Re: Growth Curve Modelling
Nyasha,
Thanks for your further explanation of what you want to do. You can easily adapt the NM-TRAN code for the model for WT based on PMA in order to compare your PMA-SEX-WT data you have with the model predictions.
I looked at the 3 papers you mention below -- Beath describes a model for weight for just the first 2 years of life after full term gestation, Cole is dealing with height not weight in a rather narrow age band around puberty, Jolicouer is also dealing with height but from 1 month to 19 years of post-natal age.Jolicouer's approach seems to be the closest to the method we described but does not include very premature and premature infants nor older adults.
You ask about derivation of the model but there really is nothing to derive. This is a top down empirical model that has the form it does because it seems to describe the data. However, our approach is based on successful empirical use of a sigmoid hyperbolic function to describe maturation of glomerular filtration rate and clearances for various drugs. The sum of asymmetrical sigmoid hyperbolic functions plus the switch to an exponential term for the last component of weight was developed by trial and error. One feature of this fully parametric approach which has weight components associated with different age ranges means that we could see that sex differences occur in very young (<1 y) and older children (> 12 y) but not in the intervening ages. Otherwise the model we propose has no direct relationship to other biological quantities. All the other descriptions of weight changes with age describe separate curves for males and females.
Best wishes,
Nick
1. Beath KJ. Infant growth modelling using a shape invariant model with random effects. Stat Med. 2007;26(12):2547-64. 2. Cole TJ, Donaldson MDC, Ben-Shlomo Y. SITAR—a useful instrument for growth curve analysis. Int J Epidemiol. 2010;39(6):1558-66. 3. Jolicoeur P, Pontier J, Pernin M-O, Sempé M. A Lifetime Asymptotic Growth Curve for Human Height. Biometrics. 1988;44(4):995-1003.
Quoted reply history
On 24/02/2011 4:11 a.m., nyashadzaishe mafirakureva wrote:
> Hi Nick and Joe,
>
> Thank you very much for your responses. Well, I must say I'm a real beginner in modelling and this is my first attempt at growth curve modelling. I was of the idea that growth curves have been developed and in use for quite sometime and hence it would be possible to find a function or functions describing growth (height, weight etc) in normal children. Using this function, I wanted to fit data from children with a specific disease and as such try to see any differences with the general population as well as the effect of treatment and other factors on growth. However, when I searched literature I couldn't find something conclusive about this. Basically there are a number of functions described for example the Preece & Baines model, the Gompertz & it's modifications, the Shape Invariant models. Some authors made use of polynomials to describe growth, although they are generally not favoured. My main challenge at the moment is undestanding and relating most of the constants/parameters in these models to biology. Hence it's difficult to pick one for my analysis. I have also gone through Nick's paper, but I am still trying to understand how you derived your equations and I would appreciate if you can help me understand that.
>
> Some of the papers I read are here;
> 1. Statist. Med. 2007; 26:2547–2564
> 2. Int. J. Epidemiol. (2010) 39(6): 1558-1566
> 3. Biometrics. 1988 Dec;44(4):995-1003.
>
> Thank you.
>
> Regards,
>
> Nyasha
>
> On Thu, Feb 17, 2011 at 8:26 PM, Nick Holford < [email protected] < mailto: [email protected] >> wrote:
>
> Hi,
>
> Thank you Joe for noticing the recent publication of a
> quantitative model relating age and sex to weight across the human
> life span (Sumpter & Holford 2011).
>
> The WHO growth charts are just that -- charts i.e. a picture with
> some lines on it. The model that Anita and I have proposed is
> quantitative and includes both fixed (post-menstrual age, sex) and
> random effects to describe the increase in weight in a sample of
> similar size to that used for the WHO charts. Although the model
> is quite empirical we noticed that sex is associated with weight
> differences only in the very young (<1 y) and after about 12 y of
> age. Children in between these ages have similar weight for age
> irrespective of sex.
