Background: Despite a large body of literature evaluating the association between recreational physical activity and epithelial ovarian cancer (EOC) risk, the extant evidence is inconclusive, and little is known about the independent association between recreational physical inactivity and EOC risk. We conducted a pooled analysis of nine studies from the Ovarian Cancer Association Consortium to investigate the association between chronic recreational physical inactivity and EOC risk.

Methods: In accordance with the 2008 Physical Activity Guidelines for Americans, women reporting no regular, weekly recreational physical activity were classified as inactive. Multivariable logistic regression was utilized to estimate the ORs and 95% confidence intervals (CI) for the association between inactivity and EOC risk overall and by subgroups based upon histotype, menopausal status, race, and body mass index.

Results: The current analysis included data from 8,309 EOC patients and 12,612 controls. We observed a significant positive association between inactivity and EOC risk (OR = 1.34; 95% CI, 1.14–1.57), and similar associations were observed for each histotype.

Conclusions: In this large pooled analysis examining the association between recreational physical inactivity and EOC risk, we observed consistent evidence of an association between chronic inactivity and all EOC histotypes.

Impact: These data add to the growing body of evidence suggesting that inactivity is an independent risk factor for cancer. If the apparent association between inactivity and EOC risk is substantiated, additional work via targeted interventions should be pursued to characterize the dose of activity required to mitigate the risk of this highly fatal disease. Cancer Epidemiol Biomarkers Prev; 25(7); 1114–24. ©2016 AACR.

It is well established that recreational physical activity is associated with decreased risks of developing breast, colon, and endometrial cancers (1, 2), but the association between physical activity and epithelial ovarian cancer (EOC) remains less clear (3, 4). Despite the publication of dozens of individual epidemiologic studies, two organizational systematic reviews have concluded that insufficient and inconsistent evidence was available in the current literature to support an association between recreational physical activity and ovarian cancer risk (4, 5).

Inconsistent epidemiologic reports of the association between physical activity and EOC risk may be the result of limitations in the physical activity and ovarian cancer literature or may be due to a complex dose–response relationship that has not yet been fully elucidated. For example, individual studies of ovarian cancer often have relatively small numbers of case subjects, especially for the less common histotypes, which could limit statistical power to detect significant associations. In fact, the largest individual study to date included 1,580 cases, but only 44 patients were diagnosed with mucinous tumors (6). A lack of consistency in the literature could also reflect that only modest decreases in risk are associated with higher levels of activity (7) and that the greatest risk is associated with inactivity (8), a construct that has been scarcely investigated as an independent exposure relative to EOC risk. To our knowledge, all but two EOC studies (9, 10) have examined recreational physical activity using arbitrary cut-off points of incrementally higher levels of activity, with the low- or no activity group identified as the reference group. Although parameterizing incrementally higher levels of activity exposure is important for detecting dose–response relationships, this approach has been associated with complex and meaningful exposure misclassification (11–13) and has precluded the establishment of a clear public health recommendation specific to ovarian cancer (4, 5). Importantly, this common methodology overlooks recreational physical inactivity as an independent public health exposure of interest (5).

To this end, since the publication of the 2008 Physical Activity Guidelines for Americans, adults have been encouraged to avoid physical inactivity, which is characterized by a lack of regular, weekly, moderate- or vigorous-intensity recreational activity (5). According to the most current data, 25% of Americans (14) and between 10.3% and 51.5% of adult women worldwide (8) are physically inactive. Given the persistence of inactivity at the population level and the hypothesis that the greatest protective benefits can be achieved by increasing activity levels at the low end of the activity continuum (8), inactive individuals could be a particularly important group to study in relationship to disease risk because of the ability for most individuals to increase the amount of activity they perform each week (8). In fact, current estimates suggest that congenital factors contributing to inactivity affect less than one percent of the population globally, implying that most individuals are capable of increasing activity levels (8).

Not only is studying physical inactivity important from a public health perspective (5, 15), it is likely assessed with less exposure misclassification (8, 16) and may also reflect physiologic pathways that exert an effect on carcinogenesis separately from pathways associated with physical activity and skeletal muscle contraction (15, 17). Thus, we conducted a pooled analysis of nine population-based case–control studies from Ovarian Cancer Association Consortium (OCAC) to investigate a novel, well-defined research question. Specifically, we sought to determine whether self-reported, chronic, recreational physical inactivity is associated with an increased risk of EOC. We evaluated the association between physical inactivity exposure and EOC risk overall, and according to subgroups based upon EOC histotype, menopause status, race, and body mass index (BMI).

OCAC study population

We obtained individual-level data from nine population-based OCAC case–control studies that had available self-report data on recreational physical inactivity throughout adulthood. Seven of the nine OCAC studies were based in the United States (18–24) and one each was based in Australia (25) and Denmark (26). Additional characteristics of the nine case–control studies included in the current analyses are summarized in Table 1. 

All individual studies obtained Institutional Review Board or research ethics committee approvals, and participants for each OCAC study provided written informed consent for all study activities. Approvals for the current analyses were obtained from the OCAC Data Access Coordinating Committee and from individual study site coordinators if additional approvals were required. Data were obtained from 8,309 patients aged 18 years or older with histologically confirmed primary borderline or invasive EOC, fallopian tube cancer, or peritoneal cancer. Patients were excluded from the current analyses if they had been diagnosed with nonepithelial ovarian cancers (sarcomas, germ cell tumors, sex-cord stromal tumors, etc.); if tumor histology was mixed or undifferentiated; or if tumor behavior or histology was missing or unknown. Controls included 12,612 women aged 18 years and older with at least partially intact ovaries and no prior histories of ovarian cancer.

Analysis variables

Epidemiologic data.

The primary OCAC epidemiologic dataset includes information that was collected through self-administered or interviewer-administered questionnaires. Available demographic, lifestyle, and clinical variables include age, race, ethnicity, education, tumor behavior, tumor histology, oral contraceptive use, family history of breast and ovarian cancer, age at menarche, number of full-term pregnancies, breastfeeding, age at menopause, hormone therapy use (estrogen-alone or combination therapy), hysterectomy, and several additional epidemiologic variables from previously published OCAC pooled analyses [i.e., tubal ligation, smoking history, alcohol use, genital powder use, NSAID use, recent BMI (1–5 years prior to diagnosis), and personal history of endometriosis; refs. 27–33].

