The Influence of Liquid Consistency and Food Texture on Swallowing: A Systematic Review Update Authors: Catriona M Steele, Isuru Dharmarathna, Jacqueline K Benfield, Simon Sollereder, Elaine Lipkin, Carlo Pedrolli, Ahmed Nagy, Kiara Goldenberg, Sana Smaoui, Dominie Yu, Alice Guo, Hiba Najeeb, Anittha Mappanasingam, Paige DiStefano, Pooja Gandhi, Julie A Y Cichero, Michelle M Simmons, Vanessa Panes, Irina Sanchez de Araujo, Camila Lirani-Silva, Renata Mancopes, Hilary Milne, Megan E Crawford, Fabíola Frizzo Guimarães, Michelle Catao Published: 2026 ABSTRACT AbstractDiet texture modification is a common intervention for oropharyngeal dysphagia. In 2015, the International Dysphagia Diet Standardisation Initiative (IDDSI) published a systematic review summarizing evidence on how liquid consistency and food texture influence swallowing behavior. This updated review synthesizes new evidence published between 2012 and 2024 addressing how differences in liquid consistency or food texture affect swallowing in humans. A PROSPERO-registered, PRISMA-guided review was conducted using the Covidence platform. A multi-engine search identified English, peer-reviewed studies reporting comparable quantitative measures across two or more consistencies defined using IDDSI terminology. Screening and full-text review were performed in duplicate. Articles were appraised using the FRONTIERS framework, and data synthesis focused on swallowing safety, efficiency, and other physiologic metrics. From 63,247 initial records, 24,196 abstracts were screened and 2,310 full texts reviewed, yielding 465 eligible studies. Of these, 70 studies (from 2017 onward) provided non-overlapping data mapped to the IDDSI Framework. Sixty-eight articles reported findings for drink levels 0–4 and 25 for food levels 5–7. Videofluoroscopy was the most common instrumental method used, followed by endoscopy and high-resolution manometry; other physiologic measures were used less frequently. Across all 25 studies reporting safety outcomes, thicker liquids improved swallowing safety compared with thinner liquids. Among studies reporting efficiency, smaller bolus volumes helped limit residue with thickened liquids. Food textures seldom elicited penetration-aspiration but were associated with increased residue in patient populations. Overall, findings reinforce the need for individualized clinical decision-making regarding texture modification, informed by careful assessment.Keywords: Deglutition, Deglutition disorders, Dysphagia, Diet texture modification, Systematic review Introduction Texture modification is one of the most common forms of intervention for oropharyngeal dysphagia (swallowing impairment) [1–3]. In 1987, the first year of the Dysphagia journal, an expert review article by Coster and Schwarz [4] introduced the concept of a “swallow-safe bolus”, namely a bolus with physical characteristics that facilitate its “unimpeded passage through the food pathway without aspiration, choking, or retention in the pharynx”. This concept provides the foundation for widespread clinical use of texture modification with the goal of promoting safe and efficient swallowing. Prior to 2017, there was no single convention with respect to the terminology used to describe levels of liquid thickness or food texture modification for clinical use [5]. The multidisciplinary International Dysphagia Diet Standardisation Initiative task force (IDDSI, https://www.iddsi.org) was founded in 2012 with the goal of promoting patient safety through a common diet texture terminology for people of all ages in all care settings and all cultures around the world. One of the first steps toward this goal was a systematic review, undertaken to identify empirical evidence describing the impact of liquid consistency and food texture on swallowing behavior in humans [6, 7]. An initial multi-engine literature search yielded 10,146 non-duplicate articles published between 1985 and January 2013. A team of ten international researchers collaborated to conduct full-text reviews for 488 of those articles, which met the review inclusion criteria. Of these, 36 articles were found to contain specific information comparing oral processing or swallowing behaviors across at least two liquid consistencies or food textures. This included 20 articles comparing swallowing on thin liquids to thicker liquid consistencies, 18 articles reporting data for foods (among which it was noted that there were “effectively no stimuli that were the same in any two or more studies”), and 3 articles describing swallowing with mixed consistencies. The 2015 review revealed two key trends with respect to the impact of thickening liquids on swallowing: (1) thicker liquids reduce the risk of penetration–aspiration; but (2) they also appear to increase the risk of post-swallow pharyngeal residue. Insufficient evidence was found to identify specific quantitative bolus properties (such as viscosity) associated with clinically meaningful reductions in penetration-aspiration. Furthermore, very little information was found regarding the effect of altering food texture on chewing, oral processing and/or swallowing, although the literature pointed to properties of hardness, cohesiveness, and slipperiness as being relevant both for physiological behaviors and bolus flow patterns. The purpose of the current manuscript is to report on an updated systematic review, considering new evidence published between 2012 and 2024 focusing on the primary question “How do differences in liquid consistency or food texture influence measures of chewing, oral processing or swallowing in humans?” This article specifically addresses three subquestions: How do differences in liquid consistency or food texture influence measures of swallowing safety (i.e., penetration and/or aspiration of material into the airway)? How do differences in liquid consistency or food texture influence measures of swallowing safety (i.e., penetration and/or aspiration of material into the airway)? How do differences in liquid consistency or food texture influence measures of swallowing efficiency (i.e., residue)? and. How do differences in liquid consistency or food texture influence measures of swallowing efficiency (i.e., residue)? and. How do differences in liquid consistency or food texture influence other measures of oropharyngeal swallowing physiology? How do differences in liquid consistency or food texture influence other measures of oropharyngeal swallowing physiology? Measures of chewing and oral processing have been excluded from these subquestions and will be addressed in a separate manuscript. We were specifically interested to identify evidence regarding the influence of bolus characteristics defined by, or unambiguously mappable to the IDDSI Framework [8, 9] (https://www.iddsi.org), which includes terminology, definitions and testing methods for the following levels: level 0 thin liquids, level 1 slightly thick liquids, level 2 mildly thick liquids, level 3 moderately thick liquids or liquidised foods, level 4 extremely thick liquids or pureed foods, level 5 minced and moist foods, level 6 soft and bite-sized foods, and level 7 regular texture foods (which includes a subcategory of level 7 easy to chew foods). The IDDSI Framework also proposes definitions and testing methods for mixed consistency foods and transitional foods (which change texture rapidly when exposed to moisture or body temperature). Methods This review was registered with PROSPERO (# CRD42024498552) and performed according to PRISMA guidelines [10, 11] using the Covidence review platform [12]. The search strategy replicated the methods used in the previous review, with the addition of new texture terms to reflect IDDSI terminology. We used multiple search engines to search for relevant articles, including Scopus, Ovid MED-LINE (R), AMED (Allied and Complementary Medicine), EMBASE, APA, and PsycINFO. The full search strategy can be found in Supplementary Material File 1. In addition, we performed a citation search to include all articles citing previously published systematic reviews related to this topic [6, 7, 13–17]. The reference lists of any additional systematic reviews identified during the search were also hand-searched for additional articles or potential relevance. The initial search included papers published from January 1, 2012 to January 24, 2024, with a subsequent update, extending the end date to May 11, 2024, prior to beginning quality appraisal and data extraction. At each stage of the project, articles were reviewed independently by two raters from an international and multidisciplinary team of 30 collaborating dysphagia researchers, clinicians and research trainees. Discrepancies were resolved based on additional review by the 1st author, who oversaw the compilation of review data and managed the project. During the title and abstract screening stage, and the subsequent full-text review stage, articles were excluded if they: were not published in English; were not published in English; were animal studies with no human data; were animal studies with no human data; did not report quantitative measures of chewing, oral processing or swallowing across more than one consistency; did not report quantitative measures of chewing, oral processing or swallowing across more than one consistency; were case reports or had a sample size of fewer than 10 participants; were case reports or had a sample size of fewer than 10 participants; were guidelines, consensus statements or Delphi studies; were guidelines, consensus statements or Delphi studies; were scoping reviews, narrative reviews, opinion papers, letters to the editor, dissertations, protocols, or conference