Effects of Side Lying on Lung Function in Older Individuals.

Key Words: Body position, Diffusing capacity, Older adults, Oxygen transport, Pulmonary function, Side lying, Recumbency, Ventilatory inhomogeneity.

Irrespective of underlying pathology, hospitalized patients often assume recumbent body positions such as supine and side lying. Patients are routinely turned side-to-side for comfort and to avoid the negative effects of recumbent static body positions such as skin breakdown and contractures. In addition, body positioning is prescribed by physical therapists to directly enhance oxygen transport and oxygenation, to minimize the risk of aspiration, and to drain pulmonary secretions.[1] Compared with the upright position, however, recumbent positions have well-documented deleterious effects on lung function, such as reduced lung volumes and capacities, increased closing volume of the dependent airways, reduced flow rates, and reduced arterial saturation.[1] These effects are accentuated with age, smoking history, obesity, breathing at low lung volumes, sedation, anesthesia, and other pharmacological agents.[2]

Although side-lying positions are commonly used clinically, the differential effects of right and left side lying on lung function compared with a reference position such as upright sitting have not been studied in detail. There have been a few reports of improved arterial oxygenation in left versus right side lying in patients with unilateral lung disease[3-5] and bilateral lung disease[6] and in patients following coronary artery bypass surgery.[6] In recumbent positions, gas exchange is improved with the healthy lung down in patients with unilateral lung disease[3] and in right side lying in patients with bilateral lung disease.[5] In patients with unilateral lung disease, the role of the inferior lung as a gas exchanger is enhanced because of the cephalad displacement of the hemidiaphragm placing it at a greater mechanical advantage.[1] In addition, the expansive forces on the superior lung that maximize gas exchange in that lung may also contribute. In patients with bilateral lung disease, gas exchange may be enhanced due to the increased volume of the right lung anatomically and less effect of cardiac compression on this lung.[1]

The purpose of our study was to replicate and extend the existing body of knowledge pertaining to the normal relationship between side lying and several lung function variables in older age groups. Given that the majority of hospitalized patients tend to be older and our belief that physical therapists need to understand normal responses as a basis for understanding abnormal responses superimposed by pathology, we studied individuals aged 50 years or older with no known history of cardiac or pulmonary disease. The bony structure of the thorax, the ventilatory muscles, the lung parenchyma, the pulmonary vasculature, and the heart are affected by the aging process.[7,8] The interrelated changes in these structures result in a gradual decline in cardiopulmonary function and gas exchange with aging that is distinguishable from the more profound loss of function that occurs as a result of disease.[7-9]

The results of our study, therefore, could yield greater understanding of the relationship between the side-lying positions and lung function, which is fundamental to (1) improving our understanding of this relationship when pathophysiology is superimposed, (2) using body positioning as a primary intervention to maintain or improve overall gas exchange in patients with cardiopulmonary dysfunction, and (3) minimizing the effects of deleterious body positions. Thus, the findings of this study will refine the principles for prescribing body positioning as opposed to routine body positioning.

Method

Research Design

A within-subject experimental design was used to examine the interrelationship between side-lying positions and indexes of lung function in 19 older subjects. Subjects ranged in age from 50 to 74 years. Subjects between 50 and 75 years of age were selected for study because this range is representative of many older hospitalized patients. Subjects were tested during 2 sessions, conducted less than 1 week apart, over a 5-month period. In each session, lung function tests were performed in 2 body positions (ie, the reference position of sitting and one side-lying position). The selection of left or right side lying on the first visit was alternated for subsequent subjects. The same tester was used for all subjects. The lung function tests included spirometric tests of forced vital capacity (FVC) and forced expiratory volume in 1 second ([FEV.sub.1]); the single-breath nitrogen ([SBN.sub.2]) test to determine inhomogeneity of ventilation from the slope of phase III ([DN.sub.2]%/L), a precursor to ventilation and perfusion matching; and a single-breath test to determine pulmonary diffusing capacity (DLCO).

Analysis of the slope of phase III, the expired alveolar plateau, of the [SBN.sub.2] test has been used for many years to measure the uniformity or homogeneity of ventilation in the lung, a requisite for optimal oxygenation.[10] The subject's exhalation is analyzed for nitrogen (80% of the composition of normal air that we breathe) after a full inspiration of 100% oxygen from residual volume. The first part of the exhaled gas from the anatomic dead space has a 0% concentration of nitrogen. Then, there is a sharp rise in nitrogen concentration as gas from the alveoli becomes mixed with gas from the dead space.

This is followed by a relative plateau as more gas is emptied from the alveoli. A steep slope of this phase indicates nonuniform distribution of alveolar gas, The greater the nonuniformity or inhomogeneity of ventilation, the greater the ventilation to perfusion mismatch, hence, deoxygenation.

In each position, spirometric tests were conducted first to provide a reference value for vital capacity (VC) for the test of DLCO. The test of DLCO was conducted before the [SBN.sub.2] test because the latter test involved inhaling 100% oxygen, which could influence the results of the DLCO test. Each session lasted approximately 2 1/2 hours. Subjects rested between tests and between repeated trials of each test.

Subjects

Nineteen subjects (11 women, 8 men), with a mean age of 62.8 years (SD = 6.8, range = 50-74), participated in the study. Nine subjects were lifetime nonsmokers, and 10 subjects were ex-smokers. The average height of the women was 159 cm (SD = 6, range = 145-168), and that of the men was 176 cm (SD = 8, range = 163-183). The average weight of the women was 56.6 kg (SD = 8.2, range = 45-76), and that of the men was 72.0 kg (SD = 11.0, range=56-93). All subjects had no known history of cardiac or pulmonary disease. The group was recruited from the university community through a public announcement.

