Hi friend,

Capital was flowing this week as Vanna, HMNC Brain Health, Wellbees and June all announced new rounds. This came alongside some significant M&A activity.

Definium and NRx reported positive news from research and regulatory approvals, while Medicaid Fraud crackdowns continue across behavioural health in the US.

Grow and Stanford launched a new AI research collaboration, new findings were released on at-home Ketamine therapy and lots, lots more.

Let’s get into it.

⏺️ Vanna Health raised $17M to expand serious mental illness care. Founded by former NIMH director Tom Insel, the San Francisco company runs a community-coaching model for patients with Serious Mental Illness, attempting to pioneer outcomes-based contracts with insurers. AlleyCorp and Health Velocity Capital co-led the Series B. (link)

⏺️ HMNC Brain Health raised $50M Series B to advance two depression drugs. Munich-based HMNC is developing Ketabon, an oral extended-release ketamine for treatment-resistant depression designed for at-home use, and Nelivabon, a biomarker-guided vasopressin antagonist targeting biologically defined depression. MEDICE led the round and received exclusive European commercialisation rights for Ketabon. Both drugs remain investigational and unapproved. (link)

⏺️ Wellbees raised $3.6M Series A led by Qatar Development Bank. Wellbee is an employee wellbeing platform with the round funding a planned expansion across Gulf states. (link)

⏺️ June Health raised $1.8M in pre-seed funding. The Toronto-based company provides comprehensive women's health services, including mental health. (link)

⏺️ PsychPlus acquired Koa Health. Houston-based PsychPlus acquired the London-based digital mental health business Koa Health. The acquisition will expand PsychPlus’ product portfolio and add an international business line — Koa’s CBT-based products are NHS-approved and available to over 4 million users worldwide. (link)

⏺️ Merakey acquired I Am Boundless. The deal brings together two behavioural health non-profits, which combined, will have over $1B in annual revenue and serve more than 50,000 people across 12 states. (link)

⏺️ Definium reported positive Phase 3 results for its LSD-based depression drug DT120. The formerly MindMed compound delivered an 8.1-point placebo-adjusted MADRS improvement at six weeks in the 149-patient Emerge trial. (link)

⏺️ NRx Pharmaceuticals received FDA expanded access for the use of NRX-101 alongside TMS in treatment-resistant depression. The oral combination of D-cycloserine and lurasidone can now be used alongside TMS ahead of full approval, giving patients earlier access to the treatment. (link)

⏺️ Vermont banned AI from independently providing mental health services. Act 156, signed on June 17th 2026, prohibits corporations from offering therapy through AI unless a licensed professional reviews and approves the care. Violations are classified as unprofessional conduct, making it one of the most restrictive US state approaches to AI in mental health to date. (link)

⏺️ A DOJ enforcement action totalled $208M in behavioural health billing fraud. The takedown includes $31.8M in TMS billing fraud as well as charges against an Illinois behavioural health CEO for $75M of fraud. (link) (link)

⏺️ States freeze behavioural health Medicaid enrolment. Related to the Medicaid Fraud takedowns, Utah halted new substance use disorder and ABA provider enrolments, Maryland extended a freeze on certain new provider enrollments, and Arizona suspended a clinic amid an FBI investigation. (link) (link) (link)

⏺️ YesCare shut down after its second bankruptcy, cutting 150 jobs. The PE-backed correctional behavioural health provider, formerly known as Corizon, served 475 facilities before filing. (link)

⏺️ Grow Therapy and Stanford launched a research partnership on AI clinical safety standards. The collaboration aims to establish clinical safety benchmarks for AI tools used in mental health care. (link)

⏺️ Penn Medicine's crisis centre reportedly cut psychiatric ED consults by 80% in six months. The model diverts patients from emergency departments into dedicated crisis care, reducing pressure on one of the most expensive and least therapeutic settings in the system. (link)

⏺️ Akron Children's launched a 24/7 mental health support tool keeping humans in the loop. The service, built with Sonar Mental Health, pairs human coaches with background AI. Patients interact only with the human coach, not the AI directly. (link)

⏺️ New Mexico launched a free teen wellness app. The app launched in Taos County following a cluster of youth suicides. The product specifically excludes AI features, reflecting concern about its effect on adolescent mental health. (link)

⏺️ Sensible Care launched on-demand virtual therapy. “SC Now” puts patients in a virtual waiting room and connects them to a provider as soon as one is free, moving closer to walk-in-style mental health access. (link)

⏺️ Headway has been removed from the list of accepted participants. Reporting from Exits & Outcomes noted that Headway is no longer on the CMMI’s list of participants for ACCESS.

