HFE C282Y Homozygosity Is Associated with an Increased Risk of Total Hip Replacement for Osteoarthritis

OSTEOARTHRITIS

HFE C282Y Homozygosity Is Associated with an Increased Risk of Total Hip Replacement for Osteoarthritis Yuanyuan Wang, MD, PhD,*,1 Lyle C. Gurrin, PhD,†,1 Anita E. Wluka, FRACP, PhD,* Nadine A. Bertalli, GDipBioEpi,† Nicholas J. Osborne, PhD,†,‡,§,¶ Martin B. Delatycki, MD, PhD,‡,§,储 Graham G. Giles, PhD,†,** Dallas R. English, PhD,†,** John L. Hopper, PhD,† Julie A. Simpson, PhD,†,** Stephen Graves, FRACS, PhD,††,‡‡ Katrina J. Allen, MD, PhD,‡,§,§§ and Flavia M. Cicuttini, FRACP, PhD* Objective: The evidence for an association between mutations in the HFE (hemochromatosis) gene and the risk of hip or knee osteoarthritis is inconsistent. Total joint replacement is considered a surrogate measure for symptomatic end-stage osteoarthritis. We examined the relationship between HFE gene mutations and risk of total hip and knee replacement using a prospective cohort study. Methods: The Melbourne Collaborative Cohort Study recruited participants between 1990 and 1994. Participants born in Australia, New Zealand, the United Kingdom, or Ireland (n ⫽ 27,848) were genotyped for the HFE C282Y mutation. Total hip and knee replacements for osteoarthritis during 2001 to 2009 were ascertained from the Australian Orthopaedic Association National Joint Replacement Registry. Hazard ratios (HR)/odds ratios (OR) and confidence intervals (CI) were obtained from Cox regression or logistic regression. Results: Compared with those with no C282Y mutation, C282Y homozygotes had an increased risk of single total hip replacement (HR 1.94, 95% CI 1.04-3.62) and bilateral total hip replacement (OR 5.86, 95% CI 2.36-14.57) for osteoarthritis, adjusting for age, sex, body mass index, and educational level. Only 3 C282Y homozygotes had single total knee replacement; the HR was 0.51 (95% CI 0.16-1.57). C282Y/H63D compound heterozygosity was not related to the risk of total hip or knee replacement. Conclusions: HFE C282Y homozygosity was associated with an increased risk of both single and bilateral total hip replacement for osteoarthritis. Crown Copyright © 2012 Published by Elsevier Inc. All rights reserved. Semin Arthritis Rheum 41:872-878

*Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, Alfred Hospital, Melbourne, Victoria, Australia. †Centre for Molecular, Environmental, Genetic and Analytic Epidemiology, School of Population Health, University of Melbourne, Carlton, Victoria, Australia. ‡Murdoch Children’s Research Institute, Melbourne, Victoria, Australia. §Department of Pediatrics, University of Melbourne, Melbourne, Victoria, Australia. ¶European Centre for Environment and Human Health, Peninsula College of Medicine and Density, University of Exeter, Exeter, UK. 储Department of Clinical Genetics, Austin Health, Heidelberg, Victoria, Australia. **Cancer Epidemiology Centre, Cancer Council Victoria, Carlton, Victoria, Australia. ††Department of Orthopaedic, Repatriation General Hospital, Daw Park, South Australia, Australia. ‡‡AOA National Joint Replacement Registry; Discipline of Public Health, School of Population Health and Clinical Practice, University of Adelaide, Adelaide, South Australia, Australia. §§Department of Gastroenterology, The Royal Children’s Hospital, Melbourne, Victoria, Australia. 1These authors contributed equally to this article. The Melbourne Collaborative Cohort Study recruitment was funded by VicHealth and Cancer Council Victoria. This study was funded by a program grant from the National Health and Medical Research Council (NHMRC; 209057), capacity building grant (251533), and enabling grant (396414) and was further supported by infrastructure provided by The Cancer Council of Victoria. Dr Wang is the recipient of NHMRC Public Health (Australia) Fellowship (NHMRC 465142). Dr Wluka is the recipient of NHMRC Career Development Award (NHMRC 545876). Dr Delatycki is the recipient of a NHMRC Practitioner Fellowship (NHMRC 546452). The authors have no conflicts of interest to disclose. Address reprint requests to Flavia Cicuttini, FRACP, PhD, Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, Alfred Hospital, Melbourne, VIC 3004, Australia. E-mail: [email protected].

