Motoneurone synchronization for intercostal and abdominal muscles: interneurone influences in two different species.

The contribution of branched-axon monosynaptic inputs in the generation of short-term synchronization of motoneurones remains uncertain. Here, synchronization was measured for intercostal and abdominal motoneurones supplying the lower thorax and upper abdomen, mostly showing expiratory discharges. Synchronization in the anaesthetized cat, where the motoneurones receive a strong direct descending drive, is compared with that in anaesthetized or decerebrate rats, where the direct descending drive is much weaker. In the cat, some examples could be explained by branched-axon monosynaptic inputs, but many others could not, by virtue of peaks in cross-correlation histograms whose widths (relatively wide) and timing indicated common inputs with more complex linkages, e.g., disynaptic excitatory. In contrast, in the rat, correlations for pairs of internal intercostal nerves were dominated by very narrow peaks, indicative of branched-axon monosynaptic inputs. However, the presence of activity in both inspiration and expiration in many of the nerves allowed additional synchronization measurements between internal and external intercostal nerves. Time courses of synchronization for these often consisted of combinations of peaks and troughs, which have never been previously described for motoneurone synchronization and which we interpret as indicating combinations of inputs, excitation of one group of motoneurones being common with either excitation or inhibition of the other. Significant species differences in the circuits controlling the motoneurones are indicated, but in both cases, the roles of spinal interneurones are emphasised. The results demonstrate the potential of motoneurone synchronization for investigating inhibition and have important general implications for the interpretation of neural connectivity measurements by cross-correlation.

A smartphone oximeter with a fingertip probe for use during exercise training: usability, validity and reliability in individuals with chronic lung disease and healthy controls.

BACKGROUND AND AIM: Telehealth is a strategy to expand the reach of pulmonary rehabilitation (PR). Smartphones can monitor and transmit oxygen saturation (SpO2) and heart rate (HR) data to ensure patient safety during home-based or other exercise. The purpose of this study was to evaluate the usability, validity and reliability of a Kenek O2 pulse oximeter and custom prototype smartphone application (smartphone oximeter) during rest and exercise in healthy participants and those with chronic lung disease. METHODS: Fifteen individuals with chronic lung disease and 15 healthy controls were recruited. SpO2 and HR were evaluated at rest and during cycling and walking. SpO2 was valid if the mean bias was within +±2%, the level of agreement (LoA) was within ±4%; HR was valid if the mean bias was within ±5 beats per min (bpm), LoA was within ±10bpm. Usability was assessed with a questionnaire and direct observation. RESULTS: The smartphone oximeter was deemed easy to use. At rest, SpO2 measures were valid in both groups (bias <2%, lower bound LoA -2 to 3%). During exercise, SpO2 measurement did not meet validity and reliability thresholds in the patients with chronic lung disease, but was accurate for the healthy controls. HR recording during exercise or rest was not valid (LoA>10bpm) in either group. CONCLUSIONS: The smartphone oximeter did not record HR or SpO2 accurately in patients with chronic lung disease during exercise, although SpO2 was valid at rest. During exercise, patients with chronic lung disease should pause to ensure greatest accuracy of SpO2 and HR measurement.

Reliability and validity of the Brief Pain Inventory in individuals with chronic obstructive pulmonary disease.

BACKGROUND: Pain is prevalent in chronic obstructive pulmonary disease (COPD) and the Brief Pain Inventory (BPI) appears to be a feasible questionnaire to assess this symptom. However, the reliability and validity of the BPI have not been determined in individuals with COPD. This study aimed to determine the internal consistency, test-retest reliability and validity (construct, convergent, divergent and discriminant) of the BPI in individuals with COPD. METHODS: In order to examine the test-retest reliability, individuals with COPD were recruited from pulmonary rehabilitation programmes to complete the BPI twice 1 week apart. In order to investigate validity, de-identified data was retrieved from two previous studies, including forced expiratory volume in 1-s, age, sex and data from four questionnaires: the BPI, short-form McGill Pain Questionnaire (SF-MPQ), 36-Item Short Form Survey (SF-36) and Community Health Activities Model Program for Seniors (CHAMPS) questionnaire. RESULTS: In total, 123 participants were included in the analyses (eligible data were retrieved from 86 participants and additional 37 participants were recruited). The BPI demonstrated excellent internal consistency and test-retest reliability. It also showed convergent validity with the SF-MPQ and divergent validity with the SF-36. The factor analysis yielded two factors of the BPI, which demonstrated that the two domains of the BPI measure the intended constructs. The BPI can also discriminate pain levels among COPD patients with varied levels of quality of life (SF-36) and physical activity (CHAMPS). CONCLUSION: The BPI is a reliable and valid pain questionnaire that can be used to evaluate pain in COPD. SIGNIFICANCE: This study formally established the reliability and validity of the BPI in individuals with COPD, which have not been determined in this patient group. The results of this study provide strong evidence that assessment results from this pain questionnaire are reliable and valid.

Patient and health care professional perspectives on using telehealth to deliver pulmonary rehabilitation.

The objective of this study was to identify the necessary features of pulmonary telerehabilitation (P-TR) from the perspectives of individuals living with chronic lung disease and health care professionals (HCPs) who deliver pulmonary rehabilitation (PR). Focus groups were carried out with patients ( n = 26) and HCPs ( n = 26) to elicit and explore their opinions about the critical elements of in-person PR and ideas for how these elements could be supported using technology. A questionnaire was used to assess technology use, PR experience, and general health status. Four key elements of PR were identified as critical to P-TR: the social aspect of PR; communicating with HCPs for education and support; using biosensors for monitoring and promoting self-knowledge; and the evolution of support with progress over time. A range of technology-enabled devices and programs were suggested as means to recreate aspects of these integral elements. Consultations with patients and HCPs suggest that users are interested in technology and want to ensure it recreates the important aspects of PR. Patients and HCPs identified similar key elements for P-TR. The opinions and suggestions of patients and HCPs should be the driving force of innovation if P-TR is to succeed in improving health outcomes.

Connections between expiratory bulbospinal neurons and expiratory motoneurons in thoracic and upper lumbar segments of the spinal cord.

Cross-correlation of neural discharges was used to investigate the connections between expiratory bulbospinal neurons (EBSNs) in the caudal medulla and expiratory motoneurons innervating thoracic and abdominal muscles in anesthetized cats. Peaks were seen in the cross-correlation histograms for around half of the EBSN-nerve pairs for the following: at T8, the nerve branches innervating internal intercostal muscle and external abdominal oblique muscle and a more distal branch of the internal intercostal nerve; and at L1, a nerve branch innervating internal abdominal oblique muscle and a more distal branch of the ventral ramus. Fewer peaks were seen for the L1 nerve innervating external abdominal oblique, but a paucity of presumed α-motoneuron discharges could explain the rarity of the peaks in this instance. Taking into account individual EBSN conduction times to T8 and to L1, as well as peripheral conduction times, nearly all of the peaks were interpreted as representing monosynaptic connections. Individual EBSNs showed connections at both T8 and L1, but without any discernible pattern. The overall strength of the monosynaptic connection from EBSNs at L1 was found to be very similar to that at T8, which was previously argued to be substantial and responsible for the temporal patterns of expiratory motoneuron discharges. However, we argue that other inputs are required to create the stereotyped spatial patterns of discharges in the thoracic and abdominal musculature.

Falls in people with chronic obstructive pulmonary disease: an observational cohort study.

STUDY OBJECTIVE: To investigate incidence, risk factors and impact of falls on health related quality of life (HRQoL) in patients with chronic obstructive pulmonary disease (COPD). DESIGN: Observational cohort study. METHODS: Patients completed these questionnaires at baseline and at 6-months: Medical Outcomes Study Short Form 36 (SF-36), Chronic Respiratory Questionnaire (CRQ), Activities Balance Confidence (ABC) Scale and a form to record demographic data, medications, co-morbidities, oxygen use, acute exacerbations, fall history and assistive device use. Physical activity was measured with the Physical Activity Scale for the Elderly (PASE) only at baseline. Fall incidence was monitored through monthly fall diaries. Patients were categorized as non-fallers (0 falls) or fallers (≥ 1 falls). RESULTS: Data from 101 patients with a forced expiratory volume in 1 s of 46.4 ± 21.6% predicted were analyzed. Thirty-two patients (31.7%) reported at least one fall during the 6-months. Fall incidence rate was 0.1 (95% CI: 0.06-0.14) falls per person-month. Fallers tended to be older (p = 0.04), female (p = 0.04) and oxygen dependent (p = 0.02), have a history of previous falls (p < 0.001), more co-morbidities (p = 0.007) and take more medications (p = 0.001). Previous falls (OR = 7.36; 95% CI: 2.39-22.69) and diagnosis of coronary heart disease (OR = 7.07; 95% CI: 2.14-23.36) were the most important predictors of falls. The Dyspnea Domain of the CRQ declined significantly more (p = 0.02) in the fallers group at 6-months. CONCLUSIONS: Patients with COPD have a high susceptibility to falls, which is associated with a worsening of dyspnea perception as related to HRQoL. Fall prevention programs in COPD are recommended.

Postural Control Is Impaired in People with COPD: An Observational Study.

PURPOSE: We investigated deficits in postural control and fall risk in people with chronic obstructive pulmonary disease (COPD). METHOD: Twenty people with moderate to severe COPD (mean age 72.3 years, standard deviation [SD] 6.7 years) with a mean forced expiratory volume in 1 second (FEV(1)) of 46.7% (SD 13%) and 20 people (mean age 68.2 years, SD 8.1) who served as a comparison group were tested for postural control using the Sensory Organization Test (SOT). A score of zero in any trial of the SOT was registered as a fall. On the basis of the SOT results, participants were categorized as frequent fallers (two or more falls) or as fallers (one fall). To explore the potential influence of muscle weakness on postural control, knee extensors concentric muscle torque was assessed with an isokinetic dynamometer. Physical activity level was assessed with the Physical Activity Scale for the Elderly. RESULTS: People with COPD showed a 10.8% lower score on the SOT (p=0.016) and experienced more falls (40) than the comparison group (12). The proportion of frequent fallers and fallers during the SOT was greater (p=0.021) in the COPD group (four of 10) than in the comparison group (two of seven). People with COPD showed deficits in knee extensors muscle strength (p=0.01) and a modest trend toward reduced physical activity level. However, neither of these factors explained the deficits in postural control observed in the COPD group. CONCLUSIONS: People with COPD show deficits in postural control and increased risk of falls as measured by the SOT. The deficits in postural control appear to be independent of muscle weakness and level of physical activity. Postural control interventions and fall risk strategies in the pulmonary rehabilitation of COPD are recommended.

Increased injury and intramuscular collagen of the diaphragm in COPD: autopsy observations.

Evidence for diaphragm injury in people with chronic obstructive pulmonary disease (COPD) has been reported, although the extent of injury and collagen accumulation post mortem have not previously been examined. In addition, it is not known whether the amount of injury and collagen are different in key regions of the diaphragm. The cross-sectional area of collagen and the percentage of abnormal myofibres in the post mortem diaphragm and psoas major were determined by computer-assisted image analysis of stained cross-sections of the diaphragm for collagen with picrosirius red and with haematoxylin and eosin for morphology. In the midcostal diaphragm of six subjects with COPD and six subjects with no significant respiratory disease, the COPD diaphragm displayed a greater cross-sectional area of collagen and percentage of abnormal myofibres (collagen: 24.2+/-1.0 versus 18.6+/-1.1%; injury: 28.4+/-7.2 versus 12.0+/-1.3%). In 18 patients with various respiratory conditions, the midcostal diaphragm displayed more collagen and abnormal myofibres than the crural diaphragm, while both costal and crural diaphragms displaying more collagen and abnormal myofibres than psoas major. This study reveals extensive injury and collagen accumulation in the chronic obstructive pulmonary diseased diaphragm, and reveals a regional pattern of injury and intramuscular collagen which may correspond to variations in diaphragm loading.

Granulocytosis and increased adhesion molecules after resistive loading of the diaphragm.

Upregulation of endothelial cell adhesion molecules, followed by an influx of granulocytes and macrophages, can contribute to exertion-induced skeletal muscle injury. The purpose of this study was to quantify circulating leukocyte subsets, diaphragm injury and infiltrating leukocyte subsets, and surface expression of vascular cell adhesion molecule (VCAM)-1 and intracellular adhesion molecule (ICAM)-1 in the diaphragm after inspiratory resistive loading (IRL). Eight New Zealand white rabbits underwent 1.5 h of IRL and seven control rabbits underwent a sham procedure. Blood samples, taken at baseline and 2, 6, 12, 24, 48 and 72 h after the onset of IRL or sham, showed that band cell counts had increased at 6 h post-IRL. Point counting of haematoxylin and eosin-stained cross-sections, sampled at 72 h post-IRL, showed greater injury in diaphragms from IRL rabbits compared with controls. Immunohistochemical processing showed increased expression of ICAM-1 and VCAM-1, and higher granulocyte and macrophage counts in IRL diaphragms than control diaphragms. Macrophages were the predominant inflammatory cells. Increased intracellular adhesion molecule-1 and vascular cell adhesion molecule-1 expression, and infiltration of granulocytes and macrophages may contribute to inspiratory resistive loading-induced diaphragm injury.

Free radical scavengers and diaphragm injury following inspiratory resistive loading.

Three groups of NZW rabbits were studied to examine the role of free radical scavengers in preventing diaphragm injury produced by inspiratory resistive load (IRL): control, IRL, and scavenger groups. An IRL (Pao: 45-55 cm H2O) was applied to the IRL and the scavenger groups on Day 1. Free radical scavengers (polyethylene glycol superoxide dismutase, N-acetylcysteine, and mannitol) were given (intravenously) to the scavenger group both before and after the IRL. All rabbits were killed on Day 3 to collect diaphragms. Point counting H&E-stained diaphragm x-sections indicated that abnormal diaphragm muscle in the IRL group was significantly greater than control (p < 0.01). However, it was significantly lower in the scavenger group than the IRL group (p < 0.05) and it did not differ from control. In vitro diaphragm physiological studies found that the twitch tension (p < 0.05) and maximal tension (p < 0.01) in the IRL group were significantly lower than control. The maximal tensions (p < 0.05) in the scavenger group were lower than control. After the fatigue protocol, diaphragmatic contractility in the scavenger group was similar to control and was better maintained compared with the IRL group. We conclude that free radical scavengers can prevent the development of diaphragm injury as evidenced by histology but the protection of diaphragm function is limited.

Diaphragm injury in individuals with airflow obstruction.

The purpose of this study was to describe the nature of diaphragm injury, to quantify the injury and number of macrophages at the light microscopic level, and to determine their association with airflow obstruction in humans. Partial-thickness diaphragm biopsies were obtained from 21 subjects going for thoracotomy surgery (FEV(1): 74 +/- 34% predicted; range: 16 to 122% predicted). Cross sections cut from frozen diaphragm were processed with H&E or processed for immunohistochemistry using the monoclonal antibody Ber-MAC3 (DAKO Corp., Carpinteria, CA) to label macrophages. Area fractions (A(A)) or the proportions of the cross- sectional area were determined by point counting all viable fields of H&E-stained diaphragm cross sections. A(A) were 66.2 +/- 9.0% for normal muscle, 17.6 +/- 7.2% for abnormal muscle, and 16.3 +/- 4.2% for connective tissue. Percent predicted FEV(1) was inversely related to the A(A) of abnormal muscle (r = -0.53, p < 0.01) and directly related to the A(A) of normal muscle (r = 0.37, p < 0.05). The number of macrophages was not related to % predicted FEV(1) (mean +/- SD: 0.41 +/- 0.18/fiber; 52 +/- 19/mm(2)). We conclude that increasing severity of airflow obstruction is associated with an increased A(A) of abnormal diaphragm and a decreased A(A) of normal diaphragm.

High-frequency oscillation and centroid frequency of diaphragm EMG during inspiratory loading.

Power spectra were derived from the diaphragm electromyogram (EMG) in anesthetized rabbits subjected to inspiratory resistive loading (IRL) with airway pressure swings of 40-60 cm H2O for 20 min to 2 h. Shifts in the centroid frequencies of the power spectra were found to be associated with the appearance of power spectral peaks in the range of 105-140 Hz, termed high-frequency oscillation, or HFO. Such peaks have been described before in phrenic nerve activity and in the diaphragm EMG. However, these peaks have not previously been connected with the shifts in centroid frequency seen during loaded breathing. Although such changes in frequency content have been taken to indicate fatigue in the diaphragm, we find that HFO can also cause a shift in centroid frequency during loaded breathing, an effect whose relation to fatigue has yet to be established.

Determinants of diaphragmatic injury.

Limb muscles can be injured during and after vigourous contractions. However, this injury is most evident under specific conditions. The strength and type of muscle contraction as well as the contractile status of the muscle are important determinants of injury. The initiating event leading to muscle injury is not clearly understood but there are several leading theories. The respiratory muscles are of obvious importance to survival, and fatigue or injury to them has been hypothesized to be prevented by various mechanisms. One such mechanism is reduced activation by the central nervous system. In this review information on the neural activation of the breathing muscles during inspiratory loading is discussed and reveals that neural activation to the diaphragm, the main inspiratory muscle, is high. Previous studies investigating the presence of muscle fatigue immediately after such inspiratory loading have shown little evidence of it. However, based on information from limb muscles, delayed or secondary muscle injury might occur and could produce deleterious effects on respiratory muscle function. Recent evidence shows that chronic low intensity inspiratory loading can produce diaphragmatic injury (Reid et al.) and secondary or delayed muscle injury can occur three days after an acute period of high intensity inspiratory loading. The results reviewed in this article suggest that the respiratory muscles, specifically the diaphragm, are not spared from injury or the results of muscle injury. Diaphragmatic function during the period of secondary muscle injury is markedly impaired and thus respiratory muscle injury is a phenomenon that warrants further investigation.

Delayed diaphragm injury and diaphragm force production.

The present study was designed to examine the effect of delayed diaphragm injury produced by inspiratory resistive loading (IRL) on diaphragm force production. On Day 1, three groups of anesthetized and intubated NZW rabbits (n = 7 in each group) were subjected to moderate IRL (Pao approximately 30 cm H2O), high IRL (Pao approximately 45 cm H2O), or no load for 1.5 h. On Day 3, the baseline twitch transdiaphragmatic pressure (Pdi) and Pdi at 10 to 80 Hz were measured during bilateral phrenic stimulation and these measurements were repeated after another IRL (high level) in all three groups. Diaphragm injury was assessed by the point-counting technique. Marked diaphragm injury was observed in the high-IRL group (p < 0.01), but no significant diaphragm injury was observed in the moderate-IRL or control groups. The baseline twitch Pdi was maintained in both IRL groups, whereas the baseline Pdi-frequency values in the high-IRL group were significantly reduced at most frequencies (p < 0.05). The decreases in twitch and Pdi at different frequencies were more pronounced after the IRL on Day 3 in the high-IRL group compared with controls. Moderate IRL did not decrease diaphragm force either before or after the high IRL on Day 3. We conclude that the diaphragm injury induced by high IRL has a significant impact on diaphragm force production and the attendant force loss produced by IRL is dependent on the intensity of inspiratory loading.

Load dependence of secondary diaphragm inflammation and injury after acute inspiratory loading.

Chronic or prolonged low-intensity loading of the inspiratory muscles has recently been shown to produce diaphragm injury. The present study was designed to examine whether an acute episode of inspiratory resistive loading (IRL) could produce secondary diaphragm inflammation and injury. On Day 1, three groups of anesthetized and intubated New Zealand White rabbits were subjected to moderate IRL (Pao of approximately 30 cm H2O), high IRL (Pao of approximately 45 cm H2O), or no load for 1.5 h. On Day 3, costal and crural diaphragms, parasternals, and gastrocnemius muscles were taken to assess injury by point counting. Normal muscle, abnormal and inflamed muscle, and connective tissue on hematoxylin and eosin-stained cross-sections were expressed as percentage of the total points for that cross-section. For the costal diaphragm, both the abnormal muscle (7.3 +/- 0.6% versus 1.1 +/- 0.2%; p < 0.001) and connective tissue (8.0 +/- 0.6% versus 5.7 +/- 0.2%; p < 0.01) in the high IRL group were higher than control, whereas in the moderate IRL group they were not significantly different from control. Total calpain-like activity was increased in the moderate IRL group but not in the high IRL group. Injury was observed in the parasternal muscles but to a lesser extent. No injury was observed in the gastrocnemius muscle. We conclude that secondary diaphragm injury occurs after acute IRL but only when the IRL exceeds the fatigue threshold.

Phrenic motoneuron firing rates before, during, and after prolonged inspiratory resistive loading.

Phrenic motoneuron firing rates during brief inspiratory resistive loading (IRL) are high, and nearly all the motoneurons are recruited. Diaphragmatic fatigue has been difficult to demonstrate during IRL. Furthermore, evidence from studies in limb muscles has shown variable motoneuron responses to prolonged high-intensity loads. We studied phrenic motoneuron firing rates before, during, and after prolonged IRL in anesthetized rabbits. Of 117 phrenic axons, only 2 axons were not recruited; 41 axons were silent during unloaded breathing but were recruited at higher loads. Silent axons showed a more rapid increase in firing rate as the load increased. Phrenic motoneuron firing rates increased throughout the period of loading, whereas airway pressure swings did not. After prolonged IRL, higher motoneuron firing rates were needed during brief loads to produce the same airway pressure. No evidence of a decline in motoneuron firing rates was seen at any point. We conclude that the respiratory muscles can be shown to demonstrate physiological responses consistent with fatigue during prolonged IRL, and activation rates are high and remain so throughout this prolonged loading.

Effect of the beta-agonist clenbuterol on dexamethasone-induced diaphragm dysfunction in rabbits.

The present study was designed to examine whether clenbuterol (CLEN) could reduce dexamethasone (DEX)-induced diaphragm dysfunction. We studied four groups of New Zealand white (NZW) rabbits, each receiving one of the following daily injections subcutaneously for 2 wk: saline (control), DEX 3 mg/kg, DEX 3 mg/kg + CLEN 2 mg/kg, and CLEN 2 mg/kg. Diaphragm fiber cross-sectional areas (CSA) were measured. Twitch transdiaphragmatic pressure (Pdi) and tetanic Pdi were measured during bilateral phrenic stimulation both before and after 60 min of inspiratory resistive loading (IRL). DEX produced a marked atrophy of type IIa and type IIb diaphragm fibers. This diaphragm atrophy was prevented by CLEN in the DEX plus CLEN group. CLEN alone increased CSAs of all three types of diaphragm fibers. Significant reductions in twitch Pdi and tetanic Pdi at all stimulation frequencies both before and after IRL were observed similarly in the DEX group as well as in the DEX plus CLEN group compared with the control animals. We conclude that DEX produces significant diaphragm atrophy and decreases diaphragmatic contractility. CLEN produces hypertrophy of the diaphragm and minimizes diaphragm atrophy induced by DEX, but it has no demonstrable protective effect on DEX-induced diaphragm dysfunction.

Some effects of vagal blockade on abdominal muscle activation and shortening in awake dogs.

1. The mechanisms of abdominal muscle activation are thought to be different during expiratory threshold loading (ETL) compared with hypercapnia. Our objectives in the present study were to determine the effects of removing excitatory vagal feedback on abdominal muscle activation, shortening and pattern of recruitment during ETL and hypercapnia. Six tracheotomized dogs were chronically implanted with sonomicrometer transducers and fine wire EMG electrodes in each of the four abdominal muscles. Muscle length changes and EMG activity were studied in the awake dog during ETL (6 dogs) and CO2 rebreathing (3 dogs), before and after vagal blockade. 2. Following vagal blockade, the change in volume (increase in functional residual capacity, FRC) during ETL was greater and active phasic shortening of all the abdominal muscles was reduced, when shortening was compared with a similar change in lung volume. Similarly, at comparable minute ventilation, abdominal muscle active shortening was also reduced during hypercapnia. The internal muscle layer was recruited preferentially in both control and vagally blocked dogs during both ETL and hypercapnia. 3. The degree of recruitment of the abdominal muscles during ETL and hypercapnia in awake dogs is influenced by vagal feedback, but less so than in anaesthetized dogs. These results illustrate the importance of the vagi and abdominal muscle activation in load compensation. However, vagal reflexes are apparently not contributing to the preferential recruitment of the internal muscle layer. In awake dogs during vagal blockade abdominal muscle recruitment still occurs by extravagal mechanisms.

Ventilatory muscle strength and endurance in myasthenia gravis.

Patients with generalized myasthenia gravis (MG) often have associated ventilatory muscle involvement. It is not known whether patients with isolated ocular muscle involvement have identifiable involvement of their ventilatory muscles. Most studies have assessed muscle involvement by measuring muscle strength; however, we hypothesized that measures of ventilatory muscle endurance may be more sensitive tests of ventilatory muscle involvement in myasthenia gravis. We studied 17 patients with myasthenia gravis (four with ocular involvement alone and 13 with varying degrees of generalized myasthenia gravis). Spirometry, ventilatory muscle strength (maximum inspiratory and expiratory pressures (MIP and MEP)) and endurance (2 min incremental threshold loading test) were measured before and 20 min after i.m. neostigmine. We compared the results with those of 10 normal controls. We found no difference between patients with isolated ocular involvement and controls. Ocular myasthenia gravis patients did not improve after neostigmine. The patients with generalized myasthenia gravis had reduced baseline ventilatory muscle strength (MIP 67 cmH2O (70% of predicted), MEP 86 cmH2O (50% of pred) and endurance (mean maximal load achieved = 246 g, mean pressure at highest load (P) = 19.4 cmH2O) compared with controls. After neostigmine, there was a significant increase in MIP in patients with generalized myasthenia gravis and a trend towards an increased MEP. As a group, the patients with generalized myasthenia gravis did not demonstrate a change in their ventilatory muscle endurance after neostigmine; however, there was considerable interpatient variability in response. We conclude that patients with isolated ocular MG have normal ventilatory muscle strength when tested conventionally.(ABSTRACT TRUNCATED AT 250 WORDS)