Potential magnitude of the misclassification of a population's trace element status due to infection: example from a survey of young Peruvian children.

To examine the effects of concurrent infection on population-based assessment of trace element status, we collected data on clinical signs and laboratory indicators of infection when obtaining blood for serum zinc, copper, and ferritin analyses in 153 Peruvian children aged 11-19 mo. Fifty-two (34.7%) of the children had some reported sign of infection and 43 (28.3%) had elevated C-reactive protein concentrations or leukocytosis. Children with any evidence of infection had marginally lower mean (+/- SD) serum zinc concentrations (7.0 +/- 2.3 vs 7.5 +/- 2.0 mumol/L, P = 0.16) and significantly greater serum copper (24.7 +/- 4.7 vs 22.7 +/- 4.2 mumol/L, P = 0.006) and serum ferritin concentrations (10.0 +/- 12.9 vs 3.9 +/- 4.4 micrograms/L, P < 0.001) than did those without infections. Infection caused an underestimation in the rate of low copper status by 1 percentage point and low iron status by 12 percentage points. Thus, the effect of concurrent infections is of variable magnitude and may differ by nutrient, nutritional status of the population, and prevalence and severity of infections.