INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

VENTILATORY AND LOCOMOTOR coupling, termed "entrainment," has been observed in humans performing rhythmic activities such as running, cycling, and rowing (1, 2, 3, 5). Rowers ranging in experience from novice to elite have been shown to entrain breathing to integer or subinteger multiples of stroke rate, including 1:1, 1.5:1, 2:1, and 3:1 (10, 13).

Rowing experience and level of training appear to be related to entrained breathing. A majority (78%) of elite rowers exhibited one of two entrainment patterns, either 1:1 or 2:1, whereas a lower proportion (30%) of untrained rowers was observed to entrain in the same patterns (11). Rowers entrained at 1:1 expired during the drive phase and inspired during the recovery phase (11, 13). Rowers entrained at 2:1 tended to inspire at or just before catch (the start of the rowing stroke) and finish (end of the drive or power portion of the stroke) (11). Mahler et al. (10) reported that novice rowers naturally developed a 2:1 entrainment pattern after an 8-mo training program that did not focus on breathing patterns.

Rowers may elect to couple ventilation to locomotion to improve athletic performance (8). Cunningham et al. (4) speculated that, at catch, the body is in a cramped position with both knees and hips flexed. Increased intra-abdominal pressure in this position may impair downward excursion of the diaphragm and therefore inspiration. Conversely, during the drive phase of the rowing stroke, the knees and hips extend and inspiration may be assisted. To date, entrainment in rowers has been studied primarily by using incremental exercise tests. Although Mahler et al. (9) showed no statistical differences in maximal physiological parameters between a 6-min "all-out" test and a progressive incremental test, the rower's goal during these two tests differs. Rowers train for and compete at a 2,000-m distance, and their primary goal is to minimize time while remaining synchronized with their teammates. In progressive-interval tests, the goal is to maintain a specific power output, and incremental increases in power are often achieved by increasing stroke rate (13).

The purpose of this descriptive study was to examine ventilation and locomotion entrainment in varsity rowers during a simulated 2,000-m race. On the basis of earlier studies (10), some variability in entrainment pattern was expected. With entrained rowers, the specific pattern of entrainment was explored, whereas with unentrained rowers the relationship between ventilation and locomotion rates was examined for insight into how locomotion affected ventilation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. The sample population consisted of 11 experienced male oarsmen (scullers and sweepers) from the University of British Columbia men's heavyweight and lightweight 8-man rowing teams. Physical characteristics of the participants are shown in Table 1. All participants were regularly practicing with the rowing team and had 7-96 mo (median = 18 mo) of rowing experience. Informed consent was obtained from all subjects, and experimental procedures were approved by the University Clinical Research Ethics Board.

Experimental protocol. Subjects were asked to refrain from ingesting food or caffeine 3 h before testing. Static pulmonary function was measured, and resting flow-volume loops were acquired in each subject in standing, upright-seated, and catch positions on the rowing ergometer. After a self-regulated warm-up of 10 min and a break of ~3 min, each subject performed a simulated 2,000-m race on a rowing ergometer (Concept II, Morrisville, VT). The fan damper on the ergometer was set at position four. All subjects were highly motivated and familiar with this type of ergometer. All data on the ergometer's display were available.

Instrumentation. Ventilatory parameters, including ventilation frequency, tidal volume (VT), and minute ventilation (E) were recorded for the duration of the test by using a pneumotachograph and nose clamp (MedGraphics CPX-D Metabolic Cart, St. Paul, MN). The pneumotachograph was calibrated with a 3-liter syringe at various expected flow rates before each test. Measured gas-exchange parameters included volume of oxygen uptake (O₂), volume of expired carbon dioxide (CO₂), and partial pressure of end-tidal oxygen (PET_O2) and carbon dioxide (PET_CO2). Carbon dioxide and oxygen analyzers were calibrated with gases of known concentration before each test. Average ventilatory and gas-exchange parameters were recorded every 5 s as was average heart rate (Polar Vantage XL). An oximeter (Ohmeda Biox 3740) attached to the earlobe measured the percentage of hemoglobin saturated with oxygen (%Sa_O2). A vasodilator nicotine cream (Finalgon, Boehringer Ingelheim) was applied to the earlobe to improve perfusion for the oximeter clip.

