Eur J Appl Physiol (2001) 85: 593±598 DOI 10.1007/s004210100469
O R I GI N A L A R T IC L E
Eric Verin á Christian Delafosse á Christian Straus Capucine MoreÂlot-Panzini á SergeõÈ Avdeev Jean-Philippe Derenne á Thomas Similowski
Effects of muscle group recruitment on sniff transdiaphragmatic pressure and its components Accepted: 17 April 2001 / Published online: 5 September 2001 Ó Springer-Verlag 2001
Abstract Measuring maximal sni pressures is an easy way of assessing inspiratory muscle strength. During a static manoeuvre, the pattern of inspiratory muscle recruitment during a sni can vary from one individual to another. We therefore assessed how voluntarily changing muscle recruitment would aect sni oesophageal, gastric and transdiaphragmatic pressures (Pes,sn, Pga,sn and Pdi,sn, respectively). Ten normal subjects (age 27± 38 years) performed natural sni manoeuvres (``nat''), and preferentially diaphragmatic (``dia'') or extradiaphragmatic (``extradia'') sni manoeuvres, after having learnt to dissociate between the inspiratory muscle groups. Abdominal displacements were monitored using a belt-mounted strain gauge. Natural patterns of muscle recruitment varied among subjects. On average, Pes,sn,nat was [median (range)] 81 (21±105) cmH2O. All of the subjects were able to modify inspiratory muscle recruitment voluntarily. Pes,sn was not signi®cantly affected by the type of manoeuvre performed, as opposed to Pdi,sn, which, as expected, increased with both the diaphragmatic and extradiaphragmatic manoeuvres
E. Verin á C. Delafosse á C. Straus á C. MoreÂlot-Panzini S. Avdeev á J.-P. Derenne á T. Similowski Laboratoire de Physiopathologie Respiratoire et Unite de ReÂanimation, Service de Pneumologie, Groupe Hospitalier PitieÂ-SalpeÃtrieÁre, Assistance Publique-HoÃpitaux de Paris, 47±83, Bd de l'HoÃpital, 75651 Paris Cedex 13, France E-mail:
[email protected] Tel.: +33-1-42176761 Fax: +33-1-42176708 E. Verin Service de Physiologie Respiratoire, Centre Hospitalier Universitaire de Rouen, Rouen, France C. Straus Service Central d'Explorations Fonctionnelles Respiratoires, Groupe Hospitalier PitieÂ-SalpeÃtrieÁre, Assistance Publique-HoÃpitaux de Paris, Paris, France C. Straus á J.-P. Derenne á T. Similowski UPRES EA 2397, Universite Paris VI Pierre et Marie Curie, Paris, France
[Pdi,sn,dia 132 (99±157) cmH2O, Pdi,sn,extradia 96 (50± 146) cmH2O, P<0.05]. Whatever the manoeuvre, there was no correlation between Pes and Pdi, but Pga and Pdi were correlated during both the diaphragmatic (r=0.82, P<0.05) and the extradiaphragmatic manoeuvre (r=0.70, P< 0.05). Pes,sn may have limitations as an index of diaphragm function, but by showing its independence from inspiratory muscle recruitment, this study contributes to its validation as a robust index of global inspiratory muscle strength that is particularly well suited for follow-up studies. This should extend to Pes,sn substitutes measured at the airway opening. Keywords Diaphragm á Respiratory muscles á Sni test á Respiratory pressures á Pulmonary function tests
Introduction Measuring inspiratory pressures during a maximal sni manoeuvre (Laroche et al. 1988; Miller et al. 1985) is a popular way of assessing inspiratory muscle strength in both physiological and clinical settings, a trend recently accentuated by the advent of the very simple sni nasal inspiratory pressure (SNIP) technique (Heritier et al. 1994) and the publication of the corresponding reference values (Fitting et al. 1996; Uldry and Fitting 1995). One of the main advantages of the sni test is that because it involves a natural manoeuvre, it is easy to understand and to perform by most people. In addition, as a dynamic manoeuvre, it is easily performed in ``dicult'' conditions (e.g. during exercise). Nevertheless, the routine practice of inspiratory muscle testing has made it obvious that not all subjects or patients use the same pattern of muscle activation to sni maximally, although this does not seem to have been the subject of systematic studies. It is thus conceivable that relatively high intrathoracic pressures can be generated during snis performed with or without the diaphragm, as is the case during static inspiratory eorts (De Troyer and Estenne 1981). Depending on the lower limit of normal values
594
considered, which can be rather low (Evans et al. 1993), sni oesophageal pressures in patients with diaphragm paralysis or diaphragm weakness are dicult to interpret. In addition, a maximal sni recruits most of the inspiratory muscle groups, but also the abdominal muscles (Nava et al. 1993). The impairment of expiratory muscles could thus alter the results of a test aimed at evaluating inspiratory muscles. This study, relying on the possibility for most normal subjects to preferentially activate the diaphragm or extradiaphragmatic muscles during inspiration (Fitting et al. 1988; Similowski et al. 1998), was designed to assess the impact of dierent patterns of muscle recruitment on sni pressures.
