Turmeric Study

Sampling Plan

A sampling plan was developed to identify commonly sold products for purchase and analysis. The sampling plan was based on an evaluation of DS reported in NHANES 2013-2014 and a detailed survey of turmeric products sold via various channels including local stores, the internet and multi-level marketing companies in 2015.

A scoring system was developed to choose products for purchase from an initial pool of 197 products which fit our criteria (commonly sold DS containing turmeric material (rhizome powder and/or extract) as a single or primary ingredient). The highest scoring products (based on NHANES reporting, in-store surveys, number of websites sold, and best-seller status) were purchased for this study.

Two lots of each product (n=54) were purchased for the study. Products were purchased from the three major sales channels: mass market retail (e.g., Walmart, CVS, Target), natural and specialty retail (e.g., GNC or Whole Foods), and direct sales (products sold exclusively via the web or by multi-level marketers like Amway or Herbalife). Two DS were purchased in bulk for use as in-house control materials to monitor laboratory performance over time. Samples were repackaged and sent for laboratory analysis in defined batches.

Products with a wide range of dosage forms were purchased: hard-shell capsules (n=39), softgels (n=4), tablets (n=7), caplets (smooth-coated tablets; n=2), and powders (n=2). Dosage forms contained dry rhizome powders only, turmeric extract only or their mixtures. Thirty-eight products contained turmeric extract only; 12 products contained turmeric extract and dry rhizome powder mixtures, and 4 products contained dry rhizome powder only. Twelve products had claims for piperine in black pepper extract, and 3 listed pepper extract without a claimed level. Most products had a label claim for the amount of TC.

Laboratory Analysis and Quality Control

Turmeric DS were analyzed for Cur, DMC, BDMC and piperine (a constituent of black pepper extract, added to improve bioavailability) using high performance liquid chromatography (HPLC) and ultra (u)HPLC with ultraviolet (UV) detection. Samples of NIST Standard Reference Material (SRM) 3299 (turmeric rhizome powder) and SRM 3300 (turmeric extract) with known levels of curcuminoids were analyzed with each batch of samples to monitor the quality of the laboratory results. In addition, product duplicates and two in-house control materials were included in each batch. The consistent results for these quality control materials obtained from the three laboratories over the course of the study gave us confidence in the results for these constituents in the commercial turmeric DS under study.

The laboratory data were reported in mg/g and were reviewed. Samples with unusually large percentage differences from label claims or high variability among the 3 laboratories or among lots of the same product were retested. The final laboratory data were sent to our consulting statistician for statistical review.

Statistical Methods

Test results for individual curcuminoids, total curcuminoids (TC) and piperine were converted into mg per serving, mg per day and percentage difference from label, if applicable. The percentage difference from the label claim was calculated for each analyzed constituent in each sample: ((analytical value – label value)/label value) × 100%. Least squares means and standard deviations (SD) were computed for each product using a SAS mixed model procedure.

Results were calculated on a per serving and per day basis, because many of the 54 products in this study had a suggested use of more than one serving per day. Suggested use was one serving per day for 34 products, two servings per day for 10 products, three servings per day for 9 products and four servings per day for one product. If the suggested use was a range (e.g., 1-2 servings per day), the maximum number of servings per day was used. The labeled and mean measured contents for TC and piperine (if labeled) in each product are reported as per serving (Appendix Table 1) and per day (Appendix Table 2).

Results and Discussion

The labeled levels for turmeric material (information required by FDA) in the 54 products showed a wide range: 200-1300 mg per serving and 285-3600 mg per day (Tables 1; 2). The mean analytical results per serving of total curcuminoids (TC) also showed a wide range (8.1- 1254 mg, Appendix Table 1; median= 222 mg). The product means for TC content measured per day ranged from 16.3-2328 mg (Appendix Table 2), with the median = 351 mg.

The measured TC content varied widely in products with the same labeled amount of turmeric material. At the most common labeled level (500 mg per serving and per day) of turmeric material, measured TC results ranged from 8 to 518 mg per serving (n=17) and from 16 to 554 mg per day (n=12). Even for a subset of products that included extracts only, varying amounts 4 of TC were found. For example, nine products labeled at 500 mg per serving of turmeric extract contained 101 to 518 mg of measured TC, and 4 products labeled at 1000 mg per serving provided from 209 to 955 mg of TC (Appendix Table 1).