Recreational physical inactivity.

Recreational physical activity data were directly acquired from each of the nine OCAC studies included in the current analysis. All nine questionnaires assessed recreational activity spanning adulthood up through the reference age, defined as the age of diagnosis among cases and the age of study entry among controls. Specifically, eight of nine questionnaires encompassed the time period spanning all decades of adulthood (i.e., age 20 through the reference date), while one study (DVE) spanned the time period from age 25 through the reference date.

The specific parameters of physical activity were inconsistently measured across studies, precluding the ability to harmonize and parameterize physical activity data in terms of frequency, intensity, or duration per session. However, in the current analysis, the exposure of interest was recreational physical inactivity, and all nine questionnaires allowed for the identification of women who self-reported engaging in no regular, weekly moderate-to-vigorous intensity recreational activity. Questionnaires from most studies (DVE, HAW, HOP, NEC, NJO, USC) utilized a global, dichotomous item assessing ever-participation in regular, weekly, recreational physical activities. For these studies, women answering “no” to the global question were classified as “inactive.” Three studies (AUS, CON, MAL) assessed recreational physical (in)activity based upon prespecified time periods spanning adulthood (i.e., by decades or a combination of decades ranging from age 20–29 through the reference age). Likewise, women reporting no regular, weekly moderate-or vigorous intensity recreational activity in all time periods prior to the reference date were classified as inactive. Furthermore, given that the most relevant (in)activity exposure period may be many years before the actual diagnosis of cancer (7), we conducted analyses designed to examine an exposure window encompassing at least two decades of adulthood prior to study entry. Thus, participants with reference dates in their 20s were excluded in sensitivity analyses, yielding a chronic inactivity exposure spanning a minimum of two decades.

Statistical analysis

Identification of confounding variables.

Based upon the definition of potential confounding (34) and their establishment as factors associated with risk of ovarian cancer, the following variables in the OCAC core dataset were prespecified as important for adjustment when estimating EOC risks: age at reference date, race (White, Black, Asian, other), use of oral contraceptives (ever, never), parity (nulliparous, 1, 2, 3, or ≥ 4 full-term births), family history of breast or ovarian cancer in a first-degree relative (yes, no, do not know), and a personal history of endometriosis (yes, no). In pooled analyses utilizing a combined dataset, we further adjusted all models for study site. We also utilized the 10% change-in-estimate approach (35) to inform the selection of additional adjustment variables. Based upon this approach, we determined that ethnicity (Hispanic, non-Hispanic), use of hormone therapy (yes or no for estrogen-only or combination estrogen/progesterone), smoking (never, current, former), alcohol use (never, ever, former), education (less than high school, high school, some college, college graduate, graduate school), talc or genital powder use (no use, genital use, nongenital use), NSAIDs use (≥ once/week, <once/week), tubal ligation prior to diagnosis (yes, no), breastfeeding (yes, no, not applicable), hysterectomy prior to diagnosis (yes, no), menopause status (pre or peri, post, do not know), and recent BMI (underweight, normal weight, overweight, or obese based upon BMI 1–5 years prior to diagnosis) were not relevant confounders. However, because obesity has an established association with physical activity and it is also associated with increased risk of some EOC histotypes, we evaluated models with and without adjustment for BMI, as obesity may be in the causal pathway of interest.

Recreational physical inactivity and EOC risk.

To account for between-study heterogeneity, we utilized a meta-analytic approach to examine the association between inactivity and EOC risk overall and according to EOC endpoints defined by tumor behavior and histology. For each of the nine studies, logistic regression analyses were conducted to estimate study-specific ORs and 95% confidence intervals (CI) for the association between chronic physical inactivity and EOC risk. Study-specific ORs and their variances were then combined via meta-analyses to estimate summary ORs and 95% CIs for all EOC endpoints. Meta-analytic analyses were conducted under random-effects or fixed-effects assumptions, depending on measures of between-study heterogeneity, which was assessed and quantified utilizing the Cochran Q-statistic and the I2 statistic (36). When evidence of significant heterogeneity was observed between studies (Q-statistic P < 0.05 or I2 value >50%), we reported a random-effects OR based upon the DerSimonian and Laird method (37), and we conducted further analyses to identify and account for source(s) of heterogeneity. However, when no significant heterogeneity was observed between studies, we reported a fixed-effects OR.

To enable well-powered subgroup analyses, we pooled individual-level data from the nine studies into a combined dataset to examine the association between inactivity and EOC risk by menopause status, race, and BMI classification. We examined associations between inactivity and EOC risk by subgroups of standard categories of BMI (i.e., underweight, normal weight, overweight, and obese) and by a dichotomous BMI classification (i.e., underweight and normal weight vs. overweight and obese). In pooled analyses, all multivariable models were adjusted by study site, and we accounted for the possibility of between-study heterogeneity by testing the significance of a cross product term for site*inactivity in all analyses.

Finally, we conducted sensitivity analyses designed to address any potential heterogeneity in the observed associations between inactivity and EOC risk that could have resulted from differences in the physical activity questionnaires among OCAC studies. Furthermore, to account for potential cultural or geographical differences in activity patterns, we excluded two studies that were not conducted in the United States (AUS and MAL).

The characteristics of the nine OCAC case–control studies included in the analyses are summarized in Table 1. The self-reported prevalence of recreational physical inactivity among the study population is presented in Table 2. Collectively, 23.7% of cases and 20.9% of controls reported a history of inactivity (P < 0.001). Inactivity rates varied by study site (10% in HAW and 48.5% in MAL), yielding a combined inactivity prevalence of 22.0% among the total study population. These estimates are very similar to national (25%) and global (10.3%–51.5%) inactivity estimates (8, 14, 38), thus enhancing the confidence in our characterization of the inactivity exposure.