Methods Subjects Ten healthy volunteers were studied (all males, aged 27±38 years), of which one had considerable experience of respiratory muscle studies and the others were either naive or quasi-so. The study was a by-product of another one for which the relevant legal and ethical clearances had been obtained. The subjects were duly informed about the study and gave their written consent to participate.
as associating a decrease in Pes with an increase in Pga and an increase in AB circumference. An ``extradiaphragmatic'' inspiration was de®ned as associating a decrease in Pes with no or little rise in Pga and a decrease in AB circumference. Once the subjects had mastered the manoeuvres, they were asked to perform ten ``diaphragmatic'' snis and ten ``extradiaphragmatic'' snis, in random order. If the tenth sni was the best of a series, additional manoeuvres were recorded until the value dropped. An interval of 3±5 normal breaths was always allowed between two consecutive sni manoeuvres. Data analysis The criteria used to select suitable snis for analysis were the occurrence of a regular upstroke and a sharp peak pressure (Heritier et al. 1994; Uldry and Fitting 1995). In addition, snis were retained for analysis only when the continuous monitoring of AB circumference con®rmed that they had been initiated from a functional residual capacity (FRC) and, when appropriate, that the target manoeuvre had been adequate. Sni pressure swings were measured from baseline to peak. Their dynamics were described in terms of time to peak pressure (Ti), total relaxation time (Te) and total sni time (Ttot). AB displacements were analysed qualitatively (expansion or de¯ation). The three best snis of a series were taken into account to study the intra-subject variability, and their mean value was used for comparisons between manoeuvres. Statistical analysis
Measurements Oesophageal and gastric pressures (Pes and Pga, respectively) were measured with two standard 80-cm balloon catheters (Marquat, Boissy-Saint-LeÂger, France) connected to linear pressure transducers (200 cmH2O, Validyne, Northridge, Calif.,USA) and passed through one nostril after topical anaesthesia of the nasal mucosa. Transdiaphragmatic pressure (Pdi) was obtained o-line by subtracting Pes from Pga. Abdominal wall displacements (AB) were assessed using a piezo-electric mechanical sensor (Nihon Kohden, Tokyo, Japan) attached to an elastic belt.
All statistical analyses were performed using the SuperAnova 4.5 software (Abacus Concept, Berkeley, Calif., USA) running on an Apple Macintosh computer. Because of large standard deviations (Table 1), a non-parametric type of analysis was chosen for maximal safety. The eect of the type of manoeuvre on sni pressures was studied using a Friedman test, followed when appropriate by a Mann-Whitney U-test. Dierences were considered signi®cant when the probability P of a type I error was 0.05 or less (i.e. P<0.05). The intra-subject variability of sni pressures was assessed by computing variances and mean coecients of variation. Correlations between Pes, Pga and Pdi were assessed using the least square method.
Procedures The subjects were ®rst asked to perform ten maximal snis without instructions other than the standard ones (``sni as sharply and strongly as you can, from the end of an normal expiration''). They were then taught to perform various inspiratory manoeuvres, sni included, with and without their diaphragm, using visual feedback from Pes, Pga and AB. A ``diaphragmatic'' inspiration was de®ned
Table 1 Eects of the type of inspiratory muscle recruitment on the pressure components of the sni. Data are presented as the median (range). (Pes,sn Sni oesophageal pressure, Pga,sn sni gastric pressure, Pdi,sn sni transdiaphragmatic pressure, Pes/Pdi ratio of oesophageal pressure to transdiaphragmatic pressure)
Suxes The sux ``sn'' henceforth denotes a sni-related pressure swing, and is completed by ``nat'', ``dia'', or ``extradia'' to designate the natural, diaphragmatic and extradiaphragmatic manoeuvres, respectively.
