Research Communication: An Observational Cohort Study on Risk of Ovarian Cancer in Women with Irritable Bowel Syndrome.

We analysed administrative claims data using Optum’s de-identified Clinformatics ® Data Mart Database, which represents ~75 million members of large commercial insurance and Medicare Advantage health plans from all 50 states. We identified women with and without an initial diagnosis of IBS between January 1, 2017 to December 31, 2020 by the presence of ≥2 ICD-10 codes for IBS in any position occurring ≥30 days apart. Women in the reference group were randomly selected among those without an IBS diagnosis. Women with a history of ovarian cancer and those without continuous enrolment ≥12 months before and after the index date were excluded. Data were collected on baseline demographics, socioeconomic status, and 28 medical comorbidities (i.e., Elixhauser comorbidities). The inverse propensity score weighting method was used to balance groups ( Supplemental Methods ). Ovarian cancer incidences at any time after the index date were identified in both groups using ICD-10 codes and the incidence rate was compared using Poisson regression. We performed weighted survival analyses to compare time to ovarian cancer incidence as the primary analysis to estimate potential cases of ovarian cancer that were misdiagnosed as IBS. The proportional hazards assumption for IBS was assessed by evaluating the Schoenfeld residuals. When the assumption was violated, the coefficient of IBS was allowed to be time-varying. Its functional form was informed by the Schoenfeld residual plot. Restricted cubic splines were used to examine how ovarian cancer risk changed with age. A sensitivity analysis was conducted to compare time to ovarian cancer using only a single ICD-10 code to define IBS to assess the potential impact of a more sensitive, less specific IBS definition and to evaluate whether a provisional IBS diagnosis influenced observed associations. We explored the effect of endometriosis, a shared comorbidity, by including a diagnosis of endometriosis prior to the index date in the model.

We identified N=79,804 women with and N=79,804 without IBS. All standardised mean differences in baseline characteristics between IBS and non-IBS participants were <0.1, indicating groups were well-balanced after inverse propensity score weighting ( Supplemental Table ). The unadjusted incidence of ovarian cancer at any time (IRR=1.52; 95% CI: 1.24-1.87) after the index date was higher in women with IBS than those without (p<0.01). The test for a trend in the Schoenfeld residual plot was non-significant, suggesting no statistical violation of proportional hazards assumption. However, visual inspection of the Schoenfeld residual plot revealed a non-linear trend, indicating that a piecewise linear function of time as the coefficient for IBS to be appropriate. Comparisons ( Figure 1A ) between IBS and non-IBS participants using a time-varying coefficient for IBS demonstrated the rate of ovarian cancer was high immediately after index IBS diagnosis and decreased over time. The rate of ovarian cancer was higher at 3 (hazard ratio [HR]=1.71, 95% CI: 1.08-2.70) and 6 months (HR=1.43, 95% CI: 1.06-1.93), but not at later time points after 8 months, in women with IBS compared to those without (p=0.02 for both). Sensitivity analysis ( Figure 1B ) among women with only a single IBS diagnostic code (N=183,518) and women without IBS (N=183,518) demonstrated similar results, but with larger effect sizes and a statistically significant association through 9 months. In the sensitivity analysis, the rate of ovarian cancer was higher in women with IBS than those without at 3 (HR=2.12; 95% CI: 1.57-2.87; p<0.01), 6 (HR=1.58; 95% CI: 1.29-1.93; p<0.01), and 9 (HR=1.23; 95% CI: 1.05-1.45; p=0.01) months, but not after 9 months. The evaluation of the association between age and ovarian cancer risk using restricted cubic splines suggested a piecewise linear effect of age with a change point at 50 years ( Figure 2 ). After modelling the age effect using the piecewise linear function, the association between increasing age and higher ovarian cancer incidence was significant only in women <50 years (HR=1.07; (95% CI: 1.04-1.10, p<0.01). This pattern was observed in both IBS and non-IBS groups and in the primary analysis. In the sensitivity analyses, risk of ovarian cancer increased both before and after 50 years; however, the magnitude of effect was greater in those <50 years. Adjusting for endometriosis did not alter the effects of IBS and age on ovarian cancer diagnosis. Ovarian cancer risk remained significantly higher during the first 8 months in women with IBS compared to those without in the absence (3-month HR=1.71, p=0.02; 6-month HR=1.43, p=0.02) of endometriosis. In the presence of endometriosis, ovarian cancer risk remained significantly higher during (3-month HR=4.20, p=0.01; 6-month HR=3.52, p=0.01) and after (HR=2.67, p=0.04 after year 1) the first 8 months.

