Impact of a hypocaloric dietary intervention on antral follicle dynamics in eumenorrheic women with obesity

Abstract

STUDY QUESTION

Do antral follicle dynamics change in women with obesity and regular ovulatory cycles after a 6-month hypocaloric dietary intervention?

SUMMARY ANSWER

After a 6-month hypocaloric dietary intervention, women with obesity and regular ovulatory cycles displayed evidence of improved antral follicle dynamics defined by the emergence of more dominant follicles, larger ovulatory follicle diameter at selection, and increased luteal progesterone concentrations compared to pre-intervention.

WHAT IS KNOWN ALREADY

Precise events in antral folliculogenesis must occur in order for natural and regular monthly ovulation. In healthy women of reproductive age, antral follicles are recruited for growth in a wave-like fashion, wherein a subset of follicles are selected for preferential growth, and typically, one dominant follicle culminates in ovulation. Women with obesity and regular ovulatory cycles display evidence of suppressed antral follicle development, as evidenced by fewer recruitment events, fewer selectable and dominant follicles, smaller diameter of the ovulatory follicle at selection, and a higher prevalence of luteal phase defects. While improvements in gonadotropin and ovarian steroid hormone concentrations after weight loss have been documented in eumenorrheic women with obesity, the precise impact of weight loss on antral follicle dynamics has not been evaluated.

STUDY DESIGN, SIZE, DURATION

A pre–post pilot study of 12 women who participated in a 6-month hypocaloric dietary intervention.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Twelve women with obesity (total body fat ≥35%) underwent transvaginal ultrasonography and venipuncture every-other-day for one inter-ovulatory interval (IOI) both before (baseline) and during the final month (Month 7) of a six-month hypocaloric dietary intervention. Participants were aged 24–34 years and had a self-reported history of regular menstrual cycles (25–35 days). Follicle number and diameter (≥2 mm) were quantified at each study visit, and individual growth profiles for all follicles ≥7 mm were determined. Blood samples were assayed for reproductive hormones. Follicle dynamics and reproductive hormone concentrations were compared pre- and post-intervention. Further, post-intervention follicle and endocrine dynamics (Month 7 IOI) were compared to an age-matched reference cohort of lean women with regular ovulatory cycles (total body fat <35%, N = 21).

MAIN RESULTS AND THE ROLE OF CHANCE

Participants lost an average of 11% of their original body weight with the hypocaloric dietary intervention. More dominant follicles were detected (≥10 mm) at Month 7 compared to baseline (0. 3 ± 0.4 versus 0.4 ± 0.5 follicles, P = 0.001), and ovulatory follicles were selected at larger diameters post-intervention (7.3 ± 2.0 versus 10.9 ± 2.6 mm, P = 0.007). Luteal progesterone concentrations were increased at Month 7 compared to baseline (5.3 ± 3.65 versus 6.3 ± 4.74 ng/ml, P < 0.0001). However, risk for luteal phase dysfunction as judged by the prevalence of a luteal phase length <10 days, integrated luteal progesterone levels <80 ng/ml or peak progesterone <10 ng/ml did not differ pre- versus post-intervention (all, P > 0.05). In Month 7, follicle dynamics and endocrine profiles were similar to the reference cohort across all measures.

LIMITATIONS, REASONS FOR CAUTION

This study does not inform on the earliest stages of ovarian follicle development and is limited to providing knowledge on the later stages of antral follicle development. This study cannot fully address causation between weight loss and sustained improvements in antral follicle dynamics. The data cannot be extrapolated to comment on potential improvements in fertility and fecundity with weight loss. The small group sizes limit statistical power.

WIDER IMPLICATIONS OF THE FINDINGS

The increasing prevalence of obesity necessitates an understanding of the mechanisms that underlie potential improvements in reproductive health outcomes with weight loss. Women with obesity and regular ovulatory cycles who undertook a 6-month hypocaloric dietary intervention demonstrated improvements consistent with benefits of lifestyle intervention on reproductive health even in those without overt signs of reproductive dysfunction. Potential improvements in the cellular makeup of follicles, which may underlie the restoration of normal follicle development and amelioration of subfertility, require further investigation.

STUDY FUNDING/COMPETING INTEREST(S)

Cornell University, President’s Council of Cornell Women, United States Department of Agriculture (Grant No. 8106), and National Institutes of Health (R01-HD0937848). B.Y.J. and H.V.B. were supported by doctoral training awards from the National Institutes of Health (T32-DK007158) and Canadian Institutes of Health Research (Grant No. 146182), respectively. The authors have no competing interests.

TRIAL REGISTRATION NUMBER

NCT01927432 and NCT01785719.

Introduction

Obesity adversely impacts ovarian function and reproductive health in women (Fedorcsák et al., 2004; Dağ and Dilbaz, 2015; Goldsammler et al., 2018; Silvestris et al., 2018). Even in those with regular menstrual cycles, aberrations in reproductive hormone dynamics have been described, including decreased luteal progesterone (P4) production (De Pergola et al., 2006; Jain et al., 2007; Yeung et al., 2013), smaller LH pulse amplitude (Jain et al., 2007), reduced anti-Müllerian hormone (AMH) concentrations (Su et al. 2008; Chiofalo et al., 2017; Oldfield et al., 2023; Olszanecka-Glinianowicz et al., 2015; Peigné et al., 2020), and lower FSH levels across the cycle (Jain et al., 2007). We recently reported that endocrine aberrations reflect disordered antral follicle development in eumenorrheic women with obesity (Oldfield et al., 2023). Namely, women with obesity and regular menstrual cycles have suppressed antral follicle dynamics as evidenced by fewer recruitment events, fewer selectable and dominant follicles as well as smaller diameters of ovulatory follicles at selection (Oldfield et al., 2023). These alterations in follicle dynamics were accompanied by reduced AMH and P4 concentrations, particularly in the luteal phase with a higher prevalence of luteal phase defects noted in women with obesity and regular cycles compared to their lean counterparts (Oldfield et al., 2023). These data support a downregulation of ovarian antral follicle development and increased likelihood of luteal phase defects as underlying factors in the subfertility (Dağ and Dilbaz, 2015; Silvestris et al., 2018) and risk for endometrial hyperplasia (Epplein et al., 2008) reported in this population.

