Table 3 Unadjusted and adjusted cross-sectional associations between exposure to screens and child language development at each time point in the EDEN cohort (N = 1,562).a

From: Exposure to screens and children’s language development in the EDEN mother–child cohort

  Cross-sectional models at age 2 years
(MacArthur-Bates CDI as outcome)
Cross-sectional models at age 3 years
(Composite language as outcome)
Cross-sectional models at age 5–6 years
(Verbal IQ as outcome)
Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted
Daily screen time
0 min 0.0 (Reference) 0.0 (Reference)
1–30 min 6.8 (1.5, 12.2) 6.2 (1.2, 11.2) 0.0 (Reference) 0.0 (Reference) 0.0 (Reference) 0.0 (Reference)
31–60 min 7.6 (1.8, 13.3) 8.7 (3.4, 13.9) 0.6 (− 1.6, 2.9) 1.5 (− 0.5, 3.5) 0.2 (− 2.8, 3.2) 1.0 (− 1.8, 3.8)
61–120 min 6.8 (0.6, 13.1) 6.4 (0.7, 12.0) − 0.6 (− 3.1, 2.0) 0.8 (− 1.5, 3.2) − 0.0 (− 2.5, 2.5) 2.0 (− 0.6, 4.6)
 > 120 min 1.7 (− 6.3, 9.8) 3.2 (− 4.8, 11.2) − 2.6 (− 6.5, 1.3) 0.1 (− 3.6, 3.7) − 1.6 (− 4.8, 1.6) 0.8 (− 2.4, 4.1)
TV on during family meals
Never 0.0 (Reference) 0.0 (Reference) 0.0 (Reference) 0.0 (Reference) 0.0 (Reference) 0.0 (Reference)
Sometimes − 2.4 (− 6.1, 1.4) − 1.4 (− 4.9, 2.2) − 4.8 (− 7.1, − 2.6) − 2.8 (− 4.9, − 0.8) − 5.0 (− 7.3, − 2.8) − 3.3 (− 5.6, − 1.0)
Often − 2.8 (− 7.0, 1.4) − 0.7 (− 4.8, 3.4) − 4.4 (− 6.8, − 2.1) − 2.9 (− 5.2, − 0.6) − 8.3 (− 11, − 6.0) − 5.0 (− 7.4, − 2.7)
Always − 13.7 (− 18.6, − 8.7) − 6.7 (− 11.8, − 1.6) − 9.8 (− 13, − 7.0) − 4.6 (− 7.4, − 1.8) − 6.9 (− 10, − 3.8) − 2.5 (− 5.7, 0.7)
  1. CDI Communicative Development Inventory, IQ intelligence quotient, HOME Home Observation Measurement of the Environment.
  2. aValues are adjusted mean differences (vs the reference group) from linear regression models conducted on multiply imputed datasets (n = 1,562). Models are of cross-sectional design, i.e., outcomes are being predicted by exposure to screens as measured at concomitant age. Exposure to screens variables (Daily screen time and TV on during family meals) were mutually adjusted for each other. Adjusted models were further adjusted for the following covariates: study center, maternal age at delivery, pre-pregnancy body mass index, tobacco and alcohol consumption during pregnancy, symptoms of postpartum depression, child sex, gestational age at birth, birthweight, breastfeeding duration, number of older siblings, parental education level, bilingual household, maternal and paternal language difficulties during childhood, household income, main caretaker, cognitive stimulating activities, HOME score. Models at ages 2 and 3 years were further adjusted for the child’s exact age at language assessment (verbal IQ scoring accounts for age). Models at age 3 years were further adjusted for schooling duration.