Mirelle Kass
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Mirelle Kass has a total of 86 co-authors. Here are the top co-authors by number of shared publications:
Michael P. Milham
4 shared publications- Giovanni Abrahão Salum
Child Mind Institute, New York, NY 10022, USA
3 shared publications - Kathleen Ries Merikangas
Genetic Epidemiology Branch, National Institute of Mental Health, Bethesda, MD
3 shared publications Lindsay Alexander
Child Mind Institute, New York, NY, USA
3 shared publications- Michael Peter Milham
Child Mind Institute, New York, USA
3 shared publications - Aki Nikolaidis
Child Mind Institute
2 shared publications
Latest Publications
- Open Access
- 14 April 2026
- Published by Elsevier in JAACAP Open
- Vol. 4 (4) , 701-712
Negative Online Experiences and Reporting Rates in Youth With Mental Health Conditions
- 14 April 2026
- Published by Elsevier in JAACAP Open
- Vol. 4 (4) , 701-712
Open AccessNegative Online Experiences and Reporting Rates in Youth With Mental Health Conditions
- Open Access
Clinical Outcomes, Privacy-Preserving Behavioral Phenotyping, and Automated Risk-Signal Characterization with Digital Journaling: A Randomized Controlled Trial and Naturalistic Study
Background: Digital journaling may serve both as a low-intensity mental health intervention and as a source of longitudinal behavioral data. Its clinical effects and its potential for privacy-preserving behavioral phenotyping and risk-signal characterization require separate evaluation. Objective: This study evaluated the clinical effects of guided and unguided digital journaling and explored whether longitudinal journaling data could characterize reproducible patterns in mood and classifier-derived distress and risk signals. Methods: Part I was an 8-week randomized controlled trial with a 1-month follow-up among 665 assigned young adults with mild-to-moderate symptoms of anxiety or depression recruited via Prolific Academic. Participants were randomized to guided journaling plus mood check, unguided journaling plus mood check, mood check only, or passive control. Available follow-up PROMIS outcomes from 423 participants contributed to the intention-to-treat models. Part II comprised exploratory, non-preregistered analyses of more than 12,000 self-initiated mood observations and classifier-derived distress and risk signals. Selected behavioral dynamics were examined in an independent general-population dataset of 16,630 users. Results: Of 665 assigned participants, 158 (23.8%) were not invited to follow-up because of incomplete onboarding/linkage or administrative issues; 84 (12.6%) were invited but provided no follow-up outcomes. Overall, 423 (63.6%) provided at least one follow-up outcome, with similar availability across conditions (61.5%–65.1%). Primary contamination-corrected analyses excluded nine passive-control participants who used the study app and yielded small anxiety estimates favoring journaling, with the largest arm-level estimate observed for guided journaling at week 12 (Cohen d =0.25). Secondary pooled estimates were approximately d=0.16 at week 8 and d=0.19 at week 12. Neither the arm-level nor pooled comparisons survived correction for multiple comparisons. Complementary Bayesian estimates were directionally consistent with lower anxiety but indicated substantial uncertainty regarding effect magnitude. We found no evidence of benefit for depression or functional impairment. In Part II, classifier-derived high-risk signals were more common among younger users (odds ratio 0.81 per year of age; P=.013). Classifier-derived distress and self-reported energy varied by time of day, and mood and risk states exhibited short-term persistence and escalation patterns. Selected behavioral dynamics were reproduced in the independent general-population dataset, although absolute rates differed between samples. Conclusions: The randomized trial yielded small, uncertain anxiety estimates favoring journaling, but these did not establish clinical benefit as no comparison survived statistical correction for multiple comparisons. No benefit was demonstrated for depression or functional impairment. Separately, the exploratory behavioral analyses showed that privacy-preserving longitudinal journaling data could be used to characterize reproducible temporal patterns in mood and classifier-derived distress and risk signals. These findings support further evaluation of digital journaling as a behavioral measurement and risk-signal characterization platform and motivate future trials of more targeted guided-journaling approaches matched to specific symptoms or therapeutic mechanisms. The study also highlights opportunities and procedural challenges in online panel-based digital trials. Trial Registration: ClinicalTrials.gov NCT07126275 .
Open AccessClinical Outcomes, Privacy-Preserving Behavioral Phenotyping, and Automated Risk-Signal Characterization with Digital Journaling: A Randomized Controlled Trial and Naturalistic Study
Background: Digital journaling may serve both as a low-intensity mental health intervention and as a source of longitudinal behavioral data. Its clinical effects and its potential for privacy-preserving behavioral phenotyping and risk-signal characterization require separate evaluation. Objective: This study evaluated the clinical effects of guided and unguided digital journaling and explored whether longitudinal journaling data could characterize reproducible patterns in mood and classifier-derived distress and risk signals. Methods: Part I was an 8-week randomized controlled trial with a 1-month follow-up among 665 assigned young adults with mild-to-moderate symptoms of anxiety or depression recruited via Prolific Academic. Participants were randomized to guided journaling plus mood check, unguided journaling plus mood check, mood check only, or passive control. Available follow-up PROMIS outcomes from 423 participants contributed to the intention-to-treat models. Part II comprised exploratory, non-preregistered analyses of more than 12,000 self-initiated mood observations and classifier-derived distress and risk signals. Selected behavioral dynamics were examined in an independent general-population dataset of 16,630 users. Results: Of 665 assigned participants, 158 (23.8%) were not invited to follow-up because of incomplete onboarding/linkage or administrative issues; 84 (12.6%) were invited but provided no follow-up outcomes. Overall, 423 (63.6%) provided at least one follow-up outcome, with similar availability across conditions (61.5%–65.1%). Primary contamination-corrected analyses excluded nine passive-control participants who used the study app and yielded small anxiety estimates favoring journaling, with the largest arm-level estimate observed for guided journaling at week 12 (Cohen d =0.25). Secondary pooled estimates were approximately d=0.16 at week 8 and d=0.19 at week 12. Neither the arm-level nor pooled comparisons survived correction for multiple comparisons. Complementary Bayesian estimates were directionally consistent with lower anxiety but indicated substantial uncertainty regarding effect magnitude. We found no evidence of benefit for depression or functional impairment. In Part II, classifier-derived high-risk signals were more common among younger users (odds ratio 0.81 per year of age; P=.013). Classifier-derived distress and self-reported energy varied by time of day, and mood and risk states exhibited short-term persistence and escalation patterns. Selected behavioral dynamics were reproduced in the independent general-population dataset, although absolute rates differed between samples. Conclusions: The randomized trial yielded small, uncertain anxiety estimates favoring journaling, but these did not establish clinical benefit as no comparison survived statistical correction for multiple comparisons. No benefit was demonstrated for depression or functional impairment. Separately, the exploratory behavioral analyses showed that privacy-preserving longitudinal journaling data could be used to characterize reproducible temporal patterns in mood and classifier-derived distress and risk signals. These findings support further evaluation of digital journaling as a behavioral measurement and risk-signal characterization platform and motivate future trials of more targeted guided-journaling approaches matched to specific symptoms or therapeutic mechanisms. The study also highlights opportunities and procedural challenges in online panel-based digital trials. Trial Registration: ClinicalTrials.gov NCT07126275 .
- Open Access
- 17 December 2025
- Published by Springer Nature in Research Square
Real-time Ecological Assessment of the Context of mental and physical Health (REACH) on Curious
Ecological Momentary Assessment (EMA) is widely used in mental health research to capture internal states and experiences as they occur in real-world settings, yet the lack of standardization in EMA content, study design, analytical approaches and data collection platforms has hindered comparability across studies and limited progress in mental health research. We have developed standardized, open-source tools for EMA with the R eal-time E cological A ssessment of the C ontext of mental and physical H ealth (REACH) at the National Institutes of Health Clinical Center, to characterize dynamic profiles of physical and mental health across the lifespan. The REACH applet is integrated into Curious, a secure, open- source, configurable, and scalable platform through a collaborative effort with the Child Mind Institute. We describe the content, materials, and methods of REACH shared under a Creative Commons license to promote access and harmonization of these tools in global settings across the life span.
- 17 December 2025
- Published by Springer Nature in Research Square
Open AccessReal-time Ecological Assessment of the Context of mental and physical Health (REACH) on Curious
Ecological Momentary Assessment (EMA) is widely used in mental health research to capture internal states and experiences as they occur in real-world settings, yet the lack of standardization in EMA content, study design, analytical approaches and data collection platforms has hindered comparability across studies and limited progress in mental health research. We have developed standardized, open-source tools for EMA with the R eal-time E cological A ssessment of the C ontext of mental and physical H ealth (REACH) at the National Institutes of Health Clinical Center, to characterize dynamic profiles of physical and mental health across the lifespan. The REACH applet is integrated into Curious, a secure, open- source, configurable, and scalable platform through a collaborative effort with the Child Mind Institute. We describe the content, materials, and methods of REACH shared under a Creative Commons license to promote access and harmonization of these tools in global settings across the life span.
