You are here

  • Home
  • Archive
  • Volume 69, Issue 5
  • A rural community intervention targeting biomass combustion sources: effects on air quality and reporting of children's respiratory outcomes

Article Text

Article menu
Environment
Original article
A rural community intervention targeting biomass combustion sources: effects on air quality and reporting of children's respiratory outcomes
  1. Curtis W Noonan1,
  2. Tony J Ward1,
  3. William Navidi2,
  4. Lianne Sheppard3
  1. 1Center for Environmental Health Sciences, Department of Biomedical Sciences, The University of Montana, Missoula, Montana, USA
  2. 2Department of Mathematical and Computer Sciences, Colorado School of Mines, Golden, Colorado, USA
  3. 3Department of Biostatistics and Environmental and Occupational Health Sciences, University of Washington, Seattle, Washington, DC, USA
  1. Correspondence to Professor Curtis W Noonan, Center for Environmental Health Sciences, Department of Biomedical Sciences, 32 Campus Drive, The University of Montana, Missoula, Montana 59812, USA; curtis.noonan{at}umontana.edu

Abstract

Objective Improvements in urban air quality are largely driven by controls on industrial and mobile source emissions, but such factors may have limited influence on many rural environments where biomass combustion (eg, wood stoves) serves as the primary source of fine particulate matter (PM2.5). The authors tracked changes in children's respiratory health during a wood stove intervention in a rural mountain valley community heavily impacted by wood smoke-derived PM2.5.

Methods Community-wide impacts on children's health were assessed by prospectively collecting surveys from parents of school children during four winter periods in Libby, Montana. Generalised estimating equations with a logit link were used to estimate the effect of reduction in ambient PM2.5 on wheeze prevalence and other reported symptoms and infections.

Results Over 1100 wood stoves were replaced with new lower emission wood stoves or other heating sources. Ambient PM2.5 was 27.6% lower in the winters following the changeout programme compared with baseline winters. There was a 26.7% (95% CI 3.0% to 44.6%) reduced odds of reported wheeze for a 5 μg/m3 decrease in average winter PM2.5. Lower ambient PM2.5 was also associated with reduced odds for reported respiratory infections, including cold (25.4% (95% CI 7.6% to 39.7%)), bronchitis (54.6% (95% CI 24.2% to 72.8%)), influenza (52.3% (95% CI 42.5% to 60.5%)) and throat infection (45.1% (95% CI 29.0% to 57.6%)).

Conclusion This wood stove intervention provided a unique opportunity to prospectively observe health benefits resulting from a targeted air pollution reduction strategy in a rural community.

  • Asthma
  • respiratory infections
  • biomass smoke
  • wood stove
  • epidemiology
  • air pollution
  • statistics

https://doi.org/10.1136/oemed-2011-100394

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What this paper adds

    • Many communities experience poor air quality due to the use of biomass combustion devices for residential heating (eg, wood stoves), but little is known about the effect of targeted interventions on this source.

    • The present study found substantial improvements in ambient air quality following a large-scale community-wide programme to replace older wood burning devices with new devices that meet more current emission guidelines.

    • Reporting of children's respiratory symptoms and infections were reduced with improvements in air quality.

    • These findings may serve as useful guidelines for predicting the effect of such targeted interventions in communities similarly impacted by biomass combustion.

    In recent decades, urban air quality has improved following the introduction of health-based ambient air standards. Such reductions have been associated with measurable health benefits. Notably, it was estimated that a decrease of 10 μg/m3 in the long-term average concentration of fine particulate matter (PM2.5) air pollution translated to a mean increase in life expectancy of 0.61 years ± 0.20.1 Improvements in urban air quality are largely driven by controls on industrial and mobile source emissions, but such factors may have limited influence on many rural environments where residential biomass combustion (eg, wood stoves) serves as a large source of PM2.5. Residential-generated biomass smoke can be a substantial PM source in communities that have a large proportion of homes using wood heating,2–7 as well as in many developing countries where biomass fuel for cooking is common.8 A limited number of epidemiological studies in communities with high levels of residential biomass smoke exposure have found morbidity and mortality impacts that are similar to those observed in communities with primarily industrial and mobile sources of PM.8 Little is known regarding the potential for beneficial health impacts of interventions targeting these domestic sources of PM in rural communities, but such communities are held to the same air quality standards as their urban counterparts.

