Introduction The 2004 Pelotas (Brazil) Birth Cohort has put strong emphasis on child development and cognition. In this work we explore how early and contemporary determinants relate to cognitive ability measured by WPPSI IQ test at age four.

Methods Early determinants explored were wealth, maternal schooling, smoking during pregnancy, birth size and neonatal morbidity, based on information collected just after birth and at 3, 12 and 24 months of age. Contemporary determinants, assessed at age four, were child stimulation and malnutrition. Interviews were done by specially trained interviewers, the WPPSI test was done by psychologists that underwent a specific training. Analyses were carried out using IQ as a continuous standardised variable through linear regression models adding fractional polynomials when non-linear relationships were detected.

Results IQ mean was 100, with s.d. of 16.5; 3.4% of 3723 children assessed were below 70. Maternal schooling and wealth had independent and linear effects on IQ. Children from mothers 0–4 years of schooling were one full s.d. below children from graduate mothers. Children in the poorest wealth quintile presented an average IQ −0.5 SD compared to the richest. Birth weight presented different effects for children small (SGA) and adequate for gestational age, negligible effect in the latter group, but for SGA the effect was strong and non-linear with low birthweight children up to −1.7 SD below normal children. Children poorly stimulated had 0.5 SD disadvantage.

Conclusion Child IQ has complex biological and social determinants, reducing the disadvantage of vulnerable children requires integrated and coordinated action.