For some consumers out there this may seem somewhat of a bonus but may not please others who’ve eaten a hemp bread sandwich everyday for a week only to get a driving ban! The European Food Safety Authority (EFSA) has looked at THC levels in 12 different categories of hemp foods, including hemp oil, breads and teas. More than 1,500 samples were collected from the Czech Republic, Germany, Italy and Romania.
Hemp Industry Daily reports
The January report from the European Food Safety Authority (EFSA) looked at THC levels in 12 different categories of hemp foods, including hemp oil, breads and teas. More than 1,500 samples were collected from the Czech Republic, Germany, Italy and Romania.
The report concluded that THC levels in people ingesting large amounts of hemp products could exceed a safety threshold for THC in food set by the group in 2015 and potentially lead to effects on the central nervous system and an increased heart rate.
“The number of consumers of hemp and hemp-based products still represents an important uncertainty,” they wrote.The report’s authors pointed out that little is known about how many people are eating such hemp products, or how much they are consuming.
Hemp Industry Daily Full Report
Here’s the EFSA – Abstract & Summary
Acute human exposure assessment to tetrahydrocannabinol (Δ9‐THC)
Delta‐9‐tetrahydrocannabinol (Δ9‐THC) is a naturally occurring psychoactive compound derived from the hemp plant Cannabis sativa. In 2015, EFSA established an acute reference dose (ARfD) of 1 μg/kg body weight (bw) for Δ9‐THC and assessed acute dietary exposure from milk and dairy products. This resulted at the most 3% and 13% of the ARfD for adults and toddlers, respectively. Following the European Commission Recommendation 2016/2115 on the monitoring of the presence of Δ9‐THC in food and the issuing of a new mandate, EFSA assessed the acute human exposure to Δ9‐THC. ‘Standard’ food categories were used as proxies for consumption of hemp and hemp‐based products. Twelve independent scenarios based on single food categories were considered and acute exposure was assessed for consumption days only for all age groups excluding infants. Occurrence data for Total‐Δ9‐THC (588 samples in total) were used for this assessment up to the highest reliable percentile for each food category. The EFSA ARfD of 1 μg/kg bw was exceeded in the adult high consumers of most considered hemp and hemp‐containing products, under the lower‐bound (LB) and upper‐bound (UB) scenario. At the UB, acute exposure in adult high consumers was estimated based on the highest reliable percentile of occurrence, for Hemp seeds (P95, up to 9 μg/kg bw), Hemp oil (P95, up to 21 μg/kg bw), Tea (Infusion) (P95, up to 208 μg/kg bw), Breakfast cereals (P50, up to 1.3 μg/kg bw), Pasta (Raw) (P75, up to 6.4 μg/kg bw), Bread and rolls (P75, up to 1.3 μg/kg bw), Bread and rolls from hemp flour (P90, up to 4.1 μg/kg bw), Cereal bars (P50, up to 0.3 μg/kg bw), Fine bakery wares (P75, up to 5.1 μg/kg bw), Chocolate (Cocoa) products (P75, up to 1.1 μg/kg bw), Energy drinks (P75, up to 0.2 μg/kg bw), Dietary supplements (P75, up to 9.9 μg/kg bw), Beer and beer‐like beverages (P90, up to 41 μg/kg bw). The use of proxies for the consumption of hemp and hemp‐containing products, the limited number of occurrence data and the analytical limitations in the quantification of Δ9‐THC represent the most important sources of uncertainty. Overall, exposure estimates presented in this report are expected to represent an overestimation of acute exposure to Δ9‐THC in the EU.
