Randomized Controlled Trial of Internet‐Delivered Cognitive Behavioral Therapy After Myocardial Infarction With Nonobstructive Coronary Arteries or Takotsubo Syndrome Authors: Philip Leissner, Runa Sundelin, Elisabet Rondung, Sophia Humphries, Claes Held, Jonas Spaak, Anders Ulvenstam, Anna Nordenskjöld, Fredrika Norlund, Lena Kövamees, Patrik Lyngå, Erik Olsson, Per Tornvall Published: 2026 ABSTRACT AbstractBackgroundPatients with myocardial infarction with nonobstructive coronary arteries (MINOCA) or Takotsubo syndrome (TS) often suffer from poor mental health and low quality‐of‐life post event, with no current evidence‐based treatments available. This study aimed to evaluate the effects of an internet‐delivered cognitive behavioral therapy program on stress and anxiety symptoms in patients diagnosed with myocardial infarction with nonobstructive coronary arteries or TS.MethodsA parallel, multicenter, randomized controlled trial with a wait‐list control group was conducted. Patients with elevated symptoms of stress or anxiety were included (the 14‐item perceived stress scale ≥25 or hospital anxiety and depression scale, anxiety subscale ≥8). The primary outcome was normalized stress and anxiety levels 10 to 12 weeks after randomization. Secondary outcomes included normalized stress and anxiety separately and symptoms as continuous measures.ResultsA total of 88 patients were randomized (treatment: 45, control: 43). The treatment group showed a greater, though not statistically significant proportion of normalized patients (odds ratio, 2.37 [95% CI, 0.88–6.38], P=0.09). Secondary analyses indicated treatment effects on anxiety normalization and on symptoms as continuous measures (effect sizes: d=0.44–0.67). There was also an interaction between treatment and diagnosis (TS versus non‐TS), favoring a treatment effect in TS patients (P=0.016).ConclusionsThe patient‐tailored internet‐delivered cognitive behavioral therapy program reduced stress and anxiety symptoms in patients with myocardial infarction with nonobstructive coronary arteries or TS and normalized anxiety symptoms in a greater proportion of treated patients than controls. Exploratory analyses indicated that patients with TS may experience greater benefit from the internet‐delivered cognitive behavioral therapy intervention, but this warrants further research.Keywords: anxiety, iCBT, MINOCA, RCT, stress, Takotsubo syndromeSubject Categories: Behavioral/Psychosocial Treatment, Digital Health Nonstandard Abbreviations and Acronyms Cardiac Anxiety Questionnaire Hospital Anxiety and Depression Scale internet‐delivered CBT Impact of Event Scale, 6 items myocardial infarction with nonobstructive coronary arteries Perceived Stress Scale, 14 items Takotsubo syndrome Clinical Perspective What Is New? Patients with myocardial infarction with nonobstructive coronary arteries and Takotsubo syndrome have increased levels of stress and anxiety and thus a need for mental health support. Patients with myocardial infarction with nonobstructive coronary arteries and Takotsubo syndrome have increased levels of stress and anxiety and thus a need for mental health support. What Are the Clinical Implications? A patient‐tailored internet‐delivered cognitive behavioral therapy can reduce symptoms of stress and anxiety but patients with Takotsubo syndrome seems to derive greater benefit from the treatment compared with myocardial infarction with nonobstructive coronary arteries. A patient‐tailored internet‐delivered cognitive behavioral therapy can reduce symptoms of stress and anxiety but patients with Takotsubo syndrome seems to derive greater benefit from the treatment compared with myocardial infarction with nonobstructive coronary arteries. The lack of evidence regarding normalization of both stress and anxiety calls for future trials to investigate whether better powered trials, or intervention modifications, can improve results before routine clinical application. The lack of evidence regarding normalization of both stress and anxiety calls for future trials to investigate whether better powered trials, or intervention modifications, can improve results before routine clinical application. Approximately 9% of all patients with a suspected myocardial infarction (MI) present with nonobstructive coronaries (MINOCA), 1 and 2% to 3% are diagnosed with Takotsubo syndrome (TS). 2 There is currently lacking evidence regarding the causes, prognoses, and treatments of these conditions, leading to considerable distress for patients. 3 Both MINOCA and TS present with chest pain, ischemic ECG changes, and elevated cardiac troponins, strongly indicative of acute MI. Although there are important differences in pathophysiology, as TS was differentiated from ischemic MI in 2018, 4 both conditions are associated with diagnostic uncertainties and lack of a clear prognosis. The clinical presentation remains similar and TS continues to be an important differential diagnosis when investigating the working diagnosis MINOCA. Furthermore, both patients with MINOCA and TS have a higher prevalence of psychiatric disorders before the event compared with those with MI due to obstructive coronary arteries. 5 , 6 Additionally, TS is reported to be triggered by stressful physical or emotional events in 70% of the cases, 4 , 7 and underlying MINOCA causes, for example, coronary microvascular dysfunction and coronary arterial spasm, are often triggered by psychological distress. 8 A recent systematic review highlighted the elevated level of mental distress after a MINOCA or TS, and that some patients do not seem to improve over time. 9 Mental distress not only affects quality of life in patients with cardiovascular disease, but it also predicts poor adherence to cardiac rehabilitation 10 and implies an increased risk of recurrent cardiac events. 11 Currently, knowledge on how to optimize cardiac rehabilitation for patients with MINOCA or TS is lacking, given their specific risk‐factor and symptom profiles. Patients with MINOCA are often offered traditional cardiac rehabilitation (designed for patients with ischemia), which has shown to be beneficial for them. 12 Patients with TS are often not offered traditional cardiac rehabilitation, assuming that they have other, yet to be formulated, needs. 7 Given the poor mental health and low quality of life associated with these conditions, psychotherapy could be a promising intervention. Cognitive behavioral therapy (CBT) has previously proved effective in improving mental health outcomes for patients with cardiovascular disease 13 and this effect has also been observed for internet‐delivered CBT (iCBT). 14 A meta‐analysis comparing guided iCBT with face‐to‐face CBT in patients with psychiatric and somatic disorders, found that the effectiveness of the 2 treatments was equivalent. 15 As both CBT and iCBT are promising treatment options for these patients, iCBT has the benefit of being more cost effective and improving access to treatment. 16 A previous example of iCBT in patients with atrial fibrillation found it effective in improving disease‐specific quality of life, reducing symptoms of cardiac anxiety and depression, and lowering disease‐specific health care consumption. 17 Despite evidence of mental distress in patients with MINOCA and TS, no studies have yet investigated the effects of psychotherapeutic interventions, digital or otherwise, specifically targeting these patient groups. Based on previous studies, patients expressed the main concerns were treatment of stress and anxiety. 3 , 18 The trial was tested in a single‐arm pilot study, including 11 participants with a recent MINOCA or TS, and the study protocol was found feasible in regard to recruitment, data collection, and intervention. 19 The current article describes the subsequent randomized controlled trial (RCT) that aimed to evaluate the effects of an iCBT intervention on self‐reported symptoms of anxiety and stress in patients who recently suffered from MINOCA or TS. METHODS The trial steering committee consisted of a number of 12 psychologists, physicians, nurses, and a patient representative. The mean age was 56 years (31–75) and 58% were male. The trial is reported in accordance with Consolidated Standards of Reporting Trials 2025. Data Availability Statement The data underlying this article cannot be shared publicly due to the General Data Protection Regulation (2016/679). The data and all intervention material will be shared on reasonable request to the corresponding author. Patients and Setting The eSMINC (E‐Health Treatment of Stress and Anxiety in Stockholm Myocardial Infarction With Non‐Obstructive Coronaries) study, was a parallel, multicenter RCT (clinicaltrials.gov NCT04178434). Trial protocol and statistical analysis plan has been previously published. 20 Inclusion criteria for the trial were (1) a suspected diagnosis of MINOCA or TS, (2) completely normal coronary arteries or stenosis at coronary angiography, <50%, (3) age 35 to 80 years with the upper age restriction removed on June 30, 2023 when 41 patients had been recruited, (4) reading and writing proficiency in Swedish, (5) access to an appropriate internet‐browsing device (eg, smartphone, smart‐pad or computer) and ability to use it, and (6) elevated levels of stress or anxiety at admission as assessed by ≥8 on the Hospital Anxiety and Depression Scale Anxiety subscale (HADS‐A) or ≥25 on the Perceived Stress Scale 14‐item version (PSS‐14). Exclusion criteria were myocarditis, spontaneous coronary artery dissection, acute pulmonary embolism, acute MI type 2 and cardiomyopathy other than TS. Patients were recruited at 7 emergency hospitals in Sweden. The study was performed in accordance with good clinical practice and the Declaration of Helsinki. It was approved by the Swedish ethical review authority (Dnr: 2018/1434‐31) and all patients gave their written informed consent. Data collection and intervention delivery were conducted via the U‐CARE internet portal, a secure internet platform with a secure login procedure. 16 The portal was used for administering the study, including randomization, data collection, automatic reminders, and delivery of the intervention with written communication between patients and psychologists. Patients were eligible for screening and inclusion into the study up to 30 days after admission to hospital. Two weeks after screening and inclusion, eligible patients completed a baseline assessment and were subsequently randomized, stratified by recruiting hospital (1:1 ratio) using permuted block randomization with a block size of 6, to either treatment or control. The 2 arms consisted of either immediate access to the iCBT program after randomization, or a wait‐list with treatment as usual before gaining access to the iCBT program after the follow‐up. Treatment as usual consisted of 2 doctor and 1 nurse follow‐up appointments. The study follow‐up was conducted 10 to 12 weeks after randomization. Due to the nature of psychological interventions, participants and intervention providers could not be blinded to group allocation. Outcome assessments were self‐reported online and could not have been influenced by researchers. Intervention The intervention consisted of a 9‐step iCBT program. This program was developed using patient research partners who themselves had experienced MINOCA or TS, 18 and it was tested in a feasibility study showing good compliance rates and was being evaluated as helpful and relevant. 19 Patients were advised to spend 2 to 3 hours on each step and to complete 1 step each week. On a mutually agreed day of the week, the patient received written feedback from a licensed psychologist. Additionally, patients were offered 2 phone calls with their psychologist to discuss the treatment plan, goals and expectations, and any practical issues. These phone calls were scheduled after step 1 and step 3. Step 1 introduced the treatment, offered information on the conditions and common reactions, and goal‐setting for the treatment. Steps 2 to 4 focused on stress management, problem‐solving, and scheduling of relaxation and recovery. Step 5 focused on values and value‐driven action. Step 6 was an introduction to exposure therapy. Steps 7 and 8 were optional based on the need and willingness to work with graded exposure therapy. Step 9 concluded the program and focused on summarizing progress and planning for sustained progress and relapse prevention. The wait‐list control group was treated as usual, depending on their final diagnosis. Patients of the study, in either group, were not excluded from cardiac rehabilitation if it was offered. Risk Assessment At baseline, suicidal ideation was assessed using the Montgomery‐Asberg Depression Rating Scale, self‐rating version. 21 Patients' mental health was also assessed using the patient health questionnaire, 4 items every week. 22 Patients showing signs of suicidality or significant deterioration were followed up by the psychologist for further risk assessment. Measures Perceived Stress Scale, 14 Items The PSS‐14 was used to assess the degree to which respondents find their lives unpredictable, uncontrollable, and overloading during the past month. 23 It consists of 14 items, graded on a 5‐point Likert scale, 0 to 4 (never, almost never, sometimes, fairly often, and very often), with the total score being 56. Internal consistency has been found to be good and test–retest reliability acceptable for intervals of <4 weeks. 24 Based on Swedish norms, a cutoff ≥25 has previously been used to discriminate between normal and heightened levels of stress. 25 , 26 Although commonly reported as a unidimensional scale, a recent publication has suggested that the PSS‐14 could be better described by 2 subscales separately, coping (the 7 positively formulated items) and distress (the 7 negatively worded items). 27 Hospital Anxiety and Depression Scale The HADS was used to assess the prevalence and intensity of symptoms of anxiety and depression during the past week. 28 It consists of 14 items, where 7 assess symptoms of anxiety (HADS‐A) and 7 symptoms of depression (HADS‐D). Each item is reported on a 4‐point Likert scale, ranging from 0 to 3, with the total score of each subscale being 21. Internal consistency is good and a cutoff value of ≥8 has been found to result in a good balance between sensitivity and specificity regarding detecting potential cases of anxiety or depression. 29 Cardiac Anxiety Questionnaire The Cardiac Anxiety Questionnaire (CAQ) was used to assess the frequency of symptoms of heart‐related anxiety. 30 The scale consists of 18 items, rated on a 5‐point Likert scale, 0 to 4 (never, rarely, sometimes, often, and always), comprising the 3 factors fear (8 items), avoidance (5 items), and attention (5 items). The authors of the questionnaire have suggested dividing the total score and subscale scores with the number of items to make them more readily comparable. There is no established cutoff and the questionnaire has not been validated against psychiatric diagnoses. The scale has previously demonstrated acceptable internal consistency, stable test–retest reliability and adequate construct validity in Swedish patients who were post‐MI. 31 Impact of Event Scale, 6 Items The Impact of Event Scale, 6 items (IES‐6), is a short form of the revised, 22‐item, impact of event scale 32 and was used to assess the intensity of subjective distress following a traumatic event. The 6 items are rated on a 5‐point Likert scale, 0 to 4 (not at all, a little bit, moderately, quite a bit, and extremely), with a possible total score of 24. The 6‐item version has demonstrated good internal consistency, a high correlation with the 22‐item version, and has been proven stable across several subgroups and populations. A cutoff value of ≥9 has been shown to predict a clinically significant score on the PCL with 0.86 sensitivity and 0.88 specificity, in indicating posttraumatic stress disorder. 