abstracts; were scoping reviews, narrative reviews, opinion papers, letters to the editor, dissertations, protocols, or conference abstracts; focused exclusively on the properties, psychometrics, or validity/reliability of scale(s), measure(s), or method(s); focused exclusively on the properties, psychometrics, or validity/reliability of scale(s), measure(s), or method(s); reported data limited to patient-reported/quality of life measures; esophageal measures; sensory or stimulus palatability/acceptability measures; bench-top measures of stimulus characteristics; or downstream health outcomes (e.g., pneumonia or malnutrition). reported data limited to patient-reported/quality of life measures; esophageal measures; sensory or stimulus palatability/acceptability measures; bench-top measures of stimulus characteristics; or downstream health outcomes (e.g., pneumonia or malnutrition). Following full-text review, we evaluated quality, methodological rigor and reporting transparency for each article using the FRONTIERS Framework critical appraisal tool (Framework for Rigor and Transparency in Research on Swallowing, https://www.frontiersframework.com; [18]). Reviewers documented whether or not authors had clearly reported elements of study methodology and conduct across 5 domains, including: universally applicable questions (i.e., study locations, data collection protocols, bolus descriptions, participant positioning and instructions, blinding, rater training, handling of missing data, statistical considerations); study participant characteristics; videofluoroscopic swallowing studies (VFSS); flexible endoscopic examinations of swallowing (FEES); and other instrumental methods of evaluating chewing, oral processing and swallowing. The full set of questions considered during review can be found in supplementary material file 2. The screening and non-instrumental/clinical assessments of swallowing, treatment techniques and patient/caregiver reported outcomes domains of the FRONTIERS Framework were not used in this review. FRONTIERS ratings were used as the basis for the final stage of review, where available details regarding study methods and findings were extracted for synthesis. Although the intent had been to perform meta-analysis where possible, wide variations in study populations and methods (particularly in the choice and definitions of the parameters measured) severely limited the extent to which data could be directly compared across studies. Companion scoping review articles will be published separately, providing a broad characterization of the body of literature included for full-text review, and appraising the quality of reporting for studies using specific methodologies (e.g., VFSS; FEES). For the subset of articles reporting data for swallowing safety, efficiency or other swallowing metrics by IDDSI level, we extracted and compared findings for the following bolus consistency groupings: Data for slightly thick and mildly thick liquids compared to thin liquids (i.e., IDDSI levels 1 and 2 compared to level 0); Data for slightly thick and mildly thick liquids compared to thin liquids (i.e., IDDSI levels 1 and 2 compared to level 0); Data for moderately thick liquids/liquidized foods and extremely thick liquids/pureed foods (i.e., IDDSI levels 3 and 4) compared to IDDSI levels 0, 1 and 2; and. Data for moderately thick liquids/liquidized foods and extremely thick liquids/pureed foods (i.e., IDDSI levels 3 and 4) compared to IDDSI levels 0, 1 and 2; and. Data for minced and moist, soft and bite-sized, easy to chew and regular foods (i.e., IDDSI levels 5, 6 and 7), including mixed consistencies and transitional foods, compared to liquids at IDDSI levels 0 through 4. Data for minced and moist, soft and bite-sized, easy to chew and regular foods (i.e., IDDSI levels 5, 6 and 7), including mixed consistencies and transitional foods, compared to liquids at IDDSI levels 0 through 4. Paired comparisons across specific food levels within the range from IDDSI level 4 to 7, including mixed consistencies and transitional foods. Paired comparisons across specific food levels within the range from IDDSI level 4 to 7, including mixed consistencies and transitional foods. The data synthesis considered information regarding healthy participants separately from information regarding people with dysphagia or specific diagnoses. To the extent possible (based on methodological reporting), the comparison of specific quantitative measures of swallowing was limited to articles in which similar operational definitions had been used for those measures. Similarly, wherever possible, extracted comparisons were limited to boluses of matched volume; where this was not possible, volume differences across consistencies were noted. Where data were reported across multiple timepoints (e.g., pre- and post-treatment), only data for the first timepoint were extracted. In cases where multiple articles from the same authors appeared to report overlapping data for the same participants, we retained the article reporting the largest sample or most complete data. Similarly, if multiple articles included overlapping data from the same control group, we retained the article reporting the largest sample of healthy controls, and limited inclusion for overlapping articles to the unique patient portions or unique parameters within the reported data. Substantial variation was found across articles with regard to the specific parameters chosen for characterizing swallowing safety or efficiency, the score thresholds considered to represent clinical concern, and the number and definitions of ordinal or categorical scores reported for these parameters. Therefore, for the purposes of data synthesis and comparison, it was necessary to classify results as reporting either the frequency of a particular type of finding (e.g., the number of participants displaying unsafe swallowing on a given consistency, using the author’s definition of “unsafe”), and/or the severity of a finding (e.g., group mean/median scores with corresponding measures of dispersion for the amount of residue on a given consistency). Furthermore, due to variations in the IDDSI levels and specific stimuli used across studies, comparisons were summarized as reporting differences between higher-numbered (i.e., thicker) versus lower-numbered (i.e., thinner) IDDSI levels within the four comparison groups outlined above (A-D). For other parameters, we extracted descriptive statistics for comparable measures reported in at least two different articles and used box plots to compare reported measures of central tendency and dispersion in cases where the same operational definitions and units had been used across studies. Results Figure 1 documents the number of articles identified for review and inclusion at each stage of the project. The lengthy list of search terms used in this review led to a vast number of articles being identified for screening and full-text review. After application of the exclusion criteria, there were 465 articles reporting quantitative data for chewing, oral processing or swallowing in humans across more than one food or liquid consistency. Of these, 70 articles [19–88] reported non-overlapping data for measures of oropharyngeal swallowing function or physiology, collected using stimuli that were described according to IDDSI terminology or unambiguously mappable to the IDDSI Framework. Six additional articles were excluded due to overlap [89–94]. Given the large number of qualifying articles in this subset, we decided not to synthesize results from additional articles in the full-text review set, for which it would be necessary to infer the mapping of stimuli to IDDSI levels. This decision had the consequence of limiting the data synthesis set to articles published beginning in 2017, aligned with release of the IDDSI Framework [8]. For articles reporting use of Varibar™, a commercially available barium product, stimuli were mapped to liquid levels on the IDDSI Framework based on information from the IDDSI website [95]: Varibar Thin = level 0 thin, Varibar Nectar = level 2 mildly thick, Varibar Thin Honey = level 3 moderately thick, Varibar Honey = level 4 extremely thick and Varibar Pudding = level 4 extremely thick. A summary table describing key characteristics of the 70 included articles can be found in supplementary material file 3. PRISMA Diagram showing the results of the search strategy and review The data synthesis set included 42 studies using VFSS, 15 using FEES, eight using HRPM, three using EMG and one study each using lingual pressure, nasal airflow and ultrasound. Although 10 studies reporting either acoustic or accelerometry data for swallowing across consistencies were included in the full-text review set [96–105], none of these described their stimuli using IDDSI terminology or testing methods. The use of IDDSI terminology was also rare amongst articles reporting EMG data. Table 1 (liquids) and Table 2 (foods) shows the extent to which data for different IDDSI levels were reported by study instrument type. Here it can be appreciated that the strong majority of studies included level 0 thin liquids in their protocols (regardless of instrument type). Inclusion of level 1 slightly thick, level 2 mildly thick and level 3 moderately thick liquid stimuli was more common in studies using VFSS, while a higher proportion of FEES studies included level 4 extremely thick liquids/pureed foods and level 7 regular food stimuli. Table 1The number of studies in the data synthesis set including different IDDSI drink levels by type of instrumentation Instrument TN0 ST1 MT2 MO3/LQ3 EX4/PU4 n (studies)% within instrumentn (studies)% within instrumentn (studies)% within instrumentn (studies)% within instrumentn (studies)% within instrumentVFSS (n = 42)3481%819%2252%2355%2048%% within consistency59%73%67%77%51%FEES (n = 15)1493%00%533%320%960%% within consistency24%0%15%10%23%HRPM (n = 8)563%00%225%00%563%% within consistency9%0%6%0%13%EMG (n = 3)267%133%133%133%267%% within consistency3%9%3%3%5%Ultrasound (n = 2)150%00%150%150%2100%% within consistency2%0%3%3%5%Lingual Pressure (n = 1)1100%1100%1100%1100%00%% within consistency2%9%3%3%0%Nasal Airflow (n = 1)1100%1100%1100%1100%1100%% within consistency2%9%3%3%3%Total (n = 70)5883%1116%3347%3043%3956%TN0 = Level 0 thin liquids; ST1 = Level 1 slightly thick liquids; MT2 = Level 2 mildly thick liquids; MO3/LQ3 = Level 3 moderately thick liquids/liquidised foods; EX4/PU4 = Level 4 extremely thick liquids/pureed foods. VFSS = Videofluoroscopic Swallow Study; FEES = Flexible Endoscopic Examination of Swallowing; HRPM = High Resolution Pharyngeal Manometry; EMG = Electromyography The number of studies in the data synthesis set including different IDDSI drink levels by type of instrumentation TN0 = Level 0 thin liquids; ST1 = Level 1 slightly thick liquids; MT2 = Level 2 mildly thick liquids; MO3/LQ3 = Level 3 moderately thick liquids/liquidised foods; EX4/PU4 = Level 4 extremely thick liquids/pureed foods. VFSS = Videofluoroscopic Swallow Study; FEES = Flexible Endoscopic Examination of Swallowing; HRPM = High Resolution Pharyngeal Manometry; EMG = Electromyography Table 2The number of studies in the data synthesis set including different IDDSI food levels by type of instrumentationInstrumentMM5SB6RG7Mixed Consistenciesn (studies)% within instrumentn (studies)% within instrumentn (studies)% within instrumentn (studies)% within instrumentVFSS (n = 42)12%410%717%00%% within consistency25%100%41%0%FEES (n = 15)17%00%853%17%% within consistency25%0%47%100%HRPM (n = 8)00%00%113%00%% within consistency0%0%6%0%EMG (n = 3)267%00%133%00%% within consistency50%0%6%0%Ultrasound (n = 2)00%00%00%00%% within consistency0%0%0%0%Lingual Pressure (n = 1)00%00%00%00%% within consistency0%0%0%0%Nasal Airflow (n = 1)00%00%00%00%% within consistency0%0%0%0%Total (n = 70)46%46%1724%11%MM5 = Level 5 minced and moist foods; SB6 = Level 6 soft-and-bite-sized foods; RG7 = Level 7 regular foods. VFSS = Videofluoroscopic Swallow Study; FEES = Flexible Endoscopic Examination of Swallowing; HRPM = High Resolution Pharyngeal Manometry; EMG = Electromyography The number of studies in the data synthesis set including different IDDSI food levels by type of instrumentation MM5 = Level 5 minced and moist foods; SB6 = Level 6 soft-and-bite-sized foods; RG7 = Level 7 regular foods. VFSS = Videofluoroscopic Swallow Study; FEES = Flexible Endoscopic Examination of Swallowing; HRPM = High Resolution Pharyngeal Manometry; EMG = Electromyography Swallowing Safety Among the 70 articles included for data synthesis, 25 reported comparative data for measures of swallowing safety across more than one IDDSI level, collected across a total of 2,596 participants. This set of 25 studies excludes eight additional articles where a floor effect was seen, with Penetration-Aspiration-Scale (PAS) [106] scores of 1 or 2 reported for all consistencies tested, suggesting safe swallowing without any meaningful opportunity to display a difference across consistencies [43, 53, 56, 60, 63, 80, 81, 86]. Three articles reported measures of swallowing safety in healthy adults [26, 39, 49]. On cup sip tasks, Garand and colleagues [39] reported that 10% of participants showed penetration on thin liquid stimuli; this dropped to 6% with mildly thick liquids, while aspiration was reported to occur in only a single participant on both of these stimuli. For 5 ml tasks with liquids, penetration was seen in 7% (thin), 6% (mildly thick), 2% (moderately thick) and 1% (extremely thick) of participants, and also in 1% on a level 7 regular food bolus (a bite of cookie). A single healthy participant was reported to aspirate on the 5 ml thin and mildly thick liquid tasks. The meta-analysis by Borders and Steele [26], included unique swallowing safety data for a sample of 78 healthy adults (source data [78]). When disaggregated, the healthy source data showed a 9% participant-level frequency of penetration on sips of thin liquid; this broke down to frequencies of 1%, 1% and 7% for PAS scores of 3, 4 and 5, respectively, with no aspiration (i.e., PAS scores of 6, 7 or 8). The frequency of penetration dropped to 1% and 4% on sips of slightly thick and mildly thick liquids, respectively, with no penetration events on 5 ml teaspoons of moderately or extremely thick liquids [26, 78]. The 30 healthy control participants in the study by Lee and colleagues [49] were reported to have mean ± standard deviation PAS scores of 1 ± 0 (i.e., no airway invasion) on both thin and moderately thick liquids. Thirteen studies of the 25 studies in the safety set reported PAS scores for patients with dysphagia on slightly or mildly thick liquids compared to thin liquids [26, 28, 32, 33, 54, 57, 58, 65, 68, 71, 82, 85, 88]. Data for IDDSI level 1 slightly thick liquids were included in two studies with a pooled sample size of 842 patients [26, 65], while mildly thick liquids were used in all 13 studies, with a pooled sample size of 1,693 patients [26, 28, 32, 33, 54, 57, 58, 65, 68, 71, 82, 85, 88]. Twenty-one studies with a pooled sample size of 1,918 patient participants compared levels 3 and/or 4 (moderately and extremely thick liquids) to levels 0, 1 and 2 [22–24, 26, 33, 37, 45, 47, 49, 54, 57, 61, 62, 64, 67, 68, 71, 82, 85, 88]. Ten studies in 551 patient participants compared foods to drink levels [22, 28, 45, 47, 57, 61, 62, 68, 82, 85], and a single study in 25 patients compared specific food items [62]. 100% of the 25 studies reporting safety results for patients with dysphagia involving PAS scores > 2 reported improved swallowing safety for at least one comparison of higher-numbered (thicker) IDDSI levels to lower-numbered (thinner) levels, either in the form of reduced severity (as represented by group mean or median scores), or reduced frequencies of scores of concern. As noted previously, studies varied in their selection of outcome variables. For comparisons of slightly and/or mildly thick liquids to thin liquids, these improvements took the form of reduced severity of PAS scores in 3/14 studies (20%, n = 941) [26, 28, 57], and reduced frequency of PAS scores of concern in 13 studies (100%, n = 1,693) [26, 28, 32, 33, 54, 57, 58, 65, 68, 71, 82, 85, 88]. Among the 21 studies reporting comparisons of moderately and/or extremely thick liquids to IDDSI levels 0/1/2, six studies (29%, n = 985) reported reduced PAS severity [22, 24, 26, 49, 57, 82], and 15 studies (79%, n = 1,778) reported reduced frequency of PAS scores of concern [23, 26, 33, 37, 45, 47, 54, 61, 62, 64, 67, 68, 71, 85, 88]. Among the 10 articles exploring foods compared to drinks, five studies (50%, n = 170) reported lower severity of PAS scores with food stimuli [22, 28, 57, 61, 82], and five studies (50%, n = 381) reported lower frequencies of PAS scores of concern [45, 47, 62, 68, 85]. In addition to studying swallowing safety in patients with multiple system atrophy using IDDSI levels 0 and 4, Mozzanica and colleagues [62] reported data for level 7 (regular food) and two mixed consistency items described as combining levels 0 and 5 (minced and moist foods), and levels 2 (mildly thick liquids) and 6 (soft and bite-sized food). Lower frequencies of penetration and aspiration were seen for the mixed stimulus containing a level 2 mildly thick rather than a level 0 thin liquid component. Swallowing Efficiency Among the 70 articles included for data synthesis, 23 studies with an aggregated sample size of 1,401 participants reported comparative data for measures of residue across more than one IDDSI level. One additional study was excluded due to an apparent floor effect [83]. Three articles [30, 41, 78] focused exclusively on healthy adult participants (pooled n = 218). Curtis et al. [30] reported FEES data from 39 community dwelling adults without dysphagia, who swallowed water (level 0), pudding (level 4) and cracker (level 7). Oropharyngeal and hypopharyngeal residue were rare in this sample, with no differences in severity reported across consistencies, but the highest frequencies of oropharyngeal residue on thin liquids, followed by extremely thick liquids/pureed foods and then regular foods (cracker). Garand and colleagues [41] reported videofluoroscopy data for 363 boluses from 101/195 healthy adults in the MBSImP™ reference sample, for whom at least one bolus with an ordinal score > 1 (i.e., residue present) had been assigned on MBSImP component 16 (pharyngeal residue). Using anatomically scaled pixel-based measures of residue, in %(C2-4)2 units, Garand et al. [41] reported a significant effect of volume (across boluses of 5 ml, 10 ml and cup sips), with increased odds of residue on larger compared to smaller volumes of thin liquid. They found no difference in residue severity, but increased residue frequency for mildly thick liquids, moderately thick liquids, and extremely thick liquids compared to thin. None of the participants displayed residue on the level 7 cracker trials. Steele et al. [78] reported videofluoroscopy data for %(C2-4)2 measures of residue on thin, slightly thick, mildly thick, moderately thick and extremely thick liquids in a reference sample of 78 healthy adults. A pattern of increasing residue severity across sips of thin to slightly thick to mildly thick liquids was noted, with smaller amounts on teaspoon-sized boluses of the thicker level 3 and 4 stimuli (i.e., ~ 5 ml). Clinical decision