Introduction

Dysphagia is a common problem in aging societies that affects up to 13 % of individuals aged 65 years and older and 51 % of institutionalized older adults [1]. Dysphagia is an important risk factor of aspiration pneumonia. The incidences of aspiration pneumonia in community- and hospital-acquired pneumonia have been reported as 60.1 and 86.7 %, respectively [2], and found to increase with age [3]. In addition, challenges related to sarcopenia as a geriatric syndrome have recently been revealed [4]. Sarcopenia is a condition characterized by loss of skeletal muscle mass and function in older adults, which is caused by aging as a primary sarcopenia factor and low activity, malnutrition, and disease as secondary sarcopenia factors [4]. Sarcopenia has been reported as an important factor of dysphagia in older adults (i.e., sarcopenic dysphagia) [5–7]. However, very few studies have examined the mechanisms responsible for the development of this disorder.

In healthy older adults, it was reported that tongue strength is associated with age, grip strength, and jump height and power [8]. This previous study revealed the effect of aging as a primary sarcopenia factor, not the effect of secondary sarcopenia factors, on tongue strength. Moreover, some studies have found that older adult inpatients undergo hospital-associated deconditioning, which is characterized by reduced physical function that occurs during acute hospitalization due to illness unrelated to specific neurological or orthopedic diseases [9, 10]. In addition, reduced muscle strength due to bed rest [11] and malnutrition have been commonly reported in older adult inpatients [12]. Taken together, these findings demonstrate that functional decline related to secondary sarcopenia is an important issue to consider for inpatients who require rehabilitation.

Against this setting, the purpose of this study was to investigate whether tongue strength is associated with muscle function and nutritional status, rather than age, and dysphagia in older adult inpatients of a rehabilitation hospital. We hypothesized that tongue strength is more strongly associated with muscle function (i.e., activity and grip strength) and nutritional status related to secondary sarcopenia than age related to primary sarcopenia. In addition, we hypothesized that tongue strength is also associated with dysphagia in older adult inpatients in rehabilitation hospitals.

Materials and Methods

Study Subjects

Subjects were consecutive older adult inpatients (65 years and older) in the post-acute phase of illness who were hospitalized for rehabilitation between October 2014 and December 2015. The research design was a cross-sectional study. Additional inclusion criteria were as follows: patients who lived independently and had no history of dysphagia prior to hospitalization in the acute care hospital; no history of cerebrovascular diseases, head and neck cancers, or underlying neuromuscular diseases such as Parkinson’s disease that might directly impair the nerves or muscles involved in tongue strength or swallowing function; mini-mental state examination (MMSE) score of 21 points or greater, reflecting mild decline in or normal cognitive function [13] (since accurate measurement of tongue strength in patients with declined cognitive function may be difficult); and presence of all upper and lower central incisors (i.e., either remaining incisors or dentures, since the tongue strength measurement probe used in the present study must be fastened to central incisors). We excluded patients who had weight restrictions due to an orthopedic condition (given effects on evaluation indices associated with muscle function); ill-fitting dentures or moving central incisors; required adjustments to the form of food due to a digestive disorder (given effects on evaluation indices associated with dysphagia); and marked swelling in body parts measured in the present study (given potential effects on evaluation indices for muscle mass). All measurements were taken within a week of hospitalization.

Tongue Strength Measurement

Tongue strength was measured using a tongue pressure measuring instrument (JMS, Hiroshima, Japan) fitted with a balloon-type disposable oral probe on one end [14]. This apparatus can measure isometric tongue pressure when the balloon is placed between the front part of the tongue and the front part of the roof of the mouth. Measurements were taken with subjects sitting in a relaxed position. Prior to measurement, each subject received the following explanation: “We will now place the balloon in the front part of your mouth. After placing the plastic pipe portion in front of the balloon stably between your upper and lower incisors, close your lips around it. Then, use your tongue to push the balloon as hard as you can against the roof of your mouth for 5 s.” Measurements were taken three times, with a 30-second break between each measurement [15]. The highest of the three measurements was defined as the maximum isometric tongue pressure (MTP). All measurements were taken by experienced speech therapists.

Assessment of Sarcopenia-Related Factors

The causes of sarcopenia were assessed as follows [4]. As a cause of primary sarcopenia, age was determined using medical records. As causes of secondary sarcopenia, three factors were assessed. The first factor was activity, which was evaluated using the barthel index (BI) [16]. The BI was assessed by experienced physical therapists. The second factor was nutritional status, which was evaluated using the Mini Nutritional Assessment—short form (MNA-SF) [17], body mass index (BMI), serum albumin levels (ALB), and controlling nutritional status (CONUT) [18]. The MNA-SF is a questionnaire survey recommended for the diagnosis of malnutrition among older adults [19], for which total scores range from 0 to 14 points. Malnutrition is indicated by scores of seven or fewer points; 8–11 points indicate a risk of malnutrition, and 12–14 points indicate a favorable nutritional status. Experienced registered dietitians conducted interviews to obtain MNA-SF scores. The CONUT score is calculated from ALB, total lymphocyte count, and total cholesterol. Total scores range from 0 to 12 points, with lower scores indicating a more favorable nutritional status. The third factor was disease, which was evaluated by the presence of cachexia based on diagnostic criteria [20] and C-reactive protein (CRP) levels as an inflammatory reaction index. Table 1 shows the evaluation items in this study by sarcopenia category and cause.

