Boswellia oleo-gum resin

Free to every reader. The safety section of every monograph is open to everyone, signed in or not. In an emergency call 000. For poisoning, call the Poisons Information Centre on 13 11 26.

15. Safety

Overall profile. Mills and Bone’s summary assessment: no major problems have been associated with ingestion of normal doses, no documented interactions with conventional drugs, and typical adverse reactions occur at low frequency and consist of diarrhoea and allergy [9]. The trial record since supports that. Vaidya 2025 found no adverse effect on clinical examination, biochemistry, haematology or ECG over 90 days [24]; Nirvanashetty 2026 reported it safe and well tolerated over 8 weeks [25]; Zhang’s network meta-analysis found no supplement, boswellia included, associated with more adverse events than placebo [20]; Holtmeier’s 52-week trial concluded "good safety profile but lack of efficacy" [28]. Three further trials read at Medicine review add nothing against it: Salter 2025 reported vital signs and safety parameters within normal limits over 120 days [55]; Giglio 2025 recorded no withdrawal in 103 patients over 8 weeks [56]; and the topical trial reported no side effect in 33 patients over 4 weeks [52].

The safety point that matters most at the counter is product variability. Measured triterpenic-acid content across sixteen retail boswellia nutraceuticals ranged from 0.4 to 35.7 per cent, and ten of fourteen commercial extracts carried 6–11 per cent undeclared citric acid while their titrimetric "total boswellic acids" figures complied (sections 10 and 16) [12,13]. A patient changing brands at the same labelled dose can therefore change actual intake by an order of magnitude. Nothing in the adverse-effect record below tells you what dose a given product delivers.

Adverse effects from trials and case reports.

Toxicology. Mills and Bone, p. 447 [9]: boswellic acids have very low acute toxicity: oral and intraperitoneal LD₅₀ above 2 g/kg in mice and rats, with no significant change in behaviour, clinical, haematological, biochemical or pathological data after chronic oral administration [9]. An extract enriched to 30 per cent AKBA gave an oral LD₅₀ above 5 g/kg in rats, was non-irritant to skin, and produced no adverse effect in a 90-day subacute study at up to 2.5 per cent of feed [9]. A 90-day gavage study of B. serrata in rats found no treatment-related toxicity [43]. Later toxicological work on standardised gum-resin extracts reached the same conclusion across acute, 90-day sub-chronic, Ames and micronucleus assays [9]. The one dissenting result: in a mouse model of experimentally induced colitis, a methanolic extract at 1 per cent of feed for 21 days produced hepatotoxic effects, supported by in vitro data [9]. That is a single finding in inflamed animals at a high dietary load, and it is recorded rather than dismissed. A PubMed search at Medicine review in September 2026 for boswellia with hepatotoxicity, liver injury or hepatitis, limited to case reports, returned no record; that is a search result, and it does not exclude unreported cases.

Contraindications. None established [9].

Precautions. Caution in patients with a known allergic tendency, on the reasoning that other resinous herbs such as myrrh cause allergic reactions and that allergic reactions have been reported for boswellia [9]. The platelet finding in section 14 — β-BA inducing calcium mobilisation, thrombin generation and aggregation in vitro [9] — has no clinical counterpart and no case report behind it, and AIP grades it theoretical; keep it in mind before surgery or alongside anticoagulants, and it is not a reason to withhold the herb. Medicine review found a study reporting that B. serrata gum resin inhibited platelet aggregation (Kokkiripati 2011, PMID 21771654) and a review of warfarin–herb interactions (Milić 2014, PMID 25233607). AIP does not hold either full text (queue #208 and #209), so neither is reported here and the direction of any platelet effect in people stays open.

Herb–drug interactions.

AIP comment on the CYP finding. It is an in vitro result, it is non-selective across six enzymes, and the compounds responsible have not been identified, so it cannot be predicted from a certificate of analysis. Given the plasma concentrations in section 14, a clinically important CYP interaction looks unlikely, but "unlikely on pharmacokinetic reasoning" is not "excluded", and no human interaction study of this drug exists. That gap is in section 18.

Pregnancy and lactation. Mills and Bone, pp. 447–448, assign Category B1 using the Australian categorisation scheme. That is the authors’ own assignment; the TGA has not categorised this herb — no increase in malformation or other harmful effect on the fetus from limited use in women, and no evidence of increased fetal damage in animal studies — and record that treatment of pregnant rats with a boswellic acid fraction at 250–1,000 mg/kg orally produced no significant effect on gestation period, litter size or birth weight, no gross morphological or skeletal abnormality, and normal second-generation offspring. They judge it likely to be compatible with breastfeeding [9]. AIP’s reading: the animal data are reassuring and the human data are thin. Harm has not been shown, and safety in human pregnancy has not been established either. Grade: animal data and expert judgement only; no human pregnancy or lactation study exists.

Children and older people. No trial has established a paediatric dose for this drug. Mills and Bone state that no information is available on safety in children and that adverse effects are not expected (p. 448) [9]. The paediatric material is thin and uncontrolled. Etzel’s 1996 narrative review recorded particular benefit, with good tolerance, in some children with chronic juvenile arthritis (p. 445) [9]. Nineteen children and adolescents with intracranial tumours, all previously treated conventionally, received palliative boswellia extract at up to 126 mg/kg/day for up to 26 months in an uncontrolled series (Janssen and colleagues, 2000, in the account at p. 447) [9]; that account reports no adverse-event data. Neither can carry a dose. Most of the osteoarthritis trials enrolled adults into their seventies [25], with no age-specific safety signal.

Overdose. No incident found in the published literature [9].

Effects on driving and machinery. None expected [9].

Australian regulatory safety position. Not scheduled in the Poisons Standard, on a full-text search of the instrument [16]. Permitted as an active in listed medicines at item 936 with no specific requirements — no dose ceiling, no route restriction and, the point that matters here, no required warning statement [15]. That absence is the TGA’s own recorded position on the ingredient and it is the strongest Australian regulatory safety statement available for this drug.

TGA safety alerts and DAEN, checked in September 2026 and again September 2026 at Medicine review, reported as they answered. The TGA’s site search and its safety alerts pages are rendered by JavaScript and returned no server-rendered results to an automated query, so no Boswellia safety alert, recall action or product defect correction could be confirmed or excluded by that route. The Database of Adverse Event Notifications could not be searched: DAEN now sits behind a Microsoft Power Apps portal at aems.tga.gov.au whose medicine search runs through authenticated cloud-flow endpoints that will not answer a non-interactive request. The 28/09/2026 re-check met the same obstacles: the safety-alert page returned its page frame with no results for the keyword, and the DAEN address redirected to the same portal. Both are open actions in section 20, and both need a person at a browser. No statement in this monograph depends on either, and nothing here should be read as "DAEN holds no report" — AIP does not know that.

Referral. Anyone considering boswellia, in any form, and particularly anyone taking anticoagulants, immunosuppressants or narrow-therapeutic-index medicines, anyone about to have surgery, and anyone pregnant or breastfeeding, should consult a degree-qualified herbalist with competencies and adequate training in pharmacognosy.