NR AMMF
AU Wientjens,D.P.W.M.; Davanipour,Z.; Hofman,A.; Kondo,K.; Matthews,W.B.; Will,R.G.; van Duijn,C.M.
TI Risk factors for Creutzfeldt-Jakob disease: a reanalysis of case-control studies.
QU Neurology 1996 May; 46(5): 1287-91
PT journal article; review; review, tutorial
AB To review the evidence for risk factors of Creutzfeldt-Jakob disease (CJD), we pooled and reanalyzed the raw data of three case-control studies. The pooled data set comprised 178 patients and 333 control subjects. The strength of association between CJD and putative risk factors was assessed by computing the odds ratio as estimate of the relative risk. The risk of CJD was statistically significantly increased for subjects with a family history of CJD (odds ratio = 19.1; 95% CI 1.1 to 348.0). Further, there was a significant association between the risk of CJD and a history of psychotic disease (odds ratio = 9.9; 95% CI 1.1 to 86.1). Although not significantly increased, there was an elevated risk of CJD for subjects with a family history of dementia, a history of poliomyelitis, subjects employed as health professionals, and subjects ever exposed to cows and sheep. No association could be shown with organ meat consumption, including brain. The negative results of this reanalysis reassures the absence of a common risk factor in all CJD patients. However, the ongoing epidemiologic surveillance of CJD in several European countries may provide more evidence to exclude any environmental exposure early in childhood.
VT
Creutzfeldt-Jakob disease (CJD) is a transmissible spongiform encephalopathy that may occur in an inherited, infectious, or sporadic form. CJD may be caused by various mutations in the prion protein gene (PRNP) on chromosome 20. [1] However, these mutations are found only in familial CJD patients, comprising not more than 15% of all CJD patients. [1] The infectious form of CJD appears to be rare among humans [2] and limited to iatrogenic transmission through neurosurgery and EEG electrode implantation, [3-6] corneal transplantation, [7] and human growth [8-10] and gonadotrophin [11] hormone. For most patients with sporadic CJD, the cause remains unknown. Studies on risk factors for CJD have yielded controversial results and do not show evidence for a common risk factor. [12-17] An increased risk of CJD has been reported for subjects with a history of infection, [12,17] surgery of the head, [13,14] trauma to the head or body, [13,14] tonometry, [14] consumption of pork, [15] and contact with various animals [16,17] in some but not all studies. The recent infection of British bovine cattle through protein supplements derived from animal carcasses has fueled speculations on the possibility of infection of humans through meat consumption and animal contact. Other than the established transmission of Kuru through cannibalistic rituals, [1,2] findings of studies on (organ) meat consumption including brain have not yielded consistent results. [12,15]
Some of the conflicting results may be explained by the small size of the individual studies yielding only limited statistical power to detect an association. To review the evidence for the risk factors previously reported, we pooled and reanalyzed the raw data of three case-control studies of CJD. As no meta-analysis of the existing case-control studies on CJD has been performed earlier, we present our results here.
Methods.
All formal case-control studies of CJD were ascertained through MEDLINE search and personal communications. Thus, four studies were identified as eligible for the reanalysis. [12-17] To increase the comparability between studies, we restricted the analysis to the three studies comprising patients diagnosed with CJD according to the criteria of Masters et al. [18] One study was performed in Japan, [13] one in the United States, [14-16] and one in the United Kingdom. [17] The pooled data set comprised 178 patients (mean plus minus SD age at diagnosis 61 plus minus 9.5 years; range 26 to 81 years) and 333 control subjects. For each study, ascertainment of patients and control subjects and data collection are discussed briefly.
Japan.