Pressure component
Variable
Pes,sn
cmH2O Variance (cmH2O)2 Coecient of variation (%)
81 (21±105) 26 (2±101) 8.1 (1.5±33)
79 (54±131) 20 (0±60) 5.6 (0±12.6)
76 (49±108) 17 (0.4±87) 4.9 (1±13)
Pga,sn
cmH2O Variance (cmH2O)2 Coecient of variation (%)
27 (±15±86) 42 (1.9±202) 20 (3±82)
53 (21±81)* 76 (17±281) 13 (1.5±39)
21 (±15±68) 42 (4±163) 15.9 (±191±30)
Pdi,sn
cmH2O 110 (90±153) 132 (99±157)* 96 (50±148) 66 (6±172) 56.7 (2±323) 68 (10±147) Variance (cmH2O)2 Coecient of variation (%) 9.2 (3±26) 6 (1±22) 9.6 (3±15) 0.77 (0.2±1.2) 0.63 (0.45±0.87) 0.80 (0.51±1.25)
Pes/Pdi
Natural manoeuvre
Diaphragmatic manoeuvre
*P<0.05 for comparison between dia and the two other manoeuvres
Extradiaphragmatic manoeuvre
595
Results Fig. 1 describes the individual data in the ten subjects. The natural ``strategy'' used to perform a maximal sni varied among the subjects, but they were all able to modify inspiratory muscle recruitment voluntarily. It is
Fig. 1 Amplitude of oesophageal, gastric and transdiaphragmatic pressures (Pes, Pga, and Pdi, respectively) during maximal snis performed without instructions (A) and with preferential activation of the diaphragm (B) or of the extradiaphragmatic muscles (C). The direction of change in abdominal circumference is shown by arrows
noteworthy that it was not possible to distinguish the one subject with a lot of experience of respiratory manoeuvres from the others in terms of the intra-subject coecients of variability. The type of manoeuvre used to sni maximally did not signi®cantly aect Pes,sn or its intra-subject variability (Table 1). As expected, Pga,sn and Pdi,sn were signi®cantly higher during the dia-
596
phragmatic manoeuvre than during the other manoeuvres (Table 1). When the subjects purposely chose a given pattern of recruitment (``dia'' and ``extradia'' manoeuvres), the variability of Pga,sn and Pdi,sn was as low as that of Pes,sn. The dynamics of Pes was not modi®ed by the type of manoeuvre used (Fig. 2). Whatever the manoeuvre considered, there was no correlation between Pes,sn and Pdi,sn. Pga,sn and Pdi,sn were correlated during both the diaphragmatic manoeuvre (r=0.82, P<0.05) and the extradiaphragmatic manoeuvre (r=0.70, P<0.05) but not during the natural manoeuvre. When all data were pooled, there was no correlation between Pes,sn and Pdi,sn (P=0.08) but, as expected, Pga,sn and Pdi,sn were correlated (r=0.69; P<0.0001; Fig. 3).
Discussion The results of this study con®rm that uncoached normal subjects do use dierent strategies to perform sni manoeuvres. Its major ®nding lies in the lack of correlation between Pes,sn and Pdi,sn when the latter is modi®ed by the ``sning'' strategy. This implies that Pes,sn (and, by extension, sni pressures measured at the airway opening, e.g. SNIP) are not as safe indicators of diaphragm function as they are robust indices of the global inspiratory performance.
