Methodological considerations regarding the relative roles of wood-smoke exposure and hypoxemia in cognitive impairment among patients with COPD

To the Editor,

We read with great interest the study by Gonzalez-Garcia et al. assessing cognitive impairment (CI) in patients with chronic obstructive pulmonary disease (COPD) at high altitude. ¹ The authors should be commended for tackling an important and understudied topic, particularly in light of the global burden of biomass fuel exposure and the increasing recognition of cognitive dysfunction as a clinically relevant comorbidity in COPD.

However, we feel that the conclusion that wood-smoke exposure and not hypoxemia is a risk factor for CI should be viewed with caution.

First, the cross-sectional design restricts the capacity to establish temporal or causal relationships. Because exposure history, oxygenation variables, and cognitive status were assessed concurrently, the results provide evidence for an association between wood-smoke exposure and CI but cannot determine whether biomass exposure led to cognitive decline or whether unmeasured factors influenced both. The title and conclusions may suggest a level of causal inference that is not supported by the study design. Associations found in cross-sectional studies are usually considered hypothesis-generating and should be confirmed in longitudinal studies before causal inferences can be drawn.

Second, the conclusion on hypoxemia may be premature. Oxygenation was assessed using resting arterial blood gases and oxygen saturation during the six-minute walk test. These measures provide important physiologic information, but may not be representative of cumulative hypoxic burden, nocturnal desaturation, sleep-disordered breathing, or long-term exposure to oxygen, all of which have been associated with cognitive dysfunction in patients with COPD. ² Therefore, the absence of a statistically significant association between the measured oxygenation variables and CI should not be interpreted as evidence that hypoxemia plays a lesser role than biomass exposure. Rather, it indicates that the specific oxygenation parameters examined in this cohort were not associated with CI.

Third, the authors correctly note limitations in exposure assessment, but the classification of wood-smoke exposure was based largely on self-reported cooking history and duration of exposure. There were no data on important determinants of the actual pollutant burden such as stove type, ventilation characteristics, household environment, daily duration of exposure and objective measurements of particulate matter or carbon monoxide. Therefore, misclassification of exposure cannot be ruled out. Furthermore, household biomass use is often associated with broader environmental and socio-economic factors that can have an independent impact on cognitive outcomes. Similarly, systematic reviews evaluating indoor air pollution and cognition have pointed to considerable heterogeneity in exposure assessment and the difficulty of separating out the effects of pollutants from contextual determinants. ³

We applaud the authors on their valuable contribution and concur that the association noted warrants further study. However, further prospective studies with detailed characterisation of exposure and longitudinal assessment of oxygenation are needed before concluding that biomass exposure is more strongly associated with cognitive impairment than hypoxemia in patients with COPD.