32 Background Information At screening, a customized questionnaire was administered to collect background data with the following questions: civil status (single, living with partner/married, living alone but in relationship, or other); highest achieved level of education (primary education, secondary education, tertiary education ≤3 years, or tertiary education >3 years); if born in Sweden (yes, no); if not born in Sweden, country of birth; if currently working at the time of the cardiac event (yes, no); if not working, which option best described their situation (retired, unemployed, student, sick leave, early retirement, homeworker/working at home without pay, or other); smoking status (never, not now but I have been a smoker before, yes I am currently a smoker); physical activity during leisure time (mainly sedentary, light, moderate, or high); psychological treatment before the cardiac event (yes, no); if experienced physical or mental strain in conjunction with the cardiac event (physical, mental, or both); if a parent or sibling has had an MI before the age of 65 years (yes, no); and if a parent or sibling has had either a MINOCA or Takotsubo (yes, no). The following information was added from the patients´ medical charts after inclusion was completed; whether MINOCA or TS was suspected following coronary angiography during the hospitalization; medical treatment for hypertension and hyperlipidemia; if the patient was diagnosed with type 1 or type 2 diabetes; if cardiac magnetic resonance imaging had been conducted; and final diagnosis at discharge from the hospital made by the discharging cardiologist. Outcome The primary outcome was a composite consisting of the proportion of patients who reported normalized ratings (ie, not heightened or not indicating potential clinical cases, see above) of both stress and anxiety, assessed by the cutoff values on PSS‐14 (<25) and HADS‐A (<8) after the intervention. Secondary outcomes were (1) the proportion of patients reporting normalized ratings on PSS‐14 and HADS‐A separately, (2) changes in the PSS‐14 and HADS‐A scores, as well as (3) changes in the scores of the PSS‐14 subscales, HADS‐D, CAQ and its subscales, and IES‐6. Sample Size Sample size was calculated based on an expectation that 45% of the patients in the wait‐list control and 75% of the patients in the treatment arm would reach normalized levels of stress and anxiety at follow‐up. Detecting such a difference with 80% power (P<0.05), required 80 patients. Subsequently, to allow for a 10% attrition rate, inclusion was set to 90 patients. Statistical Analysis All statistical analyses were performed using STATA version 18.0. Any missing data in the primary analyses were handled using multiple imputation. Total scores for missing variables were imputed using chained equations and predictive mean matching, stratified by randomization group. Descriptive data were presented as percentages and counts or means±SD. The primary analysis (proportion of patients with normalized levels of stress and anxiety) was conducted using logistic regression with group allocation as the independent variable. This analysis followed the intention‐to‐treat principle. To increase statistical power and precision of the analysis, 33 age and sex were added as covariates to the model. Additionally, baseline values of both primary outcome measures were added as covariates to account for the effect of regression to the mean. 34 Secondary analyses consisted of logistic regression of normalization of the 2 primary outcome variables (PSS‐14 and HADS‐A), separately, and linear regression analyses of all continuous measures of stress and anxiety, as well as depression (HADS‐D) and symptoms of posttraumatic stress (IES‐6). All models included the covariates age, sex, and baseline values of the respective outcome measure. Cohen's d was calculated as a measure of effect size based on pretest‐posttest‐control group design, following the dppc2 formula. 35 Cohen's d effect size of 0.2 to 0.5 is considered small, 0.5 to 0.8 moderate, and > 0.8 large. 36 Subgroup analyses with an added interaction term (group allocation × final diagnosis of TS/non‐TS) were performed to investigate if TS status may have moderated the effect of treatment. The results were visualized in forest plots, showing odds ratios (OR) with 95% CI for dichotomous outcomes and standardized estimates (Z scores) for continuous outcomes. Two sensitivity analyses were conducted. First was a per‐protocol analysis, identical to the primary analysis, but including only patients in the treatment group completing the minimum requirement of 5 steps. Second, the primary analysis was repeated with only observed (nonimputed) data. To investigate if there were any outliers, the residuals of the linear regressions of PSS‐14 and HADS‐A were visualized in box plots, where a value of >1.5 interquartile range from the first or third quartile would be considered an outlier. 37 The logistic and linear regressions of PSS‐14 and HADS‐A were then repeated after removal of identified outliers as exploratory sensitivity analyses. RESULTS A flow chart depicting the recruitment process is presented in Figure 1. Recruitment for the eSMINC study started November 1, 2021 and finished November 6, 2024. Data collection on the follow‐up (10–12 weeks after randomization) was concluded February 9, 2025. Screening and inclusion of study patients. CBT indicates cognitive behavioral therapy; HADS, Hospital Anxiety and Depression Scale; and MINOCA, myocardial infarction with nonobstructive coronary arteries; and PSS‐14, Perceived Stress Scale, 14 items. During the 3 years of recruitment, 289 patients with suspected MINOCA or TS were screened for possible eligibility at stage 1. Elevated levels of stress or anxiety were not considered at this stage. Of these, 90 patients were excluded based on criteria, 47 patients declined to participate in the study, and 22 had an unclear diagnosis or were excluded due to other reasons. The remaining 130 patients consented to participation in the