Hi Iratxe, To my knowledge, following up malnutrition treatment in the field using any body composition technique remains a research question. The most widely use and better understood technique in the field remains anthropometry collecting data on weight, height, MUAC and skinfolds thickness (see for instance http://www.ncbi.nlm.nih.gov/pubmed/18977785). There is emerging evidence using other techniques like bioelectrical impedance and densitometry in children in Cambodia, Ethiopia and Malawi. However, to my knowledge this data is still being collected/analysed. BW, Carlos

I think Carlos is correct to consider this a research issue. The last decade has seen therapeutic feeding services move towards an outpatient model with screening / case-finding done in the community by community-based volunteers. Agencies such as ALIMA has been using mothers to screen their own children (probably the ultimate decentralisation). There have also been successful experiments will delivering services by community health workers. All of this has led to the adoption of simpler and cheaper screening tools and case-definitions (i.e. MUAC rather than W/H). Work is underway on using MUAC for treatment monitoring and this has used, in research, measures of body composition during treatment. I do not foresee a move from simple measures such as uncorrected MUAC towards measures of body composition taking place unless it can be done simply and cheaply (i.e. as simply and as cheaply as MUAC) so as to allow high spatio-temporal coverage of case-finding and treatment activities.

Mark wrote: "I do not foresee a move from simple measures such as uncorrected MUAC towards measures of body composition taking place unless it can be done simply and cheaply (i.e. as simply and as cheaply as MUAC) so as to allow high spatio-temporal coverage of case-finding and treatment activities." Well, some of us are working towards developing this option, although it might take us a bit of time and effort. Stay tuned...

I won't be holding my breath!

Hi Mark and Carlos Having, many moons ago, worked in clinical dietetics in the UK and Ireland, many 'body composition' techniques were available to me in a clinical setting (surgery and paediatrics) in hospital nutrition departments. Undernutrition is significant in hospitalised patients for a variety of reasons, yet these tools still often proved impractical to use on a routine basis. Skinfold callipers are not high tech but remained unfeasible in busy wards, staff training minimal, and just no time. It remained difficult enough to get an admission weight on an adult patient that looked visibly thin. Having also worked in much more under resourced settings, I would have to cast doubt on the feasibility of these methods for assessment in the contexts you are discussing, beyond operational research and maybe a few choice locations. Always happy to be proved wrong!

Just to emphasise the need for rapidity, low cost, and simplicity (i.e. practicability) ... I am currently in discussion with a large UK-based social care provider (community-based, sheltered residential units, and inpatient units for older people, disabled people, and persons with learning disabilities) on using MUAC as a screening tool. They have recognised (as some in our community have) that, to be effective, screening for undernutrition must be universal (i.e. cover all of the at-risk population) and timely (i.e. as frequently as possible). They could not achieve this with weight, BMI (a W/H measure), or measures of body composition. If we cannot do this in the UK then I doubt we will be able to do this in programatic contexts in (e.g.) South Sudan. I'd also love to be proved wrong but it seems like very long odds to me. Good luck!