The study conducted in Japan comprised 60 cases. [13] According to the criteria of Masters et al, [18] 50% were classified as probable, 47% as definite, and 3% as transmissible CJD. For this study, neurologic and psychiatric institutions throughout Japan were requested to report patients diagnosed with CJD in the period 1975 through 1977. Patients were compared with 109 age-matched control subjects (plus minus 5 years), including 50 spouses and 59 neighborhood control subjects. In some instances, public health personnel served as neighborhood control subjects. The response rate was 80% in patients, 78% in spouses, and 93% in neighborhood control subjects. Data were collected by a structured interview. For patients, the data were obtained from the spouse. Control subjects were questioned directly.
United States.
The study performed in the United States comprised 26 patients, of which 23% were classified as probable CJD, 46% as definite, and 31% as transmissible. [14-16] Patients were ascertained through records submitted to the Laboratory of Central Nervous System Studies of the National Institutes of Health from Pennsylvania, New Jersey, Maryland, New York, and Delaware. Patients were diagnosed with CJD in the period 1970 through 1981. Two control groups were selected: hospital control subjects diagnosed with a nonneurologic disease at the hospital where the patient was diagnosed (n equals 22; matching sex and age plus minus 5 years) and relative control subjects (n equals 18; matching sex and age plus minus 10 years). The response rate was 96% in patients, 85% in hospital control subjects, and 86% in relative control subjects. Data were collected by a structured interview of the spouse (58%), child (27%), sibling (11%), or parent (4%) of the patients. Control subjects were questioned directly.
United Kingdom.
The study conducted in the United Kingdom comprised 92 patients, of whom 24% were classified as probable and 76% as definite CJD. [17] Patients diagnosed with CJD in the period 1980 through 1984 were ascertained through neurologic centers in England and Wales and through death certificates of the Office of Population, Censuses and Surveys. For each patient, two age- and sex-matched control subjects were derived from the hospital where the patient was diagnosed (n equals 184; matching age plus minus 4 years). Of the two control subjects, one suffered from a neurologic disorder other than CJD and the other from a nonneurologic disorder. The response rate was 75% in patients and is unknown for control subjects. For patients and control subjects, the data were collected by interviewing an informant. The relationship to the informant was matched in 87% for patients and control subjects. In 78% of subjects, it was the spouse, in 16% a child, and in 6% a sibling.
Analysis.
This reanalysis was based on the original raw data of the three studies. We only included risk factors assessed comparably in at least two out of three studies. The risk factors studied were family history of CJD, dementia, and Parkinson's disease; medical history, including head trauma, hospitalization for psychotic disease, history of poliomyelitis, hepatitis, jaundice, allergy, and blood transfusion; employment as health professional; exposure to animals; living in a rural area; and consumption of organ meat. Exposures of patients, relatives, and spouses may be similar for family and occupational history, animal exposure, place of residence, and diet, due to the shared environment. As this may bias the results toward the null hypothesis, relative and spouse control subjects were excluded in the analysis of these putative risk factors. Thus, the number of subjects included in the study may vary according to the number of studies and control subjects included, as well as to the missing values within a study. For each risk factor, we report the number of subjects exposed and the total number of subjects for whom data on the risk factor were available.
The strength of association between CJD and putative risk factors was assessed by computing the odds ratio (OR) as an estimate of the relative risk. [19] Relative risks were estimated by maximum likelihood methods and the 95% CI were based on the asymptotic SEs. Confounding by age, sex, and study center was taken into account by including these variables into the logistic regression model.
Results.
Family history of CJD, dementia, and Parkinson's disease for patients and control subjects are reported in the table. A positive family history of CJD was found only in CJD cases. Seven patients (4%) had a relative with CJD. An increased risk of CJD was also observed for patients with a positive family history of dementia in any relative, which was borderline significant (OR equals 1.9; 95% CI 0.9 to 4.1). When restricting the analysis to first-degree relatives, the OR remained increased (OR equals 1.8; 95% CI 0.8 to 4.0; not shown in the Table 1).
