Characteristics and Clinical Outcomes of Women with Polycystic Ovary Syndrome After Bariatric Surgery

Abstract

Background

Polycystic ovary syndrome (PCOS) commonly co-occurs with obesity, medical comorbidities, and psychiatric symptoms. Bariatric surgery is an effective treatment for co-occurring obesity and PCOS. While the incidence of PCOS declines substantially after bariatric surgery, the condition is still present for a subset of women. Examining characteristics and clinical outcomes of those with and without PCOS post-surgery may underscore potential risk factors or intervention targets.

Methods

Individuals up to four years after bariatric surgery were invited to participate in this cross-sectional survey study, which included validated measures of depression, anxiety, eating disorder pathology, and quality of life. Post-surgical weight outcomes, medical comorbidities, and mental health treatment engagement were also assessed. Regression analyses were performed to examine differences in outcomes between those with and without a PCOS diagnosis after bariatric surgery.

Results

Of the 657 female (sex assigned at birth) participants who underwent bariatric surgery, 7% (n = 46) reported having a current diagnosis of PCOS. All females identified as women. Women with PCOS were significantly younger (p < 0.001) and were more likely to endorse migraines (p < 0.007) and loss of control (LOC) eating episodes (< 0.001) since undergoing surgery. Additionally, 47.8% of women with PCOS endorsed clinically significant anxiety, compared to 25.7% of women without PCOS (p = 0.03). There were no differences in other demographic, psychiatric, or medical characteristics.

Conclusion

Despite the low prevalence of PCOS diagnoses in the four years after bariatric surgery, this subpopulation may be particularly susceptible to migraines, disinhibited eating behavior, and anxiety, although weight and cardiometabolic outcomes were comparable to those without a diagnosis of PCOS post-surgically.

Introduction

Polycystic ovary syndrome (PCOS) is the most prevalent endocrinologic condition in females (sex assigned at birth) and is one of the most common causes of female-factor infertility [1]. PCOS is characterized by menstrual irregularities (e.g., anovulation), hormone imbalances (e.g., hyperandrogenism, insulin resistance), and polycystic ovaries [1, 2]. Given complex pathophysiological consequences, PCOS increases the risk for several psychiatric and medical comorbidities [1–6]. Specifically, PCOS frequently co-occurs with obesity, diabetes, sleep apnea, cardiovascular disease, depression, anxiety, and eating disorder pathology—particularly binge eating [1–3, 7–9]. Medical and psychological features of PCOS can interact and exacerbate symptoms [1]. For example, physical manifestations of PCOS—such as hirsutism, menstrual problems (i.e., bloating, headaches/migraines), fertility concerns, and excess weight, can contribute to increased depression and anxiety [10–13]. Additionally, greater disordered eating behavior may result from maladaptive attempts to control one’s weight, cope with psychological distress, or in response to pathophysiological processes (i.e., insulin resistance, increased cortisol) driving intense hunger or food cravings [4, 8, 9]. In turn, the cycle of disordered eating may further worsen mood and physical functioning [9]. Ultimately, PCOS is a complex disorder that warrants a comprehensive and biopsychosocial approach to treatment [1, 3, 5, 9].

While not all individuals with PCOS have overweight or obesity, those with both (nearly 50–80%) are at greater risk for psychiatric and medical comorbidities [1–5, 7, 13]. Although there is preliminary support for the benefits of weight management interventions for those with PCOS [1], those with PCOS and severe obesity experience significant barriers to sustained and substantial weight loss through behavioral methods alone, given both psychiatric and pathophysiological consequences [2–6]. Thus, bariatric surgery is an efficacious potential treatment option for co-occurring obesity and PCOS [14–16]. While approximately 46% of females pursuing bariatric surgery have PCOS, the incidence declines to less than 10% post-operatively [14–16]. Accordingly, many PCOS symptoms (i.e., menstrual cycle irregularity, hormone imbalances, hirsutism, and markers of infertility) improve or resolve along with substantial postsurgical weight loss [16].

Despite bariatric surgery leading to improvements in PCOS, little is known about the small subset of those who have PCOS after undergoing bariatric surgery (~ 7–8%) [14–16]. Given medical and psychiatric comorbidities associated with PCOS in other samples, the presence of post-surgical PCOS may undermine the long-term benefits of bariatric surgery [1–6]. Identifying characteristics and clinical outcomes may highlight potential risk factors and treatment targets for this unique, potentially high-risk subpopulation after bariatric surgery. Therefore, the purpose of this study was to examine whether psychiatric and medical outcomes associated with PCOS in other samples differ between females with and without PCOS up to four years post-surgically [2, 3, 7–9, 14]. We anticipated that those with PCOS may have worse psychiatric and weight-related outcomes than those without PCOS [1–5]. However, this was largely exploratory given limited research on PCOS specifically after bariatric surgery.

Methods

Participants

The current study was part of a larger, overarching cross-sectional observational survey study conducted in 2021–2022 to examine a variety of psychological and physical health outcomes among those who underwent bariatric surgery from 2018–2021 at a large Midwestern healthcare system. The study was approved by the system’s Institutional Review Board prior to participant recruitment. A total of 765 individuals provided their informed consent and completed the entire survey for the broader research study. Given that the focus of this specific sub-study was PCOS-related outcomes, only females (sex assigned at birth) were included in analyses (n = 657), excluding 108 males from the original sample.

Measures

Demographic and Medical/Treatment Characteristics

Race, sex assigned at birth, and surgery type were extracted from participants’ electronic health records. Pre-surgical body mass index (BMI) and self-reported gender were obtained from participants’ pre-surgical psychological evaluation, which was obtained from their medical records. Percent of total weight loss (%TWL) and BMI were calculated via participants’ self-reported weight on the survey. Participants reported their current medical conditions, including whether they had PCOS (“Which of the following medical conditions are you currently diagnosed with, and/or receiving treatment for?”). Medical conditions most relevant to PCOS were included in the study—specifically diabetes, cardiovascular disease, sleep apnea, and migraines [2, 3, 12, 13, 17]. Lastly, mental health treatment seeking was assessed (“Since undergoing bariatric surgery, have you seen a mental health provider such as a therapist, psychiatrist, or another type of mental health provider?”).

Psychosocial Outcomes

Depressive and anxiety symptoms were examined through the Hospital Anxiety and Depression Scale (HADS), which includes two 7-item subscales for depression and anxiety symptoms [18]. Scores of 8 or greater on each subscale reflect clinically significant symptoms and this was therefore measured dichotomously based on this clinical cut-off. [18]. Quality of life was examined via the one-item Kemp Quality of Life Scale (“Taking everything in your life into account, please rate your overall Quality of Life on the following 7-point scale: 1—life is very distressing; to 7—life is great”) and was examined continuously [19, 20].

Eating Disorder Pathology

Overvaluation of shape and weight (OSW)—the extent to which one judges themselves based on shape and weight and an important feature of eating disorder pathology, was examined via the two-item OSW subscale of the Eating Disorders Examination—Questionnaire (EDE-Q) [21–23]. An average score was obtained and higher scores on the 6-point scale reflect greater overvaluation. We examined the score as a continuous variable, given strong psychometric properties in individuals pre- and post-bariatric surgery [22]. Loss of control (LOC) and binge eating episodes were examined via their respective individual items on the EDE-Q, which assesses whether one engaged in LOC eating and/or binge eating episodes over the last 28 days [21]. Participants responded with yes or no, and the presence of these behaviors was examined dichotomously [21–23].

Analyses

Analyses were conducted with SPSS version 29. For group differences in demographic characteristics, chi-square and the Mann–Whitney U-test (non-parametric alternative given differences in sample size) were conducted. Significant demographic differences between groups were controlled for in subsequent analyses. Differences between women with and without PCOS were examined with binary logistic regression analyses for dichotomous outcomes, and linear regression analyses for continuous outcomes. The Bonferroni correction was used to account for multiple comparisons, resulting in an adjusted significance threshold of p < 0.004 for regression analyses.

Results

Demographic Characteristics

Demographic characteristics are included in Table 1. All females included in the study identified as cisgender women. Forty-six women (7%) reported a current diagnosis of PCOS. There were no group differences in surgery type, time since surgery, or pre-surgical BMI. Women with PCOS were significantly younger (M = 33.6 years old) than women without (M = 44.6 years old) (p < 0.001). Both groups consisted of mostly White (58.6 and 63%) and Black (39.3 and 32.6%) women. The PCOS group had 4.3% American Indian/Alaska Native (AI/AN) (n = 2) compared to 0.2% (n = 1) in the non-PCOS group.

