Hi friend,

OK… FASCINATING week in mental health.

Siggy raised $6M and XRHealth acquired Living Brain and reportedly bought two companies holding assets from the former digital-therapeutics company Pear. Remember those?

LifeStance and Acadia announced strong Q3 results, both raising their full-year 2026 guidance, and Ellipsis Health acquired Kouper Health.

In the US, CMS finalised its 2027 Physician Fee Schedule, increasing the value of collaborative mental-healthcare services and Compass Pathways filed for FDA approval of psilocybin for treatment-resistant depression.

NatureHealth launched the Consumer AI and Mental Wellbeing Commission and Flourish AI published data on the performance of their AI mental-health app in a trial with 1,137 university students. Finally, new research shed light on Ketamine treatments and also on the impact of mental health interventions for veterans.

Lots to cover, so let’s get into it.

What happened in mental health last week?

Industry News:

⏺️ Siggy raised $6M to develop AI-supported mental healthcare. The New York startup, founded by the co-founders of Nurx and Journey Clinical, secured pre-seed funding in a round led by Village Global. Siggy users start by providing their history in a free AI-led conversation, before meeting a clinician by video. They can then use an AI chatbot for daily check-ins about their mood, sleep, symptoms and medication. According to the company, this information, alongside optional wearable data, feeds into the Siggy Score, a measure designed to track changes in a patient's mental health over time and help clinicians identify potential concerns between appointments. Siggy accepts commercial insurance and has begun treating paying patients. Siggy plans to use the funding to expand its clinical team and technology. (link)

⏺️ XRHealth acquired Living Brain, plus two companies holding assets from the former digital-therapeutics company Pear. XR Health announced the acquisition of Living Brain, a German medical device manufacturer that uses XR for neurological rehabilitation treatments. In the same week, reporting from Exits & Outcomes declared that XRHealth has acquired PursueCare and Harvest Bio in deals involving approximately $16.3M in shares, with further payments linked to sales targets. The assets include reSET and reSET-O, prescription digital treatments originally developed by Pear Therapeutics for substance-use disorders. The transactions bring addiction-treatment technology into XRHealth's wider digital-health portfolio. The PursueCare and Harvest Bio acquisitions have not been confirmed by XRHealth. (link)

⏺️ Ellipsis Health acquired Kouper Health to expand its AI-powered patient support platform. Ellipsis Health, a US company developing AI voice agents for healthcare, acquired Kouper Health for an undisclosed sum. Ellipsis's AI agent, Sage, uses voice analysis to identify potential signs of anxiety and depression and helps manage patient assessments, medication follow-up and other tasks. Kouper provides a care navigation platform for health systems and provider groups. Through the deal, Sage will be implemented into Kouper’s platform. (link)

⏺️ LifeStance's Q3 revenue rose 25% and profitability increased. The US outpatient mental-health provider reported third-quarter revenue of $455.9M, up from $363.8M a year earlier. Patient visits increased 17% to 2.7 million, supported by 11% growth in its clinician workforce (to 8,863 clinicians). Net income also rose significantly, from $1.1M to $22.4M, and adjusted EBITDA increased 70% to $68.4M. The company raised its full-year financial outlook, reflecting continued growth in demand and improved operating performance. (link)

⏺️ Acadia Healthcare's revenue rose 5.1% as newly opened psychiatric beds helped increase patient volumes. The US behavioural-health operator reported third-quarter revenue of $895.3M (+5% YoY), while net income fell 4.7% to $34.5M. Revenue from acute psychiatric hospitals increased 11.3%, supported by a 7.5% rise in patient volumes as new facilities and additional beds came into use. Acadia added 220 licensed beds during the quarter, including a new 144-bed hospital in Florida. The company raised its full-year revenue and earnings forecasts. (link)

⏺️ Universal Health Services signalled further outpatient acquisitions after its $835M Talkspace deal. During its third-quarter earnings call, the US hospital operator said outpatient behavioral health is a priority for the company and that it would consider additional acquisitions in that space if the price was attractive. UHS plans to develop more intensive outpatient programmes and a virtual version through Talkspace, which it acquired earlier this year. The company also wants to improve referrals between online therapy, outpatient treatment and hospital care. UHS reported that outpatient visits increased around 3% in its existing business and 5% at Talkspace during the third quarter, with management expecting faster growth as the businesses are integrated. (link)

