Ingredient science
See the latest research on supplement ingredients, from botanicals and vitamins to nootropics and proteins.
Find out what the science really says about trending ingredients, and uncover new findings on ingredients you thought you already knew well. Scientific papers can be dense for non-scientists, but our authors are scientists themselves. They turn complex research into practical, easy-to-understand product and formulation advice you can put to use right away.
Looking for the full list of health claims? You’ll find them here.
Rosa canina
Standardised rosehip powder modestly reduces osteoarthritis pain versus placebo (SMD 0.37, NNT ≈ 6). Galactolipid and polyphenol constituents downregulate NF-κB, COX/LOX, iNOS and MMPs. Limitation: only three trials were pooled, all manufacturer-funded.
Arctostaphylos uva-ursi
Arbutin is reliably absorbed and renally excreted as hydroquinone conjugates (~65% of dose; tablets equivalent to solution), which is the antibacterial rationale. Limitation: the largest placebo-controlled UTI trial found no symptom benefit and no antibiotic-sparing effect. The EMA/HMPC monograph lists bearberry leaf as traditional use only, adult women, maximum one week.
Regulatory note: EMA/HMPC/750269/2016 — European Union herbal monograph on Arctostaphylos uva-ursi (L.) Spreng., folium (Rev. 2). Regulatory status, not clinical evidence.
Sideritis scardica
Polyphenols may inhibit monoamine reuptake and raise cerebral blood flow; 950 mg/day for 28 days cut RVIP false alarms and state anxiety in over-50s. A six-day decoction spared CYP1A2, xanthine oxidase, NAT2 and UGT, with CYP2A6 activity falling in men only. Limitation: single small trials; the CYP study was open-label in 14 subjects, not randomised.
Rhodiola rosea
The standardised extract SHR-5 at 576 mg/day for 28 days beat placebo on burnout scores and attention indices and blunted the cortisol awakening response, supporting HPA-axis modulation. Limitation: across eleven trials results are contradictory and risk of bias is high or unclear throughout.
Bovine milk glycoprotein
An iron-binding glycoprotein modulating NF-κB signalling. 200 mg/day lowered IL-6 but not CRP, and cut respiratory-infection incidence in infants and children (OR 0.78) though not in adults. Versus ferrous sulfate it raised serum iron, ferritin and haemoglobin more. Limitation: doses across trials range from 35 to 833 mg, so pooled estimates are highly heterogeneous.
Hydrolysed collagen
Bioactive di- and tripeptides reach the circulation and stimulate fibroblasts and chondrocytes. Around 90 days improves skin hydration, elasticity and wrinkle depth, and reduces knee osteoarthritis pain (SMD −0.58, moderate certainty). Limitation: all knee trials were rated high risk of bias; skin trials were 95% female and largely industry-sponsored.
Forms, absorption, endpoints
Organic salts (citrate, glycinate, lactate) outperform inorganic oxide, and fractional absorption falls as the dose rises. A median 368 mg/day for three months lowered systolic blood pressure by 2.0 mmHg and diastolic by 1.8 mmHg. Limitation: the effects are small, and direct head-to-head comparisons between salt forms remain scarce.
Regulatory note: EFSA NDA Panel 2024, EFSA Journal — safety of magnesium L-threonate as a novel food and bioavailability from this source (PMID 38481468). Useful precedent for magnesium-form bioavailability dossiers.
Cobalamin
High-dose oral B12 normalises serum levels comparably to intramuscular dosing, via intrinsic-factor-independent passive diffusion and at lower cost. Serum B12 alone is insensitive: methylmalonic acid, homocysteine and holotranscobalamin detect functional deficiency better. Limitation: the Cochrane evidence is graded low quality and reports no clinical-symptom outcomes.
This class is part of our category structure, but no brief has been appraised and published yet. We add a class only once there is a peer-reviewed synthesis worth reporting.
This class is part of our category structure, but no brief has been appraised and published yet. We add a class only once there is a peer-reviewed synthesis worth reporting.
If you have any questions about how a specific ingredient works, get in touch.
