Relationship between International Dysphagia Diet Standardisation Initiative Flow Test and Consistometric Measures for Consistency Classification: An Examination of Thickened Liquids Prepared Using Starch-Based and Xanthan Gum-Based Thickening Agents

Swallowing is a complex physiological process involving a series of highly synchronized muscle activities, facilitating the passage of liquid and solid food to the stomach [1]. Dysphagia refers to difficulties in swallowing during the transport of a food bolus from the oral cavity through the pharynx and esophagus to the stomach [2, 3]. It may stem from structural and functional deficits in respiratory, nervous, and muscular systems [1, 2]. Depending on its severity, dysphagia may lead to complications that could negatively impact an individual’s quality of life [4]. The prevalence of dysphagia has been estimated to be around 8% worldwide [5]. Recent studies reported up to 61% of the institutionalized geriatric populations in Hong Kong exhibited dysphagic symptoms [6]. Without proper and timely intervention, dysphagia may escalate to life-threatening conditions such as malnutrition, dehydration, choking, airway obstruction, penetration, and aspiration pneumonia [7].

In managing dysphagia, commercial thickening agent is often used to change the thickness or consistency of liquid food to be swallowed. It is believed the altered consistency might lead to a slower flow of liquid food and thus more time for swallowing events to happen, and in turn allowing better coordination among swallowing muscles to reduce the chance of aspiration, and eventually enhancing swallowing safety [7]. Efforts have also been made to search for natural food alternatives to commercial thickeners for cost and/or nutrition reasons [cf. 8, 9]. Currently, starch-based and xanthan gum-based thickening agents are commercially available [10, 11]. Starch-based thickening agents contain modified starch granules composed of amylose and amylopectin [12‒14]. Upon contact with liquid, starch granules absorb water and swell, increasing the thickness of the mixture. However, starch-based thickeners exhibit a notable disadvantage: the resulting thickness is relatively unstable and can be influenced by external factors such as resting time, temperature, and contact with saliva [13‒16]. Although the evidence supporting the significant improvement of nutritional status and/or risk of pneumonia through the use of thickened liquids is limited [17‒19], their application remains an essential component of dysphagia management [20, 21]. Xanthan gum is a commercial biopolymer that can be used to increase liquid thickness even with a very small quantity (0.05–2%) [22]. It is widely used in bakery and salad dressing. Such thickened liquids have been shown to maintain a more stable thickness over time compared to their starch-based counterparts [13‒15]. Recent studies also reported that xanthan gum-based thickened liquids were better and safer than starch-based thickened liquids in managing dysphagic patients [14, 23, 24]. Regardless of thickener type, stability of thickened food was crucial for ensuring accurate consistency measurements. The optimal consistency for safe consumption often depends on the severity and pathophysiology of an individual’s dysphagic condition, objective measurements of liquid consistency must be established to ensure effective management.

Over the years, the use of consistometry in thickness measurement for liquid food has been common in providing standardized thickness measurements [25‒27], particularly in the food industry, in addition to other objective measurements such as fork drip test, syringe test, and ball back extrusion technique [24, 28]. A Bostwick consistometer can be used to objectively describe liquid consistency by referring to the distance the liquid traveled in the consistometer’s trough after a predetermined duration [29]. A consistometer is a simple and dependable instrument originally designed for the oil well cementing industry to measure consistency of grout [30]. Nowadays, consistometers are commonly used in the food industry to quantify consistency of viscous materials based on the distance the material flows under its own weight in a given time interval [31]. The consistometer has been used to assess the consistency of thickened drinks for individuals with dysphagia in the clinical settings [26, 27]. Researchers attempted to align consistometric measurements with traditional consistency labels such as nectar, honey, and pudding. However, practical challenges including equipment accessibility for nonprofessional stakeholders and out-patients were seen [5, 16]. The International Dysphagia Diet Standardisation Initiative (IDDSI) framework was recently developed to provide a universal standardized nomenclature for defining thickness of food and drinks, with particular emphasis on culturally sensitive terms and tool accessibility [32]. According to the IDDSI Flow Test, liquid thickness can be classified based on the amount of liquid remaining in a 10-mL BD syringe after 10 s of liquid flow [33]. Instead of relying on descriptors drawn from existing foods, consistencies are designated as levels 0 to 4, representing “thin,” “slightly thick,” “mildly thick,” “moderately thick,” and “extremely thick,” respectively.

