Abstract
Lifestyle-related factors influence risk of endometrial and ovarian cancers, but few studies have examined their joint associations with risk of these cancers. Using multivariable Cox regression models, we assessed the association of a healthy lifestyle index (HLI—a composite score (range, 0–20) involving diet, alcohol consumption, physical activity, body mass index, and smoking; higher scores represent healthier behavior) with risk of endometrial and ovarian cancers among 108,136 postmenopausal women who were recruited in the US Women’s Health Initiative study between 1993 and 1998. After a median follow-up of 17.9 years, 1,435 endometrial cancer cases and 904 ovarian cancer cases had been ascertained. Women in the highest quintile of the HLI score had a lower risk of overall, type I, well-differentiated, moderately differentiated, poorly differentiated, and localized endometrial cancer than those in the lowest quintile (for quintile 5 vs. quintile 1, hazard ratio (HR) = 0.61 (95% CI: 0.51, 0.72), HR = 0.60 (95% CI: 0.49, 0.72), HR = 0.66 (95% CI: 0.46, 0.96), HR = 0.69 (95% CI: 0.52, 0.90), HR = 0.49 (95% CI: 0.34, 0.72), and HR = 0.61 (95% CI: 0.50, 0.74), respectively). The HLI score had a weak positive association with risk of serous ovarian cancer. Our findings underscore the potential importance of a healthy lifestyle in lowering endometrial cancer risk among postmenopausal women.
Keywords: alcohol intake, BMI, diet score, endometrial cancer, healthy lifestyle index score, ovarian cancer, physical activity, smoking
Epidemiologic evidence suggests that specific lifestyle-related factors might influence risk of endometrial cancer (1–4). Obesity, in particular, has been strongly associated with increased risk of endometrial cancer (1, 2), possibly through its effects on levels of circulating estrogens and inflammatory factors (5–9). Further, some studies have suggested that foods with high glycemic load might be associated with increased risk of this cancer (1, 2), while lifestyle-related risk factors with antiestrogenic and/or antioxidant properties, such as physical activity and smoking, might be associated with lower risk (1, 3, 4, 10, 11).
It has not been well-established whether lifestyle-related factors are associated with ovarian cancer, but recent findings from the World Cancer Research Fund Continuous Update Report suggest that obesity might be associated with increased risk of ovarian cancer, particularly mucinous invasive ovarian cancer and low-grade serous ovarian cancer (12, 13). Moreover, in the Women’s Health Initiative (WHI) cohort, we provided evidence indicating that a low-fat dietary pattern might be inversely associated with ovarian cancer risk (14), and, in some studies, carotenoids and phytoestrogens have also been associated with a reduced risk of ovarian cancer (15–17). A few studies have also shown an inverse association between physical activity and ovarian cancer risk, although the associations were weak (3, 4, 12).
An individual’s lifestyle habits typically cluster (18, 19). In this respect, existing evidence purports that, in combination, lifestyle-related factors might contribute to a greater increase/decrease in risk of chronic diseases (e.g., cardiovascular diseases) than that associated with each factor individually (18, 19). However, only a few studies, using various lifestyle indices, have assessed the combined association of lifestyle-related risk factors—namely diet, alcohol consumption, physical activity, obesity, and smoking—with risk of cancers of the endometrium and ovary. Irrespective of the lifestyle index used, studies have consistently associated an overall healthy lifestyle with a reduced risk of endometrial cancer (20–23), but no associations have been observed for ovarian cancer (20–23). Differences between the associations of individual lifestyle-related factors and risk of endometrial and ovarian tumor subtypes have also been reported (6, 24, 25). However, to our knowledge, no study has investigated the combined association of these risk factors with histopathological subtypes of these cancers.
To advance our knowledge of the joint association of lifestyle-related factors with risk of endometrial and ovarian cancers, we examined the association of a healthy lifestyle index (HLI) with the risk of endometrial and ovarian cancers among women in the WHI cohort.
METHODS
Study population and design
Details of the WHI trial design and primary results have been previously published (26). Briefly, the WHI study comprised 161,808 postmenopausal women, aged 50–79 years, from major racial/ethnic groups, who were recruited at 40 US clinical centers between 1993 and 1998 (26). The WHI included an Observational Study and 4 overlapping clinical trials, including 2 hormone-therapy trials (estrogen alone or estrogen plus progesterone), a low-fat dietary modification trial, and a calcium and vitamin D supplementation trial (26). Since the completion of the original study in 2005, WHI Extension Studies (2005–2010, 2010–2020) have been initiated to gather follow-up data. All participants provided written informed consent. The study was approved by human subject review committees at the participating institutions.
For the present study, all women in the intervention group of the dietary modification arm (n = 19,541) were excluded because they were required to make dietary changes (i.e., reduce their intake of energy-dense foods while increasing their intake of fruits and vegetables and grain products), which would have skewed the diet score estimates (27). Participants were also excluded if they: 1) had implausible energy intake (i.e., <600 kcal or >5,000 kcal; n = 4,543) or 2) did not have information on follow-up time (n = 409). For analyses focused on endometrial cancer, we additionally excluded women who had history of hysterectomy (n = 52,534) or endometrial cancer (n = 4,658) at enrollment, leaving a total of 80,123 women available for analyses. For analyses of ovarian cancer, women with a history of bilateral oophorectomy (n = 26,866) or ovarian cancer (n = 3,266) at enrollment were additionally excluded, leaving a total of 107,183 women available for analyses.
Exposure and covariate ascertainment
Information on sociodemographic characteristics, menstrual and reproductive history, exogenous hormone use, anthropometric characteristics, family history, medical history, lifestyle factors, and dietary factors was collected at enrollment. A 122-item self-administered food frequency questionnaire was used to evaluate the participants’ dietary intake (28). Participants were required to record their usual frequency of intake (from “never or less than once per month” to “2+ per day” for foods and “6+ per day” for beverages) and portion size (small, medium, or large compared with the stated medium portion size). The food frequency questionnaire has been shown to be reliable, with intraclass correlation coefficients of 0.67 for retinol, vitamin C, and vitamin B12; 0.82 for fiber; 0.84 for magnesium; 0.92 for alcohol; and 0.74 for percentage of energy from fat (mean intraclass correlation coefficient = 0.76) (28). With respect to history of cigarette smoking, current and former smokers reported the age at smoking initiation, number of cigarettes smoked daily, and years of smoking; former smokers additionally reported age at quitting smoking. Weight and height were measured by trained staff at baseline. Body mass index (BMI) was computed as weight in kilograms divided by height in meters squared and categorized according to the World Health Organization’s criteria (29). Physical activity was summarized in metabolic equivalent-hours/week by multiplying the number of hours per week of leisure-time physical activity by the metabolic equivalent value of the activity and summing over of all types of activities (30).
