вход по аккаунту


Pets and rheumatoid arthritis an epidemiologic survey.

код для вставкиСкачать
Pets and Rheumatoid Arthritis
An Epidemiologic Survey
Norman L. Gottlieb, Norman Ditchek, Jeffrey Poiley, and Iris M. Kiem
The prevalence of pets in households of patients with rheumatoid arthritis (RA) is compared to arthritic (non-RA) and nonarthritic control patients
in this epidemiologic survey. During the 5-year period prior to diseaseonset RA patients had greater exposure to dogs, cats, or birds (combined), dogs, and sick animals than control patients. This finding raises
the possibility that pets might serve as a reservoir for an infectious
agent@) capable of initiating RA.
The possibility that rheumatoid arthritis
(RA) is an infectious disease remains a viable working hypothesis based on a host of
observations, particularly of an immunologic nature (14). Some encouragement in
support of this view also is found in reports
of the isolation of bacteria (5), and mycoplasma (6) from synovial fluid and synovia,
and the transmission of an agent from human RA synovia capable of inducing bone
and joint lesions in animals (7). Wide variability in results in the attempt to reproFrom the Arthritis Division, Department of Medicine, and Department of Epidemiology and Public
Health, University of Miami School of Medicine,
Miami, Florida.
SupporLed by Training Grant in Arthritis and
Rheumatic Diseases (NIH 5-T01-AM05058) and the
H Greenberg and S Oritt Funds.
NORMAN L GOTTLIEB, MD: Assistant Professor of
Medicine; NORMAN DITCHEK, MD: Former Fellow in
Rheumatology. Present address: 1674 Meridian Avenue, Miami Beach, Florida; JEFFREY POILEY, MD:
Former Fellow in Rheumatology. Present Address:
500 E Rollins Street, Orlando, Florida; IRIS M KIEM,
MPH: Assistant Professor of Biostatistics, Department of Epidemiology and Public Health.
Address reprint requests to: Dr Norman Gottlieb,
Division of Rheumatology, PO Box 875, Biscayne
Annex, Miami, Florida 33152.
Submitted for publication May 25, 1973; accepted
October 29, 1973.
duce such observations makes the continued search for new approaches to testing
the infectious hypothesis of RA (8-11)
mandatory. For example, household pets
are known reservoirs for many human infectious diseases (12) but to the authors'
knowledge no studies have sought to relate
the exposure to household animals to RA.
The current epidemiologic survey compares the prevalence of pet ownership in
the 5-year period preceeding the development of clinically recognized RA to that
in patients with other forms of arthritis
and to nonarthritic patients. The timeliness of such a query is enhanced by recently reported findings of canine polyarthritis
with clinical, pathologic and serologic features of RA (13).
One hundred and five patients with definite or
classic RA (group A), according to the criteria of
the American Rheumatism Association (14), were
randomly selected from the Arthritis Outpatient
Clinics of the University of Miami Hospital Center,+ and were interviewed by the authors. A stand*Jackson Memorial Hospital, National Children's
Cardiac Hospital, and Miami Veterans Hospital.
Arthritis and Rheumatism, Vol. 17, No. 3 (May-June 1974)
Table 2. Number of Rheumatoid Arthritis
and Control Patients who Owned Specified Pet8
Table 1. Selected Characteristics
of Study Patients
(Group A) (Group B)
No. of patients
Mean Age
Females (%)
Caucasian (%)
Staff patients
(% 1
Latex fixation
nodules (%)
(Group C)
'At onset of respective diseases
ard structured questionnaire was used to determine
the type and extent of household animal contact
in the 5-year period prior to the initial symptoms
of RA. For present purposes, positive pet exposure
was defined as the presence of an animal for more
than 3 months in the person's home. Note was
made of the breed of animal, and where possible,
its medical fate.
Two groups of control patients, numbering approximately 100 each, were interviewed simultaneously, using the same questionnaire, to determine
household animal contact in the 5-year period
prior to the onset of their respective diseases. The
105 arthritic control patients (group B) had arthritic diseases other than RA, such as osteoarthritis
(70). gout (14), post-traumatic arthritis and others.