- Open Access
- 16 October 2025
- Published by Elsevier in JAACAP Open
- Vol. 4 (1) , 124-139
Addressing Critical Workforce Shortages in Youth Mental Health: Implementation of a Multicomponent Career Development Program
- 16 October 2025
- Published by Elsevier in JAACAP Open
- Vol. 4 (1) , 124-139
Open AccessAddressing Critical Workforce Shortages in Youth Mental Health: Implementation of a Multicomponent Career Development Program
- Open Access
- 11 October 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Real-time Ecological Assessment of the Context of mental and physical Health (REACH): An open-source tool on the Curious platform (Preprint)
UNSTRUCTURED Ecological Momentary Assessment (EMA) is widely used in mental health research to capture internal states and experiences as they occur in real-world settings. Despite its widespread use, the lack of standardization in EMA content, study design, analytical approaches and data collection platforms has hindered comparability across studies and limited the progress in understanding mental health disorders. This paper describes our development of standardized, open-source tools for collecting EMA with the Real-time Ecological Assessment of the Context of mental and physical Health (REACH) app for the Rhythms and Blues Study protocol at the National Institutes of Health Clinical Center, a study that characterizes both the lifetime and dynamic profiles of physical and mental health across the lifespan with modules for procedures, sleep, context, media use, affect, negative thoughts, events, physical activity, diet, substance use, pain, headache, physical health and menstrual cycle. The REACH applet has been integrated into Curious, a secure, open- source, configurable, and scalable (formerly Mindlogger) platform through a collaborative effort with the Child Mind Institute in New York. There are four versions of EMA content in REACH: a full and short version for individuals 12 or older, and a full and short version for children ages 6 through 11. Here, we describe the content, materials, and methods of REACH shared under a Creative Commons license to promote access and harmonization of these tools in global settings across the life span particularly in the mental health field.
- 11 October 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessReal-time Ecological Assessment of the Context of mental and physical Health (REACH): An open-source tool on the Curious platform (Preprint)
UNSTRUCTURED Ecological Momentary Assessment (EMA) is widely used in mental health research to capture internal states and experiences as they occur in real-world settings. Despite its widespread use, the lack of standardization in EMA content, study design, analytical approaches and data collection platforms has hindered comparability across studies and limited the progress in understanding mental health disorders. This paper describes our development of standardized, open-source tools for collecting EMA with the Real-time Ecological Assessment of the Context of mental and physical Health (REACH) app for the Rhythms and Blues Study protocol at the National Institutes of Health Clinical Center, a study that characterizes both the lifetime and dynamic profiles of physical and mental health across the lifespan with modules for procedures, sleep, context, media use, affect, negative thoughts, events, physical activity, diet, substance use, pain, headache, physical health and menstrual cycle. The REACH applet has been integrated into Curious, a secure, open- source, configurable, and scalable (formerly Mindlogger) platform through a collaborative effort with the Child Mind Institute in New York. There are four versions of EMA content in REACH: a full and short version for individuals 12 or older, and a full and short version for children ages 6 through 11. Here, we describe the content, materials, and methods of REACH shared under a Creative Commons license to promote access and harmonization of these tools in global settings across the life span particularly in the mental health field.
- Open Access
Negative online experiences and reporting rates in youth with mental health conditions
Importance: Negative online experiences (NOE) are highly prevalent and pose significant dangers for youth with mental health and neurodevelopmental conditions. Objective: Identify risk factors of NOE and reporting barriers in youths. Design: Cross-sectional survey study (January to July 2023) and qualitative follow-up study (March to May 2023). Setting: A community-based sample from the Child Mind Institute’s Healthy Brain Network (HBN), a transdiagnostic research initiative in New York. Participants: A total of 1,009 youth aged 9-15 years (mean [SD] age, 11.79 [1.71] years; 585 [58.55%] male) completed a 310-item quantitative survey. The qualitative follow-up study (n=109) included four administrations of a 3-day online bulletin board. Exposure(s): Not applicable. Main Outcomes and Measures: Main outcomes include prevalence and risk factors of NOE and reporting behaviors and barriers. Qualitative analyses provided insights into youths’ mental models and cognitive schemas about different NOE forms. Results: Of the 1,129 enrolled participants, 1,009 participants completed the study (89.37%). Over one-fourth (265 [26.6%]) encountered NOE in the past year, with 182 [68.7%] reporting multiple incidents. However, only 53 [20.0%] reported the incident. Mental health symptoms ( p <.01) and parenting styles ( p <.001) increased the risk of NOE. Factor analyses identified three key reporting barriers: Reporting Process, Reporting Policy, and Emotional Barriers. Age, social aptitude, mental health symptoms, and parenting styles predicted the likelihood of experiencing the reporting barriers. Analyses of the qualitative follow-up study noted that youths’ decision-making process regarding reporting considers three areas: degree of malice, perpetrator’s intent, and included NOE forms. Ambiguity in these areas contributed to higher reporting uncertainty. Conclusions and Relevance: Findings highlight a gap between the prevalence of NOE and reporting rates in youths with mental health and neurodevelopmental conditions. Demographic, clinical, and parenting factors were risk factors of experiencing NOE or a reporting barrier, underscoring the need for targeted and multifaceted solutions. Potential solutions to aid in safer online spaces are proposed for policymakers, technology developers, clinicians, and educators. Key points: Question: When encountering negative online experiences (NOE), what influences reporting likelihood in youth with mental health conditions? Findings: High prevalence and low reporting rates of NOE were observed. Analyses identified Reporting Process, Reporting Policy, and Emotional as key reporting barriers. Parenting styles, mental health symptoms, and demographics were associated with increased likelihood of experiencing NOE or the identified barriers. The qualitative follow-up study provided insights into youths’ thought processes regarding NOE. Meaning: Our findings document key reporting barriers in children and adolescents with mental health and neurodevelopmental conditions and potential suggestions to aid in the development of safer online spaces.
Open AccessNegative online experiences and reporting rates in youth with mental health conditions
Importance: Negative online experiences (NOE) are highly prevalent and pose significant dangers for youth with mental health and neurodevelopmental conditions. Objective: Identify risk factors of NOE and reporting barriers in youths. Design: Cross-sectional survey study (January to July 2023) and qualitative follow-up study (March to May 2023). Setting: A community-based sample from the Child Mind Institute’s Healthy Brain Network (HBN), a transdiagnostic research initiative in New York. Participants: A total of 1,009 youth aged 9-15 years (mean [SD] age, 11.79 [1.71] years; 585 [58.55%] male) completed a 310-item quantitative survey. The qualitative follow-up study (n=109) included four administrations of a 3-day online bulletin board. Exposure(s): Not applicable. Main Outcomes and Measures: Main outcomes include prevalence and risk factors of NOE and reporting behaviors and barriers. Qualitative analyses provided insights into youths’ mental models and cognitive schemas about different NOE forms. Results: Of the 1,129 enrolled participants, 1,009 participants completed the study (89.37%). Over one-fourth (265 [26.6%]) encountered NOE in the past year, with 182 [68.7%] reporting multiple incidents. However, only 53 [20.0%] reported the incident. Mental health symptoms ( p <.01) and parenting styles ( p <.001) increased the risk of NOE. Factor analyses identified three key reporting barriers: Reporting Process, Reporting Policy, and Emotional Barriers. Age, social aptitude, mental health symptoms, and parenting styles predicted the likelihood of experiencing the reporting barriers. Analyses of the qualitative follow-up study noted that youths’ decision-making process regarding reporting considers three areas: degree of malice, perpetrator’s intent, and included NOE forms. Ambiguity in these areas contributed to higher reporting uncertainty. Conclusions and Relevance: Findings highlight a gap between the prevalence of NOE and reporting rates in youths with mental health and neurodevelopmental conditions. Demographic, clinical, and parenting factors were risk factors of experiencing NOE or a reporting barrier, underscoring the need for targeted and multifaceted solutions. Potential solutions to aid in safer online spaces are proposed for policymakers, technology developers, clinicians, and educators. Key points: Question: When encountering negative online experiences (NOE), what influences reporting likelihood in youth with mental health conditions? Findings: High prevalence and low reporting rates of NOE were observed. Analyses identified Reporting Process, Reporting Policy, and Emotional as key reporting barriers. Parenting styles, mental health symptoms, and demographics were associated with increased likelihood of experiencing NOE or the identified barriers. The qualitative follow-up study provided insights into youths’ thought processes regarding NOE. Meaning: Our findings document key reporting barriers in children and adolescents with mental health and neurodevelopmental conditions and potential suggestions to aid in the development of safer online spaces.