    We investigated Libby, Montana, a small rural mountain valley community. Similar to several communities in the region,7 the greatest contributor to ambient fine particulate matter was wood smoke.9 Domestic biomass combustion emissions coupled with cold temperature inversions common in these mountain valley communities resulted in an exceedance of the US Environmental Protection Agency's (EPA) annual PM2.5 National Ambient Air Quality Standard. The community initiated a large-scale wood stove changeout programme to replace over 1100 older wood burning devices with improved lower emission wood stoves or other heating appliances. During four winter periods, we tracked changes to ambient air quality and parent-reported frequency of respiratory symptoms and infections among school children in this community. This study was undertaken in response to recent calls for accountability studies to investigate the health impact of air quality interventions driven by regulatory standards.10

    Methods

    Study location

    Libby is a small community, occupying approximately 3.4 km2 of Lincoln County, Montana. The community sits in a deep valley floor on the Kootenai River at an elevation of 628 m with surrounding mountains rising to an elevation of 1828 m The PM2.5 non-attainment area is a 370 km2 quadrangle region extending beyond Libby proper. There is no natural gas line serving the city or the surrounding communities. County-based median household income in the year 2000 was $24 276 compared with $41 994 for the USA. The estimated number of older conventional wood burning appliances was 1175 among approximately 4200 households in the Census county subdivision.11

    In addition to the season-dependent air quality concerns, it is noteworthy that this community also has an historical legacy of environmental contamination relevant to respiratory disease. The largest mining facility in the area was a vermiculite operation beginning in the early 1890s and continuing until its closure in 1990. The mining site, located just outside of Libby, produced 80% of the world's vermiculite, a product used for insulation, soil amendments and various other products. The vermiculite from this site was naturally contaminated with amphibole asbestos, and several studies and health surveillance activities have investigated asbestos-related disease and occupational as well as environmental asbestos exposures among adults in the town of Libby and surrounding communities.12–14

    Compliance monitoring

    We assessed changes to air quality utilising compliance air monitoring data from the Montana Department of Environmental Quality at one central site over a 6-year period. The compliance site has been located on the roof of the Lincoln County Health Department in downtown Libby since 2003. The site is located nearly 1 km from the elementary school and approximately 3 km from the middle school. Previous investigations have supported the use of this sampling location as a representative site for the community.15 The compliance monitoring included running 24 h PM2.5 mass samplers (Federal Reference Method BGI PQ200s) and continuous PM2.5 samplers (Rupprecht and Patashnik, East Greenbush, New York, USA; tapered element oscillating microbalances).

    Data collection

    Evaluation of community-wide impacts on children's health was assessed by prospectively collecting surveys from parents of school children during four winter periods. Winter was defined as the months November through February. At the end of each winter period, parents were asked to report on their child's symptoms and health conditions during the previous 2 months, the type of heating used in their home and the occurrence of tobacco smoking in the home. The survey was adapted from core questionnaires on asthma, rhinitis and eczema from the International Study of Asthma and Allergies in Childhood, which was designed to screen for the prevalence and severity of asthma and allergic disease in defined populations.16 The survey for the present study included additional questions on respiratory infections based on the National Health and Nutrition Examination Survey. Efforts to encourage response included notifications sent through school newsletters, minor award incentives for returned surveys and repeat mailings.

    Other data included meteorological measures and a surrogate indicator of community influenza activity. Daily temperature, wind speed, relative humidity and precipitation data were obtained from an archived database of the Western Regional Climate Center. Community influenza activity per county population during each winter period was based on hospital diagnosis records with indication of influenza in January through March of a given year.

    Statistical analysis

    Using all survey response data, we focused on individual symptom reports of presence of wheeze as our primary outcome with other asthma-related symptoms and respiratory infections as secondary outcomes. As a surrogate for age, we used grouped school grade (eg, grades 1 and 2, grades 3 and 4, etc). To estimate the effect of average wintertime PM2.5 on wheeze prevalence and other reported symptoms and infections, adjusting for school grade, community influenza and presence of a wood stove in the home, we used a generalised estimating equation with a logit link. We further assumed that repeated measures on a given subject were exchangeable. For a subset of wood stove homes, we were able to identify date of wood stove changeout, allowing for a comparison of reported health conditions between wood stove homes that had and had not changed their stove prior to survey.

