During the last 10 years we have experienced an increasing number of referrals due to hyperferritinemia. This is probably due to increased awareness of hereditary hemochromatosis, and the availability of a genetic test for this condition. Most of these referred patients were over-weight middle-aged men with elevated ferritin levels, but without the hemochromatosis-predisposing gene mutations. We evaluated the relationship between hyperferritinemia and the metabolic syndrome in 40 patients.
Forty consecutive patients referred for hyperferritinemia were investigated. The examination programme included medical history, clinical investigation and venous blood samples drawn after an overnight fast. This resulted in 34 patients with unexplained hyperferritinemia, which were further examined. Liver biopsy was successfully performed in 29 subjects. Liver iron stores were assessed morphologically, and by quantitative phlebotomy in 16 patients.
The majority of the patients had markers of the metabolic syndrome, and 18 patients (52%) fulfilled the IDF-criteria for the metabolic syndrome. Mean body mass index was elevated (28,8±4,2), mean diastolic blood pressure was 88,5±10,5 mmHg, and mean fasting insulin C-peptide 1498±539 pmol/l. Liver histology showed steatosis and nuclear glycogen inclusions in most patients (19 out of 29). Only four patients had increased iron stores by histology, of which two could be explained by alcohol consumption. Fourteen of 16 patients normalized ferritin levels after phlebotomy of a cumulative blood amount corresponding to normal iron stores. Ferritin levels were significantly related to insulin C-peptide level (p<0.002) and age (p<0.002).
The present results suggest that liver steatosis and insulin resistance but not increased iron load is frequently seen in patients referred for suspected hemochromatosis on the basis of hyperferritinemia. The ferritin level seems to be positively associated to insulin resistance.
Citation: Brudevold R, Hole T, Hammerstrøm J (2008) Hyperferritinemia Is Associated with Insulin Resistance and Fatty Liver in Patients without Iron Overload. PLoS ONE 3(10): e3547. doi:10.1371/journal.pone.0003547
Editor: A. Tavill, MetroHealth Medical Center, United States of America
Received: April 27, 2007; Accepted: September 24, 2008; Published: October 28, 2008
Copyright: © 2008 Brudevold et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Hemochromatosis is a disease with vague symptoms, serious complications and an effective treatment. Since the genetic test for hereditary hemochromatosis became available in 1996, focus on the disease has increased . Serum-ferritin is frequently used as a first line blood test when iron overload is suspected. However, ferritin is also elevated in inflammatory states and malignant diseases .
The hemochromatosis-associated genotype (homozygous C282Y mutation of the hfe- gene) is found in about 85% of patients with hereditary hemochromatosis . The hfe-gene codes for the HFE-protein, which plays an important role in iron absorption in enterocytes .
A possible link between hepatic iron-overload and insulin resistance has been reported –, and an association of non-alcoholic steatohepatitis (NASH) and insulin resistance has also been shown . In some patients with NASH ferritin was elevated.
Since 1996 we have experienced an increasing number of referrals from general practitioners of patients with hyperferritinemia. Many patients were overweight, middle-aged men with vague symptoms such as tiredness, fatigue, myalgia and arthralgia. Liver transaminases, especially alanine amino transferase (ALT), were often elevated. Transferrin saturation was within the reference range and the HFE gene mutation was negative. Liver biopsy often showed steatosis and nuclear glycogen inclusions, which is a frequent finding in prediabetes and diabetes mellitus type 2 (DM2) . Usually there were no definite signs of iron overload or other pathology.
We have examined the relationship between hyperferritinemia, iron-load and insulin resistance in patients referred for elevated ferritin levels and suspicion of hemochromatosis. Patients were examined and treated according to best practice.
During 2001 40 patients were referred to the hematology clinic at Aalesund Hospital due to elevated ferritin levels and suspicion of hemochromatosis. This is an observational study with retrospective analysis of the results of the examination programme. All patients gave oral consent to the examination programme. Due to the retrospective observational nature of the study, the Regional ethics committee of Mid-Norway had no comment to the study, and as the examinations themselves were part of a clinical quality assurance program, they did not need formal approval from the ethics committee. However, all patients gave oral consent to the use of clinical data for research purposes.
Five patients had hereditary hemochromatosis and one had hemosiderosis secondary to excessive iron intake. Hence 34 patients (33males, 1 female) had no hfe-mutation known to predispose for hemochromatosis (neither homozygous for C282Y nor compound heterozygous, i.e. heterozygous for both C282Y and H63D) and was further examined to reveal other causes for hyperferritinemia.