Ninety-five clinic outpatients with nonarthritic illnesses (group C) such as diabetes mellitus (ZO),
hypertension (15), cardiovascular disease (25), pulmonary disease (IS), and others, based on workup
in the general medical clinic, composed the second
control group. Patients whose diseases were of probable infectious etiology were excluded from the
control groups. The authors confirmed the diagnoses of patients by reviewing their charts and,
when advisable, by physical examination. Some
characteristics of the patients composing the study
groups are given in Table 1. To increase ihe control numbers used for statistical evaluation. arthritic and nonarthritic control groups were combined
(group D) and compared to the group with RA.
T w o statistical methods were employed to ana-
Arthritis Controls
(Group (Group
No. of
Dog, cat
or bird
Other pets
Sick pet
NonComarthritic bined
Controls Controls
(Group (Group
lyse the data, the Mantel-Haenszel procedure (15)
and the chi-square test. The Mantel-Haenszel procedure subclassified data by factors controlled in
the analysis and was used because it simultaneously
controlled for observed differences in patient characteristics (age, sex, race, private versus staff status)
when determining the level of significance of differences between RA and control groups. With unmatched patient subgroups, (eg, no private patients
in nonarthritic control group) however data were
deleted and sample size reduced. Therefore the
chi-square test, a method which preserved and
compared all data from RA and control groups,
was also used to analyze the results of this study.
This test did not control for differences in patient
characteristics, but the significance of these factors
was analyzed separately by comparing the prevalence of pets in young and old (arbitrarily divided
at age 50), male and female, black and Caucasian,
and private and staff patients.
T h e total patients in RA and control
groups who reported the presence of pets
in their house during the 5 years prior to
disease-onset is listed in Table 2. Using the
Mantel-Haenszel procedure the rheumatoid
group had significantly greater exposure to
one or more dogs, cats, or birds (combined)
than did the arthritic controls (.05 > P
>.02); the difference between RA and
Arthritis and Rheumatism, Vol. 17, No. 3 (May-June 1974)
combined controls was borderline (.l > P
> .05); and no significant difference existed
between RA had nonarthritic controls. Similarly, patients with RA and significantly
greater exposure to dogs than did arthritic
controls (.05 > P > .02), but differences between the RA and combined controls (.2 >
P > .l) and nonarthritic controls were not
significant. The prevalence of reported sick
animals was significantly greater in the
rheumatoid than in the combined controls
(.05 > P > .02), with borderline differences
between RA and arthritic controls (.l > P
> .05) and nonarthritic controls (.2 > P >
Using the chi-square test the following
probability values were obtained: a)-dog,
cat, bird (combined)-RA versus arthritic
controls, P < .01; RA versus nonarthritic
controls, P < .05;RA versus combined controls, P < .01. b)-Dog-RA
versus arthritic
controls, P < .02; RA versus nonarthritic
controls, P > .2; RA versus combined controls, P < .05.c)-Sick animal-RA versus
arthritic, nonarthritic and combined controls, P < .05 for each. Differences in patient characteristics did not significantly influence the prevalence of household pets as
determined by comparing young and old,
male and female, black and Caucasian, and
private and staff patients. The P value was
greater than .2 for each factor analyzed.
There were more cats and birds in the
homes of patients who developed RA than
in the control groups, but differences were
not statistically significant, with either
method of analysis. Insufficient numbers of
other pets were present to analyze statistically. When the rheumatoid population
was divided into seronegative and seropositive groups, and those with and without
subcutaneous nodules, the number who had
pets was proportional to the size of the
Of interest was the finding that 3 patients with R A had dogs believed to manifest arthritis and that one (not included in
Table 2) was bitten by an unknown dog 2
months prior to the onset of RA. Several
breeds of dogs and mongrels were believed
to be arthritic. No animals other than dogs
in the rheumatoid group, and no animals
in the control groups, were thought to have
joint disease. Unfortunately the retrospective nature of this study did not permit
validation of the diagnosis of arthritis, or
determination of its type.
The results of this study indicate that
the presence of household pets in the 5-year
period prior to disease onset was greater in
patients who developed RA than in control
patients. This new epidemiologic finding
requires additional investigation as studies
of pet contact in patients with RA may
provide a means of validating the infectious agent hypothesis of RA.
The statistical differences noted between
RA and control groups were at lower levels
of confidence with the Mantel-Haenszel
procedure than with the chi-square test.
This was partially a consequence of reduced sample size resulting from unmatched
patient subgroups, using the former method. While age, sex, race and patient status
might possibly influence prevalence of
household pets the analysis of our study
groups failed to demonstrate an association.
