Dysphagia is a common complication of stroke and its reported frequency ranges widely from 19% to 81%¹. Complications from dysphagia include aspiration pneumonia, malnutrition, and dehydration which can lead to death^2,3. It is well known that patients with dysphagia have a high risk of airway aspiration, which can lead to aspiration pneumonia. Previous studies reported that the incidence of pneumonia in stroke patients had a wide range of 7% to 29% in those with dysphagia, and the relative risk is about 3 times higher than 2% to 8% patients without dysphagia. In particular, patients with confirmed airway aspiration are 12 times more likely to develop pneumonia than those without airway aspiration¹. Videofluoroscopic swallowing study (VFSS) is the gold standard test for patients with suspected dysphagia. It is used to determine the cause and severity of dysphagia, and also establish treatment strategies⁴. Silent aspiration cannot be detected through clinical examinations, and the fact that silent aspiration is significantly related to the occurrence of pneumonia highlights the usefulness of VFSS. Although there is no specific protocol for VFSS, it is common to use a test diet such as yogurt, porridge, rice, and liquids in Korea⁵.

Dietary modification based on VFSS results is a commonly used treatment strategy for dysphagia. If oral intake is possible for a dysphagic patient, the viscosity of the diet is adjusted to the patient. When ingesting liquids, the viscosity should be increased by mixing a starch-based swallowing aid, which is the fluid thickener, if necessary⁶. The selection of appro-priate diet according to VFSS results tends to be subjective to the clinician’s judgment. The patient’s age, underlying disease, cognitive function, pre-test dietary type, and functional state are considered in combination as numerous factors that affect the likelihood of aspiration pneumonia. Regarding patients with dysphagia after stroke, it is known that the risk of aspiration is reduced by mixing thickening agents when drinking water compared to ingesting thin liquids^6,7. However, in patients with mild dysphagia capable of adequate secretion expectoration, the criteria for the application of the fluid thickener are not clear. Therefore, the use of fluid thickeners for patients with mild dysphagia has been based on the judgements of individual clinicians.

Dysphagic patients are vulnerable to dehydration as a result of limited fluid intake⁸. A state of dehy-dration reduces salivary secretion, promotes colony formation of harmful bacteria, and makes a patient more susceptible to infection, which can be a risk factor for pneumonia. It is also associated with kidney failure, constipation, urinary tract infections, cognitive decline, and respiratory infections^9,10. Previous studies have reported that dehydration status after stroke is associated with worse long-term prognosis and higher admission costs^11-14. Dehydration occurring in dysphagic patients is associated with an increase in morbidity and mortality, and therefore the water intake of patients must be closely monitored and appropriately supplied^8,14. Meanwhile, some patients complain of the bad texture of the thickened water which results in less water intake. Insufficient water intake due to the poor texture of the thickened fluid may increase the risk of dehydration.

Blood urea nitrogen/creatinine (BUN/Cr) ratio, serum osmolality, inferior vena cava (IVC) diameter and urine specific gravity are used to evaluate the dehydrated status of a patient. Among such indica-tors, elevated BUN/Cr ratio is most commonly used to detect dehydration^11,15-17. BUN/Cr ratio 20 or higher with normal Cr is generally considered as a marker for dehydrated status¹⁸. The purpose of this study is to compare the changes in BUN/Cr ratio, an indicator of dehydrated status, depending on the application of fluid thickeners in patients with mild dysphagia.

1. Study design

We retrospectively reviewed the medical records of patients who underwent VFSS from January, 2016 to December, 2020. Each VFSS was performed by physi-cians of the Department of Rehabilitation Medicine according to a modified version of Logemann’s pro-cedure¹⁷. VFSS was performed as the patients swallo-wed several types of food: yogurt, porridge, rice and juice. Three different volumes were used (2 cc, 5 cc and cup drinking) for juice. A total of 6 dietary trials were set up: puree trial with yogurt, semisolid trial with porridge, solid trial with rice, liquid trials (2 cc, 5 cc and cup) with juice. Each food type was mixed with liquid barium to ensure the observation of bolus during the fluoroscopy procedure.

