Ventilated cookstoves associated with improvements in respiratory health-related quality of life in rural Bolivia
Journal of Public Health |
Ventilated cookstoves associated with improvements in respiratory health-related quality of life in rural Bolivia
Donee Alexander 2
Jacqueline Callihan Linnes 1
Susan Bolton 0
Timothy Larson 3
0 School of Environmental and Forest Sciences, University of Washington , Seattle, WA 98195 , USA
1 Department of Bioengineering, University of Washington , Seattle, WA 98195 , USA
2 Center for Global Health, University of Chicago , Chicago, IL 60637 , USA
3 Department of Civil and Environmental Engineering, University of Washington , Seattle, WA 98915 , USA
A B S T R AC T Background Household air pollution (HAP) from combustion of biomass fuels worldwide is linked to asthma, respiratory infections and chronic pulmonary diseases. Implementation of ventilated cookstoves significantly reduces exposure to HAP. However, improvements in concurrent respiratory health-related quality of life (HRQoL) have not been previously evaluated with a standardized questionnaire. Methods The association between woodsmoke exposure and respiratory HRQoL outcomes was evaluated using an intervention study in a rural community in Bolivia. Indoor carbon monoxide (CO) levels from traditional stoves and from cookstoves with chimneys were analyzed alongside interview results of women heads-of-households using the St. George's Respiratory Questionnaire (SGRQ) in 2009 and 1-year post-intervention. Results Pronounced improvements in respiratory HRQoL and significant reductions of household CO levels followed installation of ventilated cookstoves. Stove implementation yielded lower indoor CO values and correlated positively with improved SGRQ scores. Conclusions This is the first use of a standardized respiratory HRQoL assessment to determine the impact of ventilated cookstove implementation on reducing HAP. This preliminary study utilizes the SGRQ as a valuable tool enabling analysis of these health effects in relation to other respiratory disease states.
biomass fuel; carbon monoxide; CO; household air pollution; respiratory health-related quality of life; St; George's Respiratory Questionnaire
Nearly 2.7 billion people rely on biomass fuels (wood, crop
residues, dung and charcoal) for their primary domestic
energy needs.1 – 3 In many developing countries, burning of
biomass accounts for almost one-half of domestic energy
production and can reach as high as 95% in some
countries.1,4,5 Biomass fuels are typically burned in open-pit fires
or inefficient cookstoves without sufficient ventilation. This
practice results in exposure to health hazards including
particulate matter (PM) and carbon monoxide (CO) that are 10 –
20 times greater than World Health Organization air quality
guidelines.6 – 10 Studies have shown that household air
pollution (HAP) from indoor combustion of biomass fuels in
developing countries is linked to acute respiratory infections,
chronic obstructive pulmonary disease (COPD), asthma and
other health problems.11 – 14 In 2010, global mortality due to
HAP from solid fuels was estimated to be more than 3.5
million deaths per year.15
To date, no standard questionnaire has been used to evaluate
respiratory health related to ventilated cookstove interventions.
Standardized respiratory questionnaires provide insight into the
effects that improved air quality have on health-related quality
of life (HRQoL) in a manner that can be compared with other
respiratory disease states and interventions. The St. George’s
Respiratory Questionnaire (SGRQ) is a well-established tool
for quantifying health status in chronic pulmonary diseases and
has been shown to correlate well with disability due to
disease.16 It encompasses a broad range of respiratory diseases
and has previously been used to measure HRQoL for industrial
diseases such as pneumoconiosis.17 This study represents the
first use of the SGRQ in ventilated cookstove interventions.
While most questionnaires regarding respiratory health focus
on identifying symptoms only, the SGRQ measures the overall
respiratory HRQoL providing additional clinical insight.16 The
SGRQ allows researchers to determine the impact of disease
on a subject’s daily life and represents a valuable evaluation tool
to assess the benefits of ventilated cookstove interventions.