872

0049-0172/12/$-see front matter Crown Copyright © 2012 Published by Elsevier Inc. All rights reserved. doi:10.1016/j.semarthrit.2011.11.003

Y. Wang et al.

873

Keywords: hemochromatosis, osteoarthritis, total joint replacement, HFE, C282Y

H

ereditary hemochromatosis is an autosomal-recessive disorder of iron metabolism, predisposing an individual to inappropriately increased intestinal iron absorption. This can lead to progressive total body iron overload with secondary tissue damage in various organs such as the liver, heart, pancreas, and joints, promoting diseases such as fibrosis and cirrhosis of the liver (and ultimately hepatocellular carcinoma), arthritis, and diabetes (1). The C282Y and H36D mutations in the HFE gene are responsible for the majority of clinical cases of iron overload (2). Most (80-90%) patients with clinical hemochromatosis are homozygous for the C282Y mutation (2 copies of the C282Y mutation) and approximately 5% are compound heterozygotes, with 1 copy of each of the C282Y and H63D mutations (1,3-5). For populations of northern European descent, the prevalence of C282Y homozygosity is estimated to be 1 in 200 individuals (0.5%), with 2.4% being C282Y/H63D compound heterozygotes (3,6). Attempts have been made to characterize the progression of hereditary hemochromatosis into stages, for example, a genetic predisposition without abnormalities, iron overload without symptoms, iron overload with symptoms, and iron overload with organ damage (7,8). Those genetically predisposed to iron overload may not experience symptoms of disease. Recent results from a cohort study showed that at least 28% of male C282Y homozygotes compared with only 1% of female C282Y homozygotes developed iron overload-related disease (8). Studies have shown that patients with hereditary hemochromatosis have an increased risk of osteoarthritis (OA) and joint replacement surgery for severe OA compared with age- and sex-matched controls (9,10) or healthy subjects from a population-based cohort (11). A cross-sectional study reported significantly more frequent bilateral large joint arthropathy as defined by Kellgren and Lawrence grade ⱖ2 for persons with definite/probable hemochromatosis than for those with possible/unlikely hemochromatosis (12). There is evidence suggesting that iron load is a determinant of arthropathy in hemochromatosis (12,13). The evidence for an association between HFE C282Y and H63D mutations and the risk of arthritis is conflicting. Although some studies reported significant associations between any numbers of copies of the C282Y mutation (C282Y homozygous or heterozygous) and the risk of arthritis (8,14,15), others found no relationships (1619). A recent meta-analysis reported only weak evidence of an association between HFE genotype and arthritis risk (20). Most studies on the HFE gene mutations have been performed on clinical samples or selected populations, making it difficult to generalize the findings. Moreover, these studies have tended to concentrate on associations

between the risk of arthritis and the presence of any HFE mutation, or the presence/absence of the C282Y mutation, rather than examining the risk of disease for C282Y homozygotes, who are essentially the only ones at risk of iron overload-related disease. Using a prospective cohort study, we assessed the relationships between the C282Y mutation of the HFE gene and the risk of primary total hip and knee replacement for OA, a surrogate measure for symptomatic end-stage OA and clinically significant endpoint of OA. Given the recent data for an association between hemochromatosis and bilateral arthropathy (12), this study also examined the association of the C282Y mutation of the HFE gene with bilateral total hip and knee replacement for OA. PATIENTS AND METHODS Melbourne Collaborative Cohort Study (MCCS) From 1990 to 1994, 41,514 participants (24,469 women) aged between 27 and 75 years (99.3% were 40-69 years) were enrolled in the MCCS in Melbourne, Australia (21). Participants were recruited through the Australian Electoral Rolls (voting is compulsory for Australian citizens), advertisements, and community announcements in local media. At baseline, participants completed a questionnaire on potential risk factors including country of birth, education, smoking history, alcohol consumption, and physical activity, had weight and height measured [allowing body mass index (BMI) to be calculated], and provided a fasting blood sample. From 2003 onward, 28,046 study participants (68% of the original MCCS participants) attended for a follow-up visit. The participants were asked questions enquiring about their first joint replacement surgery: have you ever had a hip replacement? When did you Have your first hip replacement? Have you ever had a knee replacement? When did you have your first knee replacement? The purpose of doing this was to obtain the information necessary to exclude from subsequent analysis those who had had a total hip or knee replacement before 2001, so that we would be examining the incidence of total hip and knee replacement solely during the study period (20012009). Self-reported total joint replacement was not used to assess the incidence. Ascertainment of Incident Primary Knee and Hip Joint Replacement Participants with incident total joint replacement were identified from the Australian Orthopaedic Association (AOA) National Joint Replacement Registry (NJRR). The AOA NJRR commenced in 1999 and was introduced in a staged state-by-state approach, which was com-

874

pleted nationally by mid-2002. Victorian data collection commenced in 2001. The Registry monitors the performance and outcome of both hip and knee replacement surgery in Australia. It has detailed information on the prostheses and surgical technique used and the clinical situation that it was used in for both primary and revision joint replacement (22). By using detailed matching technology, it is able to determine the success or otherwise of the joint replacement surgery. Although data collection for the Registry is voluntary, it receives cooperation from all hospitals undertaking joint replacement surgery (22). The AOA NJRR validates its data by using both internal systems and external data sources. The most important external data source is state Health Department data. Validation of registry data against Health Department recorded data involves a sequential multilevel matching process. Initial validation results in over 93% of Registry records verified against health department data. Following the validation process, the retrieval of unreported records and checking of unmatched data, the Registry is able to obtain an almost complete set of data relating to hip and knee replacement in Australia (23). Identifying information for MCCS participants, including first name, last name, date of birth, and sex, was provided to the AOA NJRR to identify those who had had a primary or revision joint replacement between January 1, 2001, which is when the Registry commenced Victorian data collection, and December 31, 2009. The matching was performed on these data provided using U.S. Bureau of the Census Record Linkage Software. Assessment of HFE Genotype Beginning in 2004, 31,192 MCCS participants born in Australia, New Zealand, the United Kingdom, or Ireland (ie, nominally of northern European descent) were genotyped for the C282Y mutation in the HFE gene, using their stored baseline blood samples. MCCS participants born in southern Europe (Italy, Greece, or Malta) were not genotyped due to the lower prevalence of HFE mutations in these populations (24). Blood samples were either stored in liquid nitrogen or dried on Guthrie cards. DNA was extracted from Guthrie cards by the Chelex method and from buffy coats using a guanidinium isothiocyanate-based method (Corbett Buffy Coat CorProtocol 14102). All samples were genotyped for the single nucleotide polymorphism in the HFE gene that is responsible for the C282Y substitution in the HFE protein (rs1800562) using real-time polymerase chain reaction. Those with 1 copy of the C282Y mutation were also genotyped for the H63D mutation. Therefore, there were 4 HFE genotype groups: (1) C282Y homozygotes (2 copies of the C282Y mutation), (2) simple heterozygotes (1 copy of the C282Y mutation and no copies of the H63D mutation), (3) compound heterozygotes (1 copy each of the C282Y and H63D mutations),

HFE C282Y homozygosity associated risk of total hip replacement

and (4) other HFE genotype (no copies of the C282Y mutation and an unknown number of copies of the H63D mutation). All C282Y homozygotes plus a sample stratified by HFE genotype (n ⫽ 1438) were selected for invitation to the “HealthIron” study of iron and health from 2004 to 2007. Additional confirmatory genotyping was performed for those selected for invitation to the HealthIron participants from cheek swab samples (for those who participated in HealthIron) or from baseline plasma samples (for those who did not participate in HealthIron). For this analysis, C282Y homozygotes are those participants classified as C282Y homozygotes by the initial and confirmatory genotyping (where they agreed); otherwise, they were classified according to the results of the confirmatory genotyping (if initial and confirmatory genotyping disagreed). Further details of the genotyping were described previously (8). The study protocol was approved by The Cancer Council Victoria’s Human Research Ethics Committee and the Standing Committee on Ethics in Research Involving Humans of Monash University. All participants gave written consent for participation and for the investigators to obtain their medical records. Exclusion Criteria The incidence of total joint replacement was ascertained from 1 January 2001. To account for this in the analyses, we excluded participants (n ⫽ 1610) who had died or left Australia before 1 January 2001; at the follow-up visit had reported a primary joint replacement before 1 January 2001; had left Australia before the date of having a primary joint replacement; or had the first recorded procedure being a revision joint replacement as recorded in the AOA NJRR. This left 29,582 (95%) eligible participants. We further excluded 1734 participants who could not be genotyped (either because of missing baseline blood sample or because of insufficient DNA for genotyping) (Fig. 1). Statistical Analysis Hazard ratios (HR) for the association between HFE genotype and the risk of single total joint replacement were estimated using Cox regression with age as the time axis. Follow-up for primary total joint replacement began at 1 January 2001 and ended at the date of first primary joint replacement for OA or date of censoring. Participants were censored at either the date of first primary joint replacement performed for indications other than OA, the date of death, the date left Australia, or the end of follow-up (31 December 2009, the date that ascertainment of joint replacement by NJRR was complete), whichever came first. Odds ratios (OR) for the association between HFE genotype and the risk of bilateral total joint replacement were estimated using logistic regression.