The inclusion criteria were as follows: (1) patients diagnosed with stroke through brain magnetic reso-nance imaging or computed tomography; (2) over 18 years of age, (3) patients with mild dysphagia that underwent all the dietary trials of the VFSS exami-nation and showed penetration signs with a Penetra-tion-Aspiration Scale (PAS) of 4 or less, only at the liquid trials, and permitted to take a regular diet.

The exclusion criteria were as follows: (1) signs of penetration or aspiration during any of the puree, semisolid or solid trials of VFSS that indicate promi-nent dysphagia; (2) signs of penetration or aspiration with PAS 5 or more at the liquid trials of VFSS¹⁵; (3) those with no abnormalities found in all trials of VFSS and considered to have normal swallowing func-tions¹⁶; (4) medical history which can elevate baseline Cr levels; (5) history of degenerative nervous system disease such as parkinsonism that can cause dysphagia; (6) those with the other factors that may raise the BUN/Cr ratio, such as heart failure, gastrointestinal bleeding or intake of high protein meals.

A total of 81 patients were enrolled and classified into two groups (Group A and Group B) according to the dietary instructions from the VFSS results. Group A included patients who were allowed to freely consume water with a regular diet. Group B was comprised of patients who were enabled to take a regular diet, but recommended to add a fluid thickener when ingesting water. To verify the actual diet, the patients’ prescriptions and nursing records were reviewed, and in particular, Group B was monitored to check whether the fluid thickener was actually used as guided. Clinicians ordered a regular diet to patients with mild dysphagia who met the conditions listed above. During the 5-year study period, a total of 10 clinicians were in charge of the VFSS for about 6 months each, and the final dietary decision during the 5-year period was made by one supervising clinician. The application of the fluid thickener depended on the subjective, albeit arbitrary judgment of the clinician, which took into conside-ration a number of factors.

We reviewed the VFSS records of the enrolled patients including the PAS and severity of pharyngeal pooling at the liquid 2 cc/5 cc/cup trials^15,19,20. PAS and pharyngeal pooling in VFSS are commonly used indicators for evaluating the severity of dysphagia. PAS is an 8-point scale that categorizes the depth of penetration or aspiration, and response to airway invasion during the pharyngeal phase of swallowing in VFSS¹⁵. Pharyngeal pooling indicates post-swallow pharyngeal residue after swallowing in two locations: the valleculae and pyriform sinus^19,21. PAS is scored from 1 to 8, in which 1 indicates material does not enter the airway and 8 indicates material enters the airway, passes below the vocal folds and no effort is made to reject¹⁵. Severity of pharyngeal pooling is graded into none (no residue), mild (residue filling <25% of the height of the available space), moderate (residue filling 25-50% of the height of the available space), and severe (residue filling >50% of the height of the available space). Each grade was indicated as 0, 1, 2, and 3, and the lowest score of the liquid 2 cc/5 cc/cup trials was recorded²¹.

To analyze whether the patients of each group were dehydrated or at risk of dehydration, we calculated the serum BUN/Cr ratio through blood tests, which were performed within 1 week of the VFSS test date and 1 month after the first. A higher BUN/Cr ratio signified higher levels of dehy-dration^11,22,23. To compare the risk of aspiration pneumonia between the two groups, clinical symptoms of pneumonia such as sputum and fever of 38.0 degrees or higher, and also clinical signs of pneumo-nia from chest X-rays and elevated blood C-reactive protein (CRP) levels were reviewed within a month after VFSS. Patients with such symptoms and signs, and those that required empirical antibiotic treat-ments according to the opinion of a respiratory physician were regarded as suspected pneumonia and the number of cases was counted. We recognized the possible clinical differences of the two groups and compared the baseline characteristics: age, gender, etiology of stroke, VFSS time from the onset, current diet method, intravenous (IV) hydration, and cogni-tive function assessed by the Korean version of Mini-Mental State Examination (K-MMSE). Current diet methods were divided into three categories: (1) regular diet with free fluid intake, (2) oral intake, but dietary modification including fluid thickeners, and (3) tube feeding using a nasogastric tube. Patients who received IV hydration which may lower the BUN/Cr ratio during the follow-up period were identified, and the amount was recorded.

