Number of medications

Figure 2. Percentage of medications used **Figure 2.** Percentage of medications used


Category β SE OR 95% CI p

**Table 4.** The association between polypharmacy and dysphagia 1–4 drugs 1.144 0.491 2.842 (1.200, 8.223) 0.020

#### **3.5. Predictor interaction and impact on dysphagia** ≥ 5 drugs 1.156 0.490 3.176 (1.216, 8.299) 0.018

The impact of predictor interaction on the incidence of dysphagia according to binary logistic regression showed gender and IHD to be two significant predictors. Patients taking 5 medi‐ cations or more (i.e. polypharmacy) were more susceptible to the incidence of dysphagia, especially female patients and those with IHD.

Female cardiac patients taking 5 or more medications were about 2.2 times more prone to a high incidence of dysphagia than males taking no medications, while there was no significant impact on females taking fewer than 5 medications. This proved the impact polypharmacy has on females complaining of dysphagia (as shown in Table 5).

IHD patients taking 5 medications or more had a significantly higher (about 1.9 times) incidence of dysphagia than those free of IHD and taking no medications. However, no impact was found for IHD patients taking less than 5 medications (as shown in Table 6).

These tables show the impact of polypharmacy is going to increase the incidence of dysphagia when the interactions between predictors are taken into consideration. However, gender (female) was a higher predictor effect for dysphagia than the IHD predictor.


The reference category is males taking no medications

The reference category for the model is no dysphagia

**Table 5.** Impact of polypharmacy and gender on dysphagia


The reference category is no medications and no IHD

The reference category for the model is no dysphagia

outpatients. The incidence of dysphagia was about 2.8 and 3.2 times higher for patients taking 1–4 drugs and those **Table 6.** Impact of polypharmacy and IHD on dysphagia

#### taking ≥ 5 drugs (polypharmacy), respectively, than those taking no medications. However, the incidence of dysphagia in patients with polypharmacy was found to be higher than those taking fewer than 5 medications (as shown in Table 4). **3.6. Polypharmacy and type of dysphagia**

Dysphagia was classified according to symptoms of cardiac patients taking part in the study. Chest pain is the only symptom that showed a significant association with polypharmacy. Patients taking 5 or more medications had a significantly higher incidence of chest pain (about 2.1 times) than those without medications. Moreover, no significant effect was found for chest pain in patients taking fewer than 5 medications. Thus, polypharmacy has a greater effect on esophageal dysphagia than oropharyngeal dysphagia (as shown in Table 7).

#### **4. Discussion**

**Category** *β SE OR 95% CI p* 1–4 drugs 1.144 0.491 2.842 (1.200, 8.223) 0.020 ≥ 5 drugs 1.156 0.490 3.176 (1.216, 8.299) 0.018

No medication 1-4 drugs ≥5 drugs

Number of medications

Gender, IHD, and statins were the most significant factors that must be involved in the regression model to insure the

Gender 6.181 1 0.013 IHD 7.909 1 0.005 Statins 4.539 1 0.033

As a result of logistic regression, gender and IHD were found to be the significant risk factors involved in the high incidence of dysphagia. Female cardiac patients had incidences of dysphagia that were approximately 1.8 times higher than those of males. Patients with IHD had incidences of dysphagia that were 1.8 times higher than those without (as

> Variable β SE OR 95% CI p Gender Female 0.575 0.233 1.777 (1.148, 2.750) 0.010

> IHD Yes 0.599 0.207 1.820 (1.212, 2.731) 0.004

The reference category for the model is no dysphagia. The backward stepwise logistic regression test used was the

Patients with polypharmacy (i.e. those using 5 or more medications) have a higher incidence (45.84%) of dysphagia than

43.92% 45.84%

Binary logistic regression showed that medication use was a risk factor in the incidence of dysphagia in cardiac

Category β SE OR 95% CI p 1–4 drugs 1.144 0.491 2.842 (1.200, 8.223) 0.020 ≥ 5 drugs 1.156 0.490 3.176 (1.216, 8.299) 0.018

Categorical variables χ<sup>2</sup> df p

predictors of dysphagia could be determined (as shown in Table 2).

χ<sup>2</sup> = chi-square test; df = degrees of freedom; p = calculated probability

Male (ref.)

No (ref.)

Hosmer and Lemeshow goodness-of-fit test with χ<sup>2</sup> (N = 576) = 3.365 and p =0.186

Table 3. Predictors of dysphagia in cardiac outpatients

10.24%

other patients (as shown in Figure 2).

3.4. Polypharmacy and its impact on dysphagia

Table 2. Categorical variables included in the regression model

shown in Table 3).

The impact of predictor interaction on the incidence of dysphagia according to binary logistic regression showed gender and IHD to be two significant predictors. Patients taking 5 medi‐ cations or more (i.e. polypharmacy) were more susceptible to the incidence of dysphagia,

Female cardiac patients taking 5 or more medications were about 2.2 times more prone to a high incidence of dysphagia than males taking no medications, while there was no significant impact on females taking fewer than 5 medications. This proved the impact polypharmacy has

IHD patients taking 5 medications or more had a significantly higher (about 1.9 times) incidence of dysphagia than those free of IHD and taking no medications. However, no impact

These tables show the impact of polypharmacy is going to increase the incidence of dysphagia when the interactions between predictors are taken into consideration. However, gender

was found for IHD patients taking less than 5 medications (as shown in Table 6).

(female) was a higher predictor effect for dysphagia than the IHD predictor.

No medications (ref.)

**Figure 2.** Percentage of medications used

The reference category for the model is no dysphagia.

0%

10%

Percentage of patients

46 Seminars in Dysphagia

20%

30%

40%

50%

**Table 4.** The association between polypharmacy and dysphagia

Figure 2. Percentage of medications used

**3.5. Predictor interaction and impact on dysphagia**

especially female patients and those with IHD.

on females complaining of dysphagia (as shown in Table 5).

Many different results have been reported on the incidence of dysphagia in different areas of the world as a consequence of the number of diseases and medications that bring it about. Moreover, some studies have restricted themselves to different age groups; for example, some relate to childhood dysphagia while others relate to geriatrics [13]. Many physicians fail to take these symptoms into account either because they do not take dysphagia seriously or are unfamiliar with the factors that bring it about [14]. Siebens *et al* [15] and Croghan *et al* [16] found that morbidity and mortality were significantly higher in those with dysphagia than those without because of malnutrition and/or low quality of life [17]. Speyer *et al* [18] and Wallace *et al* [19] found that patients' self-reporting was the most effective tool for identifying dysphagia symptoms. This led to many studies being conducted on patients self-reporting with the aim of determining the incidence of dysphagia [20–22]. This was because patients were considered the main source of information to get at the data needed to conduct clinical


The reference category is no dysphagia.

**Table 7.** Association between polypharmacy and type of dysphagia

studies. Unfortunately, very few studies have reported on the risk factors relating to dyspha‐ gia.

There are a number of benefits stemming from the current study: first, establishing a new method to count the incidence of dysphagia by getting cardiac outpatients to fill in a validated, acceptable, and feasible questionnaire, especially because until now there has been no standard validated tool to report dysphagia and its symptoms capable of meeting clinical requirements [23]. Second, this study can be used to determine the type of dysphagia based on the types of symptoms by statistically correlating them into oropharyngeal and esophageal types; up until now all previous studies either depended on one type or using scales for the classification. Moreover, no study has ever been assigned to a specific clinical case such as the one here for cardiovascular diseases or investigated the interactions of dysphagia predictors.

The survey carried out by Barczi *et al* [24] stated the incidence of adult patients complaining of dysphagia ranged from 10 to 30%. The dysphagia incidence (21.2%) of the present study was in the normal range, which was considered to be a good level of incidence when other risks were taken into consideration; for example, the subjects involved in the current study were elderly cardiac patients complaining of serious diseases and using chronic multiple therapies. There is variance in the incidence of dysphagia symptoms such as cough, chest pain, and indigestion. Such differences have also been reported in previous studies [25, 26]. The reason for such differences is because the diseases affecting patients and the medications they take differ from one patient to another. Compared with other studies, cough had a higher incidence of dysphagia in the current study (46.7%), which was higher than incidence of dysphagia, and this case was similar to the compared results of Eslick *et al* [27] who reported dysphagia (16%), cough (27%) and chest pain (23%). Their study was based on different explanations of different mechanisms of dysphagia [28, 29]. The incidence of mild, moderate, and severe dysphagia in the present study was 71.31, 20.49, and 8.20%, respectively, which was similar to the results of Eslick *et al* who reported 65, 30, and 5% for mild, moderate, and severe, respectively [27].

Dobrzycki *et al* found a significant association between IHDs and dysphagia, because the shifting of parasympathetic levels increases the incidence of gastric reflex and induces cardiac problems [30]. Similarly, cardiac patients in the current study considered IHDs to be the main risk factors of dysphagia, where the incidence of dysphagia increased approximately 1.8 times in patients complaining of IHDs. Alves *et al* found gender had a significant impact on the incidence of dysphagia by measuring swallowing parameters such as velocity, intervals, number, and volume capacity. The velocity at which females swallow was found to be slower than for males, the volume capacity of females was found to be less than males, and females were found to need more time to swallow [31]. There are two reasons for this. First, males have larger oral and pharyngeal cavities than females; hence, they find it easier to swallow. Second, some studies have reported that it takes longer for the esophageal sphincter to open in females than in males [32, 33]. However, the results of the current study are in agreement with previous studies regarding the relationship between gender and dysphagia; for example, there is a higher incidence (approximately 1.8 times) of dysphagia in females than in males.

