Because of staffing reductions, we reduced the numbers of investigations (nutritional assessment, exercise tolerance testing and Neuroscope testing) which we had planned in the original version of the Protocol; in addition, we decided that we were no longer able to provide Cosmofer in the necessary time-frame for the patients who proved to be intolerant of oral ferrous sulphate.

1. We requested an increase in the number of patients to be approached for recruitment from 90 to 140, because a higher than expected number of patients had turned out not to meet the inclusion criteria at the time of planned recruitment because their post-consent blood test showed them either to be no longer anaemic by WHO standards, or no longer iron deficient (Fe saturation <18%) (ie, during the time between their previous out-patient appointment and the intended recruitment date, their haemoglobin and iron status had improved). 2. Because of Pharmacy staffing shortages, to save time and inconvenience for patients attending afternoon clinics at the Royal London or any clinics at St Bartholomew’s Hospital, we arranged for prescriptions for ferrous sulphate to be collected from the Pharmacy at the Royal London by one of the investigators before the potential participants attended the clinic to give their written consent, giving them their tablets if/when this was been obtained.

Change of name of sponsor from Barts and the London NHS Trust to Barts Health NHS Trust

Because the initial suppliers, Wockhardt Ltd, could no longer make or supply ferrous sulphate tablets, we had to obtain them from an alternative supplier, Teva UK Ltd.

1. Because of staff changes, we altered the arrangements for obtaining and countersigning patients' consent. Consent was taken by suitably trained (including GCP training) medically qualified and delegated staff at each site. At each site the consent form was countersigned by the site PI. 2. We changed our protocol to record that we were storing stool and serum samples at -40 rather than -80 degrees C. 3. As recruitment was going more slowly than initially hoped, we requested an extension of the duration of the study beyond 31 Dec 2013.

1. We minimally changed the inclusion criteria so that iron deficiency was defined now by a transferrin saturation of <18% rather than <16%. We had noticed at a monitoring visit at Chelsea and Westminster Hospital that they were using a version of the protocol showing 16% rather than 18%. 2. The faecal calprotectin assays were to be done in the routine Barts Health NHS Trust immunology laboratory rather than at Kings College Hospital (as originally planned) because the assay had now been established here as a routine clinical test. 3. The hepcidin assays were now to be undertaken in the Gastroenterology Laboratory at the University of Birmingham, under the supervision of Dr Tariq Iqbal, rather than here, because the Birmingham assay was well-established and probably the most reliable in the UK. 4. The letter to patients' GPs about their recruitment was very minimally modified.