We calculated the change in the values of BUN/Cr ratio from the initial to 1 month follow-up blood tests (△BUN/Cr ratio) of each group and compared the △BUN/Cr ratio between the two groups. We compared whether there was a difference in the initial, 1 month f/u, and △BUN/Cr ratio depending on the three current diet methods of the patients. Finally, we analyzed the correlation between the application of fluid thickeners and the △BUN/Cr ratio, and adjusted for total amount of IV hydration, time from onset, and severity of dysphagia (PAS at cup trial and pharyngeal pooling).

2. Statistical analysis

In this study, statistical analysis was performed using SPSS for Windows version 21. The demographic data and initial evaluations were analyzed by the independent t-test. The comparative analysis regar-ding the △BUN/Cr ratio between the two groups was also conducted by the independent t-test. The paired t-test was used to compare the differences of BUN/Cr ratio between the initial and 1 month follow-up blood tests of each group. One-way analysis of variance was applied to compare the BUN/Cr ratio of patients according to the three types of current diet methods. Stepwise linear regression and partial correlation analyses were applied to examine the association between the application of fluid thickeners and △BUN/Cr ratio. P-value less than 0.05 was considered statistically significant.

In this study, we evaluated 81 eligible patients: 46 were female, 47 were diagnosed with ischemic stroke and 34 were diagnosed with hemorrhagic stroke. The patients were classified into two groups according to dietary instructions: Group A was consisted of 37 patients and Group B, 44 patients. The demogra-phical and clinical characteristics of the two groups are listed in Table 1. There were no significant differences in characteristics, including age, gender, etiology, VFSS time from the onset, K-MMSE and initial blood test (CRP, BUN, Cr, and BUN/Cr ratio). The distribution of current diet methods in the two groups were as follows: regular diet with free fluid intake, Group A (8), Group B (7); oral intake, but dietary modification including fluid thickener, Group A (20), Group B (19); tube feeding using a nasogastric tube, Group A (7), Group B (17). There were significantly more patients with tube feeding in Group B (P=0.033). There was a significant difference of △BUN/Cr ratio between the two groups (P=0.033). There were 4 cases of suspected pneumonia within 1 month after the VFSS test with no significant differences between the two groups. A total of 13 patients received IV hydration at least once during the follow-up period, and all were applied within 1 L/day for less than 5 days. Comparison of the VFSS results between the two groups can be seen in Table 2. The PAS at liquid 2 cc/5 cc trials was significantly higher in Group B (P=0.02, P=0.02). There was no significant difference in the PAS at cup trial and pharyngeal pooling between the two groups.(Table 2) Comparison of the differences of BUN/Cr ratio between the initial and 1 month follow-up blood tests of each group are listed in Table 3. △BUN/Cr ratio decreased significantly in Group A (P=0.022) and slightly increased but was not statistically significant in Group B (P=0.34). Initial BUN/Cr ratio was slightly higher in patients with dietary modification including fluid thickeners, but was not statistically significant. Among the six groups sorted by initial and recommended dietary types, although not statistically significant, BUN/Cr ratio decreased most in patients who had initially taken fluid thickeners before the VFSS and had free water intake after the VFSS (△BUN/Cr ratio=−2.48±5.70, P=0.06).(Table 3)

Table 1 . Comparison of demographical, clinical characteristics, and the △BUN/Cr ratio from initial to 1 month f/u blood tests of group A and group B.

Parameter	Free fluidgroup A (N=37)	Fluid thickenergroup B (N=44)	P-value
Age (years)	63.0±16.8	66.2±14.7	0.36
Gender, male:female	17:20	18:26	0.65
Etiology, ischemic stroke:hemorrhagic stroke	21:16	26:18	0.83
Time from onset (days)	105.4±217.0	99.7±192.6	0.90
K-MMSE	20.86±7.93	18.00±8.03	0.11
Current diet methods
RD (%)	8 (21.6)	7 (15.9)	0.52
DM (%)	21 (56.7)	18 (40.9)	0.15
TF (%)	8 (21.6)	19 (43.1)	0.038*
Initial blood test
CRP	0.79±1.76	0.81±1.00	0.96
BUN	15.25±4.78	15.67±5.36	0.70
Cr	0.75±0.22	0.79±0.28	0.53
BUN/Cr ratio	20.84±5.66	20.62±5.76	0.85
1 month f/u BUN/Cr ratio	18.92±6.06	21.67±7.00	0.62
△BUN/Cr ratio	−1.91±4.87	1.05±7.36	0.033*
Suspected pneumonia within 1 month (%)	2 (5.4)	2 (4.5)	0.86
Intravenous hydration (liter)	7 (2.5±0.57)	6 (2.5±0.54)	0.53

Values are presented as mean±standard deviation.