With regard to the medical history Table 1, significantly more CJD patients than control subjects had a history of hospitalization for psychotic disease. Three of five CJD patients were hospitalized 15 years or more before the onset of CJD. No association was found with head trauma, hepatitis, jaundice, allergy, or blood transfusion. The risk of CJD associated with poliomyelitis was increased (OR 3.9; 95% CI 0.3 to 43.4). However, the significance is difficult to interpret as the number of exposed subjects is very small. Fewer CJD patients than control subjects had a history of surgery. The difference was statistically significant for abdominal surgery.
A nonsignificantly increased risk of CJD was found for subjects exposed to patient tissue through their employment as a health professional (i.e., medical doctors, nurses, dentists, laboratory workers, ambulance employees) (OR 1.5; 95% CI 0.5 to 4.1) Table 1. Exposure to cows and sheep was associated with a borderline significant increased risk of CJD (OR 1.7; 95% CI 0.9 to 3.1 and OR 1.6; 95% CI 0.9 to 2.9, respectively). Also, there was a nonsignificant increased risk of CJD for subjects who lived in a rural area at diagnosis (OR 1.4; 95% CI 0.8 to 2.4; not in Table 1). Consumption of organ meat, including brain, liver, and kidney, was not associated with an increased risk of CJD Table 1.
--------------------------------------------------------------------------------------
Table 1. Risk factors for Creutzfeldt-Jakob disease (CJD)
Patients* Control Odds ratio [95% CI]
Subjects* Crude Adjusted +
Family history
CJD 13-17 7/166 0/285 18.8 [1.0-341.2]# 19.1 [1.1-348.0]#
Dementia 14-17 15/126 17/212 1.6[0.7-3.2] 1.9[0.9-4.1]
Parkinson's disease 14-17 7/117 7/194 1.7[0.6-5.0] 1.7[0.6-5.0]
Medical history
Psychotic disease 13-17 5/111 1/219 10.3[1.2-89.8]# 9.9[1.1-86.1]#
Head trauma 13-17 27/176 55/329 0.9[0.6-1.5] 0.9[0.5-1.5]
Poliomyelitis 13-17 2/168 1/332 4.0[0.4-44.6] 3.9[0.3-43.4]
Hepatitis/jaundice 14-17 9/118 18/222 0.9[0.4-2.2] 0.9[0.4-2.1]
Allergy 13-17 39/174 64/330 1.2[0.8-1.9] 1.2[0.8-2.0]
Blood transfusion 13-17 17/174 45/328 0.7[0.4-1.2] 0.6[0.4-1.2]
Surgery
CNS 13-17 0/149 10/301 0.3[0.1-1.2] 0.3[0.1-1.2]
Eye 13-17 6/176 23/332 0.5[0.2-1.1] 0.5[0.2-1.1]
Abdominal 13-17 39/176 96/331 0.7[0.5-1.1] 0.7[0.4-1.0]#
Tonsillectomy 13-17 13/171 32/331 0.8[0.4-1.5] 0.7[0.4-1.4]
Occupational, animal,
and dietary exposure
Health professional 14-17 8/119 7/206 1.5[0.6-4.2] 1.5[0.5-4.1]
cow exposure 13-17 26/95 26/145 1.7[0.9-3.2] 1.7[0.9-3.1]
sheep exposure 13-17 26/171 25/262 1.6[0.9-2.9] 1.6[0.9-2.9]
organ meat
Brain 13-17 14/177 33/261 0.6[0.3-1.2] 0.6[0.3-1.8]
Liver 13-17 104/117 179/206 1.2[0.6-2.5] 1.3[0.6-2.6]
Kidney 13-17 61/86 135/179 0.8[0.4-1.4] 0.8[0.3-1.4]
* Number of subjects with a positive history or subjects exposed/total number of subjects in analysis.
+ Adjusted for age, sex, and study site.
# p < 0.05
--------------------------------------------------------------------------------------
Discussion.
In this reanalysis of case-control studies, the risk of CJD was statistically significantly increased for subjects with a family history of CJD. We also found a statistically significant association between the risk of CJD and the patient's history of psychotic disease. No statistically significant association could be shown for medical history other than psychosis or with organ meat consumption. Although not significantly increased, more CJD patients than control subjects had a family history of dementia, a history of poliomyelitis, an occupational history as health professional, history of exposure to cows and sheep, and had lived in a rural area.
