Assessing the effects of inspiratory muscle training in a patient with unilateral diaphragm dysfunction

Case history

A 55-year-old man was referred to the outpatient pulmonary department of our hospital because of dyspnoea during exertion and when bending forward, which had been present for at least 6 months. He reported experiencing severe symptoms of breathlessness and many of his daily activities had to be adapted or interrupt due to symptoms (as documented by the Baseline Dyspnea Index (BDI)) (table 1). Recent infectious episodes or episodes of neck or shoulder pain were absent. His medical history included systemic arterial hypertension, obesity (body mass index (BMI) 36 kg·m⁻²), and obstructive sleep apnoea for which he was treated with night-time continuous positive airway pressure therapy (8 cmH₂O). He was a former smoker (18 pack-years) who quit smoking 15 years ago. 8 months ago, he underwent abdominal surgery (transabdominal epigastric hernia repair). The presence of cardiopulmonary disease and other aetiologies, such as neuromuscular disease, was excluded. Chest radiograph showed an elevated left hemidiaphragm and impaired left phrenic nerve conduction (i.e. increased latency and compound muscle action potential (CMAP) duration) after electrical stimulation (table 1) [1].

During baseline assessments, upright forced vital capacity (FVC) was 60% of predicted, while the ratio between forced expiratory volume in 1 s and FVC (FEV₁/FVC) was 80% [2]. Static lung volumes and capacities were also decreased [3]. Taken together the results of these assessments suggest a restrictive pulmonary function impairment without expiratory flow limitation. When the FVC manoeuvre was performed in the supine position, a reduction by 24% was observed relative to sitting without severe symptoms of orthopnoea occurring during the supine assessment (table 1).

The maximal inspiratory (P_Imax) and expiratory pressures (P_Emax) were −108 cmH₂O (97% predicted) and 213 cmH₂O (176% predicted), respectively (table 2) [4, 5].

Based on these findings the diagnosis of unilateral diaphragm dysfunction (UDD) was made.

Question

Could an inspiratory muscle training (IMT) programme, in this patient with almost normal P_Imax, be indicated to improve diaphragm function and exertional dyspnoea?

Answer

To answer this question, additional assessments were performed both before and after a 6-month IMT period.

Transdiaphragmatic pressure (P_di) obtained during maximal voluntary sniff manoeuvres (sniff P_di) was 74 cmH₂O (51% predicted) (table 2 and figure 1a) [5]. A sniff P_di of <80 cmH₂O is indicative of significant diaphragmatic weakness [6]. Bilateral magnetic phrenic nerve stimulation resulted in a twitch P_di of 10.8 cmH₂O (figure 1b). During unilateral stimulation the obtained twitch P_di values were 2.4 and 6.4 (for left and right stimulation, respectively) (table 2). Twitch P_di values of <20 cmH₂O during bilateral stimulation or <10 cmH₂O during unilateral stimulation are indicative of significant diaphragmatic weakness [6]. In addition, a paradoxical response in gastric pressure (P_ga) (i.e. negative relative to end-expiratory P_ga) was observed during the maximal sniff manoeuvres (sniff P_ga) (table 2 and figure 1a).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Illustration of P_di, P_ga and P_oes pressure curves during a sniff manoeuvre a) pre- and b) post-IMT; and P_di during bilateral magnetic phrenic stimulation c) pre- and d) post-IMT.

Furthermore, a maximal incremental (10 W·min⁻¹) cardiopulmonary exercise test (CPET) (Vmax229d; SensorMedics, Yorba Linda, CA, USA) was performed on a cycle ergometer (Ergometrics 800S; SensorMedics). The patient achieved a peak work rate of 160 W (65% predicted), maximal oxygen consumption (V′_O2max) of 20.8 mL·kg⁻¹·min⁻¹ (69% predicted) and a maximal heart rate of 140 beats per min (82% of predicted maximum) [7, 8]. The peak minute ventilation (V′_E) was 59 L·min⁻¹ (60% MVV). The patient rated the intensity of his perceived “breathing discomfort” (dyspnoea) and leg discomfort on a modified Borg scale (0–10) every minute during the test [9]. He indicated 9 for dyspnoea and 7 for leg effort at peak exercise and selected dyspnoea as the main reason (80% relative to 20% leg effort) for stopping the test.

A subsequent constant work rate (CWR) endurance cycling test was performed at 70% of the peak work rate achieved during the CPET. The duration of loaded cycling was 5 min and 30 s and, again, the main reason for stopping exercise was dyspnoea (80% relative to 20% leg effort). During the CWR test, a combined diaphragm electromyogram (EMG_di)-electrode catheter with oesophageal and gastric balloons was used to assess the diaphragm activation (EMG_di/EMG_di,max), oesophageal (P_oes), gastric (P_ga) and transdiaphragmatic (P_di) pressures [10]. In addition, surface EMG signals of scalene (EMG_sca/EMG_sca,max) and rectus abdominus (EMG_abd/EMG_abd,max) muscles were recorded. The EMG_di/EMG_di,max and EMG_sca/EMG_sca,max were expressed relative to maximal activation obtained during inspiratory capacity manoeuvres. The EMG_abd/EMG_abd,max was expressed relative to the maximal activation obtained during the expiratory phase of a FVC manoeuvre [11]. The measures were assessed continuously and breath-by-breath measurements were averaged for each minute. At the end of the endurance test the EMG_di/EMG_di,max, EMG_sca/EMG_sca,max and EMG_abd/EMG_abd,max were 78%, 37% and 39%, respectively (figure 2 a–c). The P_di, P_oes and P_ga swings were 35 cmH₂O, −102 cmH₂O and −68 cmH₂O, respectively (figure 2 d–f). Of note, the P_ga swings were negative, indicating paradoxical P_ga responses during tidal (exercise) breathing.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Percentage of maximal muscle activation of a) diaphragm (EMGdi/EMGdi, max), b) scalene (EMGsca/EMGsca, max), c) abdominal (EMGabd/EMGSabd, max) muscles, and d) P_di, e) P_oes and f) P_ga, during constant work rate exercise pre- and post-IMT. Dotted lines indicate the end of the endurance test for pre- and post-IMT. Red arrows indicate swing pressures pre-IMT and black arrows indicate swing pressures post-IMT.

These additional assessments further supported the diagnosis of UDD, identified specific diaphragm weakness and quantified the degree of exertional dyspnoea.

Treatment

IMT was performed using an electronic tapered flow resistive loading device (POWERbreathe^®KH1, HaB International Ltd., Southam, UK). The programme was based on a high intensity IMT protocol as described previously [11–13]. It consisted of two to three daily sessions of 30 full vital capacity inspirations (4–5 min per session), performed 7 days a week for 6 months. The patient was able to start the training against an external load of 45 cmH₂O (peak pressure during inspiration, ∼40% P_Imax), which was initially selected as the highest load that allowed him to perform nearly full vital capacity inspirations. Once a month supervised training at the outpatient clinic was performed. This allowed review of the stored data on quantity and quality of home-based training sessions and comparison of those data with the results obtained during the supervised session. Evaluated training parameters included average mean pressure generated per breath (cmH₂O), average mean power per breath (W), average inspiratory volume per breath (L) and total external mechanical work of breathing per session (J) [14]. The training load was progressively increased over the course of the training period from 45 cmH₂O up to 62 cmH₂O. The patient was very motivated to perform his breathing exercises and the perceived improvement in symptoms during the initial weeks of his training period probably contributed to increase his motivation. He completed a total of 420 sessions over the 6-month period (2.3 sessions per day).

Post-IMT assessment

Exercise capacity

During the CPET the patient achieved the same peak work rate (160 W) and similar V′_O2max (20.6 mL·kg⁻¹·min⁻¹, 68% predicted) [8] in comparison to baseline assessments.

The endurance cycling test was repeated at the same work rate as at baseline. Exercise time increased from 6 to 20 min, thereby reaching the maximal pre-specified duration of the post-intervention test. The dyspnoea perception decreased from 9 to 2 at the timepoint at which the baseline test had to be interrupted by the patient (figure 3). For a similar V′_E at iso-work rate during cycling, less diaphragm (EMG_di/EMG_di,max) and scalene (EMG_sca/EMG_sca,max) activation were observed to generate a similar inspiratory effort (i.e. inspiratory P_oes and P_di) (figures 2 and 3). This comparable inspiratory effort resulted in higher tidal volume and lower respiratory frequency throughout exercise (figures 2 and 3). Both the reduction in the activation for a given inspiratory effort and improved mechanical response (slower and deeper breathing pattern) for a given inspiratory effort contributed to improve neuromechanical coupling. This probably contributed to reduce the intensity of perceived breathing discomfort (dyspnoea) during exercise [18]. In addition, inspiratory P_ga remained paradoxical during endurance cycling, indicating the persisting diaphragm defect.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

a) V′_E, b) Respiratory rate, c) Borg score for dyspnoea, d) Peak inspiratory flow, e) tidal volume (V_T) and f) Borg score for leg effort during constant work rate exercise pre- and post-IMT. Dotted lines indicate the end of the endurance test for pre- and post-IMT.

On the expiratory side, a decrease in EMG_abd/EMG_abd,max was observed post-IMT and both expiratory P_oes and P_ga swings decreased by about 50%, mainly due to the decrease in positive expiratory pressure generation (figure 2). This reduced reliance on abdominal muscle recruitment for active expiration suggests decreased reliance on passive recoil forces during inspiration after the intervention [19].

Considering these results, we can conclude that, for this patient, the IMT programme was effective to reduce exertional dyspnoea without improving diaphragm function. Although the IMT did not affect isolated diaphragm function, the improvements in global inspiratory muscle function and breathing pattern during exercise decreased the symptoms, indicating that this patient has benefited from the IMT programme despite his almost normal baseline P_Imax and persistent UDD. In the initial assessment, the P_Imax measurement did not detect respiratory weakness. The P_Imax manoeuvre is a global measurement reflecting global inspiratory muscle strength (combined measurement of diaphragmatic and non-diaphragmatic inspiratory muscle strength). This fact reinforces the need for specific assessments of diaphragmatic strength to identify UDD. Twitch P_di is therefore the preferred assessment to confirm diagnosis of diaphragm weakness [5].

Additional discussion points

The patient self-reported that he had neither consciously increased sport participation nor his participation in physical activities at work or during leisure time during the study period. In line with this, the absence of changes in peak work rate and V_O2max during the incremental cycling test, and similar V′_E at a given work rate during the CWR endurance cycling test, make it unlikely that improvements in cardiopulmonary fitness might have contributed to his reductions in exertional dyspnoea symptoms. In addition, no changes in body weight were observed, excluding this factor as a contributor to improvements in respiratory muscle function and symptoms.

Given the patient's continuing motivation and the fact that we observed persisting improvements in both training parameters and respiratory muscle function it was collaboratively decided to continue with the IMT intervention after the 6-month study period. In case a plateau in the training parameters or decrease in motivation are observed, it can be decided to either reduce the frequency to once daily or once every 2 days while maintaining intensity (i.e. external load) in order to maintain benefits.

How the specific effects of IMT that were observed in this specific case relate to the general population of patients with UDD is unclear. Since symptoms vary according to clinical presentation, associated conditions and the possibility of spontaneous recovery in some cases, the respiratory muscle responses to training may also vary. Despite the promising findings in this case, further studies, in addition to the few (mostly case) studies that have been published to date [20–22], are needed to elucidate the specific mechanisms by which IMT can impact on the interaction between the diaphragm and non-diaphragmatic respiratory muscles during exercise breathing in patients with UDD.