Many turmeric DS voluntarily listed the amount of TC on labels. We examined the percentage differences from labels for TC content. These ranged from -34.4 to +73.3% (7.8 ± 16.7%, mean ± SD; n=41). For 26 of 38 products (68%) the analytical results were within ±10% of the label claims for TC, which meets the USP recommendations from the monographs for turmeric tablets and capsules (USP43-NF38). Thirteen DS in our study had TC content more than 10% above labels and two DS were below -10% of label claims.
Thus, the voluntary label information about the TC content was more closely associated with the measured content than the FDA required information about the weights of turmeric materials.

We found that the measured TC content differed in different matrices. The TC content was much lower in products containing dry rhizome powders only than in products with turmeric extracts only. TC content in DS with extracts plus dry rhizome powders was, as expected, on average, lower than in DS with only extracts but higher than in dry rhizome powder only. Voluntary labeling of the amount of piperine on the labels was also closely associated with the actual content of piperine. The labeled levels for piperine ranged from 1.9-14.3 mg per serving and 1.9-15.1 mg per day. Measured piperine content in products having a labeled amount (n=12) ranged from 2.2 to 15 mg per serving (6.7 ± 4.3; mean ± SD), and 2.2 – 17 mg per day (8.2 ± 5.3). The percentage differences from labels for piperine ranged from -3.4 to +21% (8.3 ± 7%) (Appendix Table 3).

An evaluation of measured TC content for DS with and without a claim for content shows that products with the claim were more likely to have higher amounts than those without a claim (Table 1).

Conclusions

In summary, turmeric DS have a variety of label formats with a wide range of labeled weights of turmeric and labeled and measured TC concentrations. Results indicate that the FDA required information about the total weight of turmeric material may not permit accurate predictions for the content of TC. As expected, the dry rhizome powders had much lower TC content compared to the extracts. Turmeric extracts may or may not be highly concentrated 5 and some extracts may be microencapsulated (adds additional weight). Altogether, these factors explain the poor predictability of TC from the turmeric weight in many of these products. Products with voluntary labeling of TC and piperine content were relatively accurate (83% and 92% of tested products were within ± 20% of label claims, respectively). Those with claims for TC content were more likely to contain higher levels of TC than DS without a claim.

Current labeling requirements for turmeric DS may be insufficient to inform researchers and consumers about the actual content of TC in DS because the extract concentrations vary widely.

Additional statistical evaluation of the curcuminoid results using regression analysis is planned. The results for products with TC content claims, DS with 95% TC extract claims and those with trademarked curcumin ingredients (n=21) will be compared to products without those characteristics. The ratio of the three major curcuminoids will be compared to USP standards and products evaluated for possible adulteration.

Compliance with the United States Pharmacopeia (USP) Performance Standards for Dosage Forms

Compliance with the USP performance standards for DS is currently voluntary in the US. However, DSID results showing inadequate performance of green tea DS dosage forms raised concerns that DS users may not achieve health benefits expected from green tea DS (Gusev et al., 2020). The inadequate performance jeopardizes confidence in the efficacy and safety of green tea DS sold in the US and used in clinical trials.

Turmeric DS with immediate release forms were tested for dosage form performance using USP disintegration and rupture tests. Hard-shell capsules (n=39) and tablets/caplets (n=9) and were tested for disintegration and soft gels (n=4) for rupture. For each test, tablets or capsules were immersed and agitated in appropriate media and visually inspected after 15 (rupture) or 30 minutes (disintegration). A hard-shell capsule or tablet passed disintegration if there was no solid mass remaining in the vessel other than remnants of the shells or coating films. A soft gel passed the rupture test if its outer shell opened. Tablets/caplets had a 66.7% pass rate, and hard-shell capsules had an 84.6% pass rate, the rest products had inconsistent results in two lots (Table 2). All soft gels passed the rupture test (Andrews et al., 2021; Savarala et al., 2019).

Appendix