Reducing exposure to HAP is one of the main goals of
improved biomass cookstove implementation projects. However,
due to logistical difficulties and long sampling periods, PM
monitoring equipment can be problematic and expensive in
stove implementation studies.18 This study used CO
concentrations as a marker for woodsmoke exposure. CO passive
diffusion tubes have been shown to be a reliable proxy for PM ,
2.5 mm (PM2.5) in prior woodsmoke studies.11,18,19 In a recent
review of woodsmoke exposure studies, 8-h averages for CO
ranged from 1.2 to 60 mg/m3 [1– 49 parts per million
( ppm)].10 In contrast, National Ambient Air Quality Standards
set by the US Environmental Protection Agency require that
8-h average CO concentrations of 10 mg/m3 not to be
exceeded more than once per year.20
The study was conducted using a paired pre-test and post-test
assessment of HRQoL and CO levels following installation of
ventilated cookstoves. The research was conducted in the
Acacio province of northern Potos´ı, Bolivia (altitude range
2900 – 3300 m). In this region, most women cook indoors
over open-pit fires for 3 – 5 h per day without the benefit of a
chimney. Cooking patterns do not vary seasonally, and while
stoves provide heat to the kitchen, they are not used for this
purpose. Traditional, clay cookstoves are placed in the corner
of the kitchens (see Supplementary data for photo). The
primary fuel source is hardwood trees. The primary language
and culture of the region is Quechua, although some
individuals also speak Spanish. Communities range in size from 8
to 60 households. The village of Tuquiza was selected for
assessment based on the use of indoor biomass fuel for
cooking and planned provision of improved cooking stoves
with roofs and chimneys designed by the University of
Washington chapter of Engineers without Borders (EWB).
No other co-interventions occurred during the study period.
Fifty Yanayo cookstoves were provided to 80% of the
households in Tuquiza in September and October, 2009.
These stoves were provided to all households who wanted
them, regardless of whether they chose to participate in the
study. Only women who cooked indoors over open-pit fires
and who reported their overall health as ‘ok’ or ‘good’ were
eligible to be part of the study. Women who reported their
health as ‘bad’ were questioned as to the reason for their poor
health. If it was determined that the illness was unrelated to
respiratory health but prohibited the women from taking part
in daily activities, they were excluded from the study.
During the months of August and September, 2009,
preimplementation respiratory HRQoL assessments of the 31
women heads-of-households who elected to join this study
were performed using the SGRQ. CO levels were assessed in
each household for two consecutive 24-h periods prior to
intervention. Between the initial assessment and the 1-year
follow-up, 10 women dropped out of the study due to
nonrespiratory illness, death or re-location. In 2010, between the
months of August and October, a post-implementation
follow-up was performed for the remaining 21 women (35%
of the households) using the same SGRQ survey and CO
level detection methods. One woman did not sufficiently
complete the post-intervention questionnaire for analysis. In
total, 20 subjects were analyzed in the paired pre- and
The SGRQ is a well-established method to quantify health
status in pulmonary diseases. The SGRQ includes questions
related to symptoms (e.g. coughing, chest pain, etc.), activity
(e.g. walking uphill, walking at a rapid pace) and impact (e.g.
disrupted sleep, expectations of improvement in health) and
weights each of these sections for a total score. Scores range
from 0 to 100 for each SGRQ category with higher scores
indicating poorer respiratory HRQoL than lower scores. The
surveys were verbally translated from a validated Peruvian
Spanish version of the SGRQ into Quechua by a bi-lingual
interpreter during the face-to-face interview.21
Gastec brand CO passive diffusion tubes were placed in
the women’s households both pre- and post-intervention for
two consecutive, roughly 24-h periods. The time of CO tube
placement and collection were recorded. From these data,
time-weighted-average values were calculated to find CO in
ppm. The CO passive diffusion tubes were read blind and
recorded onsite by two research associates, immediately
following the sampling period.
The independent variables in this study were the
implementation status of the Yanayo cookstove and the household
CO levels. The dependent variable in the study was the
resulting SGRQ scores (symptoms, activity, impact and total).