    Results

    Figure 1 illustrates the timeline for the wood stove programme. By the end of the programme, 1147 stoves were replaced with new EPA-certified wood stoves (n=736), wood inserts (n=57), wood furnaces (n=11), pellet devices (n=170) or other non-biomass combustion heating devices such as propane, oil or electric (n=86). Some existing stoves were reconditioned to meet EPA emission guidelines (n=81) or surrendered (n=6) due to the presence of another non-biomass combustion heating source. Thus, approximately 98% of the original estimate of 1175 older conventional wood stoves were replaced with cleaner burning appliances. For tracking the timing of the wood stove changeout programme, we used cumulative percentage of completed changeouts with 1147 as the denominator. As illustrated in figure 1, winter ambient PM2.5 concentrations in Libby were 27.6% lower in the winters following the changeout programme (19.7 μg/m3 in 2007/2008 and 2008/2009) compared with the baseline winters prior to the changeout programme (27.2 μg/m3 in 2003/2004 and 2004/2005). By the end of the programme, the community was in compliance with the federal PM2.5 standard.

    Figure 1
    Figure 1

    Study overview, ambient PM2.5 and cumulative changeout of wood stoves in Libby, MT 2003–2009. Open circles are ambient PM2.5 every third day prior to November 2005 and daily thereafter. Black line = 60-day moving average for PM2.5. Grey line = cumulative wood stove changeouts. Parent-reported child health surveys (black arrows) were conducted at the end of four winter periods to collect data on respiratory symptoms and conditions in the previous 2 months.

    Summaries of the community-level variables during the study period are presented in table 1. The largest proportion of wood stove changeouts occurred between the 2005/2006 and 2006/2007 winters. Ambient temperature did not vary dramatically with a winter mean range of only 0.88°C over the four periods. As anticipated, wind speed during the winter periods was extremely low and daily wind gust maximums averaged over the winter periods never exceeded 7.4 km/h. The surrogate measure of influenza activity in the community suggested substantial differences across the 4 years with rates of influenza per 1000 population ranging from 1.54 in the 2006/2007 winter to 3.69 in the 2007/2008 winter.

    View this table:
    Table 1

    Summary of factors captured over winter study periods

    A total of 1713 surveys were returned during the study period, representing 920 children. Returned surveys corresponded to a 44%–51% response rate among parents of 1st–8th grade students. Efforts to capture parent surveys as students progressed through the high school were less successful with a 43% response rate in 2007 but falling to 21% of all attending high school students in the final year. Table 2 describes the characteristics of children and their homes among survey respondents. Wood stove use was reported in the homes of 42%–43% of students over the study period. Tobacco smoking was reported in 31%–37% of homes. Crude frequencies of reported symptoms and infections by winter are presented in table 3.

    View this table:
    Table 2

    Proportion (%) of respondents by survey year according to selected characteristics

    View this table:
    Table 3

    Frequency (%) of reporting of respiratory symptoms and infections in prior 2 months

    Table 4 presents the effect estimates for respiratory symptoms and conditions with respect to PM2.5 reductions, adjusting for school grade, presence of a wood stove in the home and community influenza activity. There was a 26.7% (95% CI 3.0% to 44.6%) reduced odds of reported wheeze for a 5 μg/m3 decrease in average winter PM2.5. Other respiratory symptoms were not strongly associated with PM2.5, while symptoms most likely to be caused by irritant effects of wood smoke were reduced, notably frequency of itchy/watery eyes (33.2% (95% CI 19.4% to 44.6%)) and sore throat (31.6% (95% CI 17.9% to 43.0%)). Lower ambient PM2.5 was also associated with reduced odds for reported cold infection (25.4% (95% CI 7.6% to 39.7%)), bronchitis (54.6% (95% CI 24.2 to 72.8%)), influenza (52.3% (95% CI 42.5% to 60.5%)) and throat infection (45.1% (95% CI 29.0% to 57.6%)).

