scope:

Background

Historically, white oils and waxes have an established use in applications related to food processing and packaging. For a large number of years, these materials were produced by acid (oleum) treatment of solvent-extracted refinery streams to yield products of the required physical and technical specification. More recently, however, this "conventional" route of manufacture has been supplanted by a hydrogenation process ("hydrotreatment"). Extensive analytical data comparing products produced by both techniques were obtained (Concawe, 1984) and formed the basis of an industry conclusion that hydrotreatment oils were, in many respects, highly similar to their acid treated counterparts.

No toxicological information was available to support the safety-in-use of either class of product, resulting in regulatory calls for studies to support the continued use of these materials. Independent studies were conducted during the 1980s by at least two companies to compare the biological effects of hydrotreated versus acid-treated oils (American Petroleum Institute, 1992). Fundamentally, these studies produced some conflicting results with respect to accumulation of the mineral oils and the effects observed in certain organs (liver, spleen and mesenteric lymph nodes). Several differences between these studies that were possibly related to the test material led to selection of a number of variables to be evaluated in this subsequent 90 day feeding study. These included crude oil type (naphthenic and paraffinic), refining process (acid-treated versus hydrogenation) and viscosity for mineral oils or melting point for waxes. Likewise, these variables for mineral oils were also identified by the EC SCF in their 1990 list of substances approved for food packaging for which additional toxicity data were required.

Sample selection

Based on information provided to Concawe on typical ranges for food-grade mineral hydrocarbon products, samples were selected to represent products currently in use in the market place in Europe. Consideration was also given to those products produced in the US since many products which contain food-grade mineral hydrocarbons are marketed internationally.

Samples were selected as follows:

1. low viscosity paraffinic, hydrogenated (P15 (H))

2. high viscosity paraffinic, hydrogenated (P100 (H))

3. low viscosity naphthenic, hydrogenated (N15 (H))

4. high viscosity naphthenic, hydrogenated (N70 (H))

5. low viscosity naphthenic, acid treated (N10 (A))

6. high viscosity naphthenic, acid treated (N70 (A))

7. hydrotreated paraffinic wax, low melting point (low melting point wax)

8. hydrotreated microcrystalline wax, high melting point (high melting point wax)

9. clay treated microcrystalline wax (high sulphur wax)

As a consequence of the different refining process, the sulphur content will be higher in the clay treated wax. Therefore it was referred to as "high sulphur wax".

In addition to an untreated control group, an oil of biological origin. was selected as another type of control to assist in determining whether other saturated oils accumulate in the body and in evaluating the effect of high dietary oil content on absorption of food and fat-soluble nutrients. Coconut oil was selected as a biologically-derived oil which is similar in terms of degree of saturation (92% saturated, 6% mono-unsaturated, 2% polyunsaturated) and carbon chain length (C8-C20) to the lowest viscosity liquid mineral oil products.

The viscosity of oils selected to represent a "low" or "high" viscosity product do not, and are not intended to, represent the extreme values for these parameters and for those products currently marketed. They are, however, "close to the extremes" and, thus, are considered representative.

Protocol considerations

The protocol design was selected to represent a compilation of those referenced by OECD, EEC and US FDA guidelines for 90 day studies.

Strain selection

The Fischer 344 rat strain was chosen because it had been used in a previous study (Baldwin et al., in press).

Dose selection

Ten-fold decremental dietary concentrations of mineral hydrocarbons, from 2% to 0.002% (w/w), were selected based on concentrations where effects were observed in previous studies. Previous studies indicated that biological effects occurred at dietary concentrations of 2.0, 0.2 and 0.02%, but that the no observable effect level (NOEL) exceeded 0.002%. Based on rat food consumption, these concentrations result in approximate average daily consumption values of 1500, 150, 15 and 1.5 mg mineral hydrocarbon/ kg body weight/day. These concentrations were considered to represent an adequate dose range for risk assessment purposes.