Potential for misclassification of micronutrient status in children participating in a Head Start program

OBJECTIVE: To evaluate relations among measures of iron and zinc status, C-reactive protein (CRP), and leukocytes in low-income children participating in the Head Start program.

DESIGN: Cross-sectional correlational study with samples collected at Head Start centers in May 2003.

SUBJECTS/SETTING: Forty-seven children (aged 3 to 5 years) attending Head Start centers in three rural communities.

MEASURES: Zinc, ferritin, CRP, and complete blood count were analyzed in nonfasting blood samples.

STATISTICAL ANALYSES: Correlations were computed among leukocyte levels, CRP levels, and measures of micronutrient status. Children having two abnormal measures (ie, leukocytes and CRP) were compared by univariate analysis of variance with children having zero or one abnormal measure.

RESULTS: Most (72%) of the children had elevated CRP levels. Four percent were anemic (hemoglobin<11.0 g/dL [<110 g/L]); 11% had low iron stores (serum ferritin<or=15 ng/mL [<or=33.7 pmol/L]); and 77% had low iron stores when a reference value that accounts for the presence of infection was used (serum ferritin<or=30 ng/mL [<or=67.4 pmol/L]). Twenty-one percent had low plasma zinc levels. Children with two measures that indicated infection had higher serum ferritin and lower plasma zinc levels than children with zero or one indicator of infection.

CONCLUSIONS: The link between measures of infection and serum ferritin levels suggests low-income preschool children with low iron stores are not identified by the <or=15 ng/mL (<or=33.7 pmol/L) criterion. The link between zinc and infection suggests possible overestimations of zinc deficiency in low-income children. Thus, in this population, the presence of infection (as indicated by CRP levels and leukocyte counts) should be determined to assess micronutrient status.