Download Study

General

Study Overview

Title:

Does deforestation increase malaria prevalence? Evidence from satellite data and health surveys (pre-analysis plan 2 of 2: All countries)

Study is 3ie funded:

No

Study ID:

RIDIE-STUDY-ID-5b6ab5379f030

Initial Registration Date:

12/01/2017

Last Update Date:

08/16/2018

Study Status:

Completed

Change History for Status

Changed On	Previous Value
08/16/2018	Ongoing

Location(s):

Angola
Bangladesh
Benin
Burkina Faso
Burundi
Cape Verde
Congo - Kinshasa
Cote Divoire
Dominican Republic
Egypt
Ethiopia
Gabon
Ghana
Guinea
Guyana
Haiti
Honduras
Indonesia
Kenya
Lesotho
Liberia
Malawi
Mali
Morocco
Mozambique
Namibia
Nepal
Nigeria
Pakistan
Peru
Philippines
Rwanda
Senegal
Sierra Leone
Swaziland
Tanzania
Timor-leste
Togo
Uganda
Zambia
Zimbabwe

Abstract:

Malaria remains a substantial health burden in many countries. Pathways through which deforestation has been found to increase malaria risk include forest-cover effects, forest-cover-loss effects, and socio-economic associations. A growing number of studies have examined whether deforestation increases malaria prevalence in humans, with the majority of these analyses concluding that it does. But these studies have been based on a small number of countries using less-than-ideal forest data at the aggregate jurisdictional level. In this paper we combine fourteen years of 30-meter resolution satellite data on forest loss with tens of thousands of surveys of malaria and fever in children in more than 40 countries. Based on methods pre-specified in a pre-analysis plan, we test the ex-ante hypotheses that forest-cover loss increases malaria prevalence and that intermediate levels of forest cover have highest malaria prevalence. We further test ex ante hypotheses related to disaggregations; i.e., that the forest-cover-loss effect is greater in Latin America and Africa than Asia, greater at earlier stages of a forest transition, and greater for smaller cuts.

Registration Citation:

Busch, J. and Bauhoff, S., 2017. Does deforestation increase malaria prevalence? Evidence from satellite data and health surveys (pre-analysis plan 2 of 2: All countries). Registry for International Development for Impact Evaluations (RIDIE). Available at: 10.23846/ridie125

Categories:

Environment and Disaster Management
Health, Nutrition, and Population
Multisector

Additional Keywords:

Cost-effectiveness analysis; Mediation analysis;

Secondary ID Number(s):

Principal Investigator(s)

Name of First PI:

Jonah Busch

Affiliation:

Center for Global Development

Name of Second PI:

Sebastian Bauhoff

Affiliation:

Center for Global Development

Study Sponsor

Name:

Norwegian Agency for International Development

Study Sponsor Location:

Norway

Research Partner

Name of Partner Institution:

Type of Organization:

Location:

Intervention

Intervention Overview

Intervention:

Tropical deforestation, generally intended to obtain land for agriculture or pasture.

Private Intervention Details:

Theory of Change:

Tropical deforestation is hypothesized to lead to higher malaria prevalence, based on a large ecological literature and a small but growing empirical literature, mostly confined to a few sites in a few countries.

Multiple Treatment Arms Evaluated?

Yes

Implementing Agency

Name of Organization:

Many organizations are carrying out deforestation

Type of Organization:

Other

Program Funder

Name of Organization:

Many organizations are funding deforestation

Type of Organization:

Other

Intervention Timing

Intervention or Program Started at time of Registration?