Procedure

We studied the following dependent variables in different body positions: (1) pulmonary function, including spirometric measures of FVC and [FEV.sub.1], (2) inhomogeneity of ventilation, as measured by the slope of phase III ([DN.sub.2]%/L) of the [SBN.sub.2] test, and (3) pulmonary diffusing capacity, based on the single-breath test adjusted for alveolar ventilation (DLCO/VA). The spirometric variables were included for the purposes of screening for airway obstruction, setting a baseline for acceptability of the DLCO test (which requires the subject to inhale a minimum of 90% of his or her VC), and evaluating changes between different positions. The [DN.sub.2]%/L and DLCO/VA provide valuable information with respect to the capacity of the lung to serve as a gas exchanger. The [DN.sub.2]%/L focuses on the process of ventilation and its distribution (ie, the movement of inspired gas from the atmosphere to the alveoli). The DLCO/VA focuses on the process of diffusion and involves the transport of gases across the alveolar-capillary membrane. The distribution of ventilation reflected by the homogeneity of ventilation within the lung, in combination with the distribution of pulmonary perfusion, is a determinant of the efficiency of gas exchange. All spirometric and lung function indexes were obtained according to the American Thoracic Society (ATS) standards.[11-13] All subjects were asked to refrain from vigorous exercise prior to testing on the day of the test session. They were also requested to avoid eating a heavy meal within 2 hours of the test and to wear comfortable, nonrestrictive clothing.

The general experimental procedure is outlined in the Figure. The protocol for session A was randomly selected for the first subject on the first visit, and the protocol for session B was used on the second visit. For successive subjects, the 2 protocols were alternated. On arrival at the laboratory for the first session, the test procedures were explained to the subjects, who then gave written consent to participate in the study. The time taken for this discussion and for the determination of individual heights and weights allowed for a rest period prior to testing. It also allowed for familiarization with the environment and the tester. In both sessions, the first test position was sitting and the second test position was either left or right side lying. The order of lung function tests, as shown in the Figure, was the same for each visit, as necessitated by spirometric prerequisites for DLCO and the potential for the 100% oxygen inhaled during the [SBN.sub.2] test to interfere with the results of the DLCO test. Both testing sessions were completed within 1 week.

Figure. General experimental protocol. Sessions A and B were alternated for successive subjects. DLCO = pulmonary diffusing capacity, [SBN.sub.2] = single-breath nitrogen.

SESSION A
            POSITION:   Sitting
            TESTS:      Spirometry
                        DLCO (after 15 rain in sitting)
            POSITION:   Left side lying
            TESTS:      Spirometry
                        DLCO (after 15 rain in side lying)
            POSITION:   Silting
            TESTS:      [SBN.sub.2] (after 15 min in sitting)
            POSITION:   Left side lying
            TESTS:      [SBN.sub.2] (after 15 min in side lying)
SESSION B
            POSITION:   Sitting
            TESTS:      Spirometry
                        DLCO (after 15 min in sitting)
            POSITION:   Right side lying
            TESTS:      Spirometry
                        DLCO (after 15 min in side lying)
            POSITION:   Sitting
            TESTS:      [SBN.sub.2] (after 15 min in silting)
            POSITION:   Right side lying
            TESTS:      [SBN.sub.2] (after 15 min in side lying)

Body position. For the sitting position, subjects were seated in a firm high-backed chair. During the tests in the sitting position, the tester ensured that the subjects maintained an upright position. For the side-lying positions, subjects lay on a flat stretcher with one pillow under the head, one pillow behind the back and against the stretcher rail, and one pillow between the knees. The tester observed the subjects throughout the rest periods and the test maneuvers to ensure the maintenance of a full side-lying position with the head and neck in line with the torso.

Subjects assumed the test position 15 minutes prior to the first trial of either the DLCO test or the [SBN.sub.2] test in each position. This rest period was included to accommodate the effects of position change on the pulmonary circulation, notably the pulmonary capillary blood volume, which are known to be time-dependent.[14,15] Although the time course and duration of position-related changes in the pulmonary circulation are not well-defined in the literature, various studies of the effect of position change on DLCO have used rest periods of 15 minutes or less.[16-18]

Measurements from the SensorMedics 2200 System. The SensorMedics 2200 System(*) has both a nitrogen analyzer, which is used in the [SBN.sub.2] test, and a rapid response multi-gas analyzer, which is used in the test of DLCO.[19] For measuring the nitrogen concentration of expired gas during the [SBN.sub.2] test, a needle valve is mounted into a mouthpiece adapter. The valve is connected to a fast-response nitrogen analyzer containing an ionization chamber. Maximum nitrogen ionization is achieved by an optimum negative pressure (created by a vacuum pump) in the analyzer. When nitrogen ionizes, it emits ultraviolet light, the intensity of which is directly related to the concentration of nitrogen. The light energy is converted into an electric signal that is translated by the computer into a concentration. The response time for this analysis is given as less than 50 milliseconds. A different method is used for measuring the carbon monoxide concentration of expired gas during the single-breath DLCO test. In this case, the measuring principle is the nondispersive infrared absorption technique. The expired gas is exposed to a beam of infrared energy and absorbs a certain amount of this energy, depending on the partial pressure of the gas (which, in turn, is dependent on the concentration of the gas). The infrared absorption is measured and converted to an electric signal that is relayed to the computer.

Calibration for flow and volume measurements was done prior to each session and according to the procedure outlined in the SensorMedics 2200 operator's manual.[19] This procedure involved using a 3-L calibration syringe to deliver room air into the system. Calibration of the nitrogen analyzer was performed a least once a week, in accordance with the terms outlined in the operator's manual. This calibration involved "peaking the needle" by opening or closing the needle valve (to allow more or less gas flow into the analyzer) to obtain the maximum percentage of nitrogen reading on room air. The vacuum pump was turned on a minimum of 20 minutes prior to calibration, in accordance with the specified calibration procedure. A linearity check of the nitrogen analyzer was performed at the start of the study and twice during the study. This check involved exposing the SensorMedics 2200 to 5 different known concentrations of nitrogen (ranging from 0.00% to room air at 79%; 100% oxygen [0% nitrogen] to air [79% nitrogen]) for analysis.