Although epidemiological studies, seeking to identify environmental factors of importance in the development of RA, have
examined trauma, climate, geography, and
other variables, review of the literature has
failed to reveal studies in which contact
with pets was examined (16-23). This is
somewhat surprising in that pets harbor a
wide variety of infectious organisms, in-
Arthritis and Rheumatirm, Vol. 17, No. 3 (MayJune 1974)
cluding bacteria, viruses, parasites, and
fungi, which are known to produce human
disease when transmitted by direct contact
or intermediate vector (12). Examples of
this might include rabies, cat scratch fever,
salmonellosis and cryptococcosis. Also there
is some evidence linking human leukemia
and other cancers to contact with pets via
spread of tumor viruses or other mechanisms (24). Perhaps the ubiquitous nature
of such animals in the American household
(25 million dogs, 30 million cats, 20 million
birds) has discouraged epidemiologic studies of this nature in the rheumatic diseases
Polyarthritis in the animal kingdom is
well documented (25). Many species of nonprimates such as swine and sheep, and more
recently nonhuman primates (gorilla), develop an inflammatory synovitis similar but
not identical to human RA (26). Some of
these are of known infectious etiology, and
cure sometimes follows the early application of appropriate antibiotics (1 1). The
identification of viral and bacterial agents
capable of causing rheumatoid-like arthritis in animals suggests the possibility of a
similar pathogenesis in man.
Whereas blood and synovial cultures,
serum antibody titers against a multitude
of infectious agents, and other studies have
failed to reproducibly identify a causative
agent in KA, such negative findings do not
exclude an infectious pathogenesis, Recent
investigations demonstrate current and continuing interest in the infectious etiology
of RA, at least to the extent that contact
with or reaction to the infectious agent
may provide the initial triggering event in
a “genetically” susceptible individual. Warren et al (7) claimed they provoked inflammatory bone and joint lesions in mice and
chicks by the intraperitoneal innoculation
of a saline homogenate prepared from hu232
man RA synovia. Furthermore, the musculoskeletal abnormalities thereby produced
were transmitted through multiple generations. These findings lend weight to the
hypothesis that RA is caused by an infectious agent and one which is transmissable
from humans to animals as well as congenitally in animals. The fact that Cole et al,
(27) were unable to reproduce and confirm
Warren’s findings adds to the confusion in
this important area of investigation.
Despite the statistically significant differences in frequency of pet contacts between
patients who develop RA and controls the
results observed in this study should be
viewed cautiously. The reasons for this include the relatively small size of the populations surveyed, and other unmatched variables such as urban versus rural residence,
private home versus apartment dwelling,
rules governing permissibility of pets, the
presence or absence of children in the
household and the difficulties involved in
recollecting remote events. The five year
period surveyed for contact with pets was
chosen arbitrarily, and may be too short
for some slow viruses, or too long for other
infectious agents (3). Lastly RA may be
multifactorial in its etiology and only certain cases associated with pet contact. Nevertheless the findings noted in this study
are of interest and suggest another, as yet
unbeaten path, for further investigation into the origins of RA.
We appreciate the helpful suggestions and
thoughtful reviews by Drs. David Howell and John
Talbott, University of Miami School of Medicine.
1 . Robinson WD: The etiology of rheumatoid
arthritis, Arthritis and Allied Conditions.
Eighth edition. Edited by JL Hollander.
Philadelphia, Lea and Febiger, 1972, pp
Arthritis and Rheumatism, Vol. 17, No. 3 (MayJune 1974)
2. Christian C: Rheumatoid arthritis-etiologic
considerations. Arthritis Rheum 7:455-466,
3. Nardo JM, Norton WL: Evidence for infectious etiologies in connective tissue disorders.
Sem Arthritis Rheum 1:2, 116-140, 1971
4. Bland JH, Phillips CA: Etiology and pathogenesis of rheumatoid arthritis and related
multisystem diseases. Sem Arthritis Rheum
1:4, 339-359, 1972
5. Stewart SM, Alexander WRM, and Duthie
JJR: Isolation of diphtheroid bacilli from
synovial membrane and fluid in rheumatoid
arthritis. Ann Rheum Dis 28:477-487, 1969
6. Williams MH: Recovery of mycoplasma
from rheumatoid synovial fluid, Rheumatic
Diseases. Edited by JJR Duthie, WRM Alexander. Baltimore, William and Wilkins,
1968, pp 172-181
7. Warren SL, Marmor L, Boak R, Liebes DM:
Transmission of an active agent from rheumatoid arthritis synovial tissue to chicks.