Yes

Start Date:

01/01/1990

End Date:

Evaluation Method

Evaluation Method Overview

Primary (or First) Evaluation Method:

Regression with controls

Other (not Listed) Method:

Additional Evaluation Method (If Any):

Difference in difference/fixed effects

Other (not Listed) Method:

Method Details

Details of Evaluation Approach:

Core: -Multivariate regression on cross-sectional data controlling for co-variates Other analyses: -Multivariate regression with fixed effects on panel data controlling for co-variates -Tests of pre-specified disaggregation hypotheses -Mediation analysis -Cost-effectiveness analysis

Private Details of Evaluation Approach:

Outcomes (Endpoints):

Primary outcomes: -fever -malaria (rapid test) -malaria (lab test)

Unit of Analysis:

Child under the age of 5 living in a rural area

Hypotheses:

Primary: Malaria is higher at higher levels of deforestation and intermediate levels of forest cover Secondary: The effect of deforestation on malaria is greater in Latin American and Africa than Asia; greater for small cut sizes than large cut sizes; and greater when forest cover is greater The effect of deforestation on malaria lasts for around seven years

Unit of Intervention or Assignment:

~5.5km x 5.5km grid cells (unit of aggregation of forest data)

Number of Clusters in Sample:

Thousands

Number of Individuals in Sample:

~470,000 for fever. ~60,000 for malaria (rapid test). ~57,000 for malaria (lab test)

Size of Treatment, Control, or Comparison Subsamples:

Unknown in advance. Some significant fraction of cells experienced positive levels of deforestation.

Supplementary Files

Analysis Plan:

Pre-analysis plan FINAL 16Nov17.docx

Other Documents:

Data

Outcomes Data

Description:

Demographic and Health Survey results of malaria and fever in children under 5.

Data Already Collected?

Yes

Data Previously Used?

Yes

Data Access:

Not restricted - access with no requirements or minimal requirements (e.g. web registration)

Data Obtained by the Study Researchers?

Yes

Data Approval Process:

Approval Status:

Treatment Assignment Data

Participation or Assignment Information:

No

Description:

Hansen et al, Science, 2013. Forest threshold=25%. Gridded to ~5.5 km x 5.5 km cells by Jens Engelmann as described in Busch and Engelmann 2014.

Data Obtained by the Study Researchers?

Yes

Data Previously Used?

Yes

Data Access:

Not restricted - access with no requirements or minimal requirements (e.g. web registration)

Data Obtained by the Study Researchers?

Yes

Data Approval Process:

Approval Status:

Data Analysis

Data Analysis Status:

No

Study Materials

Upload Study Materials:

Registration Category

Registration Category:

Prospective, Category 3: Data for measuring impacts have been obtained/collected by the research team but analysis for this evaluation has not started

Completion

Completion Overview

Intervention Completion Date:

12/31/2014

Data Collection Completion Date:

12/31/2014

Unit of Analysis:

Children under the age of 5 living in a rural area

Clusters in Final Sample:

Clustered at the unit of grid-cell. Malaria: 4,294-4,378. Fever: 22,719.

Total Observations in Final Sample:

Malaria: 56,883-60,305. Fever: 469,533

Size of Treatment, Control, or Comparison Subsamples:

The main independent variables of interest, deforestation and forest cover, were continuous variables rather than binary treatment/control

Findings

Preliminary Report:

Yes

Preliminary Report URL:

https://www.cgdev.org/publication/does-deforestation-increase-malaria-prevalence-evidence-satellite-data-and-health

Summary of Findings:

We did not find that deforestation increases malaria prevalence nor that intermediate levels of forest cover have higher malaria prevalence. Our findings differ from most previous empirical studies, which found that deforestation is associated with greater malaria prevalence in other contexts. We speculate that this difference may be because deforestation in Africa is largely driven by the slow expansion of subsistence or smallholder agriculture for domestic use by long-time residents in stable socio-economic settings rather than by rapid clearing for market-driven agricultural exports by new frontier migrants as in Latin America and Asia. Our results imply that at least in Africa anti-malarial efforts should focus on other proven interventions such as bed nets, spraying, and housing improvements. Forest conservation efforts should focus on securing other benefits of forests, including carbon storage, biodiversity habitat, clean water provision, and other goods and services.

Paper:

No

Paper Summary:

Paper Citation:

Data Availability

Data Availability (Primary Data):

No--Data not expected to be available

Date of Data Availability:

Data URL or Contact:

Access procedure:

Other Materials

Survey:

No

Survey Instrument Links or Contact:

Program Files:

No

Program Files Links or Contact:

External Link:

External Link Description:

Description of Changes:

Study Stopped

Date:

Reason: