Breath tests may lead to better reformulation: Study

The research, published in Flavour and Fragrance Journal​, suggests that using in vivo​ measurements of exhaled air may provide formulators information about the intensity and timing of aroma release for flavour components. The researchers said that such information could help industry to reformulate products to match more accepted flavour profiles.

“In this paper, commercial confectionery products and a simple commercial flavouring have been used to determine whether the concept of aroma release measurements to reformulate flavours can be applied to more complex food products,”​ said the researchers, led by Nicole Yang from the University of Nottingham, UK.

“The concept is to reformulate the flavour, so that aroma release during eating of both samples is the same … Potentially, the technique could be helpful as a tool for flavourists to quickly reformulate a flavour and then fine tune it using their skill and experience,”​ they said.

Flavour profile​

Yang and colleagues said that aroma release profiles are strongly linked to sensory assessment and overall perceived flavour sensation, and therefore may be a better predictor of sensory quality than the flavour content within a food.

The rationale, they said, is that release of flavours during eating depends on many physicochemical factors, including partition coefficients, volatility and solubility, as well as physiological factors of the individual, such as chewing rate, and saliva flow.

Therefore, they argue that the flavour profile in the intact food before eating may change significantly during consumption due to certain flavour components being released to different extents.

Yang and colleagues said that the development of new technologies has allowed the release of aroma compounds to be measured directly in vivo​ while subjects eat foods or consume drinks. During this process a portion of exhaled air is sampled continuously, and the release of aromas is monitored by analysing specific compounds characteristics.

Using such in vivo​ measurements of aroma profile, they suggested that flavour perceptions of similar products with different compositions (for example regular fat and low fat versions of products) could be matched.

Previous research using the same concentration of a single aroma compound added to regular and low fat milk showed that the aroma in low fat milk could be adjusted to match that of the regular milk using. Yang and co- workers said that this research showed the techniques was valid for use in simple systems, but noted that commercial foods typically contain multiple flavouring components.

The new research investigated the feasibility of in vivo testing on more complex foods, using two candies with different formulations and properties.

Study details​

Two candies were tested for their aroma release profile – a fat-free (pectin–sucrose gel) candy and a chewy candy containing protein and nine per cent fat.

The researchers said that these candies were chosen because it is known that a certain commercial strawberry flavouring performs well in pectin–sucrose gel but not in the nine per cent fat chewy candy.

They measured the release of the commercial strawberry flavouring in both candies when volunteers consumed them. The strawberry flavouring was reformulated so that its release profile in the chewy candy mimicked the release in the pectin-sucrose gel.

Yang and colleagues said that further sensory analyses comparing the two candies showed the reformulated flavour in the chewy candy performed significantly better than the original flavour.

“The work shows the translation of a laboratory-based concept into the commercial world and provides another tool to assist flavourists in reformulating flavourings to deliver food products with the desired sensory properties,”​ said the authors.

Source: Flavour and Fragrance Journal​
Published online ahead of print: 10.1002/ffj.2026
“Feasibility of reformulating flavours between food products using in vivo aroma comparisons”​
Authors: N. Yang, R.S.T. Linforth, S. Walsh, K. Brown, J. Hort1, A.J. Taylor