For decades, meditation research has quietly assumed that all contemplative states point in the same direction: calm down, relax, let the parasympathetic nervous system take over. A 2022 study published in Current Research in Neurobiology complicates that picture. Researchers from the National University of Singapore and Cardiff University spent time with sixteen highly experienced Vajrayana practitioners in Bhutan — monks and nuns with an average of nearly eight years of retreat training — and wired them up for heart-rate and EEG recordings while they moved through a sequence of Tibetan Buddhist practices. What they found is a meditative state built not on relaxation, but on deliberately induced arousal.
Breathing your way into a different nervous system state
The practice at the center of the study is Tummo, sometimes translated as “inner fire,” a technique that has circulated in Tibetan Buddhism for centuries as part of the Six Yogas of Naropa. It combines forceful “vase breathing” — long inhalations held with abdominal and pelvic contractions, followed by short, powerful exhalations — with visualization of flame rising through the body. The researchers also tracked Yidam, a deity self-visualization practice performed before Tummo, and Mahamudra, a non-conceptual awareness practice performed after.
The heart-rate data told a clear story. Both Yidam and Tummo produced a measurable withdrawal from parasympathetic activity — the classic marker of physiological arousal rather than relaxation. Tummo pushed this further than Yidam did: heart-rate variability in the high-frequency band dropped by 43 percent during Tummo, roughly double the 20 percent drop seen during Yidam, with the forceful breathing pattern also nudging low-frequency activity upward, a sign of rising sympathetic tone. The more wrathful the visualized deity during Yidam, incidentally, the larger the drop — practitioners visualizing peaceful deities showed a 6 percent decrease, compared with 29 percent for fully wrathful ones, suggesting the arousal response tracks the intensity of the mental content as well as the breath.
Two attention systems, not one
The EEG findings map onto a broader theory of how the brain manages attention. Sustained, “tonic” alertness — the kind cultivated in most mindfulness and open-monitoring meditation — runs through the cingulo-opercular network, a system anchored in the cingulate cortex and anterior insula that keeps attention locked onto a task by filtering out distraction. Arousal instead engages a separate, faster system: the fronto-parietal control network, tied to bursts of norepinephrine release from the brainstem’s locus coeruleus, which produces short, sharp spikes of focused, externally-directed attention rather than steady vigilance.
This is where the study’s most interesting comparison shows up. The practitioners performed Mahamudra — a state of non-dual, object-less awareness — twice: once starting from ordinary rest, and once starting from the aroused state built up by Yidam or Tummo. On paper, it’s the same instruction both times: let go of all mental content, don’t fixate on anything. In the brain, the two versions looked almost like opposite processes. Mahamudra performed after rest showed the alpha-power increases typical of relaxed, monitoring-based meditation — the same signature seen in conventional open-monitoring practice. Mahamudra performed after Tummo or Yidam showed the reverse: alpha power dropped, along with beta and gamma activity, producing a state the authors describe as heightened cortical excitability paired with sharply reduced top-down control. Among the ten Tummo-trained practitioners in the sample, this arousal-driven version of Mahamudra was the only one to show that particular signature — evidence, the authors argue, that the forceful breathing itself is doing real physiological work, not just serving as a warm-up ritual.
Why this matters beyond a Bhutanese retreat center
The takeaway isn’t that one style of meditation beats another — it’s that “meditation” isn’t a single physiological category. A practice built around breath-driven arousal recruits different neural machinery than a practice built around relaxation, even when the end goal, as with the two versions of Mahamudra here, is nominally identical. That distinction has been largely invisible to meditation science, which has tended to lump every contemplative technique into a relaxation-response framework borrowed from the 1970s.
It’s also a useful reminder of how old some of these ideas are. Long before “breathwork” became a wellness-industry buzzword, Tibetan practitioners were using structured, forceful breathing to deliberately shift autonomic state — and building entire training sequences around the order in which arousal and stillness are introduced. Wim Hof has pointed to this same lineage of breath-driven physiological control as one of the influences behind the breathing exercises at the center of the Wim Hof Method, and this study gives that broader tradition its most detailed physiological documentation yet: real-time evidence, from EEG and heart-rate recordings rather than self-report, that forceful breathing changes which brain networks are running the show.
The sample here is small and specialized — sixteen practitioners with years of training aren’t a stand-in for a general population, and Tummo experts turned out to have systematically different baseline brain activity from the other participants, which makes it hard to say how quickly these effects would show up in someone new to the practice. But as a demonstration that breath and body state can be used to deliberately steer which attentional system the brain relies on, the findings are hard to dismiss. The next open question, as the authors note, is how much of this is trainable, and how quickly.
.