Y. Wang et al.

875

RESULTS

Melbourne Collaborative Cohort Study, baseline 1990-4, n = 41,514

Demographic, lifestyle factors, and anthropometric data; Blood sample

Participants born in southern Europe (Italy, Greece, or Malta) were excluded, n = 10,322

Born in Australia, New Zealand, the United Kingdom or Ireland, n = 31,192

From 2004, genotyped for the C282Y mutation in the HFE gene, using stored baseline blood samples collected in 1990-4

Data linkage to Joint Registry, 2001-9

27,848 eligible participants for analysis

3,344 excluded due to died or left Australia prior to 2001, n = 834 self-reported joint replacement prior to 2001, n = 609 left Australia before the date of having a primary joint replacement, n = 5 first recorded procedure being a revision joint replacement, n = 162 could not be genotyped, n = 1,734

814 hip replacements and 966 knee replacements identified during 2001-9

Figure 1 Study profile.

The primary exposure variables were combinations of HFE C282Y and H63D mutations. Several variables were assessed as potential confounders of the association between HFE genotype and the risk of total joint replacement, including age (for logistic regression; adjustment by age occurs by definition in Cox regression because age was used as the time scale), sex, BMI, smoking, alcohol consumption, physical activity, and education level. A change of any estimated HR of 5% or more after inclusion of a potential confounding variable in the statistical model was considered to be indicative of confounding. The final HRs were adjusted for age, sex, BMI, and education level. To test whether associations between HFE genotype and joint replacement risk were modified by sex, interactions between sex and HFE genotype were fitted and tested using the likelihood ratio test. All statistical analyses were performed using Stata (Intercooled Stata 10.1 for Windows, Stata Corp LP, College Station, TX).

Of the 29,582 eligible participants of northern European descent, 27,848 (94%) were successfully genotyped for C282Y mutation in HFE gene. Of these, 184 (0.7%) were C282Y homozygotes, 677 (2.4%) were compound heterozygotes, 3194 (11.5%) were simple C282Y heterozygotes, and 23,793 (85.4%) were wild-type for C282Y and H63D. Table 1 displays the summary statistics of baseline characteristics by genotype group. There was little difference between the genotype groups for mean age at 1 January 2001, smoking status, daily alcohol consumption (g/d), and education level. Of the 4 genotype groups, C282Y homozygotes had the lowest BMI and a lower proportion of female participants. From 1 January 2001 to 31 December 2009, with an average of 8.4 (SD 1.8) years of follow-up per person, a total of 966 total knee replacements (mean age at surgery, 71.0; SD 7.7; range 48.9-86.6 years) and 814 total hip replacements (mean age at surgery, 70.9; SD 7.7; range 47.8-86.6 years) performed for OA were identified, giving crude incidence rates of 4.2 and 3.5 per 1000 personyears, respectively. We identified 135 participants with bilateral total hip replacement for OA and 226 participants with bilateral total knee replacement for OA. Table 2 presents the associations between the C282Y and H63D mutations of the HFE gene and the risk of single total hip and knee replacement for OA. C282Y homozygotes had an increased risk of total hip replacement for OA compared with those with no C282Y mutation [HR 1.94, 95% confidence interval (CI) 1.043.62]. C282Y homozygosity was not associated with increased risk of total knee replacement. HRs for C282Y simple and C282Y/H63D compound heterozygotes were close to unity for the risk of total hip or knee replacement. Adjustment for sex, BMI, and educational level had little influence on the unadjusted HRs and 95% CIs. Table 3 presents the associations between the C282Y and H63D mutations of the HFE gene and the risk of bilateral total hip and knee replacement for OA. Although no C282Y homozygotes in the study had bilateral total knee replacement for OA, 5 C282Y homozygotes had bilateral total hip replacement for OA. C282Y homozygotes had a nearly 6 times risk of bilateral total hip replacement compared with those with no C282Y mutation (OR 5.86, 95% CI 2.36-14.57). C282Y heterozygosity (simple and compound) was not associated with the risk of bilateral total hip or knee replacement for OA. There was no evidence for the effect modification of sex on associations between HFE genotype and risk of total hip or knee replacement (P values for the likelihood ratio test ⬎0.20). DISCUSSION We examined the association between HFE genotype and the risk of total hip and knee replacement for OA in a