In meta-analyses, we observed a significant positive association between self-reported, chronic physical inactivity and all EOC endpoints (Table 3). Specifically, among inactive women, we observed a 34% increased risk of EOC overall (OR = 1.34; 95% CI, 1.14–1.57; Fig. 1A); a 35% increased risk of invasive tumors (OR = 1.35; 95% CI, 1.14–1.60; Fig. 1B); and a 27% increased risk of borderline tumors (OR = 1.27; 95% CI, 1.10–1.46; Fig. 1C). Among the five invasive histotypes, we observed increased risks ranging between 29% and 54% among inactive women (Fig. 2A–E). Physical inactivity was also associated with a 27% and 31% increased risk of borderline serous and borderline mucinous tumors, respectively (Fig. 3A and B). Importantly, adding BMI to multivariable models did not change these estimates appreciably (Table 3).

Significant between-study heterogeneity was noted for EOC overall, invasive tumors, high-grade serous tumors, and endometrioid tumors (Table 3). The observed heterogeneity was explained by one study (NJO), and we accounted for between-study heterogeneity by presenting random-effects summary ORs. Upon excluding NJO in sensitivity analyses, associations between inactivity and EOC risk for EOC overall, invasive tumors, high-grade serous tumors, and endometrioid tumors were of comparable magnitude and remained statistically significant. For example, the OR was 1.25 and 95% CI was 1.16–1.35 for EOC overall (Pheterogeneity = 0.467).

After excluding NJO, we further compared the associations between inactivity and EOC risk based upon the questionnaire format utilized among the OCAC studies included in the analysis. We observed no significant heterogeneity in weighted point estimates among those studies utilizing one global item to assess (in)activity throughout adulthood (OR = 1.27, 95% CI, 1.16–1.39) versus the studies utilizing multiple prespecified time periods throughout adulthood (OR = 1.20, 95% CI, 1.04–1.40), yielding an overall fixed-effects OR = 1.25; 95% CI, 1.16–1.35; Pheterogeneity = 0.523. Importantly, even when adding NJO back into the analyses, there was no significant heterogeneity observed between the two formats incorporated herein (OR = 1.27, 95% CI, 1.16–1.39), Pheterogeneity = 0.154. Finally, in additional sensitivity analyses excluding two studies conducted outside the United States (AUS, MAL), the observed associations between inactivity and EOC risk were strengthened (OR = 1.43; 95% CI, 1.17–1.76), and the association remained significant after excluding NJO (OR = 1.27; 95% CI, 1.16–1.39).

In subgroup analyses, we observed no convincing evidence of effect modification by menopause status (Pinteraction = 0.483; Supplementary Table S1); race (Pinteraction = 0.337; Supplementary Table S2); or by standard BMI categories (Pinteraction = 0.082; Supplementary Table S3). Finally, when the association between inactivity and EOC risk was examined utilizing a dichotomous BMI variable (BMI < 25 or BMI ≥ 25), we observed a significant, positive association between inactivity and EOC risk among underweight/normal weight women and overweight/obese women: (OR = 1.33; 95% CI, 1.19–1.49) and (OR = 1.21; 95% CI, 1.09–1.34), respectively (Pinteraction = 0.041; Supplementary Table S4).

In the current analysis, we observed consistent evidence of a statistically significant positive association between self-reported, chronic recreational physical inactivity and all histotypes of EOC. Although published data describing the association between recreational physical inactivity and EOC risk are scant, two recent prospective studies reported the association between EOC risk and the most inactive group of women, in comparison to a reference group of women engaging in moderate amounts of activity (9, 10). Leitzmann and colleagues (10) reported an increased risk of early-stage and fatal EOC among the most inactive women (RR = 1.29; 95% CI, 0.55–3.04 and RR = 1.20; 95% CI, 0.71–2.03, respectively) and Huang and colleagues (9) reported a 29% increased risk of EOC among women who were the least active during premenopausal years (HR = 1.29; 95% CI, 0.95–7.75). Although these estimates did not reach statistical significance, they are of similar magnitude to each other and to the point estimates reported for overall EOC in our primary and sensitivity analyses.

Our analyses did not account for explicitly sedentary behaviors, such as sitting, television watching, or computer use, yet there is mounting epidemiologic evidence demonstrating a positive association between sedentary behaviors and EOC risk (39–43). Although sedentary behavior is a separate behavioral construct independent of recreational physical (in)activity, these collective findings add to the growing body of evidence that behavior patterns indicative of a sedentary lifestyle, either by way of prolonged sitting or a lack of recreational physical activity, appear to be associated with increased EOC risk. Furthermore, it is plausible that the underlying biologic mechanisms responsible for an apparent association of sedentary behavior and recreational physical inactivity with EOC risk are likely similar.

Although research has established that the incidence of certain types of tumors is reduced by regular physical activity, the biologic mechanisms relating physical (in)activity and cancer risk are not entirely understood (44). The most commonly cited mediators of these associations include changes in body fat, changes in circulating reproductive hormone levels, alterations in inflammatory cytokine and growth factor milieus, and alterations in immune function (44). Inasmuch as the current leading hypotheses about the etiology of EOC each share inflammation as an underlying mechanism (45), it is possible that inactivity increases EOC risk by way of increased systemic inflammation, decreased immune function, and increased circulating levels of sex hormones. For example, both acute and chronic physical activity produce an anti-inflammatory effect via reduced levels of inflammatory markers such as C-reactive protein and TNF (46). In addition, both mechanistic and epidemiologic evidence suggests a role for dysregulated adiponectin (an anti-inflammatory adipokine) and leptin (a proinflammatory adipokine) in epithelial ovarian carcinogenesis (47–52). There is also evidence suggesting a dysregulated adipokine milieu promotes an immunosuppressive environment via myeloid-derived suppressor cell (MDSC) induction and FoxP3+ T-regulatory cell recruitment (53), an area that has been under intense investigation in relationship to ovarian cancer etiology and prognosis (54–57). Finally, there are endocrine-related hypotheses that support the plausibility of an inverse association between activity and EOC risk by way of decreased levels of circulating estrogens and androgens, and in fact, most studies seem to suggest that exercise decreases the availability of biologically active estrogens and androgens (58).

