You don't have a sleep duration problem. You have a sleep architecture problem. The hours are on the clock, but the slow-wave depth isn't on the EEG. Cortisol stayed elevated past 10 p.m., sympathetic tone never fully released the brakes, and the brain skipped straight from light Stage 2 to fragmented REM without ever bottoming out into the delta-dominant Stage 3/4 window where actual restoration happens. Phase 1 isn't about sedating you harder. It's about reinstalling the descent — the parasympathetic drop-off at sleep onset, and the delta entrainment that should follow within the first 90 minutes. Two peptides do that work: DSIP owns the descent into deep sleep, Selank owns the anxiolytic landing pad before the descent. Run them together for four weeks before touching anything else.
What's actually happening at the cellular level
Delta Sleep-Inducing Peptide is a nine-amino-acid neuropeptide that the practitioner corpus treats as a sleep-architecture modulator rather than a sedative. It doesn't hit GABA receptors and it doesn't knock you out. The mechanism the corpus consistently cites is delta-wave entrainment: DSIP appears to interact with NMDA-receptor signaling in a way that promotes the synchronized cortical firing pattern that defines Stage 3/4 sleep. The practical observation that shows up repeatedly in high-volume peptide practices — and in the n=1 reports the corpus aggregates — is that deep sleep percentage on a tracked sleep stage chart climbs from a typical baseline of 12-18% into the 22-30% range within the first 2 weeks of nightly dosing. Critically, DSIP does not suppress REM in the way prescription hypnotics do, which is why the practitioner consensus tolerates nightly use without the architectural distortion you get from Z-drugs or benzodiazepines.
Selank works upstream of the descent. It's a synthetic tuftsin analog — a four-amino-acid peptide that modulates GABAergic tone without binding the benzodiazepine site on the GABA-A receptor. That distinction matters: you get the anxiolytic effect (reduced sleep-onset rumination, blunted sympathetic arousal in the pre-sleep window) without sedation, cognitive impairment the next morning, or the addiction/withdrawal arc that defines the benzo class. The corpus describes Selank as restoring expression of GABAergic genes that chronic stress downregulates — meaning the mechanism is partly epigenetic, not just acute receptor occupancy. That's why the effect compounds over weeks rather than tapering off.
The reason these two stack rather than redundantly sedate: Selank quiets the cortical noise that prevents sleep onset, and DSIP shapes the architecture once you're under. One is the runway, the other is the descent.
Phase 1 protocol — Week 1 through Week 4
| Compound | Dose | Route | Frequency | Evidence Tier | Notes |
|---|---|---|---|---|---|
| DSIP | 100–300 mcg | Subcutaneous | 1x nightly, 30–45 min pre-sleep | 🟢 Expert | Start at 100 mcg for first 5–7 nights. Titrate to 200 mcg if delta % on tracker hasn't moved by night 7. 300 mcg is the practical ceiling — higher doses don't add architecture, they just truncate REM. |
| Selank | 200–400 mcg | Intranasal | 1–2x/day | 🔵 Clinical | Morning dose handles daytime anxiolysis. Optional second dose 60–90 min pre-sleep on high-arousal nights. Cycle 30 days on / 10 off. |
Reconstitution and handling: DSIP comes lyophilized — reconstitute a 5 mg vial with 2.5 mL bacteriostatic water for a 2,000 mcg/mL concentration. At that concentration, 100 mcg = 5 units on an insulin syringe, 200 mcg = 10 units, 300 mcg = 15 units. Selank is typically supplied pre-mixed as a nasal spray at 100 mcg per spray, or as a powder you reconstitute and dose with a metered nasal applicator. Refrigerate both after reconstitution. Stable for 28 days reconstituted.
Injection site for DSIP: Subcutaneous abdomen-near, rotate sites within a four-quadrant pattern across the 28-day cycle. The peptide is small and well-tolerated — local irritation is rare.
Timing logic: DSIP works best when you take it 30–45 minutes before lights-out, not at lights-out itself. The peptide needs to reach the central nervous system before the first sleep cycle begins. Selank's evening dose works on a similar timeline — give the GABA modulation time to dampen sympathetic tone before you ask the brain to descend.
What you should feel — week by week
- Night 1–3: Selank effect is faster than DSIP. Expect noticeably reduced sleep-onset rumination — the "monkey mind" at the start of the night quiets. DSIP effects begin in this window for some, but the architecture shift is subtle on the first few nights and easy to miss without a tracker.
- Week 1: Subjective sleep quality improves. You wake up less between 2–4 a.m. (the cortisol-pulse window). Morning HRV starts to climb from baseline.
- Week 2: This is when the deep-sleep architecture shift becomes measurable on whatever sleep tracker you use (Oura, Whoop, Garmin, EEG-grade like Dreem). Deep sleep percentage moves from baseline into the 22–30% range. This is the milestone — if your tracker shows no change in delta/deep stage minutes by end of Week 2, your dose is too low or the substrate isn't biologically engaging. Titrate DSIP up.
- Week 3–4: Anxiolytic effect from Selank consolidates. Daytime stress reactivity drops — not because you're sedated, but because GABAergic baseline expression has shifted. Sleep onset latency continues to compress; many users report falling asleep within 10–15 minutes of lights-out where the baseline was 30–60.
What's NOT happening yet
- You haven't fixed cortisol. Phase 1 reshapes the night. The 4 a.m. cortisol awakening that defines HPA-axis dysregulation isn't addressed here — that's Phase 2 (Week 5–10) with the supporting hormone-axis stack. If you still wake at 3:30–4:30 a.m. on Week 4, that's expected. Don't stack more sleep peptides at it.
- You haven't boosted growth hormone yet. DSIP has a downstream nudge on the GH axis as a known secondary effect, but it's not a GH secretagogue. Don't expect the body-composition changes that come with the Phase 2 CJC/Ipamorelin layer.
- You haven't restored REM percentage. Phase 1 protects REM (DSIP doesn't suppress it), but it doesn't increase it. If your REM is suppressed from chronic alcohol, SSRIs, or shift work, that recovery is a longer arc.
- You're not "fixed" if you stop here. Selank is on a 30-on / 10-off cycle by design — the receptor downregulation risk at continuous use is real per the corpus. DSIP can be run more continuously, but the full reset requires the cortisol-axis layer in Phase 2.
- Your subjective sense of "I slept better" lags the data. The architecture changes before you feel the cognitive payoff. Track the numbers. Trust the data ahead of the feel.
Research describes this. Track it. Adjust.