Asthma Admissions, PM2.5 and a Proposed Low-Emission Zone
A public health unit compares age-standardised asthma emergency admissions in 2023 across four districts of a Mediterranean island. District A is rural: mean PM2.5 8 µg/m³, NO2 12 ppb, admissions 210 per 100,000. B is a residential town: 14, 25, 310. C is an inner-city area with congested roads and older housing: 22, 40, 420. D is a port area: 12, 18, 260. Smoking prevalence is 22–24% in all four districts. The unit proposes a low-emission zone in C. A councillor argues that C's higher admissions simply reflect its older, poorer population and that the low-emission zone will waste money. The unit has air-quality monitors and hospital admission data but no individual-level patient data.
1. Interpret the data and assess the councillor's claim. What confounders and what causal criteria should be considered?
Read the worked answer
The data show a positive ecological association: as PM2.5 and NO2 rise from A to C, age-standardised admissions rise from 210 to 420 per 100,000. The association is not perfect: D has lower PM2.5 and NO2 than B but lower admissions, 260 versus 310, so the gradient is not linear and area-level factors matter. The councillor is right that C differs in age structure, housing and income, and these could confound the comparison; age standardisation removes only age, not poverty, smoking, occupation or access to care. However, confounding does not explain everything: the size of the gradient, consistency across two pollutants, dose–response pattern and biological plausibility all support a causal contribution. What is missing is individual-level data, temporal data showing whether admissions rise on high-pollution days, and evidence from an intervention. The claim that the low-emission zone will waste money is not supported by the data, but neither is a precise estimate of benefit.
2. Explain the biological mechanism linking PM2.5 exposure to asthma exacerbation, and why the effect may be larger in C.
Read the worked answer
PM2.5 particles are small enough to reach the alveoli. There they activate alveolar macrophages and epithelial cells, generating reactive oxygen species and pro-inflammatory cytokines such as IL-6 and TNF-α. This increases airway inflammation, mucus secretion and vascular permeability, and enhances bronchoconstriction. In people with asthma, the inflamed airways are already hyperresponsive; PM2.5 and NO2 lower the threshold for allergen- and exercise-induced bronchoconstriction, so exposures tolerated by healthy people trigger exacerbations. NO2 also directly irritates the airway epithelium and may increase susceptibility to viral infection. The effect is plausibly larger in C because older housing with damp and poor ventilation, plus more people living near traffic, increases exposure and vulnerability. The mechanism explains why admissions can rise within days of pollution episodes, not only after years of exposure.
3. Design an evaluation of the low-emission zone. State the outcomes, comparison groups and main threats to validity.
Read the worked answer
Evaluate with a difference-in-differences design: compare C with B and D before and after implementation, using at least 12 months before and 12–24 months after, and adjust for season and weather. Primary outcome: age-standardised asthma emergency admissions. Secondary outcomes: GP consultations for asthma, inhaled corticosteroid and reliever prescriptions, school absence, and measured PM2.5 and NO2 at fixed monitors. Include a pre-specified analysis plan and check that the zone does not simply displace traffic. Because admission data are area-level, results will be ecological; individual-level linkage or a patient cohort would strengthen inference. Monitor confounders such as smoking, housing and access to care, and report uncertainty. A fall in C relative to B and D would support benefit; no change would not prove no effect if exposure reduction was small. The evaluation should be funded as part of the policy, not after it.