Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): A secondary analysis of a randomized controlled trial.

Many women with Polycystic Ovary Syndrome (PCOS) report high depression rates. The relationship between PCOS and these high depression rates is unclear. Two-component lifestyle interventions have revealed short-term effects on depression scores in this group of women. In general, 3-component interventions including diet, exercise, and cognitive behavioral therapy (CBT) are more effective in the long-term to improve emotional well-being. This has not yet been studied in women with PCOS. This study examined the effect of 20 CBT lifestyle (LS) sessions combined with a healthy diet and physical therapy with or without 9 months additional feedback through Short Message Service (SMS) via mobile phone, compared to care as usual (CAU, involving advice to lose weight). In this secondary analysis, 155 women with PCOS and a BMI above 25 kg/m2 were eligible. Depression scores decreased significantly in the LS programme compared to CAU (P = 0.045). In both the LS programme without SMS (P = 0.036) and the LS programme with SMS (P = 0.011) depression scores decreased while no change was observed in CAU (P = 0.875). Self-esteem scores improved significantly in the LS programme compared to CAU (P = 0.027). No differences in body image scores were observed in LS participants compared to CAU (P = 0.087), although body image improved significantly in both the LS without SMS (P = 0.001) and with SMS (P = 0.008) study arms. We found no significant mediating role by androgens in the relationship between LS participants and emotional well-being. Only weight-loss mediated the relationship between LS and self-esteem. To conclude, a three-component lifestyle intervention programme with or without additional SMS resulted in significant improvements in depression and self-esteem compared to CAU, in women with PCOS, obesity, and a wish to achieve a pregnancy. Testosterone, androstenedione, DHEA, insulin, HOMA-IR, and cortisol did not mediate this effect. Weight loss mediated the effects on self-esteem but not on depression and body-image. This suggests that lifestyle treatment independent of weight loss can reduce depression and body-image, but both lifestyle treatment and weight loss can improve self-esteem. Thus, a three-component lifestyle intervention based on CBT could prove successful in improving mood in women with PCOS who are overweight or obese and attempting to become pregnant.

RCT Entities:   Population Interventions Outcomes

Introduction

Polycystic Ovary Syndrome (PCOS) is a common endocrine disorder that affects 8–15% of women in their reproductive years [1-3]. The diagnosis of PCOS requires at least two of the following three criteria: (i) oligo-ovulation or anovulation (irregular or no menstrual cycle), (ii) clinical hyperandrogenism (hirsutism) and/or biochemical signs of hyperandrogenism (elevated free androgen index or elevated testosterone levels), (iii) polycystic ovarian morphology (by transvaginal ultrasound), and the exclusion of other etiologies that might cause hyperandrogenism [4]. Most women with PCOS experience one or more of the following physical symptoms in varying degrees: hirsutism (excessive body hair growth), acne, infertility, obesity, insulin resistance and dyslipidemia [5, 6].

Women with PCOS experience more depressive and anxiety complaints, have lower self-esteem, and experience a more negative body image compared to women without PCOS [7-9]. In particular, depression scores are significantly higher [7, 10] and seem to be consistently elevated throughout the lifespan of women with PCOS compared to controls [11]. A recent meta-analysis of depression rates among women with PCOS resulted in a median prevalence of depression of almost 37% compared to 14% in controls [8]. Hence, the recent international guideline on PCOS states that depressive and anxiety symptoms should be screened, assessed and managed with the requirement for awareness of emotional wellbeing [12] Women with PCOS and BMI ≥ 30 kg/m2 have significantly higher depression rates compared to women with PCOS and a healthy BMI [10]. A 5% to 10% weight loss improves many PCOS features, including psychological factors [13, 14]. It is unclear how these psychological improvements are generated and whether these psychological improvements are sustained in the long-term. One of the first lifestyle (LS) interventions in women with PCOS was developed by Clark and colleagues. This involved 6 months of seminars covering weight-related topics and resulted in significantly lower depression scores, although no control group was used [15, 16]. Thompson and colleagues developed a 20-week diet and exercise intervention and found significant improvements with respect to depression during the first 10 weeks of the intervention. It is unclear why depression scores did not improve after 10 weeks as participants continued their weight loss and PCOS symptoms improved [17]. A more recent paper demonstrated that a 16-week LS intervention programme resulted in better quality of life [18]. This LS intervention included behavioral modification strategies, although these specific strategies were not described [19]. Others found improvements in depression, health-related quality of life and self-esteem during a high-protein and low-carbohydrate diet, but not in the amount of weight loss [20, 21].

In the general population there is a bidirectional association between obesity and the odds of depression [22]. In women with PCOS the results are inconclusive: some authors concluded that women with PCOS and a higher BMI are more depressed, while others suggest the opposite. Women with PCOS still have higher odds for depressive and anxiety symptoms when matched for BMI [8]. A recent review presented potential mechanisms other than obesity for the increased depression risk in women with PCOS. Insulin resistance, increased testosterone levels, higher hirsutism scores measured by the modified Ferriman-Gallwey questionnaire, infertility due to oligo-ovulation, increased corticotrophin-releasing hormone, increased cortisol, markers of inflammation, low vitamin D status [23], and elevated Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) levels [24], may contribute to the association between PCOS and depression. Despite the evidence that women with PCOS have increased odds for depression and anxiety, there is no evidence supporting a single etiology for this increased prevalence of depression and anxiety [23]. Thus it remains unclear whether depression is related to one of the above mechanisms and if depression rates could change through weight loss.

The first-line treatment for depression is cognitive behavioral therapy (CBT) and, depending on the setting, can be combined with antidepressant treatment [25]. In the general population, long-term results are mixed; some meta-analysis found CBT to be equally effective compared to other psychological treatments while other meta-analysis found favorable results for CBT [26]. Little research has been undertaken concerning CBT among women with PCOS. A pilot study demonstrated that 8 weeks of 30 minutes’ CBT combined with 30 minutes’ LS sessions resulted in a significant improvement in quality of life but no improvements in depression scores were observed [27]. A recent randomized control trial (RCT) showed that 8 CBT group sessions of 45 to 60 minutes was effective for psychological fatigue and quality of life [28]. In the new PCOS guideline there is no referral to a specific treatment for depression in women with PCOS, and the advice is to follow regional clinical guidelines [12].

In conclusion, previous studies covered study periods of 24 weeks at most, were not randomized controlled trials, had small sample sizes, and did not use a structured CBT protocol. Hence, we investigated whether a CBT programme for women with PCOS who were overweight and obese achieved weight loss in the long-term in a large sample. The aim of this secondary analysis was to compare the changes in depression scores in a three-component CBT LS intervention (with or without SMS), with these scores in the control group. In addition, the effectiveness of additional SMS on self-esteem and body image was examined. We hypothesized that there is an interaction of androgens (testosterone, androstenedione and dehydro-epiandrosterone (DHEA)), insulin, HOMA-IR, and cortisol, on well-being scores in women with PCOS. Hence, we tested whether the relationship between lifestyle treatment and well-being is mediated by changes in androgens (testosterone, androstenedione and DHEA), insulin, HOMA-IR, and cortisol.

Methods

Patients

Women were eligible if they were diagnosed with PCOS according to the Rotterdam 2003 consensus criteria, had a BMI above 25 kg/m2, were between 18 and 38 years old, and attempting to become pregnant. Women with inadequate command of the Dutch language, severe mental illness, obesity with another somatic cause, ovarian tumors that lead to an androgen excess, adrenal diseases, had other malformations of their internal genitalia, or those that were pregnant, were not eligible for the study. Participants did not receive any fertility treatment during the study period.