These outcomes were compared pre- versus post-stove
intervention. All 20 women in the cohort reported using their
ventilated cookstoves as their sole cooking device following the
2009 intervention (see Supplementary data for additional
Questionnaire results were analyzed with the SGRQ
Analysis package for Microsoft Excel.16 Pre- and
postintervention outcomes were compared using Wilcoxon
signed-rank tests. Spearman rank correlation tests were used
to determine the relationships of SGRQ outcomes to
exposures (CO concentrations, age) using XLSTAT (Addinsoft).
Power analyses were performed using JMP software (SAS).
Table 1 summarizes the results of the study in addition to the
ages of the women in this study. Average CO levels and
all categories of SGRQ scores were improved (lower)
following the intervention. Detailed within-subject changes in each
woman’s SGRQ results from pre-intervention to
postintervention are shown in Fig. 1. Improvement in a subject’s
HRQoL is associated with a decrease in the scores from
pre- to post-intervention, while a worsening of HQRoL is
associated with an increase. After the intervention, women
using Yanayo stoves with chimneys had significantly improved
SGRQ scores than the pre-implementation scores of these
same women using traditional cookstoves without chimneys.
Table 2 describes the changes in the SGRQ and CO scores
for individuals comparing post- to pre-implementation
(negative scores are improvements). Overall, 85% (17/20) of the
women in the study had improved total SGRQ scores, 75%
Table 1 Age, pre-intervention SGRQ scores and pre-intervention
Symptoms Activity Impact Household CO level (ppm)
Pre-intervention Mean (95% confidence interval)
Post-intervention Mean (95% confidence interval)
47.2 (42.2 –54.2)
48.2 (43.2 –55.2)
58.2 (52.3 –63.0)
51.8 (41.6 –62.1)
68.4 (59.4 –77.3)
54.3 (48.0 –60.5)
10.8 (8.9– 12.7)
35.3 (25.3 –45.4)
42.4 (30.7 –54.2)
40.7 (28.1 –53.2)
29.9 (20.8 –39.1)
2.3 (1.6 –3.0)
aAll 20 participants were non-smoking female heads-of-households of
subsistence farming families interviewed during August and September
of 2009 (pre-intervention) and again in September of 2010
bLower SGRQ scores and household CO levels are considered preferable
to higher ones.
(15/20) had improved activity and impact scores and 65%
(13/20) had improved symptoms scores. While activity,
impact and total SGRQ scores had statistically significant
decreases following the stove intervention (P , 0.002,
P , 0.002 and P , 0.001, respectively), Symptoms scores
were not significantly improved. However, a breakdown of
questions contributing to the symptoms score shows that
attacks of wheezing and severity and length of attacks of chest
trouble were all significantly reduced post-implementation
(P , 0.04, 0.03 and 0.007, respectively).
Pre- and post-intervention CO concentrations in Tuquiza
confirmed the reduction of CO in all homes following
implementation of Yanayo stoves. Prior to stove
implementation, the average CO level was 10.8 ppm. Following the
intervention, CO concentrations in households decreased, on
average, to 20% of the pre-implementation levels (2.3 ppm,
P , 0.0001).
Changes in the individual subjects’ SGRQ scores were
plotted versus changes in their households’ CO levels from
pre- to post-intervention in Fig. 2. The Spearman rank
coefficients and P values were also reported on each subfigure. For
total scores, Spearman R value is 0.24 (P ¼ 0.32), Activity R
value is 0.23 (P ¼ 0.32) and Impact R value is 0.42 (P ¼ 0.7),
Symptoms R value is 20.42 (P ¼ 0.1).
Main findings of this study
In this study, the use of Yanayo cookstoves was associated
with significantly better overall HRQoL and indoor CO levels
than traditional stove use. Total, activity and impact HRQoL
scores on the standardized SGRQ improved significantly
following the use of the new stoves with chimneys. A positive
correlation with CO change was also observed for these
scores post implementation of the Yanayo stoves.