Weight loss is known to elicit beneficial impacts on reproductive health in women with obesity and irregular menstrual cycles (Lan et al., 2017). However, the effects of weight loss on reproductive outcomes in those with eumenorrhea are less defined. One study showed that even small reductions in weight (i.e. 3.8% average weight loss) induced by aerobic training improved FSH and estradiol (E2) concentrations (Al-Eisa et al., 2017). By contrast, weight loss ranging from 6% to 13% by low-energy diets showed no clinically significant improvements in FSH and/or E2 concentrations (Turcato et al., 1997; Grenman et al., 1986; Pasquali et al., 2000; Panidis et al., 2008). The type of intervention or degree of weight loss may have contributed to the discrepancies across studies, as weight loss achieved by bariatric surgery in participants with regular menstrual cycles, who lost an average of 32% of their presurgical weight, showed increased LH and P4 concentrations with partial restoration of luteal function (Rochester et al., 2009). Improved luteal function with weight loss may increase the likelihood of conception for those desiring pregnancy and provide adequate opposition of estrogen throughout the menstrual cycle, which is needed to offset heightened risks for abnormal vaginal bleeding, hysterectomy, and endometrial hyperplasia (Njoku et al., 2020; Hiller and Griesinger, 2023). Determining the precise role of weight loss on ovarian function is needed to fully address expectations related to the range of benefits of lifestyle intervention for all women living with obesity.

The primary objective of the current study was to contrast ovarian antral follicle growth and endocrine dynamics in eumenorrheic women with obesity before and after a 6-month hypocaloric dietary intervention. We hypothesized that weight loss induced by a hypocaloric dietary intervention would improve all stages of antral follicle development including an increased number of recruitment events, emergence of more dominant follicles, and a larger follicle diameter at selection. Further, we expected that any alterations in antral follicle dynamics induced by weight loss would align with improvements in reproductive hormone concentrations across the menstrual cycle particularly, improved luteal function post-ovulation.

Materials and methods

Study participants

This pilot study represents a pre-/post-analysis of a subset of female participants who consecutively completed a single-arm, 6-month hypocaloric dietary intervention as part of an ongoing registered clinical trial (NCT01785719). Participants were included in the analysis if they retrospectively met the following criteria at baseline: between the ages of 18–38 years, obesity status, regular menstrual cycle lengths and normal androgen levels. Obesity was defined by a percent total body fat ≥35% using a dual X-ray absorptiometry (DEXA) (Piqueras et al., 2021). Menstrual cycle regularity was defined as a self-reported menstrual cycle length between 21 and 35 days in the last year and confirmed post hoc using ultrasound monitoring of ovarian antral follicle development during an inter-ovulatory interval (IOI; described below). All participants included were confirmed to be normoandrogenic as defined by a total testosterone (T) of <61.5 ng/dl based on a threshold derived using an internal reference cohort. To be eligible for inclusion, participants must have had consistent and optimal visualization of both ovaries on ultrasonography. Participants were excluded if they were using medications known or suspected to interfere with reproductive function in the two months prior to the study; were pregnant or lactating in the 6 months prior to the study; had a history of premature ovarian insufficiency; or had any pre-existing confounding medical conditions (e.g. diabetes, thyroid dysfunction, hyperprolactinemia, eating disorder). All participants were also required to be at the action stage of readiness to lose weight (Kristal et al., 1999), as judged by a validated questionnaire (Duyff, 2006) and have willingness to adhere to a prescribed commercial diet for 6 months.

Ethical considerations

This study was approved by the Institutional Review Board at Cornell University and registered at ClinicalTrials.gov (NCT01785719). Informed consent was obtained from all participants before study procedures were performed. Participants were recruited from the Southern Tier of New York using advertisements for the general population. Data from the reference cohort were collected as part of another registered trial (NCT01927432) and have been reported elsewhere (Oldfield et al., 2023).

Study design

Participants were evaluated for a total of 7 months. During Month 1, participants visited the Human Metabolic Research Unit at Cornell University every other day for transvaginal ultrasonography and venipuncture. Every other day visits spanned one IOI (i.e. the baseline IOI), which represented the time from one ovulation to the subsequent ovulation. During the mid-follicular phase (Days 7–11) of the baseline IOI, participants also participated in a study visit following an overnight fast which consisted of an oral glucose tolerance test and physical examination. A subset of the data related to the baseline IOI have been previously reported (Oldfield et al., 2023). During the 5th week of the study, participants began a 6-month commercial hypocaloric dietary intervention (described below). During Month 6 of the hypocaloric dietary intervention, corresponding with Month 7 of the study period, participants resumed every-other-day visits to capture another IOI (i.e. the Month 7 IOI). A fasting study visit was repeated during the mid-follicular phase of this final Month 7 IOI.

Reference cohort

Twenty-one participants with regular menstrual cycles, normal androgen concentrations, and no evidence of obesity (i.e. total percent body fat <35%), who had completed a clinical trial (NCT01927432) with similar baseline assessments to those reported herein, served as a reference cohort (Oldfield et al., 2023).

Hypocaloric dietary intervention

Participants were assigned to Nutrisystem^® D (Nutrisystem, Inc., Fort Washington, PA), a portion-controlled, hypocaloric, and low glycemic index meal delivery system designed to help participants consume 1250–1500 calories per day. A physical activity goal of 30 min per day was also encouraged. A nutritionist met with each participant weekly to customize meal plans, manage roadblocks to weight loss as they emerged, and counsel participants on behavioral strategies for weight loss, including self-monitoring and goal setting. Changes in anthropometry (i.e. weight, and waist and hips circumference) were assessed during twice-weekly visits to the research unit. Additional body composition measures were evaluated during Month 1 (baseline IOI), after the achievement of weight loss benchmarks (i.e. at 5% and 10% initial body weight lost, if achieved before Month 7), and during the final month of the intervention (Month 7 IOI) using DEXA.

