Prevalence of iron deficiency and iron overload in the adult icelandic population

J CUnEM&anblVol.44,No. 12,PP.1289-1297, 1991 Printedin Great Britain.All rightsrrservcd

0895-4356/91 $3.00+ 0.00

copyri$ht0 1991Pcrglunon pressplc

PREVALENCE OF IRON DEFICIENCY AND IRON OVERLOAD IN THE ADULT ICELANDIC POPULATION JON J. JoNssow,‘~*~~*GUDMUNDURM. JOHANNESSO N,’ NIKULAS SIGFUSSON,~ BJARKIMAGNUSSON,~ BJARNITI-LJODLEIF~SON~ and SIGMUNDUR MAGNUS,SON~ ‘Department of Laboratory Medicine and Pathology, University of Minnesota Hospital and Clinic, Harvard St at E. River Rd, Minneapolis, MN 55455, U.S.A. and Departments of 2Clinical Biochemistry, ‘Hematology, ‘Pathology and 5Medicine, University of Iceland, Reykjavik and 6Heart Preventive Clinic in Reykjavik, Iceland (Received in revised fom 23 July 1991)

Abstract-The aim of this cross-sectional study was to estimate the prevalence of iron deficiency and overload in the adult population in Iceland, a developed Scandinavian country. The study population consisted of 4240 individual8 aged 25-74 years randomly selected from the national roster. Basic hematological, S-iron, S-total iron binding capacity (TIBC), and S-ferritin measurement8 were obtained on 2588 individual8 (61.0%). The results indicated unusually large iron stores in the adult Icelandic population and significantly larger iron stores in the rural compared to the urban population. Iron deficiency was rare except in urban premenopausal women, where 1 in 4 showed evidence of iron deficiency and 3.2% had iron deficiency anemia. Seven patients with hereditary hemochromatosis were identified from a subgroup of 1887 subjects, resulting in a prevalence of 0.37%. Two of the hereditary hemochromatosis patients had been gastrectomized. Measure8 to improve the iron balance in urban premenopausal women cannot therefore include increased iron fortification of food but must be more directed towards the target group. Epidemiology

Iron

Anemia

Ferritin

INTRODUCTION

Hereditary hemochromatosis (HH) was once assumed to be a rare disease associated with 1: 20,000 of hospital admissions and 1: 7000 of hospital deaths [l]. Studies in the last decade, using several different approaches, have revealed a considerably higher prevalence. Family studies in Utah, U.S.A. [2] and Brittany, France [3] found a gene frequency (q) of approx. 0.05 predicting that 1 in every 400 individuals is a homozygote for the still unidentified hemochromatosis gene. These results are in agreement with an autopsy report from Scotland [4]. Olsson and associates [5] examined the prevalence of iron overload among 30- to 39-year-old *Author for correspondence.

Hemochromatosis

Gastrectomy

Swedish males in a cross-sectional study, and observed that 0.5% of 623 investigated subjects had HH (q = 0.069).Biochemical screening of 11,065 blood donors in Utah, U.S.A. showed a similar prevalence [6], as have several studies in various countries on populations of Western European descent [7-91. Approximately 15 HH patients had been diagnosed prior to this study (unpublished observations) in Iceland, a country of 238,416 inhabitants [lo]. If we assume the prevalence of HH to be similar to the studies cited above a significant number of undiagnosed patients should be present in the adult Icelandic population. The prevalence of iron deficiency reflects on the nutritional state of a nation [ 1l] and estimates the efficacy of primary health care, since

1289

1290

JON J. JONSSONet al.

the condition is readily diagnosed and curable [12]. Severa; stages of iron deficiency have been defined from iron depletion and iron deficient erythropoiesis where the S-ferritin < 16 pg/l and S-TIBC saturation < 16% to the most advanced stage iron deficiency anemia [13]. In iron deficiency anemia B-hemoglobin concentration is decreased in addition to the aforementioned biochemical criteria. Thalassemia and related genetic diseases of hemoglobin formation are virtually non-existent in the Icelandic population (authors’ experience). The prevalence of anemia in the general Icelandic population should therefore correlate well with the iron status of the population. The Heart Preventive Clinic in Reykjavik (the capital city of Iceland) measured B-hemoglobin in 2955 males and 2333 females 34-61 years old during the years 1967-l 969 [ 14, unpublished observations]. The prevalence of anemia (B-hemoglobin < 130 g/l in males and < 120 g/l in females) was 2.5% in males and 9.9% in females. Specific iron parameters were not measured in this study. Recently, Steingrimsdottir [ 151 studied 198 randomly selected lo- to 1l-year-old schoolchildren in Reykjavik. Ten percent of the children were anemic (B-hemoglobin < 120 g/l) but none had concurrently decreased S-TIBC saturation and S-ferritin. Surveys on iron content or bioavailability of the modern Icelandic diet have not been reported and there are no governmental regulations in Iceland regarding iron fortification of food. However, the vast majority of our flour and cereals are imported in consumer packages from the U.S.A. and contain the same degree of iron enrichment as if marketed in that country. In countries where HH is as prevalent as recent studies have found [2-9), ascertaining the relationship betwen iron fortification of food and iron status of the population becomes essential. This study was therefore undertaken to determine the prevalence of iron deficiency and iron overload in the adult urban and rural population in Iceland. MATERIALS