BUN: blood urea nitrogen, Cr: creatinine, △BUN/Cr ratio: change in the values of BUN/Cr ratio, f/u: follow up, K-MMSE: Korean version of the mini-mental state examination, RD: regular diet with free fluid intake, DM: dietary modification including fluid thickener, TF: tube feeding, CRP: C-reactive protein.

*P<0.05.

Table 2 . Comparison of VFSS results between group A and group B.

Parameter	Free fluid group A (N=45)	Fluid thickener group B (N=57)	P-value
VFSS results
PAS (liquid 2 cc)	1.15±0.42	1.42±0.60	0.02*
PAS (liquid 5 cc)	1.34±0.60	1.62±0.56	0.02*
PAS (liquid cup)	1.79±0.63	2.00±0.83	0.18
Pharyngeal pooling	0.65±0.74	0.90±0.78	0.13

Values are presented as mean±standard deviation.

VFSS: videofluoroscopic swallowing study, PAS: penetration-as-piration scale.

*P<0.05.

Table 3 . Comparison of the initial, 1 month f/u, and △BUN/Cr ratio depending on the three types of current diet methods in each group A and B.

Parameter	Initial BUN/Cr ratio	1 month f/u BUN/Cr ratio	△BUN/Cr ratio	P-value
Free fluid Group A (N=37)	20.84±5.66	18.92±6.06	−1.91±4.87	0.022*
RD (N=8)	19.45±6.89	19.15±7.21	−0.30±3.86	0.45
DM (N=21)	21.85±5.80	19.36±6.73	−2.48±5.70	0.06
TF (N=8)	19.60±3.73	17.56±2.08	−2.03±3.19	0.11
Fluid thickener Group B (N=44)	20.62±5.76	21.67±7.00	1.05±7.36	0.34
RD (N=7)	19.54±3.22	20.52±7.71	0.97±5.80	0.67
DM (N=18)	21.88±7.92	21.90±6.73	0.02±8.61	0.99
TF (N=19)	19.81±3.73	21.86±7.73	2.05±6.80	0.20
Total (N=81)	20.84±5.66	18.92±6.06	−1.91±4.87	0.022*
RD (N=15)	19.49±5.31	19.78±7.21	0.29±4.73	0.813
DM (N=39)	21.86±6.77	20.53±6.76	−1.33±7.20	0.256
TF (N=27)	19.75±3.66	20.59±6.65	0.84±6.19	0.487
P-value (P1)	0.21	0.92	0.38

Values are presented as mean±standard deviation.

f/u: follow up, BUN: blood urea nitrogen, Cr: creatinine, △BUN/Cr ratio: change in the values of BUN/Cr ratio, RD: regular diet with free fluid intake, DM: dietary modification including fluid thickener, TF: tube feeding, P₁: one-way analysis of variance among the three groups (current diet methods).

*P<0.05.

Stepwise regression analysis showed that the △BUN/Cr ratio was possitively correlated with the application of fluid thickeners (β=0.229, P=0.020). Total amount of IV hydration, time from onset, PAS at cup trial, and pharyngeal pooling had no stati-stically significant correlation with △BUN/Cr ratio. In the partial correlation coefficient adjusted for IV hydration, time from onset, PAS at cup trial, and pharyngeal pooling, the application of fluid thicke-ners was independently associated with △BUN/Cr ratio (r=0.225, P=0.049).(Table 4)

Table 4 . Stepwise linear regression analysis for the association between the application of fluid thickener and △BUN/Cr ratio and partial correlation coefficient adjusted for IV hydration, time from onset, PAS at cup trial, and pharyngeal pooling.

Total (N=81)	△BUN/Cr ratio
	Stepwise linear regression			Partial correlation coefficient
	β	P		r	P
Fluid thickener	0.229	0.020*		0.225	0.049
IV hydration	−0.074	0.225
Time from onset (days)	−0.021	0.428
PAS (liquid cup)	−0.014	0.451
Pharyngeal pooling	0.029	0.398

Values are presented as mean±standard deviation.

