Received: 16 September 2002Revised: 22 November 2002Accepted: 1 April 2003Published online: 29 May 2003© Springer-Verlag 2003

Abstract Biomass fuels are fre-quently used in rural areas of theworld for cooking and heating fre-quently. It has been reported that theuse of these fuels causes hazardouseffects on the lungs. In this study, weevaluated the pulmonary changesdue to the use of biomass fuels in afemale population that lives in ourterritory by high-resolution comput-ed tomography (HRCT). The studyanalyzed three groups of women.The first group comprised those sub-jects who were exposed to biomasswithout respiratory symptoms (group 1; n=32). The second groupcomprised those individuals thatwere exposed to biomass and showedrespiratory symptoms, such as cough,sputum production, and dyspnea(group 2; n=30). The third group wascomposed of women who were notexposed to biomass and also had norespiratory symptoms (group 3;n=30). Women with a history of con-comitant pulmonary diseases were

excluded from the study. All groupswere examined with HRCT. Groups1 and 2 (individuals exposed to bio-mass fuels) had more pathologicfindings than group 3 (not exposedto biomass fuels). Ground-glass ap-pearance was seen in 71.9% in group1, 23.3% in group 2, and 3.3% ingroup 3. The difference between thegroups was statistically significant(p<0.05). Fibrotic bands were seen50% in group 1, 63.3% in group 2,and only 6.7% in group 3 (p<0.001).Exposure to biomass fuels was thecause or predisposing factor formany pulmonary diseases, rangingfrom chronic bronchitis to diffuselung diseases. We believe that thesepathological changes due to biomassfuels can be detected earlier byHRCT and the diseases might beprevented or treated earlier.

Keywords Biomass fuels · High-resolution computed tomography · Pulmonary disease

Eur Radiol (2003) 13:2372–2377DOI 10.1007/s00330-003-1925-5 C H E S T

Mustafa KaraSema BulutFikret TasIbrahim AkkurtZehra Seyfikli

Evaluation of pulmonary changes due to biomass fuels using high-resolution computed tomography

Introduction

Biomass fuels, such as dried cow feces with or withouthay (wood, environmental remnants, manure, etc.), arecommonly used in rural areas of the world for cookingand heating. These fuels include organic compounds,plant proteins, carbohydrates, and some other compo-nents. In houses where biomass fuels are used, carbonmonoxide, nitrogen, and sulfur dioxide levels are abovethe maximum levels of the internationally accepted val-ues. In addition, carcinogenic and toxic benzopyrene hy-

drocarbons are formed by combustion of wood, driedcow manure, and/or hay together [1].

In rural regions of Turkey, dried cow manure with orwithout hay is the main fuel used for heating. Althoughthe female population from these regions frequentlydoes not smoke, it often suffers from interstitial lung diseases, chronic obstructive pulmonary disease(COPD), lung cancer, bronchiectasis, and other lungdiseases. During childhood, breathing the exhaust ofburning biomass fuels and insufficient fresh air or oxy-gen in the houses causes detrimental effects on the res-

M. Kara · F. Tas (✉)Department of Radiology,Faculty of Medicine,Cumhuriyet University, 58140 Sivas, Turkeye-mail: ftas@cumhuriyet.edu.trTel.: +90-34-62191280Fax: +90-34-62191284

S. BulutDepartment of Radiology,Faculty of Medicine,Cumhuriyet University, 58140 Sivas, Turkey

I. AkkurtDepartment of Respiratory Disease,Faculty of Medicine,Cumhuriyet University, 58140 Sivas, Turkey

Z. SeyfikliDepartment of Respiratory Disease,Faculty of Medicine,Cumhuriyet University, 58140 Sivas, Turkey

2373

piratory functions, repeated lung infections, and COPD[2, 3, 4].

Most of the studies related to the effects of biomassfuels, are epidemiological. No study has dealt withHRCT findings of the pulmonary changes due to bio-mass fuels. We think that early or late pulmonary chang-es in subjects exposed to biomass fuels can be detectedby HRCT, even though the respiratory function tests andplain radiograms are normal.

In this study, we evaluated pulmonary changes withHRCT in a symptomatic and in a asymptomatic femalepopulation exposed to biomass fuels for a long time, andtheir findings were compared with an asymptomatic fe-male population who had never been exposed to biomassfuels.

