Underwriting Serum Ferritin
Why is this subject important?
“A high serum ferritin level may serve as a surrogate marker for poor health.”
Raymond Gambino, MD
We can say the same for an abnormally low serum ferritin, especially in males over 50 and post-menopausal females, doubly so if they are known to have microcytic/Fe-deficiency anemia or GI symptoms.
What is ferritin?
An iron-binding protein.
What is the “normal range” for ferritin?
There is some variance between laboratories (as is true with so many tests). That said, the normal ranges cited by Wallach in his widely used lab handbook are:
Normal Range ng/mL or mcg/L (µmol/L)
|6 mos-15 years||7-142 (16-320)|
|Adult female||15-120 (34-270)|
Others set ranges between 13-60 (29-135) on the low end, and 123-651 (277-1465) on the high end.
You must know the normal range of any ostensibly abnormal ferritin result reported in medical records before you can use it in risk assessment.
Why did you show ferritin values in nanograms per milliliter, micrograms per liter, and micro-moles per liter?
Because ng/mL is the measurement used in the USA and the two are measurements used elsewhere.
Hereafter we will show ng/mL, but you can easily convert to µmol/L by multiplying the ng/mL reading x 0.445.
How does menopause affect ferritin?
In one study of bodily iron stores in healthy postmenopausal US females, 3% had very low readings (< 12) and 10% had readings > 200. [Liu]
Would you expect the same pattern in premenopausal women?
No…because of the prevalence of significant menstrual blood loss, we would expect, if anything, the reverse, with low readings much more common than high.
When is ferritin testing done clinically?
It is not, as you know, a routine part of blood profile testing.
The main clinical uses of ferritin testing would be:
- Determine presence and extent of Fe (iron) overload in known/suspected primary (hereditary) or secondary iron overload disorders.
- Predict and monitor degree of iron deficiency.
- Determine response to Fe therapy or compliance with treatment.
- Differentiate Fe deficiency from other types of microcytic (low MCV) anemia.
- Monitor iron status in patients with chronic kidney disease, hemochromatosis, etc.
- As you can see, the range of uses predicts that we will often see ferritin reported in medical records.
What is the first question to ask when you see ferritin reported?
Why it was done…NOT whether or not it is normal (that’s the second question).
How common is elevated serum ferritin in the general North American population (without regard to the presence vs. absence of disease)?
6% overall. [Adams]
What is transferrin saturation (TS)?
Another measure of iron overload, more sensitive but less specific than ferritin.
How common is ferritin elevation in persons with TS elevation, again in the absence of known disease?
9% in males under 45 and 13% in females under 35, increasing to as much as 22% in males over 62 and 21% in females over 62. [Looker]
Is very high ferritin common overall among persons hospitalized, without regard to the reason?
Yes – 7% in one study [Lee, M. H.], which supports the opening observation by lab guru Raymond Gambino.
Why does ferritin increase in persons with a higher likelihood of disease?
Because it functions as an acute phase reactant…that is, a marker for inflammation.
Does inflammation affect TS in the same way?
No, it tends to lower TS, which is one reason why elevated ferritin in the presence of a normal or low TS reading is a clue to a cause other than iron overload. [Looker]
Does ferritin vary day-to-day in otherwise healthy people?
Yes, by as much as 15%, according to another lab guru named Ravel.
What factors interfere with ferritin reading validity?
- Recent transfusion or, possibly, donating blood.
- Recent ingestion of a meal with high iron content.
- Hemolytic disease may cause an artificially high reading.
- Menstruation may cause a lower-than-normal reading.
- Pregnancy, also lowering the reading.
- Use of iron supplementation therapy.
- Minimal elevation possible when taking oral contraceptives.
- Low ferritin with daily aspirin use, even at one-a-day level intake. [Haidari]
- No other commonly taken drugs are known to be associated with elevated readings according to the American Association for Clinical Chemistry’s reference book on Rx effects on lab tests.
Which lab tests are associated with elevated or below-normal ferritin?
- GGT – elevated ferritin is associated with elevated GGT, especially when heavy drinking is suspected and even in the absence of any known liver disease. [Kristenson]
- ALT and AST – but likely only in the presence of known or occult liver damage or clinically-silent hemochromatosis. [AACC reference book]
- Four pituitary hormones can exert a lowering effect: growth, follicle-stimulating, luteinizing and gonadotrophic – for what it is worth. [AACC reference book]
- A few studies have shown higher levels in persons with elevated cholesterol or triglycerides. [AACC reference book]
Does temperate alcohol intake affect ferritin?
In one study “down under,” they found this in otherwise healthy people [Rossi]:
Alcohol Intake/Day (grams/day)
Mean Ferritin Reading (ng/mL/mcg/L)
|1-10 (≤ 1 drink)||
|11-50 (1-4 drinks)||
|>50 (4+ drinks)||
Liu found similar results in postmenopausal females…ferritin increased but within the normal range at levels of 2-3/drinks per day (which is the upper level of “social drinking” stipulated for in women < 65 years old by most experts).
In other words, yes, temperate drinking (even as defined in my home state!) does increase ferritin, but not to any significant extent as far as we are concerned (that is, with the caveat that they did not segregate readings in persons consuming much higher levels!).
Hyperferritinemia (Elevated Serum Ferritin)
Is alcohol liver disease associated with high ferritin?
Does heavy intake of red meat increase ferritin levels?
Absolutely…but it would take a pretty devout carnivore to reach levels suggestive of undiagnosed disease! [Rossi]
What other factors may significantly raise ferritin?
One US study showed a higher incidence of elevations in obese persons and cigarette smokers…but the evidence for this does not appear sufficient in well persons to justify much concern in our view. [Kato]
Do highly trained athletes have higher ferritin levels than other people of the same age?
Yes, it would seem they do.
Lippi found that professional competition cyclists had mean levels of 332 ng/mL as compared to 127 ng/mL in amateur cyclists, and concluded that “professional endurance athletes have serum ferritin concentrations that are 2-fold to 3-fold higher than those matched sedentary individuals and amateur athletes.”
Do we have any clue as to why this is true?
Yes…Zotter showed that professional road cyclists use “excessive iron supplementation” – which is one form of blood doping designed to confer competitive advantage!
Maybe they should start monitoring ferritin in professional athletes?
What are the main pathological factors associated with elevated serum ferritin in the general population?
Zelber-Sagi found that both nonalcoholic fatty liver disease (NAFLD) and hyperinsulinemia (a prediabetic condition) were associated with high readings in a large unselected sample of the general population (in a country, we might add, where hereditary hemochromatosis is not very common!).
Does the presence of elevated ferritin in NAFLD correlate with an increased risk of NASH – that is, nonalcoholic steatohepatitis, the most significant type of NAFLD?
Yes…and also with an increased risk of advanced fibrosis in patients with NASH. [Trombini, Zelber-Sagi]
Thus, elevated ferritin would be an adverse finding in NASH or un-biopsied NAFLD (which could be NASH), from an underwriting perspective.
Is the same true in chronic hepatitis C?
Iron overload is significantly more common in chronic HCV than in the general population and elevated ferritin levels are found in 20-35% of patients. That said, recent studies now question at least five earlier ones that linked high ferritin to fibrosis.
Last year, Guyader and her French colleagues found that ferritin was not fibrogenic (associated with inducing fibrosis) but that is was a valid surrogate marker for greater disease severity overall.
Is high ferritin associated with known hyperglycemia?
Fumeron et al. studied 1,277 patients sorted by the presence (vs. absence) of hyperglycemia. They found that impaired fasting glucose (IFG) is associated with significantly increased readings, as compared to normoglycemic subjects.
However, this does not mean that most IFG cases will have elevated readings; rather, it only means that the average readings are greater when fasting glucose is above normal but not high enough to be considered IFG.
This is also wholly consistent with the earlier observation regarding the link between hyperinsulinemia and higher ferritin levels.
Is the same true for diabetics?
Likely if they are obese Type 2 diabetics who have NAFLD (a very common scenario, one might add).
Also in those few cases where (typically sudden onset, symptomatic) diabetes at age 45 or over is deemed to be due to hemochromatosis – with the iron overload accompanying hemochromatosis being the likely cause.
Canturk and associates found that poor control in DM was particularly likely to correlate with elevated ferritin.
Hughes reported that mean serum ferritin was also much higher in Chinese, Malay, Indian-Asian and Korean diabetics than in nondiabetics of the same ethnic origins.
What is the best clue to NAFLD in obese diabetics?
Elevated ALT, with elevated AST in some cases if the ALT-to-AST ratio is ≤1, and there is no evidence of any other prevalent cause of ALT elevation such as chronic hepatitis C (also more common in diabetics, to muddy the waters!), chronic hepatitis B, etc.
Does ferritin tend to be higher in the metabolic syndrome than in those free of this prevalent and insidious disorder?
For sure…but remember that chronic hyperglycemia and NAFLD are also quite prevalent in this syndrome.
In one study, the strongest link between elevated ferritin and metabolic syndrome was in those with hypertriglyceridemia (one of 5 criteria for the syndrome). [Vari]
Trombini found that the incidence of hyperferritinemia increased as the number of metabolic syndrome criteria increased in any given patient.
Is ferritin associated with cardiovascular disease?
“Strong epidemiological evidence is available that iron is an important factor in the process of atherosclerosis.”
B. de Valk, MD and J. J. Marx, MD
Archives of Internal Medicine
Is high ferritin associated with CAD?
Yes [You] and mostly at ages 50 and under as opposed to 60+. [Aronow, Haidari]
Has anyone shown that elevated ferritin is linked to a greater degree of ischemia-related heart impairment?
Say et al. found that hypercholesterolemic patients showed a significant correlation between magnitude of serum ferritin and probability of impaired perfusion, wall motion abnormalities and irreversible defects on imaging and angiographic studies.
Haidari in Iran found that males with angiographically-proven CAD had higher average ferritin readings than men who did not have disease…but the mean levels were well within the normal range for ferritin. Women, on the other hand, showed no evidence of an impact of ferritin in this context.
Is high ferritin linked to an increased risk of MI?
Two studies showed a definite, if modest, link between ferritin and MI. This was more pronounced if there were other CAD-risk comorbidities present. [Klipstein-Grobusch, Tuomainen]
Is ferritin linked to excess CV mortality?
Sempos reviewed 16 studies and found that 88% of them did not show a significantly increased risk.
On the other hand, van der A did establish that the highest tertile (top 3rd) of ferritin readings correlated with a 2-fold increased risk of stroke.
What is the bottom line where ferritin and CV disease is concerned?
While ferritin likely plays a role in the pathogenesis of atherosclerotic disease, it does not appear to be enough of a risk factor for us to consider in CV risk triage.
That said, we would argue that more than moderately-elevated ferritin – especially when the cause has not been found and addressed – is definitely incompatible with preferred risk!
Is elevated ferritin linked to cancer?
In the celebrated Framingham Offspring Cohort, elevated serum iron – which undoubtedly will be associated with elevated ferritin as well in many cases – was clearly a risk factor for cancer, doubly so when HDL-C was < 41. [Mainous]
Neoplastic disease is commonly cited in lab manuals as one of the chief causes of hyperferritinemia.
Should we be particularly concerned when the ferritin is mega-elevated (as in > 1000 ng/mL or > 2250 µmol/L)?
“Our findings suggest that serum ferritin extreme elevations [they used > 1000], taken by themselves, are a non-specific indicator of significant disease…”
Mark H. Lee, MD, et al.
The American Journal of Medicine
In other words, YES…it would not be incumbent upon us to offer insurance coverage in this context until a satisfactory workup was done to determine the cause of such a major elevation…
…and if no cause were established and dealt with (which is not an uncommon outcome), we would still be steadfast in postponing consideration because some life-threatening diseases may be below the threshold of discovery early on.
What other factors would make us doubly reluctant to consider mega-elevations?
- Age 65+
- Recent weight loss ≥ 10% of pre-loss weight, especially if not credibly explained (as nicely documented by Hearnshaw)
- Heavy, longtime cigarette smoking (as in pack years, not merely current use!)
- Very low or falling cholesterol (in the absence of hypolipidemic Rx)
- Low serum albumin
- Elevated GGT
What are the main causes of such mega-elevations?
- Neoplastic disease
- Alcoholic liver disease; other chronic and advanced liver pathologies
- Hematological diseases
- Chronic inflammatory diseases which are clinically active (symptomatic)
- Autoimmune diseases
- Advanced renal disease
- Symptomatic disease associated with HIV-1 infection
Hypoferritinemia (Ferritin Readings Below Normal)
What are the main causes of low serum ferritin?
- #1 is Fe deficiency anemia
- Chronic blood loss, usually with anemia present
- Disturbed iron absorption, which occurs in Crohn disease, more significant cases of celiac disease and other acute/chronic enteropathies (small bowel disorders)
What is considered to be a low ferritin in someone with depleted iron stores due to known or suspected Fe deficiency?
< 22 ng/mL. [Joosten]
When the reading gets down to < 12 ng/mL most experts concur that this is sufficient to make a diagnosis of iron deficiency.
Can asymptomatic hypoferritinemia be associated with serious disease?
In one important study, 5 of 64 asymptomatic older age patients with ferritin < 50 ng/mL who had diagnostic colonoscopy were found to have occult colon cancer and 2 more in the same cohort had large premalignant adenomatous colon polyps. [John Lee]
Among those with low ferritin who had upper GI endoscopy (132 patients) – whether symptomatic or not for undiagnosed illness – there were 2 cancers, 15 cases of advanced esophagitis and 31 patients with ulcers or erosions in either the stomach or the duodenum.
Joosten performed upper and lower GI studies on 151 older-age individuals with low ferritin (in this case, < 22 ng/mL or < 50 µmol/L) and found 18 colon cancers, 5 of which were not associated with overt anemia (low hemoglobin).
Lipato reported a 9.2% incidence of colon cancer in 414 subjects, mostly male and mean age 70, whose ferritin was < 50 ng/mL.
What is the correlation between low ferritin and hemoglobin levels?
J.G. Lee reported that patients with very low (≤ 20 ng/mL) ferritin had a mean Hb (hemoglobin) of 11.1 g/dL, while those between 21 and 50 had an average Hb reading of 12.7.
In other words, mild – rather than moderate-to-severe anemia – is the rule.
Did these patients also have low MCV readings?
Those with very low ferritin did; less so with readings in the 21-50 range…which means that ferritin is often an earlier indicator of risk than the presence of microcytosis.
Is the colon cancer risk increased if ferritin is > 100 ng/mL?
No…but the authors who reported this also maintained that older patients with readings < 100 “may benefit” from colonscopy. [Sawhney]
What is the bottom line on low ferritin?
- In Crohn disease or celiac disease, it portends a less favorable risk status.
- In younger persons (premenopausal females; males under ago 50) it probably doesn’t confer much added risk; that is, unless there are uninvestigated symptoms or other risk factors (polyp-forming syndromes, for example).
- At older ages, and progressively so at 65 and over, applicants with very low readings are – in our opinion – uninsurable, and we would maintain a high index of suspicion even between 21-50 ng/mL or 47-113 µmol/L depending on other case context factors.
Should we use ferritin as a screening test?
That said, one could fashion an argument for doing so at ages 65 and over based on all of the foregoing. Indeed, it would not surprise this underwriter if the protective value of screening with serum ferritin was higher than many of the things we do now at ages 65+.
Moreover, as it is not a tumor marker, we wouldn’t have the problems that make tests like CEA and its kindred the potential nightmares for insurers that abundant experience has shown them to be.
Note to Readers
This paper on underwriting ferritin tests contains material originally intended to be included in our CE course on HEMOCHROMATOSIS…but that course turned out to be long enough, so we saved it for this project.
The format of this course is similar to our CE and IE courses, giving you a good idea of what CE, in particular, is like. We hope you find it appealing in format, style and nature of content!
Talk to us! Did you find this article helpful? How could it be improved?
Adams. The New England Journal of Medicine. 352(2005):1769
Aronow. American Journal of Cardiology. 72(1993):347
Canturk. Endocrinology Research. 29(2003):299
Fumeron. Diabetes Care. 29(2006):2090
Gleeson. American Journal of Gastroenterology. 101(2006):304
Guyader. Journal of Hepatology. 46(2007):587
Haidari. Clinical Chemistry. 47(2001):1666
Hearnshaw. World Journal of Gastroenterology. 12(2006):5866
Hughes. Arteriosclerosis. 136(1998):25
Joosten. The American Journal of Medicine. 107(1999):24
Kato. International Journal of Vitamin and Nutrient Research. 70(2000):119
Klipstein-Grobusch. American Journal of Clinic Nutrition. 69(1999):1231
Kristenson. Drug and Alcohol Dependence. 8(1981):43
Lee, J. G. American Journal of Gastroenterology. 93(1998):772
Lee, M. H. The American Journal of Medicine. 98(1995):566
Lipato. American Journal of Gastroenterology. 99, Supplement (2004):S334
Lippi. Clinical Journal of Sports Medicine. 15(2005):356
Liu. American Journal of Clinical Nutrition. 78(2003):1160
Looker. Annals of Internal Medicine. 129(1998):940
Mainous. American Journal of Epidemiology. 161(2005):1115
Rossi. Clinical Chemistry. 47(2001):202
Sawhney. American Journal of Gastroenterology. 102(2007):82
Sempos. Annals of Epidemiology. 10(2000):441
Trombini. Journal of Hepatology.46(2007):549[editorial]
Tuomainen. Circulation. 97(1998):1461
van der A. Stroke. 36(2005):1637
Vari. Diabetes Care. 30(2007):1795
You. Clinica Chimica Acta. 357(2005):1
Zelber-Sagi. Journal of Hepatology. 46(2007):700
Zotter. British Journal of Sports Medicine. 38(2004):704
Insureintell.com February 2008