Pregnancy, GLP-1 use, and perimenopause. Reinstated 9 September 2026 at the founder’s instruction, after the previous round removed them along with the wider medical scoping. The scope given: treat these groups in a supportive, informative, factual way — not as a clinical product.
State a published reference fact, and route to a professional. Never assess the person. The worked example she gave: in a pregnancy context, iron carries an information marker reading roughly “Iron requirements are considered higher during pregnancy. If you are unsure whether your intake is sufficient, see a doctor.”
The distinction the pattern rests on: “requirements are considered higher during pregnancy” is a general, citable, public-health statement. “Your iron is low” is an assessment of an individual. The pattern stays entirely in the first category.
A convenient property of the German pregnancy guidance: the official sources themselves route to a professional for several nutrients — so quoting them accurately produces the referral automatically, rather than emealia adding a disclaimer on top. See the pregnancy section.
Germany has an official, named body publishing specific figures: the Netzwerk Gesund ins Leben, under the Bundeszentrum für Ernährung (BZfE), supported by German scientific societies and professional associations.
| Nutrient | Published recommendation | What the source itself says about professional involvement |
|---|---|---|
| Folic acid | 400 µg/day as a supplement, from planning through week 12; 800 µg/day if starting late | General recommendation, no individual diagnosis required. |
| Iodine | 100–150 µg/day throughout pregnancy | “Bei Schilddrüsenerkrankungen soll dies nur nach ärztlicher Rücksprache erfolgen” — only after consulting a doctor where thyroid disease is present. |
| Iron | No blanket supplementation | “Schwangere sollten nicht grundsätzlich Eisen-Tabletten einnehmen, sondern nur nach einem laborchemisch diagnostizierten Mangel” — only after a lab-diagnosed deficiency. The DGE states the same: iron is to be handled individually against iron status. |
| DHA | 200 mg/day, preferably from oily fish, or a supplement where fish is not eaten | General recommendation. |
| Vitamin D | 20 µg (800 IU) where sun exposure is limited | After medical consultation. |
The official German position on iron in pregnancy is explicitly that supplementation follows a lab diagnosis, not a general rule. So an emealia information marker on iron can quote a public-health source that itself says “this needs a doctor” — the referral is part of the fact, not a disclaimer bolted onto it. This is the single cleanest instance of the pattern in any of the three contexts.
| Finding | Source and strength |
|---|---|
| 22.4% received a nutritional-deficiency diagnosis within 12 months of starting (12.7% at 6 months). Vitamin D 13.6%, nutritional anaemia 4.0%, iron-deficiency anaemia 3.2%, B vitamins 2.6%, muscle loss 3.0%. | Butsch WS, Sulo S, Chang AT, et al., Obesity Pillars 2025;15:100186. Retrospective claims cohort, n = 461,382. Large and real. Caveat the authors state themselves: recorded deficiencies were more common among patients who saw a dietitian, which “may reflect both increased clinical complexity and increased opportunity for detection”. Part of that figure is detection, not incidence. The cohort is also 80.5% type 2 diabetes, mean age 52.9 — not the same population as an aesthetic weight-loss user. |
| Lean tissue as a share of weight lost: ~40% (STEP-1 DXA subset, n=95, semaglutide); ~26% (SURMOUNT-1 DXA subset, n=124, tirzepatide). | Look et al., Diabetes, Obesity and Metabolism 2025 (doi:10.1111/dom.16275). Small DXA subsets of large trials. |
| Protein guidance clustering at ≥1.2 g/kg/day (up to 1.6 in appropriate adults without kidney disease), with per-meal targets near 0.3–0.4 g/kg. One 2025 US analysis found average intake around 77 g/day. | Narrative reviews and clinical guidance — Nutrients 2026 (doi:10.3390/nu18172757); PubMed 42036071. Guidance, not trial outcomes. |
| Micronutrient risk rises as intake falls below roughly 1,200 kcal (women) / 1,800 kcal (men); iron, B12, vitamin D, calcium and thiamine are the repeatedly named shortfalls. | Same narrative-review literature. |
Never round the first row to “one in five people on Ozempic becomes deficient.” That drops the detection-bias caveat and the cohort composition, both of which the original authors put in writing.
The available synthesis is a narrative review in Nutrients (2023) — not a systematic review, and the authors do not enumerate limitations of their own evidence base, which is itself a weakness worth carrying forward. Figures it cites, drawn primarily from European guidelines:
The review also records that there is no scientific consensus on dietary soy and breast cancer, and that safety data covers soy foods rather than isoflavone supplements. Where a source says there is no consensus, emealia should say there is no consensus.
Much popular perimenopause nutrition content goes considerably further than this evidence supports. Nothing in the reviewed material establishes that a specific eating pattern relieves perimenopausal symptoms, and emealia should make no such suggestion. The defensible statements are narrow: certain published reference values differ for this age group, and bone-related nutrients appear consistently in guidance. That is the whole of it.
| Need | Status |
|---|---|
| A context setting a user turns on themselves | Partly Preferences already carry dietary context; a named context that swaps which reference values apply does not exist. |
| Information markers carrying a sourced, dated reference fact | To build The pattern is specified; the component is not built. Each string needs its source attached, not paraphrased. |
| Per-meal protein visibility (relevant to the GLP-1 guidance above) | To build Dividers across the existing band, showing contributions per meal — the design the founder corrected earlier in this collaboration. Distribution is readable from segment evenness without a per-meal target. |
| Reference values that differ by context | To build BuildTargets derives from the profile today; context-dependent reference values are a data-table change, not a new engine. |