⏺️ Study found an AI-administered PHQ-9 matched self-report scores and gained more user trust. HopeBot produced closely agreeing scores across 132 adults in the UK and China (ICC = 0.91). Of those giving feedback, 71% said they trusted the chatbot version more, citing clearer structure and a more supportive tone. This has not yet been peer reviewed. (link)

⏺️ Mental health chatbot shows positive signs of greater real-world engagement for digital mental health tools. A retrospective analysis of 5,082 paid subscribers to the Mental app found a 92.6% session-to-session return rate, with 62.4% of use outside traditional care hours. (link)

⏺️ Review found AI psychological assessment tools are improving slowly but remain poorly validated. Across 320 studies, work concentrated on depression, anxiety, suicidality, and stress. But 71% of studies used neither DSM nor ICD diagnostic criteria, and performance gains from 2014 to 2025 were only marginal. (link)

⏺️ Review concluded AI in child and adolescent psychiatry is not ready for routine clinical use. The narrative review concluded that most tools remain early-stage, that AI scribes were impractical in multiparty visits, and that chatbots offered at best modest benefit with risk of serious harm. (link)

⏺️ RCT found no single first-line treatment for paediatric anxiety was clearly superior. The SMART trial, led by Children's Hospital Los Angeles and published in the American Journal of Psychiatry, randomised children to CBT, fluoxetine, or their combination over 24 weeks. All three performed comparably overall, though child characteristics and treatment sequencing affected who responded best. (link)

⏺️ Double-blind trial found personalised brain stimulation reduced suicidal thoughts in depression. Among 60 depressed adults, HD-tACS tuned to each person's alpha brain rhythm cut suicidal ideation scores more than sham stimulation, with the effect holding at five weeks. The sample was small, and researchers noted that the effect could not be fully separated from broader antidepressant improvement. (link)

⏺️ Observational study suggested at-home subcutaneous ketamine may be safe and effective for depression, anxiety, and PTSD. Researchers reviewed charts from 3,870 patients who self-administered subcutaneous ketamine at home under telehealth supervision. After six weekly sessions, PHQ-9 scores fell from 14.64 to 6.30, GAD-7 from 13.06 to 6.09, and PCL-5 from 46.7 to 27.5, with large effect sizes. More than 80% of patients across all three conditions achieved a clinically meaningful improvement. The study was retrospective, uncontrolled, and has not yet been peer reviewed. (link)

⏺️ Swedish study found the month after a first self-harm episode carries the highest suicide risk. Among 77,647 young people aged 10 to 24, 17.3% had further self-harm within a year, 8.3% attempted suicide, and 0.3% died by suicide. Risk was concentrated in the first month after the index event. (link)

⏺️ Systematic review found structured exercise reduced PTSD symptoms across multiple populations. Across 16 randomised trials covering 663 participants, exercise produced a moderate, significant reduction versus usual care or control. (link)

⏺️ Starting psychotherapy was linked to simpler medication regimens in a large Spanish study. Among 86,502 patients in Alicante in a ten-year, real-world cohort, the median number of psychotropic medications fell from five to two after therapy began. Benzodiazepine and antidepressant use both dropped, while total dose stayed broadly stable. (link)

⏺️ Childhood trauma predicted worse recovery in depressed adolescents. In a 26-week follow-up of 150 adolescents with major depressive disorder, emotional abuse and emotional neglect were independently linked to lower odds of a positive treatment response. (link)

⏺️ Study found impulsive psychopathy traits, not psychopathy overall, predicted suicide risk. Among 458 adults at a US urban trauma centre, only the lifestyle facet of psychopathy independently predicted suicide risk after accounting for PTSD and substance use. (link)

⏺️ Review found Gulf states have a major gap in suicide research due to stigma and legal barriers. A Frontiers in Psychiatry systematic review of 50 studies across the six Gulf Cooperation Countries found suicide substantially underrepresented in the research base, driven by stigma, legal prohibitions, and systematic underreporting. Migrant workers were called out as a high-risk subgroup, with suicidal ideation reaching 68% among those facing acute financial distress(link)

That’s all for this week.

Keep fighting the good fight!

Steve Duke

Founder of Hemingway

P.S. Congratulations to Emily Sen, who correctly guessed that the graph below is the annual incidence of recorded ADHD in females by age group.

To enter this week’s competition, tell me what you think the graph below represents.