β: standardized regression coefficient, r: partial correlation coefficient, BUN: blood urea nitrogen, Cr: creatinine, △BUN/Cr ratio: change in the values of BUN/Cr ratio, IV: intravenous, PAS: penetration-aspiration scale.

*P<0.05.

This is a retrospective study that divided the participating patients into two groups according to recommendations for fluid thickeners. The criteria for prescribing a fluid thickener to patients with mild dysphagia based on the VFSS test does not seem clear. Also, to our knowledge, there are no standard clinical guidelines or criteria for mild dysphagia on the VFSS test. O’Neil et al.²⁴ proposed the Dysphagia Outcome and Severity Scale (DOSS) to classify the severity of dysphagia, and defined mild dysphagia as a patient who may exhibit one or more of the following: (1) aspiration of only thin liquids but with strong reflexive coughs to clear completely; (2) airway penetration midway to vocal cords with consistency but clears spontaneously; (3) pharyngeal pooling that is cleared spontaneously; (4) mild oral dysphagia with reduced mastication or/and oral retention that is cleared spontaneously²⁴. The authors suggest that supervision and dietary adjustments may be necessary in patients with mild dysphagia²⁴. However, the definition of mild dysphagia based on the VFSS has not yet been established, and in this study, PAS 4 or less only at liquid trials were considered as mild dysphagia based on the concept of DOSS. PAS 6 only at liquid trials may correspond to mild dysphagia according to DOSS, but was excluded in this study. With such considerations, patients who showed penetration with a PAS 4 or less only at liquid trials without any abnormality in puree, semisolid and solid trials were enrolled. This was because patients with abnormalities in puree, semisolid or solid trials necessarily required dietary modifications. Patients who showed penetration or aspiration with PAS 5 or more at the liquid trials were also excluded from this study as dietary modifications to add fluid thickeners were generally required.

In this study, there were no significant differences in baseline characteristics, including age, K-MMSE and initial blood test (CRP, BUN, Cr, and BUN/Cr ratio) between the two groups. The clinicians may have recommended fluid thickeners more frequently to the elderly or patients with severe cognitive decline, and patients in Group B were older with lower MMSE scores, however, it was not a significant difference. There were more patients in Group B who initially took tube feeding which possibly infers that clinicians recommended fluid thickener more fre-quently to such patients.

Both groups showed elevated mean BUN/Cr ratio in the initial blood tests (Group A: 20.84, Group B: 20.62). This suggests that patients with dysphagia can be easily dehydrated. A significant proportion of the patients enrolled were dehydrated, suggesting that adequate hydration is essential for the prognosis of patients with mild dysphagia^8,11,23,25. Li et al.¹¹ and Wu et al.¹³ reported that dehydration independently showed poor long-term prognosis in patients treated with thrombolysis after acute ischemic stroke. Liu et al.¹⁶ reported that dehydration is an independent in-dicator predicting the prognosis for home discharge in ischemic stroke. Deng et al.²² reported that ele-vated BUN/Cr is associated with poor outcome in acute ischemic stroke patients. Leng-Chieh Lin et al.²⁵ reported that the BUN/Cr ratio in patients with acute ischemic stroke was the sole indicator of stroke-in-evolution, and lowering the BUN/Cr ratio through proper hydration prevents stroke-in-evolution.

During the one month follow-up period of this study, 4 out of 81 patients reported suspected pneu-monia. They showed elevated CRP, fever over 38 degrees and non-purulent sputum. However, they but did not show rale, purulent sputum or prominent evidence of pneumonia on chest x-rays²⁶. All 4 patients were well-treated with empirical antibiotics and no chest computed tomography was performed. In patients with dysphagia after stroke, aspiration pneumonia is considered a fatal complication. However, this study reported a relatively few cases of pneu-monia and mild clinical courses. Patients with mild dysphagia are thought to have been able to cough properly even with small amounts of fluid passing into the airways. Although there was no significant difference in the incidence of pneumonia between the two groups in this study, since the benefits of fluid thickeners are widely known, this study is insufficient to discuss the effect of fluid thickeners on preventing pneumonia in patients with mild dysphagia.