eSMINC study and were screened at stage 2 for elevated levels of stress or anxiety (≥25 on the PSS‐14 or ≥8 on the HADS‐A). Of these, 39 patients were excluded due to low ratings on both stress and anxiety, resulting in a sample of 91 patients. Of these, 88 completed the baseline questionnaires 2 weeks later and were randomized. At screening, out of the randomized patients, 65 (74%) had elevated ratings on both PSS‐14 and HADS‐A, whereas 9 (10%) had elevated ratings only on PSS‐14, and 14 (16%) only on HADS‐A. At baseline, 89% (78 patients) reported elevated ratings on either PSS‐14 or HADS‐A, or both. Forty‐five patients were allocated to the treatment arm. Of these, 42 (93%) initiated treatment, 39 (87%) completed the predefined minimum dose of 5 steps, 20 (44%) completed the optional steps on graded exposure, and 33 (73%) completed the final step. On average, patients completed 7.2 steps. No harms or unintended events were recorded for either group. There was 1 missing observation at baseline for the outcomes PSS‐14, HADS‐D and the CAQ subscale Fear. At follow‐up, there were 7 missing observations on the measures of CAQ (including its subscales) and IES‐6, and 8 missing observations on the follow‐up measures of HADS‐A, HADS‐D, and PSS‐14 (with subscales). Patients' mean age was 63 years and 91% were female. Out of the included patients, 40 (45%) had a working diagnosis of MINOCA and 48 (55%) a working diagnosis of TS. The majority of patients got a final diagnosis of TS n=52 (59%). The remaining final non‐TS diagnoses were MINOCA (n=21; 24%), suspected MI (n=11; 12%), and other (n=4; 5%). A list of suspected and final diagnoses and a Sankey diagram are available in Table S1 and Figure S1. Further descriptive data from screening is presented in Table 1. Data from baseline on outcome variables are reported in Table 2. Table 1Patient Characteristics at ScreeningTotal N=88Treatment N=45Control N=43Female sex91% (80)91% (41)91% (39)Highest level of completed educationPrimary3% (3)0% (0)7% (3)Secondary28% (25)31% (14)26% (11)Tertiary ≤3 y31% (27)22% (10)40% (17)Tertiary >3 y38% (33)47% (21)28% (12)Born in Sweden84% (74)84% (38)84% (36)Employment statusWorking52% (46)53% (24)51% (22)Retired43% (38)44% (20)42% (18)Sick leave1% (1)0% (0)2% (1)Other3% (3)2% (1)5% (2)Current smoker2% (2)4% (2)0% (0)Physical activity during leisure timeMainly sedentary9% (8)7% (3)12% (5)Low physical activity25% (22)13% (6)37% (16)Moderate physical activity53% (47)62% (28)44% (19)High/vigorous physical activity12% (11)18% (8)7% (3)Experience of physical or mental stress in conjunction with the cardiac eventNo stress25% (22)22% (10)28% (12)Physical9% (8)9% (4)9% (4)Mental44% (39)42% (19)47% (20)Both mental and physical22% (19)27% (12)16% (7)Parent or sibling with history of myocardial infarction before 65 y24% (21)29% (13)19% (8)Treatment for hypertension50% (44)53% (24)47% (20)Treatment for hyperlipidemia31% (27)28% (12)33% (15)Treatment for diabetes8% (7)4% (2)12% (5)Previous psychological treatment34% (30)31% (14)37% (16)Cardiac magnetic resonance imaging conducted64% (56)64% (29)63% (27) Patient Characteristics at Screening Table 2Results of Linear Regressions, Baseline, Follow‐Up Scores, and Effect Sizes (Cohen's d) for Continuous OutcomesLinear regressionBaseline scoresFollow‐up scores and effect sizesβ95% CI P valueTreatment mean (SD)Control mean (SD)Treatment mean (SD)Control mean (SD)Cohen's dHADS depression−1.24−2.52–0.090.0577.8 (3.3)7.9 (3.2)5.1 (3.6)6.4 (4.1)0.37HADS anxiety−1.98−3.46 to −0.500.00910.0 (3.4)9.6 (3.1)6.2 (3.8)8.0 (3.8)0.67Perceived Stress Scale, 14 items−3.05−5.90 to −0.200.03630.4 (6.9)29.9 (6.4)23.5 (6.7)26.3 (9.2)0.48Coping−1.35−3.10–0.410.13013.1 (3.6)13.7 (3.4)10.0 (3.8)11.6 (5.0)0.30Distress−1.76−3.32 to −0.200.02817.3 (4.2)16.2 (4.0)13.5 (4.0)14.6 (5.1)0.53Cardiac Anxiety Questionnaire−0.27−0.46 to −0.070.0081.7 (0.6)1.7 (0.7)1.08 (0.6)1.4 (0.6)0.44Fear−0.41−0.64 to −0.19<0.0011.7 (0.8)1.8 (0.8)1.1 (0.7)1.5 (0.8)0.50Avoidance−0.23−0.59–0.130.2012.0 (0.9)2.0 (0.9)1.3 (1.0)1.5 (0.9)0.25Attention−0.18−0.41–0.050.1201.5 (0.7)1.5 (0.7)0.9 (0.7)1.1 (0.7)0.24Impact of Event Scale, 6 items−2.24−4.03 to −0.450.01512.8 (4.5)12.2 (5.0)7.2 (4.9)9.2 (4.3)0.56In the linear regressions, baseline values, age and sex were added as covariates. HADS indicates Hospital Anxiety and Depression Scale. Results of Linear Regressions, Baseline, Follow‐Up Scores, and Effect Sizes (Cohen's d) for Continuous Outcomes In the linear regressions, baseline values, age and sex were added as covariates. HADS indicates Hospital Anxiety and Depression Scale. Primary Analysis At follow‐up, the treatment group had a numerically greater proportion of patients with normalized ratings on both the HADS‐A and the PSS‐14 (nonimputed data: 21/40, 52.5%) compared with the wait‐list control group (nonimputed data: 15/40, 37.5%), but this difference was not statistically significant (OR, 2.37 [95% CI, 0.88–6.38], P value=0.088). Secondary Analyses When analyzing the PSS‐14 and HADS‐A outcomes separately, there were greater proportions with normalized scores in the treatment group than in the control group for both measures, but only the difference for HADS‐A was statistically significant (PSS‐14 OR, 2.18 [95% CI, 0.81–5.85], P value: 0.123; HADS‐A OR, 2.75 [95% CI, 1.01–7.51], P value: 0.049), see Figure 2. Number (and %) of normalized ratings on the primary composite outcome, as well as on the PSS‐14 and HADS‐A separately, at follow‐up. HADS indicates Hospital Anxiety and