Data analysis. Maximal physiological variables were determined in each subject and averaged. Average physiological measures were calculated from stabilized data (the first 120 s of each subject's data were excluded). Breath-by-breath volumes were calculated by integrating the ventilatory flow. Expiration was defined as positive flow or volume and inspiration as negative.

Statistics. Differences between pretest spirometry in the standing, seated, and catch positions were assessed by using a repeated-measures ANOVA. Post hoc Tukey tests were conducted to determine which conditions were significantly different. The level of significance was set at P < 0.05 for all tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All rowers completed the simulated 2,000-m ergometer test. Average completion time was 400 ± 20 (SD) s, with 199 ± 13 strokes and 365 ± 98 breaths. Pretest spirometry showed that peak flow rates in the standing (10.2 ± 1.7 l/s), sitting (10.3 ± 1.0 l/s), and catch (10.0 ± 1.6 l/s) positions were not significantly different (P = 0.91). However, there were significant differences (P < 0.001) in the maximal volume expired between the standing (5.5 ± 0.8 liters), sitting (5.5 ± 0.8 liters), and catch (5.2 ± 0.8 liters) positions. Post hoc tests revealed that catch volumes were less than both standing and sitting volumes.

Across all subjects, O₂ typically reached a plateau ~60 s into the test. The maximum respiratory exchange ratio exceeded 1.15 in 10 subjects, and peak heart rate exceeded 90% of maximum predicted heart rate (220  age) in 8 subjects, indicating that this represented a maximal test for these athletes (Table 2). The PET_CO2 typically peaked ~120 s into a test and then gradually fell for the remainder of the test. Ventilation rates also increased rapidly over the first 120 s and then increased more slowly to a maximum at the end of the test. Sa_O2 levels fell to 92% or less in eight subjects.

All rowers began their test with four to six powerful strokes and then settled into a stable rowing pattern (Fig. 1). Peak force, speed, power, and work per stroke (Table 3) then remained relatively constant until the final sprint. Ignoring the first 15 s of each test, rowers spent 44 ± 3% of each stroke in the drive phase. The SD of stroke-propulsive measures (peak force, peak power, and work per stroke) varied between 15 and 22% of their means, whereas measures of temporal consistency (stroke rate, peak chain speed, and proportion of time spent in the drive phase) varied between 5 and 7% of their respective means.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Representative force, handle speed, ventilatory (Vent) flow, and power output data for first 15 s, middle 30 s (beginning at t = 200 s), and last 15 s of test. Rower (subject 4) was entrained at a 1:1 breath-to-stroke rate for 1st portion of test and then moved into and maintained a 2:1 pattern for duration of test.

All rowers entrained at some portion of their test, although only eight rowers maintained ventilatory and locomotor entrainment for a continuous period of 120 s or more (Fig. 2). Three stable entrainment patterns were observed. One rower (subject 3) maintained a 1:1 entrainment pattern, six rowers maintained a 2:1 entrainment pattern, and one rower (subject 10) maintained a 3:1 entrainment pattern. No stable entrainment patterns were observed at subinteger multiples. Only one rower (subject 1) remained entrained (2:1) for his entire test. Two others (subjects 5 and 10) started to entrain within 30 s and then continued to be entrained for the remainder of the test.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Periods during which subjects were entrained. Thin lines, 1:1 entrainment pattern; medium lines, 2:1 pattern; heavy line, 3:1 pattern; +, time to test completion.