876

HFE C282Y homozygosity associated risk of total hip replacement

Table 1 Summary of Cohort Characteristics at Baseline C282Y Homozygote

Compound Heterozygote

184 102 (55) 62.6 ⫾ 8.9 25.6 ⫾ 3.7

677 416 (61) 62.3 ⫾ 9.0 26.4 ⫾ 4.6

3194 1955 (61) 62.1 ⫾ 9.1 26.1 ⫾ 4.1

23793 14467 (61) 62.1 ⫾ 9.0 26.2 ⫾ 4.3

99 (54) 65 (35) 20 (11)

398 (59) 216 (32) 63 (9)

1865 (58) 999 (31) 330 (10)

13728 (58) 7633 (32) 2430 (10)

39 (21) 127 (69) 14 (8) 4 (2)

191 (28) 399 (59) 59 (9) 27 (4)

880 (28) 1866 (58) 308 (10) 138 (4)

6661 (28) 13829 (58) 2236 (9) 1053 (4)

5 (3) 79 (43) 50 (27) 50 (27) 4.5 ⫾ 3.8

25 (4) 281 (41) 150 (22) 221 (33) 4.6 ⫾ 3.8

116 (4) 1438 (45) 746 (23) 892 (28) 4.5 ⫾ 3.8

863 (4) 10488 (44) 5633 (24) 6800 (29) 4.5 ⫾ 3.7

Number Female, n (%) Age entering JR cohort, yr Body mass index, kg/m2 Smoking Never Former Current Alcohol None Light Moderate Heavy Education Primary school Some high school Completed high school Degree/diploma Physical activity score

C282Y Heterozygote No C282Y Mutation

Values presented as mean ⫾ SD or number (%). JR, joint replacement. None: 0 g/d; light: 1 to 39 g/d (men), 1 to 19 g/d (women); moderate: 40 to 59 g/d (men), 20 to 39 g/d (women); heavy: 60⫹ g/d (men), 40⫹ g/d (women).

prospective cohort of healthy, middle-aged Australians over an 8-year period. We found that HFE C282Y homozygotes had an increased risk of both single and bilateral total hip replacement for OA compared with those

who had no C282Y mutation. There was no evidence of an increased risk of total knee replacement for OA in C282Y homozygotes. Data from C282Y simple heterozygotes or C282Y/H63D compound heterozygotes also

Table 2 Associations Between HFE Genotype Group and the Risk of Single Primary Total Hip and Knee Joint Replacement for Osteoarthritis Allele

C282Y Copies

Total hip replacement Codominant

Recessive With H63D

0 1 2 Trend 0 or 1 2 0 1 (no H63D) 1 (and 1 H63D) 2 Trend

Total knee replacement Codominant

Recessive With H63D

aAdjusting

0 1 2 0 or 1 2 0 1 (no H63D) 1 (and 1 H63D) 2

Cases Person-years Hazard Ratio (95% CI) Hazard Ratio (95% CI)a P Valuea 687 117 10

198,405 32,092 1505

804 10 687 95 22 10

230,497 1505 198,405 26,509 5584 1505

818 145 3 963 3 818 118 27 3

198,405 32,092 1505 230,497 1505 198,405 26,509 5584 1505

for sex, body mass index, and educational level.

