Assessment of the influence of energy under-reporting on intake estimates of four food additives

Under-reporting has been identified as an important source of uncertainty in food chemical exposure assessments. The objective of the present study was to assess the influence of under-reporting on food additive intake estimates. Dietary survey data were derived from the North-South Ireland Food Consumption Survey (2001). Data from the Republic of Ireland (n = 958) were used. Energy under-reporters were identified using a ratio of energy intakes to estimated basal metabolic rate. First, food categories (n = 26) included in an assessment of exposure of four food additives were created and patterns of food intakes (i.e. likelihood of consumption, frequency of consumption and reported portion size) between acceptable and under-reporters compared. Second, for each food additive, deterministic intake estimates for the total sample (i.e. acceptable and under-reporters), under-reporters and acceptable reporters were calculated and compared. Differential reporting of the majority of food categories between acceptable and under-reporters was recorded. Under-reporters were less likely to record the consumption of a given food and more likely to under-report the frequency of consumption and portion size compared with acceptable reporters. Food additive intake estimates amongst acceptable reporters were higher than corresponding intake estimates amongst the total sample of reporters and amongst under-reporters. With the exception of one food additive (erythrosine), ratios of upper percentile additive intakes amongst acceptable reporters to corresponding intake estimates amongst the total sample of reporters did not exceed 1.06 when results were expressed as total population or consumer-only intakes. Findings illustrated that energy under-reporting does not materially influence estimates of food additive exposure based on the four food additives studied. However, a number of situations were identified where the under-reporting might exert a more significant impact on resulting exposure estimates.