Number 52      February 1995

MAFF UK - Surveillance for pyrrolizidine alkaloids in honey

Index to MAFF UK Food Surveillance Information Sheets, 1995

Background

Pyrrolizidine alkaloids (PAs) form a large group of natural toxicants which occur in some plants of the families Boraginaceae, Compositae, Leguminosae and Ranunculaceae. Some plant species of the genera Heliotropium (family Boraginaceae) and Senecio (family Compositae) are found in the UK. PAs have been shown to cause hepatic (liver) damage and hepatocellular cancer in experimental animals. PAs have also been linked to veno-occlusive disease and chronic hepatic damage in humans at high doses.

Humans may be exposed to PAs directly, for example by the ingestion of herbal teas and herbal products. Humans may also be indirectly exposed to PAs in foods, such as milk, from animals which have consumed PA-contaminated feed. MAFF has previously instigated studies on the levels and types of PAs present in herbal products and carried out surveillance for PAs in milk.

Concern has also been raised over the possibility of foraging bees making honey from the nectar of tansy ragwort (Senecio jacobaea) which may contain PAs. Therefore, MAFF has carried out a small survey to determine the levels of PAs in UK honey produced by bees with access to ragwort. UK produced honey represents about 10 percent of the total UK honey market.

Current Surveillance

This survey was designed to obtain samples of honey during the ragwort flowering period. One hive was situated at each of three sites adjacent to where ragwort was growing. A sample of honey was taken from unsealed combs prior to the 1994 ragwort flowering season (1 sample) and at intervals during the flowering season (12 samples). Eight farmgate samples were obtained from small local producers to coincide with sampling from study hives. It was not possible to establish the sites of hives of the farmgate producers, the time at which these samples were bottled and whether the samples had been blended. In addition, two samples of a dark, waxy honey were donated by a small independent producer. This waxy honey was unpalatable and not suitable for blending with other honey.

In order to determine whether the bees had foraged on ragwort, analysis of the pollen content of the honey samples was carried out on 12 out of 13 study hive samples and all farmgate and donated samples (22 samples in total). The samples found to contain ragwort pollen were further analysed for jacoline, jacozine, jacobine, seneciphylline and senecionine, the five major PAs in ragwort, at CSL Food Science Laboratory, Norwich. Two samples, which did not contain ragwort pollen, were also analysed for PAs. The analysis for PAs is particularly difficult because, for example, pure standards are not available. However, a satisfactory method of analysis was developed at CSL Food Science Laboratory during this project.

Results

Eight honey samples (5 samples from study hives, 1 farmgate sample and 2 dark, waxy samples donated by a small honey producer) were found to contain ragwort pollen. The percentage of pollen which originated from ragwort ranged between 0.1 and 8.9. Ten samples of honey were analysed for PAs (6 study hive samples, 2 farmgate, 2 dark, waxy honey) of which 8 were the samples containing ragwort pollen. The results of this analysis are detailed in Table 1.

PAs were detected in 6 of the 8 samples containing ragwort pollen, at total concentrations ranging from 0.01 to 1.5 mg/kg. The 2 dark, waxy honey samples contained the highest total alkaloid concentrations, at 0.4 and 1.5 mg/kg respectively. Coincidentally, these dark, waxy honey samples were unpalatable and unlikely to enter the food supply through blending with other honeys. PAs were not present at detectable levels in the 2 samples which did not contain ragwort pollen grains. PAs were not detected in any of the retail samples obtained at the farmgate.

Interpretation

The presence of ragwort pollen grains in some samples indicates that bees will visit ragwort flowers. The presence of ragwort pollen grains in honey was usually associated with detectable levels of the PAs found in ragwort, but this was not always the case.

Estimates of intake of total PAs from honey have been made by assuming the maximum amount of honey consumed on any one occasion for adults (93g)¹, schoolchildren (60g)² and infants (32g)³ and are based on the highest detected level of total PAs (0.06 mg/kg) which would be likely to enter the food chain (see Table 2, below).

When considering comfrey, the Committee on the Toxicity of Chemicals in Food, Consumer Products and the Environment (COT) concluded that there was sufficient evidence linking the intake of comfrey to toxic effects in humans to warrant recommending that people should not consume preparations of comfrey which contain high levels of PAs. Comfrey tablets and capsules have been voluntarily withdrawn from sale by the health food trade but comfrey leaves for use in tea infusions are still available. Tea made from comfrey leaves contains low levels of PAs. Small amounts of PAs from this source are unlikely to do any harm. For the adult consumer, estimated daily intake of total PAs from the most contaminated, edible honey (0.006 mg/person/day) was one tenth of that from one cup of comfrey leaf tea infusion (0.06 mg/cup) and is no cause for concern. Estimated total PA intakes from the honey for schoolchildren and infants were less than for adults.

Little is known about the relative toxicities of PAs and their metabolism in humans. A MAFF-funded project on this subject is currently in progress. However, the data from this survey suggest that, compared with the ingestion of PAs from comfrey leaf tea, PA consumption from locally produced honeys is not a cause for concern, even when hives are situated near to ragwort plants. Indeed, the honey samples with the highest levels of PAs were coincidentally dark, waxy, unpalatable and unsuitable for retailing or blending with other honeys.

References

1. Gregory, J. Foster, K., Tyler, H. & Wiseman, M. (1990). The Dietary and Nutritional Survey of British Adults. (HMSO).
2. Report on Health and Social Subjects, No. 36: The Diets of British Schoolchildren. (HMSO).
3. Mills, A. & Tyler, H. (1992). Food and Nutrient Intakes of British Infants Aged 6-12 months. (HMSO).

Contact Point

Further information can be obtained from:

Mrs Chelvi Leonard
MAFF, Food Safety and Science Group
Food Contaminants Division
Room 210 Ergon House, c/o Nobel House
17 Smith Square
London SW1P 3JR

Tel: +44 (0) 171 238 5734
Fax: +44 (0) 171 238 6591

Return to Index to Surveillance Information Sheets, 1995 page
Go to top of page

These pages were last updated on 1 October 1996