Effects of combined metformin and semaglutide therapy on body weight, metabolic parameters, and reproductive outcomes in overweight/obese women with polycystic ovary syndrome: a prospective, randomized, controlled, open-label clinical trial

Abstract

Objective

This study aimed to assess the effects of metformin (MET) monotherapy versus a combination of semaglutide and MET on weight, metabolism, reproductive function, and inflammatory markers in women with polycystic ovary syndrome (PCOS).

Methods

A total of 100 overweight or obese women with PCOS diagnosed according to the Rotterdam criteria were randomly assigned to two groups: MET (1000 mg twice daily [BID] for 16 weeks) and combination therapy (COM) (1000  mg MET BID plus 1- mg semaglutide once weekly [QW] for 16 weeks). Primary outcomes, assessed at Week 0 and Week 16, included changes in anthropometric measures related to obesity, while the secondary outcomes included alterations in reproductive hormone levels, glucose and lipid metabolism, and C-reactive protein (CRP) levels, Between Weeks 16 and 40, all participants received metformin monotherapy (1000 mg BID) to evaluate pregnancy outcomes.

Result

A total of 80 participants (80%) completed the study. After 16 weeks of intervention, the COM group exhibited significantly greater reductions in body weight, BMI, and waist-to-hip ratio (WHR) compared to the MET group (all P < 0.01). The COM group experienced an average weight loss of 6.09 ± 3.34 kg, while the MET group lost only 2.25 ± 4.27 kg. The COM group also demonstrated greater improvements in testosterone (TEST), Chinese visceral adiposity index (CVAI), and CRP levels compared to the MET group. Additionally, the COM group showed higher rates of menstrual cycle recovery than the MET group. From weeks 16 to 40, the COM group demonstrated a significantly higher natural pregnancy rate than the MET group (35% vs. 15%, P < 0.05).

Conclusion

Compared to MET monotherapy, combination therapy with semaglutide and MET significantly reduced body weight, improved insulin resistance, decreased inflammatory markers, alleviated and menstrual irregularities and increased natural pregnancy rates in overweight/obese women with PCOS.

Clinical trial registration

chictr.org.cn ID: ChiCTR2400090908 (Registration time: 2024-10-15).

Introduction

Polycystic ovary syndrome (PCOS) is the most prevalent endocrine disorder among women of reproductive age, affecting their health from adolescence through menopause, with a prevalence of 11–13% [1, 2]. Characterized by hyperandrogenism, chronic oligo-ovulation, and polycystic ovarian morphology (PCOM), PCOS is frequently associated with metabolic abnormalities such as obesity, insulin resistance, and dyslipidemia. These conditions increase the risk of type 2 diabetes mellitus, cardiovascular diseases, and endometrial cancer. Additionally, ovulatory dysfunction in PCOS is a major contributor to female infertility. Obesity affects approximately 20–80% of women with PCOS, worsening metabolic and reproductive complications, including insulin resistance, menstrual irregularities, infertility, and cardiovascular issues [3, 4].

A weight loss of 5–10% can significantly improve the pathological features and long-term outcomes for patients with PCOS [5–7]. However, first-line lifestyle interventions for obesity management in these patients often lead to unsatisfactory and unsustainable results in clinical practice [8]. While metformin (MET) has demonstrated benefits in improving insulin resistance, promoting ovulation, and enhancing reproductive outcomes in patients with PCOS, its effectiveness in reducing weight among obese individuals remains limited [9, 10].

In recent years, glucagon-like peptide-1 receptor agonists (GLP-1 RAs) have emerged as a promising treatment for obese patients with PCOS [11–13]. GLP-1 RAs not only facilitate body weight loss but also address mechanisms associated with insulin resistance. For example, they enhance the expression of glucose transporters in insulin-sensitive tissues, reduce inflammation and oxidative stress, and regulate lipid metabolism. Additionally, GLP-1 RAs may improve fertility through two potential pathways: (1) by mitigating hypothalamic–pituitary suppression caused by estrogen excess linked to obesity, thereby increasing luteinizing hormone (LH) peaks; and (2) by lowering elevated LH levels associated with hyperinsulinemia and obesity [14]. The combination of MET with GLP-1 RAs, such as liraglutide or exenatide, offers several benefits in treating obese patients with PCOS. These include significant reductions in body weight and fat, improved insulin resistance, enhanced menstrual cycle regularity and ovulation rates, reduced inflammation and oxidative stress, and increased treatment efficacy, particularly in patients with limited responses to lifestyle interventions combined with MET therapy [11–13, 15]. Weekly administration of the GLP-1 RA semaglutide has been shown to consistently reduce weight in obese patients with or without type 2 diabetes mellitus. Compared to liraglutide, semaglutide offers superior weight reduction efficacy and a lower risk of gastrointestinal side effects, resulting in higher patient adherence rates [16]. A recent small-scale clinical trial demonstrated that low-dose semaglutide reduced weight, waist circumference (WC), and hip circumference (HC) in patients with PCOS [17]. However, the study did not examine the impact of low-dose semaglutide on other metabolic parameters or reproductive outcomes. To date, research on the therapeutic effects of semaglutide in PCOS has been limited studies, and studies investigating the combined effects of semaglutide and MET in obese patients with PCOS are lacking. Thus, this study aimed to evaluate the impact of combined semaglutide and MET therapy on metabolic, endocrine, reproductive disorders and inflammatory markers in overweight/obese women with PCOS.

Materials and methods

Population

A total of 100 overweight/obese women with PCOS (body mass index [BMI] ≥ 24 kg/m²) diagnosed according to the Rotterdam criteria, along with 40 healthy individuals were enrolled in this study. This single-center study, conducted at the Second Affiliated Hospital of Chongqing Medical University, adhered to the guidelines set forth in the Declaration of Helsinki and received approval from the hospital’s ethics committee (Approval No.: 2024204). All participants provided informed consent prior to enrollment. PCOS was diagnosed if at least two of the following Rotterdam criteria (2003) were met: (1) oligo-anovulation and/or anovulation, (2) clinical and/or biochemical signs of hyperandrogenism, and (3) polycystic ovaries. The inclusion criteria were as follows: (1) age between 18 and 40 years; (2) for this trial, which exclusively enrolled Chinese participants overweight was defined as BMI between 24 and 28 kg/m² and obesity as BMI ≥ 28 kg/ m² [18]; (3) fulfillment of the PCOS diagnostic criteria and absence of conditions such as congenital adrenal hyperplasia, Cushing’s syndrome, and androgen-secreting tumors; (4) lack of response to lifestyle interventions (diet and exercise) for 3 months prior to enrollment; and (5) voluntary participation with signed informed consent. Exclusion criteria were as follows: (1) allergy to GLP-1 RAs or MET; (2) pre-existing diabetes (type 1 or 2); (3) comorbid conditions such as such as congenital adrenal hyperplasia, Cushing’s syndrome, or androgen-secreting tumors; (4) abnormal liver or kidney function (alanine transaminase [ALT] > 2.5 times the upper limit of normal, estimated glomerular filtration rate [eGFR] < 60 mL/min/1.73 m²); (5) use of glucocorticoids, estrogen, progesterone, insulin, or other drugs affecting reproduction within 3 months prior to enrollment; (6) other gynecological conditions causing severe infertility or severe male infertility in the partner; (7) inability to comply with follow-up; and (8)patients planning a pregnancy within 16 weeks of study initiation or currently pregnant.

Study design

This randomized, prospective, open-label, parallel, controlled study was conducted over a 16-week period. After providing informed consent, all participants underwent baseline assessments, including measurements age, height, weight, menstrual cycle, liver and kidney function, prolactin levels, testosterone (TEST) levels, human chorionic gonadotrophin (hCG) levels and thyroid ultrasonography to exclude individuals with elevated liver enzymes, impaired renal function, hyperprolactinemia, or pregnancy. Based on these results, 16 patients were excluded and 4 were lost to follow-up, leaving 80 patients with PCOS who completed the study. Patients were randomly assigned into two groups (1:1) using computer-generated random numbers: MET monotherapy and combined therapy with MET and semaglutide (COM). MET treatment started at 500 mg daily (QD) and, in the absence of significant effects, was gradually increased by 500 mg QD every 3 days until the dose reached 1000 mg twice daily (BID). Semaglutide was initiated at 0.25 mg and if no or mild adverse effects occurred, the dose was increased to 0.5 mg at 4 weeks and 1 mg at 8 weeks Barrier contraception was recommended throughout the study. At Week 16, semaglutide was discontinued in the COM group, while MET was continued in both groups at a dose of 1000 mg BID for an additional 24 weeks. During this phase, follow-up assessments focused solely on pregnancy outcomes in women with PCOS. Patients were monitored every 4 weeks for medication adherence, safety, and adverse events.

Outcome measures

Anthropometric measurements

At baseline and Week 16, trained medical staff collected anthropometric data, including height, weight, WC, HC, and waist-to-hip ratio (WHR). Participants were instructed to wear light clothing, and height and weight were measured to the nearest 0.01 cm and 0.01 kg, respectively. BMI was calculated as weight (kg) divided by height squared (m²). WC was measured at the midpoint between the iliac crest and the lower rib margin, while HC was measured at the maximum circumference. WHR was calculated by dividing WC (cm) by HC (cm).

Biochemical parameters

After an overnight fast of at least 10 h, all participants underwent a 75-g oral glucose tolerance test (OGTT) at baseline and Week 16. Venous blood samples were collected at 0 (baseline), 30, 60, and 120 min to measure glucose and insulin levels. Fasting blood samples were also collected to measure glycated hemoglobin (HbA1c), sex hormones, lipid profile, uric acid (UA), C-reactive protein (CRP) and anti-Müllerian hormone (AMH) levels. Insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR), calculated as fasting blood glucose (FBG, mmol/L) × fasting insulin (FINS, mU/L) / 22.5. The free androgen index (FAI) was calculated as TEST level (nmol/L) / sex hormone-binding globulin (SHBG) level (nmol/L) × 100. The cardiovascular risk index (CVAI) was calculated as: − 187.32 + 1.71 × age + 4.23 × BMI (kg/ m²) + 1.12 × WC (cm) + 39.76 × log10 triglyceride (TG, mmol/L)– 11.66 × high-density lipoprotein cholesterol (HDL-C, mmol/L).

Plasma glucose levels were measured using glucose oxidase assay, HbA1c levels by high-performance liquid chromatography, and insulin levels via chemiluminescence assay. Total cholesterol (TC), HDL-C, low-density lipoprotein cholesterol (LDL-C), and TG levels were analyzed using an automated analyzer. Serum concentrations of LH, follicle-stimulating hormone (FSH), TEST, prolactin (PRL), SHBG, and dehydroepiandrosterone sulfate (DHEA-S) were measured using chemiluminescence assays.

Plasma concentrations of CRP and AMH were determined using enzyme-linked immunosorbent assay (ELISA) kits. The CRP kit (ml057570, MLBIO, China) had a detection range of 31.25–2000 pg/mL, with a detection limit of 15.63 pg/mL and intra- and inter-assay coefficients of variation (CV) of < 10%. The AMH kit (ml060605, MLBIO, China) had a detection range of 39.06–2500 ng/mL, with a detection limit of 19.53 pg/mL and intra- and inter-assay CVs of < 10%.

