Wednesday March 24 2010
The benefits of regularly taking aspirin for healthy people are unclear
“A simple aspirin can slash women's risk of breast cancer,” reported The Sun. It said that over-50s who take one pill a day can also cut their chances of getting ovarian cancer.
This study looked at painkiller use and hormone levels in 740 postmenopausal women. It found that those who regularly used aspirin may have had lower oestrogen levels than women who never or rarely used the painkillers. It did not examine cancer outcomes in women.
This study can only demonstrate trends and associations, and does not show that one thing caused the other. It involved two cross-sectional analyses, in which the women’s hormone levels were measured at the same time as their painkiller use was assessed. As such, the results cannot show which came first, or suggest that painkiller use actually affected hormone levels. The results themselves were also quite variable, for example, frequent aspirin use was associated with lower oestrogen levels in the 1988 analysis but not in the 1990 analysis.
Importantly, this study does not provide any evidence that aspirin, anti-inflammatory drugs or paracetamol reduces the risk of hormone-related cancers such as breast or ovarian cancer.
Aspirin benefits people who are at risk of cardiovascular disease. However, it is also associated with a risk of internal bleeding, and for healthy people, the benefits of regularly taking it are less clear. The results of this study alone do not support the use of aspirin or anti-inflammatories to reduce the risk of breast or ovarian cancer in postmenopausal women.
Where did the story come from?
Margaret Gates and colleagues from the Harvard School of Public Health carried out this research. The study was funded by the National Cancer Institute and published in the (peer-reviewed) medical journal Cancer Epidemiology Biomarkers and Prevention.
Withstanding the headline, the Daily Express generally gave an accurate representation of this research by discussing how regular use of aspirin was associated with lower oestrogen levels, and this in turn may be related to risk of cancer. It is not clear where The Sun’s claim that aspirin can cut the risk of both breast and ovarian cancer by up to 10% originated. The Sun also did not mention that regular aspirin use is associated with the risk of serious side effects such as internal bleeding.
Neither newspaper mentioned that this was a cross-sectional analysis, and so cannot prove that current painkiller-use is the cause of current hormone levels.
What kind of research was this?
This research investigated whether there is a relationship between the use of painkillers in postmenopausal women and their levels of oestrogen and androgens (male hormones). It is a cross-sectional analysis of data from the Nurses’ Health Study, a long-running study that enrolled 121,700 female nurses in 1976.
As it is a cross-sectional analysis, it can only indicate that there is an association between things, not whether one causes the other (so whether painkiller-use affected hormone levels).
The researchers say that, to date, research in this area has had confusing findings. They say that although some epidemiologic studies have indicated that regularly using aspirin, non-steroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen and paracetamol is associated with a reduced risk of breast and ovarian cancer, overall “the data are inconclusive and the potential mechanisms are unclear”. They refer to some studies that have suggested that, generally, aspirin and non-aspirin NSAIDs (but not paracetamol) decrease breast cancer risk by about 12-25%, but also others that have not shown a link for ovarian cancer.
This new study is limited in that it cannot tell us the women’s use of painkillers over time, and if use related to their hormone levels before they reached menopause. Although the researchers have concluded that the lower oestrogen levels of aspirin users may contribute to lower risk of hormone-related cancers (i.e. breast and ovarian cancer) this has not been investigated and cancer outcomes have not been followed.
What did the research involve?
This study used data on participants from the Nurses’ Health Study, which enrolled 121,700 female nurses, between the ages of 30 and 55, in 1976. The nurses completed lifestyle and medical history questionnaires at enrolment, and every two years afterwards. In 1989-90, a subset of 32,826 women also gave a blood sample and were asked about their menopausal status.
From this subset, these researchers selected 740 postmenopausal women (average age, 61.5) who had not used hormone therapy in the past three months, had no history of cancer and had reported their painkiller-use in the most recent questionnaire (1988 or 1990). The questionnaires had recorded the women’s frequency of aspirin, paracetamol, and other anti-inflammatory drug use (never, 1-4, 5-14, 15-21, or 22 or more days per month) and the usual number of tablets taken per day (0, 1, 2, 3-4, 5-6, or 7 or more). The blood sample was used to measure hormone levels.
The researchers analysed the association between painkiller-use and hormone levels, adjusting for a variety of possible confounders including menstrual and obstetric history, smoking and alcohol, physical activity, BMI, age at the time of sample blood sample and time of day the sample was taken.
What were the basic results?
Of the 740 postmenopausal women, 31% were regular aspirin users, 19% were regular users of non-steroidal anti-inflammatory drugs (NSAID) and 17% were regular users of paracetamol. Regular painkiller-use was slightly more common in the 1990 questionnaire compared with the 1988 questionnaire.
The researchers noted a trend in 1988 towards higher number of aspirin tablets used per month and lower oestrogen hormone levels.
The 1990 questionnaire showed no association between any frequency of use of any type of painkiller and hormone levels. However, when the use of aspirin and NSAIDs was combined, the researchers found that women who used these drugs on 15 or more days a month tended to have lower oestrogen levels (about 10.5% lower than women who never used the drugs). When looking at the women’s use of all three painkillers together, frequency of use of any painkiller was also inversely associated with oestrogen levels.
How did the researchers interpret the results?
The researchers conclude that postmenopausal women who regularly use aspirin and other analgesics may have lower oestrogen levels than women who never use the drugs. They say this could “contribute to a decreased risk of breast or ovarian cancer among analgesic users”.
This research has observed that postmenopausal women who regularly use aspirin, NSAIDs or paracetamol may have lower oestrogen hormone levels than women who never or rarely use the painkillers.
However, this study can only demonstrate trends and associations. It cannot prove that one thing caused the other. This research incorporated two cross-sectional analyses in which the women’s hormone levels were measured at the same time as their painkiller-use was assessed. As such, the results cannot show which came first, or suggest that painkiller-use actually affected hormone levels. It is unclear if the women’s current painkiller use represented a consistent pattern of use over time (i.e. whether they had used painkillers more or less frequently in preceding years). As these results were in postmenopausal women, it is also unknown whether any relationship between frequency of painkiller use and lower hormone levels in this group would be the same as that observed in premenopausal women.
The results themselves in the two data sets, 1988 and 1990, were also quite variable. Although there was an association between increased aspirin use in 1988 and lower oestrogen levels, none of the individual painkillers were associated with hormone levels in 1990.
As the authors themselves say, their study was not powered (did not have enough participants) to detect small differences in hormone levels. This was particularly true for the analyses of the highest frequency of analgesic use, for which only a few women were applicable.
Importantly, this study does not provide any evidence that aspirin, anti-inflammatory drugs or paracetamol reduces the risk of hormone-related cancers such as breast or ovarian cancer. This study did not examine cancer outcomes in these women.