|Home | About | Journals | Submit | Contact Us | Français|
Results of 3 previous studies suggest that use of aspirin is related to a reduced risk of endometrial cancer, at least in obese women (body mass index, ≥30 kg/m2). Using data obtained in a population-based, case-control study in western Washington State, the authors examined this question. Between 2003 and 2005, 410 women diagnosed with invasive endometrial cancer and 356 controls were interviewed regarding the use of aspirin and other nonsteroidal antiinflammatory drugs (NSAIDs). A history of use of NSAIDs was not associated with the risk of endometrial cancer (odds ratio=1.04, 95% confidence interval: 0.76, 1.42). The lack of association was also present specifically for use of aspirin (odds ratio=1.06, 95% confidence interval: 0.73, 1.53). NSAID use was unrelated to risk of endometrial cancer in both obese and nonobese women. Results from this study suggest that the use of aspirin or other NSAIDs has little or no influence on the risk of endometrial cancer.
Beyond the endocrinologic stimulation of endometrial proliferation, exogenous and endogenous estrogen can produce inflammation in this tissue (1), and the latter could be one reason for the adverse effect of estrogen on the risk of endometrial cancer. Conceivably, the use of antiinflammatory medication could act to reduce endometrial cancer risk.
Four previous studies have examined the relation between the use of analgesics and the risk of endometrial cancer, with inconsistent results (2–5). Two of these studies did not observe an overall association but noted that use of aspirin was associated with a reduced risk of the disease in obese women (body mass index, ≥30 kg/m2) and an increased risk in nonobese women (body mass index, <30 kg/m2) (4, 5). A third study observed that, in both obese and nonobese women, there was a reduced risk associated with the use of nonsteroidal antiinflammatory drugs (NSAIDs) or aspirin (3), whereas the results of the most recent study were null in all body mass index subgroups (2).
Using data from a population-based, case-control study, we examined the association between the use of NSAIDs and the risk of endometrial cancer.
A population-based, case-control study was conducted among residents of King, Pierce, and Snohomish counties in western Washington State. Cases were identified through the Cancer Surveillance System, a population-based tumor registry affiliated with the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute (6). Cases were women aged 50–74 years who were diagnosed with invasive endometrial cancer between July 1, 2003, and November 31, 2005. Of the 586 eligible cases, 12 could not be located, 34 were deceased before they could be contacted, and 130 refused to participate or their physician instructed us not to contact them. A total of 410 (70.0%) women were interviewed. Controls were a subset of the controls enrolled in a population-based, case-control study of ovarian cancer that has been described previously (7). Briefly, these women, aged 35–74 years and residents of a 13-county area in western Washington State, were selected by random digit dialing and frequency matched to the age and county of residence distribution of ovarian cancer cases (8). The overall control response proportion (screening response × interview response) was 69.0% (7). From these controls, women aged 50–74 years with an intact uterus, who resided in King, Pierce, or Snohomish counties and who were interviewed during the same time frame as the endometrial cancer cases, were included in the current study. Of the 365 eligible controls, 9 women refused to participate, and 356 women were interviewed.
All women provided signed, informed consent before participating in the study, and an in-person interview was conducted according to a standard protocol approved by the Institutional Review Board of the Fred Hutchinson Cancer Research Center (Seattle, Washington). Interviews pertained to events that occurred before a woman's diagnosis (cases) or a randomly selected date, reference date, during 2003–2005 (controls) and covered demographic factors; height and weight at different ages; family history of cancer; reproductive and menstrual history; history of contraceptive and noncontraceptive hormone use; history of several chronic conditions; and prescription and nonprescription medication use for pain relief, inflammation, and disease prevention. Lists of common analgesic drug names, with generic and specific brands, were used as memory aids, and color pictures of oral contraceptive and menopausal hormone therapy pills were also used to aid recall.
For the purposes of the analysis, never users were women who never used any type of NSAIDs for more than 5 days per month for at least 6 months. Participants were considered users of analgesics only if they had taken a specific analgesic drug for 5 or more days per month for at least 6 consecutive months, similar to previous studies (4, 5). For each of the analgesics, women were asked about the duration of use, age at first use, and starting and stopping dates of use. No information on dose of analgesics was obtained. In our analysis, we excluded analgesic use within a year of diagnosis/reference date to avoid the possibility that undiagnosed endometrial cancer had influenced the intake of analgesics. NSAIDs included aspirin, ibuprofen, naproxen, and cyclooxygenase-II enzyme (COX-2) inhibitors, as well as less common types, for example, Clinoril (Merck & Co., Inc., Whitehouse Station, New Jersey). As a comparison, we also looked at use of acetaminophen, a common analgesic without the antiinflammatory properties of NSAIDs.
Unconditional logistic regression was used to compute odds ratios and corresponding 95% confidence intervals. Odds ratios were adjusted for the matching variables (age at diagnosis/reference date categorized in 5-year intervals; county of residence as either King, Pierce, or Snohomish; and year of diagnosis/reference as either 2003, 2004, or 2005), as well as for body mass index 5 years before the diagnosis/reference date (continuous) and menopausal hormone use. Women were categorized as users of menopausal hormones if their use exceeded 6 months. Never users of menopausal hormones included women who never used any type of menopausal hormone therapy for greater than 6 months. Low risk users included women who had used menopausal hormone therapy but had never used unopposed estrogen and had never used sequential therapy of estrogen plus progestin with fewer than 10 days of progestin per month. Medium risk users included women who used unopposed estrogen for 0.5–<4 years regardless of recency, unopposed estrogen for 4–8 years ending more than 2 years prior to the diagnosis/reference date, or sequential therapy of estrogen and progestin (with fewer than 10 days of progestin per month) for 12 or fewer years but had never used a high-risk type of hormone therapy. High risk users included women who used unopposed estrogen for 4–8 years within 2 years of the diagnosis/reference date, unopposed estrogen for greater than 8 years regardless of recency, or sequential therapy of estrogen and progestin (with fewer than 10 days of progestin per month) for greater than 12 years, regardless of recency. If a woman met the criteria for 2 of the risk categories, she was assigned to the higher one. Adjustment for additional potential confounders such as oral contraceptive use, education, or smoking did not affect the odds ratios, and these variables were not included in the final analysis. All analyses were done by using STATA, version 10.1, software (StataCorp LP, College Station, Texas) for Macintosh.
A total of 410 cases and 356 controls were included in our study. Characteristics of the participants are shown in Table 1. Cases were more likely to have a higher body mass index, not to have used oral contraceptives, and to have taken a high risk menopausal hormone regimen or none at all. A total of 382 women (51.3%) used at least one type of NSAID for at least 5 days per month for 6 consecutive months prior to 1 year before the diagnosis/reference date. The most common type of NSAID used was aspirin (controls: 34.9%), followed by ibuprofen (controls: 28.4%), a COX-2 enzyme inhibitor (controls: 10.5%), and naproxen (controls: 6.8%). Acetaminophen was used (for at least 5 days per month for 6 months, prior to 1 year before the reference/diagnosis date) by 22.5% of the study controls.
Women who were users of any type of NSAID, including aspirin, were not at reduced risk of endometrial cancer compared with never users of any type of NSAID (odds ratio (OR)=1.04, 95% confidence interval (CI): 0.76, 1.42) (Table 2). The association was absent irrespective of the duration of use, the age at first use, or the time since first or last use of NSAIDs. Use of neither aspirin nor acetaminophen altered the risk of endometrial cancer (aspirin: OR= 1.06, 95% CI: 0.73, 1.53; acetaminophen: OR=1.11, 95% CI: 0.70, 1.77).
