From: "TKT (Thomas Klitgaard Tygesen)" <>
Subject: [NMusers] Weight normailized doses in modeling
Date: Fri, 26 Oct 2001 22:19:24 +0200

Dear nmusers:

Could some of you enlighten me with some considerations on dose
normalizations (by weight) in PK/PD modeling? I'm in the process of setting
up a dose dependant one-compartment model on insulin and have all my dose in
weigh normalized units. Moreover, I have a number of covariates which I'd
like to investigate, as to their influence on model parameters (by looking
at the eta's and the individual parameters estimates, plotted against these

However, it may be that my model should be expressed in absolute doses (e.g.
by multiplying a dose of, say, 0.3 U/kg by each subjects weight) before
running the estimation. Otherwise, my final parameters will in some way be
expressed in per kg units.

Does the per kg estimation present any particular problems (e.g. by masking
covariate effect) or is more a matter of practice? Thanks in advance.

Thomas K Tygesen
Civil engineer
Direct +45 44424960
Fax +45 44421940
Clinical Pharmacology - Pharmacometrics
Novo Nordisk A/S, Krogshoejvej 53A
2880 Bagsvaerd, Denmark





From: Nick Holford <>
Subject: Re: [NMusers] Weight normailized doses in modeling
Date: Mon, 29 Oct 2001 19:23:37 +1300


The key issue here is what is "normal" about weight normalised units? Tradition (but not necessarily reason) has decided that doses/kg are "normal". But the relationship between key pharmacokinetic parameters such as clearance and half-life is not linearly related to weight. My advice is to use the actual dose and scale the model parameters using allometric science rather than flat earth tradition.

See these references for the theory and application:
Anderson BJ, Holford NHG. Aspects of theophylline clearance in children. Anaesth Intens Care 1997;25:497-501.

Anderson BJ, McKee D, Holford NHG. Size, myths and the clinical pharmacokinetics of analgesia in paediatric patients. Clinical Pharmacokinetics 1997;33:313-327.

Anderson BJ, Woolard G, Holford NHG. A model for size and age changes in the pharmacokinetics of paracetamol in neonates, infants and children. Br J Clin Pharmacol 2000;50:125-134.

West GB, Brown JH, Enquist BJ. A general model for the origin of allometric scaling laws in biology. Science 1997;276:122-26.

West GB, Brown JH, Enquist BJ. The fourth dimension of life: fractal geometry and allometric scaling of organisms. Science 1999;284(5420):1677-9.

Nick Holford, Divn Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand tel:+64(9)373-7599x6730 fax:373-7556






Subject: Re: [NMusers] Weight normailized doses in modeling
Date: Mon, 29 Oct 2001 11:28:54 MET

Hello Thomas,

Your model (1 comp PK) has no way of "knowing" the units of the object you are
giving it as dose=AMT, so the parameters will not be scaled per kg. On the
contrary, the magnitude of your AMT entry in nonmem may be incorrect with
respect to your concentrations (which I assume are not scaled per kg) so the
parameters may simply come out wrong.

You could obtain scaled per kg parameters by setting e.g. TVCLW=THETA(1)*WT
etc. within the model (and then maybe having to account for this in the S1
scales depending on what the units of your dose finally are and whether you
also scaled V).

In terms of how you actually introduce the AMT item, if you now have it in
mg/kg for example then use S1=V/WT to scale for WT, if your concentrations are
not already scaled. In any case, obviously, the units of the your modeled and
observed concentrations must match.

Apart from the references that Nick suggests you may also want to look at the
theophylline problem in the nonmem manuals.

I hope this helps!


John Lu
Dept of Pharmacy
University of Salamanca
+ 34 647545633