A child can become taller during treatment while gaining less height than a population growth reference predicts. That is different from losing height. A historical lisdexamfetamine safety review illustrates why actual growth, expected growth and the comparison group must be specified before describing a stimulant-associated growth effect.
The source was a review, not a new growth trial
Coghill and colleagues searched English-language literature through September 16, 2013. Their 2014 systematic review identified 35 references containing lisdexamfetamine safety or tolerability data in people with ADHD, from 129 references initially identified. The review combined discussion of controlled trials, longer open-label studies and postmarketing observations; these designs do not provide equivalent causal evidence.
The clinical program commonly studied oral lisdexamfetamine doses of 30, 50 or 70 mg/day. Those are historical research regimens, not dosing instructions. Lisdexamfetamine and the d-amphetamine generated from it should also retain their separate names; a milligram amount of the prodrug is not the same amount of d-amphetamine.
The 281-child comparison used external norms
The growth analysis summarized in the review combined data from two short-term studies and a 52-week extension that enrolled participants from those studies. It involved 281 children aged 6–13 years, assessed for up to 15 months. Mean treatment duration was 265 days, with a standard deviation of 149 days. “Up to 15 months” does not mean that every child completed 15 months.
The investigators compared weight, height and body mass index with US Centers for Disease Control growth norms. At baseline, the children were taller and heavier than the reference expectations. This was therefore not a randomized comparison of 281 treated children against an otherwise identical untreated cohort followed for the same interval.
Preserve the direction and units of change
Across the summarized studies, mean weight decreased by 0.2 kg, compared with an expected increase of 3.5 kg. Mean height increased by 3.9 cm, compared with an expected increase of 4.8 cm. The weight finding included actual average loss; the height finding concerned a smaller gain than expected. Neither describes every individual child.
Among children with endpoint measurements at or beyond 12 months, the proportion with BMI at or below the fifth percentile increased from 4% to 15%. That is a specified follow-up subset. Applying those percentages automatically to all 281 children would discard the denominator qualification.
What might complicate attribution?
The review reported stronger growth effects among initially taller and heavier children, those without previous stimulant treatment and those with greater cumulative exposure. These observations require care because baseline size, treatment history, exposure duration and remaining in follow-up can be related. An external norm comparison cannot independently resolve all of them.
Longer extensions also tend to retain people who respond to and tolerate treatment. The review identified this selection issue, exclusion of major comorbidities and extremes of weight, limited duration and manufacturer sponsorship of the clinical trials as limitations. Those features constrain the generality of reassuring findings as well as adverse ones.
Can these results determine final adult height?
No. The summarized interval did not measure final adult height or establish permanent loss of growth potential. Nor does it establish that an observed slowing is harmless or will necessarily reverse. Those are longer-term questions requiring appropriate follow-up and comparison data.
For an individual record review, the useful evidence includes serial dated measurements, age-appropriate growth references, baseline trajectory, treatment exposure and other relevant health information. A population average is context for assessment, not a diagnosis of the cause or permanence of one child’s growth change.
Source
Coghill DR, Caballero B, Sorooshian S, Civil R. A systematic review of the safety of lisdexamfetamine dimesylate. CNS Drugs. 2014;28:497–511. doi:10.1007/s40263-014-0166-2. Historical review; the 281-child results above are reported through that review, rather than presented as a newly examined primary dataset.
Leave a Reply