However, the IDDSI Flow Test has limitations. For example, it may not accurately reflect IDDSI levels 3 and 4, as liquids at these thicknesses may not flow effectively in the syringe [34]. Additionally, bubbles, residue, or lumps in the liquid, which are common in drinks such as soup and freshly made juice, can lead to inaccuracies in the flow test due to blockage of the small nozzle on the syringe [35]. Classifications and reference values for different liquid consistencies associated with consistometric test and IDDSI Flow Test are shown in Table 1.

Although consistometer measures consistency and the IDDSI Flow Test measures thickness, these terms are often used interchangeably in the literature on dysphagia and diet modification [27, 36]. Consistometric measurements allow for assessments of more viscous liquids (IDDSI levels 3–4) and liquids with lumps, residue, or bubbles [25].

Previous research reported strengths and weaknesses associated with different measurement methods of liquid consistency [16, 25]. The IDDSI Flow Test appears to be more sensitive to thinner liquids; however, moderately and extremely thick liquids often demonstrate a ceiling effect, resulting in minimal or zero flow out of the syringe tip [16, 25]. In contrast, consistometry may not adequately measure thinner liquids for similar reasons [16, 25]. Regarding precision, the IDDSI Flow Test provides four levels with broad reference value boundaries, which could result in liquids with notably different flow behaviors being categorized as the same consistency level. The relatively finer measurable intervals in the consistometer may suggest higher measurement precision [27]. With the different classification nomenclature, the relationship between the consistometer’s measurements and IDDSI classification framework remains unclear.

Research examining the relationship between the IDDSI framework and consistometric measurements of liquid thickness is lacking. Recently, Côté et al. [25] reported a strong correlation between the two measurements; however, the data may not be interchangeable and might not fit well within a linear conversion model. Considering the differences in the underlying mechanism and complexity of non-Newtonian liquids, a nonlinear relationship might exist between IDDSI and consistometric measurements. Yet, several methodological concerns in the study by Côté et al. [25] could have influenced the validity of the findings and their applicability to institutional settings. First, the liquid samples were prepared using specific commercial products, and the composition of the thickening agent in each sample was not controlled. Variations in thickener type can significantly alter rheological characteristics of thickened liquids [10, 16], which might have potentially interacted with the testing instrument [25]. Consequently, generalizability of findings to other thickening agents is questionable. By not addressing starch-based and xanthan gum-based thickeners separately in the regression, the conclusion regarding the linearity relationship may be compromised. Furthermore, the samples were prepared using both water and juice. The reaction of thickeners to different base liquids can vary significantly due to differences in ingredient composition, pH values, and molecular weight [10, 37], making it unclear whether the results can be applied to specific types of beverages without standardization of the base liquid. Moreover, the study design by Côté et al. [25] did not align well with common clinical parameters. The consistency of thickened liquids was measured at a serving temperature of 8°C. Since the viscosity of starch-based thickened liquids is temperature-dependent [13, 37], results obtained at this single temperature point may not be transferable to room temperature. Additionally, the researchers used pre-prepared and refrigerated liquids rather than freshly mixing thickening agents with beverages at the point of consumption, which is more common in clinical settings. This introduced the variable of “setting time” into the equation. Starch-based thickened liquids are known to exhibit high instability during initial short periods and can absorb significant moisture during storage in humid environments [11, 37‒39]. Thus, the clinical implications of the results remain uncertain. Seeing the above issues, the present study aimed to establish (1) the relationship between the thickness measurements using IDDSI Flow Test and Bostwick consistometer, (2) the measurement limitations of each test, and (3) the adaptation of consistometry to the IDDSI framework, for thickened liquids prepared using starch- and xanthan gum-based thickening agents.

The study adopted a quantitative descriptive empirical design. Ethics clearance of research was not required as no human participants were involved in the study.

