Introduction

Breast cancer is the leading type of cancer impacting women worldwide [1,2]. Delay in diagnosis hinders women’s possibilities of less aggressive cancer treatment and survival. Women living with mental health conditions have been found to have worse health outcomes, including less adherence to screening guidelines, involvement in risky practices, and higher mortality [3]. A delay in cancer diagnosis may impact morbidity and mortality, as it represents the possibility of identifying the condition at a more advanced stage, requiring a more aggressive treatment plan [4].

Timely screening is one of the most effective measures for reducing both morbidity and mortality associated with breast cancer [5–7]. Mammography is the screening tool that has been associated the most with breast cancer survival [8,9]. While there is a clear understanding of the benefits of using mammograms for breast cancer screening, there are factors that may constitute barriers to access. These factors include age [10–12], having no health insurance, or having Medicaid coverage [13,14], lack of social support [10,15], limited geographic access to screening services facilities [16–18], as well as clinical factors to consider, such as biomarkers and comorbidities [19]. Concerning health insurance, despite Puerto Rico’s economic crisis and healthcare challenges, only 6% of the population lacks health coverage [20]. Moreover, medically indigent patients at high risk of chronic conditions, such as cancer, have access to health coverage through Medicaid [21]. However, it has been shown that cancer patients, including women with breast cancer, who are Medicaid enrollees, experience worse health outcomes [13,14,22]. These factors constitute challenges in breast cancer detection and outcomes.

Comorbidities, including mental disorders, are barriers that impact access to screening, but there is no certainty of the reach of their influence [23]. In some cases, comorbidities have been shown to promote a surveillance effect, helping to identify and treat illnesses in a timely manner [23,24], given that women are in the healthcare system, which increases the chances of accessing screening according to guidelines. On the other hand, comorbidities may create a distracting effect, deterring correct identification of signs and symptoms and leading to what has been identified as missed opportunities in primary care [25] and to late diagnoses [23].

Mental health conditions, such as depression and anxiety, can develop at any time during the cancer care continuum [1]. These may act as barriers to timely care and cancer screening [26], adversely affecting the chances of early diagnosis and decrease mortality of women with breast cancer [26–28]. Even if the incidence of cancer between women diagnosed with mental health conditions and the general population is similar, they experience increased mortality and poor quality of life [29]. It is estimated that during their lifetime, 1 in 3 women will experience depressive disorders [30]. Anxiety disorders are characterized by a sentiment of nervousness, irritability or hostility, among others, about everyday life matters [30]. It is estimated that between 2 and 11% of women suffer anxiety at any time during the past year, worldwide.

In Puerto Rico, according to the Department of Health, from July 2022 to June 2023, a total of 17,011 patients received services from mental health programs or substance abuse programs; of those, 12,750 were men and 4,065 women [31]. Canino et al (2019) estimated that the prevalence of mood disorders in Puerto Rico was 9.9% and 12.5% for anxiety disorders [32]. In 2021, the estimated prevalence of depression in Puerto Rico for women was 14.7% [33]. However, there is limited statistical information about anxiety or other disorders and other comorbidities to mental health besides substance abuse. Moreover, there is a significant gap in knowledge regarding the impact of mental health disorders in Puerto Rico. Particularly, on how these conditions interact with other health outcomes, such as cancer screening and overall healthcare access [34]. Studies about breast cancer screening in Hispanic women in Puerto Rico are more limited [35]. This study aims to explore the prevalence of anxiety disorders or affective disorders among women diagnosed with breast cancer. In addition, we evaluated whether there is a relationship between having a diagnosis of anxiety disorders or affective disorders and the use of breast cancer screening tests, mammography to be specific, in women aged 50 and over. Considering that adherence to screening is affected by other factors, the study also assessed factors that influenced access to mammography in women with a diagnosis of anxiety or an affective disorder and their impact on survival.

Methodology

Cohort definition

Cohort selection is described in Fig 1. We included women aged 50–91 years diagnosed with malignant breast cancer, confirmed histologically, residents in Puerto Rico between 2012 and 2016. Exclusion criteria included women with multiple cancer diagnoses, missing information regarding comorbidities, marital status, or stage at diagnosis. Additionally, cases that did not have continuous enrollment for 24 months before the diagnosis were excluded. Concerning mental health disorders, cases diagnosed using codes F00-F29 and F60-F99 from ICD-10 were excluded from the sample.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Population Study Selection Flowchart.

https://doi.org/10.1371/journal.pone.0357375.g001

Results

Clinical and sociodemographic characteristics.

The study population consisted of 4,708 individuals who met the inclusion criteria and were diagnosed with cancer in Puerto Rico between 2012 and 2016. Notably, 4,332 (92.0%) did not have any claims for mental health services in the twelve (12) months prior to their breast cancer diagnosis. Of the 376 women (8.0% of the total) who had a mental health condition diagnosed as identified by claims, 309 were related to a diagnosis of depression, 76 had a diagnosis of anxiety, and 16 of them presented a diagnosis of bipolar disorder, showing some overlap among conditions. Demographic characteristics are shown in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of the study population.

https://doi.org/10.1371/journal.pone.0357375.t001

Of the study population, 4,237 (90.0%) had undergone breast cancer screening, defined by having at least one mammogram in the 24 months before their breast cancer diagnosis. On the contrary, 471 (10.0%) women had not had a mammogram. In terms of the stage at the time of cancer diagnosis, 35.1% (n = 1,650) of the women had a late-stage diagnosis.

Factors associated with screening.

It was observed that having an affective or anxiety disorder was associated with undergoing a mammogram (p < 0.05) (Table 2). In the adjusted model, women with diagnoses of anxiety and affective disorders were 57.0% more likely to have a mammogram than those without a mental health diagnosis (AOR = 1.57, 95% IC: 1.01, 2.43; p = 0.046). The likelihood of undergoing mammography decreased with age and showed statistical significance for the age group of 70 years or older (AOR = 0.47, 95% CI: 0.35, 0.63; p < 0.001). Enrollment in private or Medicare health insurance increased the likelihood of getting a mammogram, compared to being enrolled in Medicaid. Dual coverage of Medicare and Medicaid, enrollees as well as Medicare, or private insurance enrollees were between 52.0% and 65.0% more likely to have a mammogram as established by the guidelines, compared to Medicaid enrollees. The comorbidity index also showed statistical significance with having undergone mammography (AOR = 1.99, 95% CI: 1.39, 2.66, p < 0.001).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Factors influencing mammography use in the 24 months prior to breast cancer diagnosis in a sample of women aged 50 years or older in Puerto Rico during the period 2012-2016.

https://doi.org/10.1371/journal.pone.0357375.t002

Factors associated with late stage.

Table 3 shows the results of the analyses describing the relationship between factors of access to services and late-stage breast cancer in women over 50 years of age between 2012 and 2016. The principal study variable, having a diagnosis of affective and anxiety disorder, did not show a statistically significant association with an advanced stage at the time of diagnosis (p < 0.05). Married women were 15.0% (AOR = 0.86, 95% CI: 0.76, 0.97; p = 0.015) less likely than unmarried women to be diagnosed at a late stage. As well as both Medicare, dual Medicare/Medicaid, enrollees were more likely to be diagnosed at an early stage (p < 0.05). Having a mammogram in the 24 months before the diagnosis of breast cancer decreases the likelihood of having a late diagnosis by nearly 50% (AOR = 0.49, 95% CI: 0.04–0.60; p < 0.001).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Factors influencing late-stage breast cancer diagnosis in a sample of women aged 50 years or older in Puerto Rico during the 2012-2016.

https://doi.org/10.1371/journal.pone.0357375.t003

Factors associated with risk of death.

