Safety evaluation of certain food additives - ipcs inchem

184 STEVIOL GLYCOSIDES (addendum)
At its fifty-first meeting, the Committee evaluated toxicological data on
stevioside and the aglycone steviol (Annex 1, reference 137) and specified further
information needed. Based on new data and information, at its sixty-third meeting
(Annex 1, reference 173), the Committee determined that the commercial material
should be known as “steviol glycosides” and established tentative specifications for
material containing not less than 95% of the total of four specified glycosylated derivatives
of steviol (i.e. stevioside, rebaudioside A, rebaudioside C and dulcoside A).
Additionally, the sum of stevioside and rebaudioside A content was specified at not
less than 70% of the four steviol glycosides.
Also at its sixty-third meeting, the Committee reviewed additional biochemical
and toxicological data on the major steviol glycosides and on the aglycone
steviol. The Committee noted that steviol glycosides are poorly absorbed and are
metabolized by the intestinal microflora by successive hydrolytic removal of glucose
units to the aglycone, steviol, which is well absorbed. Therefore, the toxicity
of the glycosides was related to the steviol content. A temporary acceptable daily
intake (ADI) of 0–2 mg/kg body weight (bw) for steviol glycosides expressed as
steviol was established on the basis of the no-observed-effect level (NOEL) 1 of 2.5%
stevioside in the diet, equal to 970 mg/kg bw per day, or 383 mg/kg bw per day
expressed as steviol, in a 2-year study in rats and with a safety factor of 200. In the
groups at 5%, final survival rates in the males and body weight gain and absolute
kidney weights in both sexes showed significant reductions compared with those
in controls. The overall safety factor of 200 incorporated a factor of 2 related to
the need for further information, to be provided by 2007, on the pharmacological
effects of steviol glycosides in humans. The Committee specified the need for
studies involving repeated exposure of normotensive and hypotensive individuals
and patients with type 1 (insulin-dependent) and type 2 (non-insulin-dependent)
diabetes to dietary and therapeutic doses. This was because the evidence available
at the time was inadequate to assess whether the pharmacological effects of steviol
glycosides would also occur at dietary exposure levels, which could lead to adverse
effects in some individuals (e.g. those with hypotension or diabetes).
Also at its sixty-third meeting, the Committee estimated international intakes
of steviol glycosides to be in the range of 1.3 (African diet) to 3.5 mg/kg bw per day
(European diet), expressed as steviol, assuming that all dietary sugars (total sugars
and honey) are replaced by steviol glycosides. The Committee acknowledged that
this was a conservative estimate and that actual intakes were likely to be 20–30%
of this figure.
At its sixty-eighth meeting (Annex 1, reference 184), the Committee considered
the information that had become available since the sixty-third meeting. This
comprised two submissions, which included a summary of published toxicological
studies and some unpublished data, additional information identified from the
scientific literature and responses intended to resolve the outstanding issues
1 Before the sixty-eighth meeting, the Committee used the term NOEL to include the current
definitions of both NOEL and no-observed-adverse-effect level (NOAEL). According to the
decision taken by the Committee at its sixty-eighth meeting (Annex 1, reference 187), this
NOEL would now be termed a NOAEL.