A main criterion to identify malaria vaccine candidates is the proof that acquired immunity against them is associated with protection from disease. The age of the studied individuals, heterogeneous malaria exposure, and assumption of the maintenance of a baseline immune response can confound these associations.

Immunoglobulin G/immunoglobulin M (IgG/ IgM) levels were measured by Luminex® in Mozambican children monitored for clinical malaria from birth until 3 years of age, together with functional antibodies. Studied candidates were pre-erythrocytic and erythrocytic antigens, including EBAs/PfRhs, MSPs, DBLs, and novel antigens merely or not previously studied in malaria-exposed populations. Cox regression models were estimated at 9 and 24 months of age, accounting for heterogeneous malaria exposure or limiting follow-up according to the antibody's decay.

Associations of antibody responses with higher clinical malaria risk were avoided when accounting for heterogeneous malaria exposure or when limiting the follow-up time in the analyses. Associations with reduced risk of clinical malaria were found only at 24 months old, but not younger children, for IgG breadth and levels of IgG targeting EBA140III-V, CyRPA, DBL5ε and DBL3x, together with C1q-fixation activity by antibodies targeting MSP119.

Malaria protection correlates were identified, only in children aged 24 months old when accounting for heterogeneous malaria exposure. These results highlight the relevance of considering age and malaria exposure, as well as the importance of not assuming the maintenance of a baseline immune response throughout the follow-up. Results may be misleading if these factors are not considered.