Sixty-two percent of all 10-s intervals (total = 446) were entrained, and 61% of all breaths (total = 3,360 breaths) occurred within these entrained intervals. For all breaths initiated in the entrained intervals, inspiration occurred most frequently during the first 40% of recovery, followed by expiration during the latter part of recovery (Fig. 3A). For all breaths initiated in unentrained intervals, a similar, although less well-defined, pattern of inhalation onset was observed (Fig. 3B). When normalized volume was superimposed onto normalized drive and/or recovery data, a similar pattern between ventilation volume and rowing cycle was observed in both entrained and unentrained intervals (Fig. 3, C and D). Inspiratory volumes were observed to be ~25% smaller for breaths initiated during middrive (0.15-0.35 stroke proportion) than throughout the remainder of the stroke. Conversely, expiratory volumes were observed to be ~17% smaller for breaths initiated during early recovery (0.4-0.6 stroke proportion) than throughout the remainder of the stroke. The distribution of the time of peak flow in the rowing cycle (Fig. 4, A and B) was similar to that observed in the ventilation-onset data, except for markedly reduced frequency of inspiratory peak flows at stroke finish in both the entrained and unentrained breaths. Peak expiratory flows were ~12% smaller in the early recovery data (0.45-0.65 stroke proportion) across both entrained and unentrained breaths (Fig. 4, C and D).

View larger version (37K): [in this window] [in a new window]   Fig. 3.   Distribution of normalized onset (A and B) and normalized volume (C and D) of expiration (defined as positive flow or volume) and inspiration (defined as negative flow or volume) for all breaths during entrained (A and C) and unentrained (B and D) intervals for all subjects (except subject 10, who entrained at 3:1). Horizontal axis in A-D, normalized stroke; 0, catch; dashed vertical line at 0.44, mean transition from drive to recovery, i.e., finish; 1, end of recovery (catch for next stroke). Histograms depict percentage of expirations and inspirations initiated during drive and recovery phases in entrained (A) and unentrained (B) portions of rowing stroke (here arbitrarily divided into 50 intervals). Thick lines in A and B, difference between expiration and inspiration data.

View larger version (40K): [in this window] [in a new window]   Fig. 4.   Distribution of normalized time to peak and normalized peak expiratory and inspiratory flow rates for all breaths during entrained (A and C) and unentrained (B and D) intervals for all subjects (except subject 10, who entrained at 3:1). See Fig. 3 for a detailed description.

It was observed that subjects who entrained at either two or three breaths per stroke modulated their inspired or expired volume during a stroke by using two dominant strategies. An example of the first dominant strategy is shown in Fig. 5 and consisted of alternating both ventilation rates (Fig. 5A) and expired volumes (Fig. 5B). Figure 5, B and D, shows that inspired volumes over the last two-thirds of this subject's test remained relatively constant (2.4 liters in early recovery and 2.7 liters in late recovery), whereas expired volumes alternated between small (2.1 liters) during recovery and large (3.2 liters) during drive. To achieve this alternating volumetric expiratory pattern, this subject used short-duration breaths during recovery (average IVR of 70 min) and long-duration breaths during drive (average IVR of 54 min) (Fig. 5A). This pattern was achieved while similar peak flow rates were maintained (Fig. 5B). Furthermore, this pattern was visible when flow-volume loops at any point in the last two-thirds of this test were examined (Fig. 5C).

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Data from an entrained subject who modulated expiratory volume when entrained at a 2:1 ratio by alternating between 2 instantaneous ventilation rates. A: instantaneous stroke and ventilation rates. B: expiratory and inspiratory peak flow and volume. C: flow-volume loops for 4 breaths taken at t = 300 s superimposed on subject's pretest seated flow-volume loop. D: absolute expired and inspired volume vs. normalized stroke.

An example of the second dominant strategy is shown in Fig. 6. This strategy achieved less-distinct alternating expiratory volumes by maintaining a regular IVR (Fig. 6A) and alternating between high peak expiratory flow rates (10.9 l/s) during the drive and low peak expiratory flow rates (7.9 l/s) during recovery (Fig. 6, B-D).