Depending on the specific test selected from the computer menu, a valve automatically switched the subject, who was connected to the system by a breathing tube and mouthpiece, between room air, 100% oxygen, or a multi-gas mixture. Throughout a test, tracings on the computer screen provided visual feedback for the subject and coaching cues for the tester. In addition, "best" results for spirometric tests (as per ATS standards) and averages for [DN.sub.2]%/L and DLCO were displayed, as were the values expressed as a percentage of the predicted values. For our study, we used the prediction equations of Crapo et al[20] for volume measurements, those of Knudson et al[21] for flow measurements, those of Buist and Ross[22] for [DN.sub.2]%/L measurements, and those of Miller et[23] ales for diffusion measurements.

Performance of spirometric tests. The maximum expiratory maneuvers were conducted according to ATS standards.[11] Positioned in either sitting or side lying, the subject was monitored throughout the test to prevent alterations in body position. Following a detailed explanation of the test, the subject was connected to the mouthpiece with the noseclip in place. After several tidal breaths, the subject was coached through the maximal forced expiration procedure. The end of the test, as defined by ATS criteria of no change in volume for at least 2 seconds following an exhalation time of at least 6 seconds, was indicated by a computer message displayed on the screen. The subject expired with maximal expiratory effort until 3 acceptable tracings were recorded (which involved a maximum of 6 trials). In accordance with ATS recommendations for test reproducibility, the 2 largest FVCs (taken from acceptable curves) varied by less than 5%. Similar reproducibility criteria were applied to the measurement of [FEV.sub.1].

From the repeated trials that met ATS criteria,[13] the largest FVC measurement and the largest [FEV.sub.1] measurement were selected for data analysis. The selected spirometric values were compared by the computer with reference values and were presented as a percentage of the predicted value.

Performance of the [SBN.sub.2] test. The [SBN.sub.2] tests were conducted according to the National Heart and Lung Institute standards.[12] Following a detailed explanation of the maneuver, the subject was connected to the mouthpiece with a noseclip in place. After several tidal breaths, the subject inhaled 2 slow, deep breaths and then exhaled completely. From residual volume (which is recognized by the computer as the point at which there is no expiratory flow for 0.5 second), the subject inhaled a slow VC breath of oxygen. The flow rate of the subsequent slow exhalation was maintained between 0.3 and 0.6 L/s for as long as possible. To assist the subject in maintaining a flow rate in this range, a resistance was inserted in the expired circuit to slow down expiratory flow and the flow rate was displayed on the computer screen during exhalation, along with the upper and lower limits of the flow rate range.

In each test position, the [SBN.sub.2] test was repeated until 3 acceptable tracings were obtained according to the National Heart and Lung Institute standards.[12] Rest periods of approximately 5 minutes were given between trials to provide time for washout of excess oxygen and thus the restoration of the normal nitrogen gradient in expired air. The [DN.sub.2]%/L was calculated by the computer from analysis of the final expirate, using the increase in nitrogen concentration over 1 L of the expired volume, between 750 mL (representing the complete dead space washout) and 1,750 mL. The average [DN.sub.2]%/L value from 3 acceptable tests was used for data analysis and was expressed as a percentage of the predicted value.

Performance of the DLCO test. The DLCO tests were conducted according to ATS standards[11] for a single-breath test. As with the other tests, the subject was connected to the mouthpiece with a noseclip in place for a short period of tidal breathing before the test maneuver. On a cue from the tester, the subject exhaled to residual volume and then quickly inhaled a VC measure of the test gas, which contained minimal, nontoxic concentrations of carbon monoxide (0.30%), methane (0.30%), and acetylene (0.30%); 20.95% oxygen; and the balance, nitrogen. On the basis of previous forced expiratory maneuvers, which gave a measure of VC in the same position, the computer displayed the minimum acceptable volume (at least 90% of the subject's VC) by a horizontal target line, which the subject could see while inhaling. The subject was coached to hold his or her breath at total lung capacity for 10 seconds, during which time he or she was encouraged to relax against the closed valve of the unit. The end of the breath hold was signaled by the crossing of the volume versus time tracing over a vertical time line, after which the expiratory valve automatically opened and the subject was coached to exhale rapidly and completely. The precise breath-hold time was calculated by the computer and based on ATS standards. The mean concentrations of methane and carbon monoxide were calculated by the computer over a 1,000-mL alveolar sample volume from 750 to 1,750 mL of the expired volume. The sample collection volume was not adjusted larger than 1,000 mL.

In each test position, the DLCO test was repeated until at least 2 acceptable tracings were obtained. A rest interval of at least 4 minutes was given between trials to allow the test gas to wash out from the subject's lungs. Before each trial, an automated procedure was initiated in which methane and carbon monoxide concentration readings were zeroed with room air. The 2 tests from which an average DLCO/VA was calculated were within 10% of each other, in accordance with ATS recommendations. 11 The average DLCO/VA was expressed as a percentage of the predicted value for data analysis. Although an adjustment for hemoglobin is desirable in the calculation of DLCO, it is not considered mandatory by the ATS[11] for subjects with no known history of cardiac or pulmonary disease and was not made in this study. In addition, based on the findings of Pistelli et al,[18] we assumed that the hemoglobin concentration was constant for every subject throughout the study. Because the 2 test sessions were conducted within 1 week, the effect of normal fluctuations in hemoglobin on DLCO was presumably minimized. Thus, we were confident that hemoglobin changes did not confound DLCO when different positions were studied.