The reason polypharmacy can be considered a significant risk factor to the high incidence of mortalities, morbidities, and serious adverse reactions, is because drug interactions increase the potential toxicity of medications, especially in elderly patients [34]. The incidence of polypharmacy detected varies widely between studies (22–82%) [35–39] as a result not only of the way in which medications are prescribed in different countries but also awareness about the risks of medications. The incidence of polypharmacy reported in the present study was considered good (45.84%) when compared with other studies. Moreover, none of these studies compared the incidence of dysphagia in healthy individuals and ill patients. Some patients in the current study were not taking any medications because they had came from other clinics for checking purposes only, making them a good standard group for comparison with those with polypharmacy and those without polypharmacy.

studies. Unfortunately, very few studies have reported on the risk factors relating to dyspha‐

**Indigestion** *β SE OR 95% CI p* 1–4 drugs -0.099 0.315 0.906 (0.488, 1.681) 0.754 ≥ 5 drugs 0.130 0.311 1.139 (0.619, 2,079) 0.676

1–4 drugs 0.040 0.291 1.041 (0.588, 1.842) 0.890 ≥ 5 drugs 0.132 0.290 1.141 (0.647, 2.014) 0.648

1–4 drugs 0.369 0.297 1.447 (0.808, 2.591) 0.214 ≥ 5 drugs 0.748 0.296 2.113 (1.182, 3.777) 0.012

There are a number of benefits stemming from the current study: first, establishing a new method to count the incidence of dysphagia by getting cardiac outpatients to fill in a validated, acceptable, and feasible questionnaire, especially because until now there has been no standard validated tool to report dysphagia and its symptoms capable of meeting clinical requirements [23]. Second, this study can be used to determine the type of dysphagia based on the types of symptoms by statistically correlating them into oropharyngeal and esophageal types; up until now all previous studies either depended on one type or using scales for the classification. Moreover, no study has ever been assigned to a specific clinical case such as the one here for

The survey carried out by Barczi *et al* [24] stated the incidence of adult patients complaining of dysphagia ranged from 10 to 30%. The dysphagia incidence (21.2%) of the present study was in the normal range, which was considered to be a good level of incidence when other risks were taken into consideration; for example, the subjects involved in the current study were elderly cardiac patients complaining of serious diseases and using chronic multiple therapies. There is variance in the incidence of dysphagia symptoms such as cough, chest pain, and indigestion. Such differences have also been reported in previous studies [25, 26]. The reason for such differences is because the diseases affecting patients and the medications they take differ from one patient to another. Compared with other studies, cough had a higher incidence of dysphagia in the current study (46.7%), which was higher than incidence of dysphagia, and this case was similar to the compared results of Eslick *et al* [27] who reported dysphagia (16%), cough (27%) and chest pain (23%). Their study was based on different explanations of different mechanisms of dysphagia [28, 29]. The incidence of mild, moderate, and severe dysphagia in the present study was 71.31, 20.49, and 8.20%, respectively, which was similar to the results of Eslick *et al* who reported 65, 30, and 5% for mild, moderate, and

cardiovascular diseases or investigated the interactions of dysphagia predictors.

gia.

No medications (ref.) **Cough**

48 Seminars in Dysphagia

No medications (ref.) **Chest pain**

No medications (ref.) The reference category is no dysphagia.

**Table 7.** Association between polypharmacy and type of dysphagia

severe, respectively [27].

Previous studies have found that polypharmacy has an effect on the incidence of dysphagia and swallowing problems. However, these studies either were reports focused on types of dosage forms, or conducted at community pharmacies for primary care patients [4, 40, 41]. The present study has demonstrated significant clinical outcomes for the relationship between polypharmacy and dysphagia, and provided evidence to show that number of medications elevates the incidence of dysphagia in cardiac outpatients. In addition to these results, the interactions of predictors were also investigated with significant positive outcomes. Females with polypharmacy had a higher incidence of dysphagia than females without polypharmacy, due to females with polypharmacy being more susceptible to adverse reactions of medications [42]. Patients complaining of IHDs had a high incidence of polypharmacy, which elevates the incidence of adverse drug reactions including dysphagia [43]. Thus, the present study has shown that the incidence of patients with polypharmacy complaining of IHDs and reporting dysphagia is high. The present study has satisfied theories about the interactions between predictors and their impact on the incidence of adverse reactions. Few studies have investi‐ gated the synergistic effect of predictors on the incidence of dysphagia, which gives the current study importance in providing a new clinical viewpoint.

A final novel result concerns the classification of dysphagia induced by polypharmacy in cardiac outpatients. There has yet to be a study determining the effects of polypharmacy on the incidence of dysphagia, let alone the type of dysphagia. The present study found that chest pain had a greater association with polypharmacy than other symptoms. By correlating the symptoms of dysphagia, the study found that polypharmacy is likely to induce a higher incidence of esophageal dysphagia due to the significant irritation (e.g. of the mucosa) of medications and physiological changes with aging. Despite being unable to pinpoint a specific medication as the main causative agent for inducing a high incidence of dysphagia, the cumulative impact of polypharmacy was a prime candidate. Therefore, the current study suggests that the effect of total number of medications (polypharmacy) is greater than the effect of the medication itself, possibly due to the interactions and adverse reactions of medications.
