Normal healthy blood donors were recruited to establish reference ranges. Healthy individuals were defined according to the donation criteria as set out by the Namibia blood transfusion service. The donation criteria states that individuals should be between the age of 16 and 65 and in good health for them to be considered for blood donation . It also states that they should have the following attributes:
- No recent alcohol intake
- No active bacterial infection
- Weigh more than 50 kg with a hemoglobin level above 125 g/L
- Lead a sexually safe lifestyle
- Enjoy general good health and
- Committed to helping others and can donate for the right reasons .
The initiative by the NaMBTS to recruit voluntary non-remunerated donors (VNRD) makes it possible to use this population to establish reference ranges. These VNRDs are conscious of their contribution to society and they generally have good healthy habits making them a better target than randomly selecting individuals outside of a healthcare establishment. The VNRDs provide blood freely and voluntarily without expecting anything in return . Research by Kalargirou et al.  established that the majority of donors donate blood out of their unselfish concern for the welfare of others. These attributes are good features that contribute to the general wellbeing of an individual. They also demonstrate the appreciation of health for both self and others.
In their work, Sdogou et al.  carried out physical examinations on their participants to help ascertain their health status. In this study, NaMBTS guidelines on donor selection  were used. The procedures include checking for vital signs such as blood pressure, weight as well as asking questions about donors' general health using a standard donor questionnaire. The donor questionnaire elicits for information on an individual's health and medical history of which the certain criteria has to be met before being allowed to donate.
The approach on using donor questionnaire is similar to what Galesloot et al.  research used where they employed a health and lifestyle questionnaire as one of their participant selection tools. These approaches ensure the participants' health status is assessed based on clinical presentation on the day of blood sample collection, as well as relevant medical history. This holistic approach minimises confounding variables such as hidden medical conditions that interfere with the physiological mechanisms involved in hepcidin metabolism.
On that premise individuals who qualified to donate blood were therefore included in the study. Deferral, whether permanent or temporary, was one of the exclusion criteria the basis of which was its perceived deviation from how "normal healthy" adult was defined. Non-consenting participants were also excluded as it is ethically immoral to coerce anyone to participate in research studies. Donors with low hemoglobin levels as well as those with recent bacterial infections were also excluded. Bacterial infections are known to upregulate hepcidin expression and consequently reduce iron bioavailability . Such participants invariably exhibit higher hepcidin levels. Excluding donors with recent infection was, therefore, mandatory for normal references to be a true reflection of the normal healthy population.
After establishing the inclusion criteria, a sample size calculation was performed using a sample size calculation for quantitative variables obtainable from cross sectional studies. The formula for sample size calculation is shown below and has been described in Charan and Biswas .
where Z1-α/2 = standard normal variate
SD = standard deviation. The SD value is obtainable from previous studies by other researchers.
d = absolute error or precision.
The standard normal variate used in this study was 1.96 using 5% type 1 error (p<0.05) and the desired absolute error/precision (d) of 6 ng/mL. The performance characteristics of the CUSABIO ELISA kit were taken into consideration in selecting the absolute error. The kit has a detection range of 4.69–300 ng/mL and a minimum detectable dose of 1.17 ng/mL, the desired absolute error of 6 ng/mL was determined as reasonable.
The standard deviation was obtained by scanning through papers with similar studies such as the work of Sdogou et al. , Pasricha et al.  and Galesloot et al. . An standard deviation of 18.94 ng/ml was used giving a sample size of 38.28. A sample size of 40 was, therefore, used to establish the normal range for hepcidin using the blood donor population in Namibia. A total of 19 female donors and 21 male donors were recruited into this study.
Approval was obtained from the University of Bath through the Research Ethics Approval Committee for Health (REACH) prior to commencement of this study. In Namibia, permission to carry out research work and ethical clearance was also granted by the Namibian Blood Transfusion Service and the Ministry of Health and Social Services in Namibia. Written informed consent was obtained from all blood donor participants before samples were collected.
Venous blood samples were collected from consenting healthy blood donors by the NaMBTS nursing staff. Before blood collection, participants' weight and vital signs such as blood pressure were determined to ensure they were safe to give blood. Normal phlebotomy procedures were followed as described in the WHO guidelines on phlebotomy procedures . The blood samples were collected into vacutainer tubes which self-regulate volumes collected according to the amount of additives contained in the tubes. Samples collected were used to perform full blood count, Hepcidin and serum iron assays.
The participants then proceeded to donate blood after samples for the hepcidin study were collected. This order of events was important in order to avoid variation in hepcidin concentrations caused by changes in blood volumes post blood donation. Hepcidin values respond to decrease in blood volume because of the consequent reduction in iron content .
The samples were also collected from donors presenting at the NaMBTS for donation in the morning in order to minimise the effects of diurnal variation in hepcidin . The timing of collections was also important in optimising sampling conditions and it also became pertinent to standardise preparation of samples for laboratory analysis.
In order to ensure sample validity samples for full blood count were processed on the same day. Samples for full blood count need to be tested as soon as possible because storage, regardless of conditions, affects the different parameters of full blood count results. This is in line with the recommendations and advice from different authors who have done some work on full blood count results and storage conditions  .
Samples for hepcidin and serum iron were centrifuged at 1000 g for 15 minutes according to the kit manufacturer's instructions . The samples were then aliquoted and frozen at –70oC to avoid loss of bioactivity of the hepcidin molecules. All samples were stored for at most two months before testing in line with the guidance from the hepcidin kit manufacturer .
On the days of the experiments samples were thawed and allowed to reach room temperature before being assayed. All the other reagents were also brought to room temperature as per the manufacturer's guidance. Once thawed the samples were spun down at 1000 g for five minutes. All reagents were prepared according to the assay procedure  and an ELISA was set-up.
Samples for the hepcidin assays were also used for assessing serum iron levels. Serum iron results were obtained to confirm the presence of normal iron levels in this group. On the day of analysis samples were thawed and working reagents were prepared immediately prior to use to ensure the validity of the assay was not compromised. The thawed serum samples were spun at 1000 g for five minutes to remove any insoluble particles that may interfere with spectrophotometric characteristics of the coloured complex when reading the absorbance of the final solution. According to the kit insert, samples needed to be separated as soon as possible. This was consistent with the requirements of the hepcidin assay protocol, therefore, the same sample preparation protocol was sufficient to ensure sample validity.
The hepcidin assays were run using an ELISA technique with a detection range of 4.69–300 ng/ml. The inter-assay and intra-assay precision for this assay was <10% and <8% respectively . The intra-assay precision was performed by the kit manufacturer. This was carried out using three samples of known hepcidin concentration. The samples were tested twenty times on one ELISA plate and the results used to assess intra assay precision . The inter-assay precision to determine precision between assays was also performed by CUSABIO® who were the kit manufacturers. On this instance three samples with a known concentration of hepcidin were tested in twenty assays and the results used to assess the inter-assay precision.
There are several methods to quantitate hepcidin which include the ELISA, mass spectometry (MS), high performance liquid chromatography (HPLC), radio immunoassays (RIA), and other novel approaches such as surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS) and liquid chromatography tandem mass spectrometry (LC/MS-MS). In their review, Macdougall et al.  discussed the different methodologies available for Hepcidin testing including the ELISA technique. In their review, a number of studies done on Hepcidin using the ELISA technique were discussed. In Macdougall et al. review , it appears Koliakari et al. used wells coated with polyclonal antibodies whereas in Ganz's work the wells were coated with monoclonal antibodies against hepcidin. They noted that Koliaraki et al. as well as Ganz et al. work on ELISA conferred that there is no cross reactivity between the hepcidin isoforms when using the ELISA technique . An ELISA based immunoassay was chosen because its performance is comparable to other technologies. It is also relatively cheap in comparison to technologies such as radioimmunoassay and mass spectrometry. These two attributes makes ELISA the right choice for an assay in the setting in which hepcidin testing was being proposed.
Full blood count, serum iron and hepcidin assays were performed on the blood samples that were collected from the participants. All three tests were performed in the laboratory. Tests for full blood count were performed on the day of sample collection while assays for hepcidin and serum iron were performed at a later date. The full blood count tests were performed using a Pentra XL 80 full blood count analyser .
The samples for Hepcidin assaying were tested in duplicate using the Hepcidin ELISA kits supplied by CUSIBIO®. Antihepcidin antibody pre-coated wells were supplied in the kit. The wells came in strips with twelve wells on each strip. One hundred microliters of prepared standards and samples were added into the wells and incubated for two hours at 37°C. At the end of the incubation period the liquid contents of the wells were decanted. One hundred microlitres of biotin labeled antibody was added to each well and the mixture incubated for 1 hour at 37°C. After which the liquid from each well was removed by aspiration. The wells were washed three times using the wash solution provided by the kit manufacturer. Subsequently, 100 µl of avidin conjugated to horse radish peroxidase (HRP) was added to each well and incubates at 37°C for 1 hour. At the end of the incubation period the HRP-avidin mixture was aspirated before the wells were washed five times with the wash solution. Ninety microlitres of the tetramethylbenzidine (TMB), a peroxidase substrate was added in the wells and incubated for at 37°C for 30 minutes.
The plates were protected from light to avoid it interfering with the color development. The color reaction was stopped by the addition of 50 µl of stop solution before reading the optical densities for each well at 450 nm. The optical density was read using a microtiter plate reader. The standards had the following hepcidin concentrations according to the manufacturer's instructions: 300 ng/ml, 150 ng/ml, 75 ng/ml, 37.5 ng/ml, 18.75 ng/ml, 9.375 ng/ml, and 4.6875 ng/ml. The logarithmic values of these concentrations were plotted against the logarithms of the optical densities obtained and this was the standard curve used to calculate sample results.