A structured examination programme was performed including medical history, clinical examination, laboratory tests and liver biopsy. Reason for taking liver biopsy was hyperferritinemia/elevated liver transaminases where examinations so far had not given any explanation for the deviations. Examinations did not deviate from what is usually recommended for follow-up of such patients. The HUNT-study  was used as reference population regarding BMI (women; 26.4 kg/m2, men; 26.5 kg/m2), waist-hip ratio (women; 0.77±0.06, men; 0.89±0.06), systolic blood pressure (women; 136.8 mmHg, men; 140.1 mmHg) and diastolic blood pressure (women; 78.8 mmHg, men; 81.9 mmHg).
Venous blood samples were drawn after overnight fast.
S-iron (reference values, women; 11–23 µmol/l, men; 14–29 µmol) was analysed using Colorimetric assay (CFAS-Roche, Mannheim, Germany). Serum-glucose (reference value 3.6–5.8 mmol/l), aspartate amino tranferase (AST) (reference values, women; <35 U/l, men; <50 U/l), alanine amino transferase (ALT) (reference values, women; <35 U/l, men; <50 U) were measured by UV-test (CFAS-Roche, Mannheim, Germany). C-reactive protein (CRP) was measured by a turbidimetric method (CFAS-Roche, reference value <6 mg/l). Hemoglobin A1C (HbA1C) (reference value, 3.5–6.5%) used ion-change-chromatography (HPLC-Biorad Variant Hb testing system, Marnes-La-Cocuette, France). Ferritin (reference values, women; 10–250 µg/l, men; 20–250 µg/l) and total iron binding capacity (TIBC)(reference value 46–80 µmol/l) were analyzed by nephelometry (Dade Behring. Marburg, Germany). Transferrin saturation (%) = s-iron/TIBCx100 (reference value, 15–45%). Fasting insulin C-peptide (reference value 240–720 pmol/l) was analysed by chemiluminescence method (DPC-immulite. Liamberis, Gwynodd, LL55 4EL, UK). HFE mutation was analyzed measuring fragment-length polymorphisms by PCR .
Liver biopsies were ultrasound-guided with 18-Gauge needles. Iron content was evaluated according to Rowe : Grade 0–1 is within normal limits, and grades 2+ to 4+ represent distinct iron increase (70–850 µmol/g dry weight).
Iron stores were also estimated by quantitative phlebotomy in 16 patients. One gram of hemoglobin contains 3.4 mg iron, and the amount of iron removed by phlebotomy can thus be calculated –. The normal amount of storage iron in males, measured by quantitative phlebotomy, has been shown to be about 750 mg –.
There have been several definitions of the metabolic syndrome (WHO 1999/NCEP 2001)–. We used the recent simplified definition recommended by the International Diabetes Federation (IDF), which emphasizes waist circumference (women; >80 cm, men; >94 cm) in addition to minimum 2 of the following criteria: HDL-cholesterol <1.03 mmol/l (men), blood pressure >130/85 mmHg, triglycerides >1.7 mmol/l, or fasting blood glucose >5.6 mmol/l .
Normal distribution of data was verified by normality plots. All normally distributed measurements are presented as mean and standard deviations (SD), otherwise median and range were used. Comparisons of normal distributed parameters were done by t-tests. Univariate and multiple forward linear regression analyses were used to investigate relations between s-ferritin and other variables. Logaritmic transformation of s-transferrin was used to obtain normal distribution. All tests were 2-sided, and a p-value <0.05 was considered significant. No adjustments for multiple testing were performed. All analyses were performed with SPSS 13.0 (SPSS Inc, USA).
Role of funding source
The funding source had no involvment in the conduction of the study.
The mean age was 53 (12.6) years.
Twenty-three percent reported DM2 among relatives, and 23% also reported some kind of iron-overload-related problem in the family. Fourteen percent knew about relatives with some kind of malignant liver disease (table 1).
Symptoms reported by the patients were vague. Fifty-eight percent reported fatigue and 17% arthralgia/myalgia.
Seventeen patients (50%) were without any medication. Most used medicines were betablockers (7 patients), calcium-channel blockers (5 patients), ACE-inhibitors (4 patients) and metformin (4 patients).
TSH, CRP, Sedimentation rate (SR), s-electrophoresis and transferrin saturation were all within reference limits (Table 2).