Ultrasonographic measurements

Antral follicle dynamics were evaluated using serial transvaginal ultrasonography in Month 1 and Month 7 using methods previously described (Baerwald et al., 2003, 2004; Rouleau et al., 2012; Jarrett et al., 2020; Oldfield et al., 2023). Scanning intervals began before ovulation (range: Cycle days 8–15) and ultrasound scans were conducted approximately every other day for one IOI, both before (baseline IOI), and during the last month of the 6-month hypocaloric dietary intervention (Month 7 IOI). As defined in Oldfield et al., an IOI captured both the luteal and follicular phases (Oldfield et al., 2023). When a large antral follicle, ∼≥16 mm, was detected, ultrasound examinations were performed daily until the fate of the large antral follicle was confirmed (i.e. ovulation or regression). Ovulation was defined as the sonographic detection of a corpus luteum during the IOI and was confirmed post hoc with a rise in serum progesterone concentrations of ≥1.5 ng/ml (Baerwald et al., 2005).

Scans were performed using a GE Voluson E8 Expert System or a GE Voluson E10 Expert System and 6–12 MHz 3D/4D transducer (GE Healthcare, Milwaukee, WI). Ovaries were imaged from their inner to outer margins in the longitudinal plane using the automated volume modality as per our standard internal protocol for image collection (Oldfield et al., 2023). Images were stored and later evaluated offline using Sante DICOM Editor (Santesoft Ltd, Athens, Greece) by three investigators that showed high agreement across all measurements performed on an independent set of images (N = 30) to justify pooling of data across raters (Average Intra Class Correlation Coefficient = 0.902). Follicles were counted and measured (in the largest cross-sectional view) for each visit using the grid system approach which ensures reliability in the measurements performed (Lujan et al., 2010).

As described previously, growth and regression profiles of individual follicles ≥7 mm were tracked using the Identity Method (Baerwald et al., 2003; Baerwald et al., 2004; Vanden Brink et al., 2013; Jarrett et al., 2020; Oldfield et al., 2023). Briefly, follicles with a diameter ≥4 mm were sketched in their relative positions to generate a map of antral follicles within each ovary. Maps were completed for each ovary at each visit of the IOI. All follicles that grew to ≥7 mm were uniquely identified, and changes in diameter were tracked from day of first detection (i.e. at 4–5 mm) to last detection (i.e. at 4–5 mm or ovulation). Growth and regression rates of each uniquely identified follicle were then calculated. Sonographic presence was defined as the interval of time between the first and last days of sonographic detection of a follicle (Baerwald et al., 2009; Jarrett et al., 2020). The growth phase was defined as the interval of time from first day of detection to the day of maximal follicle diameter (Baerwald et al., 2009), and the regression phase was defined as the interval of time from the day of maximal follicle diameter to the day of last detection (Baerwald et al., 2009; Jarrett et al., 2020). Growth and regression rates were defined as previously described (Baerwald et al., 2009; Jarrett et al., 2020).

Use of every-other-day transvaginal ultrasonography allowed for the characterization of key events in antral follicle dynamics. A recruitment event was defined when two or more small antral follicles (≥4 mm) emerged within a 3-day (or two-visit) window and went on to grow to the selectable stage (≥7 mm), alongside an increase and subsequent decrease in the number of follicles ≥5 mm. In contrast to previous reports of antral follicle dynamics in eumenorrheic participants (Baerwald et al., 2003, 2004), we define cohorts of antral follicles as recruitment events (and not as follicle waves) given our lack of consideration of FSH in the definition of a recruitment event. Dominance was morphologically defined as the growth of a follicle to ≥10 mm (Baerwald et al., 2004). Last, selection was defined by the occurrence of a (future) dominant follicle that grew ≥1 mm larger than the other follicles in the ovary and remained larger until it reached its fate (Baerwald et al., 2003).

Because there were no differences in the number of uniquely identified follicles between the left and right ovaries (data not shown) and due to general convention, follicle number and diameter data from both ovaries were combined (Baerwald et al., 2003, 2004; Vanden Brink et al., 2013; Jarrett et al., 2020). The total number and proportion of follicles detected in different diameter categories were graphed for each participant over the IOI. Diameter categories of physiologic interest (i.e. antral follicle counts [AFCs]) included: ≥2 mm (total follicle count), 2–5 mm (recruitable follicles), 6–9 mm (selectable follicles), and ≥10 mm (dominant follicles).

Biochemical measurements

Non-fasted blood samples were collected every other day during the IOI. Blood was collected into a clot-activated tube and allowed to sit at room temperature for 30–60 min. Serum was isolated by centrifugation and stored at –80°C until analysis. Chemiluminescence immunoassays (Immulite 2000, Siemens Medical Solutions Diagnostics, Deerfield, IL) were used to measure serum concentrations of FSH, LH, E2, and P4. Inter- and intra-assay coefficients of variation (CV) were as follows: FSH (4.9%, 2.6%), LH (6.2%, 3.9%), E2 (9.7%, 8.6%), and P4 (11.8%, 7.2%), respectively. Risk for luteal phase defects (LPDs) was defined by evidence of a decreased luteal phase length (<10 days) and/or biochemical measures of integrated luteal P4 < 80 ng/ml or peak P4 < 10 ng/ml (Jordan et al., 1994; Schliep et al., 2014; Practice Committees of the American Society for Reproductive Medicine and the Society for Reproductive Endocrinology and Infertility 2021).

To assess androgen and glucoregulatory status, fasted blood samples were drawn on a single day of each IOI at a standardized time such that no dominant follicles or active corpora lutea were present. Serum sex hormone binding globulin (SHBG) was measured by chemiluminescence immunoassay (inter-assay CV: 5.0%; intra-assay CV: 3.1%) and total T was measured by liquid chromatography tandem mass spectrometry (inter-assay CV: 6.4%) as previously described (Vanden Brink et al., 2016). The free androgen index (FAI) was calculated as: (total T [nmol/l]/SHBG [nmol/l]) × 100 (Vermeulen et al., 1999). Glucose was measured with a standard glucometer (Accu-Check Aviva, Roche Diabetes Care, Inc., Indianapolis, IN) and insulin was measured by chemiluminescence immunoassay (inter-assay CV: 6.2%; intra-assay CV: 4.8%). The homeostatic model assessment for insulin resistance (HOMA-IR) was calculated as: (fasting glucose [mmol/l] × fasting insulin [mIU/ml]) ÷ 22.5 (Wallace et al., 2004). AMH was measured by enzyme-linked immunosorbent assay at commercial facility (picoAMH ELISA, Ansh Labs, Webster, TX) (inter-assay CV: 5.7%; intra-assay CV: 2.9%).

