Physical activity, sedentary behavior and ovarian cancer risk in Asian populations: a scoping review and meta-analysis Authors: Jun Wang, Dongdong Wu, Xiang Guo Published: 2026 ABSTRACT BackgroundOvarian cancer (OC) exhibits significant ethnic and geographic disparities, with rising incidence in Asia contrasting global declines. While non-modifiable risk factors (e.g., genetics) are well-established, evidence for the role of modifiable factors like physical activity remains inconsistent, especially in Asian populations. This study systematically evaluates the association between physical activity, sedentary behavior, and OC risk in Asia.MethodsWe conducted a meta-analysis of seven studies (156,910 participants, 1,585 OC cases) from Japan, Korea, and China. Physical activity and sedentary behavior were assessed via self-reports. Meta-analysis was performed to pool RR estimates together with their 95% CI.ResultsRegular physical activity was associated with a 24% lower OC risk (RR = 0.76, 95% CI: 0.64–0.91; P = 0.002), despite high heterogeneity (I2 = 75%). Sedentary behaviour was associated with a 55% higher OC risk (RR = 1.55, 95% CI: 1.23–1.96; P = 0.0002).ConclusionsIn Asian populations, physical activity may protect against OC, while sedentary behavior may elevate risk. These findings underscore the need for dual public health strategies: promoting exercise and reducing sedentary time. Future research should standardize exposure assessments and include diverse Asian subpopulations to refine prevention guidelines.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12885-026-16079-y. Introduction Ovarian cancer (OC) is one of the most aggressive malignancies of the female reproductive system, associated with poor early detection rates and the worst prognosis among gynecological cancer [1, 2]. In 2020, there were about 313,959 new diagnosed cases and 207,252 associated deaths globally [3]. Projections suggest substantial increases in burden by 2040, with a near 100% rise in new cases and deaths among the low Human Development Index (HDI) countries compared to 20%–30% in the very high HDI countries [4]. The epidemiology of OC exhibits notable ethnic and geographic disparities [5, 8]. Research data from the 2020 Global Cancer Observatory (GLOBOCAN) database shows that at the national level, the incidence of ovarian cancer is associated with the gross domestic product (GDP) and human development index (HDI) [9]. While Asia has experienced a rising incidence and a trend toward younger patient populations, Europe and North America show a decline [5, 6]. OC patients in Asia exhibit different disease characteristics and sub-types from individuals of Western countries, and their incidence rates and 2-year relative survival rates also differ from those of patients in White people in Europe and America [10, 14]. From 1990 to 2019, both the age-standardized incidence rate and mortality rate of OC in China continued to rise, surpassing those in the world average, Japan and Singapore [15]. Currently, multiple established risk factors have been confirmed, which are non-modifiable factors such as race, BRCA1/2 mutation, and family history, etc. [16, 17]. Therefore, identification of modifiable risk factors is critical for reducing the OC burden [18, 19]. Physical activity, as a modifiable lifestyle factor, has been shown to reduce chronic inflammation associated with various cancers and enhance immune surveillance by increasing the number and function of monocytes, macrophages, natural killer (NK) cells, and dendritic cell [20]. A meta-analysis of 3,708,313 women indicated that diabetes is associated with an increased risk of OC, particularly among Asian populations [21]. In China, high fasting plasma glucose levels and elevated body mass index (BMI) were identified as the leading risk factors for OC mortality in both 1990 and 2019, at both global and national levels [22]. As physical activity is critically involved in regulating glycemic control and adiposity [23, 24], it seems to represent an important modifiable factor relevant to ovarian cancer risk. Sedentary behavior is defined as a state of sitting, lying or reclining without physical activity (or any state of consciousness where the body’s energy expenditure is less than or equal to 1.5 MET) [25]. The relationship between sedentary behavior and physical activities is not an "either-or" situation. Modern research further supports the view that these two factors should be regarded as two separate health risk factors, which jointly influence health outcomes [26, 27]. Although multiple studies have examined the association between physical activity and OC risk, the results of epidemiological findings remain inconsistent [17, 28, 33]. Furthermore, there have been no reports specifically for Asian populations. In this work, the data of Asian population was extracted from the published literature and systematically analyzed to try to explore the potential underlying patterns. Materials and methods Search strategy We Searched relating to OC in English and Chinese up to 2nd, February, 2025, including PubMed (https://www.ncbi.nlm.nih.gov/pubmed), Embase (https://www.elsevier.com/solutions/embase-biomedical-research), the China National Knowledge Infrastructure Databases (CNKI, http://www.cnki.net). We used the following search term: ((((((((ovarian cancer[Title/Abstract]) OR (ovary neoplasm[Title/Abstract])) OR (cancer of ovary[Title/Abstract])) AND (risk[Title/Abstract])) AND (physical activity[Title/Abstract])) OR (exercise[Title/Abstract])) OR (sedentary behavior[Title/Abstract])) OR (sitting time[Title/Abstract])) OR (watching TV[Title/Abstract]). Literature search was performed independently by two reviewers, with blinding to authors and institutions to reduce selection bias. Eligibility criteria The inclusion criteria were as follows: (1) studies reporting an association between physical activity or sedentary behavior and the risk of OC; (2) studies that directly reported risk ratio (RR) with 95% confidence intervals (CIs) or provided sufficient data to allow RR calculation; and (3) studies focusing exclusively on native Asians population. Exclusion criteria included reviews, meta-analysis, and original studies that did not provide RR estimates. In cases of duplicate publications, only the study with the most complete dataset was retained. Data extraction All relevant studies screening and data extraction were conducted independently by two researchers according to the eligibility criteria. If differences are encountered, they are resolved through discussion or by seeking the opinion of a third researcher. Data extraction includes: last name of the first author, publication year, study design, country, sample size, recruitment period, length of follow-up, physical activity measurement, RR, or OR with 95% confidence interval (CI), and adjustments. Methodological quality was assessed using Newcastle–Ottawa Scale, which awarded points for the following criteria: representativeness of the exposure group (1 point), representativeness of the non-exposed group (1 point), identification of exposure factors (1 point), non‐assessment of the outcome of interest (1 point), comparability of the exposed and non-exposed groups (2 points), assessment of outcome (1 point), adequate follow-up duration (1 point), and completeness of follow-up(1 point). Quality assessment scores were summarized in Table S2. Statistical analysis Meta-analysis was performed using RevMan 5.3 software to pool OR, RR or HR estimates together with their 95% CI. Given that ovarian cancer is a rare disease, ORs and HRs were considered approximate estimates of RR and were pooled together as relative risks. Log-transformation effect sizes and SE were calculated and used to quantitatively evaluate the association between physical activity, sedentary behavior, and ovarian cancer risk. A P value of less than 0.05 was considered statistically significant. The heterogeneity was assessed using the I2 statistic and Cochran’s Q test. An I2 value > 50% was considered indicative of substantial heterogeneity. When the heterogeneity was large, a random-effects model with the inverse-variance method was used for the meta-analysis. Results Literature search An initial database search yielded approximately 55,384 studies related to the search terms. However, only 61 studies passed the abstract screening phase, as the remaining records were excluded due to being review articles, lacking reports on OR risk, classified as case studies, or identified as duplicated. Following eligibility assessment, seven Asian studies were ultimately included in the meta-analysis to evaluate the associations between physical activity and subsequent OC risk, of which three specifically examined the association between sedentary behavior and OC risk. The search flowchart and PRISMA-ScR checklist were give in Figure S1 and Table S1. Study characteristics and quality assessment The basic characteristics of the seven included studies are presented in Table 1, with five studies examining the association between physical exercise and the risk of OC, and 3 studies were about the association between sedentary behavior and OC risk. Notably, the study by Andy H et al. (2013) investigated both physical activity and sedentary behavior in relation to ovarian cancer risk. A total of 156,910 participants, including 1,585 women diagnosed with OC, were recruited from Japan, Korea, and China and followed for 5 to 13 years. Physical activity