Recently, the number of people with dysphagia, which is defined as difficulty swallowing or the inability to swallow solid and/or liquid foods, is increasing in Korea owing to the rapid increase of the aging population [1]. It is known that people with inadequately managed dysphagia can be at risk of choking, aspiration, aspiration pneumonia, dehydration, weight loss, and malnutrition [2]. Therefore, commercial food thickeners are very useful in fluid foods for treatment of individuals with dysphagia, with the aim of increasing or modifying the fluid's viscosity to slow the flow rate of fluid transported through the pharynx, thus reducing the risk of aspiration [3]. However, the correct viscosity is very important for treating dysphagia because fluids with low viscosity may be more likely to enter the airway, resulting in aspiration, and those with high viscosity can be difficult to swallow, resulting in residue within the oropharynx, which is often aspirated [4].

Commercially available food thickeners for the management of dysphagia have starches or gums as the main active ingredient. However, it is well known that starch is hydrolyzed by the salivary amylase present in saliva, breaking it down into simple carbohydrates [2]. Therefore, it can be assumed that starch-based thickeners are highly sensitive to amylase in saliva, leading to the decreased viscosity of thickened fluids during consumption. In contrast, it can be expected that fluids thickened with gum-based thickener are less sensitive to thinning by saliva during consumption when compared to those prepared with starch-based thickeners. Even though the flow properties of thickened fluids with thickeners have been extensively studied, no research has been reported to date on the effect of human saliva on the flow properties of thickened fluids prepared with commercial food thickeners, except for Hanson et al. [5] who found that salivary amylase reduced the viscosity of drinks thickened to a custard-like consistency with starch-based thickeners marketed in the U.K. In the present study, three commercially available food thickeners, which are based on starch, gum-containing starch, and gum, are selected because they are widely used as favorable food thickeners in Korea. In addition, viscosity measurements are performed at a pudding-like consistency because this is a common target viscosity in general practice in Korea. Thus, this study focuses on the effect of human saliva on the viscosity of pudding-like thickened waters prepared with different food thickeners marketed in Korea as a function of salivary reaction time and thickener type.

The shear stress (σ) versus shear rate ($\stackrel{·}{\mathrm{\gamma }}$) data for thickened water samples added with saliva at the control (no added saliva) and 60 min after addition of saliva at 25℃ are shown in Figure 1. The experimental results of σ and $\stackrel{·}{\mathrm{\gamma }}$ of all samples, except for SB with saliva, were well fitted to the simple power law model (Eq. 1) with high determination coefficients (R² = 0.96-0.99) (data not given). In this study, the flow curves of SB with saliva could not be obtained because of the formation of an unstable structure of SB owing to the addition of saliva. Therefore, for SB samples with added saliva, we measured the viscosity (η_a,50) only at a shear rate of 50 s^-1 without the application of the power law model. Thickened samples with different thickeners also exhibited a clear trend of shear-thinning behavior. In particular, the GSB and GB samples exhibited higher shear-thinning behavior when compared to SB. There was a difference in flow curves of GSB between the control and the GSB sample with added saliva at a reaction time of 60 min (Figure 1). In addition, GSB and GB food thickeners also produced different flow properties at a 60 min of salivary reaction time after the contact with saliva.

Table 1 shows the apparent viscosity (η_a,50) of thickened waters with three different thickeners (SB, GSB, and GB) as a function of salivary reaction time (0-60 min) after the addition of saliva. The reduction in viscosity for SB and GSB on saliva addition was significant at 0 min and 20 min of reaction time, respectively. Figure 2 also shows the % changes in initial viscosity (control) values based on the control samples. For SB, the viscosity values greatly reduced after the contact with saliva irrespective of the reaction time, showing the greatest reduction of 99.9% of initial viscosity and no difference in η_a,50 values between salivary reaction times of 0 and 60 min, while for GSB, the viscosity values significantly reduced after 20 min of reaction time and then retained stable (12.2-12.7% of initial viscosity). In contrast, there was no significant difference in the viscosity values of GB between the control and the sample mixed with saliva. These results were confirmed from the photographs of the viscosity differences between the control and the thickened water with added saliva, as shown in Figure 3.

Thickened fluids prepared with commercial food thickeners are commonly used for consumption by used for dysphagic patients patients, who take a long time to eat or drink owing to swallowing difficulties. Therefore, from a clinical perspective, it is very important to observe how food thickeners perform with thickened fluids when mixed with human saliva over an appropriate salivary reaction time because the thinning of thickened fluids after contact with saliva has been reported [5]. Figure 1 shows that the GB and GSB with no added saliva showed higher shear-thinning flow behaviors and lower viscosity values compared to those for SB, possibly owing to the presence of xanthan in the food thickener [6]. Urlacher and Noble [7] also reported that such high shear-thinning behavior may be owing to the unique rigid, rod-like conformation and high molecular weight of xanthan gum. These observed results followed similar trends to those previously found in thickened fluids prepared with various xanthan gum-based thickeners [8]. A noticeable difference was observed in viscosity values between food thickeners as a function of salivary reaction time (0-60 min), as shown in Table 1 and Figure 2. Figure 3 also shows the photographs of the viscosity differences between the control and thickened water with added saliva. In comparison of η_a,50 values of the SG and GSB samples, the lower viscosity values of the samples mixed with saliva can be explained by the hydrolysis of the starch component contained in SB and GSB due to the salivary amylase present in saliva [9]. Among food thickeners, GB was not influenced by salivary amylase owing to the inactivation of amylase with gums, showing that GB is the most suitable food thickener for dysphagic patients. These results suggest that the addition of saliva had a more pronounced effect on the viscosity of thickened fluids prepared with food thickeners (SB and GSB) containing starch. Similar observations were reported by Hanson et al. [5] for food thickeners composed of corn starch alone or corn starch with the addition of gums. From these results, it was shown that dysphagic patients cannot receive thickened fluids with the correct viscosity for the swallowing safety because thickened fluids with food thickeners containing starch are greatly influenced by the type of thickeners in the presence of saliva, suggesting the need for the preparation of thickened fluids with a stable viscosity after contact with saliva in mouth. Therefore, it can be concluded that GB is a better food thickener for the treatment of dysphagic patients because the GB thickened fluids are not influenced by saliva during the consumption of fluid foods.

Changes in the viscosity of thickened waters over salivary reaction time are more pronounced in a food thickener containing starch alone owing to the complete breakdown of starch caused by the strong activation of salivary amylase. Therefore, it is desirable to use food thickeners containing gums for the treatment of patients with dysphagia because gums in food thickeners are not susceptible to digestion by amylase and provide a stable viscosity. These results will be useful for developing new food thickeners or thickening products with a stable viscosity in the presence of human saliva. However, more investigations need to be carried because the viscosity of thickened fluids with saliva can be greatly influenced by the amount of salivary amylase. In terms of clinical practice in patients with dysphagia, an additional study is also needed on various thickened fluids with saliva obtained from dysphagic patients to extend the results of this study.