Statistical analysis

All analyses were performed using JMP Pro 14.0.1 (SAS Institute, Cary, NC). Data were log-transformed if needed to meet model assumptions. Continuous cross-sectional data, obtained in Month 1 and Month 7, were compared using matched-pairs t-tests. Fisher’s exact tests were used to compare categorical variables across time points and between groups (i.e. LPDs). Longitudinal follicular and endocrine data were centralized to the day of ovulation and evaluated by: (1) normalizing the data across the IOI and (2) separately averaging the data across the luteal and follicular phases. Mixed-effect models evaluated the impact of the intervention on follicle number, follicle size populations, growth parameters, and endocrine hormones (main fixed effect: intervention status). Participant identifier was used as a random effect and day was used as a fixed effect across all models. The statistical significance threshold was set at P < 0.05.

Results

Participant characteristics

Reproductive, anthropometric, and metabolic features of participants, before and after the 6-month hypocaloric dietary intervention, are contrasted in Table 1. On average, participants were 31 ± 3 years old at baseline. Participants lost 11% ± 3.8% (mean ± standard deviation) of their initial body weight (range: 4.2–15.4%) during the 6-month intervention. Participants also experienced significant reductions in BMI, waist circumference, and total and truncal adiposity after the hypocaloric dietary intervention (All: P $<$ 0.05). Forty-two percent of participants (N = 5) experienced a shift in BMI category from ‘obese’ to ‘overweight’ and 33% (N = 4) shifted from ‘Class 2’ obesity to ‘Class 1’ obesity; 25% (N = 3) did not show a change in designation. During baseline, participants displayed a 29 ± 4-day IOI with ovulation occurring on Day 29. This translated to a 17 ± 3-day follicular phase and a 12 ± 2-day luteal phase. By contrast, during Month 7, participants displayed a 27 ± 4-day IOI with ovulation occurring on Day 27 (14 ± 3-day follicular phase and 13 ± 4-day luteal phase), consistent with a 2-day shorter cycle during the Month 7 IOI. No changes were observed in the other reproductive markers post-intervention (Table 1). However, participants did experience improvements in markers of glucose tolerance and insulin resistance, including reduced fasting insulin and HOMA-IR levels (Table 1).

Table 1.

Characteristics of the study participants.

	Inter-ovulatory interval		Reference cohort (N = 21)	Baseline versus Month 7 P-value	Month 7 versus reference P-value
	Baseline (N = 12)	Month 7 (N = 12)
Reproductive markers
Menstrual cycle length (days)	29 ± 4	27 ± 4	29 ± 3	0.0293	0.1637
Follicular phase length (days)	16 ± 4	14 ± 3	17 ± 3	0.0409	0.1749
Luteal phase length (days)	12 ± 2	13 ± 4	13 ± 2	0.6459	0.9615
Hirsutism score	4 ± 4	4 ± 4	4 ± 5	1.0000	0.1477
Total testosterone (ng/dl)	14.7 ± 7.7	20.3 ± 9.4	21.9 ± 12.7	0.0706	0.6837
Free androgen index	1.50 ± 1.10	1.58 ± 1.00	1.28 ± 0.74	0.5994	0.2235
LH: FSH	0.71 ± 0.24	0.78 ± 0.31	0.73 ± 0.32	0.3504	0.7013
Anti-Müllerian hormone (pg/ml)	4.86 ± 3.22	5.52 ± 3.94	5.94 ± 2.49	0.2989	0.7449
Luteal progesterone (pg/ml)	9 ± 4	11 ± 5	13 ± 6	0.0586	0.4158
Average antral follicle count	48 ± 19	45 ± 15	52 ± 21	0.1773	0.2708
Anthropometric markers
Weight (kg)	96.7 ± 18.4	87.2 ± 15.1	62.7 ± 11.2	0.0005	<0.0001
BMI (kg/m2)	35.3 ± 5.8	31.7 ± 4.7	22.9 ± 3.2	<0.0001	<0.0001
Percent total fat (%)	45.2 ± 4.2	43.2 ± 4.5	27.5 ± 3.7	0.0183	<0.0001
Truncal fat percentage (%)	45.1 ± 4.2	42.6 ± 4.1	23.8 ± 4.7	0.0061	<0.0001
Waist circumference (cm)	104 ± 15	95 ± 8	79 ± 8	0.0034	<0.0001
Waist:hips ratio	0.84 ± 0.06	0.82 ± 0.04	0.80 ± 0.05	0.1259	0.4197
Metabolic markers
Systolic pressure (mmHg)	112 ± 9	118 ± 19	111 ± 10	0.3825	0.2154
Diastolic pressure (mmHg)	71 ± 10	78 ± 22	68 ± 7	0.3809	0.1605
Fasting glucose (mg/dl)	89.8 ± 6.1	92.4 ± 5.5	93.6 ± 12.2	0.2184	0.9089
Fasting insulin (mIU/l)	8.32 ± 3.89	6.00 ± 3.49	4.29 ± 2.22	0.0277	0.1522
HOMA-IR	1.81 ± 0.82	1.36 ± 0.79	1.00 ± 0.56	0.0429	0.1849

Data are presented as mean ± SD.

Abbreviation: HOMA-IR, homeostatic model assessment of insulin resistance.