>
> We did not find any quantitative models published before but if
> Nyashadzaishe did find some please let us know. We would be happy
> to send the NM-TRAN code for our model and its parameters if
> anyone wants to use it.
>
> The population we used is perhaps more appropriate for simulations
> of weight in clinical trials because we used subjects who had been
> participants in clinical trials (mainly PK studies). The WHO
> population was selected to represent the ideal pattern of growth
> with what was thought to be optimal nutrition.
>
> If anyone would like to contribute data (age, weight, sex are all
> that is required for each subject) then we would be happy to
> extend our analysis. Our data tended to be concentrated in the
> under 1 year of age group so more observations in older children
> and adults would be very useful.
>
> Best wishes,
>
> Nick
>
> Sumpter AL, Holford NHG. Predicting weight using postmenstrual age
> – neonates to adults. Pediatric Anesthesia. 2011;21(3):309-15.
>
> On 18/02/2011 3:09 a.m., Standing Joseph (GREAT ORMOND STREET
> HOSPITAL FOR CHILDREN NHS TRUST) wrote:
>
> > Nyashadzaishe,
> > UK and I believe WHO growth charts are derived using LMS
> > method which you can do in R - look for papers by TJ Cole. I
> > also suspect you might get a response from a certain NHG Holford
> > telling you to look at his NONMEM method: Paediatr Anaesth. 2011
> > Mar;21(3):309-15 - I have only skimmed this paper but I think the
> > introduction gives a review of possible methods.
> > BW,
> > Joe
> >
> > ------------------------------------------------------------------------
> > *From:* [email protected]
> > <mailto:[email protected]>
> > [[email protected]
> > <mailto:[email protected]>] On Behalf Of nyashadzaishe
> > mafirakureva [[email protected]
> > <mailto:[email protected]>]
> > *Sent:* 17 February 2011 13:24
> > *To:* [email protected] <mailto:[email protected]>
> > *Subject:* [NMusers] Growth Curve Modelling
> >
> > I am trying to model growth in children (1-20 years) with a
> > particular disease condition in NONMEM, R or any other software.
> > There seem to be a lack of concensus in literature on the
> > functions (models) one can use. I am therefore looking for some
> > pointers from anyone who could have worked on something similar
> > before.
> >
> > Thank you
> >
> > -- Nyashadzaishe Mafirakureva
> >
> > ********************************************************************************************************************
> >
> > This message may contain confidential information. If you are not
> > the intended recipient please inform the
> > sender that you have received the message in error before
> > deleting it.
> > Please do not disclose, copy or distribute information in this
> > e-mail or take any action in reliance on its contents:
> > to do so is strictly prohibited and may be unlawful.
> >
> > Thank you for your co-operation.
> >
> > NHSmail is the secure email and directory service available for
> > all NHS staff in England and Scotland
> > NHSmail is approved for exchanging patient data and other
> > sensitive information with NHSmail and GSi recipients
> > NHSmail provides an email address for your career in the NHS and
> > can be accessed anywhere
> > For more information and to find out how you can switch, visit
> > www.connectingforhealth.nhs.uk/nhsmail
> > http://www.connectingforhealth.nhs.uk/nhsmail
> >
> > ********************************************************************************************************************
>
> -- Nick Holford, Professor Clinical Pharmacology
>
> Dept Pharmacology& Clinical Pharmacology
> University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
> tel:+64(9)923-6730 fax:+64(9)373-7090 mobile:+64(21)46 23 53
> email:[email protected] <mailto:[email protected]>
> http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford
>
> --
> Nyashadzaishe Mafirakureva
> +31644418939
--
Nick Holford, Professor Clinical Pharmacology
Dept Pharmacology& Clinical Pharmacology
University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
tel:+64(9)923-6730 fax:+64(9)373-7090 mobile:+64(21)46 23 53
email: [email protected]
http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford