Follow-up of large groups of patients with Parkinson disease in Denmark, the United Kingdom, and the United States have consistently shown an elevated risk for malignant melanoma.1–4 Additionally, a Danish case–control study suggested that the risk for malignant melanoma is increased years to decades before clinical manifestation of Parkinson disease.5 This seemingly bidirectional link between Parkinson disease and malignant melanoma suggests that the 2 diseases have common environmental or genetic determinants. The consideration of a shared genetic factor was supported by a recent report from the United States of an increased risk for Parkinson disease among people with first-degree relatives affected by malignant melanoma.6 To further explore the hypothesis of a genetic link, we conducted a population-based study in Denmark of the occurrence of Parkinson disease in parents and siblings of patients with early-onset malignant melanoma.
Index Patients and Relatives
From the Danish population born after 1954 and alive on 1 April 1968 (or born later), we identified 4649 patients with malignant melanoma diagnosed at age 50 years or less (early-onset melanoma) before the end of 2007 (Table 1). Data on the patients were abstracted from the files of the Danish Cancer Registry, which has recorded incident cases of cancer on a nationwide scale since 1943.7
By use of their personal identification number, index patients were linked to the files of the Danish Central Population Register to obtain information on first-degree relatives. For people born in the year 1955 or later who were alive on 1 April 1968 (ie, on the date of the start of the Central Population Register, and for all people born thereafter), the Register routinely includes a cross-reference to parents and—via the mother—to siblings.
Tracing of close relatives disclosed that 2 × 23 index patients (ie, patients with malignant melanoma diagnosed at age 50 or below) were pairwise siblings. For these 23 families, the index patient whose malignant melanoma was diagnosed earliest was regarded as the proband in the family, while the other sibling was regarded as a family case of malignant melanoma. Overall, 15,877 relatives of 4626 probands with early-onset melanoma were identified, comprising 4333 mothers (94% of all mothers), 4234 fathers (92%), 3566 sisters, and 3744 brothers (Table 2). Of the 293 probands for whom information on mothers was missing in the Central Population Register, 103 were born outside Denmark and 182 were born in the first years of the study period (1955–1959) when the cross-reference facility of the computerized civil registration system was established. The numbers of identified parents and siblings, and the associated follow-up data are shown in Table 2.
Disease Outcomes in Relatives
Parents and siblings were followed in the files of the Danish National Hospital Register through 2008 for a first hospital contact for Parkinson disease. The Hospital Register contains information on all persons admitted to nonpsychiatric hospitals in Denmark since 1977.8 Only inpatients and outpatients who received a primary diagnosis of Parkinson disease were included in this study, because a primary diagnosis is considered to be more reliable and accurate compared with a supplementary diagnosis. The period of follow-up for the occurrence for Parkinson disease (ICD-8 code 342 and ICD-10 code G20) began at the 35th birthday of the parents and siblings or 1 January 1977, whichever occurred last. Follow-up ended at the date of first hospital contact for Parkinson disease, date of death, date of emigration, or end of 2008—whichever occurred first. Relatives were also linked to the files of the Danish Cancer Registry (until the end of 2007) to identify incident cases of malignant melanoma (ICD-7 code 190 and ICD-10 code C43). Follow-up of relatives for malignant melanoma began on the date of diagnosis in the proband or the relatives' 51st birthday, and ended on the date of death, date of emigration, or end of 2007—whichever occurred first.
The observed number of parents and siblings with a first hospital contact for a primary diagnosis of Parkinson disease was compared with that expected from age-, sex-, and calendar period-specific first hospitalization rates for Parkinson disease in Denmark. National hospitalization rates were applied to the person-years of observation among parents and siblings to obtain the number of hospitalizations expected, and the hospitalization rate ratio (HRR) as calculated as the ratio of observed to expected hospitalizations for Parkinson disease. The 95% confidence interval (CI) was computed on the assumption of a Poisson distribution of the observed number of hospitalizations.9 In parallel, the observed number of cases of malignant melanoma in relatives during follow-up was compared with that expected on the basis of national incidence rates from the Danish Cancer Registry, and the standardized incidence ratio (SIR) was calculated as the ratio of observed to expected number of malignant melanoma cases.
The study was approved by the Danish Data Inspection Agency (no. 2009-41-3300).