Emerging evidence also supports the hypotheses that physical inactivity and sedentary behaviors are exposures with distinct metabolic consequences, which could be independent of physical activity and obesity-related mechanisms (15, 17, 59, 60). Our data also imply that the observed association between physical inactivity and EOC risk is mostly independent of BMI. Although adding BMI to our multivariable models yielded a slight attenuation in the observed ORs, an independent, consistent, and significant effect of inactivity remained.

Furthermore, although there appears to be a borderline statistically significant interaction between BMI classification (i.e., underweight, normal weight, overweight, obese) and inactivity in relation to EOC risk, the apparent stronger association among underweight women and weaker association among obese women were accompanied by CIs that included the null. This suggests there may be imprecision in estimating EOC risk at the extreme distribution of BMI and inactivity when they are considered jointly. Finally, when the association between inactivity and EOC risk was examined in two subgroups of BMI (i.e., <25 or ≥25), we observed a statistically significant BMI*inactivity interaction for EOC overall (P = 0.041). Although the interaction P value for EOC overall suggests this association may not be due to chance alone, we believe the associations between inactivity and EOC risk are qualitatively comparable across BMI stratum.

A key strength of our study is that our analyses were conducted with individual-level data from well-designed population-based epidemiologic investigations, yielding the first substantial analysis of the association between recreational physical inactivity and EOC risk. Furthermore, our ability to adjust for well-established risk and protective factors associated with EOC risk decreased the likelihood that the observed associations were the result of confounding. In addition, the observed associations between inactivity and EOC risk remained significant and of similar magnitude in all sensitivity analyses designed to reduce potential sources of bias. Finally, our use of chronic inactivity as the exposure of interest reduced the likelihood of reverse causation bias as a potential explanation for the observed associations reported herein.

The potential measurement error associated with self-report physical activity data categorized as a dichotomous variable is an important limitation to the current analysis. Although the dichotomous nature of the exposure variable assumes a homogenous group of activity and does not allow for an examination of the dose–response association with physical activity exposure, there are important advantages to our approach. First, although misreporting of inactivity does occur, we assume the misclassification across incremental categories of activity would be greater (8, 11, 12, 16) and more influential with respect to biasing observed associations of interest. In fact, previous research has demonstrated that the greatest concordance between direct and self-report measures of activity is found at the lowest ends of the activity continuum, with more measurement error surrounding measures of mid-to-upper levels of exposure (8, 11, 12). Utilizing a dichotomous variable avoids a potentially misleading level of accuracy in categorizing incremental levels of self-reported physical activity (8). Second, there is a body of evidence demonstrating that the use of one global question is a validated method for identifying inactive individuals (13, 61–66). Third, although it is impossible to know whether (in)activity misclassification was differential by case–control status, one tactic is to compare our findings with those from cohort studies, where self-reported (in)activity would not be subjected to recall bias (7). Among the two prospective studies providing risk estimates for inactivity (9, 10), associations were similar to those reported herein, arguing against a bias in case–control studies due to differential misclassification. Importantly, nondifferential misclassification with a dichotomous exposure variable would likely result in an underestimate of the true association between inactivity and EOC risk (67).

We also recognize that our findings may be limited by the potential for a higher prevalence of healthier women to have volunteered as controls. If so, this would have inflated our observed estimates of association between inactivity and EOC risk. Likewise, associations between EOC risk and other lifestyle factors, such as smoking, alcohol consumption, and obesity, would also show inflated risk estimates. Yet, previously published OCAC-pooled analyses utilizing data from the same studies have yielded no evidence of an association between alcohol consumption and EOC risk (28), and evidence of associations between EOC risk and smoking and obesity has been restricted to specific histotypes (27, 29). It is also possible that additional unmeasured factors that may parallel physical activity (or inactivity) in lifestyle patterns could contribute to an observed association between physical inactivity and EOC risk.

Although the goal of the current study was to examine the association between recreational physical inactivity and EOC risk, it is worth noting that previous epidemiologic studies of physical activity and EOC risk have yielded inconsistent findings based upon the type of observational study. Although the first published meta-analysis of the association between physical activity and EOC risk reported similar risk estimates for case–control studies (OR = 0.79; 95% CI, 0.70–0.85) and cohort studies (OR = 0.81; 95% CI, 0.72–0.92; ref. 6), a more recent meta-analysis reported a significant inverse association between activity and EOC risk among case–control studies (OR = 0.86; 95% CI, 0.80–0.93) but reported no association among cohort studies (OR = 1.03; 95% CI, 0.87–1.20; ref. 68).

It is also important to highlight that there are competing hypotheses regarding the shape of the dose–response physical activity curve in relationship to chronic disease risk (i.e., linear vs. nonlinear; ref. 8). In fact, prospective studies have yielded data suggesting a significant positive association between vigorous physical activity and EOC risk (9, 69, 70). Although there is biologic plausibility for a positive association between excessive vigorous exercise and increased EOC risk by way of impaired immune function (44) and exercise-induced increases in gonadotropin and androgen secretion (70), researchers have cautioned that observations of a direct association between activity and EOC risk could be due to chance (69), small cell sizes (69), and detection bias (10, 69). Importantly, nondifferential misclassification of self-reported physical activity parameterized in more than two categories can result in biased estimates away from the null (67).

In conclusion, in the first substantial analysis designed to examine chronic recreational physical inactivity as an independent exposure of interest, we observed evidence of a significant positive association between recreational inactivity and EOC risk that was consistently observed among all EOC histotypes. These data add to the growing body of literature demonstrating that physical inactivity is associated with a plethora of unfavorable health outcomes, including an increased risk for early death, heart disease, stroke, type II diabetes, and certain cancers including breast, colon, and endometrial tumors (5, 8, 14). Additional prospective epidemiologic studies are warranted to further elucidate the dose–response association between recreational physical (in)activity and EOC risk. If the apparent association between inactivity and EOC risk is substantiated, then additional work via targeted intervention studies should be pursued to characterize the dose of recreational physical activity required to mitigate the risk of this highly fatal disease.

D. Cramer has provided expert testimony for Beasley Allen Crow. M.T. Goodman is a consultant/advisory board member for Johnson and Johnson. P. Webb reports receiving a commercial research grant from BUPA. No potential conflicts of interest were disclosed by the other authors.

