Following a request from the European Commission, a revised exposure assessment of steviol glycosides from its use as a food additive was carried out, for children and adults, based on the revised proposed uses presented in the terms of reference.
Several food consumption databases were used to conduct the revised exposure assessment. For children, data from EXPOCHI project and UK NDNS survey were used. Estimates for adults were based on UK data only.
For adults, exposure estimates give a mean dietary exposure to steviol glycosides, expressed as steviol equivalents, of 1.9-2.3 mg/kg bw/day, and of 5.6-6.8 mg/kg bw/day for high level exposures (97.5th percentile), with main contributors being non-alcoholic flavoured drinks (soft drinks), tabletop sweeteners and beer and cider.
For European children (aged 1-14 years), exposure estimates give a mean dietary exposure to steviol glycosides, expressed as steviol equivalents, of 0.4-6.4 mg/kg bw/day; at the high level (95th/97.5th percentile), exposures estimates range from 1.7 to 16.3 mg/kg bw/day.
Considering the limitations of consumption data, estimates can be considered conservative as in the ANS Panel opinion (2010).
The EFSA Comprehensive database was used to identify and assess the uncertainty due to the consumption data from the food group non-alcoholic flavoured drinks (soft drinks), the main contributor for both children and adults.
From these data, no general tendency was observed regarding the difference between the consumption levels of low-calorie non-alcoholic flavoured drinks (soft drinks) and of normal-calorie soft drinks. Therefore, the consumption levels of normal-calorie soft drinks was considered as an acceptable approximation of the consumption levels of low-calorie soft drinks and thus the exposure estimates for adults were not corrected.
For children, the data from EXPOCHI for the consumption of non-alcoholic flavoured drinks (soft drinks) were found to be generally higher than the consumption of low-calorie soft drink from the EFSA Comprehensive database by a factor 2. Consequently, the data from the EFSA Comprehensive database were used to correct the consumption of this food group. The corrected exposure estimates for children high consumers (95th percentile) range from 1.0 to 12.7 mg/kg bw/day.
The revised mean exposure estimates differ only slightly from the exposure estimates given in the ANS Panel opinion (2010). By using the EFSA Comprehensive database, the upper range of high level exposure estimate decreased from a maximum of 17.2 from the ANS Panel opinion to 12.7 mg/kg bw/day for children but high consumers children exposures are still above the ADI of 4 mg/kg bw/day for several countries.