Number 77      November 1995

MAFF UK - SINGLE SEED VEGETABLE OIL SURVEILLANCE EXERCISE

Index to MAFF UK Food Surveillance Information Sheets, 1995

See also:
180: MAFF, UK - Authenticity of Olive Oils (June 1999)

Background

A wide variety of edible vegetable oils is available to the consumer, from a number of geographical origins. Certain oils (e.g. groundnut) trade at a significant premium in comparison with others (e.g. rapeseed). The price difference between various oils could provide some incentive to substitute more expensive oils with cheaper ones.

The MAFF Working Party on Food Authenticity (WPFA) decided that a surveillance exercise should be carried out to investigate whether premium oils (i.e. groundnut, sunflowerseed, maize and palm oils) which claimed to be pure were being adulterated with less expensive oils. The WPFA recommended that the purity of the samples should be assessed using a variety of analytical techniques including assessment of ¹³C/¹²C stable isotopic ratios (SIR), fatty acids, desmethylsterols and tocopherols in oils.

Samples

A total of 291 samples were collected from retailers (including wholesalers and cash and carries) and manufacturers by Trading Standards and Environmental Health Departments from: Belfast, Berkshire, Cambridgeshire, Devon, Dyfed, Humberside, Islington, Lancashire, Shropshire, and Strathclyde. In addition, a number of samples were collected by MAFF officials from retail outlets in the South East of England and East Anglia during May to September 1994. Approximately 71 percent of the samples were for sale through retail outlets. The remaining 29 percent were for catering use. The samples included the major and lesser known brands, as well as retailer own-label products. The products were manufactured in the UK and throughout Europe, and were of the following types: maize oil (79), groundnut oil (83), sunflowerseed oil (111), and palm oil (18).

Analytical Approach

The samples were analysed in two stages: screening and confirmatory.

Screening analyses: The fatty acid composition (FAC) of all oils was determined using British Standard methods BS 684:2.34 and BS 684:2.35 (equivalent to ISO 5508 and ISO 5509). The iodine values of the samples were calculated from the FAC. The SIR of all the maize oils were established. Although there are no international standards for this measurement relating to edible oils, it is a relatively simple technique and has been used routinely in many laboratories for several years. Furthermore, SIR has been used for establishing the purity of other commodities and has been used in enforcement action.

Confirmatory analyses: From the results of the screening analyses, those samples that appeared to contain in excess of 3percent of another oil (i.e. suspect samples) were analysed further to determine the desmethylsterol composition and concentration. This was conducted using BS 684:2.38 (1992), which is equivalent to ISO 6799 (1991). Particular attention was paid to the presence of elevated concentrations of brassicasterol, which is indicative of the presence of rapeseed oil. Those oils that appeared to contain 3 percent of another oil but were found not to contain unusual concentrations of brassicasterol, were further analysed to determine their tocopherol composition. The analysis was carried out using BS 684: 2.40 (ISO/DIS 9936). Particular attention was paid to the concentration of delta-tocopherol, as high levels of this are indicative of the presence of soyabean oil.

Authenticity Limits Applied and Data Interpretation

The refining operation applied to crude oils is conducted on a large scale, with one oil passing through the system immediately after another. Consequently, it is generally accepted that unavoidable adventitious contamination of one oil with another will occur during storage, transport and processing. However, if good manufacturing practice is adhered to, it is unlikely that an oil will contain more than 1-2 percent of the previous product that had passed through the refining process. Account of this has been made when assessing the analytical results.

For the purposes of this survey, it has been considered that an oil found to contain less than 3 percent of an undeclared oil is not adulterated. Samples found to contain undeclared oils at levels of between 3 percent to 5 percent may not necessarily be adulterated but contain higher levels of undeclared oils than would be expected in good manufacturing practice. Samples found to contain more than 5 percent of other oil(s) are likely to have been deliberately adulterated.

The database of purity criteria used to assess the authenticity of the samples has been produced during the past 15 years in a project part funded by MAFF. During the development of the database the oils were extracted in the laboratory from oilseeds, kernels and nuts, of known origin, and, therefore, can be guaranteed pure.

In this study oil samples have not been considered adulterated unless the chemical characteristics of the oil lie outside those of the appropriate purity criteria. For example, pure maize oil samples contain between 0 and 30 mg/kg brassicasterol, with a mean concentration of 6 mg/kg. Nonetheless, on the basis of the brassicasterol level, a sample described as maize oil was not considered impure unless it contained in excess of the maximum found in a pure maize oil (e.g. 30 mg/kg). Furthermore the refining process used to make crude oils edible changes certain chemical characteristics of the oil, in particular reducing the level of desmethylsterols and tocopherols present. The majority of the oils studied in this exercise were refined. In the absence of a database of purity criteria for refined oils, the purity of the oil samples were assessed against values for crude oils. This approach to interpreting the results will tend to underestimate the degree of adulteration in the survey samples.

Results

The results of the survey are described in the table below:

The bulk of the oils (82 percent) were found to contain less than 3 percent of another oil. However, 11 percent of samples contained 3 percent to 5 percent of an undeclared oil, and 7 percent were found to contain in excess of 5 percent of an adulterating oil. Of particular concern, was the finding that 5 percent of the oils tested were adulterated at levels of more than 8 percent. The majority of adulteration is associated with rapeseed oil.

Conclusions

This survey suggests that although the bulk of edible oils at retail are not adulterated, a significant proportion (approximately 7 percent) contain in excess of 5 percent of another oil, which in many cases has been identified as rapeseed oil.

Rapeseed oil and soyabean oil trade at lower prices to the premium products examined in this study, and, on occasions, it is clear that mixing of these with groundnut oil, sunflowerseed oil, and, in particular, maize oil has taken place. There is a commercial advantage in such action.

Follow-up Action

Individual Trading Standards and Environmental Health Departments that participated in this study are being informed of the results for the samples they collected. Once the WPFA has considered and approved a more detailed report on the findings of the exercise it will be placed in the MAFF Whitehall library. The industry, trade associations and consumer groups will be made aware of the results of the study.

Contact Point

For further information on this survey please contact:
Ms Catriona Stewart,
MAFF, Food Safety and Science Group
Food Labelling and Standards Division,
Room 323, Ergon House,
c/o Nobel House,
17, Smith Square,
London SW1P 3JR

Tel: (+44) (0)171 238 6112
Fax: (+44) (0)171 238 5782

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

These pages were last updated on 1 October 1996