Conception and design: R. Cannioto, M.J. LaMonte, H.A. Risch, L.E. Sucheston-Campbell, M.C. Pike, F. Modugno, K.B. Moysich

Development of methodology: R. Cannioto, M.J. LaMonte, H.A. Risch, K.H. Eng, L.E. Kelemen, K.B. Moysich

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Cannioto, H.A. Risch, J. Chang-Claude, E. Gower, E.V. Bandera, A. Berchuck, D. Cramer, J.A. Doherty, R.P. Edwards, B.L. Fridley, E.L. Goode, M.T. Goodman, E. Hogdall, S. Hosono, A. Jensen, S.K. Kjaer, K. Matsuo, R.B. Ness, C.M. Olsen, S.H. Olson, M.C. Pike, M.A. Rossing, P.M. Webb, N. Wentzensen, K.G. Wicklund, A.H. Wu, F. Modugno, J.M. Schildkraut, K.L. Terry, K.B. Moysich

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Cannioto, M.J. LaMonte, C.-C. Hong, L.E. Sucheston-Campbell, K.H. Eng, B. Schmalfeldt, G.R. Zirpoli, A. Berchuck, B.L. Fridley, S. Jordan, M.A. Rossing, R.A. Vierkant, L.E. Kelemen, K.B. Moysich

Writing, review, and/or revision of the manuscript: R. Cannioto, M.J. LaMonte, H.A. Risch, C.-C. Hong, L.E. Sucheston-Campbell, K.H. Eng, J.B. Szender, J. Chang-Claude, B. Schmalfeldt, R. Klapdor, A.N. Minlikeeva, G.R. Zirpoli, E.V. Bandera, A. Berchuck, D. Cramer, J.A. Doherty, E.L. Goode, M.T. Goodman, E. Hogdall, A. Jensen, S. Jordan, S.K. Kjaer, R.B. Ness, C.M. Olsen, C.L. Pearce, M.C. Pike, M.A. Rossing, E.A. Szamreta, P.J. Thompson, P.M. Webb, N. Wentzensen, S.J. Winham, A.H. Wu, F. Modugno, K.L. Terry, L.E. Kelemen, K.B. Moysich

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): R. Cannioto, A. Berchuck, R.P. Edwards, B.L. Fridley, E. Hogdall, E.A. Szamreta, P.J. Thompson, C.-C. Tseng, F. Modugno

Study supervision: A. Berchuck, S.K. Kjaer, M.A. Rossing, P.J. Thompson, F. Modugno, K.B. Moysich

The following studies were funded by the following grants:

AUS (G. Chenevix-Trench/P.M. Webb): U.S. Army Medical Research and Materiel Command (DA17-01-1-0729), National Health & Medical Research Council of Australia, Cancer Councils of New South Wales, Victoria, Queensland, South Australia, and Tasmania, Cancer Foundation of Western Australia; National Health and Medical Research Council of Australia (199600 and 400281). The grant numbers for AOCS Cancer Council funding are as follows- Multi-state application numbers 191, 211, and 182;

CON (H.A. Risch): NIH (R01-CA074850; R01-CA080742);

DOV (M.A. Rossing): NIH R01-CA112523 and R01-CA87538;

HAW (M.T. Goodman): NIH (R01-CA58598, N01-CN-55424 and N01-PC-67001);

HOP (F. Modugno, R.B. Ness, K.B. Moysich): DOD: DA17-02-1-0669 and NCI: K07-CA080668, R01-CA95023, P50-CA159981;

MAL (S.K. Kjaer): Funding for this study was provided by research grant R01-CA61107 from the NCI; research grant 94 222 52 from the Danish Cancer Society, Copenhagen, Denmark; and the Mermaid I project;

MAY/MAC (E.L. Goode): R01-CA122443

NEC (D.W. Cramer, K.L. Terry): NIH R01-CA54419 and P50-CA105009 and Department of Defense W81XWH-10-1-02802;

NJO (E.V. Bandera): NCI (NIH-K07 CA095666, R01-CA83918, NIH-K22-CA138563, and P30-CA072720) and the Cancer Institute of New Jersey; (S.H. Olson) NCI CCSG award (P30-CA008748);

USC (C.L. Pearce): P01CA17054, P30CA14089, R01CA61132, N01PC67010, R03CA113148, R03CA115195, N01CN025403, and California Cancer Research Program (00-01389V-20170, 2II0200).