Menstrual cycle and pregnancy outcomes

During the 16-week treatment period, the patients were instructed to use barrier contraception and record changes in their menstrual cycle. Menstrual irregularities were defined as oligomenorrhea (fewer than 6 menstrual periods in 12 months) and amenorrhea (cessation of menstruation for more than 6 months). Each bleeding episode was considered one menstrual cycle. Menstrual regularity at baseline was assessed based on self-reported menstrual intervals over the past 3 years, as documented in a diary review. Menstrual cycle recovery was defined as the restoration of a regular menstrual cycle. After 16 weeks of treatment, semaglutide was discontinued in the COM group, while MET at a dose of 1000 mg BID was continued in both groups for an additional 24 weeks. During this phase, patients were advised to have intercourse 2–3 times per week until pregnancy was achieved. All patients visited the hospital every 4 weeks for urinary hCG testing. Pregnancy was defined as the presence of a gestational sac on ultrasonography. The primary focus during the follow-up phase was the natural pregnancy rate. However, participants who did not achieve natural conception were allowed to pursue alternative methods, such as ovulation induction or in vitro fertilization and embryo transfer (IVF-ET). Ovulation induction and IVF-ET were collectively referred to as “other pregnancies”. Total pregnancy rate was defined as the number of pregnancies (both natural and other pregnancies) divided by the total number of participants. The natural pregnancy rate was calculated as the number of natural pregnancies divided by the total number of patients.

Statistical analysis

Data were analyzed using SPSS (V.27.0, SPSS, Chicago, Illinois, USA) and GraphPad (V.10.0) software. The Kolmogorov–Smirnov test was applied to assess the normality of data distribution. Continuous variables were expressed as the mean ± SD or median (interquartile ranges [IQRs]). For within-group comparisons, paired t-test (normal distribution) or the paired Wilcoxon test (non-normal distribution) was used. Between-group comparisons were made using the independent-sample t-test (normal distribution) and the Mann–Whitney U test (non-normal distribution). Categorical variables were presented as frequencies (percentages) and compared using the chi-square test or Fisher’s exact test as appropriate. A P-value of < 0.05 was considered statistically significant.

Results

Baseline characteristics of study participants

Eighty patients with PCOS completed this study (Fig. 1). The baseline clinical and endocrine-metabolic characteristics of the PCOS and healthy control groups are presented in Table 1. Compared to the control group, the PCOS group showed significantly higher levels of BMI, WHR, UA, ALT, TG, TC, HDL, LDL, WBCs, CRP, HbA1c, FBG, FINS, HOMA-IR, DHEA-S, TEST, FAI, PRL, LH/FSH, and AMH, while SHBG levels were lower (all P < 0.05) (Table 1). There were no significant differences between the groups in age, AST, or E2 levels. Baseline characteristics of the study patients are further detailed in Table 2. All included patients had experienced infertility for over two years without using contraception. In the MET group, 8 patients (19.05%) had a history of one live birth, and 2 patients (4.76%) had a history of two or more live births. In the COM group, 7 patients (16.67%) had a history of one live birth, and 2 patients (4.76%) had experienced two or more live births. A small proportion of patients reported a history of miscarriage. In the MET group, 9 patients (21.43%) had experienced one miscarriage, and 3 patients (7.14%) had experienced two or more miscarriages. In the COM group, 11 patients (26.19%) had a history of one miscarriage, and 2 patients (4.76%) had a history of two or more miscarriages.

Fig. 1.

Subject flow chart

Table 1.

Basal clinical and endocrinological parameters in patients with PCOS and control subjects

	Controls(n = 40)	PCOS(n = 84)
Age(years)	29 ± 5	28 ± 4
BMI(kg/m2)	20.17(19.57,21.53)	27.96(26.53,30.11) **
WHR	0.79 ± 0.04	0.90 ± 0.05**
UA(umol/L)	255.65(231.75,287.40)	343.75(317.00,370.58)**
ALT(U/L)	22.0(14.0,31.8)	27.0(20.0,35.0)**
AST(U/L)	21.0(17.5,26.0)	22.5(18.0,27.0)
TG(mmol/L)	1.00(0.65,1.23)	1.75(1.28,2.21,)**
TC(mmol/L)	4.17(3.65,4.70)	4.55(4.14,4.98)*
HDL-C(mmol/L)	1.40(1.09,1.78)	1.18(1.09,1.30) **
LDL-C(mmol/L)	2.40(1.78,2.62)	2.61(2.26,2.98) **
WBCs(×109/L)	5.32(4.56,5.75)	6.67(5.63,7.92)**
CRP(mg/L)	1.40(0.79,2.33)	3.52(2.59,4.73) **
HbA1c(%)	5.2 ± 0.4	5.4 ± 0.5*
FBG(mmol/L)	4.77(4.56,5.08)	5.12(4.79,5.59) **
FINS(mU/L)	7.47(6.03,9.77)	19.66(13.74,26.57) **
HOMA-IR	1.63(1.22,2.11)	4.49(3.09,6.55) **
SHBG(nmol/L)	35.54(22.86,55.54)	22.27(17.54,29.48) **
DHEA-S(µg/dL)	260.20 ± 55.37	291.56 ± 73.25*
E2(pg/mL)	40.76(35.50,57.50)	42.54(33.91,48.39)
TEST(ng/dL)	34.41(19.47,42.19)	78.67(71.00,87.77)**
PRL(ug/L)	11.44(7.13,17.78)	16.10(12.64,19.94)**
FAI	2.91(1.69,4.78)	11.98(9.32,14.87) **
LH/FSH	1.66(1.05,1.90)	2.09(1.47,2.62)**
CVAI	11.17(-3.01,18.60)	75.14(65.45,86.26) **
AMH(ng/mL)	3.18(1.82,4.52)	6.15(4.61,8.02) **

Data are shown as the mean ± standard deviation (SD) or median (interquartile range)

BMI: Body mass index; WHR: Waist-to-hip ratio; UA: uric acid; ALT: alanine aminotransferase; AST: aspartate aminotransferase; TG: triglycerides; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; WBCs: white blood cells; CRP: C reactive protein; HbA1c: glycosylated hemoglobin; FBG: fasting blood glucose; FINS: fasting insulin; HOMA-IR: homeostasis model assessment of insulin resistance. SHBG: Sex hormone binding globulin; DHEA-S: Dehydroepiandrosterone sulfate; E2: Estradiol; TEST: Testosterone; PRL: Prolactin; FAI: Free androgen index; LH: Luteinizing hormone; FSH: Follicle stimulating hormone; CVAI: Chinese visceral adiposity index; AMH: anti-Müllerian hormone;

The bold font indicates statistically significant between the two groups

** P < 0.01 vs. Controls;

* P < 0.05 vs. Controls

Table 2.