AFC across an IOI

Mean profiles of AFC ≥2 mm (Fig. 1A), AFC 2–5 mm (Fig. 1B), AFC 6–9 mm (Fig. 1C), and AFC ≥10 mm (Fig. 1D) are shown for participants at baseline and in Month 7 in Fig. 1. As expected, an effect of day was noted across the IOI for all follicle populations (AFC ≥2 mm, P_day = 0.0014; AFC 2–5 mm, P_day = 0.0002; AFC 6–9 mm, P_day < 0.0001 and AFC ≥10 mm P_day < 0.0001). Total AFC ≥2 mm was lower during Month 7 IOI compared to the baseline IOI (Month 7 versus baseline IOI: 44 ± 3 versus 48 ± 3 follicles; P_intervention = 0.002), which was driven by fewer 2- to 5-mm follicles (40 ± 4 versus 44 ± 3 follicles; P_intervention = 0.001). On average, an increased number of anovulatory dominant follicles was also detected (≥10 mm; 0.4 ± 0.5 versus 0.3 ± 0.4 follicles; P = 0.001) across the IOI. There were no differences in the 6- to 9-mm follicle population across the IOI in Month 7 (P_intervention = 0.816). Further, no day by intervention effect was noted for any follicle size category (AFC ≥2 mm, P_{day×intervention} = 0.9948; AFC 2–5 mm, P_{day×intervention} = 0.9985; AFC 6–9 mm, P_{day×intervention} = 0.7403 and AFC ≥10 mm P_{day×intervention} = 0.4692).

Figure 1.

Longitudinal profiles of antral follicle counts ≥2 mm (A), 2–5 mm (B), 6–9 mm (C) and ≥10 mm (D) across an inter-ovulatory interval (IOI) in women before (●) and after a hypocaloric dietary intervention (○) (mean ± SD). Day-to-day changes in antral follicle counts per diameter category were monitored using the Non-Identity Method. Mixed models showed a day effect for ≥2, 2- to 5-, 6- to 9-, and ≥10-mm follicles, and an intervention effect for ≥2-, 2- to 5-, and ≥10-mm follicles.

Reproductive hormones during an IOI

Mean profiles of reproductive hormones in baseline and Month 7 IOI are shown in Fig. 2. There were no changes in LH or E2 during Month 7 IOI compared to the baseline IOI (Fig. 2A and C, respectively; both, P_intervention > 0.05). Changes in FSH concentrations across the IOI differed in the Month 7 IOI compared to the baseline IOI, as shown in Fig. 2B (P_{day×intervention} = 0.009). Overall, P4 was higher in the Month 7 IOI compared to the baseline IOI (Fig. 2D; P_intervention = 0.013).

Figure 2.

Longitudinal profiles of LH (A), FSH (B), estradiol (C) and progesterone (D) across an inter-ovulatory interval (IOI) in women before (●) and after a hypocaloric dietary intervention (○) (mean ± SD). Day-to-day changes in hormone concentrations were monitored by serial venipuncture. Mixed models showed a day effect for LH, FSH, estradiol (E2) and progesterone (P4). An intervention effect was noted for P4 and a day by intervention effect noted FSH.

Follicle diameter and hormones by menstrual cycle phase

Mean follicle populations and hormone concentrations are presented for the follicular and luteal phases of the menstrual cycle in Table 2. In the follicular phase of the Month 7 IOI, AFC ≥2 mm, 2- to 5-mm follicle counts, and the proportion of 6- to 9-mm follicles were decreased relative to total follicle count, while the proportion of 2- to 5-mm follicles and the growth of dominant follicles were increased compared to the follicular phase of the baseline IOI (all; P < 0.05). P4 levels were significantly increased across the luteal phase in the participants after the intervention (P < 0.0001). By contrast, the number of selectable follicles, or concentrations of LH, FSH, E2, and P4 did not differ at baseline versus Month 7 IOI during the follicular phase (all; P $\ge$ 0.05). There were also no differences in luteal phase AFC ≥2 mm, recruitable follicles, selectable follicles, follicle size proportions, or LH, FSH, or E2 levels (all; P > 0.05) (Table 2).

Table 2.

Impact of day and intervention status on follicle populations and endocrine profiles during the follicular and luteal phases.

	Inter-ovulatory interval		Day effect Pa	Intervention effect Pa
	Baseline (n = 12)	Month 7 (n = 12)
Follicular phase
Antral follicle count (AFC)	49 ± 22	45 ± 16	0.014	0.003
AFC 2–5 mm	44 ± 22	40 ± 13	0.019	0.011
AFC 6–9 mm	4 ± 3	4 ± 4	0.020	0.166
Proportion 2–5 mm (%)	87.8 ± 9.2	89.7 ± 6.2	0.010	0.011
Proportion 6–9 mm (%)	10.7 ± 8.7	8.2 ± 6.2	0.071	0.005
Proportion >10 mm (%)	1.5 ± 2.1	2.3 ± 2.3	0.0001	0.014
Mean LH (mIU/ml)	9.54 ± 10.69	11.38 ± 16.47	0.0001	0.492
Mean FSH (mIU/ml)	6.17 ± 2.56	6.51 ± 3.36	0.029	0.358
Mean estradiol (pg/ml)	109.38 ± 97.60	114.63 ± 92.25	0.0001	0.096
Mean progesterone (pg/ml)	0.47 ± 0.38	0.75 ± 1.36	0.0001	0.299
Luteal phase
Antral follicle count	45 ± 16	44 ± 17	0.569	0.285
AFC 2–5 mm	43 ± 16	41 ± 16	0.097	0.186
AFC 6–9 mm	2 ± 2	3 ± 3	0.132	0.581
Proportion 2–5 mm (%)	94.0 ± 5.3	93.8 ± 5.8	0.0002	0.776
Proportion 6–9 mm (%)	5.9 ± 5.3	6.0 ± 5.5	0.0004	0.943
Proportion >10 mm (%)	0.1 ± 0.3	0.2 ± 0.9	0.698	0.131
Mean LH (mIU/ml)	6.65 ± 4.62	6.17 ± 5.07	0.0001	0.920
Mean FSH (mIU/ml)	3.83 ± 1.85	4.16 ± 2.23	0.0001	0.380
Mean estradiol (pg/ml)	114.03 ± 56.24	103.34 ± 64.34	0.0001	0.918
Mean progesterone (pg/ml)	5.29 ± 3.65	6.27 ± 4.74	0.0001	0.0001

Data are presented as mean ± SD.

^aMixed model results for day effect and intervention effect are shown.