⏺️ CMS finalised its 2027 Physician Fee Schedule, establishing national pricing for digital CBT and increasing the value of collaborative mental-healthcare services. From January, two Medicare billing codes (98978 and 98986) for supplying digital cognitive behavioural therapy through remote therapeutic monitoring will have nationally determined payment rates. However, a separate code for digital mental-health treatments (G0552) will remain contractor-priced. CMS also increased the work value assigned to psychiatric collaborative care and adjusted payments for smoking cessation and brief substance-use interventions. The increases range from 25% to 73% across the main billing codes, although actual reimbursement changes will vary. (link)

⏺️ BrainsWay partnered with Mindful Health Solutions. The medical-device company announced a collaboration with Mindful Health Solutions, whose management organisation supports a network of 27 US clinics. The partnership will promote BrainsWay's Deep TMS treatments, including its accelerated SWIFT protocol, and explore integrating at-home brain-stimulation technology developed by Neurolief. Mindful Health was acquired by Radial in August of this year. (link)

⏺️ Compass Pathways filed for FDA approval of psilocybin for treatment-resistant depression. Reporting from Psychedelic Alpha highlighted that Compass has submitted its application for COMP360, a synthetic psilocybin treatment for adults whose depression has not responded adequately to existing treatments. Earlier this month, nonprofit Usona Institute submitted an application for its own psilocybin treatment, targeting major depressive disorder more broadly. Both companies have received FDA priority-review vouchers intended to accelerate regulatory decisions. (link)

⏺️ A global commission was launched to assess the impact of consumer AI on mental health. In collaboration with NatureHealth, an international group of researchers and public-health experts has launched the Consumer AI and Mental Wellbeing Commission to investigate how AI chatbots and other consumer AI tools affect mental health. The commission will assess evidence of their benefits and risks, examine how these products should be regulated and consider their role in countries with limited access to mental healthcare. It will also incorporate the experiences of people using AI for emotional support. The commission is expected to publish its findings and policy recommendations in 2027. (link)

Research News:

⏺️ AI mental-health app, Flourish, helped limit worsening symptoms and reduce existing distress in a trial of 1,137 university students. Researchers randomly assigned students to use Flourish, an app combining an AI chatbot with psychological exercises, or continue with usual care for six weeks. Across the full sample, Flourish users had significantly better outcomes for stress, depression and anxiety, although the overall effects were small. Among students with initially low symptoms, those using Flourish experienced less deterioration, while those with elevated symptoms generally improved more. In the latter group, 52% of Flourish users experienced a meaningful improvement in depression, compared with 32% of controls. For anxiety, the figures were 51% and 38% in the control. The findings are from a working paper authored by many researchers with ties to Flourish, and longer-term impacts remain unknown. (link)

⏺️ A survey found three-quarters of US parents were uncomfortable with AI-led therapy for their children. A survey of 572 parents with children aged 5–17 found that 76% were uncomfortable with the idea of AI leading their child's behavioural therapy, while 16% were somewhat comfortable and 8% were very comfortable. Parents placed greater importance on diagnostic accuracy (59%), empathy (49%), recommendations from medical professionals (47%) and privacy (44%) than on cost or shorter waiting times. Parents who trusted AI-led healthcare for themselves were more likely to accept it for their children, but general AI use was not associated with greater comfort. The study used a non-random sample, so its findings may not represent all US parents. (link)

⏺️ Most previously published findings about AI models weakened or disappeared when tested on newer systems. Researchers from Stanford wrote an interesting piece where they revisited 50 widely cited findings about large language models and retested them across 20 models. They found that 44% of the original findings had largely disappeared, while another 34% had weakened substantially. These included problems such as fabricated references and failures in reasoning. While this was not specific to LLM performance in mental health contexts, it highlights the pace of change in the underlying technology and the challenges that poses for research. (link)

⏺️ Many psychiatric inpatients reported using AI chatbots without discussing them with clinicians. Researchers surveyed 100 adults preparing to leave a psychiatric hospital and found that 61% had used an AI chatbot, including 30% who had used one for mental-health purposes. Among those using AI for mental health, 80% viewed the experience positively and 37% had used a chatbot during a psychiatric crisis. However, only 13% had openly discussed their use with healthcare providers. The findings come from a small survey but add further evidence to the use of AI for mental health purposes and pose questions about if and how disclosure of such use should be encouraged. (link)

⏺️ A digital programme reduced anxiety during pregnancy, with some benefits lasting after birth. In a UK randomised trial involving 268 pregnant women at risk of worsening anxiety, researchers tested RELAX, an online programme designed to reduce negative thinking. Participants completed 12 short sessions over four weeks alongside usual maternity care, while the comparison group received usual care alone. After eight weeks, anxiety scores were 1.5 points lower on a 21-point scale among those offered RELAX. Only 1.5% experienced a clinically meaningful increase in anxiety, compared with 8.2% receiving usual care. Some benefits for anxiety and depression remained after birth, although the study relied on participants' own reports and did not find improvements in everyday functioning. (link)