Healthy lifestyle index
The HLI was developed based on existing scientific knowledge and on public health guidelines for cancer prevention (21, 31–35). The score is a combination of 5 common lifestyle-related factors—including diet, alcohol consumption, physical activity, BMI, and smoking—that have been associated with risk of chronic diseases including cancer (21, 31–35). For the dietary component, energy-adjusted deciles of 6 dietary components (cereal fiber, red and processed meat, the ratio of polyunsaturated to saturated fat, trans-fats, glycemic load, and fruits and vegetables) were created using the residual method (34, 36). The deciles were scored from 0 (lowest decile) to 9 (highest decile) (and vice-versa for red/processed meat, trans-fat, and glycemic load). The individual scores were then summed and categorized into quintiles (37). The healthy lifestyle index score was then constructed by summing the scores of diet (5th quintile = 4, 4th quintile = 3, 3rd quintile = 2, 2nd quintile = 1, 1st quintile = 0) and other lifestyle factors (smoking: never smoked = 4, former smoker ≤15 pack years = 3, former smoker >15 pack years = 2, current smoker ≤15 pack years = 1, current smoker >15 pack years = 0; alcohol intake: <6.0 g/day = 4, 6.0–11.9 g/day = 3, 12.0–24.9 g/day = 2, 24.0–59.9 g/day = 1, ≥60 g/day = 0; physical activity based on metabolic equivalent tasks: 5th quintile = 4, 4th quintile = 3, 3rd quintile = 2, 2nd quintile = 1, 1st quintile = 0; and BMI: <25.0 = 4, 25.0–29.9 = 3, 30.0–34.9 = 2, 35.0–39.9 = 1, ≥40.0 = 0). The final score ranged from 0 to 20, with 20 being the healthiest behaviors. The healthiest behavior was characterized by consuming a healthy diet (5th quintile), avoidance of smoking, avoidance of alcohol, high physical activity level (5th quintile), and a normal BMI (<25).
Outcome ascertainment
The outcomes were primary invasive endometrial and ovarian cancers. Information on endometrial and ovarian cancers was collected semiannually in the clinical trials groups and annually in the Observational Study group, using in-person, mailed, or telephone questionnaires. Cancer diagnoses and tumor characteristics (histological subtype, grade, and stage) were then adjudicated centrally by trained physicians, who reviewed medical records and pathology reports. Endometrial and ovarian cancer histological subtypes were defined in accordance with International Classification of Diseases for Oncology, Third Edition. For endometrial cancer, type I tumors included adenocarcinoma (not otherwise specified) or endometrioid adenocarcinoma, while type II tumors included papillary, clear cell, and serous adenocarcinomas, as well as carcinosarcomas. Histological subtypes for ovarian cancer included serous tumors and nonserous tumors, namely endometrioid, clear cell, mucinous, and other-epithelial subtypes. Tumor grade and stage were coded using the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) coding system (38). Tumor grade was classified as well, moderately, or poorly differentiated. Due to the small number of well-differentiated ovarian tumors (n = 25), this group was not included in the subtype analyses. Tumor stage was classified as localized or regional/distant metastatic. Vital status was collected through follow-up with participants and proxies and linkage to the National Death Index.
Statistical analyses
Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals for the associations between the HLI score (categorized by quintiles; participants who did not have complete information on the individual components were excluded from analyses involving the HLI) and risk of invasive endometrial and ovarian cancers. Women were followed up from their date of enrollment until the date of diagnosis of endometrial or ovarian cancer, and noncases contributed person-time from their date of enrollment until date of death, date of withdrawal from the study, date of hysterectomy (for endometrial cancer analyses), or until the end of follow-up (September 30, 2016), whichever came first. Participants were censored (noncases) if they died, withdrew from the study before the end of follow-up or had a hysterectomy during follow-up (for endometrial cancer analyses), or did not develop endometrial or ovarian cancer by the end of follow-up. After a median follow-up time of 17.9 years (interquartile range, 9.0–19.4), 1,435 endometrial cancer cases and 904 ovarian cancer cases had been diagnosed. Regression models were adjusted for age at baseline (continuous), ethnicity (white, black, Hispanic, other), education (high school or less, postsecondary or some college, graduate school or some graduate school), nonalcohol energy intake (continuous), age (years) at menarche (>12, 12–13, ≥14), parity (never been pregnant or no term pregnancy, 1, 2, 3, ≥4), combined estrogen and progestin therapy (never, former, current), unopposed estrogen therapy (never, former, current), oral contraceptive use (yes/no), age (years) at menopause (>45, 45–54, ≥55), and family history of endometrial or ovarian cancer (yes/no). The association of the HLI score with overall risk of endometrial and ovarian cancer among the subgroup of women with available clinicopathological information was also examined. For the subgroup analyses, we censored subtypes that were not in the event group of interest. Joint Cox proportional hazards models were created to simultaneously compare hazard ratios for the association between the HLI score and risk of endometrial or ovarian cancer subtypes, and the difference in these associations across subtype was assessed using a Wald test (39).
Given that hormone therapy (HT) use is a risk factor for endometrial and ovarian cancers (6, 40), we also performed analyses stratified by baseline HT status to determine whether HT use is an effect modifier for the association between the HLI score and risk of endometrial and ovarian cancers. For the stratified analyses, the stratification variable was excluded from the multivariable models. P values for interaction were computed by introducing an interaction term in the regression models and testing its coefficient with the Wald test.
In analyses involving the individual components of the HLI score, the models adjusted for the aforementioned covariates as well as the other individual components of the score.
P values for trend were calculated including the ordinal HLI variable as a continuous variable in the regression models. Use of Schoenfeld residuals showed that the proportional hazards assumption was not violated. In sensitivity analyses to assess the possibility of reverse causation, women who developed endometrial or ovarian cancer within 2 years of enrollment were excluded.
All P values were 2-sided. All statistical analyses were performed using Stata, version 14.1 (StataCorp LLC, College Station, Texas).
RESULTS
Table 1 provides a summary of the study population’s characteristics. Women with an HLI in the healthiest behavior category were slightly older, were more likely to have postcollege education and be current HT users, and had lower energy intake than those in the other HLI categories.
Table 1.
Characteristics of Study Population According to Healthy Living Index Score, Women’s Health Initiative Study, United States, 1995–2016
| Characteristic | Healthy Living Index Score | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| ≤10 | 11–12 | 13 | 14–15 | ≥16 | ||||||
| No. | % | No. | % | No. | % | No. | % | No. | % | |
| Age at entry, yearsa | 62 (57–68) | 63 (57–69) | 63 (58–69) | 64 (58–69) | 64 (57–70) | |||||
| Ethnicity | ||||||||||
| White (not of Hispanic origin) | 22,316 | 83.8 | 19,174 | 84.8 | 10,273 | 85.1 | 18,410 | 84.6 | 15,565 | 83.0 |
| Black or African-American | 2,699 | 10.1 | 1,823 | 8.1 | 820 | 6.8 | 1,391 | 6.4 | 999 | 5.3 |
| Other | 1,573 | 5.9 | 1,552 | 6.8 | 946 | 7.8 | 1,904 | 8.7 | 2,130 | 11.4 |
| Missing | 51 | 0.2 | 50 | 0.2 | 36 | 0.3 | 61 | 0.3 | 59 | 0.3 |
| Postcollege education | 6,026 | 22.6 | 6,163 | 27.3 | 3,467 | 30.2 | 7,330 | 33.7 | 7,280 | 38.8 |
| Age at menarche of <12 years | 6,496 | 24.4 | 4,782 | 21.2 | 2,542 | 21.1 | 4,355 | 20.0 | 3,734 | 19.9 |
| Nulliparous | 2,970 | 11.2 | 2,603 | 11.5 | 1,415 | 11.7 | 2,637 | 12.1 | 2,258 | 12.0 |
| Age at menopause of ≥55 years | 3,054 | 11.5 | 2,855 | 12.6 | 1,549 | 12.8 | 2,877 | 13.2 | 2,527 | 13.5 |
| WHI enrollment | ||||||||||
| OS group | 15,610 | 58.6 | 14,279 | 63.2 | 8,049 | 66.7 | 15,424 | 70.9 | 14,583 | 77.8 |
| CT group | 11,026 | 41.4 | 8,320 | 36.8 | 4,026 | 33.3 | 6,342 | 29.1 | 4,170 | 22.2 |
| HT status | ||||||||||
| Never user | 14,425 | 54.2 | 11,257 | 49.9 | 5,823 | 48.3 | 10,067 | 46.3 | 8,489 | 45.3 |
| Former user | 4,023 | 15.1 | 3,352 | 14.8 | 1,813 | 15.0 | 3,243 | 14.9 | 2,733 | 14.6 |
| Current user | 8,179 | 30.7 | 7,973 | 35.3 | 4,426 | 36.7 | 8,440 | 38.8 | 7,514 | 40.1 |
| Oral contraceptives | 11,902 | 44.7 | 9,517 | 42.1 | 5,026 | 41.6 | 8,988 | 41.3 | 7,398 | 39.5 |
| Nonalcohol energy intake, kcal/daya | 1,766.2 (1,370.0–2,239.4) | 1,586.3 (1,234.1–2,003.0) | 1,495.6 (1,167.5–1,891.7) | 1,420.0 (1,112.4–1,787.4) | 1,308.8 (1,032.1–1,634.4) | |||||
Abbreviations: CT, clinical trial; HT, hormone therapy; OS, observational study; WHI, Women’s Health Initiative.
a Values are expressed as median (interquartile range).
Table 2 shows the association between the HLI score and risk of endometrial cancer overall and according to clinicopathological characteristics. Compared with women in the lowest category of the HLI score (≤10), women in the highest quintile (≥16) had a 39% lower risk of endometrial cancer overall (hazard ratio (HR) = 0.61, 95% confidence interval (CI): 0.51, 0.72). Similarly, the uppermost quintile of the HLI score was inversely associated with risk of type 1 (HR = 0.60, 95% CI: 0.49, 0.72), well-differentiated (HR = 0.66, 95% CI: 0.46, 0.96), moderately differentiated (HR = 0.69, 95% CI: 0.52, 0.90), poorly differentiated (HR = 0.49, 95% CI: 0.34, 0.72), and localized (HR = 0.61, 95% CI: 0.50, 0.74) tumors. These associations were also observed when considering the continuous HLI score. Inverse but statistically nonsignificant associations were also observed for risk of type II or regional/distant tumors. There was no evidence to suggest heterogeneity in the associations of the score with the clinicopathological characteristics.
Table 2.
Associations Between Healthy Lifestyle Score and Risk of Endometrial Cancer Among Women From the Women’s Health Initiative Study, United States, 1993–2016
| Cancer Type | Healthy Living Index Score | P for Trend | P for Heterogeneity | Continuous, per Unit Increase in Score | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≤10a | 11–12 | 13 | 14–15 | ≥16 | |||||||||||||
| No. of Cases | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | HR | 95% CI | |||
| Overall | 431 | 276 | 0.70 | 0.60, 0.82 | 171 | 0.79 | 0.66, 0.95 | 262 | 0.65 | 0.55, 0.76 | 217 | 0.61 | 0.51, 0.72 | <0.01 | 0.94 | 0.93, 0.96 | |
| Type 1 | 338 | 216 | 0.70 | 0.59, 0.83 | 122 | 0.72 | 0.68, 0.98 | 210 | 0.66 | 0.55, 0.79 | 168 | 0.60 | 0.49, 0.72 | <0.01 | 0.94 | 0.92, 0.96 | |
| Type II | 64 | 44 | 0.75 | 0.51, 1.11 | 32 | 1.01 | 0.66, 1.56 | 41 | 0.70 | 0.47, 1.05 | 39 | 0.76 | 0.50, 1.16 | 0.17 | 0.97 | 0.97 | 0.93, 1.02 |
| Grade | |||||||||||||||||
| Well-differentiated | 84 | 47 | 0.60 | 0.42, 0.86 | 32 | 0.72 | 0.48, 1.10 | 51 | 0.62 | 0.43, 0.88 | 51 | 0.66 | 0.46, 0.96 | 0.04 | 0.94 | 0.90, 0.98 | |
| Moderately differentiated | 160 | 103 | 0.72 | 0.56, 0.92 | 70 | 0.90 | 0.68, 1.20 | 109 | 0.76 | 0.59, 0.98 | 86 | 0.69 | 0.52, 0.90 | 0.02 | 0.97 | 0.94, 0.99 | |
| Poorly differentiated | 100 | 52 | 0.56 | 0.40, 0.79 | 29 | 0.57 | 0.37, 0.86 | 52 | 0.55 | 0.39, 0.77 | 42 | 0.49 | 0.34, 0.72 | <0.01 | 0.75 | 0.92 | 0.89, 0.96 |
| Stage | |||||||||||||||||
| Localized | 338 | 216 | 0.70 | 0.59, 0.83 | 136 | 0.80 | 0.65, 0.98 | 210 | 0.66 | 0.55, 0.79 | 173 | 0.61 | 0.50, 0.74 | <0.01 | 0.94 | 0.93, 0.96 | |
| Regional/distant metastatic | 78 | 50 | 0.71 | 0.49, 1.01 | 32 | 0.83 | 0.55, 1.27 | 50 | 0.71 | 0.49, 1.02 | 43 | 0.70 | 0.47, 1.03 | 0.08 | 0.46 | 0.97 | 0.93, 1.01 |
Abbreviations: CI, confidence interval; HR, hazard ratio.
a Reference category was score of ≤10; adjusted for age at entry, education, nonalcohol energy intake, ethnicity, age at menarche, parity, combined estrogen and progesterone therapy, unopposed estrogen therapy, oral contraceptive use, family history of endometrial cancer, and age at menopause.
Exclusion of obesity from the HLI score attenuated the association between the score and risk of endometrial cancer. However, there was still a tendency towards a reduced risk of endometrial cancer (i.e., overall, type 1, high grade, localized and, to a lesser extent, intermediate tumors) with higher HLI (Web Table 1, available at https://academic.oup.com/aje).
In analyses restricted to women who had never used HT, we observed an even stronger inverse association between the HLI score and risk of endometrial cancer than that seen in the overall study population (Table 3). Among nonusers of combined estrogen and progesterone therapy and nonusers and former users of unopposed estrogen therapy, the associations were also inverse. There was evidence for heterogeneity in the associations between the HLI score and risk of endometrial cancer by HT status overall and combined estrogen and progesterone therapy status (Table 3).
Table 3.
Associations Between the Healthy Lifestyle Index Score and Risk of Endometrial Cancer According to Hormone-Therapy Status Among Women From the Women’s Health Initiative Study, United States, 1993–2016
| HT Status | Healthy Living Index Score | P for Trend | P for Heterogeneity | Continuous, per Unit Increase in Score | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≤10a | 11–12 | 13 | 14–15 | ≥16 | |||||||||||||
| No. of Cases | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | HR | 95% CI | |||
| Overall | |||||||||||||||||
| Never user | 246 | 129 | 0.63 | 0.51, 0.78 | 74 | 0.69 | 0.53, 0.89 | 98 | 0.52 | 0.40, 0.66 | 77 | 0.46 | 0.35, 0.60 | <0.01 | 0.92 | 0.89, 0.93 | |
| Former user | 69 | 33 | 0.53 | 0.35, 0.80 | 22 | 0.61 | 0.38, 0.99 | 47 | 0.72 | 0.49, 1.06 | 37 | 0.67 | 0.44, 1.03 | 0.11 | 0.96 | 0.91, 1.00 | |
| Current user | 116 | 114 | 0.94 | 0.73, 1.22 | 75 | 1.10 | 0.82, 1.48 | 116 | 0.85 | 0.65, 1.11 | 102 | 0.83 | 0.63, 1.10 | 0.17 | <0.01 | 0.98 | 0.96, 1.01 |
| Combined estrogen and progesterone therapyb | |||||||||||||||||
| Never user | 286 | 150 | 0.63 | 0.52, 0.77 | 87 | 0.67 | 0.53, 0.86 | 126 | 0.56 | 0.45, 0.69 | 102 | 0.51 | 0.40, 0.65 | <0.01 | 0.92 | 0.90, 0.94 | |
| Former user | 41 | 24 | 0.63 | 0.38, 1.05 | 21 | 0.99 | 0.58, 1.70 | 33 | 0.85 | 0.53, 1.37 | 25 | 0.70 | 0.42, 1.19 | 0.46 | 0.97 | 0.92, 1.02 | |
| Current user | 104 | 102 | 0.91 | 0.69, 1.20 | 63 | 1.00 | 0.73, 1.38 | 102 | 0.81 | 0.61, 1.07 | 90 | 0.79 | 0.58, 1.06 | 0.07 | 0.01 | 0.98 | 0.96, 1.01 |
| Unopposed estrogen therapyc | |||||||||||||||||
| Never user | 363 | 227 | 0.69 | 0.58, 0.82 | 141 | 0.80 | 0.66, 0.97 | 211 | 0.64 | 0.53, 0.76 | 171 | 0.59 | 0.49, 0.71 | <0.01 | 0.94 | 0.92, 0.96 | |
| Former user | 56 | 37 | 0.65 | 0.43, 0.99 | 18 | 0.52 | 0.30, 0.88 | 37 | 0.58 | 0.38, 0.89 | 33 | 0.56 | 0.35, 0.88 | 0.01 | 0.93 | 0.99, 0.98 | |
| Current user | 12 | 12 | 1.42 | 0.63, 3.21 | 12 | 2.30 | 1.02, 5.17 | 14 | 1.42 | 0.64, 3.17 | 12 | 1.25 | 0.53, 2.93 | 0.56 | 0.23 | 1.04 | 0.95, 1.13 |
Abbreviations: CI, confidence interval; HR, hazard ratio; HT, hormone therapy.
a Reference category was score of ≤10; adjusted for age at entry, education, nonalcohol energy intake, ethnicity, age at menarche, parity, oral contraceptive use, family history of endometrial cancer, and age at menopause.
b Also adjusted for unopposed estrogen.
c Also adjusted for combined estrogen and progesterone therapy.
There was also a tendency towards an increased risk of serous and metastatic ovarian tumors with increasing HLI score. However, no associations were observed with risk of ovarian cancer overall or with risk of the remaining clinicopathological characteristics (Table 4). The association between the HLI and ovarian cancer risk also did not vary by HT use (Table 5).
Table 4.
Associations Between Healthy Lifestyle Score and Risk of Ovarian Cancer Among Women From the Women’s Health Initiative Study, United States, 1995–2016
| Cancer Type | Healthy Living Index Score | P for Trend | P for Heterogeneity | Continuous, per Unit Increase in Score | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≤10a | 11–12 | 13 | 14–15 | ≥16 | |||||||||||||
| No. of Cases | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | HR | 95% CI | |||
| Overall | 143 | 133 | 0.88 | 0.72, 1.07 | 173 | 0.93 | 0.73, 1.19 | 168 | 1.01 | 0.82, 1.23 | 141 | 0.96 | 0.77, 1.19 | 0.84 | 1.00 | 0.98, 1.03 | |
| Nonserous | 50 | 48 | 0.79 | 0.55, 1.11 | 71 | 1.02 | 0.70, 1.49 | 60 | 0.99 | 0.71, 1.37 | 47 | 0.93 | 0.65, 1.32 | 0.87 | 1.00 | 0.97, 1.04 | |
| Serous | 58 | 60 | 1.00 | 0.74, 1.46 | 76 | 1.30 | 0.69, 1.46 | 82 | 1.30 | 0.96, 1.75 | 75 | 1.28 | 0.93, 1.76 | 0.04 | 0.41b | 1.03 | 0.99, 1.07 |
| High-grade serous | 30 | 31 | 0.94 | 0.59, 1.51 | 17 | 0.89 | 0.50, 1.59 | 38 | 1.13 | 0.71, 1.78 | 28 | 1.00 | 0.61, 1.65 | 0.73 | 1.01 | 0.95, 1.06 | |
| Grade | |||||||||||||||||
| Intermediate | 30 | 31 | 1.48 | 0.77, 2.88 | 17 | 1.84 | 0.88, 3.87 | 38 | 1.97 | 1.03, 3.77 | 28 | 1.04 | 0.47, 3.77 | 0.43 | 1.05 | 0.98, 1.13 | |
| High | 50 | 48 | 1.04 | 0.71, 1.52 | 17 | 1.22 | 0.79, 1.88 | 38 | 1.24 | 0.85, 1.80 | 28 | 1.18 | 0.79, 1.76 | 0.25 | 0.55 | 1.02 | 0.98, 1.07 |
| Stage | |||||||||||||||||
| Localized | 26 | 17 | 1.54 | 0.95, 2.49 | 34 | 0.83 | 0.41, 1.67 | 24 | 1.21 | 0.71, 2.05 | 19 | 1.08 | 0.61, 1.93 | 0.91 | 1.00 | 0.94, 1.06 | |
| Regional/distant metastatic | 97 | 93 | 0.89 | 0.69, 1.13 | 132 | 1.08 | 0.82, 1.43 | 123 | 1.17 | 0.92, 1.47 | 116 | 1.19 | 0.93, 1.53 | 0.03 | 0.15 | 1.03 | 1.00, 1.06 |
Abbreviations: CI, confidence interval; HR, hazard ratio.
a Reference category was score of ≤10; adjusted for age at entry, education, nonalcohol energy intake, ethnicity, age at menarche, parity, combined estrogen and progesterone therapy, unopposed estrogen therapy, oral contraceptive use, family history of ovarian cancer, and age at menopause.
bP for heterogeneity between serous and nonserous tumors (excluding high-grade serous tumors).
Table 5.
Associations Between the Healthy Lifestyle Index Score and Risk of Ovarian Cancer According to Hormone-Therapy Status Among Women From the Women’s Health Initiative Study, United States, 1993–2016
| HT Status | Healthy Living Index Score | P for Trend | P for Heterogeneity | Continuous, per Unit Increase in Score | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≤10a | 11–12 | 13 | 14–15 | ≥16 | |||||||||||||
| No. of Cases | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | HR | 95% CI | |||
| Overall | |||||||||||||||||
| Never user | 104 | 78 | 0.93 | 0.69, 1.25 | 43 | 0.99 | 0.69, 1.42 | 68 | 0.89 | 0.65, 1.22 | 56 | 0.87 | 0.62, 1.22 | 0.86 | 0.98 | 0.94, 1.01 | |
| Former user | 31 | 23 | 0.81 | 0.47, 1.40 | 14 | 0.91 | 0.48, 1.72 | 29 | 1.01 | 0.60, 1.71 | 25 | 0.99 | 0.57, 1.74 | 0.76 | 1.02 | 0.96, 1.09 | |
| Current user | 77 | 71 | 0.87 | 0.63, 1.20 | 43 | 0.92 | 0.63, 1.34 | 103 | 1.11 | 0.82, 1.51 | 86 | 1.02 | 0.74, 1.42 | 0.39 | 0.05 | 1.02 | 0.99, 1.06 |
| Combined estrogen and progesterone therapyb | |||||||||||||||||
| Never user | 160 | 116 | 0.84 | 0.66, 1.06 | 68 | 0.91 | 0.68, 1.21 | 129 | 0.97 | 0.76, 1.23 | 109 | 0.96 | 0.74, 1.24 | 0.94 | 0.99 | 0.96, 1.02 | |
| Former user | 18 | 18 | 1.01 | 0.52, 1.95 | 8 | 0.80 | 0.34, 1.86 | 20 | 1.10 | 0.57, 2.13 | 16 | 0.99 | 0.49, 2.03 | 0.90 | 1.03 | 0.96, 1.10 | |
| Current user | 34 | 38 | 1.01 | 0.64, 1.61 | 24 | 1.13 | 0.67, 1.92 | 51 | 1.17 | 0.75, 1.83 | 42 | 1.03 | 0.64, 1.66 | 0.69 | 0.60 | 1.02 | 0.97, 1.08 |
| Unopposed estrogen therapyc | |||||||||||||||||
| Never user | 142 | 121 | 0.95 | 0.75, 1.22 | 73 | 1.08 | 0.81, 1.44 | 125 | 0.99 | 0.77, 1.27 | 97 | 0.88 | 0.67, 1.16 | 0.54 | 0.99 | 0.96, 1.02 | |
| Former user | 28 | 18 | 0.69 | 0.38, 1.25 | 8 | 0.53 | 0.24, 1.18 | 23 | 0.83 | 0.47, 1.47 | 26 | 1.05 | 0.59, 1.86 | 0.72 | 1.03 | 0.96, 1.10 | |
| Current user | 42 | 33 | 0.78 | 0.50, 1.24 | 19 | 0.78 | 0.45, 1.35 | 52 | 1.13 | 0.74, 1.72 | 44 | 1.09 | 0.69, 1.70 | 0.30 | 0.05 | 1.02 | 0.97, 1.08 |
Abbreviations: CI, confidence interval; HR, hazard ratio; HT, hormone therapy.
a Reference category was score of ≤10; adjusted for age at entry, education, nonalcohol energy intake, ethnicity, age at menarche, parity, oral contraceptive use, family history of ovarian cancer, and age at menopause.
b Also adjusted for unopposed estrogen.
c Also adjusted for combined estrogen and progesterone therapy.
Table 6 shows that among the individual components of the HLI score, diet, physical activity, and smoking score were inversely associated with risk of endometrial cancer, while being obese was positively associated with risk. None of the individual components was associated with risk of ovarian cancer.
Table 6.
Associations Between Healthy Lifestyle Score Index Components and Risk of Endometrial and Ovarian Cancersa Among Women From the Women’s Health Initiative Study, United States, 1993–2016
| HLI Component | Endometrial Cancer | Ovarian Cancer | ||||
|---|---|---|---|---|---|---|
| No. of Cases | HR | 95% CI | No. of Cases | HR | 95% CI | |
| Diet score quintiles | ||||||
| ≤20 | 330 | 1.00 | 158 | 1.00 | ||
| 21–25 | 364 | 0.97 | 0.83, 1.13 | 211 | 1.15 | 0.93, 1.43 |
| 26–29 | 245 | 0.79 | 0.66, 0.94 | 184 | 1.21 | 0.96, 1.52 |
| 30–34 | 263 | 0.85 | 0.71, 1.02 | 177 | 1.16 | 0.92, 1.47 |
| >34 | 233 | 0.81 | 0.67, 0.98 | 174 | 1.26 | 0.99, 1.62 |
| P for trend | 0.01 | 0.11 | ||||
| Alcohol, g/day | ||||||
| <0.0 | 251 | 1.00 | 150 | 1.00 | ||
| 0.1–4.9 | 733 | 1.01 | 0.84, 1.27 | 453 | 1.06 | 0.86, 1.30 |
| 5.0–9.9 | 153 | 0.87 | 0.71, 1.07 | 119 | 1.18 | 0.91, 1.51 |
| 10.0–19.9 | 170 | 0.79 | 0.64, 0.98 | 104 | 1.05 | 0.81, 1.37 |
| >19.9 | 128 | 0.93 | 0.79, 1.09 | 78 | 1.22 | 0.92, 1.63 |
| P for trend | 0.54 | 0.20 | ||||
| Physical activity quintiles, MET-hours/week | ||||||
| ≤1.5 | 302 | 1.00 | 155 | 1.00 | ||
| 1.6–6.0 | 260 | 0.94 | 0.80, 1.17 | 175 | 1.21 | 0.98, 1.51 |
| 6.1–12.0 | 282 | 0.91 | 0.77, 1.07 | 174 | 1.10 | 0.88, 1.37 |
| 12.1–21.5 | 263 | 0.91 | 0.77, 1.08 | 180 | 1.22 | 0.98, 1.52 |
| >21.5 | 273 | 0.84 | 0.71, 0.99 | 182 | 1.10 | 0.88, 1.37 |
| Missing | 55 | 38 | ||||
| P for trend | 0.06 | 0.44 | ||||
| BMIb | ||||||
| <25.0 | 454 | 1.00 | 354 | 1.00 | ||
| 25.0–29.9 | 391 | 1.05 | 0.91, 1.21 | 302 | 0.99 | 0.85, 1.17 |
| 30.0–34.9 | 305 | 1.73 | 1.44, 2.06 | 157 | 1.10 | 0.91, 1.34 |
| 35.0–39.9 | 159 | 2.40 | 1.89, 3.05 | 52 | 1.01 | 0.75, 1.36 |
| ≥40 | 115 | 3.18 | 2.28, 4.42 | 33 | 1.29 | 0.89, 1.86 |
| Missing | 11 | 6 | ||||
| P for trend | <0.01 | 0.25 | ||||
| Smoking | ||||||
| Never | 737 | 1.00 | 445 | 1.00 | ||
| Former smoker, ≤15 pack years | 359 | 0.99 | 0.87, 1.12 | 236 | 1.03 | 0.87, 1.20 |
| Former smoker, >15 pack years | 258 | 0.90 | 0.77, 1.05 | 164 | 1.13 | 0.94, 1.36 |
| Current smoker, ≤15 pack years | 23 | 0.93 | 0.61, 1.42 | 17 | 0.88 | 0.54, 1.43 |
| Current smoker, >15 pack years | 44 | 0.72 | 0.53, 0.98 | 33 | 1.14 | 0.80, 1.64 |
| Missing | 14 | 9 | ||||
| P for trend | 0.03 | 0.31 | ||||
Abbreviations: BMI, body mass index; CI, confidence interval; HLI, health living index; HR, hazard ratio; MET, metabolic equivalent.
a Adjusted for age at entry, education, nonalcohol energy intake, ethnicity, age at menarche, parity, combined estrogen and progesterone therapy, unopposed estrogen therapy, oral contraceptive use, family history of endometrial or ovarian cancer, age at menopause, diet, physical activity, alcohol consumption, BMI, and smoking unless included as main exposure.
b Weight (kg)/height (m)2.