In 2015, the EFSA Panel on Contaminants in the Food Chain (CONTAM Panel) established an acute reference dose (ARfD) of 1 μg/kg body weight (bw) for Δ9‐tetrahydrocannabinol (Δ9‐THC) in its scientific opinion on risks for human health related to the presence of tetrahydrocannabinol (THC) in milk and other food of animal origin (EFSA,2015). In 2015, EFSA estimated acute dietary exposure only from milk and dairy products. Acute exposure to ∆9‐THC through the consumption of milk and dairy products resulted at the most in 3% and 13% of the ARfD of 1 μg/kg bw for adults and toddlers, respectively. Considering the limitations of the 2015 exposure assessment, the Commission Recommendation (EU) 2016/2115, invited Member States and food business operators to monitor the presence of Δ9‐THC, its precursors and other cannabinoids in food and to provide these data to EFSA. In October 2018, an official request was issued by the European Commission asking the EFSA Evidence Management Unit (DATA Unit) for an acute dietary exposure assessment to Δ9‐THC taking into account the new occurrence data available in the EFSA database and the updated comprehensive food consumption database. In addition, EFSA was requested to provide an overview of the available occurrence data on the precursors of Δ9‐THC and other cannabinoids in food together with information on their co‐occurrence with Δ9‐THC. By the end of December 2018, the EFSA database contained 1,866 analytical results on Δ9‐THC and other cannabinoids in food. There is documented uncertainty associated with the exact quantification of Δ9‐THC in food due to analytical methods, extraction efficiency as well as in relation to conversions related to food processing. For instance methods based on capillary gas chromatography with flame ionisation (GC‐FID) (as the official method prescribed by EC No 1122/2009) are not able to differentiate the psychoactive Δ9‐THC from its non‐psychoactive precursors delta‐9‐tetrahydrocannabinolic acids A (Δ9‐THCA‐A) and B (Δ9‐THCA‐B). In the absence of a preliminary separation step (such as a derivatisation with silanes), gas chromatography coupled with mass spectrometry (GC–MS)‐based methods are thus not specific for Δ9‐THC. In contrast, liquid chromatography (LC)‐based methods are specific for Δ9‐THC. The analytical method used for the sample preparation and analysis therefore dictates the specificity for the results reported for Δ9‐THC. Considering this, all analytical results reported as ‘Δ9‐THC’ were therefore carefully evaluated in relation to the applied analytical method. After applying the exclusion criteria and a careful reclassification of analyte according to analytical method, a total of 1,547 analytical results were available in the final data set, with 1,303 analytical results submitted by national organisations and 244 by industry). Most of the samples (n = 427) were on ‘pure’Δ9‐THC (defined as data submitted on Δ9‐THC and produced with LC‐based methods); 237 on ‘Sum of delta‐9‐Tetrahydrocannabinol and delta‐9‐Tetrahydrocannabinolic Acid’ (Sum Δ9‐THC/Δ9‐THCA) (data reported on Δ9‐THC and analysed either by GC‐based methods or unknown methods), on cannabidiol (208) and cannabinol (167). Samples were collected from 2000 until 2018, with most of the samples collected from 2016 onward. Data were provided on a variety of hemp and hemp‐based products although the amount of data of food of animal origin remained poor. Most of the analytical results reported from governmental organisations were from samples collected in Germany, Italy and the Czech Republic; industry reported samples from Germany and Romania. Considering the limited number of samples reported for ‘pure’Δ9‐THC and Sum Δ9‐THC/Δ9‐THCA, and the fact that both categories referred, although with a different extent of uncertainty, to Δ9‐THC, the two sets of data were merged into a unique group referred as ‘Total‐Δ9‐THC’ and used to perform acute exposure assessment.
The occurrence data used to assess acute exposure to Total‐Δ9‐THC was finally composed of 588 samples (covering 13 hemp and hemp‐derived products). Compared to the EFSA opinion published in 2015 (EFSA,2015), there was an increase of the number of samples as well as the number of food categories that could be used for exposure assessment to Δ9‐THC.
The latest version of the EFSA Comprehensive European Food Consumption Database (Comprehensive Database) updated in 2018, containing results from a total of 60 different dietary surveys carried out in 25 different Member States across age classes was used. Given the limited amount of data on individual consumption of hemp and hemp‐based products, it was decided to use proxies based on ‘standard’ food categories and to calculate acute dietary exposure.
Based on the occurrence data, 12 independent scenarios were considered, based on single food categories and acute exposure was assessed for all age groups, with the only exclusion of infants across European population groups. Occurrence data for Total‐Δ9‐THC were used for the acute exposure assessment up to the highest reliable percentile. Mean occurrence values were not considered representative of the distributions because in most of the cases they were positively skewed.
