Pranayama

Yes. There is fairly good modern evidence that slow, controlled pranayama-type breathing can produce measurable physiological and psychological benefits, although the evidence is much stronger for some effects than for the broader traditional claims.

The Shiva Samhita passage you quoted is specifically describing alternate-nostril breathing with breath retention—essentially a form of nāḍī śodhana/anuloma-viloma, though terminology varies by tradition.

What modern research supports

Potential effect

Evidence

What seems to be happening

Lower stress/anxiety

🟢 Moderate

Slow breathing increases parasympathetic/vagal influence and can reduce subjective stress

Lower heart rate

🟢 Moderate–strong

Particularly during and immediately after slow breathing

Higher HRV

🟢 Moderate

Slow breathing, often around ~5–6 breaths/min, can substantially increase respiratory sinus arrhythmia/HRV

Lower blood pressure

🟢 Moderate

Repeated slow-breathing practice can produce modest reductions

Improved attention/emotional regulation

🟡 Moderate

Several studies find improvements, but protocols vary

Sleep

🟡 Moderate

Breathing/relaxation practices may improve sleep quality, particularly when stress is involved

Exercise performance

🟡/🔴 Mixed

Some evidence, but much less convincing than for relaxation/autonomic effects

“Cleansing nadis,” awakening kundalini, etc.

⚪ Not established scientifically

These are traditional yogic claims rather than demonstrated physiological effects

One particularly interesting point for your practice is that you already notice relatively low resting heart rates during meditation. Slow breathing can interact strongly with the cardiovascular system through the respiratory sinus arrhythmia/baroreflex: inhalation tends to accelerate the heart slightly and exhalation slows it. At an appropriate slow breathing frequency, this oscillation becomes quite pronounced.

But the breath retention is a different matter

Your Shiva Samhita instructions aren’t simply “take slow breaths.” They say:

inhale → retain as long as possible → slow exhalation

and repeat this 20 times.

The evidence for slow breathing without prolonged retention is considerably better established than evidence for prolonged kumbhaka. Breath retention produces additional effects—CO₂ rises, oxygen falls somewhat, intrathoracic pressure changes—and those effects aren’t necessarily required to obtain the relaxation/autonomic benefits.

In fact, “as long as possible” shouldn’t be interpreted as a modern training prescription. There is no good evidence that maximizing retention provides proportionally greater health benefits, and excessive retention can cause dizziness or fainting.

There is an especially interesting connection to your meditation

If you’re interested in the Shiva Samhita from the standpoint of surat shabd/nada practice, there’s a potentially useful distinction:

Pranayama may be doing something quite concrete physiologically before meditation begins.

Slow breathing can:

slow respiration → increase vagal influence → alter HRV/baroreflex activity → reduce arousal → make sustained attention easier.

That doesn’t establish the traditional explanation of prāṇa, nāḍīs, or inner sound. But it gives a plausible physiological pathway by which an ancient breathing practice could genuinely make someone feel calmer and more inwardly focused.

And you don’t necessarily need aggressive kumbhaka to test this.

If you wanted to experiment, I’d be more interested in comparing 5–6 slow breaths/minute with gentle alternate-nostril breathing versus your normal meditation, measuring your Apple Watch heart rate/HRV and subjective calmness before and after. That would give you a surprisingly good personal experiment without trying to force the classical “as long as possible” retention.

If you’d like, I can also  look specifically at the clinical studies on alternate-nostril breathing (Nadi Shodhana) rather than pranayama in general—including blood pressure, HRV, anxiety, cognition, and whether the nostril alternation itself appears to matter.