Exclusion of women with an endometrial cancer diagnosis within 2 years of enrollment did not alter the association of the HLI score with risk of endometrial cancer overall, or with risk of type 1, poorly differentiated, and localized endometrial cancer (Web Table 2). With respect to ovarian cancer, the associations of the HLI with risk of serous and metastatic ovarian tumors disappeared (Web Table 3).
DISCUSSION
The results of this large prospective study of postmenopausal women suggest that a healthy lifestyle is associated with reduced risk of endometrial cancer overall, as well as of type 1, well-differentiated, moderately differentiated, poorly differentiated, and localized tumors. Similar inverse associations were seen among women who never used HT, as well as among nonusers and former users of opposed estrogen therapy. Diet, physical activity, and smoking were also inversely associated with risk of endometrial cancer while obesity was positively associated with risk. Further, there was a suggestion of a positive association between the HLI score and risk of serous and metastatic ovarian tumors.
To date, only 4 studies have explored the joint association between lifestyle-related risk factors and endometrial cancer (20–23). In the E3N cohort study, which used an HLI similar to that in the present study, having a high HLI score was associated with a 54% reduction in the risk of endometrial cancer (HR = 0.45, 95% CI: 0.29, 0.71) (23). Previously in the WHI cohort, using a lifestyle index based on the American Cancer Society/Cancer Prevention Guidelines, we also demonstrated an inverse association between an overall healthy lifestyle and risk of endometrial cancer (21). Other prospective studies using lifestyle indices based on the American Cancer Society guidelines (20) and on World Cancer Research Fund/American Institute for Cancer Research guidelines (22) also observed that women with the strongest adherence to the guidelines had a 23% and 60% lower risk of endometrial cancer (respectively, HR = 0.77, 95% CI: 0.62, 0.94; and HR = 0.40, 95% CI: 0.34, 0.46). Our findings also suggested that an overall healthy lifestyle might reduce risk of all endometrial cancer histopathological subtypes, although the associations were statistically nonsignificant for some subtypes.
Exogenous hormone use has been shown to alter the risk of endometrial cancer (6, 41). However, it is unknown whether HT use modulates the association between a healthy lifestyle and risk of endometrial cancer. Interestingly, in the present study, we found evidence to suggest that the associations might be modified by HT use, because the inverse associations were strongest among women who had never used any form of HT. However, more studies are needed to confirm our findings.
Epidemiologic evidence to support an association of diet with endometrial cancer is limited (1), but, in agreement with our study, several recent studies reported that a healthy dietary pattern, characterized by high intake of antioxidant-rich foods, was inversely associated with risk of endometrial cancer (42, 43). Similar to our study, others have also indicated that being physically active (1) and smoking are inversely associated with risk of endometrial cancer (44, 45). Previous studies have, however, largely failed to observe an association of alcohol consumption with risk of endometrial cancer (44, 46–48). Our study also confirmed the findings of previous studies that documented a strong positive association between obesity and risk of endometrial cancer (1). Given the strong association between obesity and endometrial cancer, it is not surprising that the results of our sensitivity analyses indicated that the association between the HLI and risk of endometrial cancer was mostly explained by level of adiposity. Nevertheless, there was still evidence to suggest that the remaining modifiable risk factors might collectively influence risk of endometrial cancer, particularly type 1 endometrial cancer.
The observed inverse association between the lifestyle-related risk factors and risk of endometrial cancer might involve a complex interaction between several biological mechanisms. Briefly, excess body fat, diets low in antioxidant-rich foods, relatively high alcohol consumption, and physical inactivity might contribute to several metabolic changes such as increased estrogen levels resulting from enhanced aromatase activity, hyperinsulinemia, increased levels of bioavailable insulin-like growth factor 1, and increased production of inflammatory markers (5–9), which might promote carcinogenesis by inducing oxidative stress, deoxyribonucleic acid damage, and mutagenesis; by inhibiting apoptosis; and by other processes that can foster tumor cell growth, proliferation, and migration (5–9). The mechanisms underlying the inverse association between smoking and endometrial cancer remain unclear, but studies have indicated that smoking might lower risk of endometrial cancer through its antiestrogenic effect (11, 44, 45).
With respect to ovarian cancer, the observed null association between the HLI score and risk of this cancer is consistent with that of the E3N study (23) and with those of studies that used scores based on adherence to the American Cancer Society and/or the World Cancer Research Fund/American Institute for Cancer Research guidelines (21, 22). In the present study, we also observed weak positive associations between the HLI and risk of serous and metastatic ovarian tumors. Unexpectedly, positive but nonsignificant associations were also observed for 2 components of the HLI: diet and physical activity. The positive associations suggest that the beneficial influence of some components of a healthy lifestyle on risk of ovarian cancer (including some subtypes) might be obscured by the influence of other components. Nevertheless, our findings might not be a true estimation of the associations between the HLI index or its components with risk of ovarian cancer, given the heterogeneity in the associations of the component risk factors across and within the various ovarian cancer subtypes (13, 25, 49–51). For example, among the nonserous and serous subtypes, obesity has been associated with risk of mucinous invasive ovarian cancer and low-grade serous ovarian cancer, respectively (13, 50). Further, cigarette smoking has been shown to be inversely associated with risk of clear cell subtypes but positively associated with risk of mucinous subtypes (25, 51).
This study has several strengths, including its large sample size, standardization of the procedures used to collect risk factor information, limited loss to follow-up, and central adjudication of pathology reports. This is also, to our knowledge, the only study to date that has explored whether the association between an overall healthy lifestyle and endometrial or ovarian cancer differs by histopathological subtypes. There are also several limitations that require consideration. Aside from height and weight, the HLI components were self-reported and therefore subject to nondifferential measurement errors. Such error might have precluded us from observing small associations with ovarian cancer risk. Moreover, we were unable to assess how change in exposure status over time influences risk of the outcomes. We also lacked information on oophorectomy during follow-up, which might have contributed to misclassification of follow-up time for individuals who had oophorectomy after baseline. Finally, the number of events for some ovarian cancer subtypes was small. Therefore, our study might not have been adequately powered to assess heterogeneity in the associations between the HLI score and these subtypes.
In conclusion, our study underscores the potential importance of maintaining an overall healthy lifestyle to lower risk of endometrial cancer. However, further studies should be conducted to substantiate our findings; these results might be useful in developing intervention strategies for the primary prevention of endometrial cancer.