    View this table:
    Table 4

    Adjusted OR* and 95% CI for effects of ambient PM2.5 reduction and use of wood stove in the home on parent-reported childhood symptoms and infections across four winter periods

    The presence of a wood stove in the home was also included in the analysis model, but this factor was not associated with increased reporting of any symptoms or health conditions with the exception of morning tightness in chest which showed lower odds with respect to having a wood stove in the home (table 3). We also evaluated the potential for the presence of a wood stove in the home to modify the relationship between ambient PM2.5 and reporting of symptoms and conditions. When stratifying the primary analysis by the presence or absence of a wood stove in the home, there were no discernable differences in the effect estimates for ambient PM2.5 and reporting of symptoms and conditions (data not shown). For a subset of survey respondents with wood stoves, we were able to capture the date when their home's wood stove was changed. Only 17% (20/115) of wood stove users has changed their stove prior to the first winter survey, 70% (100/142) had changed their stove prior to the second winter survey and 95% (160/168) had changed their stove prior to the third winter. Comparing homes that had changed wood stoves with homes that had not changed wood stoves, there was no difference in reporting of wheeze for any of the three winter survey years (data not shown). All stoves from survey homes had been changed prior to the final winter survey.

    We also assessed the impact of individually reported influenza, rather than our hospital-based indicator of community-level influenza activity, on effect estimates for asthma symptoms and other infectious conditions. Estimates for ambient PM2.5 and reporting of wheeze and infectious health conditions remained virtually unchanged (eg, 26.1% reduction on odds or reported wheeze per 5 μg/m3 decrease in average winter PM2.5 compared with 26.7%). To evaluate the potential for reporting bias as the wood stove programme progressed, parents were also asked to report on additional non-respiratory symptoms (ie, dizziness, nausea, vomiting, stomach pain and diarrhoea) that were assumed a priori to not be associated with ambient PM2.5. Overall, these results suggested no association with ambient PM2.5, although estimates for a few of the non-respiratory symptoms were sensitive to the approach to adjusting for influenza (see Supplemental table 1).

    The first winter of the survey study was highly influential because it was the only year with elevated ambient winter PM2.5, whereas PM2.5 concentrations consistently remained low in the subsequent three winters. In consideration of this influential first winter period, we conducted sensitivity analyses on the above findings, restricting the evaluation to only those children with survey data in the first winter of the study and one or more subsequent winters (n=261). The point estimates for most outcomes were similar to the primary analysis, but the CIs were wider. Odds of reported wheeze was 30% lower (95% CI −7% to 54%) for a 5 μg/m3 decrease in average winter PM2.5. Reductions in PM2.5 also remained protective for reporting of itchy/watery eyes, sore throat, bronchitis, influenza and throat infection (see Supplemental table 2).

    Discussion

    The overall stove changeout programme was comprehensive. Based on the best available data, almost 98% of the estimated number of older conventional stoves were replaced. The programme resulted in a greater than 27% reduction in winter period ambient PM2.5 and enabled the community to achieve compliance with the federal PM2.5 standard. Our primary outcome, reduction in parent-reported childhood wheeze, was associated with reductions in ambient PM2.5 (OR (95% CI) 0.73% (0.55% to 0.97%)) per 5 μg/m3 reduction in winter ambient PM2.5). Other respiratory symptoms that are commonly associated with childhood asthma were not associated with reductions in PM2.5. In communities with substantial contributions to ambient air from residential wood combustion, ambient PM has been associated with increased symptoms among asthma cohorts17 18 and asthma hospitalisations or asthma emergency department visits.19–21 Among other factors, these latter studies are not directly comparable to our study because they were conducted on the daily time scale rather than the seasonal scale.

    The frequency of most asthma-related symptoms was low in this study (≈12% or lower), but reporting of the child being woken at night due to cough was observed among 21%–31% of children across the years. Although this question has been used to assess symptoms among children with asthma, coughing at night can occur among non-asthmatic children with respiratory infections. Reporting for two other symptoms that could occur among non-asthmatic children, itchy/watery eyes and sore throat, were similarly high and strongly associated with PM2.5. These symptoms may indicate a response to the irritant effect of smoke exposure. Eye discomfort has been observed among subjects in several studies of adults living in homes with high levels of biomass smoke exposures from cookstoves.22 23 It is reasonable to speculate that the influence of wood smoke exposure on eye symptoms is relevant to the more general child population and not limited to susceptible individuals. A recent study of symptoms reported among Children's Health Study participants exposed to wildfire smoke found positive associations between smoke exposure and reporting of symptoms. Although the association with respiratory symptoms was modified by airway size, an indicator of compromised breathing, the association with eye symptom was not limited to susceptible children.24