limits for defining atypical total pharyngeal residue were proposed at the 75th percentile as follows: level 0 thin liquids (sip) > 1.7% (C2-4)2; level 1 slightly thick liquids (sip) > 1.9% (C2-4)2; level 2 mildly thick liquids (sip) > 2.2% (C2-4)2; level 3 moderately thick liquids (5 ml teaspoon) > 1.6% (C2-4)2; and level 4 extremely thick liquids (5 ml teaspoon) > 1.5% (C2-4)2. Among the 20 remaining articles in the efficiency set, 11 (55%, n = 593) reported residue data for people with dysphagia on slightly and/or mildly thick liquids compared to thin liquids [20, 28, 32, 38, 54, 55, 56, 71, 75, 82, 84], 15 (79%, n = 939) reported data for moderately and/or extremely thick liquids compared to IDDSI levels 0, 1, and/or 2 [31, 35, 37, 38, 47, 54, 56, 60, 61, 64, 71, 75, 82, 84, 86], seven (35%, n = 324) reported data for foods compared to liquids [28, 31, 35, 47, 60, 61, 82], and one (n = 25) explored differences between specific food items [62]. Of the 11 articles reporting comparable patient data for IDDSI levels 0, 1, and 2, four (36%, pooled n = 421) reported greater residue severity [28, 71, 75, 82] and six (55%, pooled n = 504) reported higher frequencies of residue with thicker consistencies [20, 28, 54, 56, 71, 75]. One article (n = 19) reported no differences in residue severity [84] and two (18%, pooled n = 66) reported no differences in residue frequency [32, 38]. One article (n = 152) [55] reported greater pyriform sinus residue severity on 5 ml thin liquid boluses compared to 10 ml mildly thick boluses in a sample of head and neck cancer patients. Among the 15 articles describing residue in patients on moderately and/or extremely thick liquids compared to thinner consistencies, four (27%, pooled n = 452) reported greater residue severity with the thicker stimuli [47, 61, 75, 82], while two (13%, pooled n = 37) reported no differences [64, 84]. Nine articles (60%, pooled n = 730) reported higher frequencies of residue with level 3 and/or 4 stimuli compared to thinner consistencies [31, 35, 37, 38, 54, 56, 71, 75, 86], while one (n = 52) [60] reported no difference. Within these comparisons, different patterns across residue locations were noted, and differences in bolus volume were highlighted as a possible factor contributing to residue severity [47, 84]. Of the seven articles reporting residue in patients on solid food stimuli, two (29%, pooled n = 72) identified greater residue severity on foods compared to liquids [61, 82]. Three articles (43%, pooled n = 152) reported higher frequencies of residue on foods [28, 31, 35], while one article (n = 52) reported no difference [62]. Kroon and colleagues [47] reported different patterns of residue for different anatomical locations and bolus volumes in 48 patients with oculopharyngeal muscular dystrophy. Mild pyriform sinus residue severity was reported for level 7 regular foods compared to both thin and moderately thick liquids. By contrast, vallecular residue severity for level 7 regular foods was greater than that seen with 10 ml thin liquid boluses, but lower than 20 ml thin and both 10 and 20 ml boluses of moderately thick liquid. Mozzanica and colleagues [62] reported worse residue severity for level 4 pureed and level 7 regular foods compared to mixed consistencies containing either a thin or mildly thick liquid component in a sample of 25 patients with neurogenic dysphagia; bolus volumes/sizes were not described in this study. Overall, worse swallowing efficiency on higher IDDSI levels, either in the form of greater residue severity or greater residue frequency, was reported in seven studies (pooled n = 556) for slightly or mildly thick liquids compared to thin [20, 28, 54, 56, 71, 75, 82], 12 studies (pooled n = 850) for comparisons of levels 3 or 4 compared to levels 0, 1 or 2 [31, 35, 37, 38, 47, 54, 56, 61, 71, 75, 82, 86], and seven studies (pooled n = 297) comparing food textures to liquid consistencies [28, 31, 35, 47, 61, 62, 82]. Other Parameters In total, 43 articles reported data for at least one additional parameter across IDDSI levels. Of these, single articles reported information regarding lingual pressures [76], the frequency of nasopharyngeal reflux [67], pulse oximetry [23], and respiratory-swallow coordination patterns [83]. Three articles reported electromyographic (EMG) data [46, 52, 70]; however, differences across these three studies in the number and locations of EMG sensors, signal segmentation and processing methods, and the EMG measures that were analyzed (i.e., mean root mean square versus peak amplitudes or durational metrics) meant that the data could not be compared. Details regarding consistency-related differences in the other parameters are described below. Number of Swallows per Bolus Consistency-specific data regarding the number of swallows per bolus was reported in three studies [30, 31, 78], two of which focused on data from healthy reference samples. In their study of videofluoroscopy data from 78 healthy adults for IDDSI levels 0, 1, 2, 3 and 4, Steele and colleagues reported that a single swallow per bolus was the predominant pattern across all consistencies, with no significant differences in the frequency of secondary, tertiary, or quaternary swallows, or the pooled category of multiple swallows per bolus across consistencies [78]. Similar results, with single swallows accounting for at least 60% of boluses were reported in Curtis et al.’s FEES study in 39 healthy adults for 5 ml, 10 ml, 20 ml and self-selected volumes of thin liquid taken by cup sip, and for 5 ml boluses of level 4 pureed foods taken by spoon as well as self-selected bites of cracker (level 7) [30]. Within the thin liquid trials, a clear association emerged between larger bolus volumes and an increased frequency of multiple swallows per bolus. Finally, in a study of 100 adults with amyotrophic lateral sclerosis (ALS), Donohue and colleagues reported a higher number of swallows per bolus on moderately thick liquid boluses compared to level 0 thin, level 4 extremely thick liquid and level 7 cracker boluses [31]. Swallow Initiation Measures Measures related to swallow initiation fell into two categories: measures describing bolus location at swallow onset (BLSO) and measures of swallow reaction time (SRT). A secondary analysis of videofluoroscopy data for 195 healthy adults by Bhutada and colleagues [21] reported no differences in BLSO between thin, mildly thick and extremely thick liquid stimuli (levels 0, 2 and 4) on cued non-sequential drinking tasks. By contrast, level 3 moderately thick liquid boluses were more likely to be located at or above the valleculae, while cracker boluses (level 7) were more likely to be located in the valleculae at swallow onset. When comparing non-cued sips of thin, slightly thick and mildly thick liquids (levels 0, 1 and 2) in a sample of 38 healthy adults under age 60, Steele and colleagues [77] reported similar frequencies of deeper BLSO (i.e., in the pyriform sinuses). By contrast, reduced frequency of deep BLSO was reported for 5 ml teaspoon boluses of level 3 and 4 (moderately thick and extremely thick liquid) stimuli. This suggests a possible contribution of bolus volume to this parameter. Similar patterns were seen in the endoscopic BLSO data for 39 healthy adults reported by Curtis et al. [30] on non-cued swallows of level 0 thin liquids in several different volume conditions, 5 ml spoons of level 4 pureed food boluses and bites of level 7 cracker. Higher BLSO, with markedly fewer boluses in the hypopharynx, was reported for level 4 extremely thick/pureed stimuli compared to thin liquid stimuli. The majority of cracker boluses were located in the oropharynx at swallow onset. No clear trend of deeper bolus position with larger volumes of thin liquid was seen in this endoscopic dataset. With regard to data from patient samples, Tuomi and colleagues [82] reported no differences in videofluoroscopic measures of BLSO between thin, mildly thick and extremely thick liquids and a level 7 cookie in a sample of 52 head and neck cancer patients. Another study [61] described BLSO as seen on endoscopy in a sample of 20 patients with neurogenic dysphagia; deeper BLSO was reported for thin liquid compared to level 4 pureed stimuli. In the same study, thin liquids and mixed consistencies containing a level 0 thin liquid component were more likely to be located in the laryngeal vestibule at swallow onset, suggesting pre-swallow penetration-aspiration, compared to mixed consistencies with mildly thick liquid component, or to level 4 pureed and level 7 regular food stimuli. The second approach to comparing swallow initiation across consistencies involved measures of SRT on videofluoroscopy, i.e., the interval between the frame showing the leading edge of the bolus passing the shadow of the ramus of mandible and the first frame of the hyoid burst movement marking initiation of the pharyngeal swallow [78]. Healthy reference data from a sample of 78 adults aged 21–82 [78] identified median (75th percentile) values of 133 (400) ms for sips of thin liquid for SRT, with stepwise increases in values across sips of slightly and mildly thick liquid and 5 ml teaspoons of moderately and extremely thick liquid, as shown at the top of Fig. 2. Two studies in patient populations (17 participants with Parkinson Disease (PD) [38] and 28 participants with COPD [56]) show similar, but not identical trends, as illustrated in the bottom two-thirds of Fig. 2. Differences across these studies with respect to bolus volumes and administration patterns, the number of task repetitions per consistency, and sample size may account for variations in results. Nevetheless, the overall pattern across these studies suggests longer SRT for mildly thick or thicker liquids compared to thin, even with smaller bolus volumes for thicker consistencies. A videofluoroscopic study by Myung and colleagues [63] in 130 subacute supratentorial stroke patients defined pharyngeal delay time (PDT) as the interval between the bolus reaching “the lower edge of the mandible” and “the appearance of the swallowing reflex”, and used a criterion of >400 ms to define prolongation of this parameter. Notably, both this definition and the threshold for defining prolonged values align with the reference value work reported by Steele et al. [78]. Myung and colleagues reported that the frequency of prolonged PDT increased from thin to moderately thick liquid stimuli, regardless of the presence/absence of oral apraxia, with further increases seen on level 5 minced and moist and level 6 soft and bite-sized food stimuli. Comparative data for measures of swallow reaction time by consistency across studies. The square data points represent median values, with tails showing the 25th and 75th percentiles. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon Timing Measures Data for consistency-related differences in various timing measures were reported in 16 articles. Oral preparation time, defined as the “time during which (the) complete bolus volume was maintained inside the mouth, before propulsion of (the) bolus through the posterior oral cavity”, was reported in two studies [59, 60], with a pattern of increasing duration from level 0 thin liquids to level 4 extremely thick/pureed to level 7 regular food stimuli seen in healthy adults as well as individuals with laryngitis and people with dysphagia after ischemic stroke. However, these same two studies (both from the same lab and published within a year of each other) reported conflicting findings for oral transit time (OTT), defined as the “time interval between the onset of the propulsive movement of the tongue tip at incisors to complete passage of the bolus tail over the tongue base”. Here, increasing durations from level 0 to level 4 to level 7 were seen in the healthy and laryngitis samples (each n = 21) in the Moda et al. study [59] but the opposite pattern was seen in the healthy (n = 19) and stroke samples (n = 33) in the Montaldi et al. study [60]. The reasons for this discrepancy are not clear; reliability data regarding the identification of boundary frames for OTT measures are not reported in either study and whether oral processing time for level 7 food stimuli was included or excluded from the OTT measures is unclear. Two additional studies described longer OTT for thicker liquid stimuli (either level 3 or 4) compared to thin liquids in 18 individuals with chronic obstructive pulmonary disease (COPD) [79] and 30 post-stroke patients with chronic dysphagia [86], while a study in 130 subacute stroke patients [63] reported higher frequencies of prolonged OTT (> 1 s) for moderately thick liquid and soft and bite-sized food stimuli compared to thin liquids, but mixed results for level 5 minced and moist food stimuli. Again, it is unclear whether food oral processing time was included or excluded from the OTT definition. Overall, the predominant trend across these studies suggests longer OTT for level 3 and 4 stimuli compared to thin liquids, but insufficient evidence to draw conclusions regarding food stimuli. Measures of pharyngeal transit time (PTT) were included in seven studies [59, 60, 63, 69, 74, 79, 86]. Here, as suggested in two prior systematic reviews [107, 108], differences in parameter definitions across studies made it challenging to compare results. The Moda et al. study [59] defined PTT as the “time interval between arrival of the bolus tail at the tongue base to complete passage of the bolus tail through the UES (upper esophageal sphincter)”, but together with the Montaldi et al. study [60], also reported data for a similar measure named pharyngeal clearance time (PCT), defined as the “time interval between arrival of the bolus head at the tongue base to complete passage of the bolus tail through the UES”. Smith-Hublou and colleagues [74] defined PTT as the “interval between bolus entry into the pharynx and UES closure” while Rockenbach et al. defined PTT as the “the time interval in seconds from the opening of the glossopalatal junction to the closure of the upper esophageal sphincter” [69]. Recognizing these variations in operational definitions together with variations in bolus volumes and study populations, it is not surprising to find mixed results reported across these studies with regard to the effect of bolus consistency on measures of pharyngeal transit. Three studies with a pooled sample including 51 stroke patients, 71 individuals with PD and 19 healthy controls [60, 69, 74] described a pattern of shorter PTT or PCT for level 4 extremely thick/pureed stimuli compared to level 0 thin liquids. Two studies (in 21 patients with laryngitis, 21 healthy controls and 18 adults with COPD) reported no apparent differences across these stimuli in PTT [59, 79] while the Wang et al. study in 30 stroke patients with chronic dysphagia [86] reported slightly longer PTT measures for level 3 moderately thick compared to level 0 thin liquid stimuli. Myung and colleagues [63] reported no differences in the frequency of prolonged PTT (> 1 s) in a sample of 130 stroke patients for level 3, 5 or 6 stimuli compared to thin liquids. Finally, white-out duration, an endoscopic timing measure that appears to bear some correspondence to videofluoroscopic measures of PTT or PCT (i.e., between 400 and 1200 ms), was reported to show no differences between IDDSI levels 0, 4 and 7 in a retrospective review of 75 FEES exams conducted across a heterogeneous group of 75 patients [25]. Another timing measure, which represents a subcomponent of measures of PTT or PCT, is the hyoid-burst-to-UES-opening (HYB-UESO) interval reported in studies using the ASPEKT Method of videofluoroscopy rating [78]. This parameter was compared between level 0 thin and level 4 extremely thick liquid stimuli in three studies, including reference data for 78 healthy adults [78] and data for 17 people with PD [38] and 28 with COPD [56]. As shown in Fig. 3, Steele et al. [78] reported similar HYB-UESO interval durations for thin and slightly thick liquid stimuli in healthy adults, with a subsequent pattern of stepwise increases in interval duration for progressively thicker liquids, despite reductions in bolus volume. In people with PD, longer durations were seen for 5 ml teaspoons of moderately and extremely thick liquid stimuli compared to thinner consistencies [38], while individuals with COPD showed longer durations for sips of mildly thick liquid and spoons of extremely thick liquid. Comparative data for measures of the hyoid-burst-to-UES-opening interval by consistency across studies. The square data points represent median values, with tails showing the 25th and 75th percentiles. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon Timing measures related to the onset and duration of laryngeal vestibule closure (LVC) were compared across consistencies in six studies [38, 56, 74, 78, 84, 87], with descriptive statistics sufficient to enable graphic comparisons in four, as shown in Figs. 4 and 5 [38, 56, 78, 87]. Here, it must be recognized that definitions of the beginning of LVC may vary slightly across studies. For example, Smith-Hublou and colleagues [74] used the term “laryngeal vestibule approximation” defined as “the first frame showing either complete contact between the base of the epiglottis and the arytenoid cartilages or the maximum level of approximation seen for the swallow”, while the other studies looked for the first frame of “most‑complete” LVC, defined as “the first frame where there is maximum approximation of the arytenoids to the laryngeal surface of the epiglottis” [78]. Notwithstanding these possible definitional variations, time-to-LVC appears relatively stable across bolus consistencies (Fig. 4). LVC duration also appears relatively invariant across thin, slightly thick and mildly thick liquids in healthy adults (Fig. 5). By comparison, shorter LVC durations are seen for level 3 moderately thick liquids [78], level 4 extremely thick liquids [78, 87] and level 7 regular food [87] stimuli, noting a possible contributing influence of smaller bolus volumes to this parameter [78]. In people with PD, mixed findings for time-to-LVC were reported, with shorter measures on 5 ml swallows of level 4 pureed stimuli compared to thin liquids in the Smith-Hublou study (n = 71) [74], but longer measures for 5 ml teaspoons of level 3 and 4 stimuli compared to sips of thinner consistencies in the Gandhi et al. study (n = 17) [38]. Differences in disease stage, severity and Parkinson Disease management approaches may well be present across these cross-sectional studies. In a sample of 28 people with COPD, Mancopes and colleagues [56] reported shorter time-to-LVC for level 3 and 4 stimuli (compared to levels 0 and 2), while Waito and colleagues [84] described the opposite pattern in a sample of 19 people with ALS. With regard to LVC duration in people with PD, Smith-Hublou et al. reported no differences between 5 ml volumes of thin and level 4 stimuli [74], while Gandhi and colleagues [38] showed a pattern of decreasing LVC duration from sips of thin to sips of slightly and mildly thick liquid, to 5 ml teaspoons of moderately and extremely thick liquid, again suggesting an influence of bolus volume. This pattern of shorter LVC duration for smaller bolus volumes of thicker liquid was also seen in the 19 ALS patients studied by Waito and colleagues [84] and in the Mancopes et al. study of 28 people with COPD [56]. Comparative data for measures of time-to-laryngeal-vestibule-closure by consistency across studies. Unless otherwise noted, the square data points represent median values, with tails showing the 25th and 75th percentiles. aThe square data points represent mean values, with tails showing one standard deviation. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon Comparative data for measures of laryngeal vestibule closure duration by consistency across studies. Unless otherwise noted, the square data points represent median values, with tails showing the 25th and 75th percentiles. aThe square data points represent mean values, with tails showing one standard deviation. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon Finally, in the category of timing measures, eight studies reported comparison data for the duration of UES opening [19, 38, 56, 59, 60, 74, 78, 84]. Here again, subtle differences in definitions were also noted across studies, in particular differences between studies using bolus events of entry and exit from the UES as the boundaries of their durational measure (e.g., [59, 60)] versus studies using boundary events of sphincter opening and closing (e.g [19, 78]). In three articles, descriptive statistics were not reported or were missing measures of dispersion [59, 74, 84]. Notwithstanding differences in study samples (i.e., 21 participants with laryngitis and 21 healthy controls; 71 participants with PD; and 19 participants with ALS, respectively), all three of these studies concurred in reporting shorter durations of UES opening for level 3, 4 and 7 stimuli (moderately thick liquids, extremely thick liquids and regular foods) compared to thinner consistencies. Figure 6 shows comparative data for the remaining studies. Inspection of the data highlights differences in bolus volume as a prominent factor that may influence UES opening duration, consistent with prior literature [109–113]. Steele et al. [78] reported stable UES opening duration in healthy participants for sips of level 0, 1 and 2 but shorter durations for smaller volumes of level 3 and 4. For 5 ml boluses, Ambrocio and colleagues [19] report data suggesting longer UES opening duration in healthy adults (n = 195) for mildly and moderately thick liquid boluses compared to thin liquids, and a further reduction for extremely thick liquids compared to thin, which is also reflected in the healthy data from Montaldi et al. (n = 19) [60]. Very similar trends are seen in the data for patient groups (i.e., 17 participants with PD [38]; 28 participants with COPD [56]; and 33 stroke patients [60], with a consistent finding of shorter durations for level 4 stimuli (compared to thinner consistencies), and a volume effect with larger volumes associated with longer UES opening duration. Comparative data for measures of UES opening duration by consistency across studies. Unless otherwise noted, the square data points represent median values, with tails showing the 25th and 75th percentiles. a The square data points represent mean values, with tails showing one standard deviation. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon. Kinematics In addition to timing measures, several studies reported measures of structural movement, including maximum distention of UES opening, pharyngeal area at maximum constriction, and measures of hyoid or laryngeal excursion. Maximum UES distention was reported in four of the studies for which UES opening duration measures were reported, namely two studies in healthy adults (pooled n = 273) [19, 78], a study in 17 adults with PD [38], and a study in 28 adults with COPD [56]. In all four cases, distention was measured as seen on lateral view videofluoroscopy at the narrowest point between C4 and C6 on the frame of maximum opening, and normalized to an anatomical scalar defined as the length of the C2-C4 cervical spine [114]. Figure 7 illustrates comparative data across these four studies. As with the previously described data for UES opening duration, volume effects appear to be influencing maximum UES diameter, with larger opening on sips [38, 56, 78] compared to 5 ml volumes of thin liquid [19] and smaller opening on 5 ml teaspoons [38, 78] of moderately and extremely thick liquids. For comparisons across stimuli within the same volume condition, the overlapping interquartile ranges in Fig. 7 suggest minimal variation in maximum UES diameter [19, 38, 56, 78]. Comparative data for measures of UES maximum distention by consistency across studies. The square data points represent median values, with tails showing the 25th and 75th percentiles. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon Anatomically-normalized measures of pharyngeal area at maximum constriction (PhAMPC) were also described in the set of studies reporting ASPEKT Method measures from videofluoroscopy [38, 56, 78, 84]. Among these, mean values for PhAMPC in 19 people with ALS showed an apparent trend towards larger PhAMPC values (i.e., a smaller degree of constriction) for slightly, mildly and moderately thick liquid stimuli compared to thin, with a further increase for extremely thick liquid stimuli, notwithstanding smaller bolus volumes for the moderately and extremely thick stimuli [84]. Median and interquartile range values were available for the remaining three studies, supporting graphical comparison as shown in Fig. 8. Here, it is notable that the data are positively skewed, with almost-complete constriction being the predominant characteristic in healthy swallowing, and a possible trend to larger PhAMPC values for sips of mildly thick liquid stimuli [78]. Poor constriction, characterized by larger PhAMPC values, is seen in both patient samples; in the 17 participants with PD sample [38], smaller volumes of moderately and extremely thick liquid stimuli are associated with better constriction, while this pattern is less obvious in the 28 participants with COPD, where the moderately thick and extremely thick liquids were taken by sip and a larger spoon (10 ml capacity), respectively [56]. Comparative data for measures of pharyngeal area at maximum constriction by consistency across studies. The square data points represent median values, with tails showing the 25th and 75th percentiles. Abbreviations: TN0 = thin liquids; ST1 = slightly thick liquids; MT2 = mildly thick liquids; MO3/LQ3 = moderately thick liquids/liquidised foods; EX4/PU4 = extremely thick liquids/pureed foods; tsp = teaspoon; tbsp = tablespoon The final set of kinematic measures reported across the studies included for data synthesis involves measures of hyoid or laryngeal excursion, described in nine studies [27, 40, 49, 50, 51, 64, 78, 79, 82]. Despite the obvious interest in the literature in hyolaryngeal measures, no two studies within this set reported measures that were directly comparable. Ordinal measures of anterior hyoid displacement in a heterogeneous sample of 119 patients undergoing videofluoroscopy [40] suggested reduced excursion for level 0 thin liquids compared to level 4 extremely thick liquid stimuli. This concurs with the pattern of greater displacement, as reflected by measures of hyoid angle, seen in a sample of 18 participants with COPD for thin compared to extremely thick liquid stimuli [79] and with one study in 52 neurologic patients [50], which reported measures of hyoid movement distance in millimeters between minimum and maximum positions in the hyoid movement trajectory. Tuomi and colleagues [82] reported data for 3 ml boluses of level 0 and level 4 stimuli that showed no differences in millimeter measures of anterior or superior hyoid movement in 52 patients with dysphagia secondary to head and neck cancer. In a study reporting pixel-based measures of laryngeal movement from videofluoroscopy, Nakao and colleagues [64] reported significantly greater superior movement distance, anterosuperior movement distance, superior mean velocity, anterosuperior mean speed, superior peak velocity and anterosuperior peak speed on moderately thick compared to thin liquid stimuli in a sample of 18 stroke patients with dysphagia. Continuous measures of hyoid movement speed or velocity in healthy adults collected using ultrasound (n = 97) [27] or videofluoroscopy (n = 78) [78] concurred in showing no differences across stimuli, while one study in 52 neurologic patients reported that instantaneous measures of peak hyoid velocity increased when comparing extremely thick to thin liquids [50]. A single study employing the Computational Analysis of Swallowing Mechanics approach to videofluoroscopy analysis [51] reported mixed patterns of biomechanical change (i.e., increased, decreased or unchanged movement) when comparing level 2 mildly thick and level 0 thin liquids in healthy participants. High Resolution Manometry Eight studies reported high resolution pharyngeal manometry (HRPM) data for different consistencies. Two studies were of swallowing in healthy adults (pooled n = 69) [36, 66]; two additional articles explored the effects of specific medications (codeine and baclofen) on swallowing in healthy adults (pooled n = 41) [42, 44]; two studies were in patient groups with comparison to healthy controls, one in individuals with PD (n = 20 patients, 20 controls) [35], and one in individuals with obstructive sleep apnea (OSA; n = 19 patients, 19 controls) [72]; and two studies were in patient groups following surgical interventions, one in 12 individuals with OSA [73], and the other in 26 patients with a jenual free flap used for circumferential pharyngeal reconstruction after resection for head and neck cancer [29]. An overview of the methods and results of these studies can be found in