A very difficult question. We know that MUAC < 230 mm is not good for the mother or the foetus (or the baby). Ideally we would want to find and support all PLWs with MUAC < 230 mm. We do not usually work in ideal setting and have to do a cost-benefit analysis. If MUAC < 230 mm in PLWs is rare then we may use a lot of scarce resources to find few cases. I think we need to work out how to do case-finding with high coverage and at low cost. The intervention is not expensive.

To be more direct ... no reference percentage that I know of.

I hope this is of some use.

A general comment about screening PLW with MUAC (or any other anthropometric index).

We should differentiate between indices which indicate a risk and indices which indicate a nutritional intervention is needed.

To make this distinction clear, let us look at maternal height. There are plenty of studies showing that short mothers have more difficult labour with higher maternal and perinatal mortality. This, however, will not be corrected by supplementation which will not make any difference for short women with a small pelvis who need caesarean section. So small maternal height is an indicator of risk but does not tell us that short pregnant women need nutritional supplementation.

Low MUAC has been shown to be associated with poorer pregnancy outcome. See:

https://www.ncbi.nlm.nih.gov/pubmed/23787989

To my knowledge though, the effect of nutritional supplementation in these low MUAC women has never been clearly shown. It is likely to have a favourable effect if selected women are malnourished and increase their MUAC following supplementation but it will have no effect if selected women just have a high risk because they are short, as this will not be corrected with food supplements (which contrasts with the situation in children who respond to treatment even when they are short).

So we should be very careful when making statements about MUAC and maybe we should look at maternal height and response to supplementation in populations where a nutritional intervention is considered.