The 8567 parents and 7310 siblings of patients with early-onset melanoma accrued 282,065 person-years of follow-up: 225,448 for parents and 56,618 for siblings. During this time, 54 relatives were hospitalized for a primary diagnosis of Parkinson disease, while 47.5 hospitalizations were expected, yielding a slightly increased HRR of 1.14 (Table 3). There were no hospital contacts for Parkinson disease among the siblings, although 1.3 were expected. Early-onset Parkinson disease is generally regarded as predominating at ages <50 or <60 years. In our study, we saw no apparent clustering of cases among parents in this age range (Table 3). The average age at first hospital contact for Parkinson disease in parents was 67.6 years.
Follow-up of parents and siblings in the files of the Danish Cancer Registry identified 1689 cases of cancer, while 1682.7 were expected (SIR = 1.00), with 1584 cases among parents (1.00) and 105 among siblings (1.13). Of these cancers, 135 (0.8%) were malignant melanoma of the skin, when 58.8 were expected, yielding a markedly increased SIR of 2.30 (Table 4). The risk was clearly increased for both parents and siblings, overall as well as at all intervals of attained age.
There was no overlap between the 54 families in which 1 member had Parkinson disease and the 135 families in which 1 member had malignant melanoma. Seven probands had had 2 primary diagnoses of malignant melanoma before the age of 51 years, but none of the relatives of these 7 families were affected by Parkinson disease.
On the basis of a combined analysis of 2 prospective cohort studies in the United States, Gao et al6 reported a 2-fold increase in the rate of Parkinson disease in people who had a first-degree relative with malignant melanoma over that of people with no family history of malignant melanoma. The familial accumulation of the 2 diseases was considerable and did not change materially after control for smoking, intake of caffeine, ethnicity, or other known or suspected environmental risk factors for Parkinson disease. The authors considered that their finding provided indirect support for the view that malignant melanoma and Parkinson disease have common genetic components. In yet another study, building on the same dataset Gao et al10 found that the polymorphic variant of the melanocortin 1 receptor (MC1R) gene, which is associated with a loss of gene function, seemed to be closely associated with red hair color as well as with development of Parkinson disease. As this polymorphic variant of the MC1R gene has been associated in several studies with malignant melanoma and other types of skin cancer, the authors argued that pigmentation may play an important role in Parkinson disease and may offer a genetic explanation for the co-occurrence of Parkinson disease and melanoma.10,11
However, in the present study from Denmark of long-term follow-up for Parkinson disease of 8567 parents and 7310 siblings of young patients with malignant melanoma, we were unable to show a clear familial link between the 2 diseases. Thus, no hospital contacts for Parkinson disease were registered for siblings of young melanoma patients, when 1.3 cases were expected, and the rate of first hospitalization for Parkinson disease among fathers was close to that of the general population. Mothers had a modestly increased risk, although the estimate was based on a limited number of outcomes. If there is a genetic link between early-onset melanoma and Parkinson disease, it would be reasonable to expect an excess number of cases of early-onset Parkinson disease; however, we observed only a modestly increased risk for Parkinson disease among parents followed through ages 35–59 years. Although it is difficult to prove a lack of association, our study was sufficiently precise to exclude (with 95% confidence) an excess relative risk of 1.5 for Parkinson disease in first-degree relatives of patients affected by early-onset malignant melanoma.
