Worldwide associations between air quality and health end-points: are they meaningful?

The World Health Organization (WHO) provides data on national indices of health, environment and economy. When we were asked, why air pollution is negatively correlated with cancer mortality, our first response (presumably the mortality data are not age-adjusted) was not sufficient to explain the paradox.

A table including all-cause, cancer and childhood mortality, life expectancy, gross national product per person, smoking prevalence, physician density and particulate matter (PM10) per country (N = 193) was developed. For explorative purposes weighted cross-sectional multiple linear regressions models were built.

Air pollution is positively correlated with infant and overall mortality and negatively with life expectancy. This might not only depict a true causal effect of PM10 because air quality is also an indicator of a country's prosperity and general state of environment. Cancer mortality is negatively correlated with PM10. However, this association turns positive when economic or health system indicators are controlled.

The World Health Organization's world-wide data sets demonstrate the large disparity of our world. A careful and professional approach is needed as interpretation is difficult, especially for lay persons. Therefore, with publicly available data WHO should also provide interpretation and guidance.