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Representative data from an entrained subject who modulated expiratory volume (less distinctly than did subject in Fig. 5) by alternating between 2 peak expiratory flow rates during drive phase while maintaining a regular ventilation rate. See Fig. 5 for a detailed description. Note that absolute flow rate vs. normalized stroke is shown (D).

Two subjects who were initially entrained at 1:1 did not maintain this ratio for the duration of their tests. One subject maintained a near-constant stroke rate and adapted his ventilation rate to meet the demands of the task (Fig. 7A). This subject achieved high levels of PET_CO2 (Fig. 7B) before abandoning 1:1 entrainment 240 s into his test. Despite decreased VT values after the subject became unentrained (Fig. 7C), E increased (Fig. 7D) and PET_CO2 decreased (Fig. 7B). The second subject maintained 1:1 entrainment for ~50 s and then entered a 120-s transition period before settling into a 2:1 breathing pattern for the remainder of the test (Fig. 8A). Although PET_CO2 (Fig. 8B) and E (Fig. 8C) remained relatively unaffected, this transition period altered the breath-by-breath VT (Fig. 8D).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Instantaneous stroke and ventilation rate (A) and physiological measures of partial pressure of end-tidal CO2 (PETCO2; B), tidal volume (VT; C), and minute ventilation (E; D) in 1 subject. Subject maintained stable 1:1 entrainment pattern for over one-half of test, after which E and stroke rate uncoupled. Note that PETCO2 decreased, E increased, and VT decreased after entrainment was abandoned.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Instantaneous stroke and ventilation rate (A) and physiological measures of PETCO2 (B), E (C), and VT (D) in 1 subject who undergoes a transition from 1:1 to 2:1 entrainment. VT and PETCO2 changed during transition, yet E remained relatively unaffected.

The kinematic data were examined graphically for differences between the entrained and unentrained intervals. No marked differences were found, i.e., the kinematics could not be used to discriminate entrainment. Peak force, peak power, and work per stroke were similarly unaffected by entrainment. Hip joint angle (trunk segment relative to the thigh segment) increased to a maximum at stroke finish, decreased only slightly during early recovery, and then decreased more rapidly to a minimum at catch (Fig. 9). Angular acceleration of the hip joint (Fig. 9) peaked at the end of the drive phase, consistent with the rowers using the rearward inertia of the torso to generate the final portion of the drive force. A secondary angular acceleration peak was observed in the early recovery phase after rowing force had dropped to zero (Fig. 9).

View larger version (21K): [in this window] [in a new window]   Fig. 9.   Mean (±SD) of hip angle and angular acceleration for 3 consecutive strokes at six 1-min intervals (18 strokes in total) by 1 subject. Solid horizontal line, 0; dashed vertical line, stroke finish, i.e., point in time when rowing force drops to 0.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first study that has described in detail the relationships among stroke rate, kinematics, and breathing frequency in well-trained athletes during a competitive simulation. All subjects in this study entrained for some portion of their test, although the duration and pattern of entrainment varied among subjects. Only integral patterns of entrainment (1:1, 2:1, and 3:1) were observed (Fig. 2). Among periods of entrainment, different subjects modulated different aspects of their ventilation. For example, subjects who alternated between two ventilatory volumes did so by either alternating between short and long breaths (Figs. 5) or alternating between low and high peak flow rates (Fig. 6).

Despite variations among subjects, the present data indicated that breaths entrained at 2:1 occurred at similar times in the stroke cycle for most rowers (Fig. 3A). This temporal pattern of 2:1 entrained breathing was consistent with previous observations in elite rowers (10). Variations in the inspired and expired volumes of unentrained breaths initiated at different times in the stroke cycle suggested that a preferred pattern of entrainment existed. Inspired volumes were smallest for breaths initiated in the middle of the drive phase, and expired volumes were smallest for breaths initiated at or immediately after finish (Fig. 3D). A similar pattern was briefly noted by Steinacker et al. (13) and was present in limited exemplar data reported by Mahler et al. (11). This pattern suggested that there were advantageous times in the stroke for large inspired and expired volumes. Moreover, when rowers were entrained, they appeared to be taking advantage of this pattern.