Supplementary Material
ACKNOWLEDGMENTS
Author affiliations: Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, New York (Rhonda Arthur, Sylvia Wassertheil-Smoller, Thomas E. Rohan); Division of Cancer Prevention and Control, Department of Internal Medicine, Ohio State University College of Medicine, Columbus, Ohio (Theodore M. Brasky); College of Nursing, University of Arizona, Tucson, Arizona (Tracy E. Crane); Division of Epidemiology, College of Public Health, Ohio State University, Columbus, Ohio (Ashley S. Felix); Department of Obstetrics and Gynecology, University of Florida, Jacksonville, Florida (Andrew M. Kaunitz); Department of Family Medicine and Public Health, School of Medicine, University of California, San Diego, La Jolla, California (Aladdin H. Shadyab); and Division of Biostatistics, Department of Public Health Sciences, University of California, Davis, Davis, California (Lihong Qi).
R.A. and T.E.R. are supported by a grant to T.E.R. from the Breast Cancer Research Foundation (grant BCRF-16-137). The Women’s Health Initiative program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, US Department of Health and Human Services (contracts HHSN268201600018C, HHSN268201600001C, HHSN268201600002C, HHSN268201600003C, and HHSN268201600004C).
We thank the Women’s Health Initiative investigators, staff, and the trial participants for their outstanding dedication and commitment.
Women’s Health Initiative Investigators. Program Office: National Heart, Lung, and Blood Institute, Bethesda, Maryland (Jacques Roscoe, Shari Ludlum, Dale Burden, Joan McGowan, Leslie Ford, and Nancy Geller). Clinical Coordinating Center: Fred Hutchinson Cancer Research Center, Seattle, Washington (Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kopperberg). Investigators and Academic Centers: Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts (JoAnn E. Manson); MedStar Health Research Institute/Howard University, Washington, District of Columbia (Barbara V. Howard); Stanford Prevention Research Center, Stanford, California (Marcia L. Stefanick); Ohio State University, Columbus, Ohio (Rebecca Jackson); University of Arizona, Tucson/Phoenix, Arizona (Cynthia A. Thomson); University at Buffalo, Buffalo, New York (Jean Wactawski-Wende); University of Florida, Gainesville/Jacksonville, Florida (Marian Limacher); University of Iowa, Iowa City/Davenport, Iowa (Robert Wallace); University of Pittsburgh, Pittsburgh, Pennsylvania (Lewis Kuller); City of Hope Comprehensive Cancer Center, Duarte, California (Rowan T. Chlebowski); Wake Forest University School of Medicine, Winston-Salem, North Carolina (Sally Shumaker). Women’s Health Initiative Memory Study: Wake Forest University School of Medicine, Winston Salem, North Carolina (Sally Shumaker). Additional information: A full list of all the investigators who have contributed to Women’s Health Initiative science appears at: https://www.whi.org/researchers/Documents%20%20Write%20a%20Paper/WHI%20Investigator%20Long%20List.pdf.
Conflict of interest: none declared.
Abbreviations
- BMI
body mass index
- CI
confidence intervals
- HLI
healthy lifestyle index
- HR
hazard ratio
- HT
hormone therapy
- WHI
Women’s Health Initiative
REFERENCES
- 1. World Cancer Research Fund/American Institute for Cancer Research Continuous Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Endometrial Cancer 2013; http://www.wcrf.org/sites/default/files/Endometrial-Cancer-2013-Report.pdf. Accessed May 4, 2017.
- 2. Reeves KW, Carter GC, Rodabough RJ, et al. Obesity in relation to endometrial cancer risk and disease characteristics in the Women’s Health Initiative. Gynecol Oncol. 2011;121(2):376–382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hildebrand JS, Gapstur SM, Gaudet MM, et al. Moderate-to-vigorous physical activity and leisure-time sitting in relation to ovarian cancer risk in a large prospective US cohort. Cancer Causes Control. 2015;26(11):1691–1697. [DOI] [PubMed] [Google Scholar]
- 4. Zhong S, Chen L, Lv M, et al. Nonoccupational physical activity and risk of ovarian cancer: a meta-analysis. Tumour Biol. 2014;35(11):11065–11073. [DOI] [PubMed] [Google Scholar]
- 5. Candido J, Hagemann T. Cancer-related inflammation. J Clin Immunol. 2013;33(suppl 1):79–84. [DOI] [PubMed] [Google Scholar]
- 6. Felix AS, Yang HP, Bell DW, et al. Epidemiology of endometrial carcinoma: etiologic importance of hormonal and metabolic influences In: Hedrick Ellenson L, ed. Molecular Genetics of Endometrial Carcinoma. Cham, Switzerland: Springer; 2017:3–46. [DOI] [PubMed] [Google Scholar]
- 7. Park J, Morley TS, Kim M, et al. Obesity and cancer-mechanisms underlying tumour progression and recurrence. Nat Rev Endocrinol. 2014; 10(8):455–465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Giovannucci E. Nutrition, insulin, insulin-like growth factors and cancer. Horm Metab Res. 2003;35(11–12):694–704. [DOI] [PubMed] [Google Scholar]
- 9. McTiernan A. Mechanisms linking physical activity with cancer. Nat Rev Cancer. 2008;8(3):205–211. [DOI] [PubMed] [Google Scholar]
- 10. Schmid D, Behrens G, Keimling M, et al. A systematic review and meta-analysis of physical activity and endometrial cancer risk. Eur J Epidemiol. 2015;30(5):397–412. [DOI] [PubMed] [Google Scholar]
- 11. Baron JA. Smoking and estrogen-related disease. Am J Epidemiol. 1984;119(1):9–22. [DOI] [PubMed] [Google Scholar]
- 12. World Cancer Research Fund/American Institute for Cancer Research Continuous Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Ovarian Cancer, 2014; http://www.dietandcancerreport.org/cup/cup_resources.php. Accessed May 4, 2017.