PAS was significantly higher at liquid 2 cc/5 cc trials in Group B. It can be inferred that clinicians had recommended fluid thickeners more frequently for patients with signs of penetration even in small amounts of fluid. On the other hand, the difference in the PAS at the cup trial between the two groups was not significant, and it seems that clinicians were relatively generous with penetration when drinking large amounts of fluid. Also, there was no significant difference in the severity of pharyngeal pooling between the two groups. If pharyngeal pooling is prominent, risk of aspiration is significant and aspi-ration can occur clinically even if PAS is normal at VFSS. In this study, the results of PAS and pharyngeal pooling indicate mild dysphagia of the enrolled patients, and yet it is notable that more than half of the patients were prescribed fluid thickeners.

In this study, it was difficult to investigate whether the current diet methods of the patients during the initial state before VFSS had an effect on dehydration. Although information regarding the dehydration risk of tube feeding compared to oral diet is limited, it may have been a confounding factor. Patients who underwent dietary modifications including fluid thic-keners for current diet methods had slightly higher initial BUN/Cr ratio. However, since it was not statistically significant, it is insufficient to conclude that these patients were more dehydrated than those who took tube feeding or regular diet with free fluid intake.

In the one-month follow-up blood test, Group A showed significant decrease in BUN/Cr ratio compared to the initial level as they drank water freely for a month. The difference of △BUN/Cr ratio between the two groups was significant. Considering that BUN/Cr ratio decreased the most in patients who had previously taken fluid thickeners but liberal water intakes after the VFSS, it can be inferred that increased water intake was the main cause of the decrease in BUN/Cr ratio in Group A. IV hydration, severity of dysphagia, and time from onset were possible confounding factors that could affect the dehydration status. However, application of fluid thickeners was still significantly correlated with △BUN/Cr ratio after adjusting for those factors. This result suggests that the dehydration condition in a patient with mild dysphagia can be alleviated by ingesting enough water through proper diet build-up, and conversely, when instructed to apply the fluid thickener, water intake may be insufficient due to poor compliance. The most likely reason is that the texture of beverages added with the fluid thickener becomes very poor. A previous study reports such disadvantages²⁷. Alves et al.²⁷ reported that it was difficult to ingest liquids added with swallowing aids even for normal people, increasing the swallowing time. Although it has been reported that fluid thickeners do not affect the water absorption rate and water bioavailability²⁸, poor flavor and texture result in poor motivation and physiologic drive to consume thickened liquids and may significantly reduce compliance with fluid intake⁷. Leibovitz et al.²⁹ reported long-term dehydration in 75% of individuals who consumed thickened liquid by oral hydration. As dysphagic patients are vulnerable to dehydration, it is important to relieve dehydration through adequate water intake. Therefore, the possi-bility of insufficient water intake due to bad texture is a possible disadvantage of the fluid thickener. Further prospective randomized studies may confirm this result. A more detailed study is required to suggest the cut-off level or guidelines to adequately prescribe fluid thickeners according to the severity of dysphagia or VFSS results.

This study has several limitations. First, its retros-pective nature and the group differences regarding the severity of dysphagia may have affected the results. Second, it was a short-term follow-up study, so it was difficult to predict the long-term effect and risk of aspiration pneumonia or dehydration. Third, other indicators for dehydration such as serum osmolality, IVC diameter, and urine specific gravity were not investigated. Therefore, elevated BUN/Cr ratio may not necessarily indicate a dehydrated status. Fourth, although the nursing records were carefully reviewed, it is unclear whether the patients strictly adhered to the prescriptions. Finally, due to the various onset time of patients during the study period, it was difficult to represent the entire group of dysphagia patients after stroke.

In conclusion, among patients with mild dysphagia that displayed minimal abnormalities only in the liquid trials of VFSS without a distinct penetration or aspiration with PAS 5 or more, a decrease in BUN/Cr ratio, which is an indicator of dehydrated status, was found in patients who drank water freely compared to those with added fluid thickeners during water intake. If fluid thickeners are prescribed for mild dysphagic patients, it is necessary to improve patient compliance with water intake to prevent dehydration.

The authors declare that there is no conflict of interest.

This study was exempted from review by the Institutional Review Board (IRB) (No: 2022-02-003). Written informed consent from the patients was waived by IRB.