The internal validity of this reanalysis depends on the validity of the individual studies. Selection bias may have occurred through nonresponse in patients and control subjects in the individual studies. Further, control selection differed considerably across studies. However, to distort the findings, nonresponse and control selection should be associated with the risk factors studied. This seems unlikely for most risk factors. However, in the case of medical history, risk factor exposure may have been associated with control selection in the largest study conducted in the United Kingdom, [17] which was hospital based. The use of hospital control subjects may have led to high exposure frequencies for medical disorders, surgery, and blood transfusion. This may explain the lower frequency of surgery in CJD patients than control subjects. Unfortunately, the a priori statistical power of an analysis limited to the two small studies from Japan [13] and the United States [14-16] was too low to yield a meaningful solution to this problem. Another source of information bias might have been biased recall of putative risk factors. Also, the data collection through informants for patients and personal interviews for control subjects might have biased the data in two of three studies. To take this bias into account, we restricted the analysis to exposures not easily forgotten by control subjects and informants. For this reason, family history of dementia was restricted to first-degree relatives and the patient's history of psychotic disease to those requiring hospitalization. Further, we checked whether the association with a risk factor was found consistently across studies. In each of the studies, a family history of dementia, history of psychotic disease, employment as a health professional, and exposure to cows and sheep was associated with an increased CJD risk. This suggests that if bias occurred, it must have occurred in all studies similarly despite the differences in the design among studies.
Another issue is that the statistical power of our analysis may have been limited for rare exposures. Although we pooled the data of all case-control studies in which the patients met the criteria for CJD by Masters et al., [18] the number of patients included in this reanalysis was still small, which is reflected in the width of the 95% CIs. Given an exposure frequency in control subjects of 10%, a significance level of 5% (two-sided), and statistical power of 90%, the smallest detectable relative risk in a study of 200 patients and 200 control subjects is 2.5. [19] If the exposure frequency is only 1%, the smallest detectable relative risk under these conditions is 8.0, [19] which is a strength of association seldom observed in epidemiologic research.
Despite the limitations, our analysis showed a significant association between history of psychotic disease and the risk of CJD. Whether psychotic disease is a risk factor for CJD or is merely an early symptom of CJD pathology is undetermined. Future studies on the relationship between CJD and psychotic disease may be relevant. Our study shows familial aggregation of CJD, which may be explained by mutations in the PRNP gene that segregate in the families. [1,20] Also, homozygosity at codon 129 of PRNP may be associated with an increased susceptibility to prion diseases. [21,22] Our reanalysis showed a borderline significant association between CJD and family history of dementia. In a number of patients with familial forms of dementia, there have been PRNP mutations, [23] and the familial aggregation of CJD and dementia may therefore be related to PRNP mutations segregating in these families. However, other genetic factors (e.g., the apolipoprotein E gene [associated with both the risk of CJD [24] and Alzheimer's disease [25]] or nongenetic factors) may underlie the relationship; studies of families in which CJD and dementia aggregate may be of interest.
Our reanalysis did not support a number of previously reported associations. When pooling the data of the three studies, there was no association of the CJD risk with head trauma or the consumption of organ meat, including brain. The latter was associated with an increased risk of CJD in the study of Bobowick et al., [12] which is not included in the present reanalysis. Based on our findings, blood transfusion did not increase the risk of CJD in the period studied. There was no significant evidence in our study of transmission of CJD among humans or from cattle to human. However, there was a nonsignificant increase in risk of CJD for subjects employed as health professionals, subjects ever exposed to cows and sheep, and living in a rural area. In our reanalysis, the power to assess a statistically significant relationship for exposure to patients and exposure to animals was low. Data on the presence of prion disorders in patients or animals to which patients were exposed were not available. Further, the individual studies were performed in the 1970s and early 1980s when information in particular on inherited prion diseases among humans and animals was not yet available, which has limited the possibility to separate genetic and nongenetic patients in the present reanalysis. The negative results of our reanalysis are important in the light of public health to reassure the population that there are no common risk factors for CJD.