Data Analysis

Descriptive statistics were calculated for age, height, and weight. Descriptive statistics were also calculated for the 4 dependent variables (ie, FVC, [FEV.sub.1], [DN.sub.2]%/L, and DLCO/VA). Individual subject test results were compared with predicted values. For each subject, the observed value was compared with the lowest acceptable normal limit (for the spirometric variables and for DLCO) or the highest acceptable normal limit (for [DN.sub.2]%/L). The limits of normal were obtained by subtracting (for the lowest limit) or adding (for the highest limit) the 95th percent confidence interval (1.65 x standard error of the estimate) from the predicted value for each subject. Coefficients of variation (CVs) were calculated to establish the reliability of the measurements.

Within-subject one-way analyses of variance (ANOVAs) were used to analyze the effect of body position on each of the dependent measures. Post hoc comparisons were made using the Duncan multiple range test. A significance level of P [is less than] .05 was selected for all statistical tests.

Results

All subjects' data fell into the acceptable limits for individuals without cardiac or pulmonary disease, based on the criteria of the ATS.[11-13] The CV is a statistic that is commonly used in the analysis of pulmonary function data to reflect the reproducibility of a measure. The CVs were in agreement with the mean norms published in the literature in the 2 sitting reference positions (Tab. 1).[24-27] We did not test the reliability of our measurements using traditional psychometric methods but rather used methods that have been widely adopted in research on pulmonary function.

Table 1. Reliability of Measurements for the Dependent Variables: Coefficients of Variation (Standard Deviation/Mean %) Within Subjects (N = 19) for the Two Sitting Positions

                Test Session A
Variable(a)      X        SD       Range

FVC              2         1      0.4-0.5
[FEV.sub.1]      3         4      1.0-3.0
DLCO/VA          2         2      0.2-5.0
[DN.sub.2]%/L   19        10      9.0-42.0

                Test Session B
Variable(a)     X        SD       Range

FVC              2         2      0.2-3.0
[FEV.sub.1]      3         2      1.0-10.0
DLCO/VA          1         1      0.3-4.0
[DN.sub.2]%/L   24        16      8.0-63.0

(a) FVC=forced vital capacity (in liters, body temperature pressure saturated [BTPS]), [FEV.sub.1]=forced expiratory volume in 1 second (in liters, BTPS), DLCO/VA=diffusing capacity for carbon dioxide per unit of alveolar volume (in milliliters of carbon dioxide per minute per millimeters of mercury per liter, BTPS), [DN.sub.2]%/L=inhomogeneity of ventilation measured by the slope of phase III of the single-breath nitrogen test. Coefficients of variation reported by other investigators: for FVC and [FEV.sub.1] , <5% (McCarthy et al[25]) and <3% (Chinn and Lee[23]); for DLCO, <4% (Du Perron et al[27]); for [DN.sub.2]%/L, 22% (McCarthy et al[25]).

The descriptive statistics for spirometric variables (FVC and [FEV.sub.1]), [DN.sub.2]%/L, and DLCO/VA for each of the test positions are summarized in Table 2, and the ANOVA results are presented in Table 3. From the sitting position to the left side-lying position during test session A, the average observed FVC decreased by 3.3%, [FEV.sub.1] decreased by 6.2%, DLCO/VA decreased by 3.0%, and [DN.sub.2]%/L increased by 25%. From the sitting position to the right side-lying position during test session B, the average observed FVC decreased by 3.2%, [FEV.sub.1] decreased by 4%, DLCO/VA decreased by 2%, and [DN.sub.2]%/L increased by 14%. Also of note was the disproportionately large range of percentage of predicted values for [DN.sub.2]%/L in the side-lying positions (a range of 409 in the left side-lying position versus 260 in the sitting position and a range of 358 in the right side-lying position versus 331 in the sitting position).

Table 2 Descriptive Statistics for the Dependent Variables in Test Session A for Sitting and Left Side Lying and in Test Session B for Sitting and Right Side Lying (N = 19)

                     Observed      % Predicted
Variable(a)          X      SD     X        SD

Test session A
  Sitting
    FVC              3.97   0.87   113      13
    [FEV.sub.1]      2.76   0.67   101      18
    DLCO/VA          4.29   0.85    98      15
    [DN.sub.2]%/L    2.07   0.87   135      70
  Left side lying
    FVC              3.84   0.85   110(b)   13
    [FEV.sub.1]      2.59   0.65    95(b)   17
    DLCO/VA          4.16   0.74    95      13
    [DN.sub.2]%/L    2.59   1.41   168(c)   99
Test session B
  Sitting
    FVC              4.00   0.94   113      14
    [FEV.sub.1]      2.74   0.68    99      18
    DLCO/VA          4.24   0.82    97      15
    [DN.sub.2]%/L    2.15   1.03   140      78
  Right side lying
    FVC              3.87   0.89   110(b)   13
    [FEV.sub.1]      2.63   0.67    96(b)   17
    DLCO/VA          4.16   0.87    95      16
    [DN.sub.2]%/L    2.46   1.52   154(c)   96

(a) FVC=forced vital capacity (in liters, body temperature, pressure-saturated [BTPS]), [FEV.sub.1] =forced expiratory volume in 1 second (in liters, BTPS), DLCO/VA=diffusing capacity for carbon monoxide per unit of alveolar volume (in milliliters of carbon dioxide per minute per millimeters of mercury per liter, BTPS), [DN.sub.2]%/L=inhomogeneity of ventilation measured by the slope of phase III of the single-breath nitrogen test.

(b) Significant difference between sitting and left side lying (test session A) and fight side lying (test session B), P <.01.

