Table 1 displays the characteristics of the 84,118 participants included in the study from the BRFSS data. Of the participants included, 408 women had endometrial cancer, 10,930 had diabetes, and 106 had both endometrial cancer and diabetes. The majority of women owned their home (69.3%, n = 58,970), reported that they were married or coupled (52.9%, n = 43,625), had healthcare coverage (90.2%, n = 77,946), were of the White race (73.4%, n = 63,200), lived in urban counties (92.9%, n = 73,153) and were employed or self-employed (50.2%, n = 37,635). Most of the participants either attended (32.6%, n = 24,160), or graduated college/technical school (30.4%, n = 33,129), had an income of $50,000 or more (50.7%, n = 32,677), and were aged between 18 and 44 years (44.2%, n = 23,560). The BMI was almost balanced between the different categories of weight with 35.7% of women reporting having normal weight (n = 26,178), 29.5% overweight (n = 22,233), 32.5% obese (n = 22,573), with only 2.3% underweight (n = 1,584). Most of the participants reported that they never smoked (65.5%, n = 50,746) and were not heavy drinkers (93.7%, n = 73,498).

Table 2 displays the frequency distribution of the levels of the different SDOH, biologic, and behavioral determinants among women with and without diabetes mellitus. The weighted chi-squared test results revealed a significant association between diabetes, the main predictor in the study, and the different variables (p < 0.001). The majority of women with diabetes owned their home (72.4%), were married or coupled (46.4%), had healthcare coverage (93.2%), were retired (39.1%), lived in urban counties (90.4%), attended college or technical school (31.7%), had an income higher than $50,000 (31.2%), and were of the White race (66.4%). Moreover, most of the women with diabetes were aged 65 or more (46.5%), obese (57.0%), never smoked (58.2%), and not heavy drinkers (97.5%).

Table 1, Supplementary Table S1 and Table 3 present the respective unadjusted and adjusted measures of associations between diabetes, SDOH, biologic and behavioral factors, and endometrial cancer expressed in terms of Cramer’s V, unadjusted and adjusted ORs. Weighted Cramer’s V showed mild effect size and magnitude of association between most of the variables and endometrial cancer (Supplementary Table S1). In addition, the unadjusted analysis revealed significant crude associations between majority of the variables and endometrial cancer, except for healthcare coverage, urban/rural areas of residency, education and income levels (Table 1, and Supplementary Table S1). With respect to the adjusted analysis, our results showed that women with diabetes had a 54% increase in the odds of endometrial cancer (approximately double the odds), compared to women without diabetes (OR 1.54; 95%CI: 1.01–2.34).

Our adjusted analysis (Table 3) also showed that indices of SDOH and biologic factors had significant associations with endometrial cancer. However, none of the behavioral factors presented a significant association with this type of cancer.

The indices of SDOH that were associated with increased odds of endometrial cancer included the level of education of attending college or technical schools with an 83% associated increase in the respective odds of endometrial cancer compared to the level of education of graduated high school (OR 1.83; 95%CI: 1.12-3.00). Moreover, the biologic factors that were also associated with an increase in the odds of endometrial cancer included older age and obesity. In this regard, women whose ages were between 45 and 54 (OR 2.75; 95%CI: 1.01–7.71), 55 and 64 (OR 4.20; 95%CI: 1.61–10.92), and 65 or older (OR 7.21; 95%CI: 2.76–18.82) were shown to have about 3- to 7-fold increase in the estimated risk of endometrial cancer compared to the younger reference age group of 18 to 44. In addition, women who were considered obese were 3 times more likely to have endometrial cancer compared to women of normal weight (OR 3.10; 95%CI: 1.96–4.90).

On the other hand, the SDOH that were associated with a decrease in the odds of endometrial cancer compared to their respective reference categories (indicated in Table 3) included women who reported renting a home or had other arrangements for homeownership, were divorced or separated, had higher ranges of income, were of Black or other races. In this regard, renting a home (OR 0.50; 95%CI: 0.28–0.88) or having other arrangements (OR 0.05; 95%CI: 0.02–0.16) for statuses of home ownership were associated with 50 and 95% lower odds of endometrial cancer compared to owning a home. Moreover, being divorced or separated as marital status was shown to be associated with a 45% decrease in the odds of endometrial cancer (OR 0.55; 95%CI: 0.30–0.99) compared to being married or coupled. Along the same lines, a higher annual income of $35,000 to $50,000 (OR 0.35; 95%CI: 0.16–0.78), and $50,000 or more (OR 0.29; 95%CI: 0.14–0.62) were income categories that were associated with respective 65 and 71% decrease in the odds of endometrial cancer compared to the lower income category of less than $15,000. As for race, Black women (OR 0.24; 95%CI: 0.07–0.84) and women of other races (OR 0.37; 95%CI: 0.15–0.88) showed respective decreases of 66 and 63% in the odds of endometrial cancer compared to White women.

Our main multivariable analysis was followed by a number of confirmatory additional analyses that took into consideration several conditions.

First, given the cross-sectional nature of our study design, we anticipated that some women might have had endometrial cancer before diabetes. Accordingly, we performed additional multivariable analysis in which we excluded women (32 women in total) who were diagnosed with endometrial cancer before their diagnosis with diabetes. Our results (not shown for all the variables) were not substantially affected by this left censoring, and diabetes was still significantly associated with endometrial cancer (OR 1.81; 95%CI: 1.11–2.94; p = 0.017).

Then we carried out new analyses in which we incorporated the age of diagnosis with diabetes in one multivariable model, and the duration of diabetes in another multivariable model, with study population being exclusive to women with diabetes (results not shown for all the variables). Our results showed that age of diagnosis with diabetes (OR 0.97; 95%CI: 0.95-1.00; P = 0.104), and duration of diabetes (OR 1.02; 95%CI: 0.99–1.05; p = 0.108) did not have significant associations with endometrial cancer.

Our original analysis assumed that the control group is comprised of women who specifically did not have endometrial cancer. To confirm these denoted associations, we conducted further analysis in which we considered our control group as women who did not have any type of cancer (results not reported for all the variables). The new results showed that diabetes continued to be a significant predictor of endometrial cancer with OR = 1.54, 95%CI: 1.01–2.35, P = 0.043.

Lastly, we carried out additional sub-analysis that focused on women who reported having more than one type of cancer and age of diagnosis was recorded for the first type of cancer (total of 2854 women). In this analysis we aimed to determine if age of diagnosis with other types of cancer was associated with the odds of endometrial cancer, along with our main predictors which included diabetes and determinants of health. Results of this analysis are presented in Table 4 and showed that age of diagnosis with other types of cancer was not significantly associated with endometrial cancer P = 0.18, but a significant association was still present between diabetes and endometrial cancer in this subpopulation of women (OR = 2.28, 95%CI: 1.02–5.12, P = 0.04).

In addition to identifying diabetes as a significant predictor of endometrial cancer in all of the aforementioned analyses, our multivariable models also revealed strong relationships between this type of cancer and several determinants of health. These indices included, but were not limited to, homeownership, marital status, education, income, age and BMI; thus, confirming the link between these determinants of health and endometrial cancer.