Antibody responses to serotype 14 of the vaccine however were higher amongst infants who were smaller at 12 months 5-FU order of age and showed slower growth between 3 and 12 months of age. In addition, infants born during July to December (the ‘hungry’ season) had higher antibody titres to serotype 14. The data from this study offer only limited support an early-life programming effect of early nutrition on antibody response to vaccination in adulthood within this environment. The observed associations between early life exposures and response to serotype 14 of the pneumococcal vaccine only

are rather difficult to explain. Globally, serotype 14 is the most important serotype causing disease worldwide, inhibitors although carriage rates vary between populations [12], [22] and [23]. Of the 4 serotypes assessed in the current study (1, 5, 14 and 23f), exposure to 23F

and 14 are most likely similar and so early exposures during infancy are unlikely to explain check details the difference. Technically, type 14 is the ‘purest’ serotype to assay, with little cross-reaction with other serotypes when measured in ELISA (D Goldblatt, personnel communication), but it is unlikely that this alone explains the observed differences. Selection of serotypes was primarily based on carriage rates amongst infants in The Gambia. However, and since it is known that pneumococcal carriage is not equally distributed between adults and children in this population, and is also variable by age (for infants) and season [24], knowledge of precise carriage rates (through nasopharyngeal swabs) at the time of vaccination may have been informative. Inclusion of additional serotypes, such as those known to elicit a ‘weak’ response may help explain this observation. Indeed, previous research has identified serotype 6B as being sensitive to modulation by infant feeding status[25], following vaccination with a conjugated vaccine. Such serotypes TCL may, therefore, be more sensitive to nutritional exposures

early in life. In interpreting the results presented here, consideration should be given to the limitations of the current study. Much of the programming literature in based on the follow up of cohorts of adults for whom records from early-life are available. In The Gambia, the UK Medical Research Council (MRC) has been maintaining demographic and health-related records for three rural villages since 1949 [26]. From 1976, these records have included detailed information on maternal and infant health, allowing the study of early-life predictors of current health within this population. However, as with many studies within this field [27], the current study suffered with considerable loss to follow up. A total of 78 (9%) of the 858 subjects born during the study period were known to have died prior to the start of fieldwork. In addition, we were only able to recruit 41% of the remaining 781 subjects available for follow up.