There were no significant differences in the prevalence of LPDs pre- versus post-intervention (all; P > 0.05). Namely, 58% of participants (N = 7) met the criteria for LPDs based on integrated luteal P4 at baseline compared to 50% that met the criteria in the Month 7 IOI (N = 6). Based on peak luteal P4, 50% (N = 6) versus 67% (N = 8) of participants met criteria for LPD at the baseline (N = 6) and Month 7 IOIs (N = 8), respectively. Likewise, 25% of participants in the baseline IOI (N = 3) and 33% of participants in the Month 7 IOI (N = 4) displayed LPDs using criteria for shortened luteal phase length. At the individual level, only two participants no longer met the criteria for LPD based on integrated P4 following the intervention. Nine participants did not change designation, and one shifted to having an LPD that was not present at baseline. Similarly, based on peak luteal P4, one participant showed an improvement, eight did not change, and three showed a shift to having an LPD after the intervention. Lastly, based on luteal phase length, three participants no longer met the criteria for LPD, five did not change, and four shifted to meeting criteria for LPD after the intervention.

Recruitment, selection, and ovulation

Table 3 summarizes the number and distribution of recruitment events experienced by the participants before and after the intervention. During the baseline IOI, 6 of 12 participants demonstrated one recruitment event and 5 of 12 participants exhibited two recruitment events. In the Month 7 IOI, 6 of 12 participants experienced two recruitment events and 2 of 12 participants displayed three recruitment events. Ultimately, this shift in the number of recruitment events from the baseline IOI to the Month 7 IOI did not reach statistical significance (P = 0.068) (Table 3).

Table 3.

Recruitment events during an inter-ovulatory interval (IOI) at baseline, Month 7, and for a reference cohort during natural cycles.

	Baseline IOI (n = 12)	Month 7 IOI (n = 12)	Reference cohort (n = 21)	Baseline versus Month 7 P-value	Month 7 versus reference P-value
Recruitment
Number of recruitment events	1 ± 0.7	2 ± 0.9	2 ± 0.9	0.0683	0.4484
Distribution of events (N, %)
0	1/12 (8%)	1/12 (8%)	1/21 (5%)
1	6/12 (50%)	3/12 (25%)	2/21 (10%)
2	5/12 (42%)	6/12 (50%)	9/21 (43%)
3	0/12 (0%)	2/12 (17%)	9/21 (43%)

Data are presented as mean ± SD or proportion (%).

The characteristics of dominant follicles before and after the intervention are presented in Table 4. In both the baseline and Month 7 IOIs, 42% of participants experienced anovulatory follicles and the incidence rates of anovulatory follicles did not differ. Of those follicles that progressed from the selectable pool to dominance, there were no differences in the maximum diameter achieved or the diameter of anovulatory follicles at selection before or after the intervention (P > 0.05). However, ovulatory follicles in the Month 7 IOI were selected at a significantly larger diameter than those in the baseline IOI (P = 0.007) (Table 4), although the day of selection did not differ after the intervention (P = 0.281). By design, all of the study participants experienced ovulatory dominant follicles, and there was no difference in the maximum diameter of the ovulatory follicles in the baseline versus the Month 7 IOIs (P = 0.134) (Table 4).

Table 4.

Follicle kinetics of dominant follicles during an inter-ovulatory interval (IOI) at baseline, Month 7, and for a reference cohort during natural cycles.

	Baseline IOI (n = 12)	Month 7 IOI (n = 12)	Reference cohort (n = 21)	Baseline versus Month 7 P-value	Month 7 versus reference P-value
Characteristics of anovulatory dominant follicles
Total number over the IOI (N)	7	8	18	0.7227	0.6558
Prevalence (% of participants)	5/12 (47%)	5/12 (42%)	12/21 (57%)	1.0000	0.6558
Prevalence in the Follicular phase (N participants, %)	4/12 (33%)	5/12 (42%)	10/21 (48%)	0.1039	0.5058
Prevalence in the luteal phase (N participants, %)	2/12 (17%)	0/12 (0%)	5/21 (24%)	0.1661	0.0212
Maximum diameter (mm)	10.5 ± 0.8	11.6 ± 1.7	10.7 ± 1.0	0.1747	0.1856
Characteristics of ovulatory dominant follicles
Total number over the IOI (N)	12	12	22a	1.0000	1.0000
Prevalence (% of participants)	12/12 (100%)	12/12 (100%)	21/21 (100%)	1.0000	1.0000
Emergence to ovulation
Growth phase (days)	14.8 ± 2.6	14.8 ± 4.1	15.4 ± 3.1	0.8552	0.5524
Growth rate (mm/day)	1.02 ± 0.23	1.18 ± 0.32	1.03 ± 0.22	0.2131	0.3631
Selection to ovulation
Diameter at selection (mm)	7.3 ± 2.0	10.9 ± 2.6	9.5 ± 1.9	0.0073	0.1251
Day of selection (day)	21.1 ± 2.9	20.6 ± 3.1	21.0 ± 3.9	0.2813	0.8592
Growth phase (days)	8.8 ± 3.2	7.5 ± 2.2	8.9 ± 1.9	0.1007	0.0814
Growth rate (mm/day)	1.17 ± 0.36	1.28 ± 0.34	1.22 ± 0.27	0.2955	0.5000
Maximum diameter of ovulatory dominant follicles (mm)	19.1 ± 2.2	20.3 ± 3.3	19.8 ± 2.9	0.1336	0.6748
Day of ovulation (days)	29 ± 4	27 ± 4	29 ± 3	0.0293	0.1637

Data are presented as mean ± SD or proportion (%).

^aOne participant ovulated two follicles.

Follicle kinetics

Complete growth and regression profiles were available for 73 uniquely identifiable follicles in the baseline IOI and 59 follicles in the Month 7 IOI. Of the uniquely identified follicles, 19 follicles and 20 follicles progressed to dominance in the baseline IOI and the Month 7 IOI periods, respectively. The kinetics of the anovulatory dominant follicles did not differ before versus after the intervention. Full growth profiles of ovulatory follicles were available for 100% of participants before and after the intervention and are summarized in Table 4. The ovulatory follicles during the baseline and Month 7 IOIs had similar growth phases and growth rates from emergence to ovulation and from selection to ovulation (all: P > 0.05) (Table 4).