Table 1.

Demographic characteristics by PCOS status after bariatric surgery

	No PCOS N = 611		PCOS N = 46		Mann-Whitney-U
	M	SD	M	SD	z	(p)
Age	44.6	10.8	33.6	8.3	− 6.6	< 0.001
Pre-surgical BMI	46.8	7.4	47.2	6.9	− 0.55	0.59
Time since surgery (days)	728.4	367.3	701.6	386.1	− 0.79	0.43
	%	n	%	n	Chi-square
Race
White	58.6	358	63.0	29	0.35
Black	39.3	240	32.6	15	0.80
American Indian or Alaska Native	0.2	1	4.3	2	16.5
Other/multiracial	2	12	0	0	0.92
Surgery type
Sleeve gastrectomy	69.4	424	65.2	30	0.35
Roux-en-Y gastric bypass	30.6	187	34.8	16

Medical Comorbidities

Rates of post-surgical medical comorbidities are included in Table 2. More women with PCOS endorsed having migraines (32 vs. 11%) (p < 0.007), although this fell short of the Bonferroni corrected significance level (p < 0.004). There were no significant group differences in diabetes, cardiovascular disease, or sleep apnea.

Table 2.

Clinical outcomes after bariatric surgery by PCOS status (controlling for age)

	No PCOS		PCOS
	M	SD	M	SD	B	SE	OR (95% CI)	p
Quality of life	6.0	1.1	5.9	1.1	− 0.14	0.17	− 0.03 (− 0.48, 0.20)	0.42
Overvaluation of shape/weight	3.6	1.9	4.0	1.8	0.30	0.30	0.04 (− 0.29, 0.90)	0.32
% Total weight loss	28.2	9.9	26.9	10.0	− 2.0	1.6	− 0.05 (− 5.1, 1.1)	0.20
Body mass index	33.5	6.6	34.5	6.6	0.51	1.0	0.02 (− 1.4, 2.6)	0.63
	%	n	%	n	B	SE	OR (95% CI)	p
Diabetes with insulin	3%		0%	0	− 17.4	59.0	0.00 (0.00)	0.99
Cardiovascular disease	2.6%	16	0%	0	− 16.2	55.1	0.00 (0.00)	0.99
Sleep apnea	14.6%	89	10.9%	5	0.22	0.51	1.3 (.46, 3.4)	0.66
Migraines	11.8%	72	30.4%	14	0.97	0.36	2.7 (1.3, 5.4)	0.007
Mental health treatment	33%	199	45.7%	21	0.48	0.33	1.6 (.85, 3.1)	0.14
Clinical anxiety symptoms	25.7%	157	47.8%	22	0.70	0.33	2.0 (1.1, 3.9)	0.03
Clinical depression symptoms	7.5%	46	6.5%	3	0.12	0.64	1.1 (0.32, 4.0)	0.85
Binge eating	10.3%	63	19.6%	9	0.79	0.43	2.2 (0.95, 5.1)	0.07
Loss of control eating	24.7%	151	45.7%	21	1.1	0.33	3.1 (1.6, 5.9)	< 0.001

Weight Loss and Psychosocial Outcomes

Weight loss and psychosocial outcomes are also in Table 2. A higher percentage of women with PCOS endorsed LOC eating episodes (46 vs. 25%) (p < 0.001). While more women with PCOS reported clinically significant anxiety (48 vs. 26%) (p = 0.03), this was not statistically significant with the Bonferroni correction (P < 0.004). Although a greater proportion of women with PCOS engaged in mental health treatment since surgery (46 vs. 33%), this was no longer significant when controlling for age. There were no significant differences in quality of life, OSW, depression, %TWL, or post-surgical BMI.

Discussion

Overall, the percentage of women with PCOS up to four years after bariatric surgery was low (7%), as demonstrated in other samples [14–16]. Women with PCOS were younger than women without PCOS, consistent with other studies [8, 24]. This age difference is also reflective of diagnostic aspects of PCOS, as menstrual irregularities are a primary criterion and thus would only include those who have not undergone menopause [1]. Additionally, there was a higher proportion of AI/AN women with PCOS compared to those without (4.3 vs. 0.2%; 66.7% of AI/AN women). Although this finding should be interpreted cautiously given the small number of AI/AN women (N = 3; 0.05%), its acknowledgment may be a step toward greater inclusion for a historically underrepresented racial group [25, 26]. Despite being a small percentage of the US population (~ 1.3%), researchers have identified a gap in understanding the comparative experience of AI/AN individuals in both bariatric surgery and PCOS research [27–31]. Relative to the general population, AI/AN individuals have disproportionately high rates of PCOS (6.9 vs. 5.2%, second only to Hawaiian Pacific Islanders) [28], obesity (48 vs. 31%), and related comorbidities such as diabetes (23.5 vs. 9.5%) [28], making bariatric surgery a valuable potential treatment option [27]. Elevated chronic disease risk in this group results from downstream effects of social and structural determinants of health—including historical trauma, healthcare inequities, economic and environmental barriers (i.e., poverty, rurality), and associated risk factors [27, 30, 31]. Cumulative underrepresentation of this marginalized group despite PCOS and weight-related disparities, may inadvertently perpetuate healthcare inequities [27, 30, 31]. Our preliminary finding that two of three (66.7%) AI/AN women had post-surgical PCOS suggests the importance of additional research on PCOS and bariatric surgery outcomes among this racial group [27, 30, 31].

Our hypothesis that those with PCOS may have poorer weight outcomes or greater medical comorbidities, consistent with non-post–surgical samples, was unsupported [1–3]. Both groups of women on average achieved successful postsurgical weight loss, defined as greater than 20% TWL (28 and 27%) [32], and had similar post-surgical BMIs. Comparable weight outcomes between groups may explain why there were no significant differences in cardiometabolic comorbidities, and why these rates were relatively low overall, consistent with other samples following bariatric surgery [14, 15]. However, women with PCOS were more likely to experience migraines (32 vs. 11%) (p < 0.007). Menstrual headaches are a commonly reported concern in those with PCOS, and this group is also susceptible to migraines [12, 17]. This relationship may be explained by the interaction of female sex hormones (i.e., estrogen), neurotransmitters (i.e., serotonin), insulin, and PCOS-related alterations in these processes [17]. Migraines have also been related to anxiety, possibly via similar mechanisms (i.e., serotonergic dysfunction) [33], and nearly half of the women with PCOS endorsed clinically significant anxiety. Ultimately, in this post-surgical context, migraines may be the medical condition most proximal to PCOS beyond associations with cardiometabolic comorbidities (i.e., diabetes, cardiovascular disease, and sleep apnea), although more research in this specific population with larger samples is warranted.

Women with PCOS were significantly more likely to endorse LOC eating episodes (46 vs. 25%) (p < 0.001). Further, while not statistically significant, more women with PCOS endorsed binge eating episodes, with trends toward significance (20 vs. 10%). Thus, following bariatric surgery, women with PCOS may have an increased risk for disinhibited eating behaviors, consistent with other PCOS participant samples [8, 9]. Greater LOC eating in this population may be explained by a combination of affective (i.e., anxiety, emotional dysregulation) and pathophysiological mechanisms (i.e., intense cravings, increased reward sensitivity to food, etc.) [4–6, 9]. For example, high rates of anxiety in this group may catalyze greater disinhibited eating as a means to self-regulate or decrease distress, or alternatively, may be a side effect of prescribed anxiety medications [4, 8, 9, 40, 41]. However, additional research is needed regarding possible relationships between anxiety, disinhibited eating, and underlying mechanisms in this population.

Although the difference in clinical anxiety symptoms was no longer significant after correcting for multiple comparisons, this trend toward significance (P = 0.03) may still have clinical relevance as it was endorsed by nearly half of the women with PCOS [4, 6, 7]. Further, both younger age and having PCOS may mutually contribute to a greater need for post-surgical mental health treatment (46 vs. 33%). Although depression is also shown to be elevated in those with PCOS, there was comparable and relatively low depression severity in both groups. This may be in part because depression is shown to improve initially after surgery, whereas anxiety symptoms can be more variable [35]. Further, infertility concerns have been shown to contribute to significant anxiety, but not depression, in those with PCOS [36]. Given that this group is a younger reproductive age, increasing fertility may have been a motivating factor to pursue bariatric surgery [16]. It is possible that PCOS-related infertility concerns may cause or exacerbate anxiety in this participant sample, particularly if actively planning or attempting to conceive. While anxiety and functional impacts of fertility-related concerns may be important to evaluate in this group post-surgically, additional research is needed.

There were no significant differences between women with and without PCOS in OSW or quality of life. OSW may be important to monitor in all women post-bariatric surgery as this is a risk factor for greater disinhibited eating and functional impairment and was relatively high in both groups [22, 23]. Lastly, quality of life was high in both groups, suggesting that regardless of PCOS status, women who undergo bariatric surgery may experience good quality of life overall up to four years post-surgically.

Limitations and Recommendations for Future Research

Despite the contributions of this study to an understudied area, there are important limitations that should be considered and addressed in subsequent research. First, there was a relatively small number of women with a known PCOS diagnosis in this sample (7%), limiting our ability to draw strong conclusions. Further, we relied on participants’ self-reported diagnosis of PCOS, other medical comorbidities, and weight outcomes. Although self-reported weight has been concordant with objective weight among those who have undergone bariatric surgery [39], this method may be less reliable for medical diagnoses—particularly PCOS. PCOS is a highly heterogenous syndrome, and many with the condition may remain undiagnosed [1]. Thus, we may not have accurately identified all patients with PCOS. Additionally, lack of information on the timing of PCOS diagnosis is a limitation, as we did not assess whether these diagnoses were maintained from pre-surgery, versus whether these were new post-surgical diagnoses. Future research studies would benefit from more rigorous investigation with greater diagnostic specificity, either in collaboration with medical providers, through medical chart review, or by including more specific assessment questions. Clarifying the timing of diagnoses, specific diagnostic criteria met, and PCOS phenotype [1, 30], with larger, more descriptive, and diverse samples would be advantageous for more thoroughly understanding PCOS outcomes after bariatric surgery. Additionally, although our goal was to compare characteristics of women with a diagnosis of PCOS after bariatric surgery to those without, future studies on characteristics of those with and without pre- to post-surgical resolution of PCOS, in addition to examining resolution of other pre-surgical medical comorbidities, may elucidate specific causes for PCOS remission or maintenance. Further, preliminary findings supplemented with existing literature suggest a need for additional research on PCOS and bariatric surgery outcomes among AI/AN women specifically, as this group may be at increased risk yet is historically underrepresented in related research [27–31]. Longitudinal research is also needed on longer-term weight outcomes in those with post-surgical PCOS, given trends toward greater anxiety and disinhibited eating as these are identified risk factors for weight recurrence [34]. Lastly, qualitative research may provide greater insight on the lived experiences of this unique subpopulation to inform specific intervention efforts [e.g., 31].

Clinical Implications

These preliminary findings integrated with prior research may loosely inform clinical practice [1]. Given the biological and psychological consequences of PCOS, this group may benefit from a multidisciplinary, biopsychosocial approach to post-surgical management [1, 5, 10–13]. Screening and treatment for migraines, disinhibited eating patterns, and anxiety may be particularly advantageous [34]. Cognitive behavioral or self-compassion-based approaches may be valuable psychotherapeutic treatment options for simultaneously addressing disinhibited eating, anxiety, and other PCOS-related concerns (i.e., poor body image) [37, 38].

Conclusions

Those with PCOS up to four years after bariatric surgery may be a relatively small group with distinct post-surgical treatment considerations. While there were similar weight and cardiometabolic outcomes, those with PCOS were younger, had higher rates of migraines and LOC eating, and nearly half had clinically significant anxiety. Future research extending this work through several of the recommended avenues may enhance long-term outcomes for those with PCOS after bariatric surgery.

Key Points.

• The prevalence of PCOS diagnoses up to four years after bariatric surgery was low (7%).

• Those diagnosed with PCOS had comparable weight loss outcomes to those without PCOS.

• More women with PCOS endorsed loss of control eating (46%) than those without PCOS (25%).

• Women with PCOS had high rates of migraines (30%) and clinically significant anxiety (48%).

Data Availability

The data are not publicly available due to medical confidentiality and to not compromise the privacy of research participants.