Breathing, craving and recovery.

A useful tool needs a clear job.

When an urge arrives, the immediate task can be very small: notice it, pause and choose the next action. Breathing is one possible part of that pause. It does not need to erase the urge to be worth exploring.

This is ORIGIN's practical interpretation, not a finding that our programmes prevent relapse. The research below concerns specific breathing interventions and populations. It does not validate ORIGIN, the Human Foundations framework or a particular app.

The clearest signal is a small reduction in craving.

A July 2026 systematic review in Addiction identified 20 studies of controlled breathing and heart rate variability biofeedback in people using alcohol, nicotine or other drugs. Fifteen contributed to meta-analyses. Across 12 studies with 696 participants, craving was lower than in control conditions: a small average effect, Hedges' g = 0.30, with a 95% confidence interval of 0.12–0.48. [1]

Confidence remains limited. The craving analysis combined randomised and non-randomised studies, outcomes were often subjective, and long-term follow-up was limited. The authors rated craving evidence as low certainty. Withdrawal findings were inconclusive, while dependence findings rested on only two small, heterogeneous studies.

No clear difference was detected between HRV biofeedback and other controlled breathing approaches. That is not proof that they are equivalent. Most importantly, less craving is not automatically less substance use, sustained abstinence or fewer relapses.

A higher HRV reading is not proof of recovery.

An October 2026 analysis examined 962 men from a military cohort during five minutes of spontaneous breathing followed by five minutes at six breaths per minute. Short-term HRV increased during paced breathing. A higher starting respiratory rate was associated with a larger response; injury status, physical activity and walking performance did not significantly moderate it. [2]

The study measured an acute physiological response, not addiction recovery or lasting psychological improvement. Everyone completed the conditions in the same order, without randomised controls. Its all-male military sample also limits generalisation.

For ORIGIN, the distinction matters: a changed physiological signal, feeling calmer and functioning better are different outcomes. We should measure and describe them separately. The study does not establish a breathing-rate threshold for personalising treatment.

Recovery extends beyond the exercise.

A July 2026 review of relapse after residential substance use treatment included 61 studies and 18,693 participants. Higher depressive symptoms and distress, and lower self-efficacy, were associated with poorer outcomes. Peer support and longer treatment duration were associated with better outcomes. [3]

Almost all included studies were cohorts. Differences in participants, treatment and follow-up were substantial, alcohol studies dominated, and confounding makes causal claims unsafe. The review cannot tell us that increasing one factor will reliably prevent relapse.

Our interpretation is to place a breathing exercise inside a wider support plan: continuity of care, human connection and a realistic next step. An app session should not become a reason to postpone established treatment or disconnect from people who can help.

Adding more techniques is not automatically better.

A September 2026 placebo-controlled experiment combined slow-paced breathing with active or sham prefrontal brain stimulation, known as tDCS. Some physiological and emotional measures responded to each intervention, but the combination did not consistently enhance outcomes and weakened some effects. [4]

This was an acute laboratory experiment, not a clinical recovery trial. It does not show that combinations are generally harmful. It does challenge the assumption that two plausible interventions must work better together. ORIGIN should test combinations rather than market their presumed synergy.

Turn the pause into one concrete action.

The following is an ORIGIN coaching suggestion, not the protocol of the reviewed studies and not a clinically validated relapse-prevention sequence.

• Notice what is happening: name the urge, situation or feeling without treating it as an instruction.
• If it feels comfortable, take a brief pause with gentle, unforced breathing. Do not force a slow rate, deepen every breath or hold your breath to chase a sensation or score.
• Choose one action from your existing recovery plan: contact a trusted person, move away from a high-risk situation or use a skill agreed with your treatment provider.
• Judge usefulness by what happens next, not only by a wearable reading. Record whether the practice helped you follow your plan.

Know the boundary.

Breathing exercises are optional supportive practices, not a substitute for medical care, psychological treatment or medication when indicated. Stop if the practice causes dizziness, faintness, escalating panic, unusual breathlessness or chest pain. Return to comfortable breathing and seek appropriate help for concerning symptoms.

Do not attempt to manage suspected alcohol or sedative withdrawal with breathwork: withdrawal can require urgent medical assessment. If there is immediate danger, use emergency services rather than an exercise.

Editorial note: this is a selected research commentary, not a systematic review or an independently peer-reviewed article. It draws on peer-reviewed papers published from July to October 2026. Evidence assessments and practical recommendations are ORIGIN's interpretation; they should not be confused with the authors' findings or formal clinical guidance.