(c) Significant difference between sitting and left side lying (test session A) and fight side lying (test session B), P <.05.

Table 3.
Results of Analyses of Variance

                        Main Effect    Error

Dependent Variable(a)   df     MS      df     MS        F       P

FVC                     3      88.67   54     16.27    5.45    .002
[FEV.sub.1]             3     117.5    54     12.22   14.28   <.001
DLCO                    3      97.9    54     34.18    1.42    .25
[DN.sub.2]%/L           3    3981.7    54   2244.3     1.77    .16

(a) FVC=forced vital capacity, [FEV.sub.1] = forced expiratory volume in 1 second, DLCO/VA = pulmonary diffusing capacity, [DN.sub.2]2%/ L = inhomogeneity of ventilation measured by the slope of phase III of the single-breath nitrogen test.

Discussion

Effect of Sitting Versus Side-Lying Positions

In the majority of studies in which spirometric measures were compared between sitting and recumbent positions, the researchers investigated FVC with subjects in sitting and supine positions rather than in side-lying positions.[18,28,29] The reduction in FVC from a sitting position to a side-lying position observed in our study is in agreement with the results of these studies, supporting our hypothesis that the effect of side lying on FVC is similar to that observed in a supine position. Behrakis et al,[30] who were among the few investigators to study the effect of side lying on lung function, reported similar findings for FVC. The decrease in FVC in recumbency may reflect both increased thoracic blood volume due to the gravitational facilitation of venous return and cephalad displacement of the diaphragm caused by abdominal encroachment. In a side-lying position, even though only the dependent hemidiaphragm is displaced, the effect on FVC appears to be similar to that observed in a supine position. Other factors include increased airway resistance and decreased lung compliance secondary to anatomical differences between the left and right lungs and shifting of mediastinal structures.[31]

The decrease in [FEV.sub.1] in a side-lying position compared with a sitting position is in agreement with the relatively few studies demonstrating changes in [FEV.sub.1] with recumbency.[17,18] Although direct comparison with these studies cannot be made because the effects of the recumbent position appear to vary between side-lying and supine positions, based on the results of our study, the similarity of results suggests that recumbent positions also limit expiratory volumes and flow. The exact cause of this apparent obstructive process, which may reflect an increase in airway resistance, a decrease in elastic recoil of the lung, or decreased mechanical advantage of forced expiration, presumably affecting the large upper airways, is not clear. In our study, care was taken to position subjects so that the head and neck were in alignment with the trunk and in a neutral position (in both sitting and side-lying positions). This neutral position was adopted to avoid extraneous stress on the upper airways. Anthonisen[32] reported that hyperextension of the neck results in an increase in [FEV.sub.1], secondary to elongation and stiffening of the trachea, thereby facilitating air flow. Conversely, the relaxed recumbent positions examined in our study may effectively shorten and increase the compliance of the airways, such that these positions limit rather than augment maximal forced expiratory effort.

The slope of phase III, [DN.sub.2]%/L, of the [SBN.sub.2] washout test reflects both interregional and intraregional differences in the lung? Even in health, the CV for [DN.sub.2]%/L is well-known to be high compared with other indices of lung function? The mechanism underlying the greater ventilatory inhomogeneity and the higher CV for [DN.sub.2]%/L in a side-lying position compared with a sitting position could reflect several factors that affect the distribution of ventilation, especially in older people. Key contributory factors include airway closure and increased pulmonary time constants, both of which are adversely affected in recumbent positions as well as with advancing age.[33] Anatomical factors such as an increase in the weight and volume of the heart with aging[34] and thus greater impingement on adjacent lung parenchyma in the left side-lying position, may also have had a role. Whether residual effects from a history of smoking affected intraregional differences in 10 subjects is unclear.

Statistically, our results do not support increased inhomogeneity of ventilation in a side-lying position, which would be predicted for an older age group. As described by Otis et al,[35] the concept of pulmonary time constants explains the variations in gas entry into independently ventilated lung units. Increased pulmonary time constants, caused by regional changes in airway resistance and compliance and leading to varying degrees of filling of lung units, can increase the inhomogeneity of ventilation.[36] From determinations of lung resistance and lung compliance in different body positions, Behrakis et al[30] concluded that pulmonary time constants overall were greater in a side-lying position than in a sitting position in young adults without cardiopulmonary impairments. This conclusion was based on the disproportionate increase in lung resistance in a side-lying position (40% greater than in a sitting position) compared with the decrease in lung compliance (10% less than in a sitting position). Thus, the product of resistance and compliance (ie, the pulmonary time constant) was greater in a side-lying position than in a sitting position. Michels et al[37] compared supine and sitting positions and found a similar position-dependent increase in resistance of the respiratory system in a group of adults without cardiopulmonary impairments aged 20 to 67 years. Furthermore, they found that, in a subgroup of young subjects, the increase was more marked in smokers than in nonsmokers. In addition, the increase was more marked with aging, such that over the age of 50 years, both nonsmokers and smokers demonstrated position-dependent increases in resistance that were of the same magnitude. Behrakis et al[30] suggested that changes in geometry of the upper airways, the aperture of the glottis, or both may Contribute to this effect. Michels et al[37] proposed that intrinsic narrowing of the peripheral airways of smokers may be more pronounced in recumbency than in a sitting position.

In light of these considerations, we would anticipate that, in comparison with the sitting position, both left and right side lying would reflect the effects of airway closure and increased pulmonary time constants in conjunction with increased inhomogeneity of ventilation. The fact that our results do not support the findings of other researchers suggests that the effect of body position on the inhomogeneity of ventilation is more variable than for the spirometric measures. Rather than reflect on the results of "outliers," our results indicate that age-related factors could have had a role. The large variability of the [DN.sub.2]%/L may have obscured a real difference between the left and right side-lying positions. Further study is needed to elucidate differential effects between left and right side-lying positions in older people and whether this effect is accentuated further with factors such as pathology, smoking history, and obesity.

