A group median can conceal a clinically relevant individual result. A small dexamphetamine lactation study, published in 2007, illustrates why a relative infant dose is an exposure estimate rather than a stand-alone safety conclusion. It also distinguishes drug transfer into milk from drug measured in an infant’s circulation.
What was measured in four mother–infant pairs?
Ilett and colleagues studied four women treated orally for ADHD, taking 15–45 mg/day in divided doses. Infant ages were 3.3–10 months. Milk was sampled across approximately 24 hours, with maternal plasma sampling varying between pairs. The calculated relative infant doses were 3.9%, 7.3%, 13.8% and 4.0%; the median was 5.7%. These percentages compare weight-adjusted infant and maternal doses. Ilett et al.
Calculations assumed milk intake of 0.15 L/kg/day. Of three infants with plasma samples, one result was below 1 µg/L and two were 2 and 18 µg/L. No overt adverse effects were identified, but only two infants received specialist developmental examinations. The authors stated that medium- and long-term consequences were unknown. Methods, Table 1 and discussion.
Why does the individual value matter?
The statement “the median was below 10%” is compatible with a value above 10% in one pair. Rewriting it as “all infant doses were below 10%” changes the finding. Reporting the four values makes the uncertainty visible without presenting the highest value as typical.
A percentage also needs its denominator. Relative infant dose compares doses per unit body weight. It is not the percentage of maternal blood concentration found in an infant, the percentage of a tablet swallowed by the infant or a probability of an adverse outcome. Those quantities require different measurements.
What does milk-to-plasma transfer tell us?
A concentration ratio describes distribution between specimens at the measured times or over a specified sampling interval. Infant dose additionally depends on the concentration of milk actually consumed and the amount consumed. Infant plasma then reflects absorption and elimination as well as intake. Collapsing these steps into one “transfer percentage” removes information needed for interpretation.
This distinction is especially useful when records contain a maternal prescription and a milk measurement but no infant sample. That evidence can support an exposure calculation under stated assumptions. It cannot supply an observed infant blood concentration or establish an outcome that was never measured.
What does a reassuring short observation exclude?
Four pairs cannot characterize uncommon events or developmental outcomes emerging over years. Conversely, the presence of measurable infant plasma drug does not itself prove that injury occurred. An exposure finding and a clinical outcome should retain separate descriptions even when both are relevant to the same assessment.
This historical study concerns known prescription dexamphetamine dosing and older infants within its small sample. It does not establish a present-day breastfeeding protocol, a universal safe threshold or a rule for nonprescribed methamphetamine exposure. Its most useful contribution is transparent measurement: individual estimates, specimen identity and the assumptions behind the calculation.
Source
Ilett KF, Hackett LP, Kristensen JH, Kohan R. Transfer of dexamphetamine into breast milk during treatment for attention deficit hyperactivity disorder. British Journal of Clinical Pharmacology. 2007;63:371–375. doi:10.1111/j.1365-2125.2006.02767.x. Observational human pharmacokinetic study.