Table 1 Evaluation items by sarcopenia category and cause
Full size table

As diagnostic criteria of sarcopenia, grip strength and muscle mass were assessed [4]. Grip strength and muscle mass can be affected by both primary sarcopenia and secondary sarcopenia. Grip strength was assessed using a digital grip strength dynamometer (Model T.K.K. 5401, Takei, Japan). Subjects were asked to sit in a relaxed position and grip the instrument as hard as they could twice for both hands. The highest measurement was used for analysis [21]. Muscle mass was assessed by measuring calf circumference (CC), a method validated for muscle mass measurement in the diagnosis of sarcopenia [22–25]. Arm muscle area (AMA) was also measured as an additional index of muscle mass [14, 26]. CC measurements and arm circumference (AC) values, required to calculate AMA, were measured on the non-dominant lower side. In subjects with fractures, measurements were taken on the healthy side. CC and AC measurements were taken with a non-elastic tape measure, which was placed around the calf or arm without compressing subcutaneous tissue and moved along the length of the calf or arm to obtain the maximal circumference. Triceps skinfold (TSF) was measured using a caliper on the same arm as was used for AC measurement. Each measurement was taken twice and the mean was determined for analysis. AMA was calculated from the AC and TSF using the following equation: [AC – (TSF × π/10)]2/4π. Physical measurements were taken with subjects in a seated position. To eliminate inter-examiner variation, an experienced rehabilitation therapist took all measurements. In addition, to avoid any variability caused by swelling due to the time of day, all measurements were taken between 5 and 6 pm. The presence of sarcopenia was confirmed by both low grip strength (<26 kg for men, <18 kg for women) and low muscle mass based on the criteria of the Asian working group for sarcopenia [27]. For low muscle mass, we referred to CC values of previous studies that examined muscle mass in adults with sarcopenia and set cut-off values of <34 cm for men and <33 cm for women [22].

Evaluation of Dysphagia Indices

Dysphagia indices included the functional oral intake scale (FOIS), which reflects oral intake level [28]; the modified water swallow test (MWST), which screens individuals for aspiration risk [29]; and the 10-item eating assessment tool (EAT-10), which assesses subjective dysphagia symptoms [30, 31]. Each of these tools has been found to be associated with dysphagia through the use of videofluoroscopic examination of swallowing function, an objective examination. These evaluations were conducted by experienced speech therapists. The FOIS ranged from Level 1 (nothing by mouth) to Level 7 (total oral diet with no restrictions). The MWST was conducted by pouring 3 ml of cold water into the seated subject’s floor of the mouth, and asking them to swallow. If the subject was unable to swallow, a score of 1 was given. If they were able to swallow but experienced dyspnea after swallowing, a score of 2 was given. If they were able to swallow but experienced coughing or wet-hoarseness after swallowing, a score of 3 was given. If they were able to swallow without those symptoms, a score of 4 was given. If they were able to perform two saliva swallows within 30 s after swallowing water, a score of 5 was given. A MWST score ≤3 is considered abnormal and indicates the presence of aspiration. The EAT-10 is a questionnaire used to screen for dysphagia. Each item is scored from 0 to 4; a score of 0 indicates no problem and 4 indicates a severe problem. An EAT-10 score ≥3 is considered abnormal and indicates the presence of dysphagia. The presence of dysphagia was determined based on one of the following: FOIS ≤5, MWST ≤3, or EAT-10 ≥ 3.

Assessment of Other Parameters

Other parameters that were assessed included sex, illness that required hospitalization, and the number of days hospitalized at the acute care hospital. This information was obtained from medical records. The occlusion of the molar region and number of remaining teeth were also evaluated given their potential effects on dysphagia [32, 33]. In addition, since the central incisors are required to stabilize the probe during MTP measurements, we divided subjects into those whose upper or lower central incisors were complete dentures (potential for low stability) and those without complete dentures (potential for high stability). The status of occlusion at the molar region, regardless of natural teeth or artificial teeth, was assessed using the Eichner index for the following groups: group A, occlusal support for molars and premolars on both sides; group B, reduced occlusal support; and group C, no occlusal support [34]. Dental evaluations were conducted by experienced dental hygienists.

Statistical Analysis

The unpaired t test (MTP and CC) and the Mann–Whitney U test (all other variables) were used to examine sex-dependent differences. The unpaired t-test was also used to examine differences in MTP according to the central incisor status. Fisher’s exact test was used to examine nominal scales. Correlations between MTP and each parameter were assessed using bivariate correlation analysis, Pearson’s product-moment correlation coefficients for parametric values, and Spearman’s rank-correlation coefficients for non-parametric values. Multiple regression analyses were also conducted with MTP and FOIS as dependent variables. In the multiple regression analysis, significant factors associated with MTP in univariate analyses were chosen as independent variables for MTP, and possible factors associated with sarcopenic dysphagia were chosen as independent variables for FOIS. Multiple regression analysis for MTP was conducted in all patients and patients aged 80 years and older to examine interaction with age. P < 0.05 was considered statistically significant. SPSS Statistics 23 Software (IBM Corporation, Armonk, NY, USA) was used for statistical analyses. Measurement values did not include missing data.

Ethical Considerations

The present study was performed according to the ethical recommendations of the Declaration of Helsinki and was approved by the Institutional Review Board of Setagaya Memorial Hospital (approval no. 2014-1). All enrolled patients provided their written informed consent.

Results

A total of 174 patients (64 men, 110 women) participated in the present study. The median age (interquartile range, IQR) was 84 (80–89) years. The median number of days in acute care was 34.5 (22.0–52.3) days. The most common reason for hospitalization was orthopedic, such as fractures of the proximal femur (61.5 %). The median (IQR) MTP was 25.2 (21.4–32.3) kPa in men and 26.7 (22.4–32.6) kPa in women, showing no sex-dependent difference. A significant correlation was not found between MTP and days in acute care (r = −0.02, p = 0.798). Sarcopenia was confirmed in 77.0 % (n = 134) of patients, and dysphagia in 32.2 % (n = 56). Subject characteristics are shown in Table 2.

Table 2 Subject demographic characteristics
Full size table

The median (IQR) MTP among patients with complete dentures was 24.3 (18.3–28.7) kPa, while that for subjects without complete dentures was 27.4 (22.4–32.6) kPa, showing a significant difference according to central incisor condition (p = 0.028). Bivariate correlation analyses between MTP and each sarcopenia-related factor are shown in Table 3. Significant correlations were found between MTP and age (r = −0.19, p = 0.012), muscle mass (CC, r = 0.45, p < 0.001; AMA, r = 0.38, p < 0.001), grip strength (r = 0.46, p < 0.001), and BI (r = 0.35, p < 0.001). Regarding the relationship between MTP and nutritional status, we identified significant correlations between MTP and MNA-SF (r = 0.44, p < 0.001), ALB (r = 0.25, p = 0.001), CONUT scores (r = −0.16, p = 0.037), and BMI (r = 0.36, p < 0.001). Significant correlations were also found between MTP and all indices of dysphagia (FOIS, r = 0.48, p < 0.001; EAT-10, r = −0.24, p = 0.002; and MWST, r = 0.36, p < 0.001) (Table 4). Multivariate linear regression analysis in all patients and patients 80 years and older revealed that MTP was significantly associated with grip strength in all patients (coefficient = 0.33, 95 % confidence interval (CI) 0.12–0.54, p = 0.002), but not patients 80 years and older (coefficient = 0.25, 95 % CI −0.01 to 0.52, p = 0.061). MTP was also significantly associated with MNA-SF in both groups (coefficient = 0.74, 95 % CI 0.12–1.35, p = 0.019; coefficient = 0.88, 95 % CI 0.20–1.56, p = 0.012, respectively). Interestingly, MTP was not significantly associated with age (coefficient = −0.05, 95 % CI −0.23 to 0.12, p = 0.560; coefficient = −0.14, 95 % CI −0.43 to 0.15, p = 0.346, respectively) (Table 5). Finally, MTP was significantly associated with FOIS (coefficient = 0.02, 95 % CI 0.00–0.15, p = 0.047) (Table 6).

Table 3 Bivariate correlation analyses between MTP and each sarcopenia-related factor
Full size table
Table 4 Bivariate correlation analyses between MTP and dysphagia indices
Full size table

1

Table 5 Multivariate linear regression analysis for MTP
Full size table
Table 6 Multivariate linear regression analysis for FOIS
Full size table

Discussion

The present study found that MTP was independently associated with grip strength and MNA-SF, even after adjusting for age, in older adult inpatients of a rehabilitation hospital. In addition, MTP was independently associated with FOIS. Although the association between tongue strength and grip strength related to primary sarcopenia has been reported in healthy older adults [8], to our knowledge, this is the first study to reveal associations between tongue strength and grip strength and nutritional status, which may be related to secondary sarcopenia in older adult inpatients.

