Guidelines WRT weight velocities with the CTC protocol can be found in the CTC manual. They are not usually s high as 15 g/kg/day and typically closer to 5 g/kg/day. Note that these are average velocities. Weight will usually follow a typical growth curve during nutritional rehabilitation (i.e. fast weight gain at the start of the treatment episode slowing over time). Data (presented by me at the 2008 CMAM conference) shows median growth of between 0.25 mm (MoH IMAM without antimicrobials) and 0.40 mm per day (NGO CMAM with full CTC protocol). Again, these are averages and MUAC follows a growth curve. Highest values in my database are a little over 0.8 mm per day over a treatment episode. I would not expect 15 mm / week. I hope this is of use.

Thank you very much, I was more concerned about the physical maximum values (ie like maximum/minimum Z scores of-6/+6 for WFA, WFH and HFA), If one has to put limit (in terms of physiologically possible) to GoW, would you recommend something like -15/+15 g/kg/day? Likewise, can we consider that MUAC value over a week cannot increase/decrease more than 80mm? Thank you again for your support, Best regards

This is not really my area of expertise. We expect different velocities at different points in treatment. Over an entire treatment episode I would not expect to see weight velocities in excess of 15 g / kg / day in a CMAM program. A MUAC gain of 80 mm / week is 1.14 cm per day. The highest that I have seen is about 0.6 cm per week (i.e. a little over 0.8 mm per day). Limits to losses are hard to estimate because we usually find ourselves close to the end-point of wasting (i.e. SAM). You may find useful data in the starvation experiments and in data relating to (e.g.) anorexia nervosa.

From Mike Golden: The question of the maximum expected rates of MUAC and weight gain is interesting. We can look at it theoretically and see if this is backed up by data from children fed under supervised conditions – i.e. from TFC data. Anorexia nervosa data are not useful in this respect as “intensive feeding” is not part of the recommended treatment of these patients. Incidentally there are very interesting differences between Anorexia nervosa and the malnourished adult patients seen in African crises – the difference probably being due to the quality of the food that is taken. Parenteral feeding data is determined to a large extent by the amount that is “pushed” in (can be very dangerous) and is not applicable. Many experiments have shown that ingestion of 5kcal in excess of maintenance amounts leads to a weight gain of 1 gram. This is when there is a mixture of lean and adipose tissue being synthesised (and a famous paper by Alan Jackson related this efficiency of tissue synthesis to changes in new muscle synthesis). However, if the child is making only lean tissue there should be a gain of 1 gram for each 2kcal (approx) above maintenance and if the child is making only adipose tissue then it is about 8 kcal/g. (note lean tissue does contain some fat and adipose tissue some water). These figures of the efficiency of “food conversion” into tissue can go substantially higher if there is increased malabsorption or the diet is not nutritionally balanced to promote tissue synthesis (like most staples/family plates). But it is unlikely that the food can be used more efficiently. Thus if we give, and the child takes 200kcal/kg/d of a diet designed to promote new tissue growth, and the energy needed for maintenance is 100kcal/kg then there are 100kcal/kg left to make new tissue and we expect a rate of weight gain of 100/5 = 20g/kg/d – on average. Of course there is spillage and not all the diet is taken by many children so the rates of weight gain seen in supervised feeding situations is generally lower. However, if the child is making only lean tissue the maximum theoretical rate of weight gain would be 50g/kg/d. We do sometimes see this at the beginning of recovery, but it is uncommon, and it slows later as more fat is being replenished so that over the whole of recovery this figure is not reached. Young babies make more hydrated tissue than older children or adults so their rates of weight gain will be higher for the same dietary intake. So what do we see in practice? In the following table I have analysed the data from children who reached the discharge criteria. I removed any values (very few) with rates of weight gain of over 40g/kg/d as more likely due to measurement/recording errors and then looked at the mean, SD, median and the 2.5 and 97.5 centiles of the rates of weight gain. It is the 97.5 centile that you are interested in. Table 1 Rates of weight gain from minimum weight to weight 28 days later. In grams/kg minimum weight/d Age n mean sd 2.5c* 97.5c* median 0-6 459 18.6 8.5 6.1 39.9 17.1 7-24 3703 14.5 6.4 5.4 31.4 13.3 25-60 2650 14.8 6.2 5.5 30.2 13.8 61-144 917 14.6 6.8 5.2 32.1 13.1 145-216 136 11.8 6.6 2.8 31.5 9.9 216+ 372 10.6 6.2 3.6 28.6 8.8 * c=centile Some of the children had discharge MUAC values as well as admission values – not as many as I would like! So I have examined the rate of MUAC increase and rate of weight gain in this group Rate of MUAC gain mm/d Age n mean sd 2.5c* 97.5c* 0-6 26 0.51 0.31 0.13 1.38 7-24 1003 0.50 0.28 0.05 1.11 25-60 914 0.56 0.33 0.03 1.27 61-144 268 0.51 0.32 0.03 1.31 145-216 51 0.44 0.24 0.00 0.98 216+ 152 0.46 0.30 0.00 1.18 Rate of weight gain g/kg/d Age n mean sd 2.5c* 97.5c* 0-6 26 19.3 8.8 6.5 37.2 7-24 1003 14.1 7.6 5.0 33.6 25-60 914 14.3 5.6 6.2 28.4 61-144 268 14.0 6.1 6.3 30.4 145-216 51 15.4 9.8 6.0 38.9 216+ 152 10.2 6.5 3.2 29.2 It would appear that the values obtained from detailed metabolic experiments and the theoretical calculations agree fairly well with these results. The highest mean rates of weight gain were in Mali with an average of just over 20g/kg/d and the maximum very high (data were censored) – this appeared to be because the children took the F100 avidly because they were thirsty (in the desert) rather than hungry – if was only later when we realised this and the rates of weight gain fell again - so the children were taking more than 200kcal/kg.d in these extreme conditions and we were failing to supply sufficient water. The very low (relative) rates of weight gain in OTP is a reflection of the extensive sharing (the minimum amount of sharing can be calculated from the parameters given above) and the fact that family food, which is invariably taken, is not properly balanced nutritionally for rapid tissue synthesis. The data from OTP do not reflect the maximum rates of weight or MUAC gain that can be achieved.