Peak inspiratory flow rates in unentrained breaths were rarely achieved at stroke finish, and peak expiratory rates were smaller and less frequent during early recovery (Fig. 4D). This pattern was even more distinct in peak flow data from entrained breaths (Fig. 4C). The near absence of peak inspiratory flow at the end of drive likely accounts for the decreased inspiratory volume for breaths initiated in the middle of the drive phase. The timing of peak flow and volume minima suggests that a greater limitation on ventilation exists in the finish position than in the catch position. Therefore, the present data suggest that a mechanism other than the cramped body position at catch (4) impairs ventilatory volumes.

Kinematic data revealed that, at stroke finish, the torso had reached its maximum extension relative to the thighs. Subsequent flexion between the trunk and thighs, visible as the second negative peak in the angular acceleration of the hip joint (Fig. 9), occurred after the rowing force had dropped to zero (at ~44% of the rowing cycle) and was likely produced by contraction of the abdominal flexor muscles. Although the activity of these muscles was not measured in this study, the timing of this contraction would correspond to the period in which few breaths achieved peak inspiratory flow (Fig. 4, C and D) and would suggest that these two phenomena were linked. Abdominal flexor muscles have been shown to assist in trunk flexion in two ways: first, by generating the force to cause acceleration and, second, by stiffening the trunk to transfer the force to the upper torso. When a subject is in a standing posture, stiffening of the torso is partially achieved through pressurization of the abdominal cavity by cocontraction of the diaphragm and abdominal muscles (7). In rowing, periodic cocontraction of these muscles at stroke finish and the resulting transient abdominal pressure increase may momentarily impair diaphragm function. This mechanism may be responsible for the paucity of peak inspiratory flow data at stroke finish.

Under resting conditions, the maximal inspired volume in the catch position was lower compared with both the standing and seated positions. Although these results appeared to support a theory that a cramped body position in catch affected lung volumes (4), the volume decrement in the catch position was only ~5%. Given that maximal VT during exercise averaged ~55% of forced vital capacity in the standing position, it was unlikely that this volume decrement limited performance.

Previous studies have observed lower E (12) and reduced VT (14) for rowing compared with cycling and running, also suggesting ventilatory impairment in rowing. Gavin et al. (6) studied 13 men during an incremental maximal cycle test and used the E/CO₂ ratio as an index for the ventilatory response to exercise. In the group defined as "high," subjects reached a maximal O₂ of 4.4 l/min, CO₂ of 5.2 l/min, and respiratory exchange ratio of 1.19, values similar to those in our study. The E/CO₂ ratio in our data was slightly higher than that in the study by Gavin et al. (6) (33.4 vs. 28.0), suggesting that despite similar levels of CO₂, our subjects had an adequate ventilatory response to exercise.

On the basis of earlier incremental exercise tests, it has been previously speculated that breathing drives locomotion (13). Under the simulated race conditions used in this study, all rowers maintained a similar stroke rate and stable power output through all but the initial and final portions of their 2,000-m test. These results were expected, given that team rowers must all row at the same stroke rate during a race. Given the observed locomotor consistency and the differing ventilatory strategies used to maintain entrainment, these rowers appeared to alter their ventilation to match locomotion under simulated race conditions. Despite increasing demands to breathe during the test, rowers maintained steady stroke rates and power output. In general, rowers may seek to breathe at times where muscle synergy produces larger volumes for a given amount of respiratory work, or alternatively, the same volume for less respiratory work. The findings of the present descriptive study support the theory that locomotion drives ventilation under simulated race conditions in rowers.