The decision tree
The calculator runs four branches in order: hard gates first, then dose range from goal × sex × diet, then form from context, then copper-offset and warning flags.
1. Hard gates (block first):
- Cold-onset goal + hours since first symptom >24 → window-closed flag. Lozenge efficacy is concentrated in the first 24 h per Hemilä's IPD meta-analysis (PMID 28480298); Cochrane 2024 (PMID 38719213) classifies later-start lozenges as low-certainty. The calculator refuses to dose into a closed window — save the lozenges for the next cold.
- Wilson disease and age-related macular degeneration are intentionally NOT in the goal enum. Wilson disease uses 150 mg elemental zinc/day (50 mg TID) for chelation-style copper management per AASLD 2023 guidance (PMID 37184530); AMD uses the AREDS2 80 mg/d formulation per the 10-year AREDS2 Report 28 outcomes (PMID 35653117). Both are physician-supervised and out of scope for a self-serve calculator.
2. Daily elemental zinc range (by goal):
- RDA gap-filling: 15–25 mg/d (men), 15–20 mg/d (women), 1.5× for vegan / PPI
- Cold-onset (within 24 h): 75–92 mg/d as acetate or gluconate lozenges, split across 6–7 lozenges, for 5–7 days
- Acne: 30–45 mg/d for ≥12 weeks
- Athlete: 25–40 mg/d ongoing during heavy training
- Deficiency-context: 20–30 mg/d, scaled 1.5× for vegan, +5 mg/d floor for PPI
The vegan / PPI 1.5× multiplier comes from the IZiNCG bioavailability adjustment: phytate from grains and legumes binds zinc and cuts absorption 40–50% on plant-heavy diets, and chronic proton-pump inhibitor use produces achlorhydria that compromises ionic-zinc absorption 30–50%.
3. Form (bisglycinate vs picolinate vs lozenge):
Bisglycinate is the gen-pop default — 60–70% bioavailable, excellent GI tolerance, no metallic-taste issues. Picolinate (60–65% bioavailable) kicks in for vegan / PPI / athlete cohorts where the chelation premium earns its keep — the picolinate chelate bypasses the gastric-acid step that PPI users have lost. Citrate is a value alternative ($$ vs $$$) with comparable absorption.
Cold protocols use zinc acetate or gluconate lozenges only — they release ionic zinc in the oropharynx where it interferes with rhinovirus replication locally. The calculator excludes citrate / tartrate lozenges because those chelate the zinc ion and nullify the local effect. Intranasal zinc (Zicam-type gels and sprays) is an absolute contraindication — permanent anosmia, FDA warning issued 2009.
4. Copper-offset gate:
Long-term zinc ≥25 mg/d induces copper deficiency by upregulating intestinal metallothionein, which sequesters dietary copper. The calculator adds a 1–2 mg copper (gluconate or bisglycinate) recommendation when daily zinc is ≥25 mg/d, duration is long (≥3 months), and copper is not already being supplemented. Caught early it is reversible; chronic deficiency causes microcytic anemia, neutropenia, and myeloneuropathy — annual ceruloplasmin + RBC monitoring is the safety net.
The gate intentionally suppresses on the cold-onset goal (acute, 5–7 days total) and when the user reports concurrent copper supplementation already.
What the calculator does NOT cover
- Wilson disease (150 mg/d, 50 mg TID) — physician-supervised chelation protocol per AASLD 2023; out of scope for self-serve dosing.
- AMD (AREDS2 80 mg/d formulation) — ophthalmology territory, paired with lutein/zeaxanthin and copper in a fixed ratio per AREDS2.
- IV zinc — reserved for parenteral nutrition and severe burn / sepsis cases. Hospital territory.
- Pediatric dosing — weight-scaled, separate evidence base, deferred to a future pediatric calculator.
Where the math stops being safe to extrapolate
The HDL-suppression / inverted-U immunity warning at ≥150 mg/d sits at the boundary of "supplement" and "drug": zinc at 150–300 mg/d for prolonged periods reliably reduces HDL cholesterol and paradoxically suppresses T-cell-mediated immunity. The calculator surfaces this as a non-blocking warning rather than a hard gate because the only legitimate indication at that dose is Wilson disease — and Wilson disease patients should be on this calculator's "see your physician" path, not its dose path.
The 24 h cold-onset window itself is a probabilistic boundary, not a step function. Hemilä's IPD analysis pooled trials with start times ranging from 0–24 h; pooled effects past 24 h are inconclusive rather than zero. The calculator chooses to close the window at 24 h because the dose × duration × ionic-zinc-side-effect tradeoff stops favoring the lozenge protocol when the recovery curve is already on its descending limb.