Empirical evidence of the public health benefits of tropical forest conservation in Cambodia: a generalised linear mixed-effects model analysis

Lancet Planet Health. 2017 Aug;1(5):e180-e187. doi: 10.1016/S2542-5196(17)30081-5. Epub 2017 Aug 4.

Abstract

Background: Potential synergies between public health and environmental protection that offer new opportunities for achieving health and sustainable development targets have been postulated. However, empirical evidence of the effect of ecosystem degradation and protection on public health outcomes is scarce, which restricts policy makers' ability to assess the net health effects of land-use change.

Methods: We used generalised linear mixed-effects models to analyse data for 35 547 households in 1766 communities from the Cambodian Demographic Health Surveys to investigate the relation between health and protected areas across deforestation gradients in Cambodia between Feb 1, 2005, and April 30, 2014. Diarrhoea, acute respiratory infection, and fever in children younger than 5 years were used as population health indicators. Dense and mixed forest coverage were derived from Open Development Cambodia, and forest loss was calculated from 2000 to 2004, 2004 to 2009, and 2009 to 2014. The incidence of non-specific illness and injury in people older than 15 years was used as a negative control. Our analyses included rich pseudo-panel data (combining cross-sectional datasets from 2005, 2010, and 2014) that accounted for socioeconomic, demographic, and behavioural characteristics, and had a negative control, approximating a quasi-experimental study design.

Findings: Deforestation of dense forest was associated with an increased incidence of diarrhoea (p=0·007), fever (p=0·0495), and acute respiratory infection in children (p=0·003). For example, a 10 percentage point increase in loss of dense forest was estimated to be associated with an increase of 14·1% (95% CI 2·6-35·8) in the incidence of diarrhoea in children younger than 5 years per household in the 2 weeks before the Cambodian Demographic Health Surveys. Protected area coverage, but not type, was associated with decreased incidences of diarrhoea (p=0·028) and acute respiratory infection (p=0·030). Apart from an association between mixed forest coverage and increased incidence of diarrhoea, forest coverage was not associated with any health outcomes.

Interpretation: Deforestation is associated with increased risk of several major sources of global childhood morbidity and mortality. Although causal mechanisms are unclear, our findings suggest that protected areas could help to alleviate the global health burden, presenting new possibilities for simultaneous achievement of public health and conservation goals.

Funding: Ministry of Education of Singapore.