To prepare the thickened liquids of different consistencies, two types of thickening agents were used: the starch-based thickening agent (ThickenUp^®, Nestlé Health Science, Switzerland) containing mainly modified food starch (maize) and the xanthan gum-based (ThickenUp^® Clear, Nestlé Health Science, Switzerland) with ingredients of xanthan gum, potassium chloride, and maltodextrin were used. They represented the most common commercial thickeners used in hospitals and elderly care centers in Hong Kong. The amount of thickening agent for preparing liquids of four consistencies conforming to IDDSI consistency labels using 200 mL of water is shown in Table 2, as per manufacturer’s recommendations. To minimize the effect of temperature on consistency [13], all preparations were carried out under room temperature (about 25°C). To encourage reproducibility of results, thickening powder was measured in grams using a household precision electronic scale (KD-321, Tanita, Japan) (±0.1 g) instead of scoops.

According to Cichero et al. [5], IDDSI Flow Test was most suitable for measuring consistency of liquids of slightly thick, mildly thick, and moderately thick, but not for extremely thick liquids. Therefore, only IDDSI consistency levels 1, 2, and 3 were considered when preparing thickened liquids in the present study, with an increment of 0.4 g for starch-based thickener and 0.6 g for xanthan gum-based thickener. In addition, restricted by the testing range of IDDSI Flow Test and Bostwick consistometer, with 200 mL of water, starch-based and xanthan gum-based thickeners with amount <7.2 g and <1.2 g, respectively, resulted in a ceiling effect in consistometric measures, while those with amount >12.0 g and >8.4 g, respectively, resulted in a floor effect in IDDSI Flow Test. Consequently, starch-based thickener with weights between 7.2 g and 12.0 g and xanthan gum-based thickener with weights between 1.2 g and 8.4 g, both resulting in 13 distinct levels of consistencies, were used.

To prepare for the thickened liquids, thickening powder of the correct amount was transferred to its individual beaker, then poured slowly into 200 mL of boiled tap water cooled to room temperature (∼25°C), and stirred clockwise using a fork for 1 min until no lump was present in the liquid to ensure uniformity in consistency. The thickened liquids were inspected and checked by another experimenter to make sure no lump was present. Then, the samples were set to rest for 1 min before further testing as per manufacturer’s instructions, which simulated the timeframe of ad hoc hydration needs. The entire experiment was carried out under room temperature (∼25°C).

Following the recommendations by IDDSI [5], a 10-mL Luer tip syringe (Model 302143, BD™, Belgium) was used for the IDDSI Flow Test measurements. The distance from the 0 mL line to the 10 mL line was measured at 61.5 mm. To begin the measurement, 10 mL of thickened liquid was placed into an empty syringe while the nozzle was blocked with a finger to prevent leakage. Upon initiating the measurement, the nozzle was released, allowing the liquid to flow for a duration of 10 s, as timed by a digital stopwatch (A168WA-1, Casio, Japan). After precisely 10 s, the syringe nozzle was blocked once again. The remaining liquid level inside the syringe was observed at eye level, and the reading was recorded to the nearest marking (±0.2 mL) at the bottom of the meniscus to minimize error. The instrument was thoroughly washed and dried after each measurement.

A standard Bostwick consistometer (CSC Scientific Company, Inc., Fairfax, VA, USA) was used. Made of stainless steel, the tool consists of two compartments separated by a spring-loaded trap door. One compartment is the liquid reservoir with a capacity of 75 mL, and the other compartment is a long slanted trough with 24 cm in length and 5 cm in width, with markings at every 0.5 cm interval [31]. For each measurement, the thickened sample was poured into the reservoir compartment. To ensure an exact volume of 75 mL, the reservoir was overfilled slightly, then leveled with a spatula. The spring-loaded trap door was then released to allow the fluid to flow along the slanted trough for 30 s, as timed by using a digital stopwatch (A168WA-1, Casio, Japan). After 30 s, a measurement was obtained by averaging the farthest point and the shortest point that the liquid had reached. After each measurement, the instrument was washed and dried thoroughly.

