Risk factors for suspected developmental delay at age 2 years in a Brazilian birth cohort

Many children are at risk of not achieving their full potential for development. Epidemiological studies have the advantage of being able to identify a number of associated factors potentially amenable to intervention. Our purpose was to identify risk factors for suspected developmental delay (SDD) at age 2 years among all children born in the city of Pelotas, Brazil, in 2004. This study was part of the 2004 Pelotas Birth Cohort. The Battelle Screening Developmental Inventory (BSDI) was administered to cohort children at age 2 years. A hierarchical model of determination for SDD with confounder adjustment was built including maternal sociodemographic, reproductive and gestational characteristics, as well as child and environmental characteristics. Multivariable analysis was carried out using Poisson regression. Prevalence ratios (PR) and 95% confidence intervals [95% CI] were calculated. In the results, 3.3% of the 3869 children studied screened positive for SDD. After confounder control, children more likely to show SDD were: those with positive BSDI at age 12 months (PR = 5.51 [3.59, 8.47]); with 5-min Apgar <7 (PR = 3.52 [1.70, 7.27]); with mothers who had <4 years of schooling (PR = 3.35 [1.98, 5.66]); from social classes D and E (PR = 3.00 [1.45, 6.19]); with a history of gestational diabetes (PR = 2.77 [1.34, 5.75]); born <24 months after the last sibling (PR = 2.46 [1.42, 4.27]); were not told child stories in the preceding week (PR 2.28 [1.43, 3.63]); did not have children's literature at home (PR = 2.08 [1.27, 3.39]); with low birthweight (PR = 1.75 [1.00, 3.07]); were born preterm (PR = 1.74 [1.07, 2.81]); with <6 antenatal care appointments (PR = 1.70 [1.07, 2.68]); with history of hospitalisation (PR = 1.65 [1.09, 2.50]); and of male sex (PR = 1.43 [1.00, 2.04]). These risk factors may constitute potential targets for intervention by public policies and may provide help to paediatricians in preventing developmental delay.