The therapist made them heal.
Even across the distance of a screen, the human bond at the heart of therapy proves to be the mechanism of healing rather than its backdrop. Researchers at Hebrew University have confirmed, with careful rigor, that guided internet-based cognitive behavioral therapy meaningfully reduces depression and insomnia — and that the quality of the relationship between therapist and patient, more than any program feature, determines whether symptoms actually improve. The finding invites a broader reckoning: as mental health care moves into digital spaces, the ancient need for felt human presence does not quietly recede.
- Depression and insomnia remain among the most undertreated conditions globally, and the promise of scalable digital therapy has long outpaced the evidence for what actually makes it work.
- A Hebrew University study of thirty-nine participants found that internet-based CBT produced significant symptom improvements — but the data revealed a critical tension between what keeps people engaged and what genuinely heals them.
- The therapeutic alliance with the human therapist predicted symptom reduction, while the patient's bond with the program itself predicted whether they stayed or dropped out — two distinct forces pulling in the same direction for different reasons.
- Patients with avoidant attachment patterns showed meaningful shifts toward greater ease with connection, suggesting that even a digitally constrained therapeutic relationship can quietly reshape deep relational habits.
- Researchers are now urging clinicians to monitor therapeutic relationships weekly in digital settings, treating the human connection not as a soft supplement but as the active clinical ingredient requiring active clinical attention.
A study led by Dr. Dina Zalaznik at Hebrew University set out to test something therapists had long suspected but rarely measured with precision: that human connection shapes outcomes even when therapy unfolds through a screen. Thirty-nine participants moved through a six-module internet-based cognitive behavioral therapy program for depression, guided weekly by supervised doctoral students who assigned homework and checked in. The modules covered psychoeducation, cognitive techniques, behavioral activation, thought challenging, optional insomnia work, and relapse prevention. Across all measured variables — depression, insomnia, and the quality of therapeutic relationships — improvements were significant.
The study's deeper contribution was in parsing what drove those improvements. Two forms of alliance were tracked: the patient's relationship with their therapist, and their relationship with the program itself. Both influenced whether someone stayed engaged or dropped out. But only the human alliance — the bond with the therapist on the other end of the screen — predicted actual symptom improvement. The program kept people coming back. The therapist made them heal.
A secondary finding added texture to the picture. Participants with avoidant attachment styles — those who habitually pull away from closeness — showed meaningful improvement in their attachment patterns over the course of treatment. Those with anxious attachment shifted less. This was notable given that internet-based therapy deliberately limits relational intensity, yet something in the work still moved avoidant participants toward greater ease with connection.
The researchers' practical recommendation is direct: in digital therapy, the therapeutic relationship is not a background condition but the primary mechanism. They call for weekly questionnaires to monitor how patients experience both their therapist and the program, with room to adjust when either relationship weakens. The findings point toward a future in which internet-based therapy is understood not as a lesser substitute for in-person care, but as a distinct modality — one where the therapist's presence, however mediated, remains the irreducible center of the work.
Researchers at Hebrew University have documented what many therapists suspected but few had rigorously tested: that the human connection matters even when the therapy happens through a screen. A study led by Dr. Dina Zalaznik, working in Prof. Jonathan Huppert's laboratory for mental health treatment and research, tracked thirty-nine participants—mostly women, averaging thirty-eight years old—as they moved through a six-module internet-based cognitive behavioral therapy program for depression. The modules covered the standard ground: psychoeducation, cognitive techniques, behavioral activation, thought challenging, optional insomnia work, and relapse prevention. Supervised doctoral students guided them weekly, assigned homework, and checked in. What emerged from the data was not a surprise, exactly, but a confirmation with weight: depression and insomnia symptoms improved significantly across the board.
The study's real insight came from examining what made people stick with the program and what actually changed their symptoms. Two kinds of alliance mattered—the relationship between patient and therapist, and the relationship between patient and the program itself. Both predicted whether someone would drop out or stay engaged. But only one predicted whether their depression actually got better. The alliance with the therapist, the human on the other end of the screen, was the one that moved the needle on symptom improvement. The program itself kept people coming back. The therapist made them heal.
Another finding surprised the researchers less but still warranted attention: attachment styles—the deep patterns people carry about closeness and distance—responded differently to the treatment. Those with avoidant attachment, people who tend to pull away from connection, showed significant improvement in their attachment scores. Those with anxious attachment, people who tend to cling, did not shift as much. This mattered because internet-based therapy, by design, limits the therapist's involvement and keeps the focus on cognitive and behavioral work rather than the relationship itself. Yet even in that constrained setting, something about the work seemed to help people with avoidant patterns move toward greater ease with connection.
The study used a culturally adapted Hebrew version of the program, recruited participants through advertisements for a free trial, and screened them with online surveys and phone interviews. Dropout rates and adherence were tracked alongside symptom change. The researchers found meaningful, noticeable improvements across all measured variables—depression, insomnia, and the quality of the therapeutic relationships themselves.
What clinicians should take from this, the researchers suggest, is straightforward: in digital therapy, the relationship is not a luxury. It is the mechanism. They recommend weekly questionnaires to monitor how patients experience both the therapist and the program, with the flexibility to adjust when those relationships weaken. The findings point toward a future where internet-based therapy is not treated as a lower-tier alternative to in-person work, but as a distinct modality with its own requirements—one where the therapist's presence, even at a distance, remains the active ingredient.
Notable Quotes
The alliance with the therapist was significantly related to symptom improvement, while both alliances played a role in treatment adherence and dropout rates— Study findings from Dr. Dina Zalaznik's research at Hebrew University
The Hearth Conversation Another angle on the story
Why does the therapist-patient relationship matter more than the patient's relationship with the program itself, if the program is doing the actual cognitive work?
Because knowing someone is paying attention to you, that your struggle matters to a real person, changes what you're willing to try. The program is the map. The therapist is the one who believes you can walk it.
But these are supervised doctoral students, not experienced clinicians. Does that change the finding?
It might actually strengthen it. If even a graduate student, checking in weekly and caring about your progress, can predict symptom improvement, that tells you something about the power of basic human attention in healing.
The avoidant attachment finding is interesting—people who pull away got better at connection through a program that doesn't focus on relationships. How does that work?
The cognitive and behavioral work itself seems to shift something. Maybe when you're not forced into intimacy, you can practice closeness at your own pace. The program gives you tools; the therapist gives you permission to use them.
What should a clinic actually do with this information?
Start asking patients weekly: Do you feel heard by your therapist? Does the program make sense to you? Then listen. If the answer is no, something needs to change before the patient disappears.