Our results suggest that women with a new IBS diagnosis may be at increased risk for ovarian cancer, but this risk varies by time from IBS diagnosis. Risk of ovarian cancer was significantly higher during the first 8-9 months after IBS diagnosis, but there were no significant differences in risk of ovarian cancer at later time points. Findings suggest that some women with ovarian cancer may initially be misdiagnosed as IBS, particularly if IBS is considered a provisional diagnosis. However, an established diagnosis of IBS does not appear to be causally linked to ovarian cancer. We explored whether the association between ovarian cancer and IBS could be related to factors such as endometriosis that independently increase the risk of both conditions. Results of our analyses revealed that although the risk of ovarian cancer was substantially higher in IBS compared to non-IBS controls among women with endometriosis, presence of endometriosis did not confound the relationship between IBS and ovarian cancer. Endometriosis was an independent risk factor for ovarian cancer. Our findings are consistent with a prior population-based cohort study, which demonstrated that endometriosis was associated with a 4-fold higher risk of ovarian cancer. 7 Risks were further elevated among those with ovarian endometriomas and/or deep infiltrating endometriosis. Epidemiological studies have demonstrated separately that the risk of IBS is 2-3 times higher in women with endometriosis than those without. 8 However, research on the relationship between endometriosis and IBS has been relatively limited to date. Whether the presence of comorbid gynaecological conditions in women with IBS should prompt earlier screening for ovarian cancer due its potential role as a shared risk factor warrants further study. Of note, we observed that increasing age was a risk factor for ovarian cancer in those younger than 50 years and that the incidence of ovarian cancer remained consistently high compared to younger ages after 50 years, which coincides with the average age of menopause. This finding suggests a potential interaction between hormonal changes and ovarian cancer risk. A few studies have examined the impact of menopause on IBS symptoms, suggesting that sex hormones modulate gastrointestinal function and visceral sensation. 9 One recent study found that most DGBI symptoms were more prevalent in pre-menopausal women than post-menopausal women, 10 possibly due to the stabilisation of chronic low oestrogen levels in the latter group. Similarly, hormonal drivers such as oestrogen exposure, which influence ovarian cancer risk, decline after menopause. Although ovarian cancer is more commonly diagnosed in older women, its age-related incidence may stabilise post-menopause, despite an increase in absolute incidence due to accumulated lifetime risk. These findings suggest the importance of hormonal- and age-related factors in the diagnostic evaluation of women with IBS. To date, the potential overlap of ovarian cancer and IBS has been poorly studied. Study strengths include the large sample size and access to claims data from both commercial insurance and Medicare Advantage plans. Limitations include the use of diagnostic codes to identify IBS cases, which may be susceptible to misclassification or reflect symptom presentations rather than confirmed diagnoses based on validated criteria. Future studies are needed to externally validate these observations and investigate the clinical impact of screening for ovarian cancer prospectively with non-invasive testing in women with IBS, particularly in those with atypical (e.g., fixed and persistent abdominal bloating) IBS symptoms. Identifying patient-specific risk factors, such as chronic pelvic pain or endometriosis, could further help develop tailored risk profiles and improve the approach to personalised care in women with IBS-type symptoms.

Ovarian cancer is a leading cause of cancer deaths among women. In the United States, approximately 20,890 women will receive a new diagnosis of ovarian cancer and 12,730 will die from ovarian cancer in 2025. 1 Despite declining incidence rates in recent years, which have been attributed to various factors such as the increased use of oral contraceptives and decreased use of menopausal hormone therapy, the death-to-incidence ratio of ovarian cancer is notably high, because ovarian cancer is generally diagnosed at an advanced stage. 2 Early diagnosis is made difficult because the disease is often asymptomatic or characterised by non-specific symptoms such as abdominal bloating or distention, nausea, altered bowel habit, or back pain. 3 These symptoms can overlap with symptoms of common non-malignant conditions such as irritable bowel syndrome (IBS), a disorder of gut-brain interaction (DGBI). Although the likelihood of missed structural gastrointestinal diseases after a clinical diagnosis of IBS is very low, 4 it is not known how often symptoms of ovarian cancer in women are misdiagnosed as IBS. Diagnostic screening for ovarian cancer is considered in some, 5 but not all, IBS guidelines. 6 To address the knowledge gap, we conducted a retrospective cohort study comparing ovarian cancer incidence in women with and women without a new diagnosis of IBS.