This article (with its slightly odd title) was first published in 1988. I put it here as an example of the persistence of beliefs without much evidence. What are called 'old wives' tales' illustrate the same principle. I don't know if reliable updates are available; I doubt it, since cancer is still not understood. The paper here is an example of one aspect of revisionism, namely taking the trouble to track sources of information. (Note: This is probably not my copyright; I was sent the paper, and decided to post it, though without a table, and with incomplete references, as an example of revisionism).
Skrabanek and McCormick took their role in Community Health very seriously. Skrabanek was unimpressed with medical students, regarding them as gullible and uncritical. Here's a short review of Skrabanek and McCormick's book Follies and Fallacies in Medicine. The book includes information on the unreliability of blood pressure measurements as people age. However it predates much AIDS criticism, and BSE, swine flu, anal sex problems, drug addictions and more errors and misinterpretations. It also ignores war material, such as 'Agent Orange'.
Big-Lies site
Journal of Clinical Epidemiology Vol. 41, No. 6, pp. 577-582, 1988
0895-4356/88 $3.00 + 0.00
Printed in Great Britain. All rights reserved
© 1988 Pergamon Press plc
CERVICAL CANCER IN NUNS AND PROSTITUTES: A PLEA FOR SCIENTIFIC CONTINENCE
PETR SKRABANEK
Department of Community Health, University of Dublin, Trinity College, Dublin, Ireland
(Received in revised form 12 November 1987)
[Skrabanek's keywords: Cervical cancer, Nuns, Prostitutes, Sexism, Wishful thinking]
INTRODUCTION
An expert answering a reader's query in a medical weekly stated about carcinoma of the cervix that "it is now well documented that the disease is rare in nuns and common in prostitutes", adding, somewhat cryptically, that "a connection between intercourse and cervical cancer was apparently first suggested in 1842" [1]. Both statements are false but widely believed to be true, presumably because they support what is believed to be proved.
RIGONI-STERN, 1842
A reference to an obscure Italian communication from 1842 has become de rigueur in the opening paragraphs of articles on the aetiology of cervical cancer, but how many authors have read the original? For example, an early culprit wrote: "Speculation on the relation of marriage to onset of cervical cancer goes back to 1842, when Rigoni-Stern proposed the non-married status of Catholic Sisters as a reason for an associated low frequency" [2]. Subsequent authors, copying from each other, gradually embellished the nun's tale, adding various invented details.
Rigoni-Stern was a Veronese surgeon and an amateur epidemiologist [3]. He analysed over 150,000 death certificates from the Veronese district for the years 1760-1839. Of 74,184 women who died, 1288 were nuns. Rigoni-Stern estimated that cancer in nuns was about five times more common than in other women, mainly because of an excess of breast cancer in nuns. He made no comments on "rarity" of cervical cancer in nuns, as cervical cancer was not distinguished from other cancers of the uterus. In fact, he recorded four deaths from uterine cancer in nuns, while the expected number (based on 361 cases in the remaining 72,896 women) was six. The very low numbers for this and other cancers in Rigoni-Stern's data suggest under-diagnosis [4].
OTHER NUN STUDIES
The evidence for the claim that cervical cancer is rare in nuns rests on the work of Gagnon [5]. He searched "medical files of an annual average of 13,000 women, covering a twenty-year period, in archives of many different convents" but he did not find a single case. However, Gagnon admitted that 1500 files were destroyed and another 2000 "could not be verified". He was "stupefied, not to say alarmed" by this negative finding and embarked on another search, this time using records from several pathological laboratories. In this search he identified three cases of cervical cancer in nuns. He concluded that "it was necessary, in my opinion, that very exceptionally at least this variety of cancer be found in virgin women".
Janet Towne, in a somewhat more reliable study, often misquoted as supporting Gagnon, stated that her own results were "quite different from those of Gagnon, in that 6 virginal women were recorded with proved cervical carcinoma, 3 having occurred in our own series of cases and 3 from the general survey" [6].
There is an interesting, rarely quoted, study from Holland, based on the Registrar-General's vital statistics from the period 1931-35: cervical cancer in nuns accounted for 2.5% of all cancer deaths in nuns (5/197), which was about the same as for wives of university teachers (2/70) and even higher than in farmers' wives (20/1183) [7].
In a survey of mortality in German nuns. Schömig found that cancer in nuns and in the general (female?) population was equally frequent (13.2 vs 13.4%, respectively) [8]. The nuns had a life expectancy about 10 years less than the general population. The frequency of genital carcinoma in nuns and in the general population was the same (23.6 vs 24.7% of all cancers, respectively). Of seven genital cancers in nuns, for which the site was specified, one was cervical cancer, four were cancers of the corpus, and two were ovarian cancers.
In a recent study on the mortality of nuns in Britain, Kinlen found 20 deaths from carcinoma of the uterus (site unspecified) against 28 expected. After 1941, when cervical cancer became to be classified separately, two nuns died of cervical cancer against 10 expected [9]. In a survey of three orders of nuns in the U.S.A., Taylor el al. found, in a cohort born between 1870 and 1889, eight carcinomas of the uterus against an expected figure of 18 [10].
Fraumeni et al, collected 5893 death certificates among 41 religious orders. "Only white, native-born, never-married sisters" were included, while those who "had performed household or manual duties, were nurses, or had served at foreign missions" were excluded. Among 1021 cancer deaths there were 102 uterine cancers (76 site unspecified, 15 cancers of the corpus, 11 cancers of the cervix) [11].
PROSTITUTES, VENEREAL DISEASE, AND CERVICAL CANCER
The paucity of good data on cervical cancer in prostitutes is even more striking than in nuns. The nineteenth-century doctors thought that uterine cancer was rare in prostitutes. Thus, for example, Drysdale wrote: "The evidence of Duchalet, Acton, Lippert, Bare of Nantes, and others, show incontestably that the health of prostitutes is above that of women in general. The only two diseases which infect them peculiarly being syphilis and scabies. Cancer of the womb is rare among prostitutes. Lippert of Hamburgh had not seen a case in eleven years among them" [12]. The best study is over 30 years old: Rojel found among 1262 patients with cervical cancer attending the Radium Centre in Copenhagen, 40 prostitutes (3.2%) and he calculated that prostitutes were four times more likely to be among the cases than among the controls [13]. All Rojel's prostitutes belonged to the lowest socio-economic stratum, but the data were not corrected for this.
Other studies are summarised in Table 1 [14-27] [Table is omitted - RW]. Only two studies deal specifically with prostitutes [19,22], though it was implied or stated in other studies, particularly those of prison populations, that a part of the clientele were prostitutes. The studies provide no evidence that cervical cancer is a more common cause of death in prostitutes than in other women. The Taiwan study explicitly contradicted the belief that "prostitution predisposes to increased rates of cervical cancer" [22]. The term "carcinoma in situ" (CIS) in these studies was used promiscuously, without histological definition and verification, and occasionally meant nothing more than a "positive smear"; yet, in the titles of these studies the term was shortened simply to "cancer". In one study, in which one third of the prisoners were alleged to be indulging in "prolonged scortatory* practices", the only case of invasive carcinoma occurred in a woman not classified as a prostitute [15].
[Footnote: *The word "scortatory" is not in English dictionaries, but appears to be derived from the Latin scorior (to whore), scorium (a concubine).]
The lack of relationship between venereal disease and cervical cancer was discussed by Gardner and Lyon [28]. However, I have included data on the prevalence of "carcinoma in situ" in patients attending VD clinics, together with some early and more recent prevalence studies on "carcinoma in situ" in various populations, for comparison (see Table 1). While none of these data are reliable and do not reflect the true incidence of either "premalignant" lesions or of invasive carcinoma, their .wide scatter casts doubt on the interpretation of uncontrolled studies used as evidence that cervical cancer is "common" in prostitutes.
LESSONS FOR HEALTH EDUCATION
The link between cervical cancer and prostitution, pace the experts' opinion, is not "well documented". Statements such as "if one were to grade women by their sexual experience, from virgin to prostitute, the incidence of cervical cancer would be related to the amount of sexual exposure" [29] are sexist and degrading. Similarly, the term "promiscuity", often used in the literature on cervical cancer, is unhelpful. According to some authors, "promiscuity" means "sexual intercourse with more than one partner" [30] or with more than two [31]. It seems that promiscuity, if it means anything, is having more sex than the investigator. In a Dutch study on cervical cancer, 83% of cases had only one partner, and after controlling for the age at first coitus, the number of partners had no separate effect on the relative risk of invasive carcinoma in screened vs unscreened women [32].