intensity in five included studies were all classified based on participants’ self-reported questionnaire responses. Among them, two studies reported clear subjective perceptions of moderate-intensity activity. Additionally, individuals with regular physical activity were also defined as a high-quality population. Given the specific characteristics of physical activity research, we adopted a priori definition and exclusion strategies to minimize misclassification bias. In contrast to physical activity, we redefined sedentary behavior as excessive television watching and prolonged sitting. Methodological quality of all included studies was evaluated using the Newcastle–Ottawa Scale (NOS). The detailed quality assessment results are summarized in Table S2.Table 1Summary characteristics of included studiesStudyLocation/periodAgecases/cohortLength of follow up (year)Type of measurementRR/OR(95%CI)Adjustment for covariatesElisabete, et al. (2012); (cohort study) [45]Japan 1990–199446–6986/45748average 7.6physical activity during leisure timeRR = 1.002 (0.618–1.626)age and study area adjustedMa et al. (2013); (cohort study) [34]China 1996–200040–70152/70258more than 9regular physical activityRR = 1.063 (0.764–1.479)age and education adjustedSoseul, et al. (2022); (case control study) [63]Korea 2008–201552.4 ± 10.3534/26557regular exercise enough to sweatOR = 0.41 (0.28–0.6)age, education and oral contraceptive useZhang, et al. (2003); (case control study) [35]China 1999–2000under 75254/9065moderate activityOR = 0.59 (0.31–1.12)age at diagnosis, education, living area, BMI, smokingAndy, et al. (2013); (case control study) [46]China 2006–200859 ± 5.68500/10005total physical activity (moderate)OR = 0.82 (0.6–1.11)age, parity, oral contraceptive use, BMIShigekazu, et al. (2018); (cohort study)Japan 1988–199040–7959/3475813high versus low levels of watching TVRR = 1.81 (0.75–4.39)age, parity, oral contraceptive use, BMI, education, etcZhang, et al. (2004); (case control study) [36]China 1999–2000under 75254/9065high versus low levels of sitting while watching TVOR = 3.39 (1–11.5)age, locality, education, family income, BMIAndy, et al. (2013); (case control study)China 2006–200859 ± 5.68500/10005high versus low levels of sittingOR = 1.07 (0.77–1.48)age, parity, oral contraceptive use, BMI Table 1Summary characteristics of included studiesStudyLocation/periodAgecases/cohortLength of follow up (year)Type of measurementRR/OR(95%CI)Adjustment for covariatesElisabete, et al. (2012); (cohort study) [45]Japan 1990–199446–6986/45748average 7.6physical activity during leisure timeRR = 1.002 (0.618–1.626)age and study area adjustedMa et al. (2013); (cohort study) [34]China 1996–200040–70152/70258more than 9regular physical activityRR = 1.063 (0.764–1.479)age and education adjustedSoseul, et al. (2022); (case control study) [63]Korea 2008–201552.4 ± 10.3534/26557regular exercise enough to sweatOR = 0.41 (0.28–0.6)age, education and oral contraceptive useZhang, et al. (2003); (case control study) [35]China 1999–2000under 75254/9065moderate activityOR = 0.59 (0.31–1.12)age at diagnosis, education, living area, BMI, smokingAndy, et al. (2013); (case control study) [46]China 2006–200859 ± 5.68500/10005total physical activity (moderate)OR = 0.82 (0.6–1.11)age, parity, oral contraceptive use, BMIShigekazu, et al. (2018); (cohort study)Japan 1988–199040–7959/3475813high versus low levels of watching TVRR = 1.81 (0.75–4.39)age, parity, oral contraceptive use, BMI, education, etcZhang, et al. (2004); (case control study) [36]China 1999–2000under 75254/9065high versus low levels of sitting while watching TVOR = 3.39 (1–11.5)age, locality, education, family income, BMIAndy, et al. (2013); (case control study)China 2006–200859 ± 5.68500/10005high versus low levels of sittingOR = 1.07 (0.77–1.48)age, parity, oral contraceptive use, BMI Summary characteristics of included studies Physical activity and OC risk in Asian population A total of five studies were included for the analysis of the association between physical exercise and the risk of OC (Fig. 1). The reported data from these studies classified individuals who engaged in leisure-time physical activity, maintained regular exercise routines, or participated in moderate-intensity exercise into a single category, which was then compared with those reporting no habitual physical activity (Fig. 1). Meta-analysis revealed substantial heterogeneity across the studies (I2 = 75%); however, the overall effect was statistically significant (Z = 3.03, P = 0.002). A random-effects model with the inverse-variance method was used for the meta-analysis. The analysis demonstrated that individuals with exercise habits had a significantly lower risk of OC compared to those without such habits (RR = 0.76, 95% CI: 0.64–0.91; P = 0.002). Given the small number of included studies, the statistical power and reliability of publication bias tests are generally considered limited and unreliable. Therefore, we did not perform a formal publication bias analysis in the present study.Fig. 1The association between physical activities and the risk of ovarian cancer. Exercise habits (experimental) vs no-physical activity (control) The association between physical activities and the risk of ovarian cancer. Exercise habits (experimental) vs no-physical activity (control) Sedentary behavior and OC risk in Asian population Only three included studies reported data on sedentary behavior. Despite the limited number of included studies, the current analysis demonstrated a statistically significant positive association between sedentary behavior and the risk of ovarian cancer (RR = 1.55, 95%CI: 1.23–1.96, P = 0.0002) (Fig. 2). Although heterogeneity across studies was substantial, the overall effect remained significant (Fig. 2).Fig. 2The association between sedentary behavior and the risk of ovarian cancer. RR > 1 indicates a higher risk associated with higher level of sedentary behavior The association between sedentary behavior and the risk of ovarian cancer. RR > 1 indicates a higher risk associated with higher level of sedentary behavior Discussion Asian women exhibit considerable heterogeneity in genetic background and sociocultural context, which may underlie regional and ethnic variations in the distribution of histological subtypes, risk factors, and biomarkers of ovarian cancer [37, 39]. This review synthesizes current evidence on the association between physical activity and ovarian cancer risk in Asian populations and reveals that, despite substantial inter-study heterogeneity, a consistent trend emerges: moderate physical activity is associated with a reduced risk of ovarian cancer [40, 42]. This pattern contrasts, to some extent, with findings from European and American populations, where large-scale prospective cohort studies have largely failed to demonstrate a significant protective effect of physical activity against ovarian cancer [28, 30, 43, 44]. Notably, Mendelian randomization analyses—leveraging genetic variants as instrumental variables—have found no strong causal association between genetically predicted physical activity levels and overall ovarian cancer risk [28]. In contrast, within the Asian setting, multiple observational studies consistently report that regular engagement in moderate- to vigorous-intensity physical activity is linked to lower ovarian cancer incidence, with a more pronounced protective effect observed among individuals who engage in at least two hours of exercise per week [34, 35, 45, 46]. Sedentary behavior, as a risk exposure pattern distinct from physical activity, is increasingly recognized in the epidemiological literature for its consistent positive association with elevated ovarian cancer risk [32, 47, 48]. In both Asian and Western populations, prolonged sitting has been repeatedly identified as a potential carcinogenic factor [32, 36, 48, 49]. The biological mechanisms underlying this association may include the accumulation of metabolic disturbance [50], such as heightened insulin resistance [51, 52], elevated chronic low-grade inflammation [53], and dysregulation of sex hormone metabolism [54, 55], which can promote tumorigenesis independently of engagement in regular physical activity. Notably, several studies report only a weak correlation between sedentary time and moderate-to-vigorous physical activity, indicating that these behaviors reflect distinct physiological and behavioral domains and are not interchangeable as health indicators [25, 26, 56, 57]. Consequently, individuals who engage in regular exercise may still face an increased risk of ovarian cancer if they concurrently accumulate high levels of sedentary time [47, 58]. This evidence underscores the importance of targeting sedentary behavior reduction as an independent public health objective, rather than assuming it is adequately mitigated by physical activity alone [25, 58]. From a biological mechanistic standpoint, there is currently no robust evidence indicating that physical activity during adolescence or the post-menopausal period disrupts normal ovarian development or promotes malignant transformation [30, 43, 44]. In contrast, a growing body of evidence supports the anti-cancer effects of regular exercise through multiple interconnected biological pathways [59, 61]. These include, among others, reducing BMI, modulating insulin and insulin-like growth factor-1 (IGF-1) signaling, maintaining homeostasis of reproductive hormones such as estrogen and progesterone, improving the