform a model of the interaction between health anxiety and the medical system. Essentially, we sought to include the context in which health anxiety occurs to improve understanding of how health anxiety develops and is maintained within the healthcare system. Although psychological theory, including affective, cognitive, and behavioral processes, undoubtedly guides understanding of health anxiety, including the role of the healthcare system improves the conceptual model and provides opportunities for intervention at the system level. Additionally, the use of qualitative data bolsters the explanatory power of prior studies relying primarily on quantitative data to model health anxiety by reflecting these models in the lives of patients with these presenting concerns in their own words.
The qualitative analyses conducted for this study suggested 3 causal themes contributing to the development and maintenance of health anxiety,1 mitigating theme, and1 response theme. The causal themes are further differentiated as moderating and direct causal themes. The causal themes are Predisposing Factors, Stimulus Events, and Maladaptive Social Interaction Factors, with Predisposing Factors and Maladaptive Social Interaction Factors functioning as moderators of the relation between Stimulus Events and the Response theme. The relation between Predisposing Factors and Stimulus Events in the production of health anxiety is consistent with the triple vulnerabilities model of psychopathology wherein a general psychological vulnerability predisposes an individual to risk of developing a psychiatric disorder that is activated by a specific learning event (ie, a Stimulus Event). 22 It appeared that these factors are also compounding, such that the more stimulus events that a patient experiences and the more prominent the predisposing factors, the more likely they are to experience health anxiety. Events in these direct contributing factor themes were reporting by all of the participants in this study, which further reinforces the importance of some key learning events to development of health anxiety. Predisposing Factors appear to have the effect of increasing or decreasing the salience of the stimulus events, consistent with its proposed role as a moderating influence on health anxiety. Similarly, Maladaptive Social Interaction Factors likely determines the influence of stimulus events also moderated by predisposing factors on health anxiety. That is, interpersonal behaviors, such as not trusting providers, leads to health anxiety in the presence of events that otherwise might not result in health anxiety alone (eg, having a negative encounter with a physician would not necessarily lead to health anxiety without mistrust). Importantly, the hypothesized moderating effect of Maladaptive Social Interactions Factors is characterized by problematic social behaviors and relationships which can be modified bi-directionally through improved patient-clinician interaction, whereas predisposing factors and the stimulus itself are perhaps more effectively addressed through a different modality, namely cognitive-behavioral therapy. 23 The proposed framework gives clinicians a target to either address the underlying psychopathology through evidence-based treatment of health anxiety or reduce the symptomatic presentation through optimization of interpersonal interactions.
This model provides several recommendations that a clinician could enact when working with a health anxious patient. One simple recommendation would be to increase clinician practice of actions in the Mitigating Factor theme. Indeed, participants in this study indicated that these behaviors are potential solutions to the conflict that arises in the patient-provider relationship in the context of health anxiety. Clinicians could address health anxiety early by identifying potential perceived violations of trust and attempt to repair this with patients by noting the perception directly and inquiring about ways to regain trust. This is, of course, not unique to patients with health anxiety, but is particularly relevant to that population based on this model.
Several participants noted that they simply wanted their providers to acknowledge that a health problem or concern might not be due to anxiety and that they do not see them as an “anxious patient.” Participants endorsed feeling less health anxious when their physician took time to listen to their concerns authentically and had a bidirectional conversation, using their clinical knowledge to give patients evidence that their concerns were not supported by available evidence. For example, 1 participant feared she had sarcoidosis in her scalp despite no classic symptoms; the participant cited that she “didn’t know if scalp sarcoidosis was possible.” In fact, this diagnosis would be extraordinarily rare, 24 yet this patient was spending many hours of her day worried that her scalp was being affected by sarcoidosis. This participant stated that she was hesitant to voice her concerns due to her belief that she would be disturbing her physician with her concerns, and hence suffered with worry about the possibility of a medical rarity. Given our findings, an open and clinically informed conversation about her concerns would have gone far in ameliorating her suffering.
A parallel recommendation is to reduce events that contribute to or interact with the Maladaptive Social Interaction Factor theme. For example, interactions in which the patient is left feeling as though they are “not being understood” and the related experience of feeling as if they are “falling through the cracks” of the medical system. Participants endorsed feeling as though medicine has become impersonal, citing that their physicians no longer had enough time to listen to their concerns. They also noted that, because of the fragmentation of medicine into isolated specialties, providers were not aware of the patient’s interacting needs and therefore could not correctly diagnose a “systemic” disease. Our analysis showed that health anxious patients believe that medicine cannot be trusted, in general. This may then give patients an opportunity to rationalize that their health anxiety and fears are warranted due to the medical system overall not being able to determine how all of their different experienced symptoms are interconnected. A remedy to these aforementioned concerns is empathic listening and authentic interest about the patient’s symptom experience and following-up with the findings from other clinicians in ways that are visible to the patient, such as in shared communications available in a patient portal.
Patients experiencing health anxiety often express rational reasons to be worried about their health, but lack understanding of the medical pathophysiology or medical system at large to understand their symptoms. They instead develop safety behaviors, particularly information-seeking, as a coping strategy in order to reduce the negative experiences of health anxiety. This lack of medical understanding combined with access to medical information through the Internet and other media increases the ease with which safety behaviors can flourish within the individual. Furthermore, our study indicates that when patients perceive that their symptom are written off as “symptoms of health anxiety,” their safety behaviors increase and health anxiety symptoms worsen. To counter this problem, clinicians could provide health education to address the direct causal factors as well as reduce experienced symptoms. The challenge to the clinician is that, without adequate trust building between clinician and patient, such a suggestion could be viewed by the patient as a dismissal of symptoms and hence exacerbating health anxiety. For this reason, we recommend reducing the intensity of health anxiety factors first through addressing any Maladaptive Social Interaction Factors and activating Mitigating Factors such as building trust and providing sufficient medical evidence. Only when enough trust is built, the clinician delivers health education to solidify the alternate explanation and encourage a psychotherapeutic approach through the appropriate modality.
