Comparison
Turmeric (curcumin) vs. Frankincense (Boswellia)
Turmeric or Boswellia – which one really helps with knee osteoarthritis?
The short answer
In knee osteoarthritis both hold up, and both with caveats. Turmeric has the broader data base: a network meta-analysis of 23 studies with 2,175 patients found -1.63 on the pain scale, and in a study with 367 patients an extract was non-inferior to ibuprofen. For Boswellia, the Cochrane review found high-quality evidence, but from only 2 studies with 85 participants; there is no comparison with ibuprofen. The sources name no study testing the two head to head, so the studies are laid side by side. The clearest difference is safety: LiverTox lists turmeric in its highest category A as a cause of liver injury, Boswellia as E, an unlikely cause. Boswellia's weak point is product quality instead: 41 percent of the products tested did not match their label.
The numbers side by side
| Axis | Turmeric (curcumin) | Frankincense (Boswellia) | Difference |
|---|---|---|---|
| Human evidenceHow much and how good the data in humans are | 6 | 6 | even |
| MechanismHow well the chain of action is understood | 5 | 5 | even |
| Safety dataHow well safety has been studied – not how safe it is | 7 | 6 | Turmeric better studied |
| Hype gap10 means: the advertising stays level with the data | 3 | 5 | Frankincense closer to the data |
| Track record of useHow much documented use in humans exists – does not measure whether it works | 9 | 7 | Turmeric in use longer |
| Rating | Supported, with caveats · Supported, with caveats | ||
“Track record of use” measures how long and how widely something has been used in humans – not whether it works. The BK-Score is a subjective assessment by Biohacking Kompakt based on published scoring rules (German). It is not a scientific rating, not the consensus of a professional society and not a medical recommendation.
Why these numbers
Turmeric (curcumin)
In knee osteoarthritis, several meta-analyses show a moderate reduction in pain versus placebo with short-term use – a network meta-analysis of 23 studies with 2,175 patients found -1.63 on the pain scale and -18.85 on the WOMAC total score, and a study with 367 patients found non-inferiority to ibuprofen. The quality behind this is weak: a critical review rated 37 of 48 endpoints as very low quality of evidence, and the EMA recognises only traditional use for Curcuma longa, not well-established use. For cancer prevention, anti-aging and the claimed liver detox there is nothing comparable; no study tests a detox endpoint. The mechanism has a bioavailability problem: many cell-culture findings rely on concentrations that are not reached orally; in humans, CRP and TNF-alpha are at least lowered. Well documented on the safety side: the DILIN case series (Am J Med 2023) describes 10 cases of curcumin-induced liver injury, one of them fatal; 7 of 10 carried HLA-B*35:01. In its opinion No. 040/2021 of 14 December 2021, the BfR found that products tested at 8.3 and 6.1 mg curcumin per kilogram of body weight were well above the ADI of 3 mg/kg.
Frankincense (Boswellia)
Evidence 6, because several double-blind RCTs and meta-analyses show a benefit in knee osteoarthritis – Cochrane 2014 with high-quality evidence for 100 mg of enriched extract over 90 days (2 studies, 85 participants, pain -17 points out of 100), Yu 2020 with 7 RCTs and 545 patients, Zhang 2025 as a network meta-analysis –, but the studies are small, short and partly close to the manufacturer (AkdÄ 2017), Bannuru 2018 calls the overall quality low and Dalmonte 2024 found no significant overall effect because of heterogeneity. Mechanism 5, because inhibition of 5-lipoxygenase and mPGES-1 is well described in the lab and in human whole blood (Siemoneit 2011, IC50 3 to 10 µM), but plasma levels in humans are often below these concentrations (Abdel-Tawab 2021) and target inhibition in the body has not been measured. Safety 6, because controlled data up to 52 weeks show no disadvantages versus placebo (Holtmeier 2011) and LiverTox considers liver injury unlikely (likelihood score E), but pharmacovigilance as for medicines is lacking. Hype 5, because frankincense is often marketed as a natural alternative to painkillers, although there is no RCT against approved NSAIDs and many products do not match their label (Meins 2016: 41 %). Use 7, because frankincense has been used traditionally in Ayurveda and widely as a dietary supplement for years without being approved as a medicine. Direction positive: in osteoarthritis the data point predominantly in one direction; in maintaining remission of Crohn's disease there was no effect.
Frequently asked questions
Is turmeric or Boswellia better for osteoarthritis?
There is no head-to-head study. Laid side by side, both reduce pain in knee osteoarthritis compared with placebo. Turmeric has more studies and a comparison with ibuprofen, but a critical review rated 37 of 48 endpoints as very low quality of evidence. Boswellia's studies are small, short and partly close to the manufacturer, and a 2024 meta-analysis found no significant overall effect because of heterogeneity. In both cases symptoms were measured, not the state of the cartilage.
Which is riskier for the liver, turmeric or Boswellia?
Based on the data, turmeric. LiverTox, the reference database of the US National Institutes of Health, lists turmeric in its highest category A as a well-documented cause of clinically apparent liver injury. The DILIN case series describes 10 cases; 5 patients needed hospital treatment and 1 patient died. The EMA sees the risk mainly with highly bioavailable, high-dose preparations. LiverTox rates Boswellia as E, an unlikely cause, although pharmacovigilance as for medicines is lacking.
Can you combine turmeric and Boswellia?
The combination is common, but the sources reviewed do not study it: whether the two together do more than either alone is open. Both topic pages mention interactions with anticoagulant drugs. For turmeric, the EMA lists possible interactions with NSAIDs, antiplatelet drugs, lipid-lowering drugs and immunosuppressants, while considering the clinical evidence for them insufficient. For Boswellia, the German consumer advice centre advises against combining it with drugs such as warfarin. Anyone taking such medicines should clarify this with a doctor.
Why does the product matter so much with turmeric and Boswellia?
With turmeric, because of absorption: curcumin barely reaches the blood, and piperine, micelles and phospholipid complexes are pharmacokinetically very different routes, not all of which have their own efficacy data. Highly bioavailable preparations in particular stand out in the liver injury cases. With Boswellia, because of content: the study results apply to a few special extracts such as 5-Loxin or Aflapin, and in an analysis of 17 products, 41 percent did not match their label.
See all ratings in the ranking (German app)
Information only, not medical advice and not a usage or dosage recommendation. Prescription-only and unapproved substances belong in a doctor's hands.