pro- and anti-inflammatory cytokine balance within the tumor microenvironment [62], and enhancing immune surveillance to suppress tumor initiation and progression [19, 20]. Notably, given the escalating global prevalence of obesity and type 2 diabetes—particularly across Asian populations—and the well-established role of these conditions as modifiable risk factors for epithelial ovarian cancer, the capacity of physical activity to serve as a non-pharmacological strategy for mitigating adiposity and improving glucose and lipid metabolism is of critical importance. Thus, although a definitive causal link between physical activity and ovarian cancer risk remains to be fully elucidated, its proven efficacy in preventing and managing obesity-related metabolic dysregulation underscores its substantial public health value. The review-analysis included in this review offers several methodological strengths, including the incorporation of multiple high-quality prospective cohort studies, large sample sizes, extended follow-up durations, and—importantly—a dedicated focus on East Asian populations (encompassing China, Japan, and South Korea) [34, 36, 45, 46, 49, 63]. This regional specificity enhances the applicability and representativeness of the findings for Asian populations. Nevertheless, several limitations warrant careful consideration. First, the number of primary studies specifically designed to examine the association between physical activity/sedentary behavior and ovarian cancer in Asian populations remains limited; most data derive from secondary analyses of large multifactorial cohorts, which may introduce measurement bias and inadequate control of potential confounders. Second, physical activity and sedentary behavior were were not uniformly assessed, often relying on self-reported measures, which introduces potential recall bias and misclassification, particularly across different cultural contexts. Given the limited number of studies for sedentary behavior, and the potential substantial heterogeneity across definitions of sedentary behavior (prolonged TV viewing vs. total sitting time), caution is warranted in the interpretation of the pooled results. Furthermore, reverse causation may be a potential limitation. Early-stage ovarian cancer often presents with atypical, non-specific, and subtle clinical symptoms that are easily neglected or misdiagnosed. As a result, women with undiagnosed early ovarian cancer may change their physical activity or sedentary patterns due to subclinical illness, which could distort the observed associations and influence the interpretation of the results. In addition, epithelial ovarian cancer exhibits distinct histological subtype distributions across ethnic groups. Clear cell carcinoma is the most prevalent subtype in Asian populations, in contrast to the predominance of serous carcinoma in Western populations, which may be driven by genetic and environmental differences. Since histological subtype can influence risk factor associations, the absence of subtype stratification in this study restricts the interpretation of our results among Asian populations. Lastly, due to constraints in data availability, subgroup analyses by subregional geographic areas, urbanization levels, or Human Development Index (HDI) within Asia could not be conducted, thereby limiting insights into potential intra-Asian disparities. Conclusions In summary, epidemiological evidence in Asian populations generally supports a direct protective effect of physical activity against ovarian cancer, with most studies reporting either an inverse association or no significant relationship. In contrast, sedentary behavior consistently demonstrates a stable positive association with increased ovarian cancer risk across nearly all available studies. This pattern underscores the importance of public health strategies that not only promote higher levels of physical activity but also actively target the reduction of prolonged sedentary time. Given the rising burden of obesity and diabetes among Asian women, encouraging moderate-intensity exercise represents not only a promising avenue for ovarian cancer prevention but also a cornerstone of broader efforts to improve overall health. Future research must prioritize well-designed, prospective investigations employing standardized assessment tools, extended follow-up periods, and inclusion of diverse Asian subpopulations to elucidate the differential effects of exercise patterns, intensities, and life-course timing on ovarian carcinogenesis. Such evidence is essential to inform the development of precise, population-specific prevention guidelines. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 1. Supplementary Material 2. Supplementary Material 3.