To establish the relationship between IDDSI Flow Test and consistometric measures, correlation and regression analyses were carried out, and Pearson product-moment correlation coefficients were calculated.

Average IDDSI Flow Test and consistometric measures of thickened liquids prepared using starch-based and xanthan gum-based thickening agents are shown in Tables 3 and 4, respectively. Scatterplots were used to depict the relationship between IDDSI Flow Test measures and consistometric measures related to the use of starch-based and xanthan gum-based thickeners, and they are shown in Figures 1 and 2, respectively.

Correlation analyses revealed that IDDSI Flow Test measures and consistometric measures associated with starch-based and xanthan gum-based thickeners were significantly correlated (r = −0.9812 and r = −0.9653, respectively) (ps < 0.001). Subsequent regression analyses revealed that, for liquids thickened using starch-based thickener, a linear relationship was found between IDDSI Flow Test and consistometric measures (R² = 0.9627). However, for liquids thickened with xanthan gum-based thickener, a quadratic relationship was obtained (R² = 0.9834) (see Fig. 1, 2). The high R-squared values associated with both thickeners indicated that, with the regression lines modeling the relationship between IDDSI Flow Test and consistometric measures, 96.27% and 98.34% of the variances were explained for starch-based and xanthan gum-based thickeners, respectively. This also indicated that the two regression lines well fitted the data obtained. The linear and quadratic regression equations obtained for starch-based and xanthan gum-based thickeners are as follows:

The present study examined the relationship between the IDDSI Flow Test, and the consistometric measures associated with thickened liquids prepared using starch-based and xanthan gum-based thickening agents. Findings allow precise conversion of liquid consistency between both measurements and mapping consistometric measures with IDDSI Flow Test measurements. This facilitates more precise measurement of consistency of thickened liquids, which warrants safer swallowing by minimizing risks of penetration and aspiration in dysphagia management. ThickenUp^® and ThickenUp^® Clear were selected for the present study as they represented the most popular thickening agents used in Hong Kong. It was believed that both thickening agents offer stability and reliability when used to alter liquid consistency. In the study, IDDSI Flow Test and consistometric measures associated with starch-based and xanthan gum-based thickeners were measured, and regression was carried out in an attempt to obtain the equations that best fit the data points.

The present data revealed a linear relationship between IDDSI Flow Test and consistometric measures for thickened liquids prepared using starch-based thickener. This finding was somehow contradictory to that reported by Côté et al. [25], despite the stronger correlation found in the present study. The use of only water in the present study might explain the higher correlation found, as compared to the many types of beverages used by Côté et al. [25]. In addition, the more homogeneous samples might have contributed to greater linearity and lower variance, which in turn allowed more accurate conversion and prediction between IDDSI and consistometric measures. A linear regression equation of C = −2.33637I + 31.146 was obtained. The negative slope in the equation implies that a higher IDDSI Flow Test value (the amount of liquid remained in the syringe after 10 s) is correlated with a lower consistometric value (the less amount of liquid flowing down the slanted trough of the Bostwick consistometer). The linear relationship between IDDSI Flow Test and consistometric measures for starch-based thickened liquid implies a proportionate increase in consistometric measure should be correlated with a reduction in IDDSI reading.

Data from the experiment on xanthan gum-based thickener revealed a quadratic relationship between IDDSI Flow Test and consistometric measures. Such findings did not seem to align with those reported by Côté et al. [25]. However, it should be noted that only linear regression was used by Côté et al. [25], which could be inappropriate to reveal the relationship between two variables that are nonlinearly related and could lead to an incorrect conclusion. The nonlinear relationship between IDDSI and consistometric measures is apparent (see Fig. 2), and polynomial regression should be used. In the present study, the quadratic regression equation was $C=−0.183278I2+0.402632I+24.1862$⁠. According to the present finding, the quadratic relationship between IDDSI and consistometric measures implies that, for xanthan gum-based thickened liquids, the same consistometric measure may be correlated with two IDDSI readings. Future studies of a more extended range are needed to reveal a clearer picture of how IDDSI and consistometric measures are related.

