Measuring long-term exposures that vary widely over time  


Combining Individual- and Group-Level Exposure Information.  Child Carbon Monoxide in the Guatemala Woodstove Randomized Control Trial


In the Guatemalan highlands, replacing open-fire woodstoves with a chimney woodstove reduced 48-hour mean child carbon monoxide exposure by 51% (95% confidence interval 46%-55%).


  • John P. McCracken
  • Joel Schwartz
  • Nigel Bruce
  • Murray Mittelman
  • Louise M. Ryan
  • Kirk R. Smith


Citation Epidemiology 2009; 20(1):127-136 Context The Randomized Exposure Study of Pollution Indoors and Respiratory Effects (RESPIRE) trial was designed to test the impact of reduced woodsmoke exposure on incidence of acute lower respiratory tract infections over about 1.5 years among Guatemalan children aged 0 to 18 months.  Carbon monoxide was used as a surrogate for woodsmoke.  Abstract Background: Epidemiology frequently relies on surrogates of long-term exposures, often either individual-level short-term measurements or group-level based on long-term characteristics of subjects and their environment.  Whereas individual-level measures are often imprecise due to within-subject variability, group-level measures tend to be inaccurate due to residual between-subject variability between groups.  Rather than choose between these error-prone estimates, we borrow strength from each by use of mixed-model prediction and we compare the predictive validity. Methods: We compared alternative measures of long-term exposure to carbon monoxide (CO) among children in the RESPIRE woodstove randomized control trial during years 2003 and 2004.  The main study included 1932 repeated 48-hour-averrage personal CO measures among 509 children from 0 to 18 months of age.  We used a validation study with additional CO measures among a random subsample of 70 of the children to compare the predictive validity of individual-level estimates (based on observed short-term exposures), group-level estimates (based on stove type and other residential characteristics), and mixed-model predictions that combine these 2 sources of information. Results: The estimated error variance for mixed-model prediction was 63% lower than the individual-level measure based on the exposure data and 58% lower than the corresponding group-level measure. Conclusions: When both individual- and group-level estimates are available but imperfect, mixed-model prediction may provide substantially better measures of long-term exposure, potentially increasing the sensitivity of epidemiologic studies to underlying causal relations. Policy Implications Due to concerns about health impacts of exposure to woodsmoke, rural Guatemala residents using open fires are being encouraged to use improved stoves. Web link Keyword(s) carbon monoxide, woodsmoke, Guatemala