supplementary material file 4. Differences in equipment (e.g., catheter diameter and sensor configuration), as well as bolus consistencies, bolus volumes, and HRPM metrics make comparisons across studies challenging. Nevertheless, some trends can be appreciated. Four studies (in healthy participants and those with OSA) reported increases in UES integrated relaxation pressure from thin liquid to thicker consistencies [36, 66, 72, 73]. Four studies (again in healthy participants and those with OSA) reported increasing UES maximum admittance from thin liquid to thicker stimuli [36, 42, 66, 72]. Conversely, thicker consistencies were associated with lower pre-deglutitive UES basal pressure in three studies in healthy participants [36, 44, 66] and with shorter UES relaxation time in three studies in healthy participants and after surgery for OSA [36, 66, 73]. Equally interesting are the predominant findings of no differences across consistencies in healthy adults and OSA patients for contractile integral pressure measures (from the velopharynx down to the hypopharynx) [36, 42, 44, 66, 73] or for measures of hypopharyngeal intrabolus (distention) pressure [42, 44, 66, 72]. Other metrics or results were measured and/or reported in single studies. Overall, these studies point to HRPM measures of UES contractility, relaxation and opening showing greater sensitivity to bolus consistency than pharyngeal contractility measures or manometric measures of bolus flow. Discussion This systematic review update synthesizes information regarding the effects of bolus consistency on measures of swallowing, both in healthy and patient populations. An extraordinary volume of scientific literature has been published since 2012, in which consistency-specific data regarding quantitative measures of swallowing function and physiology have been reported. The original review [6, 7] asked (a) whether there was evidence to support or refute a functional or behavioral change resulting from the thickening of liquids and/or texture modification of foods; (b) if yes, how many and which levels of thickening or texture modification were supported by evidence, and what the quality of that evidence was; (c) whether the literature provided trustworthy objective measures (e.g., viscosity, density, yield stress, texture analysis, or other physical measures) to guide the definition of different levels of thickened liquids and texture modified foods; (d) whether the available evidence applied across the lifespan, or was it specific to particular subpopulations, defined either by age or diagnosis; and (e) what gaps in the literature were identified regarding thickening of liquids and/or texture modification of food as a strategy to manage dysphagia? This update focuses specifically on new evidence regarding question (a), namely how do differences in liquid consistency or food texture influence measures of swallowing safety, efficiency, and other measures of oropharyngeal swallowing physiology. Although our original intent had been to synthesize data published from 2012 to 2024, we ultimately made the decision to only synthesize data from articles published since 2017, in which stimuli were described using IDDSI terminology and definitions. Additionally, this manuscript excludes focus on measures of chewing and oral processing, which will be addressed separately. This manuscript excludes articles that have been published since the cut-off date of May 11, 2024 that was established before launching the quality appraisal and synthesis stages of this project. Even with these limits, data from 70 non-overlapping articles were synthesized, reflecting almost twice the number of articles that met the inclusion criteria for synthesis in the original review [6, 7], published over a 7–8 year time frame. Among these articles, 65 (93%) reported data for thin, slightly and/or mildly thick liquids, 60 (86%) reported data for moderately and/or extremely thick liquids, and 24 (34%) reported data for foods. Notably, information regarding swallowing with slightly thick liquids is new compared to the prior review. Data for healthy participants were reported in 32 articles, while data for individuals with medical diagnoses or dysphagia were reported in 51 articles. Swallowing Safety Measures of swallowing safety were the most common outcome reported in the data synthesis set, with comparisons across consistency in 25 articles. All (100%) of these articles showed a pattern of improved safety, either in the form of reduced severity or frequency of penetration-aspiration. These findings corroborate the trend identified in the original 2015 review of clinical benefit in the form of a decreasing frequency of penetration-aspiration as liquid viscosity increases [6, 7]. This update provides further detail regarding this pattern, including a single study reporting swallowing data with mixed consistencies (a stimulus combining a Level 0 thin liquid with level 5 minced and moist solids, and a stimulus combining level 2 mildly thick liquid with level 6 soft and bite-sized solids) [61]. The results echoed those of a study described in the previous review [6, 7, 115], with higher frequencies of penetration-aspiration on the mixed consistency containing the thin liquid component. Although the majority of studies exploring safety outcomes with thickened liquids used mildly thick (n = 16), moderately thick (n = 12) or extremely thick (n = 12) stimuli, two non-overlapping studies in the safety subset included swallowing safety data for level 1 slightly thick liquids and reported these to be effective for reducing the frequency or severity of penetration-aspiration, compared to level 0 thin liquids [26, 65]. Although delineation of slightly thick liquids as an initial degree of thickening was first introduced with publication of the IDDSI Framework in 2017 [8], this level was acknowledged to be in common clinical use in pediatric dysphagia management prior to that date. In this update, only 2/25 studies reporting safety results included pediatric participants. These observations point to an opportunity to increase the clinical consideration of slightly thick liquids as a potentially effective intervention for patients of all ages who aspirate thin liquids, particularly in centers and caseloads where endoscopy is the primary instrumental method of swallowing assessment. Swallowing Efficiency In the original review [6, 7], the available evidence pointed to the possibility that the safety benefits of increased liquid thickness might be offset by an accompanying increase in residue [115–118]. This is concerning, given evidence that residue increases the risk of secondary aspiration [89]. In this update, we again found evidence of this trend across the 23 articles in the efficiency subset. Either more frequent or larger amounts of residue were reported in three of the four articles reporting efficiency comparisons for slightly thick versus thin liquids: one in healthy participants [78], one in patients with PD [38], and one in patients following radiation treatment for oropharyngeal cancer [20]. This trend was also seen in 9/13 articles comparing residue for mildly thick and thin liquids: two in healthy participants [41, 78] and seven in patient samples [20, 28, 54, 56, 71, 75, 82]. However, this update reveals important new evidence that the risk of residue depends both on bolus volume and on bolus consistency, and specifically that less residue is seen with 5 ml teaspoon-sized boluses of moderately thick and extremely thick liquids than with sips of thinner consistencies [41, 47, 78, 84]. In many of the studies included for review, the protocols involved different volumes across consistencies, making it challenging to discern the degree to which efficiency results (either favorable or worse) reflected consistency effects, volume effects, or their interaction. From a clinical perspective, the data suggest that the safety and efficiency impact of thicker consistencies should be evaluated using both small (e.g., 5 ml teaspoon) and larger (i.e., sip-sized) bolus volumes, and that information regarding expected upper limits for residue by consistency and volume in healthy swallowing [78] should be taken into consideration, in order to tailor recommendations regarding texture modification for optimal patient benefit. Other Measures of Swallowing Physiology In this review, an increase in the amount of available information regarding the effects of bolus consistency on other measures of swallowing physiology is noted compared to the 2015 review, in which single studies were typically cited as the basis for possible trends. Data were only synthesized when more than one study with comparable metrics and stimuli were available, and studies were only included if they reported data for a minimum sample size of 10 participants. Against the backdrop of these operational constraints, the synthesis dataset found new evidence across multiple studies of longer swallow reaction times for mildly thick and thicker stimuli compared to thin liquids, both in healthy and patient populations. Bolus location at swallow onset was found to be variable within consistency, both in healthy and patient populations, with no clear differences across the range from thin to mildly thick liquids. With moderately and extremely thick liquids, BLSO tended to be higher up in the oropharynx, both in healthy and patient populations. Among the reported timing measures, a longer hyoid-burst-to-UES-opening interval was seen in healthy individuals for levels 2, 3 and 4 compared to levels 0 and 1, while the opposite effect was seen for measures of UES opening duration. Measures of time-to-LVC, LVC duration and pharyngeal