The Cox proportional hazards regressions are presented in Table 4, and were calculated controlling for age, marital status, region of residence, comorbidity index, health insurance, biomarkers, stage, and mammography screening. The proportional hazards assumption was examined [37], including all variables: age, marital status, region of residence, comorbidity index, health insurance, biomarkers, mental health, and mammography screening. The global Schoenfeld residual test was applied, indicating that the proportional hazards assumption was not met (p < 0.001). Then, we performed a Schoenfeld test for each variable and identified biomarker status as the covariate that violated the proportional hazards assumption. Therefore, the Cox model was stratified by biomarker status. The model was stratified by the biomarker variable, controlling for the other variables. Following the team stratified by biomarkers, the proportional hazards assumption was met (p = 0.338).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Cox Regression Model, stratified by the variable biomarkers.

https://doi.org/10.1371/journal.pone.0357375.t004

After adjusting for potential confounders, having a diagnosis of anxiety or affective disorder did not increase the risk of death within 5 years (AHR = 1.01, 95% CI: 0.76, 1.35; p = 0.094). Married women had a statistically significant lower risk of death within 5 years (AHR = 0.76, 95% CI: 0.65, 0.88; p < 0.001). Lower risk of death was observed for users of private insurance and Medicare, and this finding was statistically significant (p < 0.05). On the other hand, comorbidity indexes, triple-negative biomarkers, and late stage are significantly associated with a higher risk ratio or likelihood of dying within the established period (p < 0.05).

Discussion

The relationship between breast cancer and mental health conditions has been documented globally, including in studies conducted in the continental United States [2,6,23,26,29]. However, there is a significant gap in research addressing this intersection within women living in Puerto Rico. To our knowledge, this is the first study in Puerto Rico to use secondary data sources to examine breast cancer screening and diagnostic patterns among women with anxiety and affective disorders. By leveraging linked cancer registry and health insurance claims data, this study analysis offers new insights into how mental health conditions may impact or be impacted by access to healthcare services, as exemplified in screening behaviors and diagnostic outcomes in this understudied population.

One of the study questions was whether having a mental health diagnosis, specifically anxiety or affective disorders, impacts breast cancer screening. Results showed a prevalence of anxiety and affective disorders of 8%, comparable to that of the general population of Puerto Rico, according to Canino et al (2019) but lower than the estimates of the Behavioral Risk Factors Surveillance System [32,33]. These differences raise questions as to whether it is the result of lack of access or lack of effective identification and diagnosis of mental health conditions. Further research is needed to answer these questions.

A diagnosis of anxiety or affective disorder emerged in this study as being associated with getting mammography as recommended in the clinical guidelines [25]. One explanation may be that to get mental health diagnoses, women must access the health care system, which could mean that these medical encounters turned into opportunities for screening for breast cancer [25]. These findings align with the idea proposed by Cunningham et al. (2015) that medical encounters provide an opportunity to address risk factors and that having comorbidities (including mental health conditions) could require consistent healthcare visits, which in turn may lead to screening opportunities [3,23,38]. Others have found no association among more common disorders, such as depression or anxiety, and mammography compliance [5]. However, associations with adherence to mammography, stage at diagnosis, and cancer treatment have been identified to be related to severe mental health conditions, for example, schizoaffective disorder [5,27], which implies the need for further research about the experience of Puerto Rico. This study excluded serious mental health illnesses such as schizophrenia, alcohol dependence, or other substance dependences, among others, which may have an impact on compliance with mammography use, as encountered by other researchers [3,27,39]. Since we identify anxiety and affective disorders based on insurance claims, we only capture women with diagnosed conditions who accessed care. This likely underestimates the true prevalence of mental illness in the study population. As we did not include women with these conditions who did not seek care, we could have underestimated the prevalence of these conditions. We may also have introduced surveillance bias, as women with greater healthcare utilization may have been more likely both to receive a mental health diagnosis and to undergo breast cancer screening.