What is already known on this topic
A study by Ferrer et al. has reported mean SGRQ symptoms,
activity, impact and total scores in healthy populations of
women at 7.8, 14.8, 4.3 and 8.2 respectively.22 The women in
the study represented a cross-section of ages, educations and
socioeconomic statuses in Spain. A study of pneumoconiosis
(due to inhalation of coal dust) in men in Hong Kong resulted
in SGRQ scores for symptom, activity, impact and total
scores of 38.0, 44.5, 34.2 and 39.4, respectively.17
In a review evaluating the health effects of woodsmoke
exposure, Naeher et al. reported average daily CO values
ranging from 1.2 – 60 mg/m3 ( 0.97 – 48.6 ppm).19 CO
values from our study fall within this range.
VENTILATED COOKSTOVES ASS OCIAT ED WITH RESPIRATORY H EALTH
Table 2 Change in SGRQ scores
D SGRQ score
222.8 (233.2 to 212.5)
29.4 (223.3 to 4.6)
227.7 (240.3 to 215.1)
224.3 (235.7 to 213.0)
P value (LSNb)
aPost-intervention score minus pre-intervention score.
bLSN (least significant number). The sample size needed to prove
statistical significance for each change in scores at an alpha of 0.05.
The actual sample size was 20 subjects.
What this study adds
According to the authors of the SGRQ, clinically significant
improvements in health status of COPD patients were
associated with a 4 point decrease in SGRQ total scores.16 While
no clinical diagnosis of disease has been made in the present
study, Total scores declined significantly, on average by more
than 22 points. This is a clear indication of substantially
improved HRQoL for women after implementation of
Yanayo cookstoves (P ¼ 0.001). It is interesting to note that
symptoms scores, which are the only impact measured by
most respiratory HRQoL studies, were not statistically
significantly improved following the stove implementation. Despite
this, a number of the specific questions related to Symptoms
measurements did show statistically significant improvements
While improvements in HRQoL due to indoor stove use
have been described in a recent study,23 our use of the
standardized SGRQ allows for analysis of the severity of these
effects in a manner that can be related to other disease states.
Both healthy and pneumoconiosis populations have lower
(better) SGRQ scores than the Tuquiza women prior to
Yanayo stove intervention. It is likely that, in addition to
respiratory disease presence, the differences in population,
environment and limited access to medical care play a role in
At the time of the post-intervention surveys, Yanayo stoves
with chimneys had been used by the participants for 1 year.
While no stove use monitoring data is available, all
participants destroyed their traditional stove upon receiving the
Yanayo cookstove and began using their Yanayo cookstoves
as their sole cooking apparatus.
R = 0.23
p = 0.32
R = 0.24
p = 0.32
This preliminary study using standardized HRQoL
assessments to measure exposure and outcomes shows promising
results suggesting that a large scale case-control study will
provide even more definitive conclusions.
Limitations of this study
The Yanayo stoves were part of an intervention that included
building a separate kitchen with corrugated metal roofing in
which the ventilated cookstove was installed. This holistic
intervention was aimed at reducing the CO and HAP in the
homes. Thus, the stoves alone cannot be considered the sole
cause of the improvements in HRQoL or CO reduction.
However, the ventilated cookstoves themselves could not
have been installed in the traditional thatch roofed homes.
We found an association between woodstove
implementation and improved HRQoL as measured by SGRQ activity,
impact and total scores. Decreases in individual subjects’
SGRQ total, activity and impact scores following stove
intervention were associated with concurrent decreases in CO in
the households according to the Spearman rank correlations.