Table 3 shows the association of endometrial cancer and NSAID use separately among women with a high body mass index (≥30 kg/m2) and other women. In neither group was the use of NSAIDs associated with a reduced risk of endometrial cancer. This lack of association was observed even when NSAIDs had been used for over 10 years or within 5 years of the diagnosis/reference date. Similarly, use of aspirin was not related to risk of endometrial cancer in either obese women (OR=1.17, 95% CI: 0.60, 2.27) or nonobese women (OR=1.00, 95% CI: 0.64, 1.57).
Although there was some variation in the size of association between use of NSAIDs and endometrial cancer according to category of hormone use (Table 4), the numbers of women in the subgroups were too small for any firm conclusion to be reached.
No association between use of NSAIDs and endometrial cancer risk was observed in our study, in either obese or nonobese women. However, the study has several limitations. Because of its sample size, a true but relatively small influence of the use of NSAIDs on endometrial cancer would have been difficult to detect. Nevertheless, the numbers of cases and controls in our study were relatively similar to those in the studies of Fortuny et al. (3) (469 cases and 467 controls) and Moysich et al. (4) (427 incident endometrial cancer cases and 427 controls). Furthermore, the prevalence of NSAID use in our study population (about 50%) was relatively high. Our study was interviewed based, and though memory aids were used to help reduce misclassification of exposure, the number of brands, different doses, and the complexity of the formulations of different analgesics make it unlikely that ascertainment of use of NSAIDs was perfectly precise. Another limitation of the current study is our lack of success in obtaining information from all eligible cases and controls. It is possible that women who were able and who agreed to participate differ in some relevant aspects from women who did not. Finally, we attempted to control for known confounders, but it is possible that unknown or unmeasured variables related to the use of NSAIDs and endometrial cancer might have influenced our results.
The present study differs in several respects from previous ones (3–5). In this study, analgesic use was considered up to 1 year before the diagnosis/reference date to avoid including women who initiated NSAID use during a period when the disease might have been present but undiagnosed. Another distinction is that never users in the current study were defined as women who did not use any type of analgesic for more than 4 days per month for at least 6 months. However, the reference group in other studies included participants who were taking analgesics other than aspirin (4, 5), NSAIDs (3), or both (2). Despite this difference, under the assumption that the use of NSAIDs lowers the risk of endometrial cancer, we should have observed a larger decrease in risk among NSAID users than formerly reported.
Obesity leads to several proinflammatory processes (9) and also increases the levels of circulating estrogens (10). Because of this, it is possible that a reduced risk of endometrial cancer associated with the use of NSAIDs might be evident among obese women only. The current study did not observe a reduced risk in users of NSAIDs, irrespective of body mass index, similar to the findings of one prior study (2). This result differs from other previous studies, which observed either a reduced risk of the disease in obese women and an increased risk in nonobese women associated with the use of aspirin (4, 5), or a reduced risk in both obese and nonobese women in users of NSAIDs (3).
Another well-characterized risk factor for endometrial cancer is the use of certain forms of estrogen-dominant menopausal hormone therapy. However, only the current study and the study by Viswanathan et al. (5) looked at whether there was reduced risk related to the use of NSAIDs among women taking hormone therapy. In our study, the use of NSAIDs among women who had used a form of hormone therapy categorized as either moderate risk or high risk of endometrial cancer was related to a modest reduction in risk, albeit with wide confidence bounds. In contrast, Viswanathan et al. observed the inverse association with use of aspirin to be most prominent among never users of menopausal hormone therapy.
From the foregoing, it is clear that the possible relation of the use of aspirin or other NSAIDs to the incidence of endometrial cancer is far from settled.
Author affiliations: Department of Epidemiology, School of Public Health, University of Washington, Seattle, Washington (Clara Bodelon, Chu Chen, Mary Anne Rossing, Noel S. Weiss); and Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington (Jennifer A. Doherty, Chu Chen, Mary Anne Rossing, Noel S. Weiss).
This work was supported by the Cancer Epidemiology Biostatistics Training Grant (grant T32 CA009168) from the National Institutes of Health and the National Cancer Institute (grant R01 CA105212-03).
Conflict of interest: none declared.