Depression Scale; and PSS‐14, Perceived Stress Scale, 14 items. When analyzing the 2 primary outcomes as continuous scales, the treatment group had lower ratings than the control group on both PSS‐14 and HADS‐A at follow‐up (P=0.04 and P=0.009 respectively). Additionally, there were also effects of treatment on the outcomes IES‐6, the PSS‐14 Distress subscale, CAQ Total, and the CAQ Fear subscale (see Table 2 and Figure 3 for an overview of continuous outcomes). The effect sizes, where there was an effect, ranged between 0.44 and 0.67. Line graph of mean values of PSS‐14 and HADS Anxiety at screening, baseline, and follow‐up. HADS indicates Hospital Anxiety and Depression Scale; and PSS‐14, Perceived Stress Scale, 14 items. Subgroup Analyses There was evidence of an interaction effect between treatment and final diagnosis (TS or non‐TS) for the primary outcome (P=0.016). In the TS group, normalization rates were 67% in the treatment group and 25% in the control group showing a clear benefit of treatment. In the non‐TS group normalization rates were 31% in the treatment group and 56% in the control group, thus indicating a greater improvement in the control group. There was also support for the same interaction when normalization of the 2 primary outcomes were analyzed separately (PSS‐14, P=0.045; HADS‐A, P=0.024) favoring a treatment effect in patients with TS. Analyses by group (TS or non‐TS), investigating both dichotomous and continuous outcomes are shown in Figures 4 and 5, as well as in Tables S2 and S3. Forest plots of subgroup analyses by diagnosis (TS or non‐TS), odds ratios of normalized ratings on the PSS‐14, HADS‐A, or both. HADS indicates Hospital Anxiety and Depression Scale; PSS‐14, Perceived Stress Scale, 14 items; and TS, Takotsubo syndrome. Forest plots of subgroup analyses by diagnosis (TS or non‐TS), standardized estimates of score changes on continuous outcomes. CAQ indicates Cardiac Anxiety Questionnaire; HADS, Hospital Anxiety and Depression Scale; IES‐6, Impact of Event Scale, 6 items; PSS‐14, Perceived Stress Scale, 14 items; and TS, Takotsubo syndrome. Further exploratory analyses and descriptive data of patients with and without TS can be seen in Tables S4 and S5, and Figures S2 through S4. Sensitivity Analyses In the per‐protocol analysis, the effect of treatment yielded similar results as the intention‐to‐treat analysis (OR, 2.46 [95% CI, 0.88–6.86], P value: 0.086). The analysis of the nonimputed data also resulted in a similar result (OR, 2.17 [95% CI, 0.80–5.97], P value: 0.130). Per‐protocol (Tables S6 and S7) and nonimputed (Tables S8 and S9) analyses for all outcomes are presented in the supplementary material. An analysis of residuals in the linear models revealed 2 patients with outlier values, 1 on both the PSS‐14 and HADS‐A, and 1 on only the HADS‐A (Figure S5). The sensitivity analysis exluding these patients showed a statistically significant treatment effect on the primary composite outcome (P=0.024), and of normalization of both included measures respectively (PSS‐14, P=0.042; HADS‐A, P=0.016). Further descriptive data of the outlier and sensitivity analyses excluding these patients are reported in Tables S10 through S12. DISCUSSION This RCT investigated the effects of an iCBT intervention on self‐reported stress and anxiety in patients who had recently experienced MINOCA or TS. This is the first study to evaluate a psychological intervention specifically designed by, and for, these patients. Regarding the primary outcome, no evidence was found for a greater proportion of normalized patients in the treatment group compared with the control group. However, the majority of secondary analyses indicate a significant symptom reduction for patients in the treatment arm. Additionally, exploratory analyses revealed that patients with TS had a significantly better outcome of treatment compared with patients without TS. In the TS group, treatment effects were observed for all but 1 of the 13 outcomes investigated in this study. To our knowledge, this study represents the first reported trial to investigate the effects of iCBT in patients with MINOCA and TS. Given the extensive evidence supporting the effectiveness of iCBT in both psychiatric and somatic disorders, 38 as well as in populations with cardiovascular disease, 14 the observed effects on symptom reductions are consistent with expectations. The effect sizes found in this trial (0.44–0.67) also align with those commonly reported for both iCBT and face‐to‐face CBT interventions. 13 , 14 Notably, our findings are comparable to those of a previous study that evaluated group‐based CBT integrated into cardiac rehabilitation for patients with obstructive coronary artery disease. 39 That study reported a between‐group difference of 2.4 points on HADS‐A, which is similar to the 2.0‐point difference observed in our trial. Both of these estimates are exceeding the minimally clinically important difference of HADS‐A, which is generally considered to be 1.7. 40 This further supports the relevance and clinical utility of CBT‐based approaches in the clinical context of cardiac care and how it can be extended to patients with MINOCA and TS. Although no evidence was found for the primary outcome, normalization of both stress and anxiety, this was an ambitious treatment goal. The expected degree of normalization was derived from unpublished findings of the SMINC‐2 (Stockholm Myocardial Infarction With Normal Coronaries Study on Diagnosis Made by Cardiac Magnetic Resonance Imaging) study, showing a 45% normalization rate for patients receiving treatment as usual, 6 months after the event. 