Acute dietary exposure was assessed at the upper bound (UB) and lower bound (LB) for Total‐Δ9‐THC for consumers only of the following hemp and hemp‐derived products:
- ‘Hemp oil’ (n = 125, at the UB occurrence level P50 = 1,890, up to P95 = 17,000 μg Total‐Δ9‐THC/kg);
- ‘Hemp seeds’ (n = 127, at the UB P50 = 390, up to P95 = 3,960 μg Total‐Δ9‐THC/kg);
- ‘Tea (Infusion)’ (n = 119, at the UB P50 = 140, up to P95 = 6,467 μg Total‐Δ9‐THC/kg);
- ‘Bread and rolls’ (‘as such’ n = 14, at the UB P50 = 85, up to P75 = 190 μg Total‐Δ9‐THC/kg) and as ‘Bread and rolls from hemp flour’ (n = 49, at the UB P50 86; up to P90 597 μg Total‐Δ9‐THC/kg);
- ‘Pasta (Raw)’ (n = 18, at the UB P50 = 310, up to P75 = 1,000 μg Total‐Δ9‐THC/kg);
- ‘Breakfast cereals’ (n = 7, at the UB P50 = 200 μg Total‐Δ9‐THC/kg);
- ‘Cereal bars’ (n = 5, at the UB P50 = 200 μg Total‐Δ9‐THC/kg);
- ‘Fine bakery wares’ (n = 24, at the UB P50 = 65, up to P75 = 776 μg Total‐Δ9‐THC/kg);
- ‘Chocolate (Cocoa) products’ (n = 19, at the UB P50 = 200, up to P75 = 400 μg Total‐Δ9‐THC/kg);
- ‘Energy Drinks’ (n = 25, at the UB P50 = 3, up to P75 = 15 μg Total‐Δ9‐THC/kg);
- ‘Beer and beer‐like beverages’ (n = 30, at the UB P50 = 14, up to P90 = 635 μg Total‐Δ9‐THC/kg);
- ‘Dietary supplements’ (n = 26, at the UB P50 = 1,115, up to P75 = 19,800 μg Total‐Δ9‐THC/kg).
Mean and high (P95) exposure estimates based on different percentiles (up the highest reliable one) of occurrence were assessed. For the scenario on ‘Hemp oil’: at the P95 occurrence level for the high consumers, at the UB the acute exposure to Total‐Δ9‐THC in ‘adults’ ranged from 3 to 21 μg/kg bw and in ‘other children’ from 7 to 59 μg/kg bw. For the scenario on ‘Hemp seeds’: at the P95 occurrence level for the high consumers, acute exposure to Total‐Δ9‐THC ranged at the UB from 1.1 to 2.9 μg/kg bw in ‘other children’ and from 2.3 to 9 μg/kg bw in ‘adults’. For the scenario on ‘Tea (Infusion)’: at the highest reliable occurrence percentile (P95) for high consumers, the acute exposure ranged from 40 to 208 μg Total‐Δ9‐THC /kg bw in adults (UB). For the scenario on ‘Bread and Rolls’: acute exposure was estimated up to the P75 occurrence level (14 samples) for ‘Bread and Rolls’ (as such) and up to the P90 occurrence level for ‘Bread and rolls from hemp flour’ (49 samples). In adults, exposure to Total‐Δ9‐THC for the high consumers was 0.3–1.3 (LB–UB) μg/kg bw in ‘bread and rolls’ vs. 1.15–1.15 μg/kg bw (LB–UB) in ‘Bread and rolls from hemp flour’ at the P75 occurrence level. The difference was mainly due to the elevated amount of left‐censored data (71%) in the category ‘Bread and rolls’ (as such). For the scenario on ‘Dietary supplements’: 26 samples were available and acute exposure was estimated up to the P75 occurrence level. At this percentile, for high consumers the exposure to Total‐Δ9‐THC varied between 1.5 and 9.9 (UB) μg/kg bw in adults. For the scenario on ‘Pasta (Raw)’: only 18 samples were available and acute exposure was estimated up to the P75 occurrence level. At this percentile for high consumers at the UB, exposure varied between 1.2 and 6.4 in adults μg Total‐Δ9‐THC /kg bw. For the scenario on ‘Breakfast cereals’: exposure was estimated up to the P50 occurrence level (7 samples). At this percentile, the exposure for high consumers ranged from 0.18 to 1.27 (at the UB) Total‐Δ9‐THC μg/kg bw in adults. For the scenario on ‘Cereal bars’: exposure was estimated only for the P50 occurrence level (5 samples). At this percentile, acute exposure to Total‐Δ9‐THC ranged from 0.19 to 0.27 (at the