    We found that parental reporting of children's infectious health conditions was associated with changes in ambient PM2.5. Reductions in ambient wood smoke PM were associated with reduced reporting of bronchitis, cold, throat infection and influenza among children. Our risk estimate for bronchitis was based on few reports (n=20 during the first winter), but these findings were consistent with a recent study in British Columbia that found residential wood smoke exposure to be associated with increased risk of outpatient and inpatient visits for infant bronchiolitis.25 A recent meta-analysis of studies evaluating biomass smoke-exposed rural populations in developing countries found a strong summary estimate for risk of acute respiratory infection (OR (95% CI) 3.5% (1.9% to 6.4%)).26 Our observed associations between PM2.5 and reporting of cold, sore throat and throat infection suggested that changes in reported health outcomes were not limited to the lower respiratory tract. Upper respiratory conditions and associated symptoms have also been observed among children exposed to smoke from nearby wildfires.27 Previous studies not specific to biomass exposure have also observed associations between PM and otitis media.28 29 We did observe a suggestive association between PM2.5 exposure and ear infection, but our study population was grade school age, and the low frequency of reporting for this condition limited our ability to precisely estimate the effect.

    The presence of a wood stove in the home was not associated with reporting of respiratory symptoms or infections in models that concurrently evaluated ambient PM2.5. This finding is consistent with our previous cross-sectional observations in this community in which parent reporting of wheeze or other asthma-related symptoms among children were not associated with the presence of a wood stove in the home.30 In a comparison of a subset of homes, we also did not find a difference in reporting of wheeze for children living in homes that had already changed their stove compared with those living in homes that had not yet changed their stove. Several factors may account for this seemingly counter-intuitive finding. First, it is possible that individuals susceptible to wood smoke effects self-select out of wood stove homes, thereby limiting the number of susceptible individuals in such homes. Second, central ambient levels of wood smoke can affect indoor exposures via infiltration, allowing for indoor wood smoke exposures among homes without wood stoves. We were unable to directly evaluate indoor versus outdoor sources of wood smoke, but other studies have described substantial residential infiltration of ambient biomass smoke.31 32 Finally, observations by our group and others suggest that the impact on indoor air quality following replacement of older wood stoves with upgraded technology wood stoves is highly variable, at times resulting in no overall improvement in indoor air quality.33–36 The reasons for these variable effects on indoor air quality are unknown, but one factor could be improper operation of the newer technology stoves. The changeout programme included training of residents in the proper usage of the installed stove, but the effectiveness of this training had not been assessed.

    Several features made this setting ideally suited for evaluating community health measures concurrent with a source-targeted intervention on ambient PM2.5. The community had a history of dramatic and predictable seasonal fluctuations in PM that put them out of compliance with federal standards. The high concentrations of PM during the cold temperature season were derived almost exclusively from one source, residential wood combustion.9 Specifically, a source apportionment study identified residential wood combustion as the source for >80% of the PM2.5 concentration measured during the heating season, and there are no significant industrial sources in the Libby air shed.2 In response to these concerns, the community managed a well-funded intervention to change out the older conventional wood stoves that achieved near-complete compliance. All these factors contributed to this rural community's success in achieving substantial reductions in ambient winter PM2.5. There are few similar wood stove programmes with documented community-wide reductions in PM and no such programmes have concurrently documented health impacts. The small population size of this community limited our ability to observe effects of small magnitude. Despite this concern and the resulting imprecision in risk estimates, we still observed reductions in frequency of reported wheeze and other symptoms and infections with CIs that supported the a priori study hypotheses. A study focused more precisely on clinical objective measures in a susceptible population such as asthmatics may have been warranted. Our survey data suggested, however, that the benefits of reduced ambient winter PM2.5 included more general infectious health conditions and were not limited to a specific population such as asthmatic children.

