How StackingRight works

An interaction rule is only added if it clears three bars. First, the mechanism must be explainable in plain language — a shared intestinal transporter, an enzyme induction, a direct binding in the gut, an additive pharmacological effect. If we cannot say why it happens, we do not ship it.

Second, there must be a source: either an NIH Office of Dietary Supplements fact sheet or a specific indexed study. That citation is displayed on every result and on every interaction page, and it is the same link we use in the page markup for machine-readable citations.

Third, the rule must be actionable. Every entry carries a severity, and where separation matters, a minimum number of hours. A rule that tells you something is true but not what to do about it is not useful.

Severe means the combination can cause therapeutic failure or a clinically significant adverse event. Vitamin K with warfarin, potassium with ACE inhibitors, and St. John’s Wort with SSRIs all sit here. These are not schedule tweaks; they need prescriber involvement.

Moderate means absorption or drug levels are meaningfully affected, but the problem is manageable with correct spacing or dose. Calcium with iron is the archetype: both work, but not in the same hour.

Mild means the effect is small or situational, usually only relevant at high doses. Synergy means the opposite of a conflict — the pairing is more effective taken together. Vitamin C with iron, and vitamin D3 with K2, are synergies.

The score starts at 100 and is reduced by five weighted terms. Critical conflicts remove up to 25 points, scaled by how many of your total item pairs they represent. Moderate conflicts remove up to 12. Intakes above the tolerable upper limit remove up to 15, scaled by the share of tracked nutrients affected. Duplicate ingredients remove up to 10. Timing conflicts remove up to 8.

Up to 5 points are added back for seven-day adherence, because a perfectly designed stack you do not take scores worse in practice than a good stack you do.

Every deduction is returned individually with a plain-language explanation, so the score is never a black box. If you cannot see why the number moved, the score has failed at its job.

Nutrient totals are computed from a reference table of the amounts each product delivers per standard serving. Where you enter an explicit dose in a unit that matches the nutrient’s own unit, that dose overrides the reference amount for the primary nutrient the product supplies.

Upper intake levels are compared against the total from every product in your stack. Where a nutrient carries a supplemental UL — vitamin A, vitamin E, niacin, folate and magnesium — the comparison applies to the supplement portion, which is what this tool measures.

The totals do not include the nutrients in your food, except for the food items you explicitly add. For nutrients such as magnesium and potassium, diet supplies most of the total, so read the supplement figure as the supplemental portion.

Each substance resolves to a timing rule with a preferred window, a with-or-without-food instruction, items to pair with and items to avoid. Where the user has expressed their own preference, that wins — the engine respects what someone will actually do.

Spacing checks compare windows rather than clock times, assuming roughly five hours between adjacent windows. Where an interaction rule requires more separation than the arrangement provides, the planner emits a warning naming the two items and which interval fell short.

It does not check drug-drug interactions between two prescription medications. That needs a full pharmacological database and clinical context, and your pharmacist has the authoritative tool. Ask them.

It does not diagnose, and it does not replace a blood test. A normal supplement schedule says nothing about whether your ferritin, B12 or vitamin D level is adequate. Only measurement does that.

It does not account for your individual genetics, absorption disorders, kidney or liver function, or pregnancy status. Those change the calculus substantially, which is why the output is framed as questions to bring to a clinician.

Health content pages carry a last-reviewed date and a reviewer line. Rules are revisited when the underlying source material changes or when a credible correction is raised.

If you find a rule that is wrong, outdated, or missing a citation, tell us. A wrong rule is worse than no rule, and the whole value of this product depends on getting the sourcing right.