The Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) has been asked to evaluate the role of tobacco additives in the addictiveness and attractiveness of tobacco products.
The criteria for dependence established in humans indicate that tobacco has a high addictive potential, but it remains difficult to assess the addictiveness of individual additives. In animal studies the addictive potency of the final tobacco product cannot be assessed. The reinforcing potency of drugs is measured after intravenous injections and suggests that the abuse liability of pure nicotine is weaker than the addictive potential of tobacco products in humans. The currently used methods to define addictiveness of nicotine and additives are thus not considered adequate.
[…]
No tobacco additives which are addictive by themselves have so far been identified. However, sugars, polysaccharides and cellulose fibres which are naturally present in tobacco, or sugars added in high quantities to most tobacco products, give rise to numerous aldehydes, such as acetaldehyde, in tobacco smoke. Acetaldehyde given intravenously is self-administered and enhances the addictiveness of nicotine in experimental animals. Additives that facilitate deeper inhalation (e.g. menthol) or inhibit the metabolism of nicotine may enhance the addictiveness of nicotine indirectly. Substances such as ammonia that increase the pH of the tobacco and the smoke, result in higher amounts of uncharged nicotine. However, it is uncertain if more nicotine is absorbed with higher smoke pH. For smokeless tobacco it seems that an increased pH enhances nicotine absorption in the mouth.
The methods used to quantify the addictive potency of additives have limitations because of technical challenges in experimentally manipulating the presence or absence of an additive in a tobacco product. Such experiments require large technical and financial resources. In addition, there are ethical issues if testing in humans is considered. Due to these limitations, the available methodologies are not considered adequate. [From the author's abstract]