Chronic obstructive pulmonary disease related to wood smoke and impact of the combined exposure to tobacco

SUMMARY BACKGROUND Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2023 highlights the need to explore aetiotypes of chronic obstructive pulmonary disease (COPD) beyond the tobacco-smoking COPD. Exposure to wood smoke (WS) is a risk factor for COPD in women, but the effect of the combined exposure to tobacco smoke (TS) in the general population and among COPD patients, and the characteristics of WS-COPD are unclear. METHOD This was an analysis of data from PREPOCOL (Prevalence of COPD in Five Colombian Cities Situated at Low, Medium, and High Altitude), a random cross-sectional population-based study (n = 5,539) focusing on the effect of combined WS and TS exposure and WS-COPD characterisation. RESULTS Prevalence of COPD was significantly higher in those exposed to both WS and TS (16.0%) than in those exposed to WS (6.7%) or TS (7.8%) only (P < 0.001). Exposure to WS was associated with COPD in men (OR 1.53, P = 0.017). WS-COPD individuals were more frequently female, older, shorter and had higher forced expiratory volume in 1 sec (FEV1) (all P < 0.05). Those exposed to both WS and TS had more symptoms and worse airflow limitation (P < 0.001). CONCLUSIONS This was the first random population-based study showing that WS is an associated risk factor for COPD also in men, and that people exposed to both WS and TS have a significantly higher prevalence of COPD. Similarly, COPD subjects exposed to both types of smoke have more symptoms and greater airflow obstruction. This suggests an additive effect of WS and TS.

Chronic obstructive pulmonary disease (COPD) remains a highly prevalent condition, causing significant morbidity and mortality worldwide. 1,2][4] Although strong evidence remains insufficient, outdoor air pollution has been considered as a risk factor of COPD. 5,6On the other hand, long-term exposure to indoor air pollution derived from burning biomass fuels, including wood smoke (WS), has been related to COPD, chronic bronchitis (CB) and airflow obstruction in women, [7][8][9][10] less consistently in men, and is currently accepted as a risk factor for COPD. 2,85][16][17][18][19] However, knowledge about the prevalence and characteristics of patients with WS-COPD, and the effect of the combined exposure to WS and TS, is limited.Some studies have shown an increased risk of COPD and worse outcomes in smokers exposed to WS, 20,21 as well as worse oxygen saturation in COPD patients exposed to both WS and TS; 16 however, information about people chronically exposed to this combination, particularly from population-based studies, is scarce.
A better characterisation of biomass-and WS-COPD is necessary, not only because around 40% of the world population (about 2.8 billion people) continue to use solid fuels as household energy source, 22 which has a large negative impact on global respiratory health, 23 but because the treatment of biomass-related COPD could be different and its prevention could imply different approaches.
The PREPOCOL Study, 24 aimed to determine COPD prevalence in Colombia, where exposure to WS was found to be a risk factor, offers a good opportunity to further explore the impact of combined WS and TS exposure on the general population and among COPD individuals, and the incompletely known characteristics of WS-COPD.

Design and population
Information was obtained from the PREPOCOL Study, 24 a random cross-sectional, population-based study conducted in urban areas of five Colombian cities. Detailed information about sample size and standardisation of measurements is provided in the original article. 24Subjects were selected using a probabilistic, two-stage clustered sampling technique.After providing informed consent, adults of both sexes, aged �40 years, who underwent high-quality spirometry and answered a respiratory questionnaire, were included.Demographic, socio-economic, clinical and spirometry variables were collected.The study and informed consent were approved by the Institutional Research Ethics Committee (Comité de Ética en Investigaci ón de la Fundaci ón Neumol ógica Colombiana) Bogot á, Colombia..

Questionnaire and spirometry
We used a Spanish version of the Standardized Respiratory Questionnaire for Epidemiologic Studies of the American Thoracic Society (ATS-DLD-78A) with the following additional questions on WS exposure: Have you ever used wood for cooking habitually?If yes, for how many years?What type of fuel do you currently use for cooking?(Including wood as an option).Habitual use was defined as most days of the week.Pre-/post-bronchodilator (BD) spirometry (MicroLoop; Micro Medical, Rochester, UK) was performed according to ATS/ERS (European Respiratory Society) recommendations.Crapo et al.'s reference values were used. 25finitions COPD was defined by a post-BD ratio of forced expiratory volume in 1 sec (FEV 1 ) to forced vital capacity (FVC) of ,0.70, 2 and CB as an affirmative answer to the question: Have you ever had cough and expectoration for three or more months a year for at least two consecutive years?Cough and phlegm were considered present if an affirmative answer was obtained for the following questions, respectively: 1) Do you usually have cough?2) Do you usually bring up phlegm from your chest, not from the back of your nose?Usually was defined as most days of the week.
Group categorisation according to exposure to wood or tobacco smoke All participants (n ¼ 5,539) and subjects with COPD (n ¼ 494) were categorised in four groups: 1) WS group: exposed to WS �10 years and to TS ,10 packs/ year; 2) TS group: exposed to TS �10 packs/year and to WS ,10 years; 3) combined group (MS): exposed to both WS �10 years and to TS �10 packs/year; and 4) not exposed to either WS or TS.The cut-off point for WS exposure (�10 years as risk factor of COPD) was based on the previously published multivariate analysis in the same population; 24 The cut-off for TS of 10 packs/year is reasonably supported for discriminating the risk of COPD. 26For comparative analyses among COPD individuals, we focused on the three exposed groups; the non-exposed group was excluded (n ¼ 35, 7.1%).

Statistical analysis
Our analysis addressed COPD prevalence and factors associated with COPD and CB in the entire population.A logistic regression model was constructed using the variables that showed P , 0.1 in the univariate analysis.Odds ratios (ORs) for COPD were estimated according to exposure to WS, TS or both using nonexposed people as reference, and adjusting by sex, age, educational level, city of residence, self-reported history of TB and occupational exposure to vapours, gases, dust, and fumes (VGDF).After the first analysis, we focused on individuals with COPD to compare the following clinical and functional characteristics of three groups according to exposure to WS, TS or both (combined): age, sex, educational level, height, body mass index (BMI), symptoms (cough, phlegm, dyspnoea, wheezing), FVC, FEV 1 and FEV 1 /FVC.As recommended by Halbert et al., we grouped the participants into two categories for age-specific estimates: �40-64 and .64years. 27ifferences between COPD groups were evaluated using the v 2 test and analysis of variance.P , 0.05 was considered for statistical significance.Interactions between WS and TS exposures were explored.The statistical software SigmaStat v3.2 (Informer Technologies, Los Angeles, CA, USA) was used.

RESULTS
Prevalence and associated factors according to exposures in the entire population A total of 5,539 participants were included.Table 1 shows the distribution of the entire study population and the people with COPD.In comparison with non-COPD, participants with COPD were older, more frequently male, exposed to both WS and TS, and were more likely to have a history of TB and occupational exposure (all P , 0.001).Of the study population, 60% was exposed to WS: 30.9% (n ¼ 1,713) to WS only and 29.8% (n ¼ 1,650) to both WS and TS (combined).There were no significant differences in exposures between the cities.
The overall prevalence of COPD was 8.9% (n ¼ 494).According to exposures, the COPD prevalence was significantly higher in the combined group (i.e., exposed to both WS and TS) (16.0%) than in the TS group (7.8%) or the WS group (6.7%) (P , 0.001).Similarly, the prevalence of CB was higher in the MS group (9.0%) than in the TS group (5.2%) or the WS group (4.2%) (P , 0.001).
After adjusting for age, sex, smoking, educational level, occupational exposures and history of TB, exposure to WS �10 years was associated with COPD in both women (OR 1.84, 95% CI 1.31-2.60;P , 0.001) and men (OR 1.53, 95% CI 1.08-2.18;P ¼ 0.017).COPD prevalence increased significantly according to the duration of exposure to WS, reaching 23.2% in those exposed �30 years (Figure ).The unadjusted and adjusted ORs for COPD were significantly higher in the MS group than in the WS or TS groups (Table 2).No statistical interactions between WS and TS exposures were identified.

Characteristics of COPD individuals according to exposure
Of the 494 participants with COPD, 35 (7.1%) had no history of exposure to either WS or TS and these individuals were not considered for the comparative analyses.Table 3 presents the demographic, clinical and spirometry characteristics of the 459 participants with COPD (47.8% of whom were women) according to exposure to WS, TS or both (combined).Those with WS-COPD were predominantly female and had significantly lower height and higher BMI than those with TS-COPD (P , 0.05).Women exposed to WS (WS-COPD and combined) were older (P ¼ 0.036) and referred more frequently to dyspnoea (P ¼ 0.004); they also had lower height and higher BMI than women with TS-COPD (P , 0.001).Educational level was significant lower in people with COPD exposed to WS (WS-COPD and combined) than in those with TS-COPD (P , 0.001).3).

DISCUSSION
This study not only demonstrates that exposure to WS for �10 years is a contributing factor to COPD in women but also establishes, for the first time in a random population-based study, that this exposure is also associated with COPD in men.Moreover, individuals exposed to both WS and tobacco smoke (TS) exhibit a significantly higher prevalence of COPD and chronic bronchitis (CB) than those exposed to either WS or TS alone.Furthermore, our study innovatively † Exposed to WS �10 years and to TS ,10 packs/year.Supporting evidence for exposure cut-off points is provided in the Methods section.‡ Exposed to TS �10 packs/year and to WS ,10 years.Supporting evidence for exposure cut-off points is provided in the Methods section.§ Exposed to both WS �10 years and to TS �10 packs/year.Supporting evidence for exposure cut-off points is provided in the Methods section.OR ¼ odds ratio; COPD ¼ chronic obstructive pulmonary disease; CI ¼ confidence interval; aOR ¼ adjusted OR; WS ¼ wood smoke; TS ¼ tobacco smoke.showed that people with COPD exposed to both WS and TS are more likely to have persistent respiratory symptoms (cough and phlegm) and significantly greater airflow limitation (lower post-BD FEV 1 % and FEV 1 /FVC%) than those exposed to WS or TS.Finally, we found that people and women with WS-COPD, as a biomass COPD type, have different characteristics than those with TS-COPD: they are older and have significantly shorter height, higher BMI and lower educational level.These findings in individuals with combined exposure to WS and TS, both in the general population (higher prevalence and associations with COPD and CB) and in individuals with COPD (worse clinical and functional outcomes) suggest an additive adverse effect of combined exposure.Although the questionnaires used do not allow us to define the precise sequence of the combined exposure by length of exposure, combined exposure was simultaneous in many of our study participants.
Although the information about the effect of the combined exposure to biomass smoke (i.e., WS þ TS) is still scarce, some study findings are in line with our results. 20,21In a smokers cohort, Sood et al. found that self-reported WS exposure was independently associated with lower FEV 1 and a higher prevalence of airflow obstruction and CB. 20In the Lovelace Smokers Cohort, individuals with WS exposure experienced a more rapid decline of FEV 1 and worse quality of life than those without WS exposure. 21Recently, Olloquequi et al. described significantly lower oxygen saturation in patients with COPD and combined exposure to biomass smoke and TS. 16L ópez-Campos et al. found that patients with one other factor in addition to TS exposure had more chronic sputum production, worse scores in the COPD Assessment Test and greater longterm oxygen therapy requirement. 28ur findings suggest confluent pathophysiologic mechanisms due to combined exposure.Sood et al. showed that WS exposure interacted with aberrant promoter methylation of the p16 or GATA4 genes, increasing the risk of COPD and lower FEV 1 in smokers. 20Awji et al. found that WS enhances TS-induced inflammation in airway epithelial cells. 29he inflammatory pathways induced by biomass smoke/WS or TS could be different, 16,19,30,31 and could interact in enhancing the negative effects of the combined exposure.
As previously described, [11][12][13]32,33 we found that people with WS-COPD are predominantly female and older, and have significantly shorter height and higher BMI than TS-COPD. Cookig in developing countries is traditionally done by women who also spend more time indoors.Here, for the first time in a random population-based study, we show that exposure to WS �10 years was associated with COPD in men as well, probably due to poor household ventilation and the proximity of the kitchen, leading to prolonged exposure to high concentrations of pollutants from WS throughout the home.This finding is similar to that previously described in a meta-analysis.34 The older age observed in patients with WS-COPD suggests a different pattern of exposure-response than in those with TS-COPD, 11 whereby exposure over a longer period of time is probably required for the development of COPD.Similarly, adjusted by age, airflow obstruction was milder in those with WS-COPD as described in other studies, 11,12,29,32,33 suggesting also a lower decline of FEV 1 .33 However, no clear explanation for the lower height and higher BMI in the general population and in women exposed to WS (both WS-COPD and in the combined group) than in those with TS-COPD could be ascertained.The use of biomass fuels for cooking is associated with low socio-economic status; it is possible that the low birth weight of children, which is linked to maternal prenatal exposure to biomass fuels, 35 is associated with reduced height in adulthood.In addition, although we did not collect data on the race/ethnic origins of the participants, in Colombia, indigenous ethnicity is more frequent in people from rural areas, who are more likely to be exposed to biomass fuels than people from urban areas, where White and mixed races are more frequent.Andean Colombian indigenous groups have lower height and higher BMI than the general Colombian population.36 The higher frequency of cough, phlegm and CB in those with WS-COPD than in those with TS-COPD has been described.12,18 The novelty of this study lies in its revelation that the occurrence of these symptoms, particularly dyspnoea in women, is further elevated among those with concurrent exposure to both WS and TS.In addition, this is the first study to show that COPD individuals exposed to both WS and TS have greater airflow obstruction (lower FEV 1 and lower FEV 1 /FVC) than those exposed to either WS or TS.These observations might indicate the cumulative effects of extensive airway damage caused by WS, such as bronchial anthracofibrosis, 12,[15][16][17][18]35,37 along with damage induced by TS in the airways and lungs, suggesting interacting pathophysiological mechanisms.[19][20][21]28 It is not clear if the results of our study can be extrapolated to COPD caused by other types of biomass fuels.
Our study strengths include its randomised design, and a large and representative number of participants in total and in each of the groups of interest according to smoke exposure: wood, tobacco and combined.The large number of participants identified with COPD, both on the whole and in each group, allowed us to reliably analyse associations, characterise groups and accurately describe differences.There were also some limitations, mainly due to the cross-sectional design of the study; as information was based on questionnaire responses about past exposures and events, we were unable to confirm causality but only suggest associations.
Data on the intensity of WS exposure (ventilation conditions, proximity of the kitchen and type of stove used) were lacking.Nevertheless, questions about the length of exposure and type of fuel routinely used for cooking make the analysis and conclusions robust.
GOLD 2023 2 and COPD Lancet Commission 3 have highlighted the relevance of risk factors for COPD other than TS, and the need for a better characterisation and understanding of these aetiotypes.Our study presents important information about a type of biomass COPD due to indoor air pollution derived from WS and the impact of the combined exposure to WS and TS, and highlights the need for research on the underlying mechanisms.

Figure .
Figure.Prevalence of COPD according to length of wood smoke exposure (n ¼ 5,539).COPD ¼ chronic obstructive pulmonary disease.

Table 1 .
Demographic and clinical characteristics of participants.
P ¼ 0.091).Individuals in the combined group were more likely to have persistent cough (P ¼ 0.018) and phlegm (P ¼ 0.001), and had significantly lower post-BD FEV 1 % and FEV 1 /FVC% (P ¼ 0.002), both in the total group and in women only (P ¼ 0.003) than in the WS or TS groups (Table

Table 2 .
Unadjusted and adjusted ORs for COPD (airflow obstruction) by exposure (n ¼ 5,539).Adjusted by sex, age, educational level, city of residence, self-reported history of TB and occupational exposure to vapours, gases, dust, and fumes. *

Table 3 .
Demographic, clinical and spirometric characteristics in COPD groups by exposure (n ¼ 459).