ref 1.05 (0.87-1.28) 1.93 (1.03-3.60) 1.12 (0.94-1.34) ref 1.91 (1.02-3.57) ref 1.04 (0.84-1.28) 1.14 (0.75-1.74) 1.93 (1.03-3.60) 1.11 (0.97-1.26)

ref 1.06 (0.87-1.29) 1.94 (1.04-3.62) 1.13 (0.95-1.34) ref 1.92 (1.03-3.59) ref 1.05 (0.84-1.30) 1.11 (0.73-1.70) 1.94 (1.04-3.62) 1.11 (0.97-1.26)

ref 1.10 (0.92-1.31) 0.49 (0.16-1.51) ref 0.48 (0.15-1.49) ref 1.08 (0.89-1.31) 1.17 (0.80-1.72) 0.49 (0.16-1.51)

ref 1.10 (0.92-1.31) 0.51 (0.16-1.57) ref 0.50 (0.16-1.55) ref 1.09 (0.90-1.32) 1.15 (0.79-1.69) 0.51 (0.16-1.57)

0.57 0.04 0.18 0.04 0.67 0.62 0.04 0.13

0.29 0.24 0.23 0.39 0.46 0.24

Y. Wang et al.

877

Table 3 Associations Between HFE Genotype Group and the Risk of Bilateral Primary Total Hip and Knee Joint Replacement for Osteoarthritis Allele

C282Y Copies

Cases

Odds Ratio (95% CI)

Odds Ratio (95% CI)a

0 1 2 2 0 1 (no H63D) 1 (and 1 H63D) 2

115 15 5 130 5 115 11 4 5

ref 0.80 (0.47-1.37) 5.75 (2.32-14.25) ref 5.92 (2.39-14.63) ref 0.71 (0.38-1.32) 1.22 (0.45-3.33) 5.75 (2.32-14.25)

ref 0.80 (0.47-1.37) 5.86 (2.36-14.57) ref 6.03 (2.43-14.95) ref 0.72 (0.39-1.33) 1.19 (0.44-3.23) 5.86 (2.36-14.56)

0 1 0 1 (no H63D) 1 (and 1 H63D)

192 34 192 26 8

ref 1.09 (0.76-1.57) ref 1.01 (0.67-1.52) 1.47 (0.72-2.99)

ref 1.10 (0.76-1.59) ref 1.03 (0.68-1.55) 1.42 (0.69-2.90)

Total hip replacement Codominant

Recessive With H63D

Total knee replacement Codominant With H63D

aAdjusting

0 or 1

P Valuea

0.42 ⬍ 0.001 ⬍ 0.001 0.29 0.74 ⬍ 0.001

0.62 0.90 0.34

for age, sex, body mass index, and educational level.

failed to show evidence of an association with the risk of total hip or knee replacement. Although individuals with hereditary hemochromatosis have a higher risk of OA compared with controls (9-11), the evidence for the relationship between HFE mutations and the risk of arthritis is neither consistent nor conclusive (8,1419). Most previous studies are cross-sectional or case-control studies that are prone to selection bias. In our prospective cohort study, the recruitment of participants was not based on the presence or absence of diagnosis or symptoms of hemochromatosis and occurred before the discovery of the HFE gene, thus reducing the potential for selective recruitment bias or reverse causation. A recent study reported that the hip was the most common localization for total joint replacement due to severe OA for patients with hemochromatosis (44 joints, 84.6%), whereas total knee replacement was less frequent (6 joints, 11.5%) (11). Another study showed definite/probable hemochromatosis was associated with more frequent bilateral hip, but not bilateral knee joint arthropathy than possible/unlikely hemochromatosis (12). Consistent with these findings, we observed an association of C282Y homozygosity with the risk of single total joint replacement for the hip but not for the knee, which was further strengthened by the findings for bilateral joint replacement. This study showed a stronger association of HFE C282Y homozygosity with bilateral total hip replacement than unilateral total hip replacement, which is consistent with the hypothesis that C282Y homozygosity explains a greater proportion of the risk of joint replacement in bilateral than unilateral cases. It is possible that the association observed for total hip replacement may be attributable to an effect of iron overload on hip bone. Iron overload has been shown to be associated with the inhibition of osteoblast metabolism, bone loss, and reduced bone mineral density (25-28), which may lead to reduced bone strength and make bone more susceptible to shape changes. There is increasing evidence