Study design

This study was approved by the Medical Research Ethics Committee of the Erasmus MC in Rotterdam; reference number MEC 2008–337. The study protocol can be found at http://dx.doi.org/10.17504/protocols.io.bfq9jmz6. The current study on emotional well-being represents a secondary analysis. At baseline, 183 participants were randomized at a 1:1:1 ratio using a computer-generated random numbers table into three arms: 1) 1-year CBT LS intervention provided by a multidisciplinary team, or 2) 1-year CBT LS intervention provided by a multidisciplinary team extended with a Short Message Service (SMS,) or 3) care as usual (CAU): encouragement to lose weight by publicly available services (i.e. diets, visiting a dietician, going to the gym, or participating in public programmes such as Weight Watchers®). The 1-year multidisciplinary LS intervention aimed at: 1) changing cognitions, 2) changing dietary habits, 3) encouraging and promoting physical activity, and 4) activating social support. It consisted of 20 group sessions of 2.5 hours over one year. During all sessions, CBT techniques were used to create awareness and to restructure dysfunctional thoughts about lifestyle (food and exercise), weight (loss) and self-esteem. More details about the intervention and an overview of the content of each session can be found in the study protocol [29]. Additional to the lifestyle programme, participants in the SMS group sent weekly self-monitored information regarding their diet, physical activity, and emotions by SMS to the psychologist. Subsequently, they received feedback on their messages to provide social support, encourage positive behavior, and empower behavioral strategies.

Outcome measures

At baseline, and at 3-, 6-, 9- and 12-months, participants attended the outpatient clinic for a standardized screening. This screening included a family and reproductive history, and a physical examination assessing anthropometric and ultra-sonographic features of the syndrome. The primary outcome of the RCT (weight) was also measured. Participants also completed questionnaires on well-being at these time points.

Well-being was measured using three instruments: depression with the Beck Depression Inventory-II (BDI-II), self-esteem with the Rosenberg Self Esteem Scale (RSES), and body image with the Fear of Negative Appearance Evaluation Scale (FNAES).

BDI-II is a validated and widely-used questionnaire in depression trials assessing the severity of depressive symptoms over the previous 2 weeks, according to the DSM-IV criteria. It is a 21-item self-report questionnaire with items rated on a 4-point scale (0–3) and summed to give a total score (range 0–63). A higher score on the BDI-II denotes more severe depression. In non-clinical populations, scores above 20 indicate depression [30]. More specifically: scores of 0–13 indicate minimal depression, 14–19 (mild depression), 20–28 (moderate depression), and 29–63 (severe depression) [31]. The National Institute for Health and Care Excellence (NICE) suggested a difference of ≥3 BDI-II points as a clinically significant effect for normal depression [32]. A recent study estimated a minimal clinically important difference (MCID) for the BDI-II of a 17.5% reduction from baseline [33].

Global self-esteem and self-acceptance was measured by the RSES [34]. This questionnaire consists of 10 questions (5 positive and 5 negative) and has been validated for the Dutch population [35]. Items are rated on a 4-point Likert scale and total scores range from 0 to 30, where a higher score indicates higher levels of self-esteem. There are no official cut-offs, although scores between 15 to 25 are considered as normal self-esteem and scores below 15 as low self-esteem in women with PCOS [36].

The brief version of the FNAES [37] is a short questionnaire consisting of 6 items that measure body image, eating attitude, and depression. The items are answered on a 5-point Likert scale, ranging from ‘not at all’ to ‘extremely’, where a higher score indicates more fear of negative evaluation by others (range 6–30). We used a translated version of the FNAES, which has been used before in PCOS [38].

All participants underwent 5 similar standardized measurements during the study period. During these measurements blood samples were collected between 8.00 and 11.00 a.m. after overnight fasting. Levels of serum testosterone, androstenedione, DHEA, and cortisol were measured with RIA (Siemens) until 2012. After 2012, liquid chromatography-tandem mass spectrometry (LC-MS/MS) was used. Homeostatic assessment of insulin resistance (HOMA-IR) was calculated from fasting insulin and glucose by the following equation: HOMA-IR = (fasting glucose (mg/dl) * fasting insulin (μIU/ml)) / 405 [39].

Analyses

The power calculation was based on the primary outcome of the lifestyle intervention: weight (kg). The method described by Aberson [25] was applied, with a power of 0.90, a 2-sided alpha of 0.025 (corrected for the interim analysis as described in the study protocol), and 5 repeated measures linearly decreasing. We observed an intercorrelation of around 0.90 between all measurements. With a ratio of 1:1:1, the required sample was 42 in each group. With an expected drop-out proportion of 40% [40], 60 participants in each group were needed for the study.

Descriptive statistics were used to characterize depression, self-esteem, and body image in this sample. Normality of the distributions was checked with Shapiro-Wilks tests. Multilevel regression models were applied for longitudinal analyses of depression (BDI-II), self-esteem (RSES), and body image (FNAES) scores. Mixed modeling can deal efficiently with missing data and unbalanced time-points [41]. This means that, additionally, patients without complete follow-ups could be included in the analyses, without imputation. This method also compensates for selective dropout, on the condition that dropout is related to variables included in the models. The analysis included 2 levels; the patients constituted the upper level and their repeated measures the lower level. The difference from ordinary linear regression is that this analysis takes into consideration that measurements belong to a given participant. The deviance statistic [42] using restricted maximum likelihood [43] was applied to determine the covariance structure, thus taking into account the situation when, e.g., the deviation at baseline was different from the deviations at follow-ups. The covariance structure was determined with deviance tests, using restricted maximum likelihood. To this end, the unstructured component, the variance component and the intercept- only covariance structures were compared amongst each other. In the case of a non-normal distribution a bootstrap procedure with 10,000 samples was performed to obtain reliable standard errors and p-values. Study group, linear and logarithmic time and interactions were included as independent variables. Cohen's d effect sizes were calculated by performing additional multilevel models on normalized outcome measures, using Blom transformations [44]. Blom transformations have the characteristic that they are standardized, thus the outcomes are Cohen's d effect sizes. Cohen valued d = 0.2 as a 'small' effect size, 0.5 as a 'medium' effect size and 0.8 as a 'large' effect size [45]. To test if androgens, insulin, HOMA-IR, and cortisol mediated the effect of LS intervention with or without SMS on emotional well-being, we used multilevel longitudinal mediation or indirect effect analyses. Paths α, β, τ and τ′ were estimated employing multilevel regression analyses. Firstly, we determined whether paths β were significant. When path β was not significant, mediation was improbable. We adjusted the Sobel-Goodman test for the indirect effect of the independent variable on the dependent variable as reported by MacKinnon and Dwyer [46], following the recommendations by Krull and MacKinnon [47] for multilevel mediation analyses. The significance of the mediated effect is given by [48]:

All analyses were performed utilizing IBM Corp (Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp).

Results

Between August 2nd 2010 and March 11th 2016, 561 eligible women were asked to participate and 209 provided written informed consent, of whom 26 were included in a pilot study. The total sample for this secondary analysis consisted of 140 women who completed the depression questionnaire, 155 who completed the body image questionnaire, and 141 who completed the self-esteem questionnaire, all at baseline (Fig 1 and Table 1). According to the Shapiro-Wilks test none of the baseline outcome variables were normally distributed. For all mixed models linear time was not significant, thus superfluous. The logarithm of time was included in all the models. The multilevel models are presented in S1 Table.

Fig 1

Consort flowchart.

BDI-II = depression, RSES = self-esteem, FNAES = body image.

Table 1

Baseline characteristics by trial group.

IQR = Interquartile range, BDI-II = depression, RSES = self-esteem, FNAES = body image. Spontaneous pregnancies observed during the study period.

	Control (CAU)	Lifestyle without SMS	Lifestyle with SMS
	Median [IQR]	Median [IQR]	Median [IQR]
BDI-II scores	11.0 [5.0–18.0]	13.5 [5.9–24.0]	12.0 [5.5–20.9]
RSES scores	23.0 [17.5–26.0]	20.0 [14.0–23.0]	20.0 [16.0–24.7]
FNAES scores	15.0 [8.0–21.5]	20.5 [13.3–23.8]	19.0 [12.0–23.3]
Age (year)	28.0 [26.0–32.0]	30.0 [27.0–33.0]	28.0 [26.0–32.0]
Attempting to conceive (months)	27.5 [15.0–59.0]	27.0 [16.0–63.5]	24.5 [11.8–36.3]
Weight (kg)	84.0 [79.0–97.3]	89.0 [80.0–103.5]	94.5 [85.3–105.8]
Height (cm)	165 [160–170]	164 [160–169]	167 [161–170]
BMI (kg/m2)	30.6 [29.3–34.3]	33.5 [30.4–36.0]	33.5 [30.9–37.1]
Waist (cm)	96 [89–109]	100 [93–107]	102 [94–110]
Hip (cm)	114 [107–122]	116 [109–124]	120 [113–129]
Waist-Hip ratio	0.84 [0.80–0.90]	0.87 [0.81–0.90]	0.84 [0.80–0.90]
Modified Ferriman–Gallwey score	3 [1–6]	4 [2–9]	3 [1–9]
Testosterone	1.48 [1.10–2.00]	1.55 [1.20–2.20]	1.49 [0.99–2.00]
Androstenedione	8.8 [5.7–13.8]	7.7 [5.3–11.0]	8.5 [5.0–13.4]
Dehydro-epiandrosterone (DHEA)	24.9 [19.1–44.2]	21.2 [14.3–27.9]	21.6 [15.2–34.9]
Insulin	88.5 [62.0–122.5]	102.5 [54.0–147.5]	87.0 [50.5–122.0]
HOMA-IR	1.10 [0.77–1.57]	1.26 [0.68–2.01]	1.10 [0.67–1.65]
Cortisol	309 [248–366]	262 [220–334]	323 [237–385]
	N (%)	N (%)	N (%)
Menstrual cycle
Oligomenorrhea	40 (88.9)	36 (80.0)	33 (70.2)
Amenorrhea	4 (8.9)	7 (15.6)	12 (25.5)
Regular	1 (2.2)	2 (4.4)	2 (4.3)
Spontaneous pregnancies	10 (16.7)	16 (26.7)	14 (23.3)
Hirsutism	11 (23.9)	16 (35.6)	14 (28.6)
Caucasian	14 (30.4)	17 (37.8)	24 (49)
Education
Low	5 (20.0)	1 (3.4)	2 (6.1)
Intermediate	15 (60.0)	16 (55.2)	20 (60.6)
High	5 (20.0)	12 (41.4)	11 (33.3)
History of depression	0 (0.0)	0 (0.0)	4 (8.2)
BDI-II > 13	19 (41.3)	22 (50.0)	21 (42.9)
BDI-II > 20	9 (19.6)	16 (36.4)	13 (27.1)

Depression

For depression a variance component covariance structure was found to be optimal (S2 Table). Depression scores decreased significantly in the LS intervention compared to CAU (Cohen's d = -0.34; p = 0.045). We observed no difference between LS with SMS and LS without SMS (Cohen's d = -0.02; p = 0.939), Table 2 and Fig 2. Over the study period, depression scores decreased in LS without SMS by 3.7 points (Cohen's d = -0.35; p = 0.036) and in the LS with SMS by 3.8 points (d = -0.37; p = 0.011). The decrease in LS with and without SMS is considered clinically significant, and both LS groups reached the MCID threshold of a more than 17.5% reduction from baseline. Within the CAU group no change in depression scores was observed (d = -0.02; p = 0.875, Table 3).

Table 2

Difference in depression, self-esteem and body image changes between study groups at 12 months.

The difference in depression, self-esteem and body image scores over time compared between lifestyle and care as usual and compared between lifestyle with SMS and without SMS. Cohen’s D 0.20 is small effect, 0.50 is medium effect and 0.80 a large effect, BDI-II = depression, RSES = self-esteem, FNAES = body image.

	Lifestyle vs Care as Usual	Lifestyle with SMS vs. Lifestyle without SMS
	Estimate	Estimate
BDI-II difference	-3.44	-1.19
Cohen's d	-0.41	0.04
P value	0.045	0.628
RSES difference	2.65	-0.24
Cohen's d	0.46	-0.04
P value	0.003	0.823
FNAES difference	-2.60	0.63
Cohen's d	-0.29	0.13
P value	0.094	0.756

Fig 2

Depression scores over time.

Table 3

Estimated depression, self-esteem and body image scores over time.

Cohen’s D: 0.20 = small effect, 0.50 = medium effect and 0.80 = a large effect.

	Group	Baseline	12 months	Change baseline—12 months
		Estimate	Estimate	Estimate	Percent	Cohen’s d	P value
Depression (BDI-II)	Care as usual (CAU)	13.3	13.1	-0.2	-1.5%	-0.02	0.875
	Lifestyle without SMS	15.5	11.9	-3.7	-23.6%	-0.35	0.036
	Lifestyle with SMS	13.2	9.4	-3.8	-29.0%	-0.37	0.011
Self-esteem (RSES)	Care as usual (CAU)	21.2	20.8	-0.4	-1.8%	0.02	0.688
	Lifestyle without SMS	18.8	21.5	+2.6	+14.0%	0.44	<0.001
	Lifestyle with SMS	19.5	21.7	+2.2	+11.2%	0.36	0.002
Body image (FNAE)	Care as usual (CAU)	15.5	14.6	-0.9	-5.5%	-0.12	0.447
	Lifestyle without SMS	18.9	15.4	-3.5	-18.5%	-0.50	0.001
	Lifestyle with SMS	18.1	14.8	-3.3	-18.1%	-0.47	0.008

Self-esteem

It was also the case that for self-esteem a variance component covariance structure was found to be optimal. Self-esteem scores improved significantly in the LS intervention compared to CAU (Cohen's d = 0.48; p = 0.027, Table 3 and Fig 3). We observed no beneficial effect for additional SMS during lifestyle treatment (Cohen's d = -0.07; p = 0.759). Self-esteem scores improved in the LS intervention without SMS by 2.6 points (Cohen's d = -0.44; p<0.001), and in the LS with SMS by 2.2 points (d = -0.36; p = 0.002). Self-esteem scores remained virtually stable within the CAU group (d = -0.02; p = 0.688), Fig 3.

Fig 3

Self-esteem scores over time.

Body image

For body image self-esteem an intercept only covariance structure was found to be optimal. We observed no difference for LS intervention compared to CAU (Cohen's d = -0.37; p = 0.087), see Table 3. Although body image scores did improve significantly within the LS intervention without SMS (d = -0.50; p = 0.001) and in LS with SMS (d = -0.47; p = 0.008). The improvement within the CAU group of d = -0.12 was not statistically significant (p = 0.447), Table 2 and Fig 4.

Fig 4

Body image scores over time.

Mediation

We tested 21 different paths β: the relationship between LS intervention and the 3 outcome measures (depression, self-esteem, and body image) with 7 potential mediators (testosterone, androstenedione, DHEA, insulin, HOMA-IR, cortisol, and weight loss), Fig 5. Only 4 paths β out of 21 turned out to be statistically significant with self-esteem; weight loss, androstenedione, testosterone, and DHEA. No significant paths were observed for insulin, HOMA-IR, and cortisol. Consequently, there was no mediation by insulin, HOMA-IR and cortisol. Notably, none of the paths with depression turned out to be statistically significant. When the potential mediators with statistical significant paths β were used, mediation was not found in either the relationship between LS (p = 0.613) and self-esteem with androstenedione, between LS with self-esteem and testosterone (p = 0.834), nor between LS with self-esteem and DHEA (p = 0.737). We also tested if weight loss mediated the effects on depression, self-esteem, and body image. First, we examined weight loss in all groups. In CAU, participants lost 2.32 kg, 4.65 kg in LS without SMS and 7.87 kg in LS with SMS (within all groups P<0.001). Second, we examined mediation in all three well-being outcomes. We found a nearly significant (p = 0.08) relationship between weight loss and self-esteem. In other words, weight loss had a nearly significant effect on the treatment-related changes in self-esteem. Weight loss appeared to be a strong mechanism by which the intervention improved self-esteem. Weight loss had no effect on the relationship between lifestyle treatment and improvements in depression or body-image.

Fig 5

Mediation effects.

Discussion

This study is the largest RCT investigating weight loss during a three-component CBT lifestyle intervention, and also the first to investigate long-term effects. All previous studies were LS interventions that lasted between 10 and 24 weeks [15–18, 49] and did not examine well-being in the long-term. We thus performed a secondary analysis of the well-being data that was collected in the RCT. We observed positive effects of LS treatment on depression scores during the entire 12 months intervention period. Others only observed short term effects that lasted for 10 weeks [17]. As discussed in the Introduction, some researchers have suggested that women with PCOS appear to have a unique risk for depression [8] that is persistent over time [11, 50], which could either be related to the condition itself, or to: weight, androgens, insulin, cortisol [22]. Hence we tested the potential mediation of androgens, insulin, HOMA-IR, and cortisol in the relationship between LS treatment and emotional well-being. Surprisingly, we found neither mediation by androgens nor by insulin, HOMA-IR, or cortisol. A nearly significant relationship was found between LS treatment and self-esteem mediated by weight loss, suggesting that the effects on self-esteem were caused by changes in weight loss. Our results suggest that the three-component intervention was the determining factor with respect to the improvements in depression and body-image, and that improvements in self-esteem were mediated by weight loss.

Compared to other LS interventions performed in women with PCOS [15, 16, 18], our intervention was the only one that was CBT-based. Previous interventions involved seminars covering weight-related topics [16, 49] or behavioral modification strategies [18]. Comparing our intervention to others is difficult because there are large differences in treatment protocols or information is lacking on which behavioral strategies are used. To optimize future research and promote treatment adherence, we used a standardized CBT protocol for all 20 sessions. During every therapy session a given topic was discussed, and the specific CBT techniques for that session were described in the study protocol [29].

In addition to the significant decline in depression scores, we also observed a clinically significant decline of ≥3 points [32] and a minimal clinically important difference (MCID) of more than the threshold of 17.5% [33] in LS interventions with and without SMS. The MCID is the optimal threshold above which individuals report feeling ‘better’. In other words, the three-component LS intervention improved depression rates while no changes in depression rates during CAU were observed. Little is known about the possible mechanism through which LS interventions achieve their effects or which components contributed the most [51]. Due to the design of our study we do not know if 1 or 2 of the 3 components (diet, exercise, CBT), or the 3 components as a whole, affected emotional well-being.

Participants in our study had lower mean self-esteem and lower body image scores compared to a previous study in women with PCOS [38]. This difference could be explained by BMI because our study population was more obese. As shown in the mediation analysis, weight loss mediated treatment effects on self-esteem, while this was not the case for the changes in depression and body-image during LS treatment. This result is in line with a meta-analysis of well-being outcomes in weight loss treatments. Only treatments that produced actual weight loss showed increased self-esteem, whereas improvements in depression were independent of weight loss. This indicates that self-esteem and depression are different constructs [52]. The improvements in body-image could be caused by the combination of CBT and group treatment, where group cohesion and social support might have played an important role. Many participants mentioned that the LS programme helped them to realize that they ‘were not alone’, emphasizing that PCOS and obesity made them feel lonely and insecure. It is known that group cohesion and social support can be strong in small groups [53] and especially where group members have similar backgrounds [54]. Other researchers have found that, especially for women with PCOS, group support is important for behavior change and reducing social isolation [55, 56]. The combination of small group treatment and one-year treatment seems beneficial for this group of women with PCOS beyond weight loss.

A strength of the current study is that we started with a population that was not severely depressed, whereas other researchers only either included [56] or excluded [57-59] severely depressed patients. [57-59]. Many LS programmes have excluded participants with symptoms of depression based on the idea that they may lose less weight than non-depressed participants [57-59]. Hence, our population might be a reliable reflection of the clinical situation where a substantial number of women with PCOS report moderate depressive symptoms [8]. It has also been suggested that depressed participants should be identified before entering an LS intervention and offered treatment for depression before entering an LS intervention [60]. Based on our findings, we consider that all women with PCOS, depressed or not depressed, can benefit from a three-component LS intervention. Moreover, in particular, participants with elevated depression scores at baseline should be selected for these interventions, since they can benefit the most from lifestyle treatment.

A limitation of our study lies in the high discontinuation rates we observed in all arms of the study. Compliance and drop-out are the most difficult aspects of any weight-reduction intervention, especially in programmes that last over 42 weeks [61]. In general weight loss programmes, dropout rates of around 40% are observed [40]. We expected to have relatively high discontinuation rates for two reasons: firstly, the intervention is demanding for participants (the intervention takes place on Monday afternoons and involves a one-year commitment) and secondly, because pregnancy, which is the ultimate goal for all participants, is considered as a reason to end study participation. Because high drop-out rates were expected in this intervention, a statistical method was chosen that could include all available data without imputation. Hence participants without a complete follow-up could also be included. This method also compensated for selective dropout, on the condition that dropout is related to variables included in the model [41].

Future research should examine whether the current LS programme could be further improved with more PCOS-related topics and/or specific CBT sessions about depressive thoughts. We have implemented the 3-component lifestyle intervention as standard care at our outpatient clinic to contribute to this development. Weight loss and depression are the biggest health concerns of women diagnosed with PCOS [62]. Based on their experiences, most women are not satisfied with the emotional support and help they receive [12, 62]. Thus, we believe that a three-component lifestyle programme should be accessible for all women with PCOS who are overweight or obese and trying to become pregnant. Three-component lifestyle interventions can contribute to a healthier weight, a better mood, and can enhance self-esteem and body image in women with PCOS.

Conclusions

A three-component LS intervention programme with or without additional SMS resulted in significant improvements in depression and self-esteem compared to CAU in women with PCOS, obesity, and a wish to achieve a pregnancy. Testosterone, androstenedione, DHEA, insulin, HOMA-IR, and cortisol did not mediate this effect. Weight loss mediated the effects on self-esteem but not on depression and body-image. This suggests that LS treatment independent of weight loss can reduce depression and body-image, whereas both LS treatment and weight loss can improve self-esteem. Hence, a three-component lifestyle intervention based on CBT can be successful in improving mood in women with PCOS who are overweight or obese and attempting to become pregnant.

Mixed models.

(TIFF)

Click here for additional data file.

Deviance tests for determination of linear and log time and covariance structure.

(TIFF)

Click here for additional data file.

31 Mar 2020

PONE-D-20-02106

Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a randomized controlled trial

PLOS ONE

Dear Mrs Jiskoot,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

The reviewers have highlighted several issues that need to be addressed including the concern over the drop out rate and the bias that may have resulted.

==============================

We would appreciate receiving your revised manuscript by May 15 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Stephen L Atkin, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files

3. We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 1 in your text; if accepted, production will need this reference to link the reader to the Table.

4. Thank you for stating the following in the Competing Interests section:

"GJ, ADL, AB, RT and JB have nothing to declare. JL has received unrestricted research grants from Ferring, MSD, Merck-Serono, Roche Diagnostics and Euroscreen. He received consultancy fees from the following companies: Euroscreen, Danone-Nutricia, Ferring, Roche Diagnostics, and Titus Healthcare."

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Results are reported from a clinical study, involving 155 obese women with polycystic ovary syndrome randomized to one of three study arms. It examined the effect of cognitive behavioral therapy lifestyle sessions combined with a healthy diet and physical therapy with or without 9 months additional feedback through SMS compared to care as usual. In women who received lifestyle sessions, depression scores were significantly lower than in controls.

Minor revisions:

1- Abstract: Define SMS.

2- Provide more precise p-values, rather than p< 0.05, p<0.01.

3- Line 170: Typographical error: Normality.

4- Line 172: Indicate the underlying covariance structure used in the mixed modeling the criteria for selecting it.

5-Cite the statistical software used for the analysis.

6- State and justify the study’s target sample size with a pre-study statistical power calculation. The power calculation should include: sample size, alpha level (indicating one or two-sided), minimal detectable difference and statistical testing method.

7- Normality is an underlying assumption for applying the Cohen's d statistic. Indicate if the distribution of the deltas (change from baseline to follow-up) for each outcome was normal.

8- As a limitation, discuss the effects of the high drop-out rate.

Reviewer #2: The study by Jiskoot et al assess the impact of CBT +/- SMS on emotional wellbeing in women with PCOS who suffer from overweight/obesity.

Notes

The study is interesting and examines an important topic. The authors have gone into a lot of efforts in designing and providing the CBT + lifestyle programme. The duration of the study is also much longer compared to other CBT studies and the authors deserve praise for that.

The study has clearly been hit by difficulty in recruitment and high drop-out rate which is unfortunate.

Comments

Main concerns.

- In the published protocol of the study ( published in Reproductive Health, Jiskoot et al. 2017); weight loss was the primary outcome while mental health was secondary outcome. However, the authors present the data in this paper as if it is a new study with mental health as the primary outcome, and weight loss is hardly mentioned. This is a protocol deviation which needs clarification.

- I feel the study will be much better if presented as a whole. Similar to the published protocol. With weight loss as the primary outcome and emotional health as secondary outcome.

- Only women who had a BMI > 25 kg/me and wanted pregnancy were included + Large number of drop outs. This affects generalisation of findings.

- It is not clear how missing data have been handled in particular around participants who got pregnant during the study.

- Data were reported for only some of the time points (baseline, 12 months). It will be better to report data at 3, 6, 9, and 12 months if available.

- Data reported do not match all the outcomes included in the protocol. This raises suspicion of selective reporting.

Other concerns:

Abstract:

- Line 34. Please change ‘’obese women with PCOS’’ to ‘’women with PCOS who have overweight/obesity’’.

- Line 36. Weight loss is mentioned in the conclusion but there is no mention of weight loss in the results section to justify this statement. It reads‘’the three-component intervention and/or weight loss is the determining factor for the improvements in emotional well-being.’’

Introduction

- Line 48. ‘’PCOS women’’. Please change to ‘’women with PCOS’’. The same applies throughout the manuscript.

- Line 50. ‘’abnormal amount of cholesterol derived lipids in their blood (dyslipidemia) characterized by high serum levels of triglycerides and low levels of HDL Cholesterol’’. Please rephrase.

- Line 60. ‘’non-obese women with PCOS’’. Please change to ‘’women with PCOS without obesity’’. The same applies throughout the manuscript.

- Line 100. ‘’overweight and obese women’’. Please change to ‘’women with overweight and obesity’’. Same applies throughout the manuscript.

- Line 101. ‘’The aim of the current study was to compare the changes in depression scores in a three-component CBT lifestyle intervention (with or without SMS), with these scores in the control group’’. This does not fit with the study protocol published in Reproductive Health, Jiskoot et al. 2017 where depression was a secondary outcome and weight loss was primary outcome.

Methods:

- Line 112, 113. Inclusion criteria. ‘’Women diagnosed with Polycystic Ovary Syndrome (PCOS) according to the Rotterdam 2003 consensus criteria and a BMI above 25 kg/m2 and a wish to become pregnant’’. Why only women who wanted to get pregnant were included? Why only women with BMI > 25 kg/m2 were included. Would that have implications on the generalisability of the study findings?

- Line113, 114. Exclusion criteria: ‘’pregnancy at baseline or follow-up’’. What about those with underlying mental health problem or on antidepressants?

- The inclusion/exclusion criteria paragraph is very small. Are there any other criteria not mentioned in paper?

- Study design. Lines 117, 118. ‘’The current study on emotional well-being represents a secondary analysis’’. Needs a bit more clarification. Secondary analysis of what? If this paper is a secondary analysis of another study, this should be made clear from the introduction rather than trying to present the data as a new study?

- Can you please clarify what were the study start and end recruitment dates?

- Other measures of quality of life were mentioned in the protocol but not reported, e.g PCOSQ, SF-36? Why not? Could these be included.

Analysis

- Statistics is better reviewed by a statistician.

- Can you please explain what is meant by intention to treat analysis in this study?

- The three groups were not equal in terms of weight or BMI? Has this been adjusted for?

- What happened to weight during the study?

- Has weight loss been considered as a mediator? If not why? can you consider it?

- Was there any difference in depression between LS+CBT vs. usual care group at baseline?

- Can you please clarify the percentage of patients who completed the programme for each questionnaire (BDI/FNAE/RSE). For example, only 46 participants out of 60 completed the BDI questionnaire at baseline in the usual care group. How many of these 46 participants completed the study?

- How many attended each visit at 3, 6, 9 and 12 months?

- In figure 1. 60 patients were assigned to each treatment group. Only 12 – 17 participants completed the study in each group (70 – 80% drop out). How do you explain the large drop out? How did you deal with missing data on follow up?

- In figure 1. More women in the LS+CBT groups got pregnant than the usual care group. How did you manage this data?

- Did any study participant receive any treatment to improve fertility?

Discussion

- Line 237. ‘’Our results suggest that the effect of the three-component intervention and/or weight loss is the determining factor for the improvement in emotional well-being scores’’. There is no mention on weight loss prior to this in the results section?

- Lines 291, 292. ‘’Therefore, we believe that a three-component lifestyle program should be accessible for all women with PCOS.’’ How can you justify this statement if all participants in the study wanted pregnancy and had a BMI > 25 kg/m2

- The limitation paragraph, should include comments on drop out, and restricting inclusion to women with BMI >25 kg/m2 who wanted pregnancy.

Conclusions:

- Line 296. ‘’obese women’’. Please change to ‘’women with obesity’’. Same applies throughout paper.

- Line 298. ‘’ Instead, the three-component intervention and/or weight loss is the determining factor for the improvements in emotional well-being.’’ Again weight loss is mentioned in the conclusions but not explained in results?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Dr Hassan Kahal

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

23 Apr 2020

April 22nd, 2020

Dear Dr. Atkin,

Thank you for reviewing our manuscript entitled ‘Successful weight reduction through a long-term cognitive behavioral therapy lifestyle intervention in polycystic ovary syndrome (PCOS): a randomized controlled trial’.

We are grateful for the detailed and constructive comments received from the reviewers, as they will help us to improve the document. You will find our specific responses in italics below, together with the original questions and comments of the reviewers, and together with the corresponding textual changes made. In the revised manuscript, we have marked these textual changes in yellow.

With kind regards,

On behalf of all the authors,

Geranne Jiskoot, MSc

Division of reproductive medicine, Department of Obstetrics and Gynecology, Erasmus MC

Wytemaweg 80, PO Box 2040, 3000 CA Rotterdam, the Netherlands

Phone: +31 (0) 6 44 184 834, E-mail: l.jiskoot@erasmusmc.nl

Reviewer #1:

Results are reported from a clinical study, involving 155 obese women with polycystic ovary syndrome randomized to one of three study arms. It examined the effect of cognitive behavioral therapy lifestyle sessions combined with a healthy diet and physical therapy with or without 9 months additional feedback through SMS compared to care as usual. In women who received lifestyle sessions, depression scores were significantly lower than in controls.

Minor revisions:

1- Abstract: Define SMS.

SMS defined as Short Message Service (SMS) via mobile phone and adjusted the sentence in the abstract.

2- Provide more precise p-values, rather than p< 0.05, p<0.01.

All the p-values adjusted more precisely in the abstract.

3- Line 170: Typographical error: Normality.

Adjusted.

4- Line 172: Indicate the underlying covariance structure used in the mixed modeling the criteria for selecting it.

Text added:

The covariance structure was determined by deviance tests, using restricted maximum likelihood. To this end, the unstructured component, the variance component, and the intercept-only covariance structures were compared with each other.

Text added to the results:

For all mixed models linear time was not significant, thus superfluous. The logarithm of time was included in all models.

For depression a variance component covariance structure was found optimal (Appendix B).

For self-esteem a variance component covariance structure was also found optimal.

For body image self-esteem an intercept-only covariance structure was found optimal.

Appendix B. Deviance tests for determination of linear and log time and covariance structure.

-2 loglikelihood df difference (χ²) df p-value

BDI-depression

linear + log 2711.087 5

linear only 2711.138 3 0.051 2 0.975

log only 2718.747 3 7.660 2 0.022

unstructured 2707.467 4

variance components 2711.138 3 3.671 1 0.055

intercept only 2723.752 2 12.614 1 0.000

Self-esteem

linear + log 2267.505 5

linear only 2265.149 3 2.356 2 0.308

log only 2272.762 3 5.257 2 0.072

unstructured 2263.733 4

variance components 2265.149 3 1.416 1 0.234

intercept only 2269.514 2 4.365 1 0.037

Body image

linear + log 2433.789 5

linear only 2432.151 3 1.638 2 0.441

log only 2437.781 3 3.992 2 0.136

unstructured 2430.285 4

variance components 2432.151 3 1.866 1 0.172

intercept only 2434.132 2 1.981 1 0.159

5. Cite the statistical software used for the analysis.

Information added with respect to the statistical software in the Analyses section.

6. State and justify the study’s target sample size with a pre-study statistical power calculation. The power calculation should include: sample size, alpha level (indicating one or two-sided), minimal detectable difference and statistical testing method.

The original sample size calculation was based on a difference between the groups of 0.45 in terms of Cohen’s d in the primary outcome variable (weight), with a power (1-beta) of 0.80 and an alpha level of 0.05 (two-sided). This resulted in 78 patients being enrolled in the lifestyle group with SMS, 78 patients in the lifestyle group without SMS, and 78 patients in the control group, a total of 234. This number was registered at the Dutch Trial Registry. During an interim power analysis we found an effect of Cohen’s d= 0.10 in the control group, whereas the lifestyle intervention group showed an effect of d= 0.52 (a difference of 0.42). Due to this large effect in the intervention group compared to the control group, we modified the original sample size calculation based on the method described by Aberson (25), with a power of 0.90, a two-sided alpha of 0.025 (corrected for the interim analysis as described in the study protocol), and 5 repeated measures linearly decreasing. We observed an intercorrelation of about 0.90 between all measurements. Maintaining a ratio of 1:1:1, the required sample was 42 in each group. With an expected drop-out proportion of 30% (26), 60 participants in each group were needed for the study.

This information was published in the study protocol paper. We agree with the reviewer that more information should have been added about the sample size calculation in this paper. A shorter version of the text above has been added.

7. Normality is an underlying assumption for applying the Cohen's d statistic. Indicate if the distribution of the deltas (change from baseline to follow-up) for each outcome was normal.

We agree that normality is an important assumption for Cohen's d. Additional analyses were applied, using Blom transformed dependent variables to recalculate the Cohen's effect sizes. These were also added to the statistical method section.

8- As a limitation, discuss the effects of the high drop-out rate.

We agree that the drop-out rate should receive more attention in the discussion section of the paper. The limitations section of the discussion now reads:

A limitation in our study lies in the high discontinuation rates we observed in all arms of the study. Compliance and drop-out are the most difficult aspects of any weight-reduction intervention, especially in programs that last over 42 weeks (61). In general weight loss programs, dropout rates of around 40% are observed (62). We expected to have relatively high discontinuation rates for two reasons: firstly, the intervention is demanding for participants (the intervention takes place on Monday afternoons and involves a one year commitment) and secondly, because pregnancy, which is the ultimate goal for all participants, is considered as a reason to end study participation. Because high-drop-out rates were expected in this intervention, a statistical method was chosen that could include all available data without imputation. Hence, participants without a complete follow-up could also be included. This method also compensated for selective dropout, on the condition that dropout is related to variables included in the model (40).

Reviewer #2:

The study by Jiskoot et al assess the impact of CBT +/- SMS on emotional wellbeing in women with PCOS who suffer from overweight/obesity.

Notes

The study is interesting and examines an important topic. The authors have gone into a lot of efforts in designing and providing the CBT + lifestyle program. The duration of the study is also much longer compared to other CBT studies and the authors deserve praise for that.

The study has clearly been hit by difficulty in recruitment and high drop-out rate which is unfortunate.

Comments

Main concerns.

1) In the published protocol of the study (published in Reproductive Health, Jiskoot et al. 2017); weight loss was the primary outcome while mental health was secondary outcome. However, the authors present the data in this paper as if it is a new study with mental health as the primary outcome, and weight loss is hardly mentioned. This is a protocol deviation which needs clarification.

We agree with the reviewer that the present presentation is ambivalent. Hence we have changed the title into: Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a secondary analysis of a randomized controlled trial. To avoid any suggestion of protocol deviation we have stressed the secondary nature of the analysis in: the introduction, the study design, the description of the outcome measures, the results section, and the first part of the discussion.

2) I feel the study will be much better if presented as a whole. Similar to the published protocol. With weight loss as the primary outcome and emotional health as secondary outcome.

Indeed, a presentation of the primary and secondary outcomes together would have its merits. However, it proved difficult to present all relevant aspects of both the primary and secondary outcomes in one paper. Combining primary and secondary outcomes in one paper in a transparent way, with an in-depth discussion of all the findings, would result in a manuscript beyond the conventional word limits of papers. We thus split the report of the investigation. A paper concerning the primary outcome (weight loss) is currently being reviewed. We agree that something should be written about weight loss and so we have added some information on this in the results section.

3) Only women who had a BMI > 25 kg/me and wanted pregnancy were included + Large number of drop-outs. This affects generalization of findings.

We agree with both reviewers that more information about drop-outs needs to be addressed by the paper; hence the limitations section in the discussion has been expanded. Information has also been added in the conclusion with respect to BMI and the desire to achieve a pregnancy: A three-component lifestyle intervention program with or without additional SMS resulted in significant improvements in depression and self-esteem compared to CAU in women with PCOS, obesity, and attempting to become pregnant.

4) It is not clear how missing data have been handled in particular around participants who got pregnant during the study.

We have expanded the multi-level analysis text to explain that drop-out or pregnancy is a problem in a complete cases analysis. The multilevel analysis is thus a more correct measure and includes all the additional data if participants have, e.g., measurements at 3 and 6 months and not at 9 and 12 months. This also implies that patients without complete follow-up can be included in the analyses, without imputation. This method also compensates for selective dropout, on the condition that dropout is related to variables included in the models.

5) Data were reported for only some of the time points (baseline, 12 months). It will be better to report data at 3, 6, 9, and 12 months if available.

Other lifestyle interventions lasted for 3 to 6 months and did not achieve long-term results. Hence we decided to concentrate on long-term results only, those found at 12 months. As the reviewer requested, information is added in figure 1 concerning the time points 3, 6 and 9 months

6) Data reported do not match all the outcomes included in the protocol. This raises suspicion of selective reporting.

7) Other measures of quality of life were mentioned in the protocol but not reported, e.g PCOSQ, SF-36? Why not? Could these be included.

We decided to focus on well-being in this paper and thus did not include quality of life measures. In addition, it transpired that the PCOSQ questionnaire was used in the study was not a correct translation. Therefore, we have submitted a new study protocol to the Medical Research Ethics Committee to validate this questionnaire. Moreover, earlier research suggested that generic QoL measures such as the SF-36 were not sensitive enough to measure changes in PCOS symptoms: this was originally the reason to include the PCOSQ. Because of the reported insensitivity of the SF-36, it was decided not to publish the SF-36 data in this paper.

Other concerns:

Abstract:

8) Line 36. Weight loss is mentioned in the conclusion but there is no mention of weight loss in the results section to justify this statement. It reads ‘the three-component intervention and/or weight loss is the determining factor for the improvements in emotional well-being.’

Information about weight-loss is given in the results section. Based on your advice at point 18, we tested weight loss as a mediator and added this analysis to the paper.

Introduction

9) Line 50. ‘’abnormal amount of cholesterol derived lipids in their blood (dyslipidemia) characterized by high serum levels of triglycerides and low levels of HDL Cholesterol’’. Please rephrase.

We have changed this sentence into: Most women with PCOS experience one or more of the following physical symptoms in varying degrees: hirsutism (excessive body hair growth), acne, infertility, obesity, insulin resistance, and dyslipidemia

9) Line 101. ‘’The aim of the current study was to compare the changes in depression scores in a three-component CBT lifestyle intervention (with or without SMS), with these scores in the control group’’. This does not fit with the study protocol published in Reproductive Health, Jiskoot et al. 2017 where depression was a secondary outcome and weight loss was primary outcome.

See also points 1 and 2. We have changed the sentence into: The aim of this secondary analysis was to compare the changes in depression scores in a three-component CBT LS intervention (with or without SMS), with these scores in the control group.

Methods:

10) Lines 112, 113. Inclusion criteria. ‘’Women diagnosed with Polycystic Ovary Syndrome (PCOS) according to the Rotterdam 2003 consensus criteria and a BMI above 25 kg/m2 and a wish to become pregnant’’. Why only women who wanted to get pregnant were included? Why only women with BMI > 25 kg/m2 were included. Would that have implications on the generalisability of the study findings?

When this study commenced, it was already known that being overweight and obesity had a negative impact on PCOS characteristics. The study was initiated to improve PCOS characteristics in a high-risk group of women with PCOS. Based on the newest latest PCOS guideline, lifestyle intervention (preferably multi-component including diet, exercise and behavioral strategies) is recommended in all those with PCOS and excess weight, for reductions in weight, central obesity and insulin resistance and not only in patients that desire a pregnancy. We believe that the results of this study are applicable for all women with PCOS.

11) Lines 113, 114. Exclusion criteria: ‘’pregnancy at baseline or follow-up’’. What about those with underlying mental health problem or on antidepressants?

12) The inclusion/exclusion criteria paragraph is very small. Are there any other criteria not mentioned in paper?

We did not use mental health scores as an exclusion criterion. Required was a reflection of a normal population of women with PCOS with a BMI > 25, thus minimizing the number of exclusion criteria employed to ensure the generalizability of our findings.

13) Study design. Lines 117, 118. ‘’The current study on emotional well-being represents a secondary analysis’’. Needs a bit more clarification. Secondary analysis of what? If this paper is a secondary analysis of another study, this should be made clear from the introduction rather than trying to present the data as a new study?

We agree with the reviewer, see answers to points 1 and 2.

14) Can you please clarify what were the study start and end recruitment dates?

The start and end dates are described in the first part of the results: Between August 2nd 2010 and March 11th 2016, 561 eligible women were asked to participate and 209 provided written informed consent, of whom 26 were included in a pilot study.

Analysis

15) Can you please explain what is meant by intention to treat analysis in this study?

We have used the definition of intention to treat by McCoy (2017): all participants who are randomized are included in the statistical analysis and analyzed according to the group they were originally assigned, regardless of what treatment (if any) they received.

McCoy, C. E. (2017). Understanding the intention-to-treat principle in randomized controlled trials. Western Journal of Emergency Medicine, 18(6), 1075.

16) The three groups were not equal in terms of weight or BMI? Has this been adjusted for?

Indeed, the lifestyle intervention group started at a higher overall weight. However, mixed modeling takes into account that participants have different baseline values, as defined in the random intercept. In the covariance structure time was also included as a random effect, allowing for participants to have a different course or slope during the study. In lifestyle intervention, especially, the slope (weight loss or weight gain) was important for each individual. Even when the lifestyle group started with a higher baseline weight and finished at about the same weight, it can be concluded that the SMS+ group, in particular, was more successful because the weight loss slope was steeper.

17) What happened to weight during the study?

Information added to the results section.

18) Has weight loss been considered as a mediator? If not why? can you consider it?

We agree that weight loss as a mediator is an interesting question. Thank you for this idea. The analysis has been added to the paper.

19) Was there any difference in depression between LS+CBT vs. usual care group at baseline?

The lifestyle intervention group started with slightly higher depression scores but this difference was not statistically significant between LS without SMS and CAU (P=0.103), LS with SMS and CAU (P=0.701), and LS with SMS and LS without SMS (P=0.189).

20) Can you please clarify the percentage of patients who completed the program for each questionnaire (BDI/FNAE/RSE). For example, only 46 participants out of 60 completed the BDI questionnaire at baseline in the usual care group. How many of these 46 participants completed the study? How many attended each visit at 3, 6, 9 and 12 months?

Figure one has been adjusted. based on your recommendations.

21) In figure 1. 60 patients were assigned to each treatment group. Only 12 – 17 participants completed the study in each group (70 – 80% drop out). How do you explain the large drop out? How did you deal with missing data on follow up?

We consider the high drop-out numbers to be normal in interventions for weight-loss and especially in programs that are time-consuming (20 Monday afternoon meetings lasting 2.5 hours over a one-year period). For many women this was difficult to combine with work or other commitments. A lifestyle program with an emphasis on changing behavior was different from what many participants expected. Most expected a traditional weight-loss program with a strict caloric deficit. A behavioral program takes more time, and personal reflection is needed in order to change behavior. Additionally, achieving a pregnancy was a reason for drop-out because weight loss is not recommended during pregnancy. We hypothesize that the high number of drop-outs is a combination of these 3 factors.

22) In figure 1. More women in the LS+CBT groups got pregnant than the usual care group. How did you manage this data?

Currently, pregnancy rates are not available for all the participants in the study and especially for participants who dropped-out. The informed consent form we did not include a statement for pregnancy follow-up. Hence a new study protocol has been submitted to the Medical Research Ethics Committee to allow us to perform the follow-up measurements. Pregnancy data is currently being collected in order to prepare a paper concerning the long-term effects of the lifestyle intervention after 5 years of randomization.

We agree that spontaneous pregnancy data may be confusing because it did not include all pregnancies. Hence it was decided to exclude the spontaneous pregnancy data from Figure 1.

23) Did any study participant receive any treatment to improve fertility?

No. This information has been added to the methods section, see line 119.

Discussion

24) Line 237. ‘Our results suggest that the effect of the three-component intervention and/or weight loss is the determining factor for the improvement in emotional well-being scores’. There is no mention on weight loss prior to this in the results section?

The conclusion has been changed to:

A three-component lifestyle intervention program with or without additional SMS resulted in significant improvements in depression and self-esteem compared to CAU in women with PCOS, obesity and attempting to become pregnant. Testosterone, androstenedione, DHEA, insulin, HOMA-IR and cortisol did not mediate this effect. Weight loss mediated the effects on self-esteem but not on depression and body-image. This suggests that lifestyle treatment independent of weight loss can reduce depression and body-image, but both lifestyle treatment and weight loss can improve self-esteem. Hence, a three-component lifestyle intervention based on CBT can be successful in improving mood in women with PCOS who are overweight or obese and attempting to become pregnant.

25) Lines 291, 292. ‘’Therefore, we believe that a three-component lifestyle program should be accessible for all women with PCOS.’’ How can you justify this statement if all participants in the study wanted pregnancy and had a BMI > 25 kg/m2 and

Line 298. ‘’ Instead, the three-component intervention and/or weight loss is the determining factor for the improvements in emotional well-being.’’ Again weight loss is mentioned in the conclusions but not explained in results?

We agree with the reviewer that this conclusion should be specific for women with PCOS, who are overweight and trying to conceive. This sentence has been changed to: Hence, we believe that a three-component lifestyle program should be accessible for all women with PCOS who are overweight or obese and trying to conceive.

26) The limitation paragraph, should include comments on drop out, and restricting inclusion to women with BMI >25 kg/m2 who wanted pregnancy.

Limitation section changed.

Furthermore, we have changed all sentences in lines 34, 48, 60 and 100 into ‘women with PCOS who have overweight/obesity’ or ‘women with PCOS’.

Submitted filename: Rebuttal PlosOne.doc

Click here for additional data file.

28 Apr 2020

PONE-D-20-02106R1

Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a secondary analysis of a randomized controlled trial

PLOS ONE

Dear Mrs Jiskoot,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

please address the reviewers comments regarding the spontaneous pregnancy

==============================

We would appreciate receiving your revised manuscript by Jun 12 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Stephen L Atkin, MD

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: Dear Professor Atkin,

The reviewers have engaged well with the review process and I feel the manuscript is now much clearer.

Comments

1. I feel spontaneous pregnancy during the study is an important information and it is worth mentioning it in the manuscript (it was included in the first submission but excluded in revised version).

2. There has been difference in spontaneous pregnancy rates (during study) between the 3 groups (10 in control group, compared to 14 and 16 in the intervention groups). All women participating in the study wanted fertility, was spontaneous pregnancy during the study considered as a mediator/confounder? If not, could it be included?

3. In the baseline data table; the median weight was 84, 89, and 94.5kg in the control, LS, and LS+SMS groups respectively. While the median BMIs were 30.6, 33.5, and 33.5 kg/m2 respectively. I find it strange that the SL+SMS group was 5kg heavier than SL group but with similar BMI. Can this result be double checked?

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Dr Hassan Kahal

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

11 May 2020

May 11th, 2020

Dear Dr. Atkin,

Thank you for reviewing our manuscript entitled ‘Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a secondary analysis of a randomized controlled trial’.

We are grateful for the detailed and constructive comments received from the reviewers, as they will help us to improve the manuscript. You will find our specific responses to the questions and comments of the reviewers (in italics) and the corresponding textual changes made. In the revised manuscript, we have marked these textual changes in yellow.

With kind regards,

On behalf of all authors,

Geranne Jiskoot, MSc

Division of reproductive medicine, Department of Obstetrics and Gynecology, Erasmus MC

Wytemaweg 80, PO Box 2040, 3000 CA Rotterdam, the Netherlands

Phone: +31 (0) 6 44 184 834, E-mail: l.jiskoot@erasmusmc.nl

1. I feel spontaneous pregnancy during the study is an important information and it is worth mentioning it in the manuscript (it was included in the first submission but excluded in revised version).

We included the information about spontaneous pregnancies in Table 1. The difference between LS without SMS compared to CAU was not significant (P=0.816), LS with SMS compared to CAU (P=0.198) and LS with SMS compared to LS without SMS (P=0.130).

2. There has been difference in spontaneous pregnancy rates (during study) between the 3 groups. All women participating in the study wanted fertility, was spontaneous pregnancy during the study considered as a mediator/confounder? If not, could it be included?

We agree with the reviewer that this could be an interesting possibility. Therefore, we have checked if spontaneous pregnancy mediated the effects of the lifestyle treatment on well-being scores. We found no mediation in the relationship between depression with spontaneous pregnancy (P=0.766), nor between self-esteem with spontaneous pregnancy (P=0.570) and body image with spontaneous pregnancy (P=0.890).

3. In the baseline data table; the median weight was 84, 89, and 94.5kg in the control, LS, and LS+SMS groups respectively. While the median BMIs were 30.6, 33.5, and 33.5 kg/m2 respectively. I find it strange that the SL+SMS group was 5kg heavier than SL group but with similar BMI. Can this result be double checked?

We double checked the data and included the SPSS output for weight, height and BMI below. The median weights as reported in the paper are correct. The LS with SMS is taller and heavier, therefore they have the same BMI as LS without SMS. We have added information about height in Table 1.

Statistics

group HEIGHT WEIGHT BMI

1 Controle N Valid 60 60 60

Missing 0 0 0

Percentiles 25 160.00 79.00 29.336

50 165.00 84.00 30.662

75 170.00 97.25 34.375

2 SMS- N Valid 63 63 63

Missing 0 0 0

Percentiles 25 160.00 80.00 30.469

50 164.00 89.00 33.695

75 169.00 105.00 36.506

3 SMS+ N Valid 60 60 60

Missing 0 0 0

Percentiles 25 161.25 85.25 31.227

50 167.00 94.50 33.577

75 170.00 105.75 37.071

Submitted filename: Rebuttal PlosOne.doc

Click here for additional data file.

15 May 2020

Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a secondary analysis of a randomized controlled trial

PONE-D-20-02106R2

Dear Dr. Jiskoot,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Stephen L Atkin, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

21 May 2020

PONE-D-20-02106R2

Long-term effects of a three-component lifestyle intervention on emotional well-being in women with Polycystic Ovary Syndrome (PCOS): a secondary analysis of a randomized controlled trial

Dear Dr. Jiskoot:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Stephen L Atkin

Academic Editor

PLOS ONE