The uterus responds to cardiovascular disease endothelial inflammation like other end organs, yet it is not typically viewed in this way. Insulin resistance (IR) and other markers of metabolic syndrome are well established risk factors of cardiovascular disease-related end organ damage.^1,2,3 Whilst IR and metabolic disorders are established risk factors for polycystic ovarian syndrome (PCOS),⁴ they also play a role in other causes of abnormal uterine bleeding (AUB). Abnormal uterine bleeding is described as menstrual flow outside of normal volume, duration and regularity.⁵ Historically, the focus of IR within the field of gynaecology is primarily dedicated to PCOS, yet gynaecologic literature indicates that risk factors for uterine fibroids, endometrial polyps and polycystic ovarian syndrome include obesity, hypertension and metabolic syndrome.⁶ However, few studies have examined the presence of IR amongst women with abnormal uterine bleeding as a whole.

Insulin is a key anabolic hormone involved in multiple metabolic pathways in the body, including glucose uptake and lipid storage, cell growth and proliferation in the body and nitric oxide (NO) synthesis by endothelial cells.^2,7,8 Insulin resistance occurs when increasingly larger amounts of insulin are required to regulate the body’s metabolic pathways, and even in the face of these increasing insulin levels, cellular activity is diminished compared with a normal person.^2,7,8 Insulin resistance is a multifactorial disease that has become an epidemic in the United States (US). Although the number of adults affected by IR is not known, one in three adults in the US has prediabetes and an additional 10.5% of adults has diabetes mellitus type 2 (DM2).^2,9 These numbers are expected to double in the next 10 years because almost 90.0% of Americans have risk factors for IR.⁹

Although angiogenesis is a relatively infrequent event in most adult tissues, it occurs repeatedly in the uterus in order to support the monthly cycle of growth, shedding and repair of the endometrial lining.¹⁰ Both the process of angiogenesis and endometrial repair are tightly regulated by oestrogen and progesterone, which involve NO, platelets, immune cells and multiple growth factors.^{6,10,11,12,13} Decreased NO production, because of IR, results in an increase in the shear stress on endothelial cells because of diffuse vasoconstriction.^2,8 Moreover, decreased NO, hyperglycaemia and the state of IR itself promote the release of inflammatory cytokines and free radical formation, which leads to further endothelial damage, surrounding tissue hypoxia, and, eventually, cellular apoptosis.^2,8 The resulting vascular permeability is worsened by high levels of circulating glucose that damages the endothelial cell glycocalyx.^2,8,14 Together, these vascular changes allow substrates to leave the blood and form plaques in the tunica intima layer of the blood vessel.⁸ This thickening of the blood vessels leads to increased vascular resistance, chronic low-grade inflammation and atherosclerosis.² Atherosclerosis is a widely accepted cause of numerous microvascular and macrovascular diseases that are significantly associated with IR.^{2,15,16,17,18} Studies have demonstrated that women with fibroids and PCOS are more likely to have risk factors for cardiovascular disease,⁶ but AUB and its other causes are not currently thought of as an atherosclerotic-related end organ disease.

Applying this knowledge to the uterus, chronic low-grade inflammation likely reduces the spiral arteries’ ability to meet the high metabolic demands of the endometrium. This leads to an overexpression of angiogenic cytokines and dysfunctional angiogenesis with excessive uterine vessel formation and dilation in an attempt to increase the amount of uterine blood flow.^10,12,19,20 It also results in functionally impaired proliferative and secretory phases of the menstrual cycle where the endometrial lining is not properly repaired during cyclical shedding.²⁰ Similar inflammatory cytokines and fibrosis-promoting growth factors seen in atherosclerotic disease encourage smooth muscle metaplasia and scarring in the uterus.^{19,21,22,23,24,25,26,27} In fact, Moss and Benditt (1975) revealed the similar growth and behaviour of the atherosclerotic plaque cell with the leiomyoma cell.²⁸ The sentinel event postulated in both uterine leiomyoma and atherosclerotic plaque development reveal smooth muscle proliferation from a single progenitor myocyte.²⁹ Features suggesting related pathology were also highlighted in other studies that revealed similar clonality and ability to fibrose and calcify.^30,31

This molecular and immunologic dysregulation leads to several forms of AUB that result in commonplace gynaecologic pathology. Abnormal uterine bleeding is the most common reason for outpatient gynaecologic consultation in the US.³² In 2011, the International Federation of Gynecology and Obstetrics (FIGO) created a uniform system of classification to describe uterine bleeding disorders called PALM-COEIN (polyp, adenomyosis, leiomyoma, metaplasia, coagulopathy, ovulatory, endometrial, iatrogenic and not otherwise specified).³³ Whilst the approach to treating PCOS-related obesity and menstrual irregularity includes lifestyle changes by reducing cardiovascular risk factors, there is little information on how this management strategy can prevent or improve AUB from endometrial polyps, adenomyosis, leiomyoma or endometrial causes. On the other hand, the majority of treatment options for AUB caused by the development of morphologic pathologies such as endometrial polyps, adenomyosis and leiomyoma often include medical, hormonal or surgical management.^4,34

The purpose of the present study is to describe the prevalence of IR and other predictors of cardiovascular disease amongst women overall with AUB in a suburban population in Southern California. Understanding this prevalence may point to relationships between metabolic and cardiovascular health and manifestations of gynaecologic disease that can lead to sustainable strategies for treatment and prevention.

