Assessment of the effects of ethnicity, home language, and socioeconomic status on neurodevelopmental performance in a multiethnic population

Objective
To investigate a unique, multicultural population to allow us to examine the association of ethnicity, socioeconomic status (SES) and spoken language at home on performance using developmental assessment tools.

Design
Analysis of a prospective cohort study.

Setting
Hamad Maternity Hospital, Doha, Qatar.

Methods
271 participants included infants born at Hamad Maternity Hospital between September 2016 and September 2017 (Q-Prem cohort). At 24 months (corrected for gestational age), children were assessed with both the language-based Bayley Scales of Infant and Toddler Development III (Bayley-III) and Babyscreen, a non-verbal tablet-based cognitive assessment tool. Bayley-III cognitive, motor, and language composite scores and Babyscreen scores were the main outcome measures.

Results
Median Bayley-III scores in all groups were lower than expected norms (cognitive composite 95 (IQR: 90 to 105), motor 91 (IQR: 85 to 97), language 89 (IQR: 79 to 94)). In contrast, Babyscreen scores were similar to published norms for term infants: mean 13 (IQR: 11 to 15). Children whose main reported language at home was not English had decreased odds for a satisfactory score of >90
for the Bayley-III cognitive and motor composite scores (OR 0.20 (95% CI 0.03 to 0.77) and OR 0.23 (95% CI 0.04 to 0.79) respectively). In contrast, reported language at home had no impact on Babyscreen performance. Higher SES was associated with a mild increase in odds of a satisfactory result on Bayley-III language assessment (OR 1.07 (95% CI 1.01 to 1.13)) but had no effect on Babyscreen. Maternal ethnicity had no significant effect.

Conclusions
Performance on developmental assessment may be affected by the language spoken at home rather than true cognitive ability. SES contributed to a lesser extent and maternal ethnicity had no effect. These results highlight the need for non-language-based assessment tools in multicultural cohorts to better compare true cognitive ability.