RESUMEN
OBJECTIVE: To assess the within-subject and between-subject coefficients of variation (CV) of iron, zinc and copper concentrations in the milk of Guatemalan mothers. METHODS: We performed a cross-sectional study in lactating women who had delivered a healthy infant 1 to 6 months previously in two low-income peri-urban areas (San Bartolome Milpas Altas and Ciudad Peronia) and a low-income rural area (San Juan Chamelco) in Guatemala. Women infested with Ascaris lumbricoides or Trichuris trichiura received a single dose of albendazole (400 mg) or placebo. Two weeks after treatment, milk samples were collected on 3 or 4 consecutive days. Trace element concentrations in milk were measured by inductively coupled plasma/atomic emission spectrometry. RESULTS: The instrumental error of the inductively coupled plasma/atomic emission spectrometry method, expressed as SD, was 0.04, 0.27 and 0.02 mg/L for iron, zinc and copper, respectively. Concentrations in milk samples collected from 47 mothers on 3 or 4 consecutive days, expressed as mean +/- SD, were 0.28 +/- 0.13, 2.03 +/- 0.37 and 0.29 +/- 0.07 mg/L for iron, zinc and copper, respectively. The within-subject CV was 46.1%, 18.2%, and 22.8% and the between-subject CV was 61.2%, 48.3% and 31.7% for iron, zinc and copper, respectively. Stage of lactation, infestation with intestinal parasites and residential area had a significant influence on milk zinc, copper and iron concentrations. CONCLUSIONS: One sample of milk is sufficient to give a reliable estimate of the zinc concentration in milk. Two samples taken on consecutive days are required for a reliable estimate of iron and copper concentrations.