Scientists have used statistical sleight of hand in their claim that the Pill is the safest drug in history. But their own data shows breast cancer risks that are alarming.
In the 35 years since the Pill and hormone therapy have been on the market, doctors have not only denied that exogenous (outside) hormones can cause cancer, but have even argued that they have a protective effect. At least four major studies have demonstrated a link between the Pill and the epidemic of breast cancer in the West (and a number of others, the same link between hormone replacement therapy and breast cancer). The latest damning news comes from a study by the Department of Epidemiology at the Netherlands Cancer Institute, (The Lancet, 24 September 1994). However, once they reached their conclusion, the Dutch researchers, like most of the others before them, began back pedalling like mad, minimizing the overall implications of their findings by emphasizing that the increased risk mainly occurred among certain sub groups and concluding that their study was, in effect, good news: "Our findings accord with the mass of evidence that [oral contraceptive use] by women in the middle of their fertile years has no adverse effect on breast cancer risk."
Dr Ellen Grant, author of Sexual Chemistry and The Bitter Pill, participated in the first major British study of the Pill as an enthusiastic advocate until she began noticing a plethora of side effects developing. Her protests were ignored or hushed up. In the following article she argues that the link between breast cancer and hormones in the Pill and hormone replacement therapy has been evident in the major studies all along.
For the best part of two centuries we have known that sex hormones cause cancer in hormone dependent tissues like the breast. The latest Dutch report confirms that the Pill causes breast cancer. How then has the fact that the breast cancer epidemic caused by prescribed hormones been so skillfully played down?
Very simply, epidemiologists only talk about an increased cancer risk if the victims have taken the hormones for longer than usual. As the women most vulnerable to side effects soon stop taking hormones, the evidence becomes muddled. Most of the studies have major flaws in the way they have been designed. Nevertheless, no matter how badly done and underestimated the risk, one hard fact comes through: there is a direct, irrefutable correlation between the taking of female hormones and breast cancer.
Breast cancer is the commonest female cancer and the main cause of death in women up to age 55, but the risk goes on increasing with age, reaching a peak among 80-90 year olds. Countries with a high dairy fat consumption have a high and escalating breast cancer incidence, but they are also the same countries which have been vigorously prescribing hormones. Japan has never sanctioned the Pill, and few Japanese women get breast cancer.
In 1971 M P Vessey and R Doll published the first paper about breast disease in women taking the Pill. Fairly small numbers were analyzed, but the results were clear. Pill takers with breast cancer were more likely to have stayed on it for longer (more than two years) than other women (Cancer 1971; 28: 1395-9). The opposite was true for women with benign disease. Sixty two per cent developed lumps during their first year of Pill taking, and only 14 per cent of women with lumps went on taking the Pill for more than two years, compared with 33 per cent of controls. Most sufferers gave up, leaving very few longer users with lumps.
This blip in the data launched what soon became the widespread belief among epidemiologists and then the medical profession that women on the Pill for at least two years are "protected" from breast disease. The fact that it is the women with side effects, including sore breasts, who stop early after a month or two, within the first year or before two or four years are up was not mentioned, and has not been taken into account since.
Large Pill studies can be misleadingly reassuring. Younger, healthier, higher social class women are used, while women with known Pill related diseases, like thrombosis or breast disease were discouraged from participating, thus skewing the data. Furthermore, scientists have been sloppy about maintaining control groups or adequately controlling for factors that might confuse their results. The biggest ever study, by the Royal College of General Practitioners in the UK, examined 47,000 women, of which 49 per cent were on the Pill when they first enrolled in 1968-69. However, over the course of the study, another 26 per cent were swapped from Pill taking to the control group. By the end of the study only 25 per cent hadn't taken the Pill. In most studies it is customary for the study population to be split roughly in half.
However, what the study didn't ask was whether these same women had had fertility drugs, hormone replacement therapy or stilbestrol, all of which could raise breast cancer risk. Estrogens were available without prescription in the 1940s and 50s. Half of all new mothers took stilbestrol to dry up their milk supply until 1968, and more than a third are still given hormones or bromocriptine to block prolactin. Other new mothers are injected or implanted with the long acting progestogens or given progestogen only Pills. Both groups of women would then have an increased risk of breast cancer, but might be labelled as never user controls in Pill studies.
The net effect of this "tainting" was false reassurance; the incidence of breast cancer appeared more similar in both controls and women on the Pill, thereby painting a false picture of safety.
By 1979, the RCGP study had reported that breast lumps were twice as likely on the new lower doses of progestogen (Clinics in Obste & Gynae, 1984; II: 769-86).
Nevertheless, the authors claimed they had confirmed "a reduced incidence of benign breast neoplasia becoming apparent after two years of usage." There was no mention of the fact that of the 6,324 completely "new" pill takers enrolled, only 799 or 12.6 per cent remained on the Pill in the third year of Pill taking. Furthermore, those who'd dropped out had more breast disease than the control group of those who'd never taken the Pill. The claim that the Pill prevented breast disease was clearly nonsense, when nearly 90 per cent of new users had given up before the "magic" two years supposedly required for prevention ended.
This was an important mistake. It meant that doctors no longer refused to prescribe hormone pills to women who already had lumpy breasts or to give renewed prescriptions to those who got lumps when they started the Pill.
At this point, women over 35 were warned off because of their higher (by 10 times) risk of thrombosis. Over the next 10 years single women under 25 became the main Pill takers (10 per cent in 1971, compared with 90 per cent by 1981). This switch had enormous consequences. As younger starters can apparently tolerate the Pill for longer this increased the number of women likely to develop breast cancer any time up to 40 years later (see box, p 2).
The consequent temporary decrease of Pill taking among 25-44 year olds gave false reassurance. In the Seventies, there was massive publicity that the Pill was causing thrombosis, strokes and heart attacks, but little was said about cancer, which kills twice as many women under 55. As fewer 35-44 year olds were given the Pill or stilbestrol, their increase in breast cancer since 1962 fell from a rise of 46 per cent in 1974 to 14 per cent in 1984. By 1988 the increase has risen again to 37 per cent, due to the effect of early age Pill taking and latent cancer development coming through. (Women between 25-34 years old had a similar pattern.)
In 1983 Malcolm Pike and his team published results from California (The Lancet, 1983; II: 359-29), showing women developing breast cancer before they were 37 had used the Pill for longer than controls (an average of 49.6 months, compared with 39.2 months). This was a similar result to Doll and Vessey's 1971 paper. Pike concluded that women who have taken the Pill for six or more years before they are 25 have a four to five times greater risk of breast cancer.
In May 1989 the UK National Case Control Study Group, known as the Oxford/Family Planning Association Study, also found that young women who had taken combined Pills were more likely to get breast cancer before the age of 36 (The Lancet, 1989; I: 973-82). Nevertheless, the authors muddied this clear result by postulating that "there may be some protective effect of progestogen only pills".
Among the two groups women of compared, 90 per cent in both groups had used the Pill. Of the 755 cases with breast cancer diagnosed before age 36, two per cent were younger than 25, while the rest were mostly diagnosed at 32 to 35, after they had one or two children. One in 20 had had a biopsy for benign breast disease before the breast cancer was diagnosed and one in 10 had a mother or sister with a history of breast cancer. The cases had taken the Pill for a year longer than the controls. Pill use for greater than four years increased the risk by 1.43 times; and, for eight years the increase was 1.75 times, rising to 3.1 times six years after stopping.
Among the few cases on progestogen only pills (POPs), nine out of 10 had developed breast cancer after taking them for less than 24 months. Only 14 cases, compared with 22 controls, had stayed on POPs for more than two years.
It seems to me that this means that fewer women with a tendency to develop breast cancer can tolerate progestogens. This is quite the opposite of protection only a very short time seems long enough to induce breast cancer. (The average use was only 13 months.) Nevertheless, the "protection" claim firmly stuck.
In the recent Dutch study, 97 per cent of the 918 Dutch women with invasive breast cancer, diagnosed before age 36, had taken the Pill. For those starting before age 20 the increase was 3.5 times risk. Longer use increased the risk, with trends increasing sharply for younger and older takers.
The authors were wrong to claim that their data supported evidence that 25-39 year olds had no risk. Their largest group of cases in this study were aged 36-45, had taken the Pill for less than four years and had a 1.4 increased risk.
Furthermore, the authors admit they may have underestimated the risk due to selection bias. Women who take the Pill and became ill did not reply, nor, of course, did those who died. Among the controls, the never takers also tended not to reply. All this non reporting again skewed the results.
In England and Wales breast cancer registrations have increased by 69 per cent between 1962 and 1988, while deaths have increased by 30 per cent to 1992. For all responders, both cases and controls, longer term Pill use of greater than four years showed an overall increase of 77.3 per cent for women younger than 35, compared with 60.9 per cent for those aged 36-45 and only 38 per cent of women aged 46-54. The implications are horrific as most women now start the Pill before they are 20 and most of them take it for more than four years. One in nine women in the US are developing breast cancer.
Klim McPherson, an epidemiologist at the London School of Tropical Medicine, has estimated that because of the latent period, this risk could increase to one in four among early age longer term Pill takers. The risk is increased by 30 times if a near relation has the disease due to breast cancer genes. Any menopausal estrogens further ups their risk 3.5 times.
Furthermore, breast cancer increases rapidly with age (Br J Hosp Med, 1989; 41: 590 and 42: 159). In the Oxford/FPA Pill study breast cancer death rates doubled over 10 years but doubled in only three years among older women followed in the British HRT trial (see WDDTY vol 4, no 11 and Sexual Chemistry).
!ADr Ellen Grant
Adapted from Sexual Chemistry (Cedar, lb5.99/$11.99).