    The prospective approach of this study was an important strength for tracking the impact of a large-scale community wood stove changeout programme. The quasi-experimental design, however, limited our ability to account for changes over time that occurred independently of the wood stove programme. The inclusion of a control community would have allowed for better characterisation of time-dependent variation in ambient conditions and other factors. To partially address this concern, we were able to evaluate our results with respect to changes in meteorology and community-level influenza frequency. Because this was a natural rather than purposive design, we also had limited ability to control the timing of the study observations. Ideally, we would have captured an additional year of survey data prior to the start of the wood stove changeout programme to more fully describe baseline symptom and disease frequency.

    Other limitations impact on the generalisability of our health outcomes findings. Our exposure assessment with respect to health outcomes was limited to ambient measures, and a great deal of variability in personal-level exposure would be anticipated due to different home environments and activity patterns. When evaluating one such individual-level home characteristic, the current study found no differences in health outcomes when stratifying the analyses according to the presence or absence of a wood stove in the home. As with any survey-based study, our findings were also highly dependent upon the population that chose to participate as well as the accuracy with which the participants understood and accurately responded to the symptom and disease questions. Based on our findings that risk estimates were similar when our analysis was restricted to those who had responded in the most influential year and in at least one of the subsequent years, we are less concerned about a differential impact in non-participation over time. Nevertheless, the overall response rates were low across the years (<50%), so our findings may not generalise to the entire childhood population in the community. Inaccuracies could also be reflected in differential reporting bias due to community knowledge of the wood stove programme or bias due to the type of heating in a respondent's home. In an attempt to evaluate the potential for biased reporting, we collected information on symptoms that were not anticipated to be associated with PM exposure. In general, we found that reporting of non-respiratory symptoms such as nausea and vomiting were not associated with the presence of a wood stove in the home nor with changes in ambient PM2.5.

    Conclusions

    It is often difficult to track the impact of any one programme or policy on a given air pollutant due to complex airsheds, multiple sources and other factors that change over time. Our results were consistent with observations from urban areas with a greater mixture of particulate matter sources that have shown beneficial health impacts following air quality improvements.1 37–40 This community-wide wood stove changeout programme demonstrated a successful approach for achieving substantial and sustained reductions of ambient PM2.5, and these reductions were associated with beneficial health impacts. Other rural wood smoke-impacted communities share similar features,7 and the overall magnitude of change in PM2.5 (≈28% reduction) may serve as a reasonable estimate for intervention programme planning in smoke-impacted communities.

    Acknowledgments

    The authors would like to thank the parents of Libby children for their participation in this study. We also thank Libby school administrators and staff, the Montana Department of Environmental Quality, Saint John's Lutheran Hospital and the Lincoln County Health Department for assistance with data collection. During this study, we received important advice from the HEI Research Committee and the HEI Science staff. Most importantly, we would like to thank Kathi Hooper, Director of Lincoln County Environmental Health Department, for her tireless support of this study.