While not statistically significantly correlated with the drastic
CO reductions seen in the 1-year timeframe of this study, the
correlations might be detectable in a longer case-controlled
study. It is also possible that other factors influenced the
relationship between quantitative measurement of CO levels and
true personal exposure such as the placement of the monitor
in the home, wind direction, backdraft from chimneys,
changed seating positions and a woman’s time away from the
While a positive correlation between reductions in CO and
SGRQ activity, impact and total scores occurred, there was a
weak suggestion of the opposite relationship for the SGRQ
symptoms scores. One reason may be the small number of
subjects. According to a retrospective power analysis, a
follow-up study would require at least 79 participants to
determine statistical significance (P , 0.05) of the symptoms
scores. However, because statistically significant
improvements occurred for other scores, it is also possible that the
intervention had no effect on respiratory symptoms. This
would agree with a previous HRQoL study that found large
improvements in non-respiratory symptoms, such as eye
irritation, but little change in respiratory symptoms following
ventilated cookstove implementation.23
As a paired pre-test post-test design, this preliminary study
did not include non-intervention control subjects. It is
therefore possible that the completion of the questionnaire a
second time could account for the changes seen in the SGRQ
scores. However, given the positive correlation with the CO
readings taken in the subject’s homes and the 1-year period
between the tests it is unlikely that the changes in SGRQ
scores are due to repetition of the questionnaire.
Finally, it is possible that education on the health effects of
HAP may have influenced participants’ answers to the
SGRQ. As a self-reported measurement, participants could
have artificially inflated or deflated their HRQoL, either
consciously or unconsciously, during the study. Potentially,
survey responses to the SGRQ could have been biased by a
subjects desire to receive a stove or other further assistance.
Participation in this study, however, was not required for
obtaining a Yanayo stove, making this bias unlikely. Subjects
did not receive the score results from either their pre- or
postintervention questionnaires, nor were there any ‘right’ or
‘wrong’ answers. Thus, it is unlikely that learning effects
occurred during the study unless subjects felt that they ‘ought’ to
have experienced changes due to the intervention. Of all the
SGRQ categories, symptoms scores might be expected to be
affected by potential learning effects as these could most easily
be associated with reduction in HAP. However, symptoms
were the only measurement that did not significantly improve.
Here, we report the first use of a validated respiratory
HRQoL instrument, the SGRQ, to determine the impact of
reducing HAP through the implementation of cooking stoves
with chimneys. Women cooking on new Yanayo stoves had
significantly better respiratory HRQoL than prior to the
implementation when these same women cooked on
traditional stoves. However, post-implementation HRQoL was
still worse than that of healthy populations, and is instead,
more similar to previous research findings in populations
suffering from pneumoconiosis. The use of Yanayo stoves
reduced air pollution exposures as measured by CO levels
and was associated with an improvement in SGRQ scores.
Our results suggest that HAP from poorly vented stoves may
lead to detrimental effects on women’s respiratory HRQoL,
and may be ameliorated by improving stove technology and
ventilation. The use of a larger case-controlled study following
participants for a longer period will likely yield even further
concrete evidence correlating HRQoL improvements with
decreases in HAP.
As organizations expand their efforts to improve global
health and indoor air quality, further research on the impact
of ventilated cookstoves is necessary to ensure that both HAP
and HRQoL are enhanced by these efforts. Because of its
focus on HRQoL aspects beyond symptoms, the SGRQ
presents a valuable evaluation tool to assess the benefits of
ventilated cookstoves and to compare these to other respiratory
Supplementary data are available at PUBMED online.
The authors would like to thank the women of Tuquiza who
were the subjects of this study, Nathaniel Wilson for helping
with data collection, the University of Washington chapter
The Osberg endowment for providing partial funding support.
1 World Resources Institute UNEP, United Nations Development Programme . World Resources 1998 - 1999: A Guide to the Global Environment . Oxford: Oxford University Press, 1998 .
2 Ezzati M , Kammen D. The health impacts of exposure to indoor air pollution from solid fuels in developing countries: knowledge, gaps, and data needs . Environ Health Persp 2002 ; 110 : 1057 - 68 .
3 Reddy A , Williams R , Johansson T. Energy after Rio: Prospects and Challenges . New York : United Nations Publications , 1996 .