Baseline characteristics of the trial participants

Variable	MET (n = 42)	COM(n = 42)	P value
Age(years)	29 ± 4	28 ± 5	NS
Weight(kg)	70.55(65.15,75.70)	71.80(67.55,78.43)	NS
BMI(kg/m2)	27.88(26.54,29.67)	28.19(26.46,30.35)	NS
WHR	0.90 ± 0.04	0.89 ± 0.05	NS
TG(mmol/L)	1.77 ± 0.57	1.79 ± 0.69	NS
TC(mmol/L)	4.57 ± 0.65	4.53 ± 0.97	NS
HDL-C(mmol/L)	1.18 ± 0.17	1.18 ± 0.16	NS
LDL-C(mmol/L)	2.58 ± 0.52	2.68 ± 0.77	NS
CRP(mg/L)	3.37(2.57,4.72)	3.68(2.78,4.75)	NS
HbA1c(%)	5.3 ± 0.5	5.4 ± 0.5	NS
FBG(mmol/L)	4.96(4.79,5.36)	5.33(4.82,5.7)	NS
FINS(mU/L)	15.03(12.57,23.07)	24.35(15.80,31.14)	< 0.05
HOMA-IR	3.26(2.62,6.21)	5.33(3.73,8.00)	< 0.05
SHBG(nmol/L)	24.77(15.06,38.29)	21.11(17.95,26.27)	NS
DHEA-S(µg/dL)	294.61 ± 72.00	288.52 ± 75.21	NS
E2(pg/mL)	42.07(33.77,48.32)	43.11(34.53,48.46)	NS
TEST(ng/dl)	78.07 ± 15.63	79.84 ± 9.86	NS
PRL(ug/L)	16.63(12.99,20.28)	14.26(12.38,19.28)	NS
FAI	11.22(7.15,14.75)	12.87(10.49,14.95)	NS
LH/FSH	1.95(1.47,2.59)	2.14(1.37,2.70)	NS
CVAI	74.84(64.34,82.02)	75.14(64.95,89.29)	NS
AMH(ng/mL)	6.19 ± 2.58	6.72 ± 2.38	NS

Data are shown as the mean ± standard deviation (SD) or median (interquartile range)

BMI: Body mass index; WHR: Waist-to-hip ratio; TG: triglycerides; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; WBCs: white blood cells; CRP: C reactive protein; HbA1c: glycosylated hemoglobin; FBG: fasting blood glucose; FINS: fasting insulin; HOMA-IR: homeostasis model assessment of insulin resistance. SHBG: Sex hormone binding globulin; DHEA-S: Dehydroepiandrosterone sulfate; E2: Estradiol; TEST: Testosterone; PRL: Prolactin; FAI: Free androgen index; LH: Luteinizing hormone; FSH: Follicle stimulating hormone; CVAI: Chinese visceral adiposity index; AMH: anti-Müllerian hormone

Post-Treatment outcome comparisons

Anthropometric changes after treatment

After 16 weeks of treatment, both the COM and MET groups showed significant reductions in body weight and BMI (both P < 0.01) (Table 3). The COM group lost an average of 6.09 ± 3.34 kg, while the MET group lost 2.25 ± 4.27 kg (Fig. 2) (Table 4). Only the COM group showed a significant reduction in WHR (Table 3). Compared to the MET group, the COM group had lower body weight and BMI after treatment (both P < 0.01).

Table 3.

Metabolic and endocrine features of patients at baseline and after 16-week treatment

	Metformin		P	Metformin + Semaglutide		P
	Before treatment	After treatment		Before treatment	After treatment
Weight(kg)	70.55(65.05,75.30)	67.90(64.63,72.88)	< 0.01	73.06 ± 7.45	66.97 ± 7.28	< 0.01
BMI(kg/m2)	27.88(26.49,29.48)	27.25(25.59,28.52)	< 0.01	28.19(26.36,30.25)	25.63(24.28,27.74)	< 0.01
WHR	0.90 ± 0.04	0.90 ± 0.04	NS	0.89 ± 0.05	0.87 ± 0.05	< 0.01
TG(mmol/L)	1.81 ± 0.56	1.76 ± 0.50	NS	1.82 ± 0.69	1.71 ± 0.58	NS
TC(mmol/L)	4.56 ± 0.67	4.42 ± 0.69	NS	4.53 ± 0.99	4.28 ± 0.4	NS
HDL(mmol/L)	1.18 ± 0.17	1.19 ± 0.21	NS	1.19 ± 0.16	1.13 ± 0.23	< 0.05
LDL(mmol/L)	2.58 ± 0.53	2.53 ± 0.45	NS	2.70 ± 0.78	2.43 ± 0.60	< 0.05
CRP(mg/L)	3.55 ± 1.25	3.36 ± 1.33	NS	3.86 (3.26,4.81)	3.30 (2.46,4.14)	< 0.01
HbA1c(%)	5.3 ± 0.5	5.1 ± 0.6	< 0.05	5.5 ± 0.5	5.2 ± 0.4	< 0.01
FBG (mmol/L)	4.93(4.79,5.27)	4.89(4.58,5.16)	NS	5.28 ± 0.59	4.87 ± 0.40	< 0.01
FINS (mU/L)	14.72(12.38,23.32)	10.85(7.36,15.61)	< 0.01	24.79(17.16,31.62)	18.03(9.66,25.23)	< 0.01
HOMA-IR	3.23(2.61,5.91)	2.34(1.62,3.44)	< 0.01	5.54 (3.84,8.16)	3.89 (2.01,6.06)	< 0.01
SHBG(nmol/L)	25.44(15.92,39.00)	35.39(19.61,53.38)	< 0.01	21.11(18.32,25.86)	55.91(30.02,90.85)	< 0.01
DHEA-S(µg/dL)	295.45(253.08,325.85)	267.06(222.20,319.18)	NS	283.35(230.04,338.63)	255.95(221.09,296.53)	< 0.05
E2(pg/mL)	42.07(33.50,48.35)	44.47(35,86.39)	< 0.05	42.60(34.06,48.52)	48.80(35.18,77.69)	< 0.05
TEST(ng/dL)	79.86(64.62,90.79)	71.38(65.64,77.87)	< 0.05	79.74 ± 10.00	64.85 ± 15.67	< 0.01
PRL (ug/L)	16.63(12.98,20.33)	17.86(14.72,21.13)	NS	15.35 ± 4.96	19.49 ± 4.29	< 0.01
FAI	11.10(7.07,13.74)	8.60(5.45,13.82)	< 0.01	12.87(10.69,14.89)	4.29(2.30,8.32)	< 0.01
LH/FSH	1.95(1.49,2.61)	1.84(1.54,2.18)	NS	2.14(1.4,2.8)	1.66(1.28,2.39)	NS
CVAI	74.84(65.74,83.33)	66.90(55.17,81.45)	< 0.01	75.41(67.46,88.93)	58.93(49.79,71.92)	< 0.01
AMH(ng/mL)	5.66(4.33,7.80)	4.26(3.34,6.16)	< 0.01	7.17(4.88,8.18)	4.51(2.52,5.82)	< 0.01