20 Sample size calculations were subsequently calculated based on an additional 30% absolute increase in the treatment group. This corresponds to a relative risk of 1.67, an OR of approximately 3.67, and a numbers needed to treat of ∼3 to 4, which is clinically meaningful for a binary outcome such as remission. 41 In the present study, the normalization rate was 37.5% for the control group and 52.5% for the treatment group, a 15% absolute increase (numbers needed to treat = 6.7). The absolute rate of normalization without treatment was lower than in the SMINC‐2 study, which may be explained by the fact that the follow‐up was scheduled 3 months after the event, not 6 as in SMINC‐2. Given the strong evidence base of iCBT, and the results of this study, there is an indication that patient‐tailored iCBT may effectively reduce symptoms of stress, anxiety, and posttraumatic stress for patients with MINOCA or TS. Integrating these findings in clinical practice would likely lead to significant improvements of mental health for these patients and may lead to improved prognosis and reduced health care costs in the long run. However, we have an important reservation regarding our results for patients without TS (many of whom were included in the study with suspected MINOCA). The interaction between treatment and the diagnosis of TS versus non‐TS, indicates that the treatment effect in this study was driven by the patients with TS. Although it should be clear that these exploratory analyses were not prespecified, the differences between groups were substantial. These findings are intriguing and may reflect underlying differences between the conditions, which warrant further investigation. For example, patients with TS have a higher prevalence of psychiatric disorders, 42 which might make them a more appropriate target group for psychological treatments. This has also been specifically called for in a recent review of TS, stating that psychological treatment may improve both prognosis and quality of life. 43 Although this study provides evidence for the latter, the former remains the be seen. Furthermore, a closer look also indicates that the group without TS has a greater spontaneous recovery than patients with TS, among the wait‐list controls (See Figures S2 through S4). This could suggest that there are differences in treatment as usual between the 2 groups that might influence the mental health status after the cardiac event. Many of the patients without TS have MINOCA or even a late diagnosis of MI with obstructed coronaries as their final diagnosis. These are by definition patients with MI, and in Sweden these patients are offered cardiac rehabilitation in a standardized manner, which is not the case for patients with TS. Given the results of these subgroup analyses, it is difficult to draw a confident conclusion about the effectiveness of the treatment for patients without TS. The specific results for patients with TS also need to be confirmed in future studies. Strengths and Limitations The study has several strengths: The preparatory work for this RCT involved active patient research partners in the development of the intervention and the design of the study. 18 The iCBT was also tested and found feasible in a following study. 19 As a result, the compliance rates were high confirming the results of the feasibility trial. 19 Comparatively, 73% of the patients completed all of the mandatory steps of the program, which is higher than the average amount (65%) of completers in iCBT treatments. 44 There were also few missing observations on the outcome variables, suggesting a minimal risk of responder bias and a high internal validity. The study was a multicenter RCT incorporated in regular care and thus has high external validity. A major strength of the present study was the use of internet as delivery format for CBT. iCBT has been shown to be equally effective as face‐to‐face CBT under certain circumstances, 45 especially regarding more specific target populations. 46 With iCBT having the advantage of being more flexible toward the patients' scheduling needs and geographical circumstances, it might be especially helpful for patients with rare conditions such as MINOCA and TS, as it allows for greater catchment areas and centralization of expertise. One limitation was the use of a dichotomous primary outcome based on 2 thresholds of 2 continuous scales. It is well known that dichotomization of outcomes has lower statistical power and increases the risk of type II errors 47 , 48 and also risks obscuring potentially important effects of the 2 components. 49 Additionally, as we observed that 10 patients had normalized ratings on both outcome measures at baseline, these could not show a treatment effect on the primary outcome. Instead, we have the benefit of continuous measures as secondary outcomes, where a majority indicate both statistically and clinically significant levels of symptom reductions. When investigating the treatment effect on the 2 measures separately, we found no effect on the dichotomized stress scale but a significantly higher proportion of participants had normalized their anxiety levels following treatment. Although the cutoff used for anxiety is well established, 29 the validity of the PSS‐14 threshold has less support. Another limitation is that our sample consisted of a majority of female patients (91%), which reflects the sex differences seen in with TS 50 but is not representative of patients with MINOCA, where there are no major sex differences. 51 , 52 This may be partially explained by the fact that men generally show a lower tendency for help‐seeking behavior. 53 , 54 Although these characteristics reflect the epidemiology of the conditions and typical participation patterns in psychological trials, it is nonetheless a limitation to the generalizability of these findings. On the other hand, the predominance of female patients in the sample may also be considered a strength, as TS primarily affects these patients. According to The Lancet Women and Cardiovascular Disease Commission, cardiovascular disease in women remains “understudied, under‐recognised, underdiagnosed, and undertreated”, 55 and with this study we add valuable information to the understanding of women's heart disease. The heterogeneity in pathophysiology between MINOCA and TS is a limitation. MINOCA is in itself a diverse condition, caused by various pathologies with different causes, whereas TS is a more definite diagnosis without known pathology. This may also have influenced why the patients with TS seemed to benefit more from the intervention than the patients with MINOCA. Our inclusion of both groups has historical and clinical reasons, as previously described. 18 , 19 Although screening and support for mental health issues should be made available for all cardiac conditions, as promoted by the new European Society of Cardiology consensus statement, 56 future trials could consider evaluating these groups separately. CONCLUSIONS In this study, a patient‐tailored iCBT intervention initiated early after the cardiac event was not found to normalize stress and anxiety at 12 to 14 weeks after a MINOCA or TS. However, secondary analyses indicated that the intervention significantly reduced symptoms of perceived stress, anxiety, and posttraumatic stress, and normalized anxiety in a greater proportion of treated patients than controls. Additional, subgroup analyses indicated that patients with TS may experience greater benefit from the intervention. These findings are encouraging and highlight the potential of iCBT for improving mental health in this understudied patient population including more women than men. Further research is needed, in particular, focusing on the psychological needs of patients with TS. Sources of Funding This study was supported by grants from the Swedish Research Council (2018‐02655), The Swedish Heart and Lung Association (Riksförbundet Hjärt‐Lung) (Grant nr 2020:21) and The Swedish Heart and Lung Foundation (20190286, 20200221, 20230225). None of the funding bodies had any influence on the study design, implementation, data analysis, or interpretation. Disclosures Jonas Spaak reports speaker honoraria from Bayer, AstraZeneca, Boehringer Ingelheim, Novo Nordisk, and Sanofi and principal investigator for clinical trials for Boehringer Ingelheim, Novo Nordisk, Astra Zeneca, and Neko Health. The remaining authors have no disclosures to report. Supporting information Tables S1–S12 Figures S1–S5 CONSORT Checklist Acknowledgments We want to thank Johan Lindbäck and Lars Lindhagen at Uppsala Clinical Research Center for their statistical consultation and Örjan Sundin for his contributions in the early preparatory stages of the trial. We would also like to extend our gratitude to the research staff at hospitals for their help with recruiting, and to the patients who have contributed to this research. Author Contributions: Philip Leissner: writing—review and editing, writing—original draft, visualization, validation, software, methodology, investigation, formal analysis, data curation, conceptualization. Runa Sundelin: writing—review and editing, writing—original draft, visualization, resources, investigation, conceptualization. Elisabet Rondung: writing—review and editing, methodology, conceptualization. Sophia Humphries: writing—review and editing, conceptualization. Claes Held: writing—review and editing, conceptualization. Jonas Spaak: writing—review and editing, resources, conceptualization. Anders Ulvenstam: writing—review and editing, resources. Anna Nordenskjöld: writing—review and editing, resources. Fredrika Norlund: writing—review and editing, methodology, conceptualization. Lena Kövamees: writing—review and editing, resources, methodology, conceptualization. Patrik Lyngå: writing—review and editing, supervision, funding acquisition, conceptualization. Erik M. G. Olsson: writing—review and editing, supervision, software, resources, project administration, methodology, investigation, funding acquisition, conceptualization. Per Tornvall: writing—review and editing, supervision, resources, project administration, methodology, investigation, funding acquisition, conceptualization. This article was sent to Ajay K. Gupta, MD, MSc, PhD, FRCP, FESC, Senior Associate Editor, for review by expert referees, editorial decision, and final disposition. Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.125.047311 For Sources of Funding and Disclosures, see page 12. See Editorial by xxx. References Associated Data Supplementary Materials Tables S1–S12 Figures S1–S5 CONSORT Checklist Data Availability Statement The data underlying this article cannot be shared publicly due to the General Data Protection Regulation (2016/679). The data and all intervention material will be shared on reasonable request to the corresponding author.