J O U R N A L O F P U B L I C H E A LT H 4 5 6 7 8 9 Desai M , Mehta S , Smith K. Indoor Smoke from Solid Fuels: Assessing the Environmental Burden of Disease at National and Local Levels . Geneva, Switzerland: Organization WH , 2004 .
Environ Health Perspect 2006 ; 114 : 373 - 8 .
Ezzati M , Mbinda B , Kammen D. Comparison of emissions and residential exposure from traditional and improved biofuel stoves in rural Kenya . Environ Sci Technol 2002 ; 34 : 578 - 83 .
Zhang J , Smith K (eds). Indoor air pollution: formaldehyde and other carbonyls emitted from various cookstoves . In: Indoor Air 1996 : Proceedings of the 7th International Conference on Indoor Air Quality and Climate , Nagoya, Japan, 1996 .
WHO. Air quality guidelines: global update 2005 . Geneva, Switzerland: WHO, 2005 .
Bull World Health Organ 2000 ; 78 : 1078 - 92 .
10 Naeher L , Brauer M , Lipsett M et al. Woodsmoke health effects: a review . Inahal Toxicol 2007 ; 19 : 67 - 106 .
11 Bruce N , McCracken J , Albalak R et al. Impact of improved stoves, house construction and child location on levels of indoor air pollution exposure in young Guatemalan children . J Expo Anal Environ Epidemiol 2004 ; 14 (Suppl. 1): S26 - 33 .
12 Ezzati M , Kammen D. Quantifying the effects of exposure to indoor air pollution from biomass combustion on acute respiratory infections in developing countries . Environ Health Perspect 2001 ; 109 : 481 - 8 .
13 Smith K , Mehta S , Maeusezahl M. Indoor smoke from household solid fuels . In: Ezzati MR , Lopez AD , Murray CJL (eds). Comparative Quantification of Health Risks: Global and Regional Burden of Disease Due to Selected Major Risk Factors. Geneva, Switzerland: WHO , 2004 , 1435 - 93 .
14 Smith KR , Samet JM , Romieu I et al. Indoor air pollution in developing countries and acute lower respiratory infections in children . Thorax 2000 ; 55 : 518 - 32 .
15 Lim SS , Vos T , Flaxman AD et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990 - 2010: a systematic analysis for the Global Burden of Disease Study 2010 . Lancet 2013 ; 380 : 2224 - 60 .
16 Jones PW , Quirk FH , Baveystock CM . The St George's respiratory questionnaire . Respir Med 1991 ; 85 ( Suppl . B): 25 - 31 ; discussion 3- 7 .
17 Tang WK , Lum CM , Ungvari GS et al. Health-related quality of life in community-dwelling men with pneumoconiosis . Respiration 2006 ; 73 : 203 - 8 .
18 McCracken J , Smith K. Emissions and efficiency of improved woodburning cookstoves in highland Guatamala . Environ Int 1998 ; 24 : 739 - 47 .
19 Naeher L , Smith K , Leaderer B et al. Carbon monoxide as a tracer for assessing exposures to particulate matter in wood and gas cookstove households of highland Guatemala . Environ Sci Technol 2001 ; 35 : 575 - 81 .
20 National Air Quality Standards (NAAQS) [database on the Internet] . U.S. Environmental Protection Agency , 1990 [cited 3 October 2007 ]. http://www.epa. gov/air/criteria.html (6 March 2013 , date last accessed) .
21 Jones PW. St George's Respiratory Questionnaire Manual. London: St George 's University of London, 2009 , 14 .
22 Ferrer M , Villasante C , Alonso J et al. Interpretation of quality of life scores from the St .: George's Respiratory Questionnaire. Eur Respir J 2002 ; 19 : 405 - 13 .
23 Diaz E , Bruce N , Pope D et al. Self-rated health among Mayan women participating in a randomised intervention trial reducing indoor air pollution in Guatemala . BMC Int Health Hum Rights 2008 ; 8 : 7 .