Comparison with a non-obese reference cohort

Month 7 IOI data were compared to an age-matched (29 ± 6 years versus Month 7: 31 ± 3 P = 0.2986) reference cohort of participants with regular menstrual cycles but without evidence of obesity (N = 21) (Oldfield et al., 2023). As shown in Table 1, characteristics during the Month 7 IOI did not differ from those of the reference cohort for any reproductive marker. By contrast, the groups differed in weight, BMI, percent total fat, trunk fat percentage, and waist circumference, as expected (all; P < 0.001). After the intervention, the metabolic markers during the Month 7 IOI were comparable to those of the reference cohort, including blood pressure, fasting glucose, fasting insulin, and HOMA-IR levels (all; P > 0.05). Additionally, all elements of antral follicle development (i.e. number of recruitment events, diameter at selection, or number of dominant follicles) appeared similar when comparing participants evaluated during the Month 7 IOI to the reference cohort and no statistically significant differences were detected (all; P > 0.05) (Tables 3 and 4).

Discussion

This study represents the first comparison of ovarian antral follicle dynamics in women with obesity and regular menstrual cycles before and during the final month of a 6-month hypocaloric dietary intervention, in which weight loss approximated 10.6%. Our findings are consistent with improved antral follicle development with weight loss as evidenced by an increased number of dominant follicles across the IOI, a larger follicle diameter at selection, and increased post-ovulatory P4 production. While luteal P4 concentrations were increased during the final month of the intervention, this increase did not result in any improvement in the prevalence of meeting criteria for risk of LPD. Follicle dynamics post-weight loss were similar to that demonstrated in a reference cohort of lean women with regular ovulatory cycles. However, P4 concentrations in women with obesity remained lower than that reported in their lean counterparts. Together, these findings suggest that modest weight loss induced by a 6-month hypocaloric dietary intervention may be sufficient to normalize antral follicle dynamics but not luteal function.

In the present study, we used serial transvaginal ultrasonography to characterize antral follicle dynamics. This approach allowed for a more comprehensive evaluation of ovarian function after a hypocaloric dietary intervention and confirmed improvements in folliculogenesis that could only be speculated by previous studies involving endocrine and metabolic assessments in a similar study population (Grenman et al., 1986; Turcato et al., 1997; Pasquali et al., 2000; Moran et al., 2007; Panidis et al., 2008; Rochester et al., 2009; Nikokavoura et al., 2015; Al-Eisa et al., 2017; MacKintosh et al., 2019). In Month 7, the total number of antral follicles across the IOI declined by 7% relative to baseline, which was driven by fewer 2- to 5-mm follicles. We posit that the decline in small antral follicles is attributed to fewer follicles transitioning into the recruitable pool. Improvements in glucoregulation post-weight loss could underlie the changes in follicle populations noted in our study. Insulin has been implicated in conditions of follicle excess such as polycystic ovary syndrome (PCOS). Insulin can act as a co-gonadotropin with LH on theca cells to produce more local androgens which are known to be necessary for follicle activation (Gervásio et al., 2014; Franks and Hardy, 2018) and could conceivably explain the larger recruitable follicle pool noted in those with obesity. It is possible that the improvements in fasting insulin and HOMA-IR noted after the dietary intervention in our study dampened the paracrine actions of ovarian androgens leading to fewer total and 2- to 5-mm AFC across the IOI. However, it is important to note that follicular excess persisted post-intervention consistent with hypotheses of obesity-promoting manifestations of a PCOS-like phenotype of metabolic origins.

We noted a trend toward increased recruitment events after the hypocaloric dietary intervention, with the number of recruitment events during the Month 7 IOI being comparable to that of the reference cohort. A transition toward an increasing number of recruitment events may suggest improved fertility potential after weight loss, as a higher number of follicle waves (or recruitment events) has been shown to reflect fertility potential in bovine models (Ahmad et al., 1997; Townson et al., 2002). Follicle waves are known to be FSH dependent (Baerwald et al., 2004). We showed that changes in FSH across the IOI differed by intervention status, most notably with participants exhibiting increased pre-ovulatory FSH production during the Month 7 IOI. After the intervention, participants also demonstrated a clear increase in FSH levels on the day of emergence of the ovulatory follicle cohort. This was in contrast to the baseline IOI, in which the increase in FSH occurred earlier (baseline IOI versus Month 7 IOI: 11 ± 1 versus 13 ± 1 days, respectively; P = 0.045) and did not coincide with the day of emergence of the ovulatory follicle (baseline IOI versus Month 7 IOI: 15 ± 1 versus 13 ± 1 days, respectively; P = 0.039). This lag between the rise in FSH and emergence of the ovulatory follicle during the baseline IOI could reflect a reduced receptivity of medium-sized antral follicles to FSH. FSH receptor expression can be regulated by a variety of factors including activins and inhibins whose concentrations are known to be impacted by obesity and metabolic status (Eldar-Geva et al., 2001; De Pergola et al., 2006; Zaragosi et al., 2010; Kuo et al., 2018). Similar to other studies (Grenman et al., 1986; Turcato et al., 1997; Pasquali et al., 2000; Rochester et al., 2009), we did not notice a global increase of FSH concentrations after weight loss. Rather, our data are more consistent with better alignment between FSH and follicle development post-weight loss which would not have been captured by earlier studies that did not employ more frequent sampling.

Follicle selection occurred at larger diameters in Month 7 IOI compared to baseline, and the increased diameter at selection was comparable to that of the reference cohort. We previously showed that the follicles of participants with obesity were selected at a smaller diameter compared to non-obese participants (Oldfield et al., 2023), which we posited might reflect earlier acquisition of LH receptors and an untimely transition to LH-dependent growth versus those without obesity (Zeleznik, 2004). In anovulatory conditions associated with obesity, premature acquisition of LH receptors in antral follicles is posited to occur owing to increased insulin signaling in both granulosa and theca cells (Willis et al., 1998; Poretsky et al., 1999). Improvement in the metabolic state of participants with weight loss may have been sufficient to alter the size at selection (7.3 mm in baseline IOI versus 10.9 mm in Month 7 IOI) suggesting insulin may be the main factor regulating premature follicle selection in those with obesity. After the intervention, participants had 28% lower concentrations of fasting insulin and demonstrated an improvement in HOMA-IR levels, with both levels comparable to that of the reference cohort. Together, our data provide evidence that metabolic factors may play a substantial role in follicle selection and LH receptor acquisition even in the context of eumenorrhea.