area at maximum constriction showed general stability across consistencies in healthy participants, with mixed results in patient populations. Predominant volume rather than consistency effects were seen for measures of UES opening duration and maximum UES distention, both in healthy and patient participants. Measures of hyoid and laryngeal excursion were challenging to synthesize due to differences in measurement units and definitions; however, minimal differences were seen across consistencies when anatomical normalization was employed. Finally, in comparison to the previous review [6, 7], this update contains data from an entirely new area of swallowing research, in which high resolution pharyngeal manometry has been used to study pressure-flow dynamics, both in healthy adults and individuals with dysphagia. This new area of literature suggests that bolus consistency influences are present in measures of UES contractility (in the form of higher pressures as well as greater admittance reflecting greater luminal distention with thicker consistencies), but are less obvious in measures of pharyngeal contractility or bolus flow. Limitations As with any study, the results of this systematic review update need to be interpreted in the context of several acknowledged limitations. Chief among these is the fact that the data synthesis article set has an end date of May 11, 2024. Given the enormous number of studies identified in the search, and the work involved in quality appraisal and data extraction, it was not possible to perform a further search update after this date. The criteria imposed as study inclusion criteria also limit the dataset to some extent. These include the omission of studies that did not use IDDSI terminology to describe stimuli (which was the case for the majority of articles reporting EMG, acoustic and accelerometry data), studies published in languages other than English, studies with sample sizes of < 10 participants, studies that focused primarily on validation of novel scales, and studies in which the exploration of different bolus consistencies within a study protocol was dictated in a contingent fashion (e.g., studies using the Volume-Viscosity Swallowing Test [119]), thereby precluding direct comparison of the severity or frequency of parameters reflecting impairment across consistencies. Similarly, we excluded studies using standard protocols or analysis methods in which results are not broken out separately by consistency, such as the MBSImP™ [120], Dynamic Grade of Swallowing Toxicity (DIGEST™) [121, 122] or Fiberoptic Endoscopic Dysphagia Severity Scale (FEDSS) [123]. We acknowledge that these exclusion criteria may have introduced a potential selection bias. A further limitation is that many of the reviewed studies described their stimuli using IDDSI terminology, but did not report specific results for IDDSI tests of stimulus properties to confirm the appropriateness of those labels. Finally, although we did not specify any age limitations for study inclusion, the final data synthesis set included only two studies reporting data for children [65, 88]. Despite identifying a very large number of articles for data synthesis, coverage of different consistencies, and of bolus volumes within and across these consistencies, varied across articles. For this reason, true meta-analysis of quantitative data was not possible, limiting our ability to estimate effect sizes and directly compare results across studies. Where possible, comparisons were synthesized for matching stimuli and volumes, but in some cases information was only reported at the consistency level. Similarly, the variables that were included varied widely across included articles. Even within a domain such as swallowing safety or efficiency, some articles chose to report measures of parameter severity (for example, greater residue) while others reported the frequency of findings suggesting impairment in those parameters (for example, the frequency of PAS scores above a specified threshold). The absence of results for a given parameter in this review reflects a lack of reporting rather than an absence of consistency-related findings. To the extent possible, comparisons were limited to studies using the same definitions for parameters of interest. However, particularly in cases where frequencies of impairment were reported, variability in definitions, and in the thresholds used to define scores representing impairment or clinical concern was noted. Conclusions In this manuscript, we report the main results of a systematic review update project, exploring differences in measures of oropharyngeal swallowing function and physiology across bolus consistencies as defined by the IDDSI Framework. In comparison to the original review [6, 7], this update reveals substantial growth in the number of published scientific articles containing relevant data. This article synthesizes data from 70 articles with sample sizes of at least 10 participants, reporting quantitative data for at least two bolus consistencies. Main findings from the review can be summarized as follows: The frequency and severity of penetration or aspiration of material into the airway during swallowing is highest for thin liquids (IDDSI level 0) or mixed consistencies containing a thin liquid component, and reduces with incremental degrees of thickening, beginning with slightly thick liquids (IDDSI level 1). The frequency and severity of penetration or aspiration of material into the airway during swallowing is highest for thin liquids (IDDSI level 0) or mixed consistencies containing a thin liquid component, and reduces with incremental degrees of thickening, beginning with slightly thick liquids (IDDSI level 1). In healthy swallowing, both the frequency and amount of pharyngeal residue increases across incremental degrees of thickness from thin (level 0) to mildly thick (level 2), when taken by sip. Further increases are not seen with level 3 and 4 moderately and extremely thick liquids, with this pattern attributed to smaller volumes, where these stimuli have been served by teaspoon. Mixed results in patients suggest that careful evaluation is needed to ensure that recommendations for thicker consistencies to improve swallowing safety do not negatively impact swallowing efficiency. In healthy swallowing, both the frequency and amount of pharyngeal residue increases across incremental degrees of thickness from thin (level 0) to mildly thick (level 2), when taken by sip. Further increases are not seen with level 3 and 4 moderately and extremely thick liquids, with this pattern attributed to smaller volumes, where these stimuli have been served by teaspoon. Mixed results in patients suggest that careful evaluation is needed to ensure that recommendations for thicker consistencies to improve swallowing safety do not negatively impact swallowing efficiency. Measures of time to swallow initiation become longer as liquid consistency becomes thicker, both in healthy and patient populations. Measures of time to swallow initiation become longer as liquid consistency becomes thicker, both in healthy and patient populations. Measures related to the timeliness and duration of laryngeal vestibule closure, and of pharyngeal area at maximum constriction show stability across bolus consistencies in healthy and patient populations. Measures related to the timeliness and duration of laryngeal vestibule closure, and of pharyngeal area at maximum constriction show stability across bolus consistencies in healthy and patient populations. Measures of the duration and extent (i.e. distention) of upper esophageal sphincter opening are primarily influenced by bolus volume rather than bolus consistency, in both healthy and patient populations. Measures of the duration and extent (i.e. distention) of upper esophageal sphincter opening are primarily influenced by bolus volume rather than bolus consistency, in both healthy and patient populations. High resolution pharyngeal manometry, a technique that is new since the previous review, shows that measures of upper esophageal sphincter contractility and distention are more sensitive to the effects of bolus consistency than measures of pharyngeal contractility or bolus flow. Measures of contractile integral pressure throughout the pharynx and of hypopharyngeal intrabolus pressure display minimal variation across consistencies. High resolution pharyngeal manometry, a technique that is new since the previous review, shows that measures of upper esophageal sphincter contractility and distention are more sensitive to the effects of bolus consistency than measures of pharyngeal contractility or bolus flow. Measures of contractile integral pressure throughout the pharynx and of hypopharyngeal intrabolus pressure display minimal variation across consistencies. This review also identifies several gaps in the scientific literature, which require further research. In particular, there is a paucity of evidence regarding the effects of bolus consistency on swallowing in children and very little information regarding swallowing with mixed consistencies. In studies with adults, this review reveals variability in the types of parameters chosen for measurement. The development of guidance regarding core outcomes to measure when evaluating the impact of bolus consistency on swallowing, taking both expected values in healthy swallowing and bolus volume into consideration, would help to move the field forward. It would also be worthwhile to establish an open-access living review database, to keep appraised of new information in an efficient manner. Supplementary Material Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00455-026-10977-w.