⏺️ School-based mindfulness training showed little long-term benefit for pupils' mental health. A UK randomised trial involving 8,376 pupils and 679 teachers across 85 schools found that mindfulness lessons offered no significant advantage over standard social-emotional education in reducing depression risk or behavioural difficulties after two years. Pupils in the larger study group reported a small improvement in wellbeing, while teachers reported lower stress and a more positive school environment. However, these benefits were modest, and some were not significant in additional analyses. Follow-up was also disrupted by COVID-19, with only 42% of pupils in the larger group providing data. The findings suggest that universal mindfulness programmes may offer limited additional value over existing school-based mental-health education. (link)

⏺️ Mood stabilisers were linked to fewer relapses than antipsychotics alone in early bipolar disorder. A blog post from Mental Elf analysed a recent paper from Szmulewicz et al (2026) that reviewed health records from 371 patients following a first episode of mania with psychosis. Over two years, the estimated risk of relapse was 42.3% among those taking mood stabilisers, 42.2% among those receiving mood stabilisers alongside antipsychotics and 49.5% among those taking antipsychotics alone. The differences were driven partly by fewer depressive relapses. Treatment was not randomly assigned, so other differences between patients may have influenced the results. (link)

⏺️ Statin use was linked to a 20% lower risk of depressive symptoms in older adults. Researchers analysed data from more than 20,000 older adults followed for a median of 8.4 years, including 17,231 without depression at the start of the study. Among those initially free of depression, statin use was associated with a 20% lower risk of developing depressive symptoms, with stronger associations among people taking statins for longer and those with greater genetic susceptibility to depression. However, statins were not associated with an overall improvement among people who already had depression. The study was observational, and the researchers called for randomised trials to confirm the findings. (link)

⏺️ International experts reached agreement on just one-third of recommendations for ketamine and esketamine treatment. A panel of 21 psychiatrists convened by the American Society of Clinical Psychopharmacology reviewed 105 questions about using ketamine and esketamine to treat mood disorders, reaching consensus on only 35 (33%). Experts remained divided on issues including high-dose ketamine infusions, oral and intramuscular formulations, ketamine-assisted psychotherapy and treatment beyond major depression. The authors called for further research to establish clearer standards for prescribing and monitoring. We will be releasing a special report on ketamine treatment for treatment-resistant depression very soon, so if this is of interest to you, make sure to keep an eye out for our announcement. (link)

⏺️ Brain stimulation showed promise for extending ketamine's antidepressant effects. A small randomised trial in Taiwan tested whether intermittent theta-burst stimulation (iTBS), a form of non-invasive brain stimulation, could help sustain the benefits of ketamine in people with treatment-resistant depression. Researchers assigned 38 patients who had initially responded to ketamine to receive either five days of active brain stimulation or a placebo procedure. Those receiving active stimulation had significantly lower depression scores overall during follow-up, which lasted up to 49 days. Remission rates were also higher at certain points, but the treatment did not significantly extend the time patients maintained their initial response. (link)

⏺️ Suicide prevention programmes reduced suicide attempts by 32% among US military personnel and veterans. A systematic review of 47 studies, including 41 in a meta-analysis covering more than 5.5 million participants, found that psychosocial interventions reduced the risk of suicide attempts by 32% and self-harm by 21% compared with control groups. The programmes did not significantly reduce suicidal thoughts, and there was insufficient evidence to determine whether they prevented suicide deaths. The findings suggest that psychological and behavioural interventions can reduce suicidal behaviours even when they do not produce measurable improvements in suicidal thoughts. (link)

⏺️ US veterans who used Veterans Health Administration services had lower suicide rates after leaving the military. Researchers analysed records from nearly 650,000 US Army veterans who had served in Iraq or Afghanistan. The age-adjusted annual suicide rate was 28.8 per 100,000 person-years among those who used Veterans Health Administration services, compared with 35.9 among nonusers, a difference of approximately 20%. The gap was greatest during the first two years after military separation. However, the observational study cannot establish whether using VA services directly reduced suicide risk. (link)

That’s all for this week.

Keep fighting the good fight!

Steve Duke

Founder of Hemingway

P.S. If you find these updates valuable, consider becoming a Hemingway Pro Member.

Keep reading