Postprint version. Published in Early Human Development, Volume 53, Issue 3, January 1, 1999, pages 251-269.
Copyright Elsevier Science Ireland Ltd.
The definitive version is available at http://dx.doi.org/10.1016/S0378-3782(98)00080-2.
Coffee is commonly given daily to toddlers in Guatemala. Possible negative effects of coffee ingestion on cognitive development and sleep patterns were assessed in 132 children 12–24 months of age who had received coffee for >2 months and were iron deficient on at least one indicator. Children were stratified by initial hemoglobin (A=anemic, Hgb <10.5 g/dl; NA=‘non-anemic', Hgb ≥10.5 g/dl) and were randomly assigned to an experimental group (S=substitute consisting of sugar and coloring), and a control group (C=continuation of coffee) (42 C-NA; 53 S-NA; 18 C-A; and 19 S-A). Anemic children were provided Fe supplements for 2–3 months. Compliance was assessed every 2 weeks. After 5 months, testers masked to treatment group and anemia evaluated children with the Bayley Scales of Infant Development II in a central location. Scores were the Mental Development Index (MDI), the Psychomotor Development Index (PDI), and scales from the Behavior Rating Scale (BRS). The child's sleep in the previous 24 h was assessed with a set of standardized sleep questions to the care giver on the first visit and every 2 weeks thereafter. No significant effects of treatment on test scores or BRS ratings were found. In the 24 h period reported on at the final visit, children in the Substitute group slept more during the night and overall (night plus naps) than children in the Coffee group, a difference not found at the first visit. No differences were found in sleep difficulty or number of times waking at night. Women's reported coffee intake per day during pregnancy was associated with lower BRS ratings, even after controlling for SES and child age. The effects of postnatal coffee ingestion in Guatemala were seen for sleep duration, but not for cognitive development. Prenatal coffee ingestion was negatively associated with Behavior Rating Scales and should be investigated further.