There was also interest in responding to the question of whether there was a relationship between having comorbid anxiety or affective disorders upon getting a late-stage breast cancer diagnosis. Findings in this study showed no significant relationship with stage at diagnosis. Similarly to studies in London, UK, which did not show an association between the stage at time of diagnosis and anxiety or depression [40]. Results showed no statistically significant relationship with the risk of survival, in response to the question, “Does having anxiety or affective disorders impact 5 years of survival?” These findings are important, and they can be interpreted as evidence that breast cancer screening awareness measures are reaching different segments of the population. They may also indicate that healthcare providers are effectively using each patient’s encounter to promote screening services. Mental health conditions can be developed or diagnosed at any time during the process [41], still it is important to consider a connection of treatment impacting mental health [42]. Notwithstanding, it is vital to continue researching, considering areas that have not been explored, such as the impact of having a breast cancer diagnosis and developing a mental health illness [2,43] as well as the impact of severe mental health conditions on breast cancer screening, treatment, and survival in Puerto Rican women.

Additionally, our findings demonstrated that factors such as age, marital status, health insurance, comorbidities, and biomarkers may also influence screening services such as mammography, thus impacting timely cancer diagnosis. It is essential to consider other age-related variables that may affect screening behavior and quality of life, such as mobility, cognitive decline, or healthcare access. One of the risks of developing cancer, understandably, is age, however, concerning Puerto Rico it is an essential factor to consider. The population of Puerto Rico is rapidly aging [44], representing a challenge for public health as well as for the health system, with a shortage of physicians and the economic crisis [45,46].

Marital status was also associated with screening behaviors and stage at diagnosis; being married was significantly associated with a ~ 14–15% lower likelihood of being diagnosed at a late stage. This might be due to better social support, healthier lifestyles, and increased healthcare access [18,39]. Married women have been shown to have a higher likelihood of diagnosis, which may be associated with having an additional support structure, having, in turn, a better lifestyle, diet, exercise, and access to health care. Conversely, other studies related being married to delayed screening, due to fear of the impact of a positive diagnosis on their marriage [47,48]. This suggests cultural factors that should be explored in future studies, as well as the need to consider sociological analyses of the social roles assigned to women applied to the experience of breast cancer.

Medicaid enrollees had less access to mammography, according to our study findings. This is consistent with studies in the mainland United States, where Medicaid enrollees and uninsured women face greater challenges in timely breast cancer screening and treatment [13,14]. One question emerges from these findings: whether other elements have an additional impact, such as distance to services or referral from primary care physicians to screening. Furthermore, Medicaid beneficiaries with comorbid mental health conditions are at even higher risk of inadequate screening and higher mortality rates [14].

Clinical factors such as biomarkers and comorbidities emerged, as expected, as a significant factor for assessing risk, defining prognosis, and impacting survival [49]. This calls for further investigations concerning the experiences of women in Puerto Rico. But comorbidities, particularly considering as previously stated, the challenges faced by Puerto Rico’s health system, may impact access to much-needed services.

Conclusion

Our study found that anxiety and affective disorders were associated with an increased likelihood of breast cancer screening among Puerto Rican women aged 50 and older. This finding may suggest that these mental health conditions may act as a secondary effect of increased medical encounters in this context. This is a noteworthy finding given the limited research examining the intersection of mental health and cancer screening behaviors in this population as a starting point on the discussion. As the burden of both breast cancer and mental health conditions grows, it is essential to consider the broader implications for healthcare delivery. Future research should explore the impact of mental health status following a breast cancer diagnosis, as well as the quality of care and life experienced by cancer patients living with co-occurring mental health disorders. Diverse methodological approaches may help ensure the inclusion of women who lack health insurance or access to primary care in future studies to improve cancer care. These insights are critical to informing culturally relevant, integrated approaches to cancer prevention and survivorship care in Puerto Rico and similar settings.

Acknowledgments

We are especially grateful to Dr. Estela S. Estape and Dr. Rachel M. Rodríguez, from the Alliance for Clinical and Translational Research, who provided invaluable feedback.