Data are shown as the mean ± standard deviation (SD) or median (interquartile range)

BMI: Body mass index; WHR: Waist-to-hip ratio; TG: triglycerides; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; WBCs: white blood cells; CRP: C reactive protein; HbA1c: glycosylated hemoglobin; FBG: fasting blood glucose; FINS: fasting insulin; HOMA-IR: homeostasis model assessment of insulin resistance. SHBG: Sex hormone binding globulin; DHEA-S: Dehydroepiandrosterone sulfate; E2: Estradiol; TEST: Testosterone; PRL: Prolactin; FAI: Free androgen index; LH: Luteinizing hormone; FSH: Follicle stimulating hormone; CVAI: Chinese visceral adiposity index; AMH: anti-Müllerian hormone

Fig. 2.

Body weight (kg) of women with PCOS at baseline and after 16 weeks of treatment with COM or MET. Data are presented as mean ± SD. *P < 0.05; ** P < 0 0.01

Table 4.

Participants’ changes of parameters and menstrua cycle recovery in each group after 16-week treatments

	MET	COM	P
Weight(kg)	-2.25 ± 4.27	-6.09 ± 3.34	< 0.01
BMI(kg/m2)	-1.28(-2.37,0.88)	-2.38(-3.40,-1.65)	< 0.01
WHR	-0.02(-0.03,0.02)	-0.04(-0.05,0.00)	< 0.01
TG(mmol/L)	-0.05 ± 0.38	-0.11 ± 0.81	NS
TC(mmol/L)	-0.02 ± 0.21	-0.27(-0.66,-0.09)	< 0.01
HDL-C(mmol/L)	-0.02 ± 0.15	-0.06 ± 0.18	< 0.05
LDL-C(mmol/L)	0.00(-0.12,0.09)	-0.35(-0.63,0.15)	< 0.05
CRP(mg/L)	0.03(-0.47,0.24)	-0.49(-1.29,-0.07)	< 0.05
HbA1c(%)	0.2 ± 0.5	-0.3 ± 0.4	NS
FBG(mmol/L)	-0.18 ± 0.51	-0.41 ± 0.54	NS
FINS(mU/L)	-5.66 ± 6.31	-6.68 ± 11.95	NS
HOMA-IR	-1.40 ± 1.51	-2.05 ± 2.88	NS
SHBG(nmol/L)	11.97 ± 26.49	34.83 ± 33.64	< 0.01
DHEA-S(µg/dL)	-15.93 ± 99.34	-20.75 ± 66.23	NS
E2(pg/mL)	9.06(-11.38,43.41)	9.01(-8.44,40.87)	NS
TEST(ng/dL)	-6.47 ± 14.37	-14.88 ± 16.08	< 0.05
PRL(ug/L)	0.75(-3.22,4.70)	3.10(0.11,7.62)	< 0.05
FAI	-2.31(-6.75,0.59)	-6.82(-9.65,-4.97)	< 0.01
LH/FSH	-0.19 ± 1.20	-0.45 ± 1.21	NS
CVAI	-6.00(13.90,1.79)	-14.99(-22.59,-11.63)	< 0.01
AMH(ng/mL)	-1.40 ± 2.54	-2.39 ± 2.01	NS
Menstrua cycle recovery	17(42.3%)	29(72.5%)	< 0.01

Data are shown as the mean ± standard deviation (SD) or median (interquartile range) or frequencies (percentage)

BMI: Body mass index; WHR: Waist-to-hip ratio; TG: triglycerides; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; WBCs: white blood cells; CRP: C reactive protein; HbA1c: glycosylated hemoglobin; FBG: fasting blood glucose; FINS: fasting insulin; HOMA-IR: homeostasis model assessment of insulin resistance. SHBG: Sex hormone binding globulin; DHEA-S: Dehydroepiandrosterone sulfate; E2: Estradiol; TEST: Testosterone; PRL: Prolactin; FAI: Free androgen index; LH: Luteinizing hormone; FSH: Follicle stimulating hormone; CVAI: Chinese visceral adiposity index; AMH: anti-Müllerian hormone

Changes in glucose and lipid metabolism parameters

After 16 weeks of treatment, both the COM and MET groups showed significant reductions.

in FINS, HbA1c, HOMA-IR, and CVAI (all P < 0.05) (Table 3). The COM group demonstrated a more significant reduction in CVAI compared to the MET group, but there were no significant differences in the reductions of FINS, HbA1c, or HOMA-IR between the groups (all P > 0.05) (Table 4). The COM group also showed a significant reduction in FBG (P < 0.01) (Table 3).

