Pulmonary rehabilitation in Iranian outpatients with mustard gas lung disease: a randomised controlled trial


The use of chemical weapons against Iranian combatants and civilians during the Iran-Iraq war in the 1980s resulted in long-term health problems for many tens of thousands of people.1 Respiratory complications of mustard gas exposure include tracheal stenosis, tracheomalacia, bronchiectasis and constrictive bronchiolitis2–4 leading to cough, sputum and breathlessness, associated with limitation of exercise capacity and difficulties with day to day activities. As is the case with other long-term respiratory conditions,5 fatigue anxiety and depression are also common6 7 while clinical symptoms and quality of life are only poorly correlated with measures of lung function.8 The clinical picture is thus similar to that observed in chronic obstructive pulmonary disease (COPD).9 10

Exercise capacity and breathlessness in lung disease are determined by a combination of cardiorespiratory impairment, skeletal muscle dysfunction and changes to the threshold at which symptoms are tolerated.11 12 Breathlessness, such as pain, can be modified by past experience and expectations and there is evidence of central adaptations in people with lung disease.13 A vicious cycle of breathlessness, immobility and deconditioning can worsen patients’ condition and make them more susceptible to the development of physical inactivity-related multimorbidity including hypertension, diabetes and osteoporosis. Most of the people exposed to mustard gas in the 1980s are now in their 50s or older.1 A sustainable health system needs to prevent the development of future health problems—physical activity interventions have a key role in this.14

Skeletal muscle impairment is a common feature of lung diseases including COPD and pulmonary fibrosis.15–18 It can occur early in the course of lung disease; quadriceps weakness is present in about 30% of patients with COPD and is associated with reduced levels of physical activity.17 It directly influences exercise performance,19 is associated with poor health status20 and is an independent predictor of healthcare utilisation21 and mortality.22 Musculoskeletal disorders are also highly prevalent in COPD.23

Pulmonary rehabilitation (PR) is a holistic approach for people limited by lung disease, defined as ‘a comprehensive intervention based on a thorough patient assessment followed by patient-tailored therapies that include, but are not limited to, exercise training, education on your lung disease or condition and how to manage it, and behaviour change, designed to improve the physical and psychological condition of people with chronic respiratory disease and to promote the long-term adherence to health-enhancing behaviours’.24 There is an extensive evidence base for PR in patients with COPD25 and a growing body of research in other lung diseases.24 PR can improve muscle strength, exercise capacity, health status, symptoms of anxiety and depression as well as reducing hospital readmission rates. It has been shown to alter breathlessness perception13 and also boosts COPD patients’ understanding of their condition and ability to self-manage.

A marked increase in susceptibility to fatigue is also observed in COPD, with a more rapid decline in performance during continuous or repeated bouts of exercise26 or with non-volitional testing.27 This is associated with muscle fibre-type shift28 reduced oxidative capacity27 29 and increased skeletal muscle adiposity.30 This fibre-type shift is itself independently associated with increased mortality risk.31

Although mustard gas lung disease shares many clinical features with COPD, it is not known whether PR is effective at improving quality of life and exercise capacity in people with the condition.


We conducted an assessor-blind, parallel-group randomised controlled trial comparing PR to usual care to assess the effect of this intervention in mustard lung patients with limiting breathlessness, as outlined in the previously published protocol.32 Participants were people with exposure to mustard gas contemporaneously documented in the 1980s, with symptomatic physician-diagnosed mustard gas lung disease following that, residing in Tehran and its suburbs.

Inclusion criteria included forced expiratory volume in 1 s <80%, MRC dyspnoea score ≥3.33 Exclusion criteria were (1) any type of debilitating clinical condition preventing the patient from participating in the rehabilitation programme, such as arthritis, (2) any type of clinical condition that would endanger the patient during the physical exercises, such as uncontrolled cardiac diseases, (3) the presence of other non-chemical warfare-related pulmonary diseases such as asthma, (4) active malignancy and (5) severe cognitive disorder and psychiatric disease that is associated with a memory disorder. The PR, trial activities and other clinical services were delivered in Tehran’s Khatam-ol-Anbia Hospital.


In addition to their usual care, the intervention group took part in a 6-week PR programme, which included a combination of physical exercises with endurance and strength components, tailored to each individual’s condition and characteristics at baseline at levels which were increased as individuals progressed through the programme.32 The PR programme also included education and psychosocial support. Participants underwent eighteen sessions (6 weeks, three sessions a week). Each session included 60 min of exercise, of which 15 min was spent on warming up. Exercises were taught and subsequently supervised by a physiotherapist. A general physician was also present to address any clinical issues arising.

The control group continued their usual medication, with management determined by their clinical team. All participants received standard advice about the importance of physical activity and were encouraged to exercise regularly.

Study outcomes

The primary outcome of the study was cycle endurance time at 70% of baseline workload (Tlim70), measured at the end of 6 weeks. An increase of >105 s in the endurance time has been defined as the minimum clinically important difference in pharmacological studies and was used as a threshold for responder analysis.34

An initial symptom-limited incremental exercise test (IET) was used to determine the Wmax of each participant at baseline.35 After 3 min of acclimatisation and 1 min of unloaded cycling, the workload was increased by 10W per minute until the patient could not continue, the peak workload (Wmax) being defined as the highest which could be sustained for 60 s. A practice constant work rate test was performed approximately 1 hour after the IET at 70% maximum workload. Three min unloaded cycling (0 Watt) were allowed for respiratory adaptation and warm up. At a subsequent session, 2–4 days later, the constant workload test was repeated and the outcome of this was taken as the baseline measure of Tlim70.

Six min walk distance (6MWD) was also assessed in accordance with American Thoracic Society guidelines, including a practice walk36 and additional secondary outcomes were Medical Research Council (MRC) dyspnoea score, step count, quadriceps bulk and strength, anthropometrics, anxiety and depression, as well as health-related quality of life.

Baseline measures included anthropometrics, spirometry, body plethysmography, and gas transfer (Dlco) measured in accordance with international guidelines.35 37 38 To assess quadriceps function, participants performed 30 repetitions of a maximum voluntary contraction with an angular velocity of 90/s using Isokinetic Test Rehabilitation System (MPS21) in an upright sitting position, with an angle of 90° at the thigh joint.39 Strength was assessed as the quadriceps muscle peak torque (Newton-metre (Nm)). The endurance of the quadriceps muscle is defined as the total work done (joules) during 30 consecutive muscular contraction repeats.40 To reduce the effects of learning at the baseline assessment, the test was performed twice and the highest value was taken.

Rectus femoris cross-sectional area (RFCSA) was assessed by B-mode ultrasonography using an 8 MHz 5.6 cm linear probe placed perpendicular to the longitudinal axis of the thigh in the superior segment, three-fifths of the distance from the anterior superior iliac spine to the superior patellar border, similar to de Bruin’s method.41

Daily physical activity was measured using the PD724 triaxial pedometer (Tanita, Tokyo, Japan)) worn for a week before baseline and outcome visits. The pedometer displays a cumulative step count for each day and also retains the step counts from the preceding 7 days in its memory allowing the average step count for the preceding week to be evaluated.42 The first and last days of recording were discarded to remove bias due to travel to the study centre.

Health status was assessed using a validated Farsi version of the St. George’s Respiratory Questionnaire (SGRQ).43 The MCID for the SGRQ is estimated as four points.44 Participants also completed the Farsi translation of the Hospital Anxiety and Depression Scale (HADS).45

Because of logistical issues and resource limitations, the analyses of blood tests that had been planned in the original protocol were not performed and lung function tests were not repeated at the end of the PR programme. Borg scores for breathlessness and fatigue during the walk tests are not reported and finally a number of other data items labelled as ‘secondary outcomes’ in the original trial register were in fact baseline measures used to characterise the population only.

Study allocation

After the completion of baseline measures, participants were randomly allocated to PR or usual care arms using a computer-generated random sequence. Outcome measures were assessed by staff who were blind to treatment allocation.

Sample size and statistical analysis

Assuming a 105 s improvement in the Tlim in the treatment arm and a 20 s improvement in controls with an SD of 100, 90% power and a significance level of 5% required 60 patients to complete the study. The data were cleaned and assessed for outliers and missing variables. Outcomes were compared using ANCOVA adjusted for baseline parameters, unless otherwise stated and on an intention to treat basis. Effect sizes are based on estimated marginal means using ANCOVA controlling for baseline level of the variable of interest. An alpha of <0.05 was taken to indicate statistical significance. No multiple testing adjustments were made. Data were analysed using SPSS V.26 Statistics and STATA (V.15).

Patient and public involvement

Patients were not involved in the design, or conduct, reporting, or dissemination plans of the research.


Study participants

60 Iranian men with a diagnosis of mustard gas lung disease and an MRC dyspnoea score of 3 or more, mean(SD) age 52.7 (4.4) years took part in the study (figure 1), with 31 individuals allocated to the PR arm and 29 to usual care (table 1).

Figure 1
Figure 1

CONSORT diagram describing patient flow through the study. CONSORT, Consolidated Standards of Reporting Trials.

Table 1

Baseline participant characteristics

Impact of PR

Follow-up data at 6 weeks were available for 28 in the PR group and 26 usual care. PR was associated with an improvement in exercise capacity—the primary endpoint, cycle endurance time at 6 weeks, increased from 377 (140) s to 787 (343) s with PR, whereas in the usual care arm the values were 495 (171) s to 479 (159 s), respectively, giving an effect size of +383 (231) s (p<0.001) (figure 2). Using an MCID of >105 s gave responder rates of 87% vs 41%, respectively (p<0.001).

Figure 2
Figure 2

Impact of pulmonary rehabilitation on study outcomes. (A) Cycle endurance time (p<0.001), (B) 6 min walk distance (6MWD) (p<0.001), (C) St George’s Respiratory Questionnaire (SGRQ) (p=0.001), (D) Hospital Anxiety and Depression Scales (HADS)-Depression (p=0.008). Boxes indicate 25th–75th percentile, central line is the median, whiskers are upper and lower adjacent values, outliers are values beyond the upper and lower adjacent values.

There were also improvements in 6MWD+103.2 (63.6–142.9) (p<0.001), MRC dyspnoea score −0.36 (−0.65 to −0.07) (p=0.016), quadriceps function and depression (HADS depression) −2.00 (−3.45 to −0.54) (p=0.008) together with improvements in measures of health status (SGRQ total) −6.23 (−11.70 to −0.758) (p=0.026) (table 2). 6MWT responder analysis using MCID of 30 m demonstrated response rates of 83.9% (group A) vs 24.1% (group B) (p<0.001). PR was associated with a numerical but not statistically significant improvement in daily step count +1389.7 (−3254.1 to 6033.5)

Table 2

Effect of pulmonary rehabilitation in people with mustard gas lung disease

Longer-term effect of PR

Data at 1 year were also available for 24 (77%) of participants in the intervention arm and 14 (48%) of those receiving usual care. Given this marked disparity in follow-up, caution is needed for any comparison and data on health resource utilisation and exacerbation frequency were incomplete and therefore not reliable, preventing health economic analysis. Although a numerical improvement from baseline was apparent in cycle endurance time; PR 388.3 (142.4) s to 505.8 (282.6) s vs UC 494.3 (151.4) s to 505.0 (258.7) s the difference was not statistically significant, nor were there statistically significant differences in exercise or muscle parameters (table 3).

Table 3

Comparison of outcomes at 1 year between study arms


The main outcome of this study was that, in people who are limited by breathlessness due to lung disease caused by mustard gas exposure, PR produces clinically and statistically significant improvements in exercise capacity assessed both by cycle endurance and 6MWD, as well as improving health status, depression, breathlessness and quadriceps strength and endurance.

Significance of findings

There is already compelling evidence to support the use of PR as an approach to improve outcomes in a range of respiratory conditions so the findings in this population are not unexpected. However, the new availability of objective data in this specific patient group, showing that improvements following PR substantially exceed accepted minimum clinically important differences, should encourage commissioners of healthcare to ensure that this form of intervention is made available to these patients.

By 12 months, there were no statistically significant differences between groups compared with baseline, suggesting that the effects of the intervention might have worn off. Caution is needed as there were relatively low numbers available for follow-up. However, the question of how to sustain the benefits of PR in people with lung disease is a well-established problem. Several possibilities follow which will need to be subjects for future work in this patient population. Will a longer initial programme be more likely to produce sustained behaviour change? What is the best strategy to sustain improvement through maintenance programmes or reviews? Might the use of technology to prompt or monitor physical activity and an ongoing exercise programme be effective?

Study limitations

The study was a randomised controlled trial with assessors blinded to study allocation so the findings should be robust. There was a good completion rate for the primary endpoint assessing impact immediately following completion of the programme, but the substantial drop-out by 1 year makes the question of longer-term impact less clear. It would have been desirable to include a more sophisticated measure of physical activity to understand the impact that PR had on participants’ experience of physical activity in daily life.46 47 We did not capture measures of social isolation and other social and cultural impacts of PR, which may allow people to undertake activities which they were prevented from due to breathlessness or anxiety.5 48 49

In terms of generalisability, caution is needed when extrapolating to people with less severe disease, though as is the case in COPD, it is likely that PR will be of benefit to any individual with functional limitation. There is no reason to think that women with lung disease caused by mustard gas would not also benefit from PR, but our study population was all male. Although the majority of people exposed to mustard gas during the 1980s were male combatants,2–4 civilian populations were also affected (eg, in the bombing of Sardasht),50 and the use of chemical weapons against Kurdish civilians in Iraq by the Iraqi regime is also well documented.

We powered the study using an MCID of 105 s for the cycle endurance time, however, a recent (2023) publication suggests that for bronchodilator interventions the MCID is 1.13 min (68 s), whereas for exercise interventions (as in the present paper) it is 0.93 min (56 s). The authors of this consensus paper suggest adopting 60 s.51


These results support the provision of PR for people with lung disease caused by exposure to mustard gas, though questions remain as to how to best sustain the initial benefit associated with completing a course.

Supplemental material

This post was originally published on https://bmjopen.bmj.com