Title: Statistical and epidemiological methods for assessment of correlates of antibody mediated immunity to P. falciparum malaria in children

PhD Student: Samuel Bosomprah

Supervisors: Dr. Paul Milligan (Reader in Epidemiology and Medical Statistics, Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, United Kingdom) and Dr. Daniel Dodoo (Senior Scientist, Noguchi Memorial Institute for Medical Research, College of Health Sciences, University of Ghana, Ghana)

Collaborators: Dr. Michael Theisen (Statens Serum Institut, Copenhagen, Denmark)

Background: Immuno-epidemiological cohort studies have been used to study the role of malaria antigen-specific antibody in protecting from clinical malaria. Antibody levels are measured at baseline before a period of malaria surveillance, to investigate the association of particular pre-existing antibody responses with clinical immunity. This approach presents several statistical challenges. Estimates of association may be affected by the impact of the relatively short half life of naturally induced malaria antibodies; exposure is an important confounder but it is not clear what the best measures of exposure are or to what extent inadequate adjustment for confounding effects of exposure is responsible for the lack of consistency in published results; statistical analysis of antibody concentrations is complicated by the presence of zero (below detection threshold) values, while analysis of the presence/absence of antibodies is also problematic because the zero group are contaminated by an unknown number of false negatives owing to the limited sensitivity of measurement. Cohorts kept under active surveillance for malaria receive antimalarial treatments but the effects of these treatments are rarely taken into account in analyses. These methodological challenges may partly explain the difficulty in establishing clear independent relationships between antigen specific antibody responses and protection from clinical malaria and may have contributed to the lack of consistency in results from immunoepidemiological studies. This project aims at developing and applying appropriate statistical and epidemiological methods for the analysis of such studies. The work will include meta-analysis of cohort studies completed under the multi-site Afro-Immuno-Assay (AIA) project. It is envisaged the successful candidate will also contribute to statistical work for malaria vaccine trials coordinated by AMANET.