Materials and methods

This study enrolled 92 non-smoking women. The participantswere divided into three groups. Group 1 consisted of 32 womenand their ages ranged from 24 to 65 years. They had also used bio-mass fuels at least for 10 years and had no respiratory symptoms.Group 2 consisted of 30 women and their ages ranged from 25 to70 years. They had also used biomass fuels at least for 10 yearsand had respiratory symptoms such as cough, sputum production,and dyspnea. The participants of groups 1 and 2 were selectedfrom the individuals who lived in a rural area and had regularlyused dried cow manure with hay for heating and cooking in theirhouses. Group 3 consisted of 30 women and their ages rangedfrom 27 to 69 years. They had never used biomass fuels and livedin flats in an urban area. The HRCT examinations were performedin groups 1 and 3 for the aim of the research and in group 2 forclinical indications.

All participants were informed about the study and ethical ap-proval and their permission for the participation were obtained.The participants were asked for smoking habit, medications, aller-gy, kind of biomass fuel, frequency, and style of the usage of the

biomass fuels. All of the women were asked for respiratory symp-toms and physical examinations were made especially for tubercu-losis and allergy symptoms.

High-resolution computed tomography examinations were per-formed with a spiral CT machine (Picker, PQS model, Cleveland,Ohio). The HRCT images were obtained with 2-mm slice thick-ness and 10-mm intervals. The other scan parameters were120 kV, 150–200 mA, and 2-s scan time. Images were reconstruct-ed in bone algorithm and examined in the lung window. The win-dow levels ranged between (−300 HU) and (−700 HU), windowwidths ranged between 1600–2000 HU. Contrast medium was notused. Five observers evaluated the HRCT examinations. Presenceor absence of radiological abnormalities such as the interface sign,ground-glass appearance, fibrotic bands, peribronchovascularthickening, air cysts, bullae, pleural thickening, nodular radio-opacities, mosaic pattern, bronchiectatic changes, and curvilineardensities were recorded.

Statistical analysis

Statistical analysis of the data was done by chi-square test andanalysis of variance (ANOVA) using SPSS software program. Sta-tistically, results were expressed as mean±standard deviation (SD),and a p value of less than 0.05 was accepted as significant.

Results

The mean age was 49.3±12.4 years in all participants,and it was 47.2±12.6, 53.4±12.4, and 47.3±11.6 years ingroups 1, 2, and 3, respectively. There was no significantdifference between the groups when the mean age wastaken into account (k>0.05). The results in terms of dif-ferences in the presence of the various HRCT findingsare presented in Table 1.

Table 1 Comparisons of the high-resolution computed tomography(HRCT) findings of the groups. CWG comparison within the groups, IS interface sign, GGA ground-glass appearance, FB fibrotic

band, PBVT peri-bronco-vascular thickening, AC-B air-cyst bullae,PT pleural thickening, NR nodular radiopacity, MP mosaic pattern,BC bronchiectatic change, CD curvilinear density, + present, −− absent

IS GGA FB PBVT AC–B PT NR MP BC CD

− + − + − + − + − + − + − + − + − + − +

Group 1N 19 13 9 23 16 16 23 9 32 0 23 9 25 7 30 2 30 2 25 7% 59.4 40.6 28.1 71.9 50 50 71.9 28.1 100 0 71.9 28.1 78.1 21.9 93.7 6.3 93.7 6.3 78.1 21.9

Group 2N 25 5 23 7 11 19 23 7 26 4 14 16 23 7 22 8 22 8 29 1% 83.4 16.6 76.7 23.3 36.7 63.3 76.7 23.3 86.7 13.3 46.7 53.3 76.7 23.3 73.3 26.7 73.3 26.7 96.7 3.3

Group 3N 29 1 29 1 28 2 29 1 29 1 27 3 29 1 30 0 29 1 29 1% 96.7 3.3 96.7 3.3 93.3 6.7 96.7 3.3 96.7 3.3 90 10 96.7 3.3 100 0 96.7 3.3 96.7 3.3

CWG p<0.001 p<0.01 p<0.001 p<0.05 p>0.05 p<0.001 p>0.05 p<0.05 p<0.01 p<0.05Groups 1, 2 p<0.05 p<0.001 p>0.05 p>0.05 p<0.05 p<0.05 p>0.05 p<0.05 p<0.05 p<0.05Groups 1–3 p<0.05 p<0.001 p<0.001 p<0.01 p>0.05 p>0.05 p<0.05 p>0.05 p>0.05 p<0.05Groups 2, 3 p>0.05 p<0.05 p<0.001 p<0.05 p>0.05 p<0.001 p<0.05 p<0.01 p<0.05 p>0.05

2374

Interface sign

Each group was evaluated with regard to interface sign,which is the earliest and most frequent finding of inter-stitial lung disease (Fig. 1a). Interface sign incidencewas 40.6% in group 1, 16.6% in group 2, and 3.3% ingroup 3. The differences between groups 1 and 2 and 1and 3 were significant (p<0.05) and between groups 2and 3 were insignificant (p>0.05).

Fibrotic bands

Thick interlobular septum (linear opacity) incidence was50% in group 1, 63.3% in group 2, and 6.7% in group 3(Fig. 1b). The differences between groups 1 and 3 and 2and 3 were statistically significant (p<0.001), but no dif-ference was found between groups 1 and 2 (p>0.05).

Ground-glass appearance

The incidence of ground-glass appearance was 71.9% ingroup 1, 23.3% in group 2, and 3.3% in group 3 (p<0.01;Fig. 2).

Peribronchovascular thickening

Incidence of peribronchovascular thickening was 28.1%in group 1, 23.3% in group 2, and 3.3% in group 3. Thedifferences between groups 1 and 3 and 2 and 3 werestatistically significant (p<0.05), but they were not sig-nificant between groups 1 and 2 (p>0.05).

Cystic lesions

We did not see any cystic lesion in group 1. The occur-rence of cystic lesions was 13.3% in group 2 and 3.3% ingroup 3. The difference between groups 1 and 2 was sta-tistically significant (p<0.05), but they were not signifi-cant between groups 2 and 3 and 1 and 3 (p>0.05).

Pleural thickening

Incidence of pleural thickening was 28.1% in group 1,53.3% in group 2, and 10% in group 3. The differencesbetween groups 1 and 2 and groups 2 and 3 were statisti-cally significant (p<0.05), but they were not significantbetween groups 1 and 3 (p>0.05).

Fig. 1 a Supine position high-resolution computed tomography(HRCT) slice of a 66-year-old symptomatic woman with interfacesign. b Prone-position HRCT slice of the same patient with sub-pleural lines and fibrotic bands

Fig. 2 An HRCT slice of a 65-year-old woman from the exposedgroup, without symptoms. Bilateral ground-glass appearance isshown

Bronchiectasis

The incidence of bronchiectasis was 6.3% in group 1,26.7% in group 2, and 3.3% in group 3 (Fig. 4). The dif-ferences between group 1 and 2 and 2 and 3 were signifi-cant (p<0.05), but these were not between groups 1 and 3(p>0.05). Interestingly, we found that most of these bron-chiectatic changes in the first two groups were formed bytraction bronchiectasis. This type of bronchiectasis devel-ops secondary to pulmonary fibrosis. Pulmonary fibrosiswas 50% in group 1 and 63.3% in group 2.

Subpleural lines

Subpleural lines (curvilinear increased densities) wereobserved in 21.9% of group 1 and 3. 3% of groups 2 and3 (Fig. 1b). The differences between groups 1 and 2 and1 and 3 were significant (p<0.05), but no significant dif-ference between groups 2 and 3 (p>0.05) was noted.

Discussion

In developing and underdeveloped countries, most of thepopulation lives in the rural region and uses biomass fu-els such as wood, dried cow manure, hay, and plant rem-nants. Use of biomass fuels for heating and cookingcauses harmful gases and increases suspended air parti-cles in the houses. Many reports indicate that air pollu-tion level in the houses, in which biomass fuels are used,is extremely high [2, 3, 4].

Most of the pollutants in biomass smoke have harm-ful effects on the health [4, 5]. Location, poor aerationsystem of the kitchens and the kind of the fuel is impor-tant for the health of the housewives. Generally, in ruralareas, women cook in closed and insufficiently ventilat-ed spaces. All of the fuels used for cooking contain mainrespiratory irritant smoke particles, including nitro-ox-ides, sulfur dioxide, and non-inflammable hydrocarbons.These smoke particles cause chronic bronchitis and airobstructions [6].