After 16 weeks of treatment, only the COM group showed significant reductions in HDL and LDL levels (both P < 0.05) (Table 3). There were no significant changes in TG or TC levels in either treatment group (Table 3).

Changes in sex hormones and AMH after treatment

After 16 weeks of treatment, both the COM and MET groups showed significant reductions in E2, TEST, and FAI, alongside a significant elevation in SHBG (all P < 0.05) (Table 3). Only the COM group showed a significant reduction in DHEA-S and a significant increase in PRL (Table 3). The COM group demonstrated more significant reductions in TEST, and FAI, along with a greater increase in SHBG compared to the MET group (all P < 0.01) (Table 4). AMH levels decreased significantly in both groups (P < 0.01) (Table 3), but there was no significant difference in the extent of AMH reduction between the groups (Fig. 3) (Table 4).

Fig. 3.

AMH(ng/ml) of women with PCOS at baseline and after 16 weeks of treatment with COM or MET. Data are presented as mean ± SD. *P < 0.05; ** P < 0.01

Changes in inflammatory markers after treatment

After 16 weeks of treatment, CRP levels decreased significantly in the COM group (P < 0.01) (Table 3), while the MET group showed no significant reduction.

Menstrual improvement and pregnancy outcomes

All enrolled patients with PCOS exhibited oligomenorrhea at baseline, with no cases of amenorrhea in the study cohort. At week 16, compared to the MET group, the COM group had a higher rate of menstrual cycle recovery (42.3% vs. 72.5%) (P < 0.01) (Table 4). After discontinuing semaglutide, 33 participants became pregnant during the 24-week follow-up period, with 13 (32.5%) in the MET group and 20 (50.0%) in the COM group. There was no significant difference in the total pregnancy rate between the groups (P > 0.05) (Table 5). However, the natural pregnancy rate was significantly higher in the COM group (35% vs. 15%) (P < 0.05) (Table 5).

Table 5.

Pregnancy outcomes of participants after treatments at week 40

Pregnancy outcome	MET(n = 40)	COM(n = 40)	P
Natural pregnancy	6(15.0%)	14(35.0%)	P < 0.05
Other pregnancy	8(20.0%)	3(7.5%)	NS
Total pregnancy	14(35.0%)	17(42.5%)	NS

Data are shown as frequencies (percentage)

Other pregnancy is referred to ovulation induction and IVF-ET

Total pregnancy is referred to natural and other pregnancy

Changes in outcomes based on weight loss response

While a 5–10% reduction in body weight is widely accepted as beneficial for improving clinical symptoms in patients with PCOS [14], greater weight loss may offer further advantages. A clinical study indicated that women with PCOS treated with low-dose COM who achieved over 10% weight loss showed notable improvement in menstrual irregularities [17]. Moreover, a post hoc analysis of the LOOK AHEAD trial revealed that a weight loss of at least 10% was linked to a reduction in cardiovascular events among individuals with obesity [19]. Participants were divided into three groups based on weight loss: high responders (> 10% weight loss), low responders (5–10% weight loss) relative to a high response, and non-responders (< 5% weight loss). High responders had lower BMI, WHR, FBG, FINS, HOMA-IR, TEST and FAI and higher SHBG levels compared to non-responders. Additionally, high responders showed better improvements in menstrual regularity and pregnancy rates compared to non-responders (P < 0.05) (Table 6). Nevertheless, compared to non-responders, low responders only showed significantly lower levels of BMI, TEST and FAI (P < 0.05) (Table 6). Compared to low responers, high responders exhibited significantly lower BMI, WHR, FBG and HOMA-IR levels, and also showed a significantly higher rate of menstrual cycle recovery (P < 0.05) (Table 6).

Table 6.

Parameter changes (week 16) and pregnancy outcome (week 40) by weight responses post-treatments

	Highly responsive patients with PCOS (Weight Loss ≥ 10%)	Low-responsive patients with PCOS (Weight Loss 5- 10%)	Non-responsive patients with PCOS (Weight Loss < 5%)
Number of patients	25	30	25
ΔBMI(kg/m2)	-3.35 ± 0.44*#	-1.91 ± 0.40*	0.44 ± 0.94
ΔWHR	-0.03 ± 0.05*#	-0.01 ± 0.04	-0.01 ± 0.03
ΔFBG (mmol/L)	-0.55 ± 0.60*#	-0.19 ± 0.44	-0.18 ± 0.50
ΔFINS (mU/L)	-9.32 ± 10.18*	-5.50 ± 7.44	-3.80 ± 7.44
ΔHOMA-IR	-2.80 ± 2.56*#	-1.43 ± 2.30	-0.99 ± 1.65
ΔSHBG (nmol/L)	31.82 ± 40.03*	26.06 ± 26.44	11.77 ± 27.29
ΔDHEA-S(µg/dL)	-23.29 ± 76.81	-15.89 ± 88.91	-16.35 ± 87.73
ΔE2(pg/ml)	17.13(-2.08,112.97)#	-1.77(-14.09,23.15)	8.85(-9.52,43.12)
ΔTEST (ng/dl)	-15.40 ± 16.84*	-9.14 ± 16.31*	-4.20 ± 12.60
ΔPRL (ug/L)	2.77 ± 5.59	2.05 ± 9.07	1.68 ± 7.19
ΔFAI	-6.31 ± 6.32*	-6.21 ± 6.53*	-2.26 ± 6.37
ΔLH/FSH	-0.34 ± 1.02	-0.35 ± 1.19	-0.31 ± 1.30
ΔCRP (mg/L)	-0.71 ± 1.59*#	-0.22 ± 0.98	-0.03 ± 0.33
ΔAMH(ng/mL)	-2.69 ± 2.32*#	-1.75 ± 1.92	-1.27 ± 2.64
Menstrual cycle recovery rate	22(81.5%)*#	14(51.9%)	10(38.5%)
Pregnancy rate	14(51.9%)*	11(40.7%)	6(20.1%)

Data are shown as the mean ± standard deviation (SD) or median (interquartile range) or frequencies (percentage)

BMI: Body mass index; WHR: Waist-to-hip ratio; FBG: fasting blood glucose; FINS: fasting insulin; HOMA-IR: homeostasis model assessment of insulin resistance. SHBG: Sex hormone binding globulin; DHEA-S: Dehydroepiandrosterone sulfate; E2: Estradiol; TEST: Testosterone; PRL: Prolactin; FAI: Free androgen index; LH: Luteinizing hormone; FSH: Follicle stimulating hormone; AMH: anti-Müllerian hormone;

The bold font indicates statistically significant between the two groups;

Low Responsive Patients with PCOS is relative to Highly Responsive Patients with PCOS

^*P < 0.05 vs. Non-Responsive Patients with PCOS;

^#P < 0.05 vs. Low Responsive Patients with PCOS

Adverse events

No serious adverse events or deaths occurred, and no participants withdrew due to adverse events. The most common adverse events were mild to moderate gastrointestinal issues, including nausea, vomiting, diarrhea, and bloating. A few participants reported headaches and fatigue. Most gastrointestinal adverse events were mild to moderate and improved within the first month, with the COM group experiencing more adverse events than the MET group.

Discussion

This study assessed the effects of combined MET and semaglutide therapy on reproductive outcomes in patients with PCOS. The findings demonstrated that the combination therapy (COM group) outperformed MET monotherapy in reducing BMI, androgen levels, insulin resistance, and menstrual irregularities in overweight/obese patients with PCOS. Additionally, the natural pregnancy rate was significantly higher in the COM group compared to the MET group.

Obesity and insulin resistance are prevalent in patients with PCOS [20]. Obesity, affecting 20–80% of patients with PCOS, is a key therapeutic target [1]. Both obesity and insulin resistance disrupt the hypothalamic–pituitary–ovarian (HPO) axis, resulting in ovulatory dysfunction, menstrual irregularities, and negative pregnancy outcomes [21, 22]. Furthermore, these conditions can create a self-reinforcing cycle that worsens metabolic disturbances While lifestyle interventions are advised, MET remains the first-line treatment for metabolic abnormalities in PCOS. However, MET’s impact effects on weight and BMI reduction is limited [23]. In contrast, GLP-1 RAs have shown superior efficacy in reducing weight, improving glucose and lipid metabolism, and enhancing long-term cardiovascular and renal outcomes, making them effective in managing type 2 diabetes mellitus and obesity [24]. Once-weekly administration of the GLP-1 RA semaglutide has demonstrated greater efficacy in weight reduction and a lower incidence of adverse effects compared to liraglutide, contributing to higher patient adherence rates [25]. This study is the first to evaluate the combined efficacy of semaglutide and MET therapy in treating obese/overweight patients with PCOS. The results indicate that the combination therapy significantly reduced body weight and FBG levels compared to MET monotherapy, though it did not show superior efficacy in improving insulin resistance. Some studies have suggested that HOMA-IR, a common measure of insulin resistance, exhibits high variability and low sensitivity in patients with PCOS [26]. Therefore, we used CVAI as an alternative measure of insulin resistance [27]. CVAI was utilized as an alternative measure of insulin resistance in the COM group, indicating that combination therapy with semaglutide and MET is more effective at enhancing insulin resistance in patients with PCOS. In this study, the average weight loss in the COM group was 6.09 kg, differing from the 5.0 kg reduction reported by Blundell et al. following 12 weeks of 1.0 mg semaglutide treatment [28]. This discrepancy may be attributed to the synergistic effects of MET and semaglutide as well as the longer treatment duration in this study. Additionally, a significant reduction in WHR was noted in the COM group, suggesting that the combination therapy more effectively regulates central obesity. Notably, when participants were categorized into high-response, low-response (relative to high-response), and non-response groups based on weight loss, the high-response group showed more significant improvements in metabolic markers (FBG, FINS, HOMA-IR, and TEST) and menstrual regularity. This underscores the importance of weight loss in overweight/obese patients with PCOS and may inform the development of targeted weight loss strategies for this patient population.

Hyperandrogenism is both a clinical manifestation and a central pathophysiological mechanism of PCOS. Elevated androgen levels in PCOS result from increased steroidogenesis in theca cells [29], as well as altered regulation of the HPO axis, with insulin and inflammatory factors playing a significant role. Persistent hyperandrogenism leads to endometrial dysfunction and ovulatory disorders, contributing to menstrual irregularities and infertility. Therefore, reducing androgen levels is critical in managing PCOS management. A prior study on low-dose semaglutide reported improvements in menstrual regularity in patients with PCOS presenting > 10% weight loss, though changes in sex hormone levels were not assessed [17]. This study demonstrated that both combination therapy with semaglutide and MET and MET monotherapy reduced androgen levels and facilitated menstrual cycle recovery. In our study, the COM group showed greater improvement in androgen levels, aligning with previous studies on exenatide and liraglutide in overweight/obese women with PCOS [15, 30]. Liu et al. reported reduced TEST levels and improved estrous cycles in rats with PCOS after 4 weeks of semaglutide treatment, consistent with the findings of this study. They found that semaglutide inhibited the overexpression of cytochrome P450 17A1 (CYP17A1) and steroidogenic acute regulatory protein (StAR), while enhancing the expression of cytochrome P450 19A1 (CYP19A1), thus alleviating hyperandrogenism and improving estrous cycle regularity [31]. These mechanisms may account for the therapeutic effects observed with semaglutide and MET combination therapy in this study.

Some patients with PCOS exhibit characteristic PCOM. AMH, a peptide secreted by granulosa cells of early-developing follicles [32], is frequently elevated in patients with PCOS, peaking between the ages of 20–25 years [33]. High AMH levels contribute to ovulatory dysfunction and menstrual irregularities, by inhibiting the recruitment of developing follicles into the growing pool, causing follicular arrest and disrupting dominant follicle formation. AMH also decreases FSH sensitivity in follicles, impairing dominant follicle development, and inhibits aromatase, which reduces androgen-to-estrogen conversion, thereby exacerbating hyperandrogenism [34]. Recent international guidelines (2023) have recommended using AMH as a surrogate marker for PCOM, with a reduction in AMH levels serving as an indicator of treatment efficacy in patients with PCOS [33]. Based on validation data from a large-scale retrospective multicenter study, a cut-off of 3.2 ng/mL for AMH was proposed [35], although this threshold remains a topic of ongoing debate. Notably, AMH levels are generally not elevated in obese individuals. Research on GLP-1 RAs and AMH is limited, though animal studies have shown that exenatide reduces AMH levels in diabetic rats [36]. A clinical study demonstrated that 6 months of liraglutide treatment reduced AMH levels in patients with PCOS but this reduction was not significantly different from that induced by placebo [11]. In this study, AMH levels decreased in both MET and COM groups, though no significant difference was observed between the two groups. MET may reduce AMH levels by improving insulin resistance, reducing follicular arrest, and promoting spontaneous ovulation. Although semaglutide demonstrates greater efficacy in weight reduction, improving insulin resistance, and decreasing androgen levels, it was not more effective than MET in reducing AMH levels, in this study. This may be due to possibly owing to the temporary ovarian function recovery (e.g., increased antral follicle recruitment) or fluctuations in the HPO axis induced by semaglutide, which could partially counteract the metabolic improvements. Additionally, while MET may have a lesser effect on weight-loss, it may reduce AMH levels by directly inhibiting ovarian androgen synthesis and regulating follicular development, leading to similar reductions in AMH levels as COM, albeit through different mechanisms.

PCOS remains the leading cause of ovulatory dysfunction in women of reproductive age, with many patients seeking treatment to improve fertility. MET, a first-line treatment for PCOS, enhances pregnancy rates but offers limited benefits [10]. Clinical studies have shown that GLP-1 RAs, when combined with MET, can further improve reproductive outcomes in patients with PCOS. A study of 176 overweight/obese patients with PCOS, demonstrated that exenatide treatment led to a higher natural pregnancy rate compared to MET (43.60% vs. 18.70%) [37], which aligns with our findings. Another study of 28 obese patients with PCOS found that liraglutide combined with MET resulted in higher pregnancy rates per embryo transfer compared to MET monotherapy (85.7% vs. 28.6%) [38]. However, the biological role of GLP-1 RAs and their effects on pregnancy outcomes in PCOS require further investigation. In overweight/obese individuals, semaglutide has been shown higher weight loss, fewer adverse effects, and result in greater adherence rates compared to liraglutide [16]. This study is the first to evaluate pregnancy outcomes in patients with PCOS treated with semaglutide plus MET. Compared to MET monotherapy, the combination therapy significantly increased natural pregnancy rates. This improvement may be attributed to the regulatory effects of semaglutide on the HPO axis, which reduce insulin resistance and inflammation, thereby promoting ovulation and enhancing pregnancy potential. Although no significant improvement in total pregnancy rates was observed in the COM group, the trend toward enhanced fertility suggests that semaglutide is a promising treatment for metabolic impairment-related infertility and reproductive disorders. Larger-scale studies with extended follow-up periods may provide clearer evidence of significant improvements in total pregnancy rates. Additionally, the mechanisms through which GLP-1 RAs improve pregnancy outcomes in patients with PCOS should be further explored.

Chronic low-grade inflammation is another hallmark of PCOS, often accompanied by overweight/obesity. CRP, a key inflammatory marker, can predict the risk of cardiovascular disease and diabetes. A study indicated that CRP levels remained significantly elevated in patients with PCOS even after adjusting for age and BMI [39], which aligns with our findings. Animal studies have demonstrated that semaglutide upregulates the adenosine monophosphate activated protein kinase (AMPK)/ Sirtuin 1 (SIRT1) signaling pathway while inhibiting the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway in ovarian tissue, thus reducing inflammation and improving sex hormone levels in mice with PCOS [32]. Our study showed that semaglutide combined with MET significantly reduced CRP levels after 16 weeks of treatment, whereas MET monotherapy did not produce this effect. Morin-Papunen et al. reported that 6 months of MET treatment significantly reduced CRP levels in patients with PCOS [40], which contradicts our findings. This discrepancy may be attributed to the longer treatment duration in their study. We speculate that semaglutide combined to MET improves insulin resistance, lipid profiles, androgen levels, menstrual irregularities, and natural pregnancy rates by reducing inflammation.

This study has several limitations. First, it was conducted at a single center with a small sample size and a short treatment duration. As an open-label study, it was prone to performance and detection biases. However, the use of objective clinical outcomes helped mitigate detection bias. Larger, multicenter studies with longer follow-up periods are needed to validate these findings. Second, our study exclusively enrolled patients with infertility durations exceeding 2 years without contraceptive use. Further research is required to explore the effects of treatments in patients with shorter or longer durations of infertility. Additionally, continuing MET after semaglutide discontinuation might have influenced pregnancy outcomes. Long-term studies are necessary to assess the effects of GLP-1 RAs on overall pregnancy rates and assisted reproductive outcomes, such as IVF-ET.

Conclusion

In conclusion, 16 weeks of combination therapy with semaglutide and MET was more effective than MET monotherapy in reducing body weight and WHR, improving sex hormone levels, and enhancing reproductive potential in overweight/obese Chinese women with PCOS. The combination therