Preiksaitis, Harold G., and Catherine A. Mills.Coordination of breathing and swallowing: effects of bolus consistency and presentation in normal adults. J. Appl. Physiol. 81(4): 1707–1714, 1996.—Respiration and swallowing were recorded simultaneously by inductance plethysmography, submental electromyography, and a throat microphone in 10 normal subjects during eating and drinking tasks that included single boluses of varying volume (5–20 ml) and consistency presented with a syringe and cup, a 200-ml drink taken with and without the use of a straw, and a sandwich meal. Swallows were associated with a brief swallow apnea (SA) lasting ∼1 s. Swallow effects on the duration or tidal volume of the preswallow, postswallow and swallow-associated breathing cycles varied depending on bolus characteristics and presentation. Expiration before and after the SA was the preferred pattern with all drinking and eating tasks. Inspiration followed SA in <5% of single-bolus swallows, but this pattern increased significantly with a 200-ml drink administered by cup or by straw and during a sandwich meal (23.8 ± 5.2, 27.0 ± 2.6, and 16.3 ± 2.7%, respectively). Hence, the swallow-associated breathing pattern seen with single-bolus swallows may not reflect that associated with regular eating and drinking behavior. This finding implies that the risk of aspiration may be reduced by teaching patients prone to aspiration to simplify the complex behavior of eating and drinking to a series of single-bolus swallows.
the dual role of the proximal aerodigestive tract as a conduit for food and an airway has led to several anatomic and functional adaptations that serve to protect the airway during swallowing. Numerous studies have examined the interaction of breathing and swallowing (1, 9, 10, 12, 13, 15, 17-20, 22). Although methods vary among studies, several characteristics of this interaction are consistently observed. 1) The timing of swallows in relation to the respiratory cycle is not random. Rather, >70% of swallows begin during the expiratory phase of the respiratory cycle. 2) Most swallows end with expiration as well, with this tendency being enhanced by the presence of a bolus (15, 20). 3) Swallowing is associated with a brief swallow apnea (SA), which has a duration of ∼1–2 s. The consistency of these observations has led to the speculation that the precise coordination of breathing and swallowing may be an important mechanism to prevent aspiration (2, 6,9, 15, 17, 18, 22).
Eating and drinking are complex behaviors involving postural movements; specific and varied ways of introducing food and drink of varying texture, flavor, consistency, and temperature; mastication and bolus preparation; and finally deglutition itself. In contrast, with few exceptions (8, 22), the coordination of breathing and swallowing has been studied in isolation from this complex activity, usually after a fixed bolus of water is introduced into the oral cavity with a syringe. In some studies (8, 9, 15, 17, 18), the swallow bolus was introduced with a spoon, straw, or cup, but a systematic examination of the impact that the method of introducing the bolus has on the swallow-related breathing pattern has not been made. Furthermore, the effect of bolus characteristics on the swallow-associated breathing pattern has not been adequately explored. Several studies have demonstrated that the dynamics of the oropharyngeal swallow can be influenced by bolus characteristics such as volume and consistency (2, 14). The extent to which the modulating influence of bolus characteristics extends to the breathing-swallowing interaction is largely unknown. In a very brief study by Clark (1), it was reported that a single solid-bolus swallow is associated with a breathing pattern that is qualitatively similar to that of a liquid bolus. Bolus volume has been shown in some subjects to influence the pattern of swallow-associated breathing (15), whereas others have found no effect of bolus volume on the timing of this interaction for small-volume single-bolus swallows (3–20 ml) but an increase in the frequency of swallows that ended with inspiration when subjects were asked to drink 100 ml water through a straw (9). Beyond these observations, the effect of bolus characteristics on the pattern of breathing at the time of the swallow has not been explored in detail.
We hypothesized that if the consistent pattern of swallow-related breathing is an important airway protective mechanism, that pattern should be maintained regardless of bolus characteristics and method of bolus introduction and it should be maintained during normal eating behavior. The aim of the present study was to examine the coordination of breathing and swallowing in normal subjects under these various conditions.
Ten healthy subjects (6 women and 4 men) aged 19–25 yr were studied. Subjects were of average build, and none had a history of dysphagia, gastrointestinal, neurological, or pulmonary disease. All subjects provided written informed consent, and the study protocol was approved by the University of Western Ontario Review Board for Health Sciences Research Involving Human Subjects.
Recording respiration and swallowing.
Inductance plethysmography using a Respitrace Calibrator (Ambulatory Monitoring, Ardsley, NY) was used to record respiration. The oscillator was set to DC mode to avoid potential artifacts of the automatic baseline-zeroing feature of the plethysmograph. The plethysmograph was calibrated according to the manufacturer’s recommended procedure by using a fixed-volume (800-ml) Spirobag. Swallows were recorded by submental electromyography (EMG) by using a bipolar surface electrode coupled to an AC amplifier (Biocommunication Electronics, Madison, WI) and integrated on-line (time constant 60 ms) (15). A second method of recording swallows utilized a throat microphone that was coupled to a Stand Alone Medical Monitoring Interface (SAMMI, Vitalog Monitoring, Redwood City, CA). The Respitrace, integrated surface EMG, and microphone signals were digitized (50 Hz) and recorded directly to a personal computer equipped with the MP100 data-acquisition system (Biopac Systems, Goleta, CA). The system software (AcqKnowledge) was configured so that the integrated EMG, throat microphone, rib cage movement, abdominal movement, and the sum of the latter two were displayed simultaneously on separate channels.
The subjects were seated comfortably. After a 10-min period of quiet breathing to allow subjects to adjust to the recording equipment, they performed the Respitrace calibration sequence. After this, they were randomly given 5-, 10-, and 20-ml challenges of water in a cup that they manipulated themselves and by a syringe. The sequence was repeated a minimum of four times. Subjects were then given 5-ml challenges of Kool-Aid-flavored water, the consistency of which had been altered by the addition of 0, 5, 10, and 20 ml of Thickit (Quick Thick) per 100 ml water to yield consistencies characterized as thin liquid, thick liquid, syrup, and semisolid, respectively. After this, a solid bolus consisting of 2 g of arrowroot cookie was administered. This sequence was repeated twice. During both of these sequences, subjects were asked to avoid excessive body movements and to attempt to swallow each bolus in a single swallow. Only solitary nonrepeated swallows were subsequently analyzed, and piecemeal (rapid repetitive) swallows were excluded. Subjects were given a 200-ml drink of water from a cup and from a cup with a straw. Subjects were again instructed to avoid excessive body movements and were asked to drink continuously until the cup was empty. Otherwise, subjects were allowed to drink as they would under normal conditions. The drinking tasks were repeated twice in each subject. Finally, subjects ate a sandwich of their choice, being instructed to eat at a normal pace but not to put the sandwich down once eating began. Otherwise, subjects were allowed to eat as they would under normal conditions.
Data collection and analysis.
Data for swallows and breathing were collected and analyzed by using a personal-computer program (AcqKnowledge) that allowed precise measurement of respiratory intervals and volumes. Swallows were identified by the EMG signal as well as the recording of a typical double-peak swallow “sound” recorded by the throat microphone (16, 19). In addition, subjects were positioned in the room such that the occurrence of a swallow could be confirmed visually by observing for typical movement of neck structures, whereupon the computer event marker could be activated without the subject’s awareness. All measurements were made by using the output of the sum channel of the inductance plethysmograph as shown in Fig.1. The swallow apnea (SA) was measured as the total interval during which there was no detectable respiratory movement at the time of the swallow. No attempt was made to time the SA duration relative to the microphone or EMG signal because of variability of the latter two related to the type of bolus and swallowing task. The duration and tidal volume of the following intervals were determined: the respiratory cycle containing the swallow (swallow-associated breathing cycle; BCsw), the respiratory cycle immediately before the BCsw(BCpre-sw), and the respiratory cycle immediately after the BCsw(BCpost-sw). The onset of inspiration was taken as the beginning of a respiratory cycle. The duration and tidal volume of the average breathing cycle (BCavg) were measured as the mean of 10 quiet breathing cycles selected from an interval of time during which the subject was not aware that data were being collected.
Statistical comparisons were made by using SigmaStat (Jandel Scientific, San Rafael, CA). All data were treated as paired. Data sets to be compared were initially tested for normal distribution (Kolmogorov-Smirnov test) and equality of variance (Levene median test). If the data passed the normality and variance testing, Student’s t-test was used for comparisons between two data sets, and a one-way repeated-measures analysis of variance was applied when more than two samples were compared. Alternatively, if the data failed the normality or variance testing, Friedman’s repeated-measures analysis of variance on ranks was used. Post hoc testing was done using the Student-Newman-Keuls test. Dunnett’s method was used for multiple comparisons to a control value. Additional details regarding statistical test applied are provided in the legends of Figs. 2, 3, 5, and 6 and the footnotes of Tables 2, 3, and 5. Except where indicated, data are shown as means ± SE for 10 subjects.P < 0.05 was considered significant.
Table 1 shows the age and sex of the 10 study subjects and the tidal volume and duration of their average breathing cycles.
Effects of bolus volume and method of delivery on the single-bolus swallow-associated breathing pattern.
A total of 233 single-bolus water swallows were analyzed. Figure 1shows a sample record of a 5-ml water swallow. Seventy-two percent of these swallows occurred during the expiratory phase of the breathing cycle, as shown in Fig. 1, and consisted of expiratory airflow followed by a period of SA and concluded with expiratory airflow before normal respiration resumed. Table 2 summarizes the effects of varying bolus volumes administered by syringe or cup on the swallow timing relative to the breathing cycle. The swallow-associated breathing pattern was similar for syringe or cup swallows. In both cases, at a volume of 20 ml there was a tendency for more swallows to begin in inspiration, but only in the case of cup swallows was this effect significant (P = 0.03). Expiratory airflow immediately after a swallow was the preferred pattern (>92.5%) with both syringe and cup swallows regardless of bolus volume. Each swallow was associated with a brief SA lasting 0.46–1.37 s and a 32–71% prolongation in the BCsw when compared with the BCavg. Figure2 shows the duration of SA and the BCsw for both syringe and cup swallows with 5-, 10-, and 20-ml volumes of water. SA duration was similar for syringe and cup swallows at all bolus volumes. The duration of the BCsw of cup swallows was significantly longer than the duration of the BCsw of syringe swallows at all three volumes but was unaffected by bolus volume for each method of presentation (Fig. 2, Table 3). Table 3summarizes the effects of bolus volume and method of presentation on the duration and tidal volume of the BCsw, BCpre-sw, and BCpost-sw. Compared with BCavg, BCsw was prolonged for all bolus volumes administered by syringe, whereas the tidal volumes of the BCsw for 5- and 10-ml boluses were significantly less than those of the BCavg. In contrast, for each bolus volume, BCsw was more prolonged for cup swallows than for syringe swallows (Fig. 2, Table 3) and the tidal volumes of the BCsw tended to be larger as well, although not significantly different from the BCavg. A further difference in syringe and cup swallows was that the tidal volumes of both the BCpre-sw and BCpost-sw were less than the resting breath tidal volume for 5- and 10-ml boluses administered by syringe (Table 3). Beyond this, no other difference in the BCpre-sw or BCpost-sw compared with BCavg was identified.
Effect of bolus consistency on the swallow-associated breathing pattern.
Of a total of 100 swallows of varying bolus consistency, 77% began and ended with expiration. As shown in Table 4, bolus consistency had no significant impact on the respiratory phase in which swallowing occurred; both swallows started and ended with expiration as the preferred pattern across a range of consistencies. Figure 3 shows the effect of bolus consistency on SA and BCsw duration. Neither SA nor BCsw duration was significantly altered over a range of consistencies until a solid-bolus (cookie) challenge was provided. The cookie challenge produced a 55% longer SA and a 17% greater prolongation of the BCsw compared with the 5-ml thin-liquid bolus. Table 5 summarizes the effects of bolus consistency on the duration and tidal volume of the BCpre-sw, BCsw, and BCpost-sw. BCsw was similarly prolonged compared with BCavg for all bolus consistencies (Table 5). As noted with 5- to 20-ml syringe swallows, the tidal volume of the BCsw was ∼80% of the BCavg tidal volume for consistencies ranging from thin liquid to syrup but was not significantly different for the semisolid or solid bolus. Neither the duration nor tidal volume of the BCpre-sw or BCpost-sw was affected by bolus consistency.
Breathing pattern during drinking.
Examples of the breathing pattern recorded during the two drinking tasks, 200 ml of water taken directly from a cup and with the use of a straw, are shown in Fig. 4,A andB, respectively. For the cup drinking task, 188 swallows were analyzed, with 32 additional swallows being rejected because they occurred in groups associated with a single SA and were thus considered to be piecemeal swallows. For the straw drinking task, 312 swallows were analyzed, with 26 being rejected as piecemeal swallows. Figure 5 shows the timing of swallows in relation to the breathing pattern during these two drinking tasks and compares this with single-bolus swallows taken directly from a cup (5 and 10 ml) or administered by syringe. In each condition, swallow onset during the expiratory phase remained the preferred pattern, with swallows beginning in inspiration accounting for only 25–30%. However, with both 200-ml drinking tasks, the percentage of swallows that were immediately followed by inspiration was significantly increased when compared with single-bolus swallows administered by syringe or cup. The SA duration was similar for single-bolus swallows (cup and syringe) and both 200-ml drinking tasks (Fig. 6).
Breathing pattern during a sandwich meal.
A portion of the record demonstrating the breathing pattern observed during a sandwich meal is shown in Fig. 7. As shown in Fig. 5, similar to single-bolus swallows (syringe or cup) and both 200-ml drinking tasks, swallows during the sandwich meal usually began in expiration (87.2 ± 3.8%). However, as with both 200-ml drinking tasks, breathing resumed with the inspiratory phase more frequently (16.3 ± 2.7%) for swallows during the sandwich meal than for either type of single-bolus swallow (Fig. 5). The SA duration for sandwich swallows was similar to single-bolus swallows (syringe or cup) and both 200-ml drinking tasks (Fig. 6). As with the 200-ml drinking tasks, the irregularity of the breathing pattern during the sandwich meal did not permit further analysis of the BCsw or the BCpre-sw and BCpost-sw cycles.
Previous studies have demonstrated that swallowing is associated with a pattern of breathing consisting of a brief period of apnea that usually begins and ends with expiration (1, 9, 10, 12, 13, 15, 17, 18, 22). This pattern is remarkably consistent and has led to the suggestion that the precise coordination of breathing at the time of the swallow is important to protect the airway from aspiration (2, 6, 9, 15, 17,18, 22). The present study supports this contention because the pattern of expiration-SA-expiration was found to be preserved over a range of bolus volumes and consistencies, whether the bolus was introduced with a syringe, cup, or straw or during a sandwich meal. However, significant quantitative differences were observed, particularly the increase in frequency of swallows ending with inspiration during a 200-ml drink using a straw or directly from a cup and during a sandwich meal. These findings suggest that single-bolus swallows, which have been most frequently studied, may not be representative of the swallow-related breathing pattern occurring during normal eating and drinking behavior. Assuming that normal subjects tend to favor timing swallows such that expiratory airflow follows directly after the SA because this pattern is inherently safer and avoids aspiration, the present data suggest that the risk of aspiration during some eating behaviors may be increased because of the increased frequency of swallows ending with inspiration.