Mean values for body mass index, waist circumference, waist/hip ratio and diastolic blood pressure were above cutoff values set by IDF regarding criteria for the metabolic syndrome.
Thirty-two out of 34 patients (94%) had elevated insulin C-peptide indicating hyperinsulinemia/insulin resistance.
Mean glucose was slightly elevated (6.41±2.2 mmol/l), s-triglyceride was in upper normal range, mean HDL-cholesterol was in lower normal range, and mean ALT was elevated.
Hfe- mutation analyses showed that 22 patients were heterozygous for C282Y, and twelve had neither C282Y nor H63D. There were no compound heterozygotes.
None of the present genetic combinations are known to predispose for iron overload, although heterozygous C282Y has been associated with elevated serum iron markers .
Liver histology is reported in table 3.
The majority of patients had liver steatosis and nuclear glycogen inclusions. Hemosiderin was within normal limits in 86% of the patients. The one patient with hemosiderin content grade 3 was diagnosed with steatohepatitis and surplus alcohol consumption (carbohydrate deficient transferrin>5%). The other patient with alcoholic steatohepatitis had also cirrhosis and iron storage grade 2 and had earlier suffered from heavy alcohol intake. Two other patients had microscopic sign of iron overload (grade 2), without any other obvious pathology. However the amount and distribution was not typical for hemochromatosis (not hepatocytic).
Sixteen patients were offered phlebotomy due to symptoms described earlier. Mean volume of whole blood needed for reaching low normal ferritin level (<100 µg/l) was 2500 ml (1300 ml–4450 ml). Accordingly approximately 1150 mg iron was removed for reduction of s-ferritin to low normal levels.
In univariate linear regression analysis log s-ferritin was significantly related to age (p<0.002) and insulin C-peptide (p<0.002). These parameters were also significantly related to s-ferritin in multivariate regression analysis (p<0.002) (Fig. 1 and 2). There was no significant relation between s-ferritin and transferrin saturation, number of phlebotomies required to normalize s-ferritin, ALT, LDL-cholesterol, HDL-cholesterol, trigyserides, diastolic blood pressure or BMI.
Eighteen of thirty-four (52%) patients fulfilled the IDF-criteria. In the 18 men fulfilling the criteria relative to the other male patients median ferritin was 603 µg/l (379–1300) compared to 515 µg/l (283–871) (p = 0.05), transferrin saturation was 36% compared to 44% (p<0.05) and mean insulin C-peptide was 1732 pmol/l compared to 1162 pmol/l (p = 0.001).
All liver biopsies showed steatosis in patients fulfilling the IDF-criteria, with only one showing an iron content corresponding to grade 2, the rest grade 0–1 (normal).
Hepatic iron overload was present in only 4 patients (4/29). These had also slightly elevated transferrin saturation. In two of them increased iron stores were explained by alcohol consumption. Liver histology, transferrin saturation, and number of phlebotomies needed to normalize s-ferritin, indicated normal iron-load in the remaining population. There was no sign of inflammation or malignancy. Steatohepatitis was found in three patients. One of these was diagnosed as NASH.
The majority of the study population showed markers of the metabolic syndrome. BMI, waist-hip ratio, diastolic blood pressure, ALT, HDL-cholesterol, fasting plasma glucose and insulin C-peptide showed deviations typically seen in the metabolic syndrome. Elevated insulin C-peptide levels indicate hyperinsulinemia and insulin resistance –. Liver histology showed steatosis and nuclear glycogen inclusions, a typical finding in prediabetes and DM2.
Those patients who met the criteria for the metabolic syndrome had slightly higher ferritin and C-peptide values, but lower transferrin saturation than the rest of the study population. One of these had iron storage grade 2, the rest grade 0–1 in liver biopsies.
An association between hyperferritinemia and insulin resistance in patients with different types of liver pathology has been reported earlier. In the article by Moirand et al  the patients had overt hepatic iron overload, as the median liver iron concentration was 85 µmol/g dry weight (normal <36 µmol/g).
A relationship between non-alcoholic fatty liver disease and insulin resistance , and a relationship between s-ferritin and metabolic markers of insulin resistance but without an evaluation of liver-histology –, has previously been reported. In 1992 Dinnen et al reported that hyperferritinemia was a feature of newly diagnosed DM2 , and they later presented results indicating that DM2 did not associate with iron overload . This finding has recently been confirmed by others . Inflammatory mechanisms were assumed to be responsible for the hyperferritinemia in these diabetic patients.
Our patients differed from those in the above mentioned studies by being mostly non-diabetic, without hepatic iron overload or other sign of liver pathology other than steatosis and nuclear glycogen inclusions. Our patients had more expressed markers of the metabolic syndrome than in the above studies, and mean ferritin levels were higher. This tendency was even more striking in the group with the metabolic syndrome. In addition ferritin levels correlated well with fasting insulin C-peptide.
Elevated ferritin levels in our patients seemed not to be due to iron. As elevated insulin C- peptide level is a marker of hyperinsulinemia and indirectly insulin resistance, insulin resistance might be the cause of elevated ferritin levels in our patients, either directly or due to the steatosis itself.
Patients treated with phlebotomy experienced reduction in s-ferritin, but after a while the level gradually increased. Those who changed lifestyle (weight management, increased physical activity) seemed to experience more stable s-ferritin reduction. This has to be evaluated in further studies, and in particular correlation needs to be observed between a durable fall in serum ferritin in response to weight control and exercise, and a correction in insulin resistance and reversal of fatty liver.
The finding of steatosis and nuclear glycogen inclusions further support this theory, as it is a common finding in obesity and prediabetes. We did not find signs of primary liver disease.
The frequency of hyperferritinemia not related to iron load is remarkable, as is the extreme male preponderance. Insulin resistance and metabolic syndrome is more prevalent in males , but not to the extent seen in this group. Confounding factors may have influenced patient selection, but hormonal effects and increased iron loss in females may also have been of importance. Selection bias on part of the referring physician is not possible to exclude.
Possible explanations of hyperferritinemia among our patients might be that by binding iron, ferritin protects against formation of free hydroxyl groups, it could also be due to endothelial/subendothelial inflammation – or simply leakage of ferritin from hepatic cells.
The frequency of hyperferritinemia in Norwegian unselected patient populations with insulin resistance is unknown, as are gender difference.
The greatest limitation of the study is the relatively small patient population, and male preponderance. The strength though, is the unselected patient population from a primary care hospital, and the performance of liver histology from most of the patients.
In conclusion, the present study suggests that s-ferritin elevation in our patients is a marker of the metabolic syndrome with hepatic steatosis and insulin resistance, and not of iron overload. The direct pathogenic mechanism, however, remains unknown.
We want to thank our colleagues for their valuable support and advice: Bjørn Westre, MD, Department of Pathology, Staale Skare, MD, Department of Medicine and Kaare Michelsen, MD, PhD, Department of Laboratory Medicine, Aalesund Hospital, and Arne Aasberg, MD, PhD, Department of Laboratory Medicine, Henrik Hjort-Hansen, MD, PhD, Section of Hematology, Sven Carlsen, MD, PhD, Section of Endocrinology, Department of Medicine, St.Olavs Hospital, Trondheim, Norway.
Conceived and designed the experiments: RB. Performed the experiments: RB. Analyzed the data: RB TH JH. Contributed reagents/materials/analysis tools: RB. Wrote the paper: RB TH JH.
- 1. Feder JN, Gnirke A, Thomas W, et al. (1996) A novel MHC class 1-like gene is mutated in patients with hereditary haemochromatosis. Vat Genet 13: 399–408.
- 2. Worwood M (1986) Serum ferritin. Clin Science 70: 215–20.
- 3. Merryweather-Clarke AT, Pointon JJ, Shearman JD, Robson KJH (1997) Global prevalence of putative haemochromatosis mutations. J Med Genet 34: 275–77.
- 4. Parkkila S, Waheed A, Britton RS, et al. (1997) Immunhistochemistry of HLA- H, the protein defective in patients with hereditary hemochromatosis, reveals unique pattern of expression in gastrointestinal tract. Proc Nat Acad Sci USA 94: 2534–9.
- 5. Moirand R, Mortaji AM, Loréal O, Paillard F, Brissot P, et al. (1997) A new syndrome of iron overload with normal transferrin saturation. Lancet 349: 95–97.
- 6. Mendler M-H, Turlin B, Moirand R, Juanolle AM, Sapey T, et al. (1999) Insulin resistance-associated hepatic iron overload. Gastroenterology 117: 1155–63.
- 7. Ferrannini E (2000) Insulin resistance, iron and the liver. Lancet 355: 2181–2182.
- 8. Marchesini G, Brizi M, Morselli-Labate AM, Bianchi G, Bugianesi E, et al. (1999) Association of nonalcoholic fatty liver disease with insulin resistance. The American Journal of Medicine 107: 450–55.
- 9. Caramia F, Ghergo FG, Branciari C, Menghini G (1968) New aspects of hepatic nuclear glycogenosis in diabetes. J Clin Pathol 21: 19–23.
- 10. Holmen J, Midthjell K, Krüger Ø, et al. (2003) The Nord-Trøndelag Health Study 1995–97 (HUNT 2): Objectives, contents, methods and participation. Norsk Epidemiologi 13: 19–32.
- 11. Searle JW, Kerr JFR, Halliday JW, Powell LW (1987) Iron storage disease. In: McSween RNM, Anthony PP, Scheuer PJ, editors. Pathology of the liver, 2nd edn. Edinburgh: Churchill Livingstone.
- 12. Worwood M (1994) Laboratory determination of iron status. In: Brock JH, Halliday JW, Pippard MJ, Powell LW, editors. Iron metabolism in health and disease. London: W.B Saunders. pp. 449–76.
- 13. Jandl JH (1996) Blood: Textbook of hematology. 2nd ed. Boston: Little Brown.
- 14. Walters OG, Miller FM, Worwood M (1973) Serum ferritin concentration and iron stores in normal subjects. J clin pathol 26: 770–2.
- 15. Tavill AS (2001) Diagnosis and management of hemochromatosis. Hepatology 33: 1321–8.
- 16. WHO (1999) Definition, diagnosis and classification of diabetes mellitus and its complications. Geneve: WHO.
- 17. Expert panel on detection, evaluation and treatment of high blood cholesterol in Adults (2001) JAMA 285: 2486–97.
- 18. Alberti KGMM, Zimmet P, Shaw J (2005) The metabolic syndrome - a new worldwide definition. Lancet 366: 1059–61.
- 19. Pietrangelo A (2004) Hereditary haemeochromatosis – a new look at an old disease. N Engl J Med 350(23): 2383–97.
- 20. Polonsky KS, Rubenstein AH (1984) C-peptide as a measure of the secretion and hepatic extraction of insuline. Pitfalls and limitations. Diabetes 33: 486–94.
- 21. Ferrannini E, Haffner SM, Mitchell BD, Stern MP (1991) Hyperinsulinemia; the key feature of cardiovascular and metabolic syndrome. Diabetologia 34: 416–22.
- 22. Tuomainen TP, Nyyssónen K, Salonen R, Tervahauta A, Korpela H, et al. (1997) Body iron stores are associated with serum insulin and blood glucose concentrations. Diabetes Care 20(3): 426–28.
- 23. Fernández-Real JM, Ricart-Engel W, Arroyo E, Balanca R, Casamitjana-Abella R, et al. (1998) Serum ferritin as a component of the insulin resistance syndrome. Diabetes Care 21: 62–8.
- 24. Dinneen SF, O'Mahony MS, O'Brien T, Cronin CC, Murray DM, et al. (1992) Serum ferritin in newly diagnosed and poorly controlled diabetes mellitus. IJMS 161: 636–38.
- 25. Dinneen SF, Silverberg JD, Batts KP, O'Brien PC, Ballard DJ, et al. (1994) Liver iron stores in patients with Non-Insulin-Dependent Diabetes Mellitus. Mayo Clin Proc 69: 13–15.
- 26. Hernández C, Lecube A, Carrera A, Simó R (2004) Soluble transferrin receptors and ferritin in type 2 diabetic patients. Diabet Med 22: 97–101.
- 27. Hsueh WA (2003) New insight into understanding the relation of type 2 diabetes mellitus, insulin resistance and cardiovascular disease. Am J Cardiol 92: 10–16.
- 28. Sjoholm Å, Nystrom T (2005) Endothelial inflammation in insulin resistance. Lancet 365: 610–12.
- 29. Bonora E, Kiechl S, Willeit J, Oberhollenzer F, Egger G, et al. (2003) Carotid atherosclerosis and coronary heart disease in the metabolic syndrome; prospective data from the Bruneck study. Diabetes Care 26: 1251–7.
- 30. Alexander CM, Landsman PB, Teutsch SM, Haffner SM (2003) NCEP-defined metabolic syndrome, diabetes and prevalence of coronary heart disease among NHANES III participants age 50 years and older. Diabetes 52: 1210–4.