AND METHODS

The study population consisted of 4240 individuals divided into two groups participating in epidemiological studies on cardiovascular risk factors by the Heart Preventive Clinic in Reykjavik, Iceland. The first group (designated A) consisted of 3000 individuals randomly selected from the 1982 National Roster to par-

ticipate in the “MONICA” (“Monitoring trends and determinants in cardiovascular disease”) study of the World Health Organization [16]. The standardized international “MONICA” study does not include measurement of iron parameters. They were specifically added on the Icelandic study population for the purposes of this study. Half of the individuals in group A were selected from urban Reykjavik with a total population of 86,092 [lo]. The other half came from Arnessysla, a mostly rural county in the southern part of Iceland, with a total population of 10,188. The group A study population was between 25 and 74 years old, subdivided into 5 lo-year age subgroups of equal size and even sex distribution. Group A was investigated from June to September 1983. The second group studied (group B) consisted of a random sample from the National Roster of young adults living in Reykjavik and born in the years 1940, 1944, 1945, 1949, 1950 or 1954. A subsample of this group, 669 men and 571 women, were specifically investigated as to iron status from July 1984 to March 1985. Recruitment for both groups A and B was by written invitation to participate in a study of cardiovascular disease risk factors. Non-respondents were sent a repeat written invitation and finally contacted by telephone by the Heart Preventive Clinic staff. Blood samples were collected into Vacutainer@ (Becton Die kinson, Rutherford, NJ) tubes between 8.15 and 10.30 a.m. from fasting subjects (at least 10 h). Venous stasis was applied only in difficult cases and then for less than 2 min. Subjects sat in group A and laid supine in group B for 5 min before venipuncture. Bhemoglobin, B-red cell count (B-RBC) and Bmean cell volume (B-MCV) were analyzed on a Coulter S-Plus IV@ according to manufacturer recommendations [ 171 within 6 h of blood collection. Serum samples for S-iron, S-total iron binding capacity (S-TIBC) and S-ferritin were kept at -20°C until assayed in batches within several weeks of collection. S-iron [day-to-day coefficient of variation (C.V.) 4%] and S-TIBC (C.V. 5%) were measured on a centrifugal analyzer (Multistat III, Instrumentation Laboratory Inc.) using ferrozine as chromogen [18]. S-ferritin (C.V. 7% at 22 pg/l) was assayed with a radioimmunoassay kit, which uses antibodies against splenic ferritin (Amersham International plc, Amersham, U.K.). Except for patients with iron overload, no attempt was made by the investigators to define

Prevalence of Iron Deficiency and Overload in Adult Icelanders

causes of hematological or iron abnormalities. Results on individual study subjects were sent to their family physicians for further evaluation if necessary. Iron overload was defined as follows: S-TIBC saturation 260% and S-ferritin 2 380 pg/l in males and 2 220 pg/l in females OT S-TIBC saturation 2 50% and S-ferritin 2 600 pg/l in both sexes. The S-ferritin values of 380 pg/l in males and 220 pg/l in females are the upper reference limits at the National Hospital of Iceland Hematology laboratory. Patients who consistently fulfilled the iron overload criteria on repeated testing were further evaluated. A detailed history was obtained and physical examination performed. Further laboratory testing included B-reticulocytes, B-ESR, S-haptoglobin, S-total bilirubin, S-LDH, S-ASAT, S-ALP, S-y GT, S-HB, Ag, S-protein electrophoresis, P-PT, P-APTT, S-amylase, S-uric acid, S-creatinine, S-FSH, S-LH, S-TSH, S-Ts, S-T,, S-testosterone and an oral glucose tolerance test. Routine urine analysis was obtained and U-porphyrins screened for by fluorescence after ethyl acetate: acetic acid extraction. An ECG, chest X-ray and ultrasound investigation of the liver were also performed. Liver biopsies were obtained from patients with iron overload (using a Menghini 2mm i.d. needle) and evaluated histologically after hematoxylin-eosin, Perls, Masson trichrome and reticulin staining. Iron content was classified according to Scheuer et al. [20]. Quantitative weekly phlebotomies ( x 450 ml) were performed in patients with critical iron overload until B-hemoglobin fell below 110 g/l. Storage iron was calculated with a correction factor of 4mg/day due to iron absorption during the phlebotomy period. Patients considered to have definite HH after the aforementioned studies were typed for HLA class I A and B antigens using lymphocytotoxic assays. Iron studies on first

1291

degree relatives of HH patients were obtained when possible. INSULTS

Response rates are shown in Table 1. The mean ages of subjects studied in group A were 51 .O years for urban males, 51.7 years for rural males, 49.6 years for urban females and 50.6 years for rural females. The mean ages of group B subjects were 36.7 years for males and 37.0 years for females. Means and distributions of laboratory parameters relevant to iron status are shown in Table 2 for group A and in Table 3 for group B. Urban males had significantly decreased B-RBC, increased B-MCV, higher S-TIBC and lower S-ferritin than rural males. Urban females had a significantly lower B-hemoglobin, B-RBC, S-iron, S-TIBC saturation and S-ferritin than rural females. These differences in laboratory values could not be explained by the slightly higher mean age of the rural subjects. The overall differences in iron parameters indicate considerably larger iron stores in the rural population. If we assume that 1 pmol/l of S-ferritin represents 8 mg of storage iron [21], then the rural males had on average 344 mg and females 98 mg more iron stores than their urban counterparts. Urban and rural populations in group A demonstrated similar age related correlations in laboratory values. B-hemoglobin remained unchanged in males between 25 and 55 years of age but declined steadily in the older age groups, resulting in a 3 g/l difference in means between individuals of 55 and 75 years of age. B-RBC decreased with increasing age in males, especially between 55 and 75 years of age. The average decrease in B-RBC was 4 x 10’ cells/l/year over the entire age span. S-iron and S-IIBC decreased slightly with increasing age in males leaving S-TIBC saturation essentially

Table 1. Response rates classified according to group, residence and sex Group--Residence A-Urban A-Rural

Response t%l Males (n = 1500) 14.8 65.8

Investigated (%) 65.0 59.1

Males (n = 649)

B-All

urban

63.4

55.6

Response (%) Females (n = 1500) 15.5 13.3 Females (n = 571) 65.1

Investigated (%) 12.0 57.1 51.6

The “Response” column is the percentage of invited individuals examined at the Heart Preventive Clinic in Reykjavik in cardiovascular studies. The “Investigated” column is the percentage of invited individuals in whom B-hemoglobin, B-RBC, B-MCV, S-iron, S-TIBC and S-ferritin were obtained. The percentage figures are corrected for mortality and migration between selection of the study group and execution of the study.

JON J. JONSSONef al.

1292

Table 2. Means, medians and distributions of laboratory parameters relevant to iron status in Urban Parameter Males B-hemoglobin (g/l) B-RBC (lO’*/l) B-MCV (fl) * S-iron @mol/l) S-TIBC @mol/l) S-TIBC saturation (%) S-ferritin &g/l) Females B-hemoglobin (g/l) B-RBC (lO’2/l) B-MCV (fl) S-iron @mol/l) S-TIBC @mol/l) S-TIBC saturation (%) S-ferritin 01g/l)

group A

Rural

Mean

Median

95% Range

Mean

Median

95% Range

151.0 4.81 92.2 18.1 55.1 33.7 126

151 4.88 92.1 18 54 33.3 135

133-169 4.15-5.56 84.5-101.8 8-32 39-15 14-62 21-600

151.9 4.94 91.1 18.1 53.8 34.3 169

152 4.92 91.0 18 54 33.3 170

134170 4.28-5.56 84.1-97.9 7-31 39-12 ltil 31-760

134.3 4.35 91.8 16.7 56.5 30.4 52.1

135 4.35 91.8 16 51 29.2 50

112-154 3.614.96 81.5-101.0 6-32 39-8 1 9-62 9-318

137.2 4.44 91.4 17.8 56.5 32.2 64.3

137 4.44

‘12G157 3.18-5.10 84.c99.7 7-33 39-78 11-61 13-520

91.4

17 54 31.7 62

p < 0.05 No

Yes Yes No

Yes No

Yes Yes Yes No Yes No

Yes Yes

Means are arithmetic means except for S-ferritin, where geometric means are used due to the positively skewed distribution of values. Arithmetic means for S-ferritin @mol/l) were 171 in urban males, 227 in rural males, 83 in urban females and 100 in rural females. The 95% distribution intervals were determined with the percentile row (sample order statistic) method [19]. The last column (p < 0.05) indicates whether the difference between the means of rural and urban groups is statistically significant at the 5% level using a two-sided t-test for independent samples.

unchanged. S-ferritin increased steadily with increasing age of males on average 2.2 pg/l/year (standard error of coefficient = 0.38). B-hemoglobin means were 132.9 g/l in premenopausal women ( ~45 years of age) and 137.9 g/l in postmenopausal women ( > 50 years of age). The B-RBC averages were correspondingly 4.3 1 x lOI cells/l in premenopausal women vs 4.47 x lOI cells/l in postmenopausal women. The B-MCV was larger in postmenopausal women, 91.9 fl or 0.8 fl above the premenopausal value. S-iron decreased slightly with increasing age in premenopausal women, on average 0.23 pmol/l/year, but the S-TIBC was unchanged. The means for S-iron and S-TIBC were 0.2 and 4.1 pmol/l, respectively, lower in postmenopausal than premenopausal women. S-ferritin decreased 0.8 pgg/l/year in premenopausal women, but this was not statistically significant. In contrast, S-ferritin increased sharply in postmenopausal women, on average 3.9 pg/l/year (standard error of coefficient = 0.89).