Fig. 2 Time decomposition (A) and rates of rise and fall (B) of the Pes component of the snis (Pes/Ti and Pes/Te, respectively) performed without instructions (black bars), with preferential activation of the diaphragm (white bars) or of extradiaphragmatic muscles (grey bars). The error bars represent 1 SD. (Ti Time to peak pressure, Te total relaxation time, Ttot total sni time)
Methodological considerations The pressure values found in our subjects are within the normal range (Evans et al. 1993), with low coecients of variation (Table 1): it thus seems reasonable to assume that the sni manoeuvres were adequately performed with maximal eorts. It is clear from Fig. 1 that the various individuals did not perform homogeneously with respect to the manoeuvres required from them. For example, subject numbers 3 and 4 increased their gastric pressure more during the extradiaphragmatic manoeuvre than during the diaphragmatic one. As this was associated with a decrease in AB circumference, it suggests strong abdominal muscle recruitment during the sni. We acknowledge that only measurement of the diaphragm electromyogram (EMG) through an oesophageal probe would have provided certainty about the reality of higher diaphragm activation during the ``diaphragmatic manoeuvres''. Nevertheless, the reality of the changes in muscle recruitment between the diaphragmatic and extradiaphragmatic manoeuvres is attested to by the corresponding AB displacements. An inspiratory AB expansion is the hallmark of a diaphragmatic contraction, whereas abdominal muscle contraction is ex-
Fig. 3 Relationship between the sni Pes and sni Pdi (A) and between the sni Pga and Pdi (B), during maximal snis performed without instructions (triangles, nat) and with preferential activation of the diaphragm (circles, dia) or of extradiaphragmatic muscles (squares, extra dia). In B, the dotted line indicates the regression line, all data points pooled, and the dashed lines denote the corresponding 95% con®dence interval of the regression
597
pected to decrease AB circumference. Therefore, the increase in Pga,sn (and in Pdi,sn) observed in our subjects with the ``diaphragmatic'' manoeuvre, being associated with AB expansion in all cases, must have been related mainly to diaphragm contraction. AB expansion does not rule out some degree of co-contraction of the abdominal muscles, which could have partially contributed to the rise in Pga,sn, but it suggests strongly that diaphragm activation pre-empted the activation of abdominal muscles. The reasoning is less simple regarding an inspiratory AB de¯ation, which can result from either a purely extradiaphragmatic inspiratory contraction or from a combined contraction of extradiaphragmatic inspiratory muscles, the diaphragm, and abdominal muscles. During the extradiaphragmatic manoeuvre, all subjects exhibited AB indrawing. In three cases (Fig. 1), this was associated with a negative Pga,sn, hence with very low or absent abdominal muscle activation and either a complete lack or a very low degree of diaphragm activation with respect to rib cage muscle activation. In the other cases, the conjunction of a positive Pga,sn and AB indrawing suggests a dominant action of abdominal muscles and their contribution to Pdi (Gandevia et al. 1990). The lower degree of diaphragm activation observed during the extradiaphragmatic manoeuvre than during the diaphragmatic manoeuvre is supported by the fact that Pga,sn,dia, associated with AB expansion, was almost systematically greater than Pga,sn,extradia. It could be hypothesised that the diaphragm was maximally recruited during maximal snis and that Pga,sn increased only due to abdominal contraction during the diaphragmatic sni, thereby increasing abdominal impedance. Nevertheless, it is unlikely that the subjects would have achieved maximal diaphragm activation during spontaneous, uncoached eorts (Hershenson et al. 1988). To sum up, we wish to emphasise that beyond the subjects' capability in performing the various manoeuvres, the important fact to consider is that changing the pattern of respiratory muscle recruitment resulted in major variations in Pdi,sn, mainly due to the contribution of Pga, and was not accompanied by changes in Pes,sn. Snis were retained for analysis only when the monitoring of the AB trace showed that they had been initiated from the same point in the respiratory cycle, namely FRC. Therefore, it is unlikely that changes in lung volume or AB con®guration played a signi®cant role in the changes observed in Pga,sn or Pdi,sn with the dierent manoeuvres. And the ®nal methodological issue is that the lower variability associated with Pes,sn as compared to Pga,sn or Pdi,sn probably represents a bias, because the snis retained to assess variability were selected as the best ones in terms of Pes. Possible mechanisms to explain the stability of Pes,sn The lack of in¯uence of inspiratory muscle recruitment on Pes,sn is a striking feature of this study. Several
tentative hypotheses to explain it could be proposed. One possibility would be the interruption of the sni inspiratory command by aerent inputs depending on the intrathoracic pressure, its rate of rise, or a combination of both types of information. Another possibility would be that the magnitude of Pes,sn that it is possible to develop in response to a maximal sni inspiratory command is limited by the mechanical properties of the lung and chest wall. The second hypothesis is made more likely by the ballistic-like nature of the sni manoeuvre, which, conversely to the case for static inspiratory manoeuvres, does not involve a sustained command. The lack of in¯uence of muscle recruitment on the time dynamics of Pes,sn, which we observed to proceed in the same direction, supports the idea that Pes,sn is determined by the central command. Both of the above mechanisms would easily explain our results if the limiting Pes,sn could be reached by either of the inspiratory muscle groups independently. Of note, this would imply a substantial underestimation of the actual inspiratory muscle strength in normal subjects. Implications for the interpretation of sni pressures The idea that diaphragm function may be better evaluated by sni manoeuvres than by static inspiratory efforts is relatively common. This contention, which is supported by the ®nding that the peak amplitude of diaphragm EMG activity can be higher during snis than during certain static inspiratory eorts (Nava et al. 1993), stems originally from the ®nding that Pdi,sn can be higher than maximal static transdiaphragmatic pressure (Pdi,max; Laroche et al. 1988; Miller et al. 1985), although con¯icting results have been reported in normal subjects (Heijdra et al. 1993) and in patients (Heritier et al. 1991). Abdominal muscles can be strongly activated during a sni manoeuvre (this study; see also Nava et al. 1993), and such an activation during an inspiratory manoeuvre can contribute to the build-up of Pdi (Gandevia et al. 1990). Therefore the use of Pdi,sn as an index of diaphragm function must be interpreted with caution. Our ®ndings may explain some of the variance found in healthy volunteers. Polkey et al. (1997), for instance, reported a relationship between Pdi,sn and an increasing gastric pressure contribution (as evidenced by a falling Pes,sn/Pdi,sn ratio). In the present study, Pes was not dependent upon the type of manoeuvre used to perform the sni, whereas Pga was (Fig. 1, Table 1). It follows, therefore, that Pes,sn would not be correlated with Pdi,sn (Fig. 3). In other words, Pes,sn could not have been used to predict Pdi,sn in our subjects. To our knowledge, the value of Pes,sn in the prediction of Pdi,sn has not been speci®cally assessed in the literature. Indeed, several studies have shown that sni mouth pressure, Pes,sn or SNIP are adequate methods with which to distinguish between patients with inspiratory muscle weakness and normal subjects (Fitting et al. 1996; Hughes et al. 1998; Laroche
598
et al. 1988), and that there is a good ``operative'' agreement between these measurements and Pdi,sn. However, the numerical relationship between the two parameters has not been studied in detail. A concordance between Pdi,sn and Pes,sn in terms of diagnostic value, in spite of the lack of a numerical relationship, is easy to explain because diseases aecting a given inspiratory muscle group in isolation are relatively rare. In neuromuscular diseases in particular, several muscle groups are generally involved together. The independence of Pes,sn with respect to muscle group recruitment can in fact be viewed as a major advantage for the sni test, because it should make it a good marker of the progression of diseases that aect respiratory muscles as a whole (see the recent study by Fitting et al. 1999). Problems could arise when studying diseases involving the diaphragm independently of other respiratory muscles (e.g. neuralgic amyotrophy) or vice versa (e.g. abdominal muscle dysfunction in paraplegia or rib cage muscle dysfunction in lesions of the upper spinal cord). In such conditions, Pes,sn could be relatively preserved, leading to an underestimation of the degree of dysfunction of the involved muscle. Pes,sn could also underestimate inspiratory muscle involvement in patients with slowly installed diaphragm dysfunction who have developed compensatory mechanisms at the level of the inspiratory neck muscles (Attali et al. 1997). A very good illustration of the potential diagnostic importance of our ®ndings is provided by the patterns observed in subject number 10 (Fig. 1). When uncoached, this subject developed a Pdi,sn value of less than 85 cmH2O, which is below the lower limit of the normal range de®ned by Evans et al. (1993), but his Pes,sn was within the normal range. A negative Pga,sn and AB indrawing suggested weak diaphragm activation. When taught to use his diaphragm (positive Pga,sn, AB expansion), this subject developed a perfectly normal Pdi,sn, without any change in Pes,sn. Testing him with sni-Pdi without looking at the Pes/ Pga components would have led to an undue suspicion of diaphragm weakness. In conclusion, Pes,sn seems to have limits as an indicator of diaphragm function, but does appear to be a robust global index of inspiratory muscle function, independent of their pattern of activation. Our ®ndings, however, must be veri®ed in diseased conditions. This is likely to apply to derived indices such as SNIP, and supports their value as useful follow-up tools. In practice, it appears important to con®rm that Pdi partitioning is unchanged when Pdi,sn is to be used as a repeated measure of strength. Work is needed to establish the bene®t of combining the assessment of abdominal displacement with the measurement of an inspiratory pressure during a sni, in both physiological and clinical studies.
Acknowledgements This study was supported by Association pour le DeÂveloppement et l'Organisation de la Recherche en Pneumologie (ADOREP), Paris, France. S. Avdeev was a scholar of a French-Russian exchange grant (SocieÂte de Pneumologie de Langue FrancËaise, Association FeÂdeÂrative Nationale pour le Traitement aÁ Domicile de l'Insusance Respiratoire Chronique, and Laboratoires Servier, Paris, France).
References Attali V, Mehiri S, Straus C, Salachas F, Arnulf I, Meininger V, Derenne J-P, Similowski T (1997) In¯uence of neck muscles on mouth pressure response to cervical magnetic stimulation. Am J Respir Crit Care Med 156:509±514 De Troyer A, Estenne M (1981) Limitations of measurement of transdiaphragmatic pressure in detecting diaphragmatic weakness. Thorax 36:169±174 Evans SA, Watson L, Cowley AJ, Johnston ID, Kinnear WJ (1993) Normal range for transdiaphragmatic pressures during snis with catheter mounted transducers. Thorax 48:750±753 Fitting JW, Bradley TD, Easton PA, Lincoln MJ, Goldman MD, Grassino A (1988) Dissociation between diaphragmatic and rib cage muscle fatigue. J Appl Physiol 64:959±965 Fitting JW, Heritier F, Uldry C (1996) Evaluation de la force musculaire inspiratoire par la pression nasale lors du sni. Rev Mal Respir 13:479±484 Fitting JW, Paillex R, Hirt L, Aebischer P, Schluep M (1999) Sni nasal pressure: a sensitive respiratory test to assess progression of amyotrophic lateral sclerosis. Ann Neurol 46:887±893 Gandevia SC, McKenzie DK, Plassman BL (1990) Activation of human respiratory muscles during dierent voluntary manoeuvres. J Physiol (Lond) 428:387±403 Heijdra YF, Dekhuijzen PN, van Herwaarden CL, Folgering HT (1993) Dierences between sni mouth pressures and static maximal inspiratory mouth pressures. Eur Respir J 6:541±546 Heritier F, Perret C, Fitting JW (1991) Maximal sni mouth pressure compared with maximal inspiratory pressure in acute respiratory failure. Chest 100:175±178 Heritier F, Rahm F, Pasche P, Fitting JW (1994) Sni nasal inspiratory pressure. A noninvasive assessment of inspiratory muscle strength. Am J Respir Crit Care Med 150:1678±1683 Hershenson MB, Kikuchi Y, Loring SH (1988) Relative strengths of the chest wall muscles. J Appl Physiol 65:852±862 Hughes PD, Polkey MI, Kyroussis D, Hamnegard CH, Moxham J, Green M (1998) Measurement of sni nasal and diaphragm twitch mouth pressure in patients. Thorax 53:96±100 Laroche CM, Mier AK, Moxham J, Green M (1988) The value of sni esophageal pressures in the assessment of global inspiratory muscle strength. Am Rev Respir Dis 138:598±603 Miller JM, Moxham J, Green M (1985) The maximal sni in the assessment of diaphragm function in man. Clin Sci 69:91±96 Nava S, Ambrosino N, Crotti P, Fracchia C, Rampulla C (1993) Recruitment of some respiratory muscles during three maximal inspiratory manoeuvres. Thorax 48:702±707 Polkey MI, Harris ML, Hughes PD, Hamnegard CH, Lyons D, Green M, Moxham J (1997) The contractile properties of the elderly human diaphragm. Am J Respir Crit Care Dis 155:1560±1564 Similowski T, Straus C, Attali V, Duguet A, Derenne JP (1998) Cervical magnetic stimulation as a method to discriminate between diaphragm and rib cage muscle fatigue. J Appl Physiol 84:1692±1700 Uldry C, Fitting JW (1995) Maximal values of sni nasal inspiratory pressure in healthy subjects. Thorax 50:371±375