The following recommendations summarize the clinical implications of these data. First, providers can decrease the effect of health anxiety on the patient-provider relationship by developing trust through active listening, including giving the patient an opportunity to fully express their concerns and complaints without interruption or rebuttal, and reflecting these back in summarized form. A second recommendation stems from the conversation surrounding their concerns, in which the provider actively validates the patient’s concern that the feared symptoms or sensations could be due to a physical illness without defaulting to the anxiety explanation. In cases where the patient has a documented history of anxiety, such as in a problems list within the electronic medical record, the provider should acknowledge that this is a previously noted problem, but that anxiety is not the first explanation considered. The third recommendation draws from a more general approach to management of anxiety in which the provider offers reassurance through medical expertise, including the low likelihood of a dreaded outcome, while accurately stating that absolute certainty is not realistic. This is consistent with management other anxiety and related disorders in which there is an explicit acknowledgement that certainty is not possible, but living with uncertainty is a reasonable goal. A caveat to providing reassurance is that this should applied judiciously to prevent reinforcement of habitual reassurance-seeking behavior. One method of facilitating this is to offer scheduled follow-up visits that are independent of symptom report to decrease reinforcement of visits as an anxiety-reducing behavior (ie, a safety behavior). Those scheduled follow-ups are then faded over time. 25 A fourth recommendation is to address potential violations of trust directly, by stating the perception that trust or confidence may be waning and inquiring openly about ways to regain the patient’s trust and confidence. Finally, providers may further ingrain trust and confidence that they are not narrowly seeing the patient as “just anxious” by demonstrating awareness of other clinicians’ evaluations and documentation to ensure that the patient is known as a “whole patient” with interacting problems and concerns. This can be done in the visit by referencing prior clinician care or in the documentation visible to the patient after consulting with the other clinicians in the patient’s care team.
The results of our analysis suggest that health care clinicians play an important role in the management of health anxiety and that their interactions with patients can function as both a critical stimulus event and source of poor social interaction to worsen health anxiety, and as a potential salve to reduce the effects of these prior events.
Limitations
This study was a model building effort using qualitative data. Qualitative data is a useful starting point for theory development that can be subsequently tested with other approaches, including quantitative modeling or empirical evaluation through formal experimental methods. 26 Of course, qualitative data alone has equal value to quantitative when considering model building, as quantitative data alone also does not adequately capture a phenomenon. 15 The model proposed in this study would be strengthened by a follow-up study comprised of quantitative evaluation of the model in a sample of severely health anxiety patients. Relatedly, we use the term “causal” when referring to the themes and subthemes identified in this study with recognition that outside of an experimental manipulation, causality remains theoretical at best. Similarly, the positioning of Predisposing Factors and Maladaptive Social Interaction Factors as moderators does not suppose a temporal sequence of these themes and indeed they may have a mediating effect as much as a moderating effect if tested in a quantitative framework. Nonetheless, the present work provides a starting point for consideration of how the interactions between the individual and their medical care and larger systems influence the observed behaviors. Another consideration for future investigation is identification of organizational characteristics of primary care and other medical settings where health anxiety poses risks to quality and satisfaction with care, including through consideration of forma organizational theory. Finally, we used a purposeful random sample of health anxious individuals for this study, though a stratified sample may be important to employ to account for variation in experience across other domains that may influence patient-provider relationship and access to medical care (eg, socioeconomic status, race, and ethnicity).
Conclusions
Health anxiety is a complex psychiatric problem involving several intersecting facets of a patient’s life. Our analysis highlights the role of characterological and information processing biases, personal history with illness, and interpersonal interactions with medical providers and others in their life that contribute to the development and maintenance of health anxiety. These themes culminate with the classic features of health anxiety, but can be mitigated with intervention from medical providers leveraging aspects of the patient-provider relationship. The results of this study substantively add to the understanding of health anxiety and its interactions with the healthcare and social environments with data provided by patients’ lived experiences navigating the healthcare system. The practical benefits include identification of modifiable factors that providers can target to effectively engage with these patients for a better outcome.
Supplemental Material
sj-docx-1-jpc-10.1177_21501319231214876 – Supplemental material for Challenging Interactions Between Patients With Severe Health Anxiety and the Healthcare System: A Qualitative Investigation
Supplemental material, sj-docx-1-jpc-10.1177_21501319231214876 for Challenging Interactions Between Patients With Severe Health Anxiety and the Healthcare System: A Qualitative Investigation by Robert E. Brady and Armando N. Braz in Journal of Primary Care & Community Health
Supplemental Material
sj-docx-2-jpc-10.1177_21501319231214876 – Supplemental material for Challenging Interactions Between Patients With Severe Health Anxiety and the Healthcare System: A Qualitative Investigation
Supplemental material, sj-docx-2-jpc-10.1177_21501319231214876 for Challenging Interactions Between Patients With Severe Health Anxiety and the Healthcare System: A Qualitative Investigation by Robert E. Brady and Armando N. Braz in Journal of Primary Care & Community Health
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute of Mental Health [K23MH116367 to R.B.].
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
