The study population comprised patients with iron deficiency anaemia and transfusion-related anaemia that showed a difference in body iron content, who were being followed up in the Haematology Department of Internal Medicine at Erciyes University Medical Faculty and Kayseri Training and Research Hospital (Kayseri, Turkey) between September 2010 and October 2011. The control group was formed of age and gender-matched subjects with no history of underlying disease. The transfusion-related anaemia group, with an excessive amount of body iron, comprised of 20 patients diagnosed with myelodysplastic syndrome or β thalassemia from erythrocyte suspension of ≥30 units. The iron deficiency anaemia group comprised of 25 patients and the control group comprised of 21 healthy subjects were evaluated as within the reference values of TS and ferritin levels. Any cases who were pregnant; had a history of acute or chronic inflammatory disease, malignant disease, infection or renal failure; had any underlying disease, such as malabsorption syndrome, which affected serum iron and ferritin levels; or any disease associated with the hair or scalp that affected the hair iron concentration, were excluded from the study.

Serum iron, TIBC, TS, ferritin and hair iron concentration were examined in all the cases together with haemoglobin, haematocrit and erythrocyte markers, such as mean erythrocyte volume (MCV), mean erythrocyte haemoglobin (MCH), MCH concentration (MCHC) and erythrocyte distribution width (RDW). TS was calculated using the formula of serum iron / iron-binding capacity × 100. The diagnosis of iron deficiency anaemia was made according to the World Health Organization criteria (5). Haemoglobin was accepted as <13 g/dl in males and <12 g/dl in females. Iron deficiency anaemia cases were classified according to reference intervals, as serum iron low, TIBC high; transferrin saturation <15% and serum ferritin <15 ng/dl.

Fasting blood samples were taken in the morning. Full blood count was assayed using an automatic electronic full blood count analyser (Siemens Advia 2120; Siemens Healthcare Diagnostics, Deerfield, IL, USA), serum iron and iron-binding capacity using a biochemical analyser (Siemens Advia 1800 Chemistry System; Siemens Healthcare Diagnostics) and ferritin using a hormone analyser ADVIA Centaur XP (Siemens Healthcare Diagnostics, Deerfield, IL, USA).

Prior to the hair sampling process, no history of cosmetic hair application, such as lightening, dyeing or perm, was found in any of the cases. The hair samples in the patient and control groups were obtained with stainless steel scissors from the occipital area as far as the proximal section. Each sample collected was stored in an autosealable polythene bag and washed sequentially with acetone, deionized water and acetone, using three successive portions. Subsequently, the samples were dried in an oven at 80˚C. Each hair sample, weighing 250 mg, was transferred into a digestion vessel of 10 ml capacity and concentrated nitric acid (5 ml) was pipetted into the vessel. The samples were diluted with 65 % (v/v) nitric oxide in a microwave oven (Berghof Speedwave, Eningen, Germany). Analytic techniques were applied with Agilent 7500a Inductively Coupled Mass Spectrometer (Agilent Technologies, Tokyo, Japan). Digestion solutions were diluted and analyzed on an Agilent 7500a series ICP/MS. For the quality control, the concentration of internal standards 200 ppb (9Be,45Sc,103Rh,208Bi) and reference materials were run together with the samples and before the analyses tune parameters were cotrolled. To obtain the detection limit of each element the axis cut point of the calibration line was used. It was also important to ensure the linear rank of the methodology analyzing different standards with low and high known concentrations of each element (0, 1, 5, 10, 20, 30, 40 and 50 ppb). At least five different reference materials covering all the elements in the study were used. Duplicate samples were also used in order to determine precision of the analysis. For the iron element a minimum of three standards were used to cover the analytical working range of the instrument. Ultrapure water was used to prepare calibration standards and blanks, and three replicate determinations were performed for each sample. From all the sample results, a reagent blank was subtracted. Detection limits were calculated as three times the standard deviation for the reagent blanks. LODs value (in microgram per milliliter) was 0.0008 (Fe). Hair concentrations were evaluated as micrograms per gram.

Data were analysed using SPSS v20 software statistical program (IBM Corp, Armonk, NY, USA). In the evaluation of statistical differences between groups of numerical values with normal distribution, one-way analysis of variance was used. In the statistical evaluation between groups of parameters such as age, haemoglobin, MCV, MCH, MCHC, RDW, serum iron, hair iron levels and TS, which did not return to normal distribution despite logarithmic transformation, the Kruskal-Wallis test was used. In all the statistical tests, a value of P<0.05 was considered to indicate a statistically significant difference.

The mean age was 33 (range, 14–48 years) in the iron deficiency anaemia group, 22 (range, 15–75 years) in the transfusion-related anaemia group and 28 (range, 13–42 years) in the control group. The degree of statistical significance of these findings together with the haematological and biochemical findings of the cases is shown in Table I. According to this, a statistically significant difference was found between the groups in all parameters except for age (Table I).

The mean hair iron ⁵⁶Fe and ⁵⁷Fe concentrations of the iron deficiency group were measured as 5.08 and 6.03 µg/g, respectively, and in the transfusion-related anaemia group these values were 28.88 and 29.44 µg/g, respectively. In the control group, the mean hair iron ⁵⁶Fe and ⁵⁷Fe concentrations were measured as 12.01 and 17.62 µg/g, respectively. The highest hair iron concentration of the ferritin value (89.41 µg/g) was observed in transfusion-related anaemia patients, whereas the lowest hair iron concentration (0.77 µg/g) was determined in the iron deficiency anaemia group. The differences between the three groups with respect to hair iron ⁵⁶Fe and ⁵⁷Fe concentrations were found to be statistically significant (Figs. 1 and 2).

Correlation analysis applied to the parameters showed a strong positive correlation between hair iron ⁵⁶Fe and ⁵⁷Fe (r=0.776, P<0.001). A statistically significant positive correlation was determined between hair iron ⁵⁶Fe and ⁵⁷Fe concentrations and serum iron, ferritin level, transferrin saturation, MCV and MCH values. A statistically significant negative correlation was determined between hair iron ⁵⁶Fe and ⁵⁷Fe concentratons and TIBC. The results of the correlation analysis are shown in Table II.