that joint shape is modifiable and that nonoptimal joint geometry is a risk factor for developing OA (29-31). It is likely that the effect of bone geometry on OA is seen early and obviously at the hip. The relationship between iron overload and hip bone shape will need to be confirmed by further work. Our study has several strengths. We had almost complete ascertainment of total hip and knee replacement because AOA NJRR collects the most complete set of data relating to hip and knee replacement in Australia and few participants left the country. We had extensive information on other factors that might confound the associations, none of which showed great variation between HFE genotypes (Table 1). Limitations include our inability to determine whether the associations with genotype were mediated through body iron stores, because data on baseline serum ferritin and transferrin saturation were not available for most of those with a total joint replacement. Genotyping errors, such that people with no or only 1 C282Y mutation were classified as C282Y homozygotes, would have attenuated the associations. The limited number of C282Y homozygotes among those with a total hip or knee replacement is another limitation of the study—some of the risk estimates have wide confidence intervals. We do not have complete and reliable joint replacement data for the study population before 2001. Although we excluded those MCCS participants who reported a joint replacement before 1 January 2001 at the follow-up visit, this information may be unreliable and is only known for 68% of the original cohort. As a result, some misclassification of joint replacement status might have occurred, although it is likely to have been nondifferential in relation to the HFE genotype and, thus, to have resulted in underestimates of the strength of any observed associations. Another limitation of this study is that we did not have any other assessment for OA such as symptoms or radiology or magnetic resonance imaging. Although total joint replacement is an efficacious procedure for

878

the treatment of symptomatic end-stage OA, the decision to proceed with the surgery may be affected by socioeconomic factors, patient perception, preference, and willingness (29,30). We have adjusted for education level in the analysis, which may have accounted for some differences in terms of socioeconomic status. We have also done analysis adjusting for self-description of health status and physical function limitation in the regression analyses, which did not alter the results (data not shown). HFE C282Y homozygosity is associated with increased risk of both single and bilateral total hip replacement for OA, whereas the association is not observed at the knee. The reason for this is unknown and the findings need to be confirmed by other studies.

HFE C282Y homozygosity associated risk of total hip replacement

14.

15. 16. 17.

18.

ACKNOWLEDGMENTS We would especially like to thank data manager Lisa Ingerson and statistician Nicole Pratt from the Australian Orthopaedic Association National Joint Replacement Registry, and Georgina Marr from Cancer Council Victoria.

19. 20.

REFERENCES 1. Olynyk JK, Trinder D, Ramm GA, Britton RS, Bacon BR. Hereditary hemochromatosis in the post-HFE era. Hepatology 2008;48(3):991-1001. 2. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, et al. A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nat Genet 1996; 13(4):399-408. 3. Burke W, Reyes M, Imperatore G. Hereditary haemochromatosis: a realistic approach to prevention of iron overload disease in the population. Best Pract Res Clin Haematol 2002;15(2):315-28. 4. Imperatore G, Pinsky LE, Motulsky A, Reyes M, Bradley LA, Burke W. Hereditary hemochromatosis: perspectives of public health, medical genetics, and primary care. Genet Med 2003;5(1):1-8. 5. Adams PC, Barton JC. Haemochromatosis. Lancet 2007; 370(9602):1855-60. 6. Gurrin LC, Bertalli NA, Dalton GW, Osborne NJ, Constantine CC, McLaren CE, et al. HFE C282Y/H63D compound heterozygotes are at low risk of hemochromatosis-related morbidity. Hepatology 2009;50(1):94-101. 7. Pietrangelo A. Hereditary hemochromatosis—a new look at an old disease. N Engl J Med 2004;350(23):2383-97. 8. Allen KJ, Gurrin LC, Constantine CC, Osborne NJ, Delatycki MB, Nicoll AJ, et al. Iron-overload-related disease in HFE hereditary hemochromatosis. N Engl J Med 2008;358(3):221-30. 9. Richette P, Ottaviani S, Vicaut E, Bardin T. Musculoskeletal complications of hereditary hemochromatosis: a case-control study. J Rheumatol 2010;37(10):2145-50. 10. Crooks CJ, West J, Solaymani-Dodaran M, Card TR. The epidemiology of haemochromatosis: a population-based study. Aliment Pharmacol Ther 2009;29(2):183-92. 11. Sahinbegovic E, Dallos T, Aigner E, Axmann R, Engelbrecht M, Schoniger-Hekele M, et al. Hereditary hemochromatosis as a risk factor for joint replacement surgery. Am J Med 2010;123(7):659-62. 12. Carroll G, Breidahl W, Bulsara M, Olynyk J. Hereditary haemochromatosis (HH) is characterized by a clinically definable arthropathy that correlates with iron load. Arthritis Rheum 2011; 63(1):286-94. 13. Valenti L, Fracanzani AL, Rossi V, Rampini C, Pulixi E, Varenna M, et al. The hand arthropathy of hereditary hemochromatosis is

21. 22.