    References

    1. ↵
      1. Pope CA 3rd.,
      2. Ezzati M,
      3. Dockery DW
      . Fine-particulate air pollution and life expectancy in the United States. N Engl J Med 2009;360:376–86.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Ward TJ,
      2. Rinehart L,
      3. Lange T
      . The 2003/2004 Libby, Montana PM2.5 Source Apportionment Research Study. Aerosol Sci Tech 2006;40:166–77.
      OpenUrlCrossRefWeb of Science
    3. ↵
      1. Fairley D
      . Daily mortality and air pollution in Santa Clara County, California: 1989–1996. Environ Health Perspect 1999;107:637–41.
      OpenUrlPubMedWeb of Science
    4. ↵
      1. Larson T,
      2. Gould T,
      3. Simpson C,
      4. et al
      . Source apportionment of indoor, outdoor, and personal PM2.5 in Seattle, Washington, using positive matrix factorization. J Air Waste Manag Assoc 2004;54:1175–87.
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Manchester-Neesvig JB,
      2. Schauer JJ,
      3. Cass GR
      . The distribution of particle-phase organic compounds in the atmosphere and their use for source apportionment during the Southern California Children's Health Study. J Air Waste Manag Assoc 2003;53:1065–79.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. McGowan JA,
      2. Hider RN,
      3. Chacko E,
      4. et al
      . Particulate air pollution and hospital admissions in Christchurch, New Zealand. Aust N Z J Public Health 2002;26:23–9.
      OpenUrlCrossRefPubMed
    7. ↵
      1. Ward T,
      2. Lange T
      . The impact of wood smoke on ambient PM2.5 in northern Rocky Mountain valley communities. Environ Pollut 2010;158:723–9.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Naeher LP,
      2. Brauer M,
      3. Lipsett M,
      4. et al
      . Woodsmoke health effects: a review. Inhal Toxicol 2007;19:67–106.
      OpenUrlPubMedWeb of Science
    9. ↵
      1. Ward TJ,
      2. Palmer CP,
      3. Noonan CW
      . Fine particulate matter source apportionment following a large woodstove changeout program in Libby, Montana. J Air Waste Manag Assoc 2010;60:688–93.
      OpenUrlCrossRefPubMed
    10. ↵
      Health Effects Institute Accountability Working Group. Assessing Health Impact of Air Quality Regulations: Concepts and Methods for Accountability Research. Cambridge, MA: Health Effects Institute Accountability Working Group, 2003.
    11. ↵
      1. Eagle B,
      2. Houck JE
      . Phase I of the Libby, Montana Woodstove Changeout Program. Beaverton, OR: OMNI Environmental Services, Inc., 2007.
    12. ↵
      1. Noonan CW,
      2. Pfau JC,
      3. Larson TC,
      4. et al
      . Nested case-control study of autoimmune disease in an asbestos-exposed population. Environ Health Perspect 2006;114:1243–7.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Peipins LA,
      2. Lewin M,
      3. Campolucci S,
      4. et al
      . Radiographic abnormalities and exposure to asbestos-contaminated vermiculite in the community of Libby, Montana, USA. Environ Health Perspect 2003;111:1753–9.
      OpenUrlPubMedWeb of Science
    14. ↵
      1. Larson TC,
      2. Antao VC,
      3. Bove FJ
      . Vermiculite worker mortality: estimated effects of occupational exposure to Libby amphibole. J Occup Environ Med 2010;52:555–60.
      OpenUrlCrossRefPubMed
    15. ↵
      Montana Department of Environmental Quality. An E-BAM Study to Determine the Validity of Particulate Measurements at the Libby-Courthouse Annex Air Monitoring Site. Helena, Montana: Department of Environmental Quality, Air Resources Management Bureau, 2009.
    16. ↵
      1. Asher MI,
      2. Keil U,
      3. Anderson HR,
      4. et al
      . International Study of Asthma and Allergies in Childhood (ISAAC): rationale and methods. Eur Respir J 1995;8:483–91.
      OpenUrlAbstract
    17. ↵
      1. Slaughter JC,
      2. Lumley T,
      3. Sheppard L,
      4. et al
      . Effects of ambient air pollution on symptom severity and medication use in children with asthma. Ann Allergy Asthma Immunol 2003;91:346–53.
      OpenUrlPubMedWeb of Science
    18. ↵
      1. Yu O,
      2. Sheppard L,
      3. Lumley T,
      4. et al
      . Effects of ambient air pollution on symptoms of asthma in Seattle-area children enrolled in the CAMP study. Environ Health Perspect 2000;108:1209–14.
      OpenUrlPubMedWeb of Science
    19. ↵
      1. Lipsett M,
      2. Hurley S,
      3. Ostro B
      . Air pollution and emergency room visits for asthma in Santa Clara County, California. Environ Health Perspect 1997;105:216–22.
      OpenUrlPubMedWeb of Science
    20. ↵
      1. Norris G,
      2. YoungPong SN,
      3. Koenig JQ,
      4. et al
      . An association between fine particles and asthma emergency department visits for children in Seattle. Environ Health Perspect 1999;107:489–93.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Sheppard L,
      2. Levy D,
      3. Norris G,
      4. et al
      . Effects of ambient air pollution on nonelderly asthma hospital admissions in Seattle, Washington, 1987–1994. Epidemiology 1999;10:23–30.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Diaz E,
      2. Smith-Sivertsen T,
      3. Pope D,
      4. et al
      . Eye discomfort, headache and back pain among Mayan Guatemalan women taking part in a randomised stove intervention trial. J Epidemiol Community Health 2007;61:74–9.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Romieu I,
      2. Riojas-Rodriguez H,
      3. Marron-Mares AT,
      4. et al
      . Improved biomass stove intervention in rural Mexico: impact on the respiratory health of women. Am J Respir Crit Care Med 2009;180:649–56.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Mirabelli MC,
      2. Kunzli N,
      3. Avol E,
      4. et al
      . Respiratory symptoms following wildfire smoke exposure: airway size as a susceptibility factor. Epidemiology 2009;20:451–9.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Karr CJ,
      2. Demers PA,
      3. Koehoorn MW,
      4. et al
      . Influence of ambient air pollutant sources on clinical encounters for infant bronchiolitis. Am J Respir Crit Care Med 2009;180:995–1001.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Po JY,
      2. FitzGerald JM,
      3. Carlsten C
      . Respiratory disease associated with solid biomass fuel exposure in rural women and children: systematic review and meta-analysis. Thorax 2011;66:232–9.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Kunzli N,
      2. Avol E,
      3. Wu J,
      4. et al
      . Health effects of the 2003 Southern California wildfires on children. Am J Respir Crit Care Med 2006;174:1221–8.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Brauer M,
      2. Hoek G,
      3. Smit HA,
      4. et al
      . Air pollution and development of asthma, allergy and infections in a birth cohort. Eur Respir J 2007;29:879–88.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Brauer M,
      2. Gehring U,
      3. Brunekreef B,
      4. et al
      . Traffic-related air pollution and otitis media. Environ Health Perspect 2006;114:1414–18.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Noonan CW,
      2. Ward TJ
      . Environmental tobacco smoke, woodstove heating and risk of asthma symptoms. J Asthma 2007;44:735–8.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Allen R,
      2. Larson T,
      3. Sheppard L,
      4. et al
      . Use of real-time light scattering data to estimate the contribution of infiltrated and indoor-generated particles to indoor air. Environ Sci Technol 2003;37:3484–92.
      OpenUrlPubMed
    32. ↵
      1. Barn P,
      2. Larson T,
      3. Noullett M,
      4. et al
      . Infiltration of forest fire and residential wood smoke: an evaluation of air cleaner effectiveness. J Expo Sci Environ Epidemiol 2008;18:503–11.
      OpenUrlCrossRefPubMed
    33. ↵
      1. Allen R,
      2. Leckie S,
      3. Millar G,
      4. et al
      . The impact of wood stove technology upgrades on indoor residential air quality. Atmos Environ 2009;43:5908–15.
      OpenUrlCrossRefWeb of Science
    34. ↵
      1. Noonan CW,
      2. Ward TJ,
      3. Navidi W,
      4. et al
      . Assessing the Impact of a Wood Stove Replacement Program on Air Quality and Children's Health. Research Report 162, Boston, MA: Health Effects Institute, 2011.
    35. ↵
      1. Ward TJ,
      2. Boulafentis J,
      3. Simpson J,
      4. et al
      . Lessons learned from a woodstove changeout on the Nez Perce Reservation. Sci Total Environ 2011;409:664–70.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Ward TJ,
      2. Palmer C,
      3. Bergauff M,
      4. et al
      . Results of a residential indoor PM2.5 sampling program before and after a woodstove changeout. Indoor Air 2008;18:408–15.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Pope CA 3rd.,
      2. Rodermund DL,
      3. Gee MM
      . Mortality effects of a copper smelter strike and reduced ambient sulfate particulate matter air pollution. Environ Health Perspect 2007;115:679–83.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Clancy L,
      2. Goodman P,
      3. Sinclair H,
      4. et al
      . Effect of air-pollution control on death rates in Dublin, Ireland: an intervention study. Lancet 2002;360:1210–14.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. Hedley AJ,
      2. Wong CM,
      3. Thach TQ,
      4. et al
      . Cardiorespiratory and all-cause mortality after restrictions on sulphur content of fuel in Hong Kong: an intervention study. Lancet 2002;360:1646–52.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Laden F,
      2. Schwartz J,
      3. Speizer FE,
      4. et al
      . Reduction in fine particulate air pollution and mortality: extended follow-up of the Harvard Six Cities study. Am J Respir Crit Care Med 2006;173:667–72.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Funding Funding was provided by Health Effects Institute (#4743-RFA04-4/06-4) and National Center for Research Resources (NCRR) (COBRE P20RR017670).

    • Competing interests None.

    • Ethics approval Ethics approval was approved by University of Montana Institutional Review Board.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Data will be made available per the requirements and procedures of the funding agency, Health Effects Institute.