UB) μg/kg bw in adults with high consumption. For the scenario on ‘Fine bakery wares’: exposure was estimated up to the P75 occurrence level (24 samples). At this occurrence percentile, the exposure to Total‐Δ9‐THC in adults, for high consumers ranged from 1.7 to 5.1 μg/kg bw at the UB. For the scenario on ‘Chocolate (Cocoa) products’: acute exposure was estimated up to the P75 occurrence level (19 samples). At this occurrence percentile, the exposure to Total‐Δ9‐THC in adults, for high consumers ranged from 0.37 to 1.07 μg/kg bw at the UB. For the scenario on ‘Energy drinks’: acute exposure was estimated up to the P75 occurrence level (25 samples). For high consumers at the P75 of occurrence, exposure in adults varied between 0.14 and 0.25 (UB) μg/kg bw. For the scenario on ‘Beer and Beer‐Like Beverages’: acute exposure was estimated up to the P90 occurrence level (30 samples). At this percentile, the acute exposure to Total‐Δ9‐THC at the UB, for high adult consumers varied from 7 to 41 μg/kg bw. Overall exposure estimates were in line with those performed by the BfR (2018). The EFSA ARfD of 1 μg/kg bw was exceeded at the UB in adult consumers for most of the hemp and hemp‐derived products. Main sources of uncertainty in current exposure assessment are represented by: (i) the use of proxies of hemp and hemp‐based products on consumers of ‘standard’ food categories, (ii) the limited occurrence data set used to feed the exposure scenarios (ranging from 5 up to 127 per food category) represents a source of uncertainty related to the representativity of the data set, (iii) well‐known methodological limitations in the exact determination of ∆9‐THC levels in different foods due to a number of factors (variable selectivity towards ∆9‐THC of different analytical methods (e.g. GC‐based vs. LC‐based; the extraction efficiency of ∆9‐THCA‐A and ∆9‐THC for different food types and methods; conversion of ∆9‐THCA‐A into ∆9‐THC during food processing and cooking)). In the present report, a 100% conversion of ∆9‐THCA‐A into ∆9‐THC was considered, however an attempt to assess exposure to ‘pure’Δ9‐THC was carried out in order to evaluate the level of conservativeness of this assumption. Overall, the inclusion of occurrence data on Sum Δ9‐THC/Δ9‐THCA reduced the exposure estimates and its uncertainty since more samples were available for the assessment. Considering all the above‐mentioned factors, exposure estimates presented in this report are expected to represent an overestimation of the acute exposure to Δ9‐THC in ‘single food’ scenarios on consuming days. It would be desirable to encourage further research to obtain sensitive, validated (including interlaboratory validation) and Δ9‐THC specific methods to be translated to reliable official methods. Studies on the stability of Δ9‐THC, as well as on the conversion of Δ9‐THCA during food processing, including cooking, are required using state‐of‐the‐art Δ9‐THC specific methods. Data providers and, in general, the scientific community working in the field, should be encouraged to avoid the misclassification and the submission to EFSA of data on Total‐Δ9‐THC (based on unspecific GC methods) as ‘Δ9‐THC’. Member states should be encouraged to collect and submit to EFSA more occurrence data (based on selective methods) for Δ9‐THC in food and especially of animal origin, including dairy products, eggs and meat of animals fed with hemp and hemp‐derived products. Consumption data on hemp‐derived products are needed to refine the exposure scenarios.
Click On Image To Read Full Report