CSCaroline M. Swords

Caroline M. Swords

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Caroline M. Swords has a total of 61 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Incorporating Objective Behavioral and Biological Measures in a Large-Scale Decentralized Mobile Health Trial for Depression: Randomized Controlled Trial

    by Wendy S-Y Lau, Caroline M Swords, Margaret W Thairu, Nelson Hidalgo, Raquel Tatar, Hadley Rahrig, Christine D Wilson-Mendenhall, Daniel W Grupe, Robin I Goldman, Nathaniel J Vack, et al.

    Background: Decentralized (ie, fully remote) clinical trials (DCTs) offer a scalable and ecologically valid approach to evaluating digital mental health interventions (DMHIs). Yet, incorporating objective measures of well-being within DCTs poses novel methodological and feasibility challenges that can limit the rigor and precision of data collection in the field. Objective: We recently conducted the Behavior, Biology, and Well-being (BeWell) Study, a large-scale, 3-arm randomized DCT evaluating the Healthy Minds Program app (Humin)—a self-guided DMHI that includes didactic well-being information and meditation components. The current paper’s objective is to describe the methodological approach to collecting and processing a multimodal battery of psychological (questionnaires and structured clinical interview), biological (blood and stool assays), and behavioral (stimulus-elicited affect [SEA] task, video log, and Mnemonic Similarity Task [MST]) measures, longitudinally across baseline, the 4-week intervention period, and 17-week follow-up. Secondarily, we evaluate its feasibility. Methods: The BeWell Study methodology is described. Descriptive statistics characterized recruitment and retention, as well as data completion and data quality. Binary logistic regressions evaluated whether recruitment sources, demographics, and depression severity predicted measure completion and data quality. Results: From June 1, 2022, through July 15, 2024, 52,052 individuals started the screening questionnaire. Of these, 2.22% (n=1156 participants; mean age 38.69, SD 12.14 years; women=75.50%; non-Hispanic White=68.88%) completed the screener, met eligibility criteria (including elevated depressive symptoms [Patient Health Questionnaire-9 [PHQ-9] score≥5]), completed baseline measures, were enrolled, and were randomized. Participants were recruited from Craigslist (694/1154, 60.14%), a US-based advertisement website, and university email messages (107/1154, 9.27%); participants were enrolled from every US state. Across the study, 84% (971/1156) of participants were retained (completed ≥1 measure at 17-week follow-up). Across time points, ≥81.40% (941/1156) of participants completed questionnaires and ≥75.17% (869/1156) completed biological samples. Of those completing questionnaires, ≥84.14% (796/946) completed SEA and video log tasks and ≥91.75% (868/946) completed the MST. Of completed assessments, high-quality or analyzable data rates across all time points were ≥97.89% (926/946) for questionnaires, ≥90.20% (773/857) for SEA, ≥84.60% (725/857) for video logs, ≥94.20% (1056/1121) for the MST, ≥98.84% (849/859) for dried blood spots, and ≥99.43% () for microbiome samples. Greater depressive symptoms at screening, but not age, predicted completion of baseline measures (odds ratios 0.96‐0.99; P <.001). For enrolled participants, recruitment sources and baseline depressive symptoms were not predictive of data completion or quality. Older age and having a college degree predicted significantly higher rates of data completion across all data types (questionnaire, behavioral, and biological) at week 4 or week 17 follow-ups. Conclusions: This methodological approach offers one model of incorporating objective measures of well-being into DCTs testing DMHIs. Implications and limitations of our approach are discussed.

  • Open Access

    Mapping the methodological landscape of social media detox intervention through a scoping review of evidence relevant to social media restriction policies

    by Elombe Calvert, Shrutangi Vaidya, Gabrielle Ilagan, Caroline M Swords, Simon B Goldberg, Christopher Ferguson, Jake Linardon, John Torous

  • Open Access

    Designing an Affective Mobile Probe to Measure Smile Dynamics in Depression

    by Nelson Hidalgo Julia, Robert Lewis, Craig Ferguson, Joshua Angulo Lopez, Hahrin Jung, Rosalind Picard, Simon Goldberg, Raquel Tatar, Wendy S.-Y. Lau, Caroline Swords, et al.