Anemia was very rare in males but more frequent among women (Table 4). The higher prevalence of anemia in urban compared to rural women was statistically significant (p < 0.05). The prevalence of anemia in group A was 6.3% in premenopausal women compared to 1.9% in postmenopausal women. Urban premenopausal women in both groups A and B had an overall 10.7% prevalence of anemia. The prevalence of iron deficiency is shown in Table 5. Definite iron deficiency anemia was defined as anemia in conjunction with both S-TIBC saturation < 16% and S-ferritin < 12 pg/l[13,22,23]. The 7 women with definite iron deficiency anemia in group A were all ~45 years of age. Therefore, the overall prevalence of definite iron deficiency anemia in the 563 urban premenopausal women investigated (from both groups A and B) was 3.2%. Pregnant women were not excluded from the study. In group B, 7 women (2%) indicated that they

Table 3. Means, medians and distributions of laboratory parameters relevant to iron status in group B Females

Males Parameter

Mean

Median

95% Range

Mean

Median

95% Range

B-hemoglobin (g/l) B-RBC (1012/1) B-MCV (fl) S-iron @mol/l) S-TIBC @mol/l) S-TIBC saturation (%) S-ferritin @g/l)

150.9 4.91 91.0 17.9 47.4 39.2 125.2

151 4.96 90.8 17 48 36.1 136

133-168 4.33-5.59 84.5-98.8 5-35 30-66 l&92 19-537

131.2 4.33 90.4 14.8 50.3 30.9 30.4

132 4.31 90.6 15 51 29.6 30

110-151 3.12-4.95 80.1-98.9 3-32 30-75 571 4-180

See. footnote to Table 2. Arithmetic means for S-ferritin were 165 and 45 pmol/l for males and females, respectively.

Prevalence of Iron Deficiency and Overload in Adult Icelanders Table 4. Prevalence of anemia classified according to group, residence and sex Group-Residence

Males (%)

Females (%)

1.8 0.9 1.4

6.9 1.9 10.5

A-Urban A-Rural B-Urban

Anemia is defined according to WHO as B-hemoglobin < 130 g/l in males and < 120 g/l in females [311.

were or could be pregnant in prescreening for X-ray studies (as part of cardiovascular risk studies). None of these 7 women had definite iron deficiency anemia. Information on the pregnancy rate in group A premenopausal women was not available but was probably comparable to group B. The 2 males with definite iron deficiency anemia were 72 and 73 years old, respectively. The percentage of study subjects with biochemical evidence of iron deficiency, i.e. with S-TIBC saturation < 16% and/or S-ferritin < 12 pg/l varied from 3.6% in rural males in group A to 25.8% in urban premenopausal women in group B. Seven individuals fulfilled the criteria for iron overload on repeated testing (Table 6) and underwent further evaluation (Table 7); 5 of the patients were from Reykjavik and 2 from Arnessysla. The first patient was an asymptomatic 68-year-old female with a normal menstrual history and a mother of 2 children. Three first degree male relatives had normal iron parameters.

1293

The second patient, a 63-year-old male, had no specific symptoms suggestive of iron overload. He underwent a 65% partial resection of the stomach (Bilroth type I gastrectomy) 10 years prior to this study because of a benign peptic ulcer. The patient admitted to excessive alcohol intake in the past but it was reportedly currently negligible. Laboratory investigation revealed elevated S-yGT 93 U/l (reference limit ~50 U/l). He had four siblings, including a 60-year-old brother with S-TIBC saturation of 79% and an S-ferritin value of 461 cl/l. The 2 other brothers of similar age had normal S-TIBC saturations but increased S-ferritin values of 383 and 455 pg/l, respectively. A 62-yearold sister and a daughter had normal iron parameters. A 34-year-old son had normal S-TIBC saturation but an S-ferritin value of 646 ,ug/l. The third patient was a SZyear-old male complaining of lassitude. He admitted to moderate weekend alcohol drinking. Liver biopsy revealed grade 3 storage iron, moderate fatty changes and slight fibrosis in periportal areas. After investigating 5 of his 6 siblings we identified a 45-year-old brother with S-TIBC saturation of 67% and an S-ferritin value of 995 pg/l. The fourth patient was a 50-year-old male with polyarthritis affecting the second and third metacarpophalangeal joints most severely consistent with hemochromatosis associated

Table 5. Prevalence of iron deficiency in non-anemic and anemic subjects

Subgroup

No evidence for iron deficiency (%)

S-TIBC saturation <16% (%)

S-ferritin < 12 pg/l (%)

Definite iron deficiency (%)

1.0 94.5

0.6 3.3

0.0 0.4

0.2 0.0

0.9 95.5

0.0 3.4

0.0 0.2

0.0 0.0

4.1 83.6

1.3 8.4

0.2 0.9

1.3 0.2

1.2 92.0

0.7 5.9

0.0 0.0

0.0 0.2

Group A urban males (n = 479)

Anemic Non-anemic Group A rural males (n = 446)

Anemic Non-anemic Group A urban females (n = 535)

Anemic Non-anemic Group A rural females (n = 427) AllCUliC

Non-anemic Group B males (n = 372)

Anemic Non-anemic

0.3

0.0

0.0

9El

6.2

1.9

0.5

3.6 70.6

1.8 10.5

1.8 3.9

3.3 4.5

Group B females (n = 333)

Anemic Non-anemic

Anemia is de&d as B-hemoglobin c 130 g/I in males and < 120 g/l in females. The “No evidence for iron deficiency” column includes subjects with S-TIBC 2 16% and S-ferritin 2 12pg/l. The “S-TIBC saturation < 16%” column includes subjects with S-TIBC < 16% but S-ferritin > 12 pg/l. The “S-ferritin < 12 fig/l” column includes subjects with S-ferritin < 12 pg/l but S-TIBC saturation 2 16%. The “Detinite iron deficiency” column includes subjects with S-TIBC saturation < 16% and S-ferritin < 12 fig/l.

JONJ. JONSWNet al.

1294

Table6. Numberand percentageof patientswithiron overloadin group A Iron overload on initial screening Subgroup Urban Urban Rural Rural

males (n = 479) females (n = 535) males (FI= 446) females (n = 427)

Iron overload upon repeated testing

Number

%

Number

%

6 4 5 1

1.25 0.75 1.12 0.23

3 2 2 0

0.63 0.37 0.45 0.0

The criteria for iron overload were S-TIBC saturation 260% and S-ferritin 2 380 pg/l in males and a220 pg/l in females or S-TIBC saturation > 50% and S-ferritin 2600 pg/l. The 15 individuals with iron overload in the initial study were retested at a different time and the number and percentage of patients repeatedly fulfilling the aforementioned criteria are denoted in the second half of the table.

group A had slightly lower response rates than their urban counterparts. The mean B-hemoglobin and B-RBC values were similar to previously determined values for the adult Icelandic population [ 14,27,28, Heart Preventive Clinic unpublished observations] or published in the international literature [29]. Excluding individuals with evidence for iron deficiency (MCV < 80 fl, S-TIBC saturation < 16% or S-ferritin < 12 pgg/l) did not significantly change the B-hemoglobin means in any of the subgroups studied (data not shown). Not unexpectedly, our B-hemoglobin values were therefore similar to the results from the American Second National Health and Nutrition Examination Survey (NHANES II) [30], which were reported after eliminating individuals with laboratory values suggesting iron deficiency. The lower B-hemoglobin mean in the urban compared to rural women in group A could, however, in part, be explained by less iron stores in the urban population. The prevalence of anemia in the subgroups of adult Icelandic men 0.9-1.8% was in the i’ange observed for northwestern Europe [31] and slightly lower than found in the U.S.A. [32]. The 6.9 and 1.9% prevalence of anemia in urban

arthropathy [24]. He underwent Bilroth type II gastrectomy for a peptic ulcer 28 years prior to this study. All of his 4 siblings had normal iron parameters. Patients Nos 5 and 6 were asymptomatic and no underlying cause for iron overload was identified. Patient No. 7 was an alcoholic with mild hepatomegaly, grade 3 storage iron, mild fatty changes and minimal patchy portal tract fibrosis in the liver biopsy. The last 3 patients did not undergo phlebotomy treatment by the investigators. DISCUSSION Response rates were comparable to other epidemiological studies conducted by the Heart Preventive Clinic in Reykjavik using similar methods of recruitment [25,26]. Identifiable reasons for non-participation in previous studies at that institute have included difficulty in contacting the randomly selected individuals, or inability of the subjects to participate because of time constrains, hospitalization or institutionalization. Singles as a group have had lower response rates, as have individuals recently seen by a physician. In this study rural subjects in

Table 7. Overview of the seven patients with iron overload in group A Case No.

5 6 7

Age WI

Sex

S-TIBC saturation (%)

S-ferritin 01molil)

Liver iron grade

68 63 52 50

F M M M

82 85 77 84

900 1050 740 800

3 4 3 3

54 62 57

F M M

89 54 70

350 1080 615

3 2 3

Iron stores (g) 3.3 3.8 3.8 2.7

-

Clinical remarks Asymptomatic Gastrectomized Lassitude Arthritis, gastrectomized Asymptomatic Asymptomatic Alcoholic

HLA class A3, B7, B18 A3, B7, A25, B18 Al, B15, A2, Bl7 Aw19, B7, A25, B18 -

The S-TIBC saturation and S-ferritin values are averages from 3 independent samples. Liver iron grade was estimated according to the method of Scheuer et al. [20]. Iron stores were calculated after serial phlebotomy with a 4mg/day correction factor for iron absorption. Patients Nos 5-7 did not undergo serial phlebotomy by the investigators.

Prevalence of Iron Deficiency and Overload in Adult Icelanders

and rural women, respectively, in group A and 10.7% in urban premenopausal women in this study is consistent with observations on a few other populations [32,33] but considerably lower than the 13% average prevalence of anemia among women in developed countries reported by DeMaeyer and Adiels-Tegman [3 11. The S-ferritin means were higher than generally found for other populations [34-391 although the difference is small for premenopausal women. Our results are comparable to results from a study on elderly subjects in Sweden [40], where iron fortification of flour has been in effect for decades [33]. Interestingly similar S-ferritin values were observed in a recent study on middle-aged men in the Faroe Islands, a neighboring small community in the North Atlantic Ocean [41]. The S-TIBC saturation was also a few percentages higher than usually observed in Caucasian populations in developed countries [35,38,39,42]. Therefore, the three independent laboratory parameters of iron stores in a population, i.e. prevalence of anemia in women, S-ferritin and S-TIBC saturation, all indicated that adult Icelanders had unusually large iron stores. Probable explanations for this observation are the relatively high protein (16.4% of total caloric intake) and low carbohydrate (40.5%) content of the Icelandic diet [43] in combination with iron enrichment of imported flour and cereals. Comparisons between populations could however be influenced by subtle differences in selection of the study population or laboratory methodology. The reason(s) for larger iron stores in the rural compared to the urban population need to be studied. A dietary survey did not reveal any significant difference in food composition between residents in Reykjavik and Arnessysla [43]. We are also not aware of any difference in the preparation of food, e.g. use of iron skillets, which could explain larger iron stores in the rural populations. Consistent with the iron storage results we found iron deficiency to be rare except in urban premenopausal women. Among women in group B, 7.8% had definite iron deficiency, i.e. both S-TIBC saturation < 16% and S-ferritin < 12 pg/l. In the same group, 25.8% of the women had some evidence of iron deficiency, i.e. S-TIBC saturation < 16% and/or S-ferritin < 12 /&g/l. Four patients with definite HH and 3 very likely to have the disease were identified by examining 1887 individuals in group A. The prevalence estimate for HH in this population

1295

was 0.21% if only the 4 patients with definite diagnosis are counted but 0.37% if we include all 7 individuals that repeatedly fulfilled the iron overload criteria. The design of our study, with young adults from 25 years of age included as well as half of the subjects being women, tends to underestimate the prevalence of HH, because both of these groups have incomplete penetrance for the abnormal HH allele(s) using currently accepted clinical and biochemical criteria [2,44-461. Taking these characteristics of our study population into consideration, the prevalence of the hemochromatosis allele in the study population was probably higher than this study demonstrated. Our study on the Icelandic population extends the recent findings that HH is far more common in nations of Western European descent than previously thought [2-91. Variability in the hemochromatosis gene frequency in Western European populations has, however, been observed since at least three studies on Scandinavian populations have found a lower prevalence of the disease f47-491. HLA-typing was done on 4 hemochromatosis patients in this study. Three patients had both HLA B7 and B18, and 2 of these had HLA A3. This suggests that the A3, B7 haplotype generally found in increased frequency in hemochromatosis patients [50] could also be in linkage disequilibrium with the hemocrhomatosis allele in the Icelandic population. Except for 1 patient with arthritis our HH patients did not have clinical symptoms suggestive of HH. The gastrointestinal symptoms in the second patient probably resulted from the gastrectomy rather than a complication of HH [51]. The lack of clinical signs was to be expected, given the moderate accumulation of storage iron in our HH patients [2,52]. Interestingly, 2 of the iron overload patients were partially gastrectomized, which generally results in decreased storage iron [53]. We are aware of only one partially gastrectomized patient described previously in the literature with iron overload due to HH [54]. Apparently the hemochromatosis allele is penetrant despite partial gastrectomy. We feel that alcohol consumption could not explain the iron overload in patients Nos 2 and 3. Asserting homozygosity for the hemochromatosis allele in patient No. 7 was more uncertain in light of the controversial association between alcholism and iron overload [55]. The fatty change and minimal fibrosis in his liver were probably the consequence of alcohol toxicity.

JONJ. JONWIN

1296

Venesection therapy has been shown to improve symptoms and decrease mortality associated with HH [56-581. The finding of asymptomatic patients with significant iron overload in this study underscores the importance of being on the alert for this disease and raises the question whether biochemical screening of middle-aged and elderly males is worthwhile. Before that question can be definitely answered the natural history of HH, especially in its milder form, needs to be better defined. This paper examines the iron balance among adults in Iceland, an industrialized Scandinavian nation. We demonstrate that iron deficiency is not a public health problem in adult Icelanders, except for urban premenopausal women where approx. 1 in 4 shows evidence of iron deficiency and 3.2% have iron deficiency anemia. This is observed concurrently with a few individuals of the adult population accumulating critical levels of iron overload. Measures to improve the iron balance in young urban women cannot therefore include general measures like increased iron fortification of food but must be more directed towards the urban premenopausal women target group. Acknowledgements-This work was supported by the National Science Fund of Iceland. We would like to thank the staff at the Heart Preventive Clinic in Reykjavik and technologists in the clinical laboratories at the National Hospital of Iceland for their help in carrying out the study. The contribution of Drs R. Skuladottir and G. Williams at the National Hospital of Wales, Cardifl, U.K. in confirming the liver biopsy findings is acknowledged.

REFERENCES 1. Finch SC, Finch CA. Idiopathic hemochromatosis, an iron storage disease. A. Iron metabolism in hemochromatosis. Medicine (Baltimore) 1955; 34: 381-430. 2. Cartwright GE, Edwards CQ, Kravitz K, Skolnick M, Amos DB, Johnson A et al. Hereditary hemochromatosis. Phenotypic expression of the disease. N Engl J Med 1979; 301: 175-179. 3. Beaumont C, Simon M, Fauchet R, Hespei JP, Brissot P, Genetet B et al. Serum ferritin as a possible marker of the hemochromatosis allele. N Engl J Med 1979; 301: 169-174. 4. MacSween RNM, Scott AR. Hepatic cirrhosis: a clinico-pathological review of 520 cases. J ClIn Pathol 1973; 26: 936-942. 5. Olsson KS, Ritter B, Rosen U, Heedman PA, Staugaard F. Prevalence of iron overload in central Sweden. Acts Med Scnnd 1983; 213: 145-150. 6. Edwards CQ, Griffen LM, Goldgar D, Drummond C, Skolnick MH, Kushner JP. Prevalence of hemochromatosis among 11,065 presumably healthy blood donors. N Engl J Med 1988; 318: 1355-1362. 7. Borwein ST, Ghent CN, Flanagan PR, Chamberlain MJ, Valberg LS. Genetic and phenotypic expression of hemochromatosis in Canadians. Clin Invest Med 1983; 6: 171-179.

et al.

8. Ballot D, Meyer TE, Bothwell TH, Bezwoda WR, Green A, Baynes RD et al. Idiopathic haemochromatosis. Family studies and results of a pilot prevalence survey. Afr Med J 1987; 71: 639-642. Elliott R. Tait A. Lin BPC. Smith CI. Dent OF. Prevalence of hem&hromatos& in a random sample of asymptomatic men. Aust NZ J Med 1986; 16:491-495. Statistical Abstract of Iceland 1984. In Statistics of Iceland, Vol. II. Reykjavik, Iceland: Statistical Bureau of Iceland; 1984: 82. Baker SJ, DeMaeyer EM. Nutritional anemia: its understanding and control with special reference to the work of the World Health Organization. Am J CIin Nutr 1979; 32: 368-417. Jacobs A. Iron deficiency and iron overload. CRC Rev Hematol One01 1985; 3: 143-183. Cook JD, Finch CA. Assessing iron status of a population. Am J Clin Nutr 1979; 32: 2115-2119. Davidsson D, Sigfusson N, Olafsson 0, Bjomsson OJ, Thorsteinsson Th. Hemoglobin, hematokrit, MCHC og sokk. (Hemoglobin, hematocrit, MCHC, and ESR.) In Health Survey in the Reykjavik Area 1%7-‘68. Report A IV. Reykjavik, Iceland: Icelandic Heart Association; 1976. Steingrimsdottir L. Born og jam. (Children and iron). Heilbrigdismal 1985; 2: 29-32. WHO MONICA Project Principal Investigators. The World Health Organization MONICA project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. J ClIn Epidemiol 1988; 41: 105-114. Coulter Counter Model S-Plus IV with Data Terminal. Product Reference Manual No. 4235153D. Hialeah, FL; 1982. Persijn JP, Van der Slik W, Riethorst A. Determination of serum iron and latent iron-binding capacity (LIBC). Clin Cbim Acta 1971; 35: 91-98. Shultz EK, Willard KE, Rich SS, Connelly DP, Critchfield GC. Improved reference-interval estimation. Clin Cbem 1985; 31: 1974-1978. Scheuer PJ, Williams R, Muir AR. Hepatic pathology in relatives of patients with haemochromatosis. J Pathol Bacterlol 1962; 84: 53-64. Walters GO, Miller FM, Worwood M. Serum ferritin concentration and iron stores in normal subjects. J Clln Pathol 1973; 26: 770-772. Bainton DF, Finch CA. The diagnosis of iron deficiency anemia. Am J Med 1964; 37: 62-70. Hercberg S, Galan P. Assessment of iron deficiency in uouulations. Rev Eoidemiol Sante Publiaue 1985: 33:

s

9.

10.

11.

12. 13. 14.

15. 16.

17.

18. 19.

20. 21.

22. 23.

i28-239.

Hamilton EBD, Bomford AB, Laws JW, Williams R. The natural history of arthritis in idiopathic haemochromatosis: progression of the clinical and radiological features over ten years. Q J Med 1981; 50: 321-329. 25. Bjornsson OJ, Davidsson D, Olafsson H, Olafsson 0, Sigfusson N, Thorsteinsson Th. Participants, invitation, response etc. In Health Survey in tbe Reykjavik Area-Men. Stages I-III, 1967-1969, 1970-1971 and 1974-1976. Report ABC XVIII. Reykjavik, Iceland: Icelandic Heart Association; 1979. 26. Bjornsson G, Bjomsson OJ, Davidsson D et al. Participants, invitation, response etc. In Health Survey h the Reykjavik Area-Women. Stages I-III, l%f&1%9, 1971-1972 and 1976-1978. Report ABC XXIV. Reykjavik, Iceland: Icelandic Heart Association; 1982. 27. Bjornsson OJ, Davidsson D, Filippusson H, Olafsson 0, Sigfusson N, Thorsteinsson Th. Distribution of haematological, serum and urine values in a general population of middle-aged men. The Reykjavik Study. Seand J Soe Med 1984; Suppl 32: 3-12. 24.

Prevalence of Iron Deficiency and Overload in Adult Icelanders V. Algengismork nokkurra blod28. Thorsteinsson rannsokna. (Reference values for several blood tests.) Ic&ndic Med J 1982; Suppl 13: 52-59. 29. Elwood PC, Hughes J, Abernethy R, Davies R, Gough R, Johnson AP et al. An international haematological survey. BulI WHO 1976; 54: 87-95. 30. Yip R, Johnson C, Dallman PR. Age-related changes in laboratory values used in the diagnosis of anemia and iron deficiency. Am J Clht Nutr 1984; 39: 427-436. 31. DeMaeyer E, Adiels-Tegman M. The prevalence of anemia in the world. World Health Stat Q 1985; 38: 302-316. 32. Dalhnan PR, Yip R, Johnson C. Prevalence and causes of anemia in the United States, 1976 to 1980. Am J Clin Nutr 1984: 39: 437-445. 33. Hallberg L, &ngtsson C, Garby L, Lennartsson J, Rossander L, Tibblin E. An analysis of factors leading to a reduction in iron deficiency in Swedish women. Bull WHO 1979; 57: 947-954. 34. Wonvood M. Serum ferritin. CRC Crit Rev Clio Lab Sci 1979; 10: 171-204. 35. Valberg LS, Sorbie J, Ludwig J, Pelletier 0. Serum ferritin and the iron status of Canadians. Can Med J 1976; 114: 417-421. 36. Milman N, Strandberg Pedersen N, Visfeldt J. Serum ferritin in healthy Danes: relation to marrow haemosiderin iron stores. Dan Med Bull 1983; 30: 115-120. 37. Milman N, Andersen HC, Strandberg Pedersen N. Serum ferritin and iron status in ‘healthy’ elderly individuals. Sand J Clia Lab Invest 1986; 46: 19-26. 38. Cook JD, Skikne BS, Lynch SR, Reusser ME. Estimates of iron su8iciency in the US population. Blood 1986;68: 726-731. 39. Cook JD, Finch CA, Smith NJ. Evaluation of the iron status of a population. Blood 1976; 48: 449-455. 40. Qvist I, Norden A, Olofsson T. Serum ferritin in the elderly. Stand J CIht Lab Invest 1980; 40: 609-613. 41. Mihnan N, Thomsen H, Mathiassen B. Serum ferritin, iron status and plasma ascorbic acid in 40- to 49-yearold males in the Fame Islands. ScantI J Clht Lab Invest 1990; 50: 559-564. 42. Vellar OD, Sondena ER, Bay IG. Hematologiske verdier i tilfeldige utvalg av den voksne befolkning. (Hematological values in randomly selected adults.) Norw Med J 1977; 12: 549-551. 43. Sigurdsson G, Einarsdottir S, Sveinsdottir E. Islenskt mataraedi. (Icelandic diet.) Icelandic Med J 1976; 62: 87-95.

1297

44. Dadone MM, Kushner JP, Edwards CQ, Bishop DT, Skolnick MH. Hereditary hemochromatosis. Analysis of laboratory expression of the disease by genotype in 18 pedigrees. Am J CBn PathoI 1982; 78: 196-207. 45. Borwein S, Ghent CN, Valberg LS. Diagnostic e5cacy of screening tests for hereditary hemochromatosis. Can Med Assne J 1984; 131: 8955901. 46. Bassett ML, Halliday JW, Ferris RA, Powell LW. Diagnosis of hemochromatosis in young subjects: predictive accuracy of biochemical screening tests. Gastmenterology 1984; 87: 628-633. 47. Lindmark B, Eriksson S. Regional differences in the idiopathic hemochromatosis gene frequency in Sweden. Acta Med Scmd 1985; 218: 299-304. 48. Karlsson M, Ikkala E, Reunanen A, Takkunen H, Vuori E, Makinen J. Prevalence of hemochromatosis in Finland. Acta Med !&ml 1988; 224: 385-390. 49. Hallberg L, Bjom-Rasmussen E, Jungner I. Prevalence of hereditary haemochromatosis in two Swedish urban areas. J M Med 1989: 225: 249-255. 50. Simon M, Brissot P. The genetics of haemochromatosis. J Hepatol 1988; 6: 116-124. 51. Jansen PLM. Thien Th. Lamers CBHW. Yan SH. Reekers P, Strijk S. Exo&ine pancreatic insufh~iency and idiopathic haemochromatosis. Post@ Med J 1984; 60: 675-678. 52. Valberg LS, Lloyd DA, Ghent CN Flanagan PR, Sinclair NR, Stiller CR et al. Clinical and biochemical expression of the genetic abnormality in idiopathic hemochromatosis. Gastroenterology 1980; 79: 884-892. 53. Lloyd DA, Valberg LS. Serum ferritin and body iron status after gastric operations. Am J Dig Dls 1977; 22: 598-604. 54. Parfrey PS, Tumbull AL, Pollock DJ, Bamett M. Iron overload despite partial gastrectomy. P&grad Med J 1979; 55: 747-750. 55. Irving MG, Halliday JW, Powell LW. Association between alcholism and increased hepatic iron stores. Alcohol Clin Exp Res 1988; 12: 7-13. 56. Niederau C, Fischer R, Sonnenberg A, Stremmel W, Tramp&h HJ, Strohmeyer G. Survival and causes of death in cirrhotic and in noncirrhotic patients with primary hemochromatosis. N Engl J Med 1985; 313: 1256-1262. 57. Bomford A, Williams R. Long term results of venesection therapy in idiopathic haemochromatosis. Q J Med 1976; 45: 61 l-623. 58. Conte D, Pipemo A, Mandelli C. Fargion S, Cessna M, Brunelli L et al. Clinical, biochemical and histological features of primary haemochromatosis: a reportof 67 cases. Liver 1986: 6: 310-315.