Our data cannot exclude an environmental source of infection that might have occurred early in childhood. The ongoing systematic epidemiologic surveillance of CJD in several European countries might provide evidence of such exposure as residential histories throughout life are studied. [26] These extensive studies with similar design will be highly comparable and therefore will be able to resolve the problem of selection bias and of the limited statistical power of our reanalysis.
Acknowledgment
This research was carried out under the auspices of Concerted Action on the Epidemiology of CJD, by means of a grant of the European Community. We would like to thank Hanneke den Breeijen and Michael Koenders for their help in data management and data analysis and Jane Mackenzie and Wakana Sakurai for their technical assistance.
REFERENCES [Click here for reference links. (24 references linked.)]
1.Prusiner SB, Hsiao KH. Human prion diseases. Ann Neurol 1994;35:385-395.
2.Brown P, Cathala F, Raubertas RF, et al. The epidemiology of Creutzfeldt-Jakob disease: conclusion of a 15-year investigation in France and review of the world literature. Neurology 1987;37:895-904.
3.Masters CL, Richardson EP Jr. Subacute spongiform encephalopathy Creutzfeldt-Jakob disease: the nature and progression of spongiform change. Brain 1978;101:333-344.
4.Will RG, Matthews WB. Evidence for case-to-case transmission of Creutzfeldt-Jakob disease. J Neurol Neurosurg Psychiatry 1982;45:235-238.
5.Thadani V, Penar PL, Partington J, et al. Creutzfeldt-Jakob disease probably acquired from a cadaveric dura mater graft. J Neurosurg 1988;69:766-769.
6.Bernouilli C, Siegfried J, Baumgartner G, et al. Danger of accidental person to person transmission of Creutzfeldt-Jakob disease by surgery. Lancet 1977;1:478-479.
7.Duffy P, Wolf J, Collins G, et al. Possible person to person transmission of Creutzfeldt-Jakob disease. N Engl J Med 1974;290:692-693.
8.Powell-Jackson J, Weller RO, Kennedy P, et al. Creutzfeldt-Jakob disease after administration of human growth hormone. Lancet 1985;2:244-246.
9.Gibbs CJ, Joy A, Heffner R, et al. Clinical and pathological features and laboratory confirmation of Creutzfeldt-Jakob disease in a recipient of pituitary-derived human growth hormone. N Engl J Med 1985;313:734-739.
10.Brown P, Gajdusek DC, Gibbs CJ, et al. Potential epidemic of Creutzfeldt-Jakob disease from human growth hormone therapy. N Engl J Med 1985;313:728-731.
11.Cochius JI, Mack K, Burns RJ, et al. Creutzfeldt-Jakob disease in a recipient of human pituitary gonadotrophin. Aust NZ J Med 1990;20:592-593.
12.Bobowick AR, Brody JA, Matthews MR, et al. Creutzfeldt-Jakob disease: a case-control study. Am J Epidemiol 1973;98:381-394.
13.Kondo K, Kuroiwa Y. A case-control study of Creutzfeldt-Jakob disease: association with physical injuries. Ann Neurol 1982;11:377-381.
14.Davanipour Z, Alter M, Sobel E, et al. Creutzfeldt-Jakob disease: possible medical risk factors. Neurology 1985;35:1483-1486.
15.Davanipour Z, Alter M, Sobel E, et al. A case-control study of Creutzfeldt-Jakob disease: dietary risk factors. Am J Epidemiol 1985;122:443-451.
16.Davanipour Z, Alter M, Sobel E, et al. Transmissible virus dementia: evaluation of a zoonotic hypothesis. Neuroepidemiology 1986;5:194-206.