Arch Intern Med 128:619-622, 1971
8. Clasener HAL, and Biersteker PJ: Significance of diphtheroids isolated from synovial
membrane of patients with rheumatoid arthritis. Lancet 2: 1031-1033, 1969
9. Schumacher HR, Katz SL, Kundsin RB: T h e
continued search for viral and other infectious agents in the systemic rheumatoid diseases: study of early lesions. Proceedings of
the Conference on Atypical Virus Infections
-Possible Relevance to Animal Modes and
Rheumatic Disease, December 1968, Atlanta,
Georgia. Edited by C Christian, P Phillips,
R Williams. T h e Arthritis Foundation 1823, 1971
10. Sharp JT: Mycoplasma and arthritis. Arthritis Rheum 13:263-271, 1970
11. Decker JL: T h e relationship of mycoplasma
(PPLO) to rheumatoid arthritis and related
diseases: a working conference. Bull Rheum
Dis 16:412413, 1966
12. Carithers HA: Pets in the home: incidence
and significance. Pediatrics 21:840-848, 1958
13. Lewis RM, Bore1 Y: Canine rheumatoid arthritis, a case report. Arthritis Rheum 14:1,
67-74, 1971
14. Ropes MW, Bennett GA, Cobb S, et al: Revision of diagnostic criteria for rheumatoid
arthritis. Bull Rheum Dis 9: 175-1 76, 1958
15. Mantel N, Haenszel W: Statistical aspects of
the analysis of data from retrospective studies of disease. J Nat Cancer Inst 22:719-748,
16. Engel A: Rheumatoid arthritis in US adults,
Population Studies of the Rheumatic Diseases. Proceedings of the Third International Symposium, Excerpta Medica Foundation,
June 1966. Edited by P Bennett, P Wood,
1968, p p 83-89
17. Cobb S, Kasl S: Epidemiologic contributions
to the etiology of rheumatoid arthritis, with
special atten tion to psychlogical and social
factors. 16 p p 75-82
18. Lawrence JS: The epidemiology of rheumatic diseases, Textbook of T h e Rheumatic
Diseases. Edited by WSC Copeman. Edinburgh and London, E and S Livingstone Ltd,
1969, 163-181
19. Lawrence JS, Behrend T , Bennett PH, et al:
Geographical studies on rheumatoid arthritis. Ann Rheum Dis 25:425-432, 1966
20. Wolfe AM: The epidemiology of rheumatoid arthritis: a review. Part I: Surveys. Bull
Rheum Dis 19:518-523, 1968
21. Masi AT: T h e epidemiology of rheumatoid
arthritis: a review. Part 11: Incidence and
diagnostic criteria. Bull Rheum Dis 19:524529, 1968
22. Lewis-Faning E: Report on an inquiry into
the aetiological factors associated with rheumatoid arthritis. Ann Rheum Dis 9 Suppl:
1-94, 1950
23. Allander E: A population survey of rheumatoid arthritis. Acta Rheum Scand (suppl) 15:
24. Bross ID, Bertell R, Gibson R: Pets and
adult leukemia. J Am Public Health Assoc
621520-1531, 1972
25. Pearson CM: Arthritis in animals. Arthritis
and Allied Conditions, Eighth edition.
Edited by J L Hollander. Philadelphia, Lea
and Febiger, 1972, p p 195-207
26. Brown TMcP, Clark HW, Bailey JS, et al: A
mechanistic approach to treatment of rheu-
Arthritis and Rheumatism, Vol. 17, No. 3 (MayJune 1974)
matoid type arthritis naturallyoccurring in a
gorilla. Trans Am Clin Climatol 82:227-247,
27. Cole BC, Ward JR, Smith CB: Studies on
the infectious etiology of rheumatoid arthritis. Arthritis Rheum 16:191-198, 1973
Arthritis and Rheumatism, Voi. 17, No. 3 (May-Jumm 1974)
Без категории
Размер файла
367 Кб
survey, epidemiology, pets, arthritis, rheumatoid
Пожаловаться на содержимое документа