By design, all of the participants in our study exhibited at least one instance of dominance during each IOI as per the manifestation of a dominant ovulatory follicle. However, multiple dominant follicles can emerge during an IOI (Baerwald et al., 2003, 2004). Across the Month 7 IOI, participants exhibited an increased number of anovulatory dominant follicles, specifically in the follicular phase. This propensity for morphologic dominance could be interpreted as a positive impact of weight loss as it suggests that the intraovarian environment was sufficiently favorable to enable transition to the next functional stage of folliculogenesis. However, it is important to note that once a follicle reached dominance, as defined by a diameter ≥10 mm, whether ovulatory or anovulatory, we noted no differences in the growth kinetics or maximum diameters achieved by the dominant follicles during baseline and Month 7 IOIs. Additionally, we noted a shortening of the IOI post-intervention. This decrease in IOI length was associated specifically with a shortening of the follicular phase, with ovulation occurring 2 days earlier compared to baseline. Despite this change in length post-intervention, all IOIs remained within the normative range suggesting that the shorter IOI length may be clinically insignificant. There were no differences in reproductive hormone concentrations before and after the intervention once dominance was achieved (data not shown). Collectively, the emergence of more dominant follicles from the selectable pool, alongside maintenance of a normal IOI length, suggests that the suppression of morphologically dominant follicle development previously reported in eumenorrheic participants with obesity may be reversible with weight loss.

Weight loss has been shown to improve luteal function in eumenorrheic women as judged by increased P4 concentrations, albeit data are controversial. In a study involving bariatric surgery, P4 levels increased by 55% in those that had lost 32% of their initial presurgical body weight (Rochester et al., 2009). By contrast, studies associated with lower levels of weight loss (5.6–15.3%) induced by a diet and/or surgical intervention did not result in increased P4 concentrations (Pasquali et al., 2000; MacKintosh et al., 2019). Our study is consistent with the conclusion that modest reductions in body weight (10.6%) can result in increases in P4. Further, we previously reported an increased prevalence of LPDs in participants with obesity and regular cycles compared to their non-obese counterparts (Oldfield et al., 2023). In the current study, there were no differences in the prevalence of LPDs before and after the intervention and prevalence of LPDs in Month 7 IOI remained higher compared to the reference cohort. Therefore, despite the increase in luteal P4 production after the intervention, the degree of weight loss induced by the 6-month hypocaloric intervention used in this study was not sufficient to reduce the portion of participants meeting criteria for LPD, suggesting more weight loss may be needed. That said, we appreciate that the diagnosis of LPD is largely clinical and at best, our data reflect risk for LPD and not necessarily a confirmed diagnosis.

This prospective study had several strengths. We included a well-characterized, otherwise healthy study population, recruited from the general population. Participants were at a stage of weight loss readiness and had high likelihood of retention in a dietary intervention study. The prescribed 6-month hypocaloric dietary intervention was efficacious, with the vast majority of participants transitioning to a more favorable BMI class. In general, 5% is considered a clinically meaningful reduction in weight and has been shown to be associated with improvements in metabolic status (Williamson et al., 2015), as well as improved pregnancy outcomes (Stang and Huffman, 2016; Practice Committee of the American College of Obstetricians and Gynecologists, 2019). This minimum amount of weight loss was achieved by all but one participant who showed only a 4.2% peak reduction in weight. To eliminate factors that could confound follicle dynamics in the context of obesity, we excluded for hyperandrogenism which is known to alter follicle dynamics (Jarrett et al., 2020). Our use of serum testosterone levels to confirm the normoandrogenic status of the participants was important as clinical markers, such as hirsutism, have not been shown to be a consistent proxies for current androgen levels (Ewing and Rouse, 1978; Legro et al., 2010; Vanden Brink et al., 2016). This study also had limitations. The sample size is relatively small with the intensive nature of the data collection and analyses impacting the feasibility of conducting these types of studies. Not all women with obesity can be assumed to have ovulatory problems or defects in folliculogenesis and larger trials are needed to corroborate our pilot findings. Also, with 100% of the participants identifying as Caucasian and 92% identifying as not Hispanic or Latino, our findings cannot be generalized to other populations, as both obesity rates (Petersen et al., 2019) and ovarian reserve (Bleil et al., 2014) are known to differ by race and ethnicity. Further, the intervention employed in this study should be considered short-term. We are unable to comment on whether any improvements in follicle and endocrine dynamics noted with weight loss are sustained in the long term.

In summary, hypocaloric dietary intervention improved antral follicle dynamics even in the context of regular ovulatory cycles. Improvements in antral follicle development aligned with favorable changes in metabolic status and endocrine profile, including post-ovulatory P4 production. These data, alongside our findings that obesity suppresses folliculogenesis in women with regular ovulatory cycles, support that weight loss can have an immediate and positive impact on antral follicle development in women with obesity. Additional research is needed to determine whether normalization of antral follicle dynamics in response to weight loss persists and whether full resolution of luteal function can be achieved with sustained lifestyle intervention.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Author’s roles

Participation in study design (A.L.O., B.Y.J., and M.E.L.), data collection (H.V.B., M.E.L., A.L.O., F.E.C., and B.Y.J.), image analysis (A.L.O., F.E.C., and R.E.R.), data analysis and interpretation (A.L.O., R.E.R., and M.E.L.), and manuscript drafting and critical discussion (A.L.O., F.E.C., B.Y.J., R.E.R., H.V.B., and M.E.L.).

Funding

Cornell University, President’s Council of Cornell Participants, United States Department of Agriculture (Grant No. 8106), and National Institutes of Health (R01-HD0937848). B.Y.J. and H.V.B. were supported by doctoral training awards from the National Institutes of Health (T32-DK007158) and Canadian Institutes of Health Research (Grant No. 146182), respectively.

Conflict of interest

The authors report no conflict of interest.