1.
Friedenreich
CM
,
Neilson
HK
,
Lynch
BM
. 
State of the epidemiological evidence on physical activity and cancer prevention
.
Eur J Cancer
2010
;
46
:
2593
604
.
2.
Kruk
J
,
Czerniak
U
. 
Physical activity and its relation to cancer risk: updating the evidence
.
Asian Pac J Cancer Prev
2013
;
14
:
3993
4003
.
3.
Cannioto
RA
,
Moysich
KB
. 
Epithelial ovarian cancer and recreational physical activity: a review of the epidemiological literature and implications for exercise prescription
.
Gynecol Oncol
2015
;
137
:
559
73
.
4.
WCRF/AICR. Food, Nutrition, Physical Activity, and the Prevention of Ovarianc Cancer 2014
.
Continuous update project report ovarian cancer 2014 summary
.
Available from
: http://www.aicr.org/continuous-update-project/reports/ovarian-cancer-2014-report.pdf.
5.
United States Department of Health & Human Services.
2008 physical activity guidelines for Americans
.
Washington, D.C.
:
Office of Disease Prevention and Health Promotion
; 
2008
.
6.
Olsen
CM
,
Bain
CJ
,
Jordan
SJ
,
Nagle
CM
,
Green
AC
,
Whiteman
DC
, et al
Recreational physical activity and epithelial ovarian cancer: a case-control study, systematic review, and meta-analysis
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
2321
30
.
7.
Moorman
PG
,
Jones
LW
,
Akushevich
L
,
Schildkraut
JM
. 
Recreational physical activity and ovarian cancer risk and survival
.
Ann Epidemiol
2011
;
21
:
178
87
.
8.
Bull
FC
,
Armstrong
TP
,
Dixon
T
,
Ham
S
,
Neiman
A
,
Pratt
M
. 
Physical inactivity
.
Comparitive quantification of health risks: global and regional burden of disease attributable to selected major risk factors.
Geneva, Switzerland
:
WHO
; 
2004
. p.
729
881
.
9.
Huang
T
,
Eliassen
AH
,
Hankinson
SE
,
Okereke
OI
,
Kubzansky
LD
,
Wang
M
, et al
A prospective study of leisure-time physical activity and risk of incident epithelial ovarian cancer: impact by menopausal status
.
Int J Cancer
2016
;
138
:
843
52
.
10.
Leitzmann
MF
,
Koebnick
C
,
Moore
SC
,
Danforth
KN
,
Brinton
LA
,
Hollenbeck
AR
, et al
Prospective study of physical activity and the risk of ovarian cancer
.
Cancer Causes Control
2009
;
20
:
765
73
.
11.
Troiano
R
. 
Limitations of self-report in physical activity and obesity research
. In:
Bouchard
C
,
Katzmarzyk
PT
, editors.
Physical activity and obesity
, 2nd Edition.
Champaign, IL
:
Human Kinetics
; 
2010
:
34
7
.
12.
Troiano
RD
,
Dodd
KW
. 
Differences between objective and self-report measures of physical activity. What do they mean?
J Kor Soc Meas Eval
2008
;
10
:
31
42
.
13.
Tudor-Locke
CE
,
Myers
AM
. 
Challenges and opportunities for measuring physical activity in sedentary adults
.
Sports Med
2001
;
31
:
91
100
.
14.
Moore
LV
,
Harris
CD
,
Carlson
SA
,
Kruger
J
,
Fulton
JE
. 
Trends in no leisure-time physical activity–United States, 1988–2010
.
Res Q Exerc Sport
2012
;
83
:
587
91
.
15.
Sanchis-Gomar
F
,
Lucia
A
,
Yvert
T
,
Ruiz-Casado
A
,
Pareja-Galeano
H
,
Santos-Lozano
A
, et al
Physical inactivity and low fitness deserve more attention to alter cancer risk and prognosis
.
Cancer Prev Res
2015
;
8
:
105
10
.
16.
Celis-Morales
CA
,
Perez-Bravo
F
,
Ibanez
L
,
Salas
C
,
Bailey
ME
,
Gill
JM
. 
Objective vs. self-reported physical activity and sedentary time: effects of measurement method on relationships with risk biomarkers
.
PLoS One
2012
;
7
:
e36345
.
17.
Fiuza-Luces
C
,
Garatachea
N
,
Berger
NA
,
Lucia
A
. 
Exercise is the real polypill
.
Physiology
2013
;
28
:
330
58
.
18.
Bandera
EV
,
King
M
,
Chandran
U
,
Paddock
LE
,
Rodriguez-Rodriguez
L
,
Olson
SH
. 
Phytoestrogen consumption from foods and supplements and epithelial ovarian cancer risk: a population-based case control study
.
BMC Womens Health
2011
;
11
:
40
.
19.
Goodman
MT
,
Lurie
G
,
Thompson
PJ
,
McDuffie
KE
,
Carney
ME
. 
Association of two common single-nucleotide polymorphisms in the CYP19A1 locus and ovarian cancer risk
.
Endocr Relat Cancer
2008
;
15
:
1055
60
.
20.
Ness
RB
,
Dodge
RC
,
Edwards
RP
,
Baker
JA
,
Moysich
KB
. 
Contraception methods, beyond oral contraceptives and tubal ligation, and risk of ovarian cancer
.
Ann Epidemiol
2011
;
21
:
188
96
.
21.
Pike
MC
,
Pearce
CL
,
Peters
R
,
Cozen
W
,
Wan
P
,
Wu
AH
. 
Hormonal factors and the risk of invasive ovarian cancer: a population-based case-control study
.
Fertil Steril
2004
;
82
:
186
95
.
22.
Risch
HA
,
Bale
AE
,
Beck
PA
,
Zheng
W
. 
PGR +331 A/G and increased risk of epithelial ovarian cancer
.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
1738
41
.
23.
Rossing
MA
,
Cushing-Haugen
KL
,
Wicklund
KG
,
Doherty
JA
,
Weiss
NS
. 
Menopausal hormone therapy and risk of epithelial ovarian cancer
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
2548
56
.
24.
Terry
KL
,
De Vivo
I
,
Titus-Ernstoff
L
,
Shih
MC
,
Cramer
DW
. 
Androgen receptor cytosine, adenine, guanine repeats, and haplotypes in relation to ovarian cancer risk
.
Cancer Res
2005
;
65
:
5974
81
.
25.
Merritt
MA
,
Green
AC
,
Nagle
CM
,
Webb
PM
. 
Talcum powder, chronic pelvic inflammation and NSAIDs in relation to risk of epithelial ovarian cancer
.
Int J Cancer
2008
;
122
:
170
6
.
26.
Glud
E
,
Kjaer
SK
,
Thomsen
BL
,
Hogdall
C
,
Christensen
L
,
Hogdall
E
, et al
Hormone therapy and the impact of estrogen intake on the risk of ovarian cancer
.
Arch Intern Med
2004
;
164
:
2253
9
.
27.
Faber
MT
,
Kjaer
SK
,
Dehlendorff
C
,
Chang-Claude
J
,
Andersen
KK
,
Hogdall
E
, et al
Cigarette smoking and risk of ovarian cancer: a pooled analysis of 21 case-control studies
.
Cancer Causes Control
2013
;
24
:
989
1004
.
28.
Kelemen
LE
,
Bandera
EV
,
Terry
KL
,
Rossing
MA
,
Brinton
LA
,
Doherty
JA
, et al
Recent alcohol consumption and risk of incident ovarian carcinoma: a pooled analysis of 5,342 cases and 10,358 controls from the ovarian cancer association consortium
.
BMC Cancer
2013
;
13
:
28
.
29.
Olsen
CM
,
Nagle
CM
,
Whiteman
DC
,
Ness
R
,
Pearce
CL
,
Pike
MC
, et al
Obesity and risk of ovarian cancer subtypes: evidence from the Ovarian Cancer Association Consortium
.
Endocr Relat Cancer
2013
;
20
:
251
62
.
30.
Pearce
CL
,
Templeman
C
,
Rossing
MA
,
Lee
A
,
Near
AM
,
Webb
PM
, et al
Association between endometriosis and risk of histological subtypes of ovarian cancer: a pooled analysis of case-control studies
.
Lancet Oncol
2012
;
13
:
385
94
.
31.
Sieh
W
,
Salvador
S
,
McGuire
V
,
Weber
RP
,
Terry
KL
,
Rossing
MA
, et al
Tubal ligation and risk of ovarian cancer subtypes: a pooled analysis of case-control studies
.
Int J Epidemiol
2013
;
42
:
579
89
.
32.
Terry
KL
,
Karageorgi
S
,
Shvetsov
YB
,
Merritt
MA
,
Lurie
G
,
Thompson
PJ
, et al
Genital powder use and risk of ovarian cancer: a pooled analysis of 8,525 cases and 9,859 controls
.
Cancer Prev Res
2013
;
6
:
811
21
.
33.
Trabert
B
,
Ness
RB
,
Lo-Ciganic
WH
,
Murphy
MA
,
Goode
EL
,
Poole
EM
, et al
Aspirin, nonaspirin nonsteroidal anti-inflammatory drug, and acetaminophen use and risk of invasive epithelial ovarian cancer: a pooled analysis in the ovarian cancer association consortium
.
J Natl Cancer Inst
2014
;
106
:
djt431
.
34.
Szklo
M
,
Nieto
J
. 
Indentifying non causal associations: confounding
. In:
Szklo
M
,
Nieto
J
, editors.
Epidemiology: beyond the basics
.
Sudbury, MA
:
Jones & Bartlett Learning
; 
2007
:
151
82
.
35.
Maldonado
G
,
Greenland
S
. 
Simulation study of confounder-selection strategies
.
Am J Epidemiol
1993
;
138
:
923
36
.
36.
Higgins
JP
,
Thompson
SG
,
Deeks
JJ
,
Altman
DG
. 
Measuring inconsistency in meta-analyses
.
BMJ
2003
;
327
:
557
60
.
37.
DerSimonian
R
,
Laird
N
. 
Meta-analysis in clinical trials
.
Control Clin Trials
1986
;
7
:
177
88
.
38.
Hallal
PC
,
Andersen
LB
,
Bull
FC
,
Guthold
R
,
Haskell
W
,
Ekelund
U
. 
Global physical activity levels: surveillance progress, pitfalls, and prospects
.
Lancet
2012
;
380
:
247
57
.
39.
Patel
AV
,
Hildebrand
JS
,
Campbell
PT
,
Teras
LR
,
Craft
LL
,
McCullough
ML
, et al
Leisure-time spent sitting and site-specific cancer incidence in a large U.S. Cohort
.
Cancer Epidemiol Biomarkers Prev
2015
;
24
:
1350
9
.
40.
Patel
AV
,
Rodriguez
C
,
Pavluck
AL
,
Thun
MJ
,
Calle
EE
. 
Recreational physical activity and sedentary behavior in relation to ovarian cancer risk in a large cohort of US women
.
Am J Epidemiol
2006
;
163
:
709
16
.
41.
Xiao
Q
,
Yang
HP
,
Wentzensen
N
,
Hollenbeck
A
,
Matthews
CE
. 
Physical activity in different periods of life, sedentary behavior, and the risk of ovarian cancer in the NIH-AARP diet and health study
.
Cancer Epidemiol Biomarkers Prev
2013
;
22
:
2000
8
.
42.
Zhang
M
,
Xie
X
,
Lee
AH
,
Binns
CW
. 
Sedentary behaviours and epithelial ovarian cancer risk
.
Cancer Causes Control
2004
;
15
:
83
9
.
43.
Hildebrand
JS
,
Gapstur
SM
,
Gaudet
MM
,
Campbell
PT
,
Patel
AV
. 
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
:
1691
7
.
44.
Walsh
NP
,
Gleeson
M
,
Shephard
RJ
,
Gleeson
M
,
Woods
JA
,
Bishop
NC
, et al
Position statement. Part one: immune function and exercise
.
Exerc Immunol Rev
2011
;
17
:
6
63
.
45.
Ness
RB
,
Cottreau
C
. 
Possible role of ovarian epithelial inflammation in ovarian cancer
.
J Natl Cancer Inst
1999
;
91
:
1459
67
.
46.
Kasapis
C
,
Thompson
PD
. 
The effects of physical activity on serum C-reactive protein and inflammatory markers: a systematic review
.
J Am Coll Cardiol
2005
;
45
:
1563
9
.
47.
Dupont
J
,
Reverchon
M
,
Cloix
L
,
Froment
P
,
Rame
C
. 
Involvement of adipokines, AMPK, PI3K and the PPAR signaling pathways in ovarian follicle development and cancer
.
Int J Dev Biol
2012
;
56
:
959
67
.
48.
Mor
G
,
Visintin
I
,
Lai
Y
,
Zhao
H
,
Schwartz
P
,
Rutherford
T
, et al
Serum protein markers for early detection of ovarian cancer
.
Proc Natl Acad Sci U S A
2005
;
102
:
7677
82
.
49.
Visintin
I
,
Feng
Z
,
Longton
G
,
Ward
DC
,
Alvero
AB
,
Lai
Y
, et al
Diagnostic markers for early detection of ovarian cancer