Mapping between measures from IDDSI Flow Test and Bostwick consistometer was largely limited by the floor and ceiling effects of both measures. Consistometric measurements of consistency of thinner liquids prepared by either thickening agent were found to be limited by the Bostwick consistometer which had a trough of length of 24 cm. Liquids of consistency level 0 (thin) consistently exhibited a travel distance of greater than 24 cm in the 30 s of flow time, and thus was associated with a reading of >24 cm (exceeded the maximum reading of Bostwick consistometer) [16]. This can be considered as a major limitation of the present study. A longer version of the Bostwick consistometer is needed for future study that could extend and include thinner consistency. At the other end of the spectrum, the current IDDSI Flow Test failed to estimate consistency of thicker liquids. Liquids of consistency level of 4 (extremely thick) were not appropriate for IDDSI Flow Test [16, 25]. Even given 10 s to flow, the liquid simply stayed in the syringe and no liquid dripped out of it, resulting in failure in mapping of IDDSI consistency level 4. Because of such limitations, only consistency levels of 1, 2, and 3 were examined in the study (see Tables 3, 4). However, using the regression equations, these consistency levels could be interpolated.

Based on the regression equations, consistometric measures can easily be estimated for different consistency levels. However, due to the aforementioned limitations, only IDDSI consistency levels 1, 2, and 3 could be validly estimated using these equations. According to IDDSI, consistency levels 1, 2, and 3 correspond to 1–4 mL, 4–8 mL, and >8 mL of liquid left in the syringe after dripping for 10 s (see Table 1). Based on this, consistometric measures corresponding to different consistency levels are estimated and shown in Table 5.

The present study only examined the flow behavior of starch-based and xanthan gum-based thickened liquids using a consistometer. It is understood that both thickeners exhibit very different flow characteristics and the use of consistometer may not be able to reveal fully the flow characteristics of liquids [40]. Yet, the value of the study lies in the fact that IDDSI is commonly used by speech therapists for managing dysphagic patients, and consistometry is simple to use and to understand.

Apart from the limitations stated above, several other limitations in the present study can be identified. First, thickened liquids were prepared using only water as the dispersion medium, and only ThickenUp^® and ThickenUp^® Clear, both of which were manufactured by Nestlé Health Science, were used. This might have limited the generalizability of thickened liquids consumed by dysphagic patients. Another limitation of the study relates to the use of only linear and quadratic regressions to describe the relationship between IDDSI and consistometric readings for starch-based and xanthan gum-based thickeners, and the regressions were based on limited amount of data. Other higher order regression may be tried in relating such relationships using a large scale of data.

A third limitation of the study is the apparent floor and ceiling effects of both IDDSI and consistometric measurements, which can be found in Figures 1 and 2. Data should therefore be interpreted with caution. In addition, other aspects of liquid food including acidity, serving temperature, and texture were not considered. Future studies involving more base liquids and other thickening agents should be carried out in order to obtain a wider picture of the association between IDDSI Flow Test measures and consistometric measures.

The present study revealed a very strong correlation between liquid consistency measures obtained using IDDSI Flow Test and Bostwick consistometer. In addition, according to regression analyses, starch-based and xanthan gum-based thickeners were found to exhibit a linear and quadratic relationship between IDDSI Flow Test and consistometric measures, and the regression equations were $C=−2.33637I+1.146and$$C=−0.183278I2+0.402632I+24.1862$⁠, respectively. Seeing the limited range of consistency seen in the study, future studies are needed to better map consistometric measure to IDDSI levels, especially for liquids of IDDSI levels 3 and 4.

No ethics were required for the study as it did not involve human or animal participants.

The authors have no conflicts of interest to declare.

The project was funded by the Education Faculty Research Fund, Faculty of Education, University of Hong Kong (Reference No. 000250337).

Mingyue Xiong: substantial contributions to the conception and design of the study. Nelson Ng and Brian Siu: data collection, preliminary analysis, and preliminary preparation of the manuscript. Manwa L. Ng: substantial contributions to the conception and design of the study, administration of the project, data and statistical analyses, and manuscript preparation.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.