23. 24. 25. 26. 27.

28.

29.

30.

31.

strongly associated with iron overload. J Rheumatol 2008; 35(1):153-8. Ross JM, Kowalchuk RM, Shaulinsky J, Ross L, Ryan D, Phatak PD. Association of heterozygous hemochromatosis C282Y gene mutation with hand osteoarthritis. J Rheumatol 2003;30(1): 121-5. Carroll GJ. HFE gene mutations are associated with osteoarthritis in the index or middle finger metacarpophalangeal joints. J Rheumatol 2006;33(4):741-3. Beutler E, Felitti VJ, Koziol JA, Ho NJ, Gelbart T. Penetrance of 845G ¡ A (C282Y) HFE hereditary haemochromatosis mutation in the USA. Lancet 2002;359(9302):211-8. Alizadeh BZ, Njajou OT, Hazes JM, Hofman A, Slagboom PE, Pols HA, et al. The H63D variant in the HFE gene predisposes to arthralgia, chondrocalcinosis and osteoarthritis. Ann Rheum Dis 2007;66(11):1436-42. Sherrington CA, Knuiman MW, Divitini ML, Bartholomew HC, Cullen DJ, Olynyk JK. Population-based study of the relationship between mutations in the hemochromatosis (HFE) gene and arthritis. J Gastroenterol Hepatol 2006;21(3):595-8. Waalen J, Felitti V, Gelbart T, Ho NJ, Beutler E. Prevalence of hemochromatosis-related symptoms among individuals with mutations in the HFE gene. Mayo Clin Proc 2002;77(6):522-30. Ellervik C, Birgens H, Tybjaerg-Hansen A, Nordestgaard BG. Hemochromatosis genotypes and risk of 31 disease endpoints: meta-analyses including 66,000 cases and 226,000 controls. Hepatology 2007;46(4):1071-80. Giles GG, English DR. The Melbourne Collaborative Cohort Study. IARC Sci Publ 2002;156:69-70. Graves SE, Davidson D, Ingerson L, Ryan P, Griffith EC, McDermott BF, et al. The Australian Orthopaedic Association National Joint Replacement Registry. Med J Aust 2004; 180(Suppl 5):S31-4. Australian Orthopaedic Association National Joint Replacement Registry. Annual Report. Adelaide: AOA; 2010. Lucotte G, Dieterlen F. A European allele map of the C282Y mutation of hemochromatosis: Celtic versus Viking origin of the mutation? Blood Cells Mol Dis 2003;31(2):262-7. Yamasaki K, Hagiwara H. Excess iron inhibits osteoblast metabolism. Toxicol Lett 2009;191(2-3):211-5. Weinberg ED. Role of iron in osteoporosis. Pediatr Endocrinol Rev 2008;1(Suppl 6):81-5. Sinigaglia L, Fargion S, Fracanzani AL, Binelli L, Battafarano N, Varenna M, et al. Bone and joint involvement in genetic hemochromatosis: role of cirrhosis and iron overload. J Rheumatol 1997;24(9):1809-13. Guggenbuhl P, Deugnier Y, Boisdet JF, Rolland Y, Perdriger A, Pawlotsky Y, et al. Bone mineral density in men with genetic hemochromatosis and HFE gene mutation. Osteoporos Int 2005; 16(12):1809-14. Doherty M, Courtney P, Doherty S, Jenkins W, Maciewicz RA, Muir K, et al. Nonspherical femoral head shape (pistol grip deformity), neck shaft angle, and risk of hip osteoarthritis: a casecontrol study. Arthritis Rheum 2008;58(10):3172-82. Gregory JS, Waarsing JH, Day J, Pols HA, Reijman M, Weinans H, et al. Early identification of radiographic osteoarthritis of the hip using an active shape model to quantify changes in bone morphometric features: can hip shape tell us anything about the progression of osteoarthritis? Arthritis Rheum 2007;56(11): 3634-43. Bredbenner TL, Eliason TD, Potter RS, Mason RL, Havill LM, Nicolella DP. Statistical shape modeling describes variation in tibia and femur surface geometry between Control and Incidence groups from the osteoarthritis initiative database. J Biomech 2010;43(9):1780-6.