Many authors from different regions of the world re-ported the relationship between the exposure to biomassfuels with the occurrence of the chronic bronchitis andCOPD [2, 3, 4,7,8]. In our study, HRCT examinationsshowed that the incidence of bronchiectatic changes was26% in symptomatic patients (group 2), 6.3% in asymp-tomatic patients (group 1), and 3.3% in the controls(group 3). Bronchiectatic changes, observed in the bio-mass-exposed groups, were traction bronchiectasis. Thistype of bronchiectasis develops secondary to the pulmo-nary fibrosis.

Radiographic findings of women living in the ruralareas of the Africa were observed in one study. In thatstudy, miliary appearances and massive fibrosis were ob-

2375

Nodular opacities

Occurrence of nodular radioopacity was 21.9% in group1, 23.3% in group 2, and 3.3% in group 3 (Fig. 3). Thedifferences between groups 1 and 3 and between 2 and 3were significant (p<0.05). The difference betweengroups 1 and 2 was not significant (p>0.05).

Mosaic pattern

Incidence of mosaic pattern was 6.3% in group 1, 27.6%in group 2, but it was not seen in group 3. There weresignificant differences between groups 1 and 2 and 2 and3 (p<0.05), but the difference between groups 1 and 3was not significant (p>0.05).

Fig. 3 An HRCT slice of a 34-year-old woman from the exposedgroup without symptoms. Nodular pleural thickening is shown

Fig. 4 An HRCT slice of a 60-year-old symptomatic woman, withbilateral cystic bronchiectasis and peribronchial thickening

served. These patients had no active tuberculosis, but thepneumoconiosis was present in lung biopsies. It hadbeen suspected that these findings were caused by smokeexposure from biomass fuels and silica particles inhaledduring manual corn grinding between stones. This lungis described as the “hut lung” by the authors [9]. In pro-gressive massive fibrosis, opacities greater than 1 cm,nodular appearances, cavitations, and bullous emphyse-ma may be seen [10].

To our knowledge, there are no previous reports onHRCT findings in subjects exposed to biomass fuelsmoke. In our study the incidence of the interstitial nod-ule in the first two groups (exposed groups) was barelyhigher than that of the control group. The incidence ofthe fibrotic bands was 50% in group 1, 63.3% in group 2,and 6.7% in group 3. The difference between the groupswas significant (p<0.001). The fibrotic bands, mostlyseen in the symptomatic group, may be the sign of therestrictive phase of the disease.

According to a study from Saudi Arabia, women ex-posed to biomass smoke might be at risk for COPD [11].In another study from India, Behera et al. [12] evaluated3318 non-smoking women exposed to biomass fuels,with pulmonary function tests. They found that biomassexposure might cause parenchymal disorders. Etiologyof the respiratory complaints in Mexican women, ex-posed to biomass smoke for at least 10 years, was inves-tigated, and most of the chest radiographs of the patientswere found to be normal, but sometimes increased local-ized reticular or reticulonodular infiltrations were seen.Fourteen of 22 patients who had anthracosis were inves-tigated with bronchoscopy. All of them had mucosal ery-thema, edema, dilated mucous glands, as well as acuteand chronic airway inflammations. Bronchus biopsysamples showed chronic bronchitis varying from medi-um to severe changes. Frequently, anthracosis was seenin the lung parenchyma around the small airways and ar-teries. Muscular hypertrophy was observed in the arteriesand arterioles. These findings, seen in open-lung biopsysamples, were similar to the interstitial lung diseases dueto exposure to other inorganic dusts. In some patients,cystic cavity formation was observed [13] which is thelate finding of interstitial lung disease.

In our study, occurrence of ground-glass appearance,which shows the potentially treatable limit of interstitiallung disease, was 71.9% in group 1, 23.3% in group 2,and 3.3% in the control group. The difference betweenthe groups was statistically significant (p<0.01). Ground-glass appearance was seen frequently in group 1 (71.9%)and represents a sign of alveolar wall (alveolar septum)thickening, and preservation of the secondary lobules.Ground-glass appearance may be due to the early phaseof interstitial lung disease and it was less frequently seenin the symptomatic group (23.3%).

Honeycomb appearance, which is seen in the latestperiod of the interstitial lung diseases, was found in 2 pa-

tients of the symptomatic group. Incidence of the mosaicpattern, which is also a finding of advanced interstitiallung diseases, was 6.3% in the asymptomatic group,26.7%, in the symptomatic group, and it was not encoun-tered in the control group. The difference was significant(p<0.05). The occurrence of vascular diameter alterationand bronchiectasis allow to differentiate mosaic patternfrom ground-glass appearance and these findings indi-cate the last phase of interstitial lung disease.