A total of 77 participants completed the measured variables and were included in the analysis. As shown in Table 1, the majority of participants were either Hispanic or white women, a mean age 37, with an average of two prior births. Interestingly, 90.0% of subjects were either current or former smokers. Approximately two-thirds of participants had a WHR > 0.80 with the average WHR being 0.86. Almost half of participants were obese, with the average BMI being 31.82 kg/m². Two-thirds of participants had a FI > 10 μiU/mL, with 38.0% having a FI > 15 μiU/mL, but only 22.0% of participants had an A1c ≥ 5.7%. The majority of participants also had dyslipidaemia, with 68.0% having high total cholesterol levels and an additional 38.0% of participants having high LDL levels.

Table 2 shows the difference in characteristics between the two groups when IR was defined as FI > 10 μiU/mL. Ethnicity was significantly related to IR, with all African-Americans and Middle Easterners being classified as IR, whilst none of the Asian participants were classified as IR (p = 0.027). The IR group was significantly more likely to be obese (60.8% vs 15.4%; p < 0.001) and have low HDL levels (41.2% vs 19.2%; p = 0.03) than the non-IR group. There were no statistical differences between the two groups with regard to the other variables analysed, but it is worth noting that no participants in the non-IR group had a history of DM2 or elevated triglycerides levels, and more than 62.0% of participants in both groups had high total cholesterol levels.

Table 3 depicts the difference in characteristics between the two groups when IR was defined as FI > 15 μiU/mL. The IR group was significantly more likely to be obese (75.9% vs 27.1%; p < 0.001), have low HDL levels (48.3% vs 25%; p = 0.016) and have high triglycerides levels (13.8% vs 2.1%; p = 0.032) compared with the non-IR group. Although there were no statistical differences between the two groups regarding the other variables analysed, again, no Asian participants were classified as IR and more than 62.0% of participants in both groups had elevated total cholesterol levels.

By univariate regression analysis, factors significantly associated with IR in women with AUB were BMI, WHR and HDL levels. These associations persisted regardless of the FI level used to define IR (see Table 4). Ethnicity, Cr, triglycerides levels, total cholesterol levels, LDL, history of DM2, history of hypertension, history of migraines and smoking were not significantly associated with IR in women with AUB in either of the univariate regression analyses. When multivariate regression analysis was performed using FI > 10 μiU/mL to define IR, BMI (standardised coefficient β: 1.07, p = 0.015) and WHR (standardised coefficient β: 1.94, p = 0.045) were independent predictors of IR in women with AUB. Regression analysis was not performed using FI > 15 μiU/mL.

This study shows that the prevalence of IR and other predictors of cardiovascular disease are higher amongst women from this diverse suburban community with overall AUB (66%) than amongst the general female US population.⁹ The BMI and WHR were shown to be independent predictors of IR in women with AUB. The odds of a woman with AUB having IR were seven times higher for every one-unit increase in BMI over 30 kg/m² when controlling for other variables. By the same convention, the odds of a woman with AUB having IR were nearly four times higher for every 0.1 unit increase in WHR after 0.8. Low HDL levels, elevated triglycerides levels and ethnicity were also significantly associated with IR in women with AUB, depending on how IR was defined.

Similar to prior studies on AUB, the present study’s participants had a high prevalence of risk factors for cardiovascular disease, including smoking (90%), obesity (45%), an elevated WHR (62%) and dyslipidaemia.^{16,24,26,42,43,44,45,46} The majority of participants did not have an elevated A1c or a history of hypertension or DM2, but this lack of association has been observed in prior studies on overall AUB as well.^28,29,43,44 This may indicate that whilst subjects may not have current diabetes or hypertension, elevated fasting insulin levels may be a precursor to abnormal uterine bleeding. Prior studies have found a relationship between AUB and elevated triglycerides levels, but this relationship was only significant in the present study when participants were IR using the higher FI cut-off of > 15 μiU/mL.^37,44 Interestingly, the significant inverse relationship between IR and HDL levels in women with AUB in the present study was not seen in prior studies on overall AUB.^43,44 Subjects with normal HDL levels, which are considered cardioprotective, were almost 70% less likely to be IR.⁴⁰

In addition, although a significant relationship between BMI and WHR and AUB has been reported in multiple prior studies, the relationship between BMI and WHR, IR and AUB has previously only been shown to be significant in prior studies on AUB because of PCOS.^{24,42,43,44,45,46,47}

There were strengths and limitations of this study. One strength was the inclusion of the ethnically diverse patient population, which can improve generalisability of the findings. Another strength was the broad definition of AUB used intentionally to evaluate the overall prevalence of IR and metabolic risk factors amongst women with this problem. However, this broad term also made it difficult to compare the present study’s results with the results of prior studies, which often only evaluated one specific cause of AUB. However, it is worth noting that although the developmental causes of AUB (polyps, leiomyoma, PCOS, etc.) may be different, the inflammatory factors that lead to each manifestation may be related. One potential weakness of the study was the large population of smokers or former smokers in the study, which on its own carries the risk of IR.

Furthermore, as this was a cross-sectional study, conclusions can only be made between IR and other cardiovascular disease risk factors amongst women with AUB. Another limitation of this study was the high attrition rate, with 41 (35%) of the participants excluded from the analysis because of incomplete laboratory values. Although still included in the analysis, an additional four (3%) of the participants did not complete the required lipid panel and an additional 16 (13.6%) of the participants did not undergo evaluation of their WHR. This selection bias likely affects how generalisable the results from the present study are to women with AUB. Finally, attempts to limit investigator bias were made by using retrospective self-reports to diagnose AUB, but prior studies have shown these may not be accurate.⁴⁸