Transfer of cannabinoids into the milk of dairy cows fed with industrial hemp could lead to Δ⁹-THC exposure that exceeds acute reference dose

The resurgence of the industrial hemp sector has led to the increased use of hemp-derived products, including animal feed. However, concerns exist regarding the transfer of cannabinoids, such as Δ⁹-tetrahydrocannabinol (Δ⁹-THC), from hemp feed into animal products and the potential risks to human consumers. Δ⁹-THC is a psychoactive cannabinoid, while others like cannabidiol (CBD) have varying pharmacological activity. While the cultivation of industrial hemp with low Δ⁹-THC content is permitted in the European Union, the extent of cannabinoid transfer into animal products remains unclear. Previous studies are limited by the inability to differentiate between Δ⁹-THC and its non-psychoactive precursor, Δ⁹-tetrahydrocannabinolic acid (Δ⁹-THCA), which is often much more abundant in hemp. This study aims to quantify the transfer of various cannabinoids from industrial hemp silage into cow's milk, assess the potential effects on animal health, and determine the risk to human consumers.

Existing literature on cannabinoid transfer into animal products is scarce, with few experimental data specifically on Δ⁹-THC transfer into cow's milk. Analytical techniques often fail to distinguish between Δ⁹-THC and Δ⁹-THCA, hindering accurate risk assessment. Studies on the effects of hemp feeding on ruminants show varied results regarding feed consumption, with some reporting no effects, while others observed reductions. The potential impact of other phytochemicals in hemp, besides cannabinoids, on animal health also needs consideration. Previous research on human exposure to cannabinoids via breast milk after marijuana use has demonstrated transfer of Δ⁹-THC and CBD, but data on other cannabinoids and metabolites is limited. The lack of comprehensive data highlights the need for this study to fill in the knowledge gaps and provide a more accurate picture of potential risks.

This study involved a feeding experiment with ten lactating Holstein Friesian dairy cows, divided into two groups (low and high hemp supplementation). The cows were fed a partially mixed ration (PMR), where corn silage was progressively replaced with two types of hemp silage: hemp silage A (low cannabinoid concentration, whole plant) during an adaptation period, and hemp silage E (higher cannabinoid concentration, leaves, flowers, and seeds) during an exposure period. A subsequent hemp-free depuration period followed. Daily feed intake, milk yield, respiratory and heart rates, body temperature, and animal behavior were monitored. Blood plasma and fecal samples were collected from a subset of cows. A new HPLC-MS/MS-based analytical technique was developed to precisely quantify various cannabinoids, including Δ⁹-THC, Δ⁹-THCA, Δ⁹-THCV, CBD, CBN, CBDV, 11-OH-Δ⁹-THC, and THC-COOH, in milk, blood plasma, and feed samples. A two-compartment toxicokinetic model was developed to simulate cannabinoid transfer from feed to milk, accounting for absorption, distribution, metabolism, and excretion. The model used experimental data to estimate transfer rates at steady state for each cannabinoid.

Feeding cannabinoid-rich hemp silage (E) significantly decreased feed intake and milk yield in both groups. Respiratory and heart rates also decreased significantly, and cows exhibited behavioral changes such as tongue play, increased yawning, salivation, and altered gait. These effects were reversible upon cessation of hemp feeding. Measurable levels of Δ⁹-THC, Δ⁹-THCA, Δ⁹-THCV, CBD, CBN, and CBDV were detected in cow's milk. Δ⁹-THC concentrations reached up to 316 µg/kg milk, while CBD concentrations reached up to 1,174 µg/kg milk. The transfer rate of Δ⁹-THC to milk was estimated at 0.20% ± 0.03% at steady state. Using the EFSA RACE tool, the estimated exposure to Δ⁹-THC via milk and dairy products exceeded the ARfD of 1 µg/kg body weight for several consumer groups, particularly children and infants. The ARfD was exceeded up to 120-fold for high consumers in the high hemp group. Even with low hemp silage, the ARfD was exceeded for some infant consumers.

The study's findings indicate that feeding cannabinoid-rich industrial hemp silage to dairy cows can lead to significant transfer of cannabinoids, notably Δ⁹-THC, into milk. The observed adverse effects on animal health, including reduced feed intake and milk production, confirm the potential for cannabinoids to impact animal physiology. The exceedance of the ARfD for Δ⁹-THC in human consumers raises concerns about consumer safety, especially for vulnerable groups like children and infants. The high levels of CBD in the milk also warrant further investigation into its potential risks, though current data are insufficient for risk assessment. Factors such as hemp variety, plant parts used for silage, and harvest time influence cannabinoid concentration and hence the risk. The study highlights the necessity for cannabinoid analysis of hemp feed before use to ensure animal health and consumer safety.

This study demonstrates that feeding cannabinoid-rich industrial hemp silage to dairy cows leads to a transfer of cannabinoids into milk, resulting in potential human health risks exceeding the ARfD for Δ⁹-THC in certain consumer groups. The observed negative impacts on cow health further emphasize the need for careful monitoring and regulation. Future research should focus on the long-term effects of hemp consumption in dairy cows, a more detailed investigation into CBD’s potential risks, and further refinement of the toxicokinetic model to incorporate metabolic pathways and better understand the fate of cannabinoids in the gastrointestinal tract.

The study utilized a relatively small number of cows, potentially limiting the generalizability of the results. The lack of consistent urine data due to analytical challenges hindered a more comprehensive assessment of cannabinoid excretion. While the study controlled for many variables, the impact of other phytochemicals in hemp silage on animal health could not be fully isolated. The extrapolation of results using EFSA RACE tool relies on existing consumption data which might have variability. Further research with larger sample sizes and improved analytical methods could strengthen these findings.