- 13. Tworoger SS, Huang T. Obesity and ovarian cancer In: Pischon T, Nimptsch K, eds. Obesity and Cancer. Cham, Switzerland: Springer; 2016:155–176. [DOI] [PubMed] [Google Scholar]
- 14. Prentice RL, Thomson CA, Caan B, et al. Low-fat dietary pattern and cancer incidence in the Women’s Health Initiative Dietary Modification Randomized Controlled Trial. J Natl Cancer Inst. 2007;99(20):1534–1543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Qu XL, Fang Y, Zhang M, et al. Phytoestrogen intake and risk of ovarian cancer: a meta- analysis of 10 observational studies. Asian Pac J Cancer Prev. 2014;15(21):9085–9091. [DOI] [PubMed] [Google Scholar]
- 16. Cramer DW, Kuper H, Harlow BL, et al. Carotenoids, antioxidants and ovarian cancer risk in pre- and postmenopausal women. Int J Cancer. 2001;94(1):128–134. [DOI] [PubMed] [Google Scholar]
- 17. Hua X, Yu L, You R, et al. Association among dietary flavonoids, flavonoid subclasses and ovarian cancer risk: a meta-analysis. PLoS One. 2016;11(3):e0151134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Pronk NP, Anderson LH, Crain AL, et al. Meeting recommendations for multiple healthy lifestyle factors: prevalence, clustering, and predictors among adolescent, adult, and senior health plan members. Am J Prev Med. 2004;27(2 suppl):25–33. [DOI] [PubMed] [Google Scholar]
- 19. Pronk NP, Peek CJ, Goldstein MG. Addressing multiple behavioral risk factors in primary care. A synthesis of current knowledge and stakeholder dialogue sessions. Am J Prev Med. 2004;27(2 suppl):4–17. [DOI] [PubMed] [Google Scholar]
- 20. Kabat GC, Matthews CE, Kamensky V, et al. Adherence to cancer prevention guidelines and cancer incidence, cancer mortality, and total mortality: a prospective cohort study. Am J Clin Nutr. 2015;101(3):558–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Thomson CA, McCullough ML, Wertheim BC, et al. Nutrition and physical activity cancer prevention guidelines, cancer risk, and mortality in the Women’s Health Initiative. Cancer Prev Res (Phila). 2014;7(1):42–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Romaguera D, Vergnaud AC, Peeters PH, et al. Is concordance with World Cancer Research Fund/American Institute for Cancer Research guidelines for cancer prevention related to subsequent risk of cancer? Results from the EPIC study. Am J Clin Nutr. 2012;96(1):150–163. [DOI] [PubMed] [Google Scholar]
- 23. Dartois L, Fagherazzi G, Boutron-Ruault MC, et al. Association between five lifestyle habits and cancer risk: results from the E3N cohort. Cancer Prev Res (Phila). 2014;7(5):516–525. [DOI] [PubMed] [Google Scholar]
- 24. Goodman MT, Tung KH. Alcohol consumption and the risk of borderline and invasive ovarian cancer. Obstet Gynecol. 2003;101(6):1221–1228. [DOI] [PubMed] [Google Scholar]
- 25. Gram IT, Lukanova A, Brill I, et al. Cigarette smoking and risk of histological subtypes of epithelial ovarian cancer in the EPIC cohort study. Int J Cancer. 2012;130(9):2204–2210. [DOI] [PubMed] [Google Scholar]
- 26. Women’s Health Initiative Study Group Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials. 1998;19(1):61–109. [DOI] [PubMed] [Google Scholar]
- 27. Ritenbaugh C, Patterson RE, Chlebowski RT, et al. The Women’s Health Initiative Dietary Modification trial: overview and baseline characteristics of participants. Ann Epidemiol. 2003;13(9 suppl):S87–S97. [DOI] [PubMed] [Google Scholar]
- 28. Patterson RE, Kristal AR, Tinker LF, et al. Measurement characteristics of the Women’s Health Initiative food frequency questionnaire. Ann Epidemiol. 1999;9(3):178–187. [DOI] [PubMed] [Google Scholar]
- 29. World Health Organization Obesity and overweight, 2014. http://www.who.int/mediacentre/factsheets/fs311/en/. Accessed July 8, 2014.
- 30. McTiernan A, Kooperberg C, White E, et al. Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women’s Health Initiative cohort study. JAMA. 2003;290(10):1331–1336. [DOI] [PubMed] [Google Scholar]
- 31. Thomson CA, Van Horn L, Caan BJ, et al. Cancer incidence and mortality during the intervention and postintervention periods of the Women’s Health Initiative Dietary Modification trial. Cancer Epidemiol Biomarkers Prev. 2014;23(12):2924–2935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Johnson KC, Miller AB, Collishaw NE, et al. Active smoking and secondhand smoke increase breast cancer risk: the report of the Canadian Expert Panel on Tobacco Smoke and Breast Cancer Risk (2009) [Report]. Tob Control. 2011;20(1):e2. [DOI] [PubMed] [Google Scholar]
- 33. Yang HP, Brinton LA, Platz EA, et al. Active and passive cigarette smoking and the risk of endometrial cancer in Poland. Eur J Cancer. 2010;46(4):690–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. McKenzie F, Biessy C, Ferrari P, et al. Healthy lifestyle and risk of cancer in the European Prospective Investigation into Cancer and Nutrition Cohort Study. Medicine (Baltimore). 2016; 95(16):e2850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. World Cancer Research Fund/American Institute for Cancer Research Diet, Nutrition, Physical Activity and Cancer: a Global Perspective. Continuous Update Project Expert Report, 2018. https://www.wcrf.org/dietandcancer. Accessed June 7, 2018.
- 36. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124(1):17–27. [DOI] [PubMed] [Google Scholar]
- 37. McKenzie F, Ellison-Loschmann L, Jeffreys M, et al. Healthy lifestyle and risk of breast cancer for indigenous and non-indigenous women in New Zealand: a case control study. BMC Cancer. 2014;14:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Curb JD, McTiernan A, Heckbert SR, et al. Outcomes ascertainment and adjudication methods in the Women’s Health Initiative. Ann Epidemiol. 2003;13(9 suppl):S122–S128. [DOI] [PubMed] [Google Scholar]
- 39. Xue X, Kim MY, Gaudet MM, et al. A comparison of the polytomous logistic regression and joint cox proportional hazards models for evaluating multiple disease subtypes in prospective cohort studies. Cancer Epidemiol Biomarkers Prev. 2013;22(2):275–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Hunn J, Rodriguez GC. Ovarian cancer: etiology, risk factors, and epidemiology. Clin Obstet Gynecol. 2012;55(1):3–23. [DOI] [PubMed] [Google Scholar]
- 41. Brinton LA, Felix AS. Menopausal hormone therapy and risk of endometrial cancer. J Steroid Biochem Mol Biol. 2014;142:83–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Filomeno M, Bosetti C, Bidoli E, et al. Mediterranean diet and risk of endometrial cancer: a pooled analysis of three Italian case-control studies. Br J Cancer. 2015;112(11):1816–1821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Ricceri F, Giraudo MT, Fasanelli F, et al. Diet and endometrial cancer: a focus on the role of fruit and vegetable intake, Mediterranean diet and dietary inflammatory index in the endometrial cancer risk. BMC Cancer. 2017;17:757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Zhou B, Yang L, Sun Q, et al. Cigarette smoking and the risk of endometrial cancer: a meta-analysis. Am J Med. 2008;121(6):501–508.e3. [DOI] [PubMed] [Google Scholar]
- 45. Felix AS, Yang HP, Gierach GL, et al. Cigarette smoking and endometrial carcinoma risk: the role of effect modification and tumor heterogeneity. Cancer Causes Control. 2014;25(4):479–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Friedenreich CM, Speidel TP, Neilson HK, et al. Case-control study of lifetime alcohol consumption and endometrial cancer risk. Cancer Causes Control. 2013;24(11):1995–2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Fedirko V, Jenab M, Rinaldi S, et al. Alcohol drinking and endometrial cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Ann Epidemiol. 2013;23(2):93–98. [DOI] [PubMed] [Google Scholar]
- 48. Zhou Q, Guo P, Li H, et al. Does alcohol consumption modify the risk of endometrial cancer? A dose-response meta-analysis of prospective studies. Arch Gynecol Obstet. 2017;295(2):467–479. [DOI] [PubMed] [Google Scholar]
- 49. Crane TE, Khulpateea BR, Alberts DS, et al. Dietary intake and ovarian cancer risk: a systematic review. Cancer Epidemiol Biomarkers Prev. 2014;23(2):255–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Olsen CM, Nagle CM, Whiteman DC, et al. Obesity and risk of ovarian cancer subtypes: evidence from the Ovarian Cancer Association Consortium. Endocr Relat Cancer. 2013;20(2):251–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Collaborative Group on Epidemiological Studies of Ovarian Cancer Ovarian cancer and smoking: individual participant meta-analysis including 28,114 women with ovarian cancer from 51 epidemiological studies. Lancet Oncol. 2012;13(9):946–956. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.