In our recently published paper (open access version here) we describe the development of a multivariate measure of physical environmental deprivation for the 278 municipalities of Portugal, and demonstrate its strong relationship with mortality rates.
Our aim was to develop a measure that would aid understanding of the spatial inequalities in health that are persistently observed in Portugal but still unexplained. We intended that the resulting summary measure would be understandable, straightforward, ready-to-use, and available to policy makers, academics and other interested parties. To this end, we have made the multiple deprivation measure available to download from this page.
Following CRESH’s approach to quantifying multiple environmental deprivation in the UK and New Zealand we identified environmental factors relevant for population health in Portugal: air pollution (nitrogen dioxide, particulate matter and carbon monoxide), climate, drinking water quality (nitrates and trihalomethanes), green space availability and proximity to industry. We combined these datasets to produce an index – PT-MEDIx – that classified municipalities into six classes from the least to the most environmentally deprived.
We found a clear pattern in the spatial distribution of multiple environmental deprivation across Portugal – the sparsely-populated rural areas were least environmentally deprived, and urban and industrial settings had the highest levels (see map below).
Map: Spatial distribution of PT-MEDIx scores across Portuguese municipalities.
After taking account of population age structure and socioeconomic deprivation differences between the municipalities we found that PT-MEDIx was significantly associated with all-cause mortality and cancer (see graphs below): municipalities with the poorest physical environments had higher mortality rates than those with a more advantageous environment. The cancer result followed a dose-response pattern – one of the key requirements to prove causality in epidemiological research. Putting it in raw numbers, physical environmental deprivation was responsible for an annual excess of 6000 deaths.
Graphs: The relationship between PT-MEDIx and all-cause mortality (left) and cancer mortality (right). The mortality rate ratio (MRR) for each PT-MEDIx class is given relative to municipalities with a PT-MEDIx score of zero (MRR=1.0). The results account for age, sex and socioeconomic deprivation, and error bars indicate 95% confidence intervals.
Of course, no study is flawless. We were not able to include environmental factors known to affect population health, such as noise or indoor radon exposure. We also used relatively large geographical units of analysis, which might obscure some within-area heterogeneity. All these limitations, however, have resulted from data unavailability and confidentially issues, details that fall beyond our radius of action.
We have demonstrated the importance of area-level environmental deprivation for population health in Portugal. Hopefully, evidence from our Portuguese, UK and New Zealand studies will raise awareness among academics and policy makers that physical environment matters for health.
Reference: Ribeiro AI, Pina MdFd, Mitchell R. Development of a measure of multiple physical environmental deprivation. After United Kingdom and New Zealand, Portugal. The European Journal of Public Health. 2015. http://dx.doi.org/10.1093/eurpub/cku242. Published here. Open-access version here.