No independent association was observed between MTP and age in the multivariate linear regression analysis in either group, although weak associations between MTP and age were detected in bivariate correlation analyses. In addition, grip strength (used as a diagnostic criterion of sarcopenia) was more strongly associated with BI and MNA-SF as indices of secondary sarcopenia factors in this study than age as a primary sarcopenia factor in bivariate correlation analyses. Thus, for older adult inpatients in rehabilitation, the effects of secondary sarcopenia, rather than those of primary sarcopenia, on reduced tongue strength might be more important.

MTP was significantly correlated with muscle mass based on correlation coefficients in the present study. However, multiple regression analysis did not show a significant association. Thus, the association between MTP and grip strength may be more relevant than that with muscle mass in older adult inpatients in rehabilitation. In a previous study of healthy older adults, tongue strength was associated with grip strength, but not muscle mass, using simple regression analysis [8]. This may indicate that tongue strength is more closely correlated with muscle mass in older adult inpatients of rehabilitation hospitals than that of healthy older adults.

We also found that MTP was independently associated with grip strength in all patients, but not with BI. Handgrip strength has been shown to be a valid surrogate measurement of overall muscular strength [35]. Thus, our findings suggest that decrease in overall muscle strength, which may result from bedrest during hospitalization, is more important as a factor than the actual performance of activities of daily living in the reduction of tongue strength. A possible explanation for the lack of independent association with BI is that because the present study included many orthopedic patients, BI may be more affected by orthopedic disease itself than sarcopenia.

In the present study, MTP was independently associated with the MNA-SF score. In one previous study of older adult inpatients at an acute care hospital, tongue strength was independently associated with ALB, but not with the MNA-SF score [14]. Low ALB is a known marker of inflammatory reactions in acute patients [36]. In addition, while validated screening tools such as MNA-SF are indispensable for the diagnosis of malnutrition, ALB is not [19]. Thus, the association between tongue strength and nutritional status was not directly addressed in the previous study. In the present study, MTP was not independently associated with ALB. Although tongue strength may have been affected by acute inflammation in the previous study, this effect could be decreased in patients of a rehabilitation hospital. Thus, we surmise that this discrepancy was caused by differences in subject characteristics. Given a previous report that physical function markedly improved when nutritional therapies were fortified in a rehabilitation hospital [37], malnutrition is a potential factor that affects tongue strength in inpatients of rehabilitation hospitals. In addition, as suggested by multivariate linear regression analysis, it may be that the greater the respondent’s age, the greater the effect of malnutrition on tongue strength.

Very few patients showed cachexia, and none had neuromuscular disorders. Moreover, although the mean CRP value was lower than the reference value, it was not associated with MTP in the present study. Thus, the effects of disease-related sarcopenia on measurements taken in the present study were thought to be minimal.

MTP was significantly correlated with indices of dysphagia: FOIS, EAT-10, and MWST. Although some studies have reported relationships between tongue strength and dysphagia in healthy older adults, acute older adult inpatients, and long-term care residents [14, 21, 38], we found that tongue strength was associated with dysphagia indices as well as sarcopenia in older adult inpatients of a rehabilitation hospital. In the multiple regression analysis with FOIS as a dependent variable, grip strength was not a significant independent factor, although it was strongly associated with MTP. It has been reported that aspiration findings from a fiberoptic endoscopic examination of swallowing correlated with isometric tongue strength, but was not associated with grip strength [21]. Consistent with this, our findings demonstrate that, relative to muscle strength, tongue strength itself is more relevant as a dysphagia-related factor.

Several study limitations are worth noting. First, as the present study employed a cross-sectional research design, causal relationships could not be determined. Second, we did not use dual-energy X-ray absorptiometry or bioelectrical impedance analysis to measure muscle mass, but instead used CC, thus diagnostic accuracy is lacking. Third, because the measurements of tongue strength were taken only at the front of the tongue, the association with tongue strength at the back of tongue is unclear. Fourth, because we did not use videofluoroscopic examination or fiberoptic endoscopic evaluation of swallowing, data related to dysphagia lacked an objective physiological measure of swallowing. Finally, because baseline factors prior to hospitalization were not measured, the degree of the effect of secondary sarcopenia on tongue strength during hospitalization is unclear.

Conclusion

The present study found that tongue strength was independently associated with muscle function (as measured by grip strength) and nutritional status (as measured by MNA-SF), both of which could be related to secondary sarcopenia, in older adult inpatients of a rehabilitation hospital. Tongue strength was also associated with dysphagia as measured by FOIS in this patient population. To maintain and improve tongue strength in association with sarcopenic dysphagia, exercise therapy and nutritional therapy interventions, as well as direct interventions to address tongue strength, may be effective in dysphagia rehabilitation in older adult inpatients. Furthermore, grip strength and MNA-SF can be used as indices of tongue strength in older adult inpatients of a rehabilitation hospital.