- Open Access
Robust Parental Preferences in Mental Health Screening in Youth From a Multinational Online Survey
Importance: Screening youth for mental disorders may assist in prevention, promote early identification, and reduce related lifetime impairment and distress. Objective: The goal was to survey parents about their comfort and preferences for pediatric mental health screening, as well as factors associated with these preferences. Design: The online survey was available July 11-14, 2021 on Prolific Academic. Analyses were conducted from November 2021 to November 2022. Setting: Online survey. Participants: The survey was administered to English-speaking parents with at least one 5-21-year old child at home. The sample included 972 parents, aged 21 and older, from the United States ( n =265), United Kingdom ( n =282), Canada ( n =171), and Other Countries ( n =254). Exposure(s): None. Main Outcome(s)/Measure(s): Parental preferences regarding the screening content, implementation preferences, and screener reviewing preferences of pediatric mental health screening were assessed in a novel survey. Mixed effects logistic models were employed to evaluate factors that influence parental comfort levels. Results: Parents, aged 21 to 65 ( M =39.4; 62.3% female), supported annual mental health screening for their child and preferred reviewing the screening results with professional staff (e.g., physicians). Parents preferred parent-report over child self-reports, though they were generally comfortable with both options. Despite slight variations based on country of residence, screening topic, and child’s age, parents were generally comfortable discussing all 21 topics. The greatest comfort was with sleep problems; the least comfort was with firearms, gender identity, suicidality, and substance use/abuse. Conclusions/Relevance: Our data indicated that parents support annual parent- and child self-report mental health screening in primary care settings, but comfort levels differ according to various factors, such as screening topic. Parents preferred screening to occur in the healthcare office and to discuss screening results with professional staff. In addition to parental need for expert guidance, the growing awareness of child mental health needs highlights the importance of addressing mental health concerns early via regular mental health screenings. KEY POINTS: Question: What are parents’ attitudes towards pediatric mental health screening in primary care settings? Findings: The vast majority of parents surveyed online ( N=972) expressed comfort with the screening of children for mental health concerns in the primary care setting. Variations in comfort were noted in relation to age of child and topics included. Parents expressed a preference for parent report over child report, as well as for reviewing screening results with professional medical staff. These findings were robust to the country of residence (e.g., United States, Canada, United Kingdom). Meaning: Our findings document parental preferences that should be incorporated to enhance the feasibility of mental health screening in primary care settings.
Open AccessRobust Parental Preferences in Mental Health Screening in Youth From a Multinational Online Survey
Importance: Screening youth for mental disorders may assist in prevention, promote early identification, and reduce related lifetime impairment and distress. Objective: The goal was to survey parents about their comfort and preferences for pediatric mental health screening, as well as factors associated with these preferences. Design: The online survey was available July 11-14, 2021 on Prolific Academic. Analyses were conducted from November 2021 to November 2022. Setting: Online survey. Participants: The survey was administered to English-speaking parents with at least one 5-21-year old child at home. The sample included 972 parents, aged 21 and older, from the United States ( n =265), United Kingdom ( n =282), Canada ( n =171), and Other Countries ( n =254). Exposure(s): None. Main Outcome(s)/Measure(s): Parental preferences regarding the screening content, implementation preferences, and screener reviewing preferences of pediatric mental health screening were assessed in a novel survey. Mixed effects logistic models were employed to evaluate factors that influence parental comfort levels. Results: Parents, aged 21 to 65 ( M =39.4; 62.3% female), supported annual mental health screening for their child and preferred reviewing the screening results with professional staff (e.g., physicians). Parents preferred parent-report over child self-reports, though they were generally comfortable with both options. Despite slight variations based on country of residence, screening topic, and child’s age, parents were generally comfortable discussing all 21 topics. The greatest comfort was with sleep problems; the least comfort was with firearms, gender identity, suicidality, and substance use/abuse. Conclusions/Relevance: Our data indicated that parents support annual parent- and child self-report mental health screening in primary care settings, but comfort levels differ according to various factors, such as screening topic. Parents preferred screening to occur in the healthcare office and to discuss screening results with professional staff. In addition to parental need for expert guidance, the growing awareness of child mental health needs highlights the importance of addressing mental health concerns early via regular mental health screenings. KEY POINTS: Question: What are parents’ attitudes towards pediatric mental health screening in primary care settings? Findings: The vast majority of parents surveyed online ( N=972) expressed comfort with the screening of children for mental health concerns in the primary care setting. Variations in comfort were noted in relation to age of child and topics included. Parents expressed a preference for parent report over child report, as well as for reviewing screening results with professional medical staff. These findings were robust to the country of residence (e.g., United States, Canada, United Kingdom). Meaning: Our findings document parental preferences that should be incorporated to enhance the feasibility of mental health screening in primary care settings.
- Open Access
- 3 January 2022
- Published by Elsevier in Journal of Psychiatric Research
- Vol. 147, 291-300
Heterogeneity in COVID-19 pandemic-induced lifestyle stressors predicts future mental health in adults and children in the US and UK
- 3 January 2022
- Published by Elsevier in Journal of Psychiatric Research
- Vol. 147, 291-300
Open AccessHeterogeneity in COVID-19 pandemic-induced lifestyle stressors predicts future mental health in adults and children in the US and UK