The evidence behind the techniques in relif

How to read this page

Every study cited here is evidence about a technique category — slow breathing, grounding and relaxation, cognitive restructuring, expressive writing, conversational agents. None of it is evidence about relif. relif has never been the subject of a clinical trial, holds no regulatory clearance, and is not a medical device. When you see “evidence-informed” on this site, that is the distinction it is carrying.

We also try to report the awkward parts of each source, because a page that only quotes the flattering sentence is not an evidence page.

Paced and exhale-weighted breathing

This is the best-supported category in the app. In a randomised controlled study published in Cell Reports Medicine in 2023, Balban and colleagues at Stanford randomised 108 participants to five minutes a day of one of three breathwork practices or mindfulness meditation, for 28 days. The cyclic sighing arm — a double inhale followed by a long, slow exhale — produced greater daily mood improvement and a significantly larger reduction in respiratory rate than the meditation arm, with the mood benefit growing over the course of the month.

The mechanism has its own literature. Zaccaro and colleagues’ 2018 systematic review in Frontiers in Human Neuroscience concluded that slow-breathing techniques “act enhancing autonomic, cerebral and psychological flexibility”, linking increased heart-rate variability and parasympathetic activity with EEG changes (more alpha, less theta) and with self-reported calm and comfort in healthy subjects.

What this does not show. Both studies used healthy volunteers doing daily practice, not people mid-panic. Neither tested a phone app. Neither supports a claim that a breathing exercise stops a panic attack, and this site does not make one. relif’s breathing exercise implements the patterns — 4-7-8, box, 3-3-3 and coherent 5-5 — with an animated pacer and phase-synced haptics, and nothing more is claimed for it than that.

Grounding and relaxation techniques

The NIH’s National Center for Complementary and Integrative Health summarises the relaxation-technique literature and reports that a 2018 review found relaxation therapy reduced symptoms across generalised anxiety disorder, social anxiety disorder and panic disorder, and that heart-rate-variability biofeedback shows promise for self-reported stress and anxiety.

The same page is the reason we include a caution most app marketing omits: relaxation techniques are “generally considered safe for healthy people”, but there are rare reports of increased anxiety, intrusive thoughts or fear of losing control, and some techniques “might cause or worsen symptoms in people with epilepsy or certain psychiatric conditions, or with a history of abuse or trauma”. If a technique makes you feel worse, stop it.

What this does not show. The specific 5-4-3-2-1 sensory script that appears in every anxiety app — including ours — is a clinical convention with very little direct trial evidence of its own. It sits inside a broader relaxation and attention-shifting literature. We describe it as something people commonly use, not as a proven intervention.

Cognitive reframing

Cognitive behavioural therapy is the most heavily studied psychotherapy there is. Hofmann and colleagues’ 2012 review in Cognitive Therapy and Research examined 269 meta-analyses and concluded that “the evidence-base of CBT is very strong”, with the strongest support of all for anxiety disorders — large effect sizes for obsessive-compulsive disorder and at least medium effect sizes for social anxiety disorder, panic disorder and PTSD.

What this does not show. That is evidence for therapist-delivered CBT, not for a self-guided reframing exercise in an app. relif’s Thought Reframe is an eight-step question flow inspired by Byron Katie’s The Work. It borrows the shape of cognitive restructuring — examine the thought, test it, consider an alternative — without being a CBT protocol, and we do not describe it as therapy.

Expressive writing and journaling

The expressive-writing paradigm developed by James Pennebaker asks people to write about a difficult experience for 15–20 minutes across three to five sessions. Baikie and Wilhelm’s 2005 review in Advances in Psychiatric Treatment catalogues the reported outcomes: fewer stress-related medical visits, improved immune markers and lung function, better mood, reduced depressive symptoms and fewer intrusion and avoidance symptoms.

What this does not show. The same review states plainly that psychological benefits “are not as robust or as consistent as those for physical health”, that findings vary a lot across clinical populations, and that writing about something painful often makes people feel worse immediately afterwards before any longer-term benefit appears. relif’s journals are prompt-led and short — gratitude, achievements, free writing — which is a gentler format than the classic trauma-writing protocol, and also a less studied one.

AI-delivered conversational support

The 2023 systematic review and meta-analysis by Li, Zhang, Lee, Kraut and Mohr in npj Digital Medicine pooled 15 randomised controlled trials covering 1,744 participants. AI-based conversational agents improved depression with a Hedge’s g of 0.64 (95% CI 0.17–1.12) and psychological distress with a g of 0.70 (95% CI 0.18–1.22) — moderate effects, with wide confidence intervals.

What this does not show. The authors are direct about the biggest gap: with only six studies reporting follow-up and wide variation in follow-up length, they could not meta-analyse long-term effects, and “the lasting effects of CA interventions remain unclear”. The trials also tested rule-based and scripted agents alongside generative ones; they are not a verdict on today’s large language models. And none of them tested Liam. Liam is an AI companion, not a therapist, and we make no efficacy claim for it.

What “21 days” means here

relif’s Journeys run for 21 days. That is a structure — three weeks of daily practice, with milestones at days 7, 14 and 21 — chosen because it is long enough to build a rhythm and short enough to finish. It is not a claim that habits take 21 days to form. That figure comes from a 1960s popular-psychology book, not from habit research, and the research that does exist found a far longer and far more variable timeline. Any page that tells you “science says 21 days” is repeating folklore.

Limitations — read this part

• No trial of relif. There is no randomised controlled trial of relif, no published outcome data on relif users, and no regulatory clearance. Nothing on this site should be read as evidence that relif works.
• Category evidence does not transfer automatically. A five-minute daily breathwork protocol studied in a lab is not the same as tapping a breathing exercise on a phone once a week.
• Effect sizes come with wide intervals and short follow-ups. The AI-companion meta-analysis could not establish durability at all.
• Publication bias is real in this field, and small digital-mental-health trials are frequently run by the people who built the product.
• Techniques can backfire. The NIH notes rare increases in anxiety and specific cautions for people with epilepsy, some psychiatric conditions, or a trauma history.
• relif does not diagnose, treat, cure or prevent any condition, and it is not a substitute for professional care. If anxiety is affecting your daily life, speak to a qualified healthcare professional; the NIMH overview of anxiety disorders is a reasonable place to start.

Editorial policy

Who writes these pages. The relif team writes them. We are a small product team, not clinicians, and this site carries no “medically reviewed by” badge because there is no clinician reviewing it. Claiming one would be exactly the kind of borrowed authority this page exists to avoid.

How we source. Product facts — feature limits, prices, platform versions, what is encrypted — come from the app and its store listings, not from our own marketing. Health and research claims are taken from peer-reviewed papers or official bodies (NIH, FDA, regulators), read directly rather than via a summary, and linked so you can check them. Where a source is weak, mixed or contradicted, we say so on the page.

What we will not publish. Outcome claims about relif (“calmer in two minutes”). Efficacy claims resting on studies of other products. Diagnostic or treatment language. Star ratings or review counts for our own app. Security superlatives like “military-grade”. Comparative claims we cannot substantiate.

Corrections. Every content page carries a “last reviewed” date, and we re-check crisis helpline numbers quarterly. If you find an error — especially a claim that overstates what relif does — email support@myrelif.com and we will correct the page.

Keep reading

• Guided breathing in relifThe four presets, custom patterns, and how the haptic pacing works.
• Liam, the AI companionWhat Liam is, what it remembers, and its safety architecture.
• Thought ReframeThe eight-step reframing flow and where the questions come from.
• Panic attacks: what people do in the momentHow these techniques are used when something is happening right now.

Try the techniques for yourself

Free to download on iOS and Android

Wellness Disclaimer: relif is a wellness and self-care companion designed to support your emotional wellbeing. It is not a medical device, does not provide medical advice, and is not a substitute for professional healthcare, therapy, or counseling. If you are experiencing a medical or mental health emergency, please contact your local emergency services or a crisis helpline immediately.

Editorial policy: Written by the relif team from primary sources — the product itself, store listings and peer-reviewed or official publications. Not medical advice. Read our editorial policy and sources.