Epidemiological research cannot prove causation. Observations which may have a bearing on hypotheses about the aetiology on cervical cancer should not be used for imputing causation and for blaming the victim. It is also dangerous to use the results of case-control studies as a basis for mass intervention measures. For example, when it was believed that the cause of cervical cancer was smegma ("proved" experimentally by inserting equine smegma into murine vaginas), the editor of the JAMA called for circumcision of all infants of poor parents, as it would be "more practical and thorough" than to teach the proletariat "good penile hygiene" [33]. Similarly, when the health educators convinced themselves that cervical cancer was directly related to a high frequency of coitus, a Senior Medical Officer from the British Department of Health announced that "the time was ripe for a campaign"; in the same breath he warned against "a very real danger... in fostering the idea that [cervical] cancer... may be associated with venereal disease" [34]. A few years later, it is now argued by some epidemiologists that cervical cancer is not only "associated" with venereal disease but that it is a venereal disease. Only a minority still resists the notion: "although this is called a venereal disease and the press have associated it with promiscuity, in fact the greatest risk factor... is that of age and related to all sexually active women" [35]. It is uncharitable to accuse the Press of spreading false rumours, when the Press lifted their story directly from the epidemiological literature. And if a question of priority for the claim that cervical cancer is a venereal disease should ever arise, then Jean Astruc, an eighteenth-century French physician, should be considered, as he included among the causes of uterine cancer "injection of semen tainted with lues" and "venereal virus" [36].
The link between cervical cancer and misbehaviour preoccupies some experts: one epidemiologist showed that patients with cervical cancer were seven times more likely to have first coitus on the ground than in bed. and he provided details of the relative risks for 22 different ways of masturbating [37].
It is not helpful to argue with an assumption which remains to be proved. It begs the question. In one study, three women with cervical cancer said that they had only one partner: the investigators, believing in the promiscuity theory, disbelieved their informants: "the most likely explanation is that either husband or wife had in fact more than one sexual partner" [38]. The widespread decline in the incidence and mortality of cervical cancer in developed countries in the last 50 years has been interpreted as due to "less recourse to prostitutes than the older generation" [39]. It would be equally logical to argue that the decline was due to a general increase in chastity [40].
If cervical cancer were a venereal disease, the consequences might include: (1) screening and treating (?) healthy male carriers; this could be made compulsory before entering into a marriage contract; (2) screening for other venereal diseases at the time of the cervical smear: (3) exclusion from screening programmes of monogamous women, provided that their husband is "negative", as "strictly monogamous couples ... have negligible risk" [38]; (4) the end of mass screening programmes, since only women with a "promiscuous" past, or married to "promiscuous" husbands who do not use condoms, would be at risk; (5) a positive smear would be a smear on the woman's character; (6) resurrection of the popular belief, fought against by health educators for decades, that cancer is an infectious, transmissible disease.
Before all this happens, more work perhaps should be done on unresolved issues, such as (i) the aetiology of cervical cancer, and (ii) the role of viruses, if any, in the aetiology. It is also important that cases of carcinoma of the cervix in virgins are carefully documented; it seems that gynaecologists are aware of such cases but they may be inhibited from publishing them, fearing that they would not be believed. One of the reviewers of this paper stated that "there is no question that cervical cancer can and does occur in women who have not been engaged in sexual activity". This opinion is in a startling contrast with views of others, e.g. Maisin's: "the nuns and the women who remain virgins never develop cervical cancer" [41].
Lest this communication be misrepresented, I wish to make it clear that I do not intend to imply that cervical cancer is in no way related to what the old gynaecologists (such as Gagnon) used to call "cervicitis"—an ill-defined term encompassing some normal conditions and also lesions due to infectious, chemical, and other agents. Mine is a moral tale and not a contribution to the enigma of the aetiology of cervical cancer. Some authors still believe firmly in herpes simplex virus type 2 (HSV-2) as a causative factor in cervical cancer; they are now in the minority. As pointed out in a recent authoritative review: "the most informative prospective investigation revealed no relationship between HSV-2 and subsequent cervical neo-plasia" [42]. Similarly, a recent Lancet editorial stated that "the strong association between sexual activity and cervical cancer has encouraged the search for a sexually transmissible agent that could initiate or promote cervical neoplasia. Spirochaetes, spermatozoa, smegma, Trichomonas vaginalis, Chlamydia trachomatis, and HSV-2 have all come under suspicion, but proof of carcinogenesis has been lacking in every case" [43].
The latest of the putative venereal culprits is a human papilloma virus. It is however by no means clear that its only mode of transmission is a sexual contact: about 40% of normal oral biopsies in one study showed the presence of HPV-16 DNA, i.e. the type believed to be causally associated with cervical carcinoma [44]. HPV-16 has been found as often in cervical biopsies in normal women as in women with cervical cancer, if age-adjustment was carried out [45]. The presence of HPV-16 in cervical tissue does not correlate with lesions clinically diagnosed as CIN [46]. The frequency of HPV infection in the cervix decreases with age, while the frequency of invasive carcinoma increases with age [47]. As the Lancet editorialist concluded: "the high prevalence of papilloma-virus infection in women with cytologically and colposcopically normal cervices casts further doubts on the oncogenic role of these viruses" [43].
CONCLUSION
The epidemiological evidence on the prevalence of cervical cancer in nuns and prostitutes is of very poor quality and neither supports nor contradicts the belief that cervical cancer is a venereal disease. The evidence is so poor that it should not be used as additional "evidence" for a hypothesis which remains to be proved. Failure to distinguish hypotheses from facts delays clarification of the problem of the aetiology of cervical cancer.
Acknowledgements—I wish to thank the Wellcome Trust for support.
REFERENCES [Only 15 of 47 included]
1 Drife JO. Br Med J 1984; 288: 992.
2 Rotkin ID. Relation of adolescent coitus to cervical cancer risk. JAMA 1962; 179: 386-491.
3. Cislaghi C. Rilettura delle note del dott. Rigoni-Stern sulla frequenza del cancro, Verona 1842. Epidemiol Prevenzione 1978; No. 3: 48-54
4. [Rigoni-Stern]. Fatti statistici relativi alle malattie cancerose che servirono di base alle poche cose dette dal dott. Rigoni-Stern il di 23 settembre alia Sottosezione di chirurgia del IV Congresso degli scienzati Italiani. Giornale per Servire ai Progressi della Patologia e della Terapeutica (Venezia) 2 1842; (2nd series): 507-517.
5. Gagnon F. Contribution to the study of the etiology and prevention of cancer of the cervix of the uterus. Am J Obstet Gynecol 1950; 60: 516-522.
6. Towne JE. Carcinoma of the cervix in nulliparous and celibate women. Am J Obstet Gynecol 1955; 69: 606-613.
7. Versluys JJ. Cancer and occupation in the Netherlands. Br J Cancer 1949; 3: 161-185.
8. Schömig G. Die weiblichen Genitalkarzinome bei sexueller Enthaltsamkeit. Strahlentherapie (München) 1953; 92: 156-158.
9. Kinlen LJ. Meat and fat consumption and cancer mortality: a study of strict religious orders in Britain. Lancet 1982; i: 946-949.
10. Taylor RS, Carroll BE, Lloyd JW. Mortality among women in three Catholic religious orders with special reference to cancer. Cancer 1959; 12: 1207-1225.
11. Fraumeni JF, Lloyd JW, Smith EM, Wagoner JK. Cancer mortality among nuns: role of marital status in etiology of neoplastic disease in women. J Natl Cancer Inst 1969: 42: 455-468.
12. Drysdale C. On the medical aspects of prostitution. Med Press 1866; i: 123-125.
13. Rojel J. The Interrelation between Uterine Cancer and Syphilis. A Patho-Demographic Study. Copenhagen: A Busck; 1953.
14. Nelson RB, Hilberg AW. The diagnosis of unsuspected cancer of the cervix. J Natl Cancer Inst 1951: 11: 1081-1089.
15. Pereyra AJ. The relationship of sexual activity to cervical cancer. Cancer of the cervix in a prison population. Obstet Gynecol 1961; 17: 154-159.
First uploaded 2014-01-02. Document from University of Dublin. HTML Rae West.