17.Harries-Jones R, Knight R, Will RG, et al. Creutzfeldt-Jakobdisease in England and Wales, 1980-1984: a case-control study of potential risk factors. J Neurol Neurosurg Psychiatry 1988;51:1113-1119.
18.Masters CL, Harris JO, Gajdusek DC, et al. Creutzfeldt-Jakob disease: patterns of worldwide occurrence and the significance of familial and sporadic clustering. Ann Neurol 1979;5:177-188.
19.Schlesselman JL. Case-control studies. New York: Oxford University Press, 1982.
20.Collinge J, Brown J, Hardy J, et al. Inherited prion disease with 144 base pair gene insertion. 2. Clinical and pathological features. Brain 1992;115:687-710.
21.Collinge J, Palmer MS, Dryden AJ. Genetic predisposition to iatrogenic Creutzfeldt-Jakob disease. Lancet 1991;337:1441-1442.
22.Palmer MS, Dryden AJ, Hughes JT, et al. Homozygous prion protein genotype predisposes to sporadic Creutzfeldt-Jakob disease. Nature 1991;352:340-341.
23.Collinge J, Owen F, Poulter H, et al. Prion dementia without characteristic pathology. Lancet 1990;336:7-9.
24.Amouyel P, Vidal O, Launay JM, et al. The apolipoprotein E alleles as major susceptibility factors for Creutzfeldt-Jakob disease. Lancet 1994;344:1315-1318.
25.Strittmatter WJ, Saunders AM, Schmechel D, et al. Apolipoprotein E: high avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer's disease. Proc Natl Acad Sci USA 1993;90:1977-1981.
26.Alperovitch A, Brown P, Weber T, et al. Incidence of Creutzfeldt-Jakob disease in Europe in 1993. Lancet 1994;343:918.
Copyright 1996 by Advanstar Communications Inc.
Supported by the Netherlands Institute for Health Sciences (NIHES), the Slow Virus Infection Research Group Japan, and the Japan Ministry of Health and Welfare.
Received August 21, 1995. Accepted in final form November 1, 1995.
IN Indem sie die 178 Krankendaten und 333 Kontrolpersonendaten dreier Studien zusammenfaßten und erneut analysierten, versuchten die Autoren, Risikofaktoren für die Creutzfeldt-Jakob-Krankheit zu finden. Sie fanden eine eindeutig erbliche Komponente und eine deutliche Häufung bei Menschen, die schon einmal psychotisch auffällig waren. Etwas häufiger als normal wurde die Creutzfeldt-Jakob-Krankheit bei Menschen beobachtet, in deren Familien es Demenzen oder Poliomyelitis gegeben hatte, oder die im Gesundheitswesen oder immer mit Rindern gearbeitet hatten. Keinen Zusammenhang fanden sie mit dem Verzehr von Fleisch und Gehirn. Bei der geringen Zahl von Vegetariern und weil Infektionen auch durch Milchprodukte, Arzneimittel und Kosmetika erfolgen können, ist dies auch kein Wunder. Auch diese Meta-Fall-Kontroll-Studie erkannte Bluttransfusionen nicht als Risikofaktor für die Creutzfeldt-Jakob-Krankheit. Allerdings handelte es sich bei einem Teil der Kontrollen ausgerechnet um Krankenhauspatienten, die natürlich mit erhöhter Wahrscheinlichkeit Bluttransfusionen erhalten hatten.
ZR 28
MH Animal; Case-Control Studies; Cattle; Creutzfeldt-Jakob Syndrome/*epidemiology/*etiology/genetics; Environmental Exposure; Europe; Human; Japan; MEDLINE; Meat; Odds Ratio; Reference Values; Risk Assessment; Risk Factors; Sheep; Support, Non-U.S. Gov't; United States
AD Department of Epidemiology & Biostatistics, Erasmus University Medical School, Rotterdam, The Netherlands.
SP englisch
PO USA
OR Prion-Krankheiten 8