You are here

Toxicity of tartrazine

Scientific review report

1 February 2014

Book pagination

7. Hypersensitivity and intolerance in humans

7.1. History

The major controversy in the field of artificial food colours is the suggestion first made in the 1920s that artificial food colours and additives, including tartrazine, may have detrimental effects on children inducing 'hyperactivity' (Burrows 2009, cited by Arnold et al., 2012). A specific hypothesis relating to this relationship was developed in 1973 by Feingold (1975) who proposed that hyperactivity and learning problems in children were due to certain foods and food additives as well as foods containing natural salicyclates. The work was criticised by the medical profession. However, his hypothesis was accepted by many parents following media reports (comprehensively reviewed by Arnold et al., 2012).

Subsequent investigations failed to demonstrate a link with hyperactivity due to deficiencies of many of these studies (Arden & Ram, 2001; Arnold et al., 2012; Conners et al., 1976; Dipalma 1990; Khera & Munro 1979; Levy et al., 1978; Rowe 1988; Settipane 1977). In trying to identify the biological process(es) underlying this potential relationship, several hypotheses were put forward based on neurochemical, genetic or allergic/immunological mechanisms. However, no definitive biological process was established (reviewed by FDA/CFSAN 2011).

Over the years, interest waned but was revived in 2004 with a study published by Schab & Trinh (2004; see below).

7.2. Recent human studies

The following are brief summaries of reviews or studies conducted during the last 12 years.

In 2001, Ardern & Ram reviewed 18 relevant studies in humans, but found they were inconclusive as none of them conducted tartrazine challenge or avoidance in diet nor did they significantly alter asthma outcomes. Schab & Trinh (2004) conducted a meta-analysis of findings from previously conducted clinical trials that attempted to show a definitive relationship between consumption of artificial food colours, including tartrazine, and behavioural changes in children. The study was, however, deficient in analysing objective measures of behaviour such as clinical/psychological evaluations, activity monitoring or behavioural testing; and was universally criticised for its emphasis on behaviour ratings as reported by parents, teachers or clinicians (Arnold et al., 2012; Elhkim et al., 2007; FDA/CSAN 2011; Watson 2008).

Thus, no clear relationship between ingestion of food colours (including tartrazine) and the development of attention deficit hyperactivity symptoms in children (Arnold et al., 2012) or development of intolerance reaction (Elhkim et al., 2007) was established. The use of non-standardised diagnosis, questionable sample selection, imperfect blinding and non-standardised outcome measures utilised by previous investigators may have been key factors contributing to the ambiguity surrounding tartrazine consumption and these reactions.

Interest in the relationship between food colours and hyperactivity in children was revived again in 2007 when McCann et al. (2007) published a study conducted at the Southampton University. The study implied that mixtures of certain artificial food colours and sodium benzoate could increase the mean level of hyperactivity profile behaviours in two age groups of children (3-4 years old and 8–9 years old) from the general population. However, lack of consistency in the results with respect to the age and sex of the children and the type of observer (parent, teacher, or independent assessor); the unknown clinical relevance of the effects measured; the use of mixtures in the study and lack of information on dose-response resulted in the European Food Safety Authority (EFSA) rejecting suggestions of a direct link between artificial food colours and hyperactivity (Watson 2008).

Furthermore, in response to a petition from the Center for Science in the Public Interest (CSPI 2008) arising from the Southampton study, the FDA also conducted a thorough review of the McCann et al. 2007 study as well as publications of previously conducted clinical trials (33 trials), including the 2004 meta-analysis by Schab and Trinh (2004).

The concerns of the FDA neurotoxicology/toxicology panel were in line with the previously identified deficiencies of McCann et al. 2007 study by the EFSA (Watson 2008). Thus, the FDA concluded that a causal relationship between exposure to tartrazine (and other colour additives) and hyperactivity in children in the general population could not established.

Neither EFAS nor the FDA found any compelling evidence to alter current regulations on acceptable daily intakes of tartrazine in foods, drugs and cosmetics.

Top of page

Book pagination