The effect of recumbency on diffusing capacity is conflicting in the literature and may reflect age-related factors. The lack of a difference in DLCO/VA between sitting and either left or right side lying in older subjects in our study is consistent with the work of Stam et al,[38] although they used supine rather than side lying as the recumbent position. These findings are in contrast to those of investigators[39-41] who found increases in DLCO/VA of up to 15% from a sitting position to a recumbent, usually supine, position. This discrepancy may be explained by a greater unevenness of DLCO/VA through the lung in a side-lying position than in a supine position (due to an increased transverse diameter of the chest compared with anteroposterior diameter and the greater vertical gradient in a side-lying position than in a supine position). Furthermore, few investigators have targeted an older population for the study of position-related changes.

Some authors have reported that DLCO decreases with increasing age. Georges et al[9] attributed the decrease to a decline in Dm (the membrane component of DLCO) after the age of 40 years and to a decrease in pulmonary capillary blood volume after the age of 60 years. In addition, the anatomic changes associated with aging that affect Dm and pulmonary capillary blood volume may account for the apparent differences in response to a position change as compared with a young population. Brody and Thurlbeck,[33] for example, described a loss of alveolar surface area, a possible decrease in number of pulmonary capillaries, and an increase in the inner diameter of alveoli, which may affect the mixing of gases by diffusion, as morphologic changes accompanying aging. These changes may reduce the increase of pulmonary capillary blood volume from a sitting position to a recumbent position and may be reflected by a constancy of DLCO between positions. Further study of underlying mechanisms, however, is needed in light of reports that there is no difference in the distribution of pulmonary perfusion between older and younger individuals and little change in pulmonary capillary density.[42] Similarly, further study is needed to explain these discrepancies in the literature.

Effect of Left Versus Right Side-Lying Positions

The similarity of the results for left side lying and right side lying are in agreement with the results of the study by Behrakis et al,[30] one of the few studies comparing lung function in the side-lying positions. These investigators also reported no differences between the side-lying positions for VC, expiratory reserve volume, static and dynamic compliance of the lung, resistance of the lung, or pulmonary time constants in a young group of subjects. In other studies of the side-lying positions, however, there was either the selection of only one of the side-lying positions, usually right side-lying,[43-46] or no reported differentiation between left and right side-lying positions.[47-49]

In younger adults without cardiopulmonary impairments, there is no reason to suspect a difference in lung function between left and right side-lying positions. In an older population, however, the age-related variation in cardiopulmonary status (eg, increase in weight and volume of the heart,[7,8] changes in mediastinal compliance) may result in differences in function between left and right side-lying positions when compared with a sitting position. Furthermore, in the presence of cardiac or pulmonary conditions, position-related effects on cardiopulmonary status may be accentuated in left and right side-lying positions. Zack et al[5] for example, reported that arterial oxygen tension was generally higher in right side lying than in left side lying in 13 patients (ages not reported) with equally distributed bilateral lung disease, whereas there was no difference in arterial oxygen tension between side-lying positions in 6 control subjects (mean age = 25 years). This effect was attributed to the smaller volume of the left lung and compression of the heart on the left lung in left side lying. The morphological changes in the lung that are associated with aging could have a similar effect if the changes are unequally distributed between the left and right lungs; however, there is no evidence to suggest that this is the case. Regional age-related changes in the lung have not been reported, except for a greater degree of emphysematous change (considered as a normal part of the aging process in nonsmokers) in the lower zones compared with the upper zones of the lung.[33]

Clinical Implications

Physical therapists should anticipate the physiological effects of side-lying body positions when managing their patients, who may assume such positions for comfort and rest, may be placed in these positions to avoid the negative effects of static body positions, or may be placed in specific therapeutic body positions to augment arterial oxygenation or drain pulmonary secretions. Based on our results, we contend that physical therapists should consider the predictable reduction in lung volumes, lung capacities, and flow rates when placing patients in side-lying positions. Furthermore, based on a comparison of findings in the literature, we believe that therapists should also consider the less predictable changes in diffusing capacity and homogeneity of ventilation.

The adverse effects of recumbent positions can be accentuated by the following factors: cardiopulmonary pathology, age, smoking history, obesity, breathing at low lung volumes, sedation, anesthesia, oxygen, and other pharmacological agents.[2] Thus, patients who may be scheduled for even minor medical or surgical procedures may be at risk because of these confounding factors. We believe it is essential that physical therapists be able to identify patients who are at risk to ensure that they are appropriately monitored and that upright positions are encouraged over recumbent positions as much as possible. In our view, therapeutic recumbent positions should be prescribed judiciously between treatments, and early intervention should be instituted if necessary. In addition, body positioning should not be used injudiciously or without appropriate monitoring, even when administered routinely for comfort and for avoidance of the negative effects of prolonged static positioning. Without due precautions, a patient who is apparently at low risk can readily become a patient who is at high risk.

Finally, of considerable clinical importance is the fact that a greater physiologic understanding of the strong and direct effects of body positioning on cardiopulmonary function and gas exchange has helped to explain both improvements previously attributed to conventional "chest" physical therapy and the increasing number of negative outcomes in response to this time-honored treatment. We believe that the results of studies evaluating interventions for patients with cardiopulmonary disorders (eg, measures of lung function and gas exchange) have been confounded by the effects of body positioning (or body positioning in combination with mobilization), irrespective of the effect of airway clearance.[6] Because optimal lung function is associated with the upright positions,[2] the results of our study lend further support to exploiting the therapeutic benefits of these positions.

Conclusions

Forced vital capacity and [FEV.sub.1] were decreased equally in left and right side-lying positions in older individuals without cardiopulmonary disorders, whereas no corresponding change was observed in diffusing capacity and inhomogeneity of ventilation in either side-lying position. Conflicting reports in the literature suggest that, compared with spirometric measures, the effect of recumbency on diffusing capacity and homogeneity of ventilation is less predictable and that intervening variables need to be identified. Our results support the idea that body position has a profound effect on lung function and respiratory mechanics. In recumbency, factors contributing to impaired lung function may include external compression of the chest wall, impingement of the abdominal contents on the diaphragm, compression of airways and blood vessels by the heart, and the age-related increase in the mass and volume of the heart. Further studies are warranted to elucidate the mechanisms of the apparent increase in air flow resistance and reduced lung compliance in side-lying positions.

The absence of any change in diffusing capacity in older people without cardiopulmonary impairments supports the idea that side lying may not induce the comparable changes in pulmonary capillary blood volume and venous return that are reported in the literature and that are responsible for the associated increase in diffusing capacity in a supine position. Although no increase in inhomogeneity of ventilation was observed in our study, this finding does not minimize the well-known detrimental effects of recumbency on functional residual capacity and associated arterial desaturation.

Our results have implications for both routine and therapeutic positioning of hospitalized patients with or without cardiac or pulmonary conditions. The physical therapist needs a thorough knowledge of all factors that affect all steps in the oxygen transport pathway in order to prescribe this body positioning efficaciously (ie, with maximal benefit and least risk).

Acknowledgment

We gratefully acknowledge the donation of equipment by Summit Technologies Inc for the study.

(*) Summit Technologies Inc, 840-71H Ave SW, Ste 900, Calgary, Alberta, Canada T2P 3G2.

References

[1] Dean E. Effect of body position on pulmonary function. Phys Ther. 1985;65:613-618.

[2] Ross J, Dean E. Body positioning. In: Zadai C, ed. Clinics in Physical Therapy: Pulmonary Management in Physical Therapy. New York, NY: Churchill Livingstone Inc; 1992:79-98.

[3] Remolina C, Khan AU, Santiago TV, Edelman NH. Positional hypoxemia in unilateral lung disease. N Engl J Med. 1981;304:523-525.

[4] Sonnenblick M, Melzer E, Rosin AJ. Body positional effect on gas exchange in unilateral pleural effusion. Chest. 1983;83:784-786.

[5] Zack MB, Pontoppidan H, Kazemi H. The effect of lateral positions on gas exchange in pulmonary disease: a prospective evaluation. Am Rev Respir Dis. 1974;110:49-55.

[6] Dean E. Invited commentary on "Are incentive spirometry, intermittent positive pressure breathing, and deep breathing exercises effective in the prevention of postoperative pulmonary complications after upper abdominal surgery? A systematic overview and meta-analysis." Phys Ther. 1994;74:10-15.

[7] Zadai C. Pulmonary physiology of aging: the role of rehabilitation. Topics in Geriatric Rehabilitation. 1985;1:49-57.

[8] Knudson RJ. Physiology of the aging lung. In: Crystal RG, West JB, eds. The Lung. New York, NY: Raven Press; 1991:1749-1759.

[9] Georges R, Saumon G, Loiseau A. The relationship of age to pulmonary membrane conductance and capillary blood volume. Am Rev Respir Dis. 1978;117:1069-1078.

[10] Levitzky M. Pulmonary Physiology. 2nd ed. New York, NY: McGraw-Hill Book Co; 1986:77-79.

[11] American Thoracic Society. Single breath carbon monoxide diffusing capacity (transfer factor): recommendations for a standard technique. Am Rev Respir Dis. 1987;136:1299-1307.

[12] Martin R, Macklem PT. Suggested Standardization Procedures for Closing Volume Determination (Nitrogen Method). Bethesda, Md: National Heart and Lung Institute; 1973.

[13] American Thoracic Society. Standardization of spirometry: 1987 update. Am Rev Respir Dis. 1987;136:1285-1298.

[14] Lewis ML, Christianson LC. Behavior of the human pulmonary circulation during head-up flit. J Appl Physiol. 1978;45:249-254.

[15] Hirasuna JD, Gorin AB. Effect of prolonged recumbency on pulmonary blood volume in normal humans. J Appl Physiol. 1981;50: 950-955.

[16] Malmberg R. Pulmonary gas exchange at exercise and different body postures in man. Scand J Respir Dis. 1966;47:92-102.

[17] Norregaard O, Schultz P, Ostergaard A, Dahl R. Lung function and postural changes during pregnancy. Respir Med. 1989;83:467-470.

[18] Pistelli R, Fuso L, Muzzolon R, et al. Factors affecting variations in pulmonary diffusing capacity resulting from postural changes. Respiration. 1991;58:233-237.

[19] SensorMedics 2200 Pulmonary Function Laboratory Operator's Manual. Anaheim, Calif: SensorMedics Corp; 1989.

[20] Crapo RO, Morris AH, Gardner RM. Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis. 1981;123:659-664.

[21] Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve: normal standards, variability, and effects of age. Am Rev Respir Dis. 1983;113:587-600.

[22] Buist AS, Ross BB. Quantitative analysis of the alveolar plateau in the diagnosis of early airway obstruction. Am Rev Respir Dis. 1973;108: 1078-1087.

[23] Miller A, Thornton JC, Warshaw R, et al. Single breath diffusing capacity in a representative sample of Michigan, a large industrial state: predicted values, lower limits of normal, and frequencies of abnormality by smoking history. Am Rev Respir Dis. 1983;127:270-277.

[24] Bates DV. Respiratory Function in Disease. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1989:106-151.

[25] McCarthy DS, Craig, DB, Cherniack RM Intraindividual variability in maximal expiratory flow-volume and closing volume in asymptomatic subjects. Am Rev Respir Dis. 1975;112:407-411.

[26] Chinn DJ, Lee WR. Within- and between-subject variability of indices from the closing volume and flow volume traces. Bull Eur Physiopathol Respir. 1977;13:789-802.

[27] Du Perron MC, Korobaeff M, Drutel P. Valeur et reproducibilite des mesures de TLCO et de ses composantes chez le sujet sain. Bull Eur Physiopathol Respir. 1976;12:443-451.

[28] Craig DB. Effects of position on expiratory reserve volume of the lungs. J Appl Physiol. 1960;15:59-61.

[29] Agostini E, Hyatt RE. Static behavior of the respiratory system. In: Fenn WO, Rahn H, eds. Handbook of Physiology, Section 3--Volume III: The Respiratory System; Part 2: Mechanics of Breathing. Baltimore, Md: Williams & Wilkins; 1986:113-130.

[30] Behrakis PK, Baydur A, Jaeger MJ, Milic-Emili J. Lung mechanics in sitting and horizontal body positions. Chest. 1983;83:643-646.

[31] Dean E, Hobson L. Cardiopulmonary anatomy. In: Frownfelter D, Dean E, eds. Principles and Practice of Cardiopulmonary Physical Therapy. 3rd ed. St Louis, Mo: Mosby; 1996:23-51.

[32] Anthonisen NR. Tests of mechanical function. In: Fenn WO, Rahn H, eds. Handbook of Physiology, Section 3--Volume III: The Respiratory System; Part 2: Mechanics of Breathing. Baltimore, Md: Williams & Wilkins; 1986:753-784.

[33] Brody JS, Thurlbeck WM. Development, growth, and aging of the lung. In: Fenn WO, Rahn H, eds. Handbook of Physiology, Section 3--Volume III: The Respiratory System; Part 2: Mechanics of Breathing. Baltimore, Md: Williams & Wilkins; 1986:378-380.

[34] Astrand PO, Rodahl K. Textbook of Work Physiology. New York, NY: McGraw, Hill Inc; 1970:321-340.

[35] Otis AB, McKerrow CB, Bartlett RA, et al. Mechanical factors in distribution of pulmonary ventilation. J Appl Physiol. 1956;8:427-443.

[36] Bates DV. Respiratory Function in Disease. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1989: chaps 2 and 5.

[37] Michels A, Decoster K, Derde L, et al. Influence of posture on lung volumes and impedance of respiratory system in healthy smokers and nonsmokers. J Appl Physiol. 1991;71:294-299.

[38] Stam H, Kreuzer JA, Versprille A. Effect of lung volume and positional changes on pulmonary diffusing capacity and its components. J Appl Physiol. 1991;71:1477-1488.

[39] Bates DV, Pearce JF. The pulmonary diffusing capacity: a comparison of methods of measurement and a study of the effect of body position. J Physiol. 1956;132:232-238.

[40] Ogilvie CM, Forster RE, Blakemore WS, Morton JW. A standardized breath holding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Clin Invest. 1957;36:1-17.

[41] Stokes DL, Maclntyre NR, Nadel JA. Nonlinear increases in diffusing capacity during exercise by seated and supine subjects. J Appl Physiol. 1981;51:858-863.

[42] Kronenberg RS, L'Heureux P, Ponto RA, et al. The effect of aging on lung perfusion. Ann Intern Med. 1972;76:413-421.

[43] Ross J, Dean E, Abboud RT. The effect of postural drainage positioning on ventilation homogeneity in healthy subjects. Phys Ther. 1992;72:794-799.

[44] Roussos CS, Fixley M, Genest J, et al. Voluntary factors influencing the distribution of inspired gas. Am Rev Respir Dis. 1977;116:457-467.

[45] Roussos CS, Martin RR, Engel LA. Diaphragmatic contraction and the gradient of alveolar expansion in the lateral posture. J Appl Physiol. 1977;43:32-38.

[46] Chevrolet JC, Emrich J, Martin RR, Engel LA. Voluntary changes in ventilation distribution in the lateral posture. Respir Physiol. 1989;38: 313-323.

[47] Kaneko K, Milic-EmiliJ, Dolovich MB, et al. Regional distribution of ventilation and perfusion as a function of body position. J Appl Physiol. 1966;21:767-777.

[48] Hazlett DR, Watson RL. Lateral position test: a simple, inexpensive, yet accurate method of studying the separate functions of the lungs. Chest. 1971;59:276-279.

[49] Roussos CS, Fukuchi Y, Macklem PT, Engel LA. Influence of diaphragmatic contraction on ventilation distribution in horizontal man. J Appl Physiol. 1976;40:417-424.

F Manning, MD, PT, is Family Medicine Resident, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada. This study was completed in partial fulfillment of the requirements for Dr Manning's Master of Science degree.

E Dean, PhD, PT, is Professor, School of Rehabilitation Sciences, University of British Columbia, T325-2211 Wesbrook Mall, Vancouver, British Columbia, Canada V6T 1Z3 ([e-mail]). Address all correspondence to Dr Dean.

J Ross, PT, is Section Head, Critical Care, Rehabilitation Services, Vancouver Hospital, Vancouver, British Columbia, Canada.

RT Abboud, MD, FRCPC, is Professor, Division of Respiratory Medicine, Faculty of Medicine, University of British Columbia, and Director, Lung Function Laboratory, Vancouver Hospital and Health Sciences Centre, Vancouver, British Columbia, Canada.

This study was approved by the Ethics Review Committee of the University of British Columbia.

This study was supported, in part, by funding from the Canadian Lung Association.

This article was submitted December 4, 1997, and was accepted January 11, 1999.