The sleep reactivity score was used as a predictor in a linear regression, with internalizing symptom scores over the course of a 2-year follow-up as the outcome variable.
In this model, we controlled for baseline internalizing symptoms and mean EMA scores of sleep quality and stress.
Random effects from a multilevel model (within-person daily stress predicting nightly sleep) comprised the sleep reactivity score.
It was great presenting recent work at Harvard Psychiatry Day!
Higher sleep reactivity, a trait describing how much sleep is impacted by daily stress, is associated with many psychological and physical dx’s.
We quantified this with EMA, finding that it predicts future new internalizing symptoms.
This Deep Dive covers work by @hadarfisher.bsky.social @nigeljaffe.bsky.social @diegopizzagalli.bsky.social @christianwebb.bsky.social exploring whether smartphone passive sensors & LLMs can track activation in adolescents w/ anhedonia receiving behavioral activation therapy
youtu.be/K1Lo50QBcQM?...
Now out in npj Digital Medicine 🎉
www.nature.com/articles/s41...
Our systematic review and meta-analysis examines how well language-based models detect depression from text.
We reviewed 123 studies (40,000 + observations) using NLP and machine learning.
We're hiring! The Emotion and Psychopathology in Context Lab at Harvard invites applications for a full-time RA/lab manager to start in 2026 summer. Ideal for students preparing for graduate study (e.g., clinical psych). Please consider sharing & applying! jobs.smartrecruiters.com/HarvardUnive...
Podcast cover for JAMA Psychiatry Author Interviews, dated February 18, 19 minutes. Title: Predicting Adolescent Response to School-Based Mindfulness. A blue square displays 'Psyc' and initials 'JN'. A purple play button is at the bottom.
John Torous, MD, speaks with Christian A. Webb, PhD, of Harvard Medical School and McLean Hospital, about the limits of population-level prediction and the need for more potent and targeted interventions for #YouthMentalHealth.
🎧 Listen now:
ja.ma/3OSGiIw
New paper from our team: Can ML identify which adolescents benefit most from school-based mindfulness?
In 8,376 students (MYRIAD trial), models detected statistically significant but clinically negligible differential effects (d ≈ .07–.08).
Precision prevention is hard...
Interesting to see how our lab's recruitment sources have shifted over the last 3 years (see plot).
FB/Instagram used to pull 100+ responses/month, now much less productive. Methods like flyers and mailing lists remain steady.
Curious if any other labs are seeing this or have found new methods?
Sleep quality and suicide ideation: Mediating role of rumination and depressive symptoms — new paper out, led by former Hunter College undergrad Sindy Kauch, with support from @thegraduatecenter.bsky.social PhD student Lucy Liotta and our co-authors. Link to read here: rdcu.be/e3ip2
Box. Sleep and Circadian Rest-Activity Parameters describing: Sleep Onset and Wake-Up Time, Sleep Period Time (SPT), Wake After Sleep Onset (WASO), Total Sleep Time (TST), Sleep Efficiency, Midsleep Phase (MSF), Sleep Regularity Index (SRI) text@39, Composite Phase Deviation (CPD)text@40, Activity L5 and M10text@41, Relative Amplitude of Activity (RA)text@41.
Actigraphy-derived measures of sleep phase variability and activity amplitude identified individuals at higher risk of relapse in #Depression, supporting their potential as scalable biomarkers in ongoing monitoring.
bit.ly/3OzXlPo
New preprint! We investigate the risk of algorithmic bias (across race/ethnicity, gender, and their intersection) in machine learning models predicting suicide attempts across 3 clinical settings in over 1.2 million patients.
osf.io/preprints/ps...
Interested in research on depression in youth? New opening for a full-time research assistant in our lab (webbslab.com) to work on a range of projects focused on the causes & treatment of depression in teens. Come join our growing lab. Please RT and share! Apply⬇️
webbslab.com/job-postings
JAMA Psychiatry article: "Functional Somatic Disorders in Individuals With a History of Sexual Assault" by Sofie A. Jacobsen, Marie W. Petersen, and B. Wellnitz, published online November 12, 2025.
Adults exposed to sexual assault were at higher risk of developing functional somatic disorders, including single-organ and multiorgan subtypes—over five years, finds this cohort study. ja.ma/4qKzTgd
A Practical Guide to Self-Organising Maps in Psychological Research: https://osf.io/dy6z2
💫 Excited to share new preprint a systematic review & meta-analysis of 123 studies (40k+ ppl) on how well language-based models detect depression from text.
www.researchsquare.com/article/rs-8...
Younger adolescents and preteens have become an alarming portion of patients admitted for serious suicide attempts, in addition to those dying by suicide. Refinement in screening and treatment is required.
2/2
I am pleased to share our paper showing that daily assessment improves the precision of #self-harm and #suicide prediction beyond the data available at the start of a hospital admission. psycnet.apa.org/record/2026-...
Network-informed interventions for psychopathology
Journal Club by René Freichel
go.nature.com/43ispIc
"...Taken together, this would suggest that public distrust of science does not seem to be a simple dismissal of scientific knowledge, but instead a reconfiguration of epistemic trust."
Still unknown: what works, and how do we scale communications to those who do not actively seek out new research?
"Critics (rightfully) apply a high level of scrutiny to science, but simultaneously fail to apply the same standards to pseudoscience...
...Pseudoscience exploits these differing epistemic standards and often positions itself as rescuing true, authentic science from the establishment..."
No one can deny that we need better ways of approaching science communication, particularly right now. Great read below, no matter your field!
"Leveraging the epistemic authority of the 'common man' can then make science seem less elitist and more aligned with community values."
Investigating Risk and Protective Factors for Suicidal Ideation and Attempts in Bi+ Youth: A Multi-Method Study: https://osf.io/9x4h2
I’ve spent the last 8 years(!) working from the position that HiTOP relies too much on analyses of traditional diagnoses, baking in limitations of the DSM, and that we need to move to symptom-level analyses to fix it
It turns out that rebuilding HiTOP from the ground up doesn’t change much 💀
1/
Paraphrasing @emollick.bsky.social and @dkthomp.bsky.social here: AI is (so far) much better at assisting with job-specific tasks than replacing entire careers.
Research and science appear to be no exception.
Such a cool paper!
On the common approach of comparing those w/ and w/o given DSM/ICD diagnoses (e.g., case-control):
"[These] models often fail not because of the quality of the computational approach, but because the diagnostic labels they target do a poor job of carving nature at its joints."
New paper in Journal of Affective Disorders led by Olivia Losiewicz: Daily spread of positive affect is associated with subsequent well-being: A study of idiographic network analysis and emotional inertia. www.sciencedirect.com/science/arti...
From this new post by OpenAI: 0.15% of users (something like 9M people given public numbers) show signs of suicidal intent in their ChatGPT chats each week
But there seems to be progress in making ChatGPT respond appropriately to mental health issues. openai.com/index/streng...
Thanks to @bbrfoundation.bsky.social for highlighting our lab's study (led by @hadarfisher.bsky.social & @nigeljaffe.bsky.social) using📱smartphone sensors + LLM-derived text ratings to track behavioral activation and symptom change in teens with anhedonia.
bbrfoundation.org/content/smar...
Man sitting on couch with head in hands with therapist holding clipboard reading "Mariners fan 25 years YIKES"
DSM-5 Now Includes Mariners Fanhood as a Legitimate Form of Depression: tinyurl.com/5e8a65zt