In the HRCT studies, subpleural lines (curvilinear in-creased densities) can be seen in interstitial lung disease,pulmonary congestion, and dependent parts of the nor-mal lungs. These densities were observed in 21.9% ofthe asymptomatic group, and in 3.3% in both the symp-tomatic and control groups. On the other hand, incidenceof peribronchovascular thickening, which occurs as theresult of the pathologic changes in the axial interstitium,was 28.1% in the asymptomatic group, 23.3% in thesymptomatic group, and 3.3% in the control group. Highincidence of the subpleural lines and the peribronchovas-cular thickening in asymptomatic group indicates thatthese are the early-phase findings of the interstitial lungdiseases.

Frequently, pleural thickening is seen secondary topleuritis [10]. In our study, occurrence rate of the pleuralthickening in HRCT scans was 28.1% in asymptomaticgroup, 53.3% in symptomatic group, and 10% in thecontrol group. Pleural thickening in the asymptomaticgroup was mostly focal and nodular, and the maximumdiameter at the thickest point was 1 cm. In the symptom-atic group, the pleural thickening was diffuse, and itsthickness reached to 2 or 3 cm in some areas. Withinpleural thickening, calcifications were frequently en-countered.

Air cysts larger than 1 cm are defined as “bullae,” andif they were in the subpleural locations, they were de-fined as “bleb” [10]. In the present study, incidence ofair cysts was 13.3% in the symptomatic group and 3.3%in the control group, but no cystic lesion was encoun-tered in the asymptomatic group. Presence of many cys-tic lesions in the symptomatic group may be due to ad-vanced stage of the interstitial lung disease.

Conclusion

In conclusion, it can be said that exposure to biomass fu-els is likely to cause many lung diseases from interstitiallung diseases to COPD. Our results suggest that, in allstages of the disease, pulmonary parenchymal changesdue to biomass exposure can be properly evaluated andfollowed by the HRCT examination.

2376

2377

References

1. Koning HW, Simth KR, Last JM(1985) Biomass fuel combustion andhealth. Bull World Health Organization63:11–26

2. Pandey MR (1984) Domestic smokepollution and chronic bronchitis in arural community of the hill region ofNepal. Thorax 39:337–339

3. Demirtas N, Seyfikli Z, Topcu S (1999)The relationships between traditionalbiomass combustion and developmentof COPD in women of Sivas area. Respir Dis 10:156–158 [in Turkish,with English abstract]

4. Ozbay B, Uzun K, Arslan H, Zehir I(2001) Functional and radiological im-pairment in women highly exposed toindoor biomass fuels. Respirology6:255–258

5. WHO/IDRC (1993) International studygroup on indoor air pollution andchildhood pneumonia: impact of inter-ventions to reduce indoor air pollutionon childhood pneumonia and otherhealth problems. Revised, pp 1–20

6. Behera D, Jindal SK (1991) Respirato-ry symptoms in Indian women usingdomestic cooking fuels. Chest100:344–346

7. Padmanati S, Arora S (1976) Sex dif-ference in chronic cor pulmonale inDelhi. Br J Dis Chest 70:251–259

8. Malik SK (1985) Exposure to domesticcooking fuels and chronic bronchitis.Indian J Chest Dis Allied Sci27:171–174

9. Grobbelaar JP, Bateman ED (1991) Hutlung: a domestically acquired pneumo-coniosis of mixed etiology in ruralwomen. Thorax 46:334–340

10. Daehnert W (1996) Differential diag-nosis of chest disorders. Radiologic re-view manuel. Williams and Wilkins,Baltimore pp, 280–281

11. Dossing M, Khan J, Al Rabiah F(1994) Risk factors for chronic ob-structive lung disease in Saudi Arabia.Respir Med 88:519–522

12. Behera D, Jindal SK, Malhotra HS(1994) Ventilatory function in non-smoking rural Indian women using dif-ferent cooking fuels. Respiration61:89–92

13. Sandoval J, Salas J, Martinez GuerraML (1993) Pulmonary arterial hyper-tension and cor pulmonale associatedwith chronic domestic wood smoke in-halation. Chest 103:12–20