Without exception, previous studies have demonstrated that most swallows in a variety of conditions begin and end in expiration. The three remaining swallow-associated breathing patterns occur with lesser and variable frequency. Those that begin during inspiration and end in expiration are the next most frequent, being reported in about one-fifth of all swallows. Swallows ending with inspiration are least frequent, and those that begin and end with inspiration are rare, accounting for no more than 3% of all swallows in normal subjects. These latter two patterns are of particular interest because of the theoretically increased risk of aspirating any bolus residue that may not have been completely cleared during the swallow when respiration resumes with inspiration. Nishino and Hiraga (12) demonstrated that the frequency of swallows that are followed by inspiration is increased when the level of consciousness is suppressed, as during recovery after general anesthesia. They speculated that failure of the normal coordinating mechanisms may be one factor that contributes to the increased risk of aspiration when consciousness is impaired. Their observations also emphasize the importance of a central nervous system mechanism in controlling this interaction (11). On the other hand, at least two previous studies have demonstrated a decrease in the frequency of swallows that are followed by inspiration if a bolus is present (15, 20). This observation demonstrates that sensory information about the presence of a bolus has the ability to feed back to the central control mechanism and modulate the timing interaction. The putative pathways that are involved in the coordination of swallow-breathing events are likely affected in certain neuropathological conditions. Thus breathing and swallowing in a group of adult patients with neurological disorders that were associated with an increased risk of aspiration demonstrated an increased prevalence of swallows that were immediately followed by inspiration (18). A similar tendency was observed in children with cerebral palsy (5).
A second characteristic of the breathing-swallowing interaction that may be relevant to aspiration is the apneic period that occurs during the swallow itself. One could envision that the premature resumption of respiration before the swallow bolus has been adequately cleared might result in aspiration. The importance of this factor is emphasized by studies that have demonstrated that the duration of glottic closure during the swallow does not correspond completely with the SA period and that the coordination of SA and vocal cord kinetics may be further impaired in patients with a tracheostomy (21). Although previous studies have demonstrated that the bolus volume or consistency can influence the dynamics of the oropharyngeal swallow (2, 14), we did not find that the duration of SA was significantly affected by either of these factors, which in this study were the method of introducing the bolus (syringe vs. cup) or eating behavior (200-ml drink or sandwich). A slight prolongation of the SA associated with a cookie swallow when compared with a 5-ml water bolus was identified, but this is of doubtful clinical significance. This finding is at variance with a previous study in normal subjects where 50% were found to increase the duration of SA as bolus volume increased (15). Reasons for this discrepancy may be in part methodological because in the previous study (15) SA duration was defined based on nasal airflow rather than on a measurement of abdominal or chest wall movement, and the possible occurrence of mouth breathing, especially during the introduction of the bolus, cannot be completely ruled out. A second possibility may relate to the subjects themselves, because 3 of 10 of our subjects also showed a pattern of increase in SA duration with bolus volume, but overall this effect was not significant.
In agreement with previous studies (12, 13, 15), we found that the most significant swallow-induced alterations in the breathing pattern were observed in the breathing cycle that contained the swallow. The total duration of the BCsw was increased relative to BCavg for all boluses tested, whereas the tidal volume of the BCsw was decreased by ∼20% for smaller volume boluses (5–10 ml) administered by syringe. On the other hand, no significant effects on the duration of the breath immediately before and immediately after the BCsw were identified, and an effect on the tidal volume of these breathing cycles was apparent only for 5- or 10-ml boluses administered by syringe. In previous studies, different effects of the swallow on the BCpre-sw and BCpost-sw breaths have been observed. Nishino et al. (13) found a significant increase in the tidal volume of the postswallow breath. However, subjects in that study were fitted with a mouthpiece and noseclips, and swallows were induced by the injection of 1 ml of water via the mouthpiece and were recorded with the subjects in a supine posture. These encumbrances are likely to interfere with the accurate reproduction of the conditions of a normal swallow because they can influence respiration (3) and potentially the interaction of breathing and swallowing as well (10). On the other hand, in intubated unconscious patients recovering from general anesthesia, Nishino and Hiraga (12) found that neither the duration nor tidal volume of the postswallow breath was different from control breaths. In another study (15), in which respiration was monitored by nasal airflow, a slight decrease in the duration of the preswallow breath was noted in the presence of a bolus. Although nasal airflow is known to accompany oral airflow (19), the use of inductance plethysmography, as in the present study, may provide a more accurate and unobtrusive means to monitor breathing related to swallowing. Thus methodological differences may account for these discrepancies.
The most important observation in the present study is the increased occurrence of swallows that end with inspiration during more complex eating and drinking behavior as compared with single-bolus swallows. On average, single-bolus swallows were followed by inspiration in <5%, whereas this pattern was seen in 16% of swallows recorded while subjects fed themselves a sandwich and 18–26% of swallows recorded during a 200-ml drink from a cup or with a straw. Smith et al. (22), in a similar study also using inductance plethysmography to monitor the breathing pattern of seven male subjects while an assistant fed them two donuts and two glasses of water, also reported that most swallows begin and end during expiration. Their reported results implied that swallows ending with inspiration were rare, but precise quantitative data were not provided. In contrast to the present study, subjects were fed by an assistant, and no comparison is made of swallows during drinking vs. solid-bolus syringe swallows or single-bolus cup swallows. In another study, Martin et al. (9) examined the swallow-associated respiratory pattern also using inductance plethysmography and compared single-bolus swallows of 3, 10, and 20 ml, delivered by using a syringe, with 100 ml water, administered with a straw. Their findings were similar to those of the present study: the majority of single-bolus swallows begin in expiration, inspiration after SA was never seen, and bolus volume had no impact on these parameters or on SA duration. On the other hand, Martin et al. found that during the straw drinking task, the majority of their subjects took multiple swallows uninterrupted by respiration, resulting in a prolonged SA. Nonetheless, they too noted an increase in the occurrence of inspiration after SA with the use of straw. The mechanism underlying the difference in the respiratory pattern associated with single-bolus swallows and more complex eating and drinking tasks is not known.
The majority of our subjects, unlike the subjects in the study of Martin et al., did not demonstrate a pattern of prolonged apnea during which multiple swallows occurred with the large-volume swallow task. This difference underscores one of the unavoidable difficulties encountered in trying to study complex behaviors such as eating and swallowing without significantly altering them in the laboratory setting. Although we gave minimal instructions to our subjects and attempted to influence as little as possible their execution of the swallowing tasks, we cannot rule out the possibility that subjects may respond to subtle inadvertent cues and alter their behavior on that basis. Factors such as fatigue, thirst, distractibility, or the order of drinking and eating tasks could potentially influence the outcome.
Despite these limitations of currently available methods for evaluating the interaction of breathing and swallowing, the results of this study demonstrate that the swallow-associated breathing pattern observed with single-bolus swallows may not predict the pattern seen during normal eating and drinking behavior. The mechanism underlying the difference in the respiratory pattern associated with single-bolus swallows and more complex eating and drinking tasks is not known. We speculate that this could be due to mental distraction or a loss of focus on the swallowing task that may occur during regular eating and drinking and is less likely to occur with a controlled single-bolus swallow. Further studies would be required to assess this possibility and the potential clinical implications. For example, these observations provide an additional rationale for adopting measures that reduce complex eating behavior to a series of single-bolus swallows in subjects prone to aspiration (4, 7). On the other hand, the use of aids such as a simple drinking straw in swallowing-impaired patients needs to be reconsidered in light of the theoretically increased risk for aspiration that may result from modifying the swallow-associated breathing pattern. Whether these theoretical considerations are of practical importance in the clinical setting now requires similar studies to be done in patient groups at increased risk for aspiration.
This work was supported by the Physicians Services Incorporated Foundation. H. G. Preiksaitis is a recipient of an Ontario Ministry of Health Career Scientist Award.
Address for reprint requests: H. G. Preiksaitis, Dept. of Medicine, St. Joseph’s Health Centre, 268 Grosvenor St., London, Ontario N6A 4V2, Canada.
- Copyright © 1996 the American Physiological Society