  • Open Access

    Naturalistic psychedelic use and changes in depressive symptoms

    by Otto Simonsson, Peter S. Hendricks, Caroline M. Swords, Walter Osika, Simon B. Goldberg


  • Open Access

    Methodological quality in randomised clinical trials of mental health apps: systematic review and longitudinal analysis

    by Jake Linardon, Qiang Xie, Caroline Swords, John Torous, Shufang Sun, Simon B Goldberg

    Question: This study investigated the methodological rigour of randomised controlled trials (RCTs) of mental health apps for depression and anxiety, and whether quality has improved over time. Study selection and analysis: RCTs were drawn from the most recent meta-analysis of mental health apps for depression and anxiety symptoms. 20 indicators of study quality were coded, encompassing risk of bias, participant diversity, study design features and app accessibility measures. Regression models tested associations between year of publication and each quality indicator. Findings: 176 RCTs conducted between 2011 and 2023 were included. Methodological concerns were common for several quality indicators (eg, <20% were replication trials, <35% of trials reported adverse events). Regression models revealed only three significant changes over time: an increase in preregistration (OR=1.27; 95% CI 1.10, 1.46) and reporting of adverse events (OR=1.32; 95% CI 1.11, 1.56), and a decrease in apps reported to be compatible with iOS and/or Android (OR=0.78; 95% CI 0.64, 0.96). Results were unchanged when excluding outliers. Results were similar when excluding three high-quality studies published between 2011 and 2013, with additional evidence for an increase in modern missing data methods (OR=1.22; 95% CI 1.04, 1.42) and studies reporting intention-to-treat analysis (OR=1.20; 95% CI 1.03, 1.39). Conclusions: Findings provide minimal evidence of improvements in the quality of clinical trials of mental health apps, highlighting the need for higher methodological standards in future research to ensure the reliability and generalisability of evidence for these digital tools.


  • Open Access

    Longitudinal associations of naturalistic psychedelic use with psychotic and manic symptoms

    by Otto Simonsson, Simon B. Goldberg, Walter Osika, Cecilia U. D. Stenfors, Sankalp Chaturvedi, Caroline M. Swords, Jayanth Narayanan, Peter S. Hendricks

    Background: Findings from contemporary clinical trials suggest that psychedelics are generally safe and may be effective in the treatment of various psychiatric disorders. However, less is known about the risks associated with psychedelic use outside of medically supervised contexts, particularly in populations that are typically excluded from participation in clinical trials. Methods: Using a preregistered longitudinal observational research design with a purposive sample of US residents between 18 and 50 years old (N=21,990), we investigated associations between self-reported naturalistic psychedelic use and psychotic and manic symptoms, with emphasis on those with psychiatric histories of schizophrenia or bipolar I disorder. Results: The follow-up survey was completed by 12,345 participants (56% retention), with 505 participants reporting psychedelic use during the 2-month study period. In covariate-adjusted regression models, psychedelic use during the study period was associated with increases in the severity of psychotic and manic symptoms. However, such increases were only observed for those who reported psychedelic use in an illegal context. While increases in the severity of psychotic symptoms appeared to depend on the frequency of use and the intensity of challenging psychedelic experiences, increases in the severity of manic symptoms appeared to be moderated by a personal history of schizophrenia or bipolar I disorder and the subjective experience of insight during a psychedelic experience. Conclusions: The findings suggest that naturalistic psychedelic use specifically in illegal contexts may lead to increases in the severity of psychotic and manic symptoms. Such increases may depend on the frequency of use, the acute subjective psychedelic experience, and psychiatric history.


  • Open Access

    Are ecological momentary assessment measures of intervention change worth the trouble?: Evaluation in four digital mental health trials (Preprint)

    by Christian Webb, Lori Hilt, Caroline Swords, Daniel Bolt, Hadar Fisher, Simon Goldberg

    Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Ecological momentary assessment (EMA) is increasingly being incorporated into intervention studies to acquire a more fine-grained and ecologically valid assessment of change. The added utility of including relatively burdensome EMA measures in a clinical trial hinges on several psychometric assumptions, including that these measure are: (1) reliable, (2) related but not redundant with conventional self-report measures (convergent and discriminant validity), (3) sensitive to intervention-related change, and (4) associated with a clinically-relevant criterion of improvement (criterion validity) above conventional self-report measures (incremental validity). Using data from 4 app-based meditation trials (N = 412), we examined the reliability, validity, and sensitivity to change of conventional self-report and EMA measures of improvement in rumination. Conventional self-report and EMA measures of rumination were only modestly correlated, particularly with regards to change over time, which may be due to their lower reliability. Changes in rumination were larger for conventional self-report than EMA. Notably, change in both self-report and EMA rumination each accounted for unique variance in depressive symptom improvement, demonstrating incremental predictive validity. Conventional self-report and EMA measures of rumination provide distinct and clinically meaningful information. Researchers using EMA should consider the psychometric properties of their measures and the precise construct they intend to capture.