The strengths of our study include the use of a national, computerized register of familial relationships during the past 50 years and national registers with historical coverage of cancer incidence and hospital discharge diagnoses. As the Danish Cancer Registry records data from the entire population of the country, virtually all cases of malignant melanoma in people born in 1955 or later were identified, as were more than 93% of parents of the people involved; this minimizes the possibility of bias due to selection of study subjects. Also, unbiased ascertainment of cases of Parkinson disease was obtained by linking relatives to the files of a national hospital register kept primarily for administrative purposes. However, data limitations required that the study be restricted to relatives of early-onset (<50 years of age) melanoma, which had 2 important consequences. First, this limited the statistical power of the study, particularly when assessing the hospitalization rate ratio for Parkinson disease among siblings of melanoma patients. Second, this also reduced the opportunity to generalize our findings to apply to relatives of older melanoma patients as well. A modest 1.3 expected cases of Parkinson disease reflects the fact that only 7% of the follow-up time of siblings was allocated to the age group 50+, ie, to the age group in which the incidence of Parkinson disease starts to rise. In the study reported here, we had no information on the smoking habits of the melanoma patients and their first-degree relatives. Smoking is inversely associated with risk for Parkinson disease and is thus a potential confounder in studies of risk factors for this disease. As the occurrence of malignant melanoma appears to be unrelated to smoking, it is unlikely that the lack of information on smoking habits affected our analysis of a familial link between Parkinson disease and malignant melanoma.12
The apparent lack of a familial link between malignant melanoma and Parkinson disease is also not due to errors in our register-based mapping of nuclear families in Denmark, as, in the same families, we found a strong tendency for accumulation of cases of malignant melanoma, similar to that found in a large population-based study in Sweden of familial risks for cutaneous melanoma.13 The relative importance of melanoma susceptibility genes could not be tested in this study. Sunburn and exposure to solar radiation are the main environmental causes of malignant melanoma, and shared sun-bathing habits of families may explain part of the excess.14 Inherited traits, such as number of benign naevi, skin type and pigmentation, and hair color may explain another part of the excess, acting by modulating the response to solar ultraviolet radiation.15,16
A further observation in this study was that there was no overlap between the families affected by Parkinson disease and those affected by malignant melanoma. Overall, our study indicates that if a shared genetic predisposition for malignant melanoma and Parkinson disease exists, it is of minor importance for the age range investigated and that the well-described link between the 2 diseases in the same individual is of environmental rather than genetic origin.
1. Møller H, Mellemkjær L, McLaughlin JK, Olsen JH. Occurrence of different cancers in patients with Parkinson's disease. BMJ.
2. Olsen JH, Friis S, Frederiksen K, McLaughlin JK, Mellemkjaer L, Møller H. Atypical cancer pattern in patients with Parkinson's disease. Br J Cancer.
3. Driver JA, Logroscino G, Buring JE, Gaziano M, Kurth T. A prospective cohort study of cancer incidence following the diagnosis of Parkinson's disease. Cancer Epidemiol Biomarkers Prev.
4. Becker C, Brobert GP, Johansson S, Jick SS, Meir CR. Cancer risk in association with Parkinson disease: a population-based study. Parkinsonism Relat Disord.
5. Olsen JH, Friis S, Frederiksen K. Malignant melanoma and other types of cancer preceding Parkinson disease. Epidemiology.
6. Gao X, Simon KC, Han J, Schwarzschild MA, Ascherio A. Family history of melanoma and Parkinson disease risk. Neurology.
7. Storm HH, Michelsen EV, Clemmensen IH, Pihl J. The Danish Cancer Registry—history, content, quality and use. Dan Med Bull.
8. Andersen TF, Madsen M, Jørgensen J, Mellemkjaer L, Olsen JH. The Danish National Hospital Register. A valuable source of data for modern health sciences. Dan Med Bull.
9. Bailer JC, Ederer F. Significance factors for the ratio of a Poisson variable to its expectation. Biometrics.
10. Gao X, Simon KC, Han J, Schwarzschild MA, Ascherio A. Genetic determinants of hair color and Parkinson's disease risk. Ann Neurol.
11. Landi MT, Kanetsky PA, Tsang S, et al. MC1R, ASIP, and DNA repair in sporadic and familial melanoma in the Mediterranean population. J Natl Cancer Inst.
12. International Agency for Research on Cancer. Tobacco smoke and involuntary smoking. In: IARC Monographs on the Evaluation of Carcinogenic Risks to Humans.
Vol 83. Lyon:International Agency for Research on Cancer; 2004.
13. Hemminki K, Zhang H, Czene C. Familial and attributable risks in cutaneous melanoma: effects of proband by age. J Invest Dermatol.
14. International Agency for Research on Cancer. Solar and ultraviolet radiation. In: IARC Monographs on the Evaluation of Carcinogenic Risks to Humans.
Vol 55. Lyon: International Agency for Research on Cancer;1992.
15. Landi MT, Bacarelli A, Tarone RE, et al. DNA repair, dysplastic nevi, and sunlight sensitivity in the development of cutaneous malignant melanoma. J Natl Cancer Inst.
16. Bishop JN, Harland M, Randerson-Moor J, Bishop DT. Management of familial melanoma. Lancet Oncol.