.
Clin Cancer Res
2008
;
14
:
1065
72
.
50.
Cuello-Fredes
M
,
Kato
S
,
Abarzua-Catalan
L
,
Delpiano
A
,
Trigo
C
,
Garcia
K
, et al
Leptin promotes a more aggresive behavior of ovarian cancer cells: a potential explanation for a worse prognosis in obese ovarian cancer patients: igcs-0095 ovarian cancer
.
Int J Gynecol Cancer
2015
;
25 Suppl 1
:
S67
.
51.
Diaz
ES
,
Karlan
BY
,
Li
AJ
. 
Obesity-associated adipokines correlate with survival in epithelial ovarian cancer
.
Gynecol Oncol
2013
;
129
:
353
7
.
52.
Wu
MM
,
Chen
HC
,
Chen
CL
,
You
SL
,
Cheng
WF
,
Chen
CA
, et al
A prospective study of gynecological cancer risk in relation to adiposity factors: cumulative incidence and association with plasma adipokine levels
.
PLoS One
2014
;
9
:
e104630
.
53.
Okwan-Duodu
D
,
Umpierrez
GE
,
Brawley
OW
,
Diaz
R
. 
Obesity-driven inflammation and cancer risk: role of myeloid derived suppressor cells and alternately activated macrophages
.
Am J Cancer Res
2013
;
3
:
21
33
.
54.
Charbonneau
B
,
Moysich
KB
,
Kalli
KR
,
Oberg
AL
,
Vierkant
RA
,
Fogarty
ZC
, et al
Large-scale evaluation of common variation in regulatory T cell-related genes and ovarian cancer outcome
.
Cancer Immunol Res
2014
;
2
:
332
40
.
55.
Godoy
HE
,
Khan
AN
,
Vethanayagam
RR
,
Grimm
MJ
,
Singel
KL
,
Kolomeyevskaya
N
, et al
Myeloid-derived suppressor cells modulate immune responses independently of NADPH oxidase in the ovarian tumor microenvironment in mice
.
PLoS One
2013
;
8
:
e69631
.
56.
Govindaraj
C
,
Scalzo-Inguanti
K
,
Madondo
M
,
Hallo
J
,
Flanagan
K
,
Quinn
M
, et al
Impaired Th1 immunity in ovarian cancer patients is mediated by TNFR2+ Tregs within the tumor microenvironment
.
Clin Immunol
2013
;
149
:
97
110
.
57.
Obermajer
N
,
Muthuswamy
R
,
Odunsi
K
,
Edwards
RP
,
Kalinski
P
. 
PGE(2)-induced CXCL12 production and CXCR4 expression controls the accumulation of human MDSCs in ovarian cancer environment
.
Cancer Res
2011
;
71
:
7463
70
.
58.
Nock
N
,
Berger
NA
. 
Obesity and cancer: overview of mechanisms
. In:
Berger
N
, editor.
Cancer and energy balance, epidemiology and overview
:
New York, NY
:
Springer
; 
2010
:
129
79
.
59.
Byers
T
. 
Physical activity and gastric cancer: so what? An epidemiologist's confession
.
Cancer Prev Res
2014
;
7
:
9
11
.
60.
Lynch
BM
. 
Sedentary behavior and cancer: a systematic review of the literature and proposed biological mechanisms
.
Cancer Epidemiol Biomarkers Prev
2010
;
19
:
2691
709
.
61.
Li
S
,
Carlson
E
,
Holm
K
. 
Validation of a single-item measure of usual physical activity
.
Percept Mot Skills
2000
;
91
:
593
602
.
62.
Milton
K
,
Bull
FC
,
Bauman
A
. 
Reliability and validity testing of a single-item physical activity measure
.
Br J Sports Med
2011
;
45
:
203
8
.
63.
Rose
S
,
Elley
CR
,
Lawton
BA
,
Dowell
AC
. 
A single question reliably identifies physically inactive women in primary care
.
N Z Med J
2008
;
121
:
U2897
.
64.
Schechtman
KB
,
Barzilai
B
,
Rost
K
,
Fisher
EB
 Jr.
Measuring physical activity with a single question
.
Am J Public Health
1991
;
81
:
771
3
.
65.
Smith
BJ
,
Marshall
AL
,
Huang
N
. 
Screening for physical activity in family practice: evaluation of two brief assessment tools
.
Am J Prev Med
2005
;
29
:
256
64
.
66.
Weiss
TW
,
Slater
CH
,
Green
LW
,
Kennedy
VC
,
Albright
DL
,
Wun
CC
. 
The validity of single-item, self-assessment questions as measures of adult physical activity
.
J Clin Epidemiol
1990
;
43
:
1123
9
.
67.
Rothman
K
,
Greenland
S
,
Lash
TL
. 
Validity in Epidemiologic Studies
. In:
Rothman
K
,
Greenland
S
,
Lash
TL
, editors.
Modern epidemiology
, 3rd Edition.
Philadelphia, PA
:
Wolters Kluwer/Lippincott Williams & Wilkins
; 
2012
:
128
47
.
68.
Zhong
S
,
Chen
L
,
Lv
M
,
Ma
T
,
Zhang
X
,
Zhao
J
. 
Nonoccupational physical activity and risk of ovarian cancer: a meta-analysis
.
Tumour Biol
2014
;
35
:
11065
73
.
69.
Anderson
JP
,
Ross
JA
,
Folsom
AR
. 
Anthropometric variables, physical activity, and incidence of ovarian cancer: the iowa women's health study
.
Cancer
2004
;
100
:
1515
21
.
70.
Chionh
F
,
Baglietto
L
,
Krishnan
K
,
English
DR
,
MacInnis
RJ
,
Gertig
DM
, et al
Physical activity, body size and composition, and risk of ovarian cancer
.
Cancer Causes Control
2010
;
21
:
2183
94
.
71.
Bodelon
C
,
Cushing-Haugen
KL
,
Wicklund
KG
,
Doherty
JA
,
Rossing
MA
. 
Sun exposure and risk of epithelial ovarian cancer
.
Cancer Causes Control
2012
;
23
:
1985
94
.
72.
Lo-Ciganic
WH
,
Zgibor
JC
,
Bunker
CH
,
Moysich
KB
,
Edwards
RP
,
Ness
RB
. 
Aspirin, nonaspirin nonsteroidal anti-inflammatory drugs, or acetaminophen and risk of ovarian cancer
.
Epidemiology
2012
;
23
:
311
9
.
73.
Wu
AH
,
Pearce
CL
,
Tseng
CC
,
Pike
MC
. 
African Americans and hispanics remain at lower risk of ovarian cancer than non-hispanic whites after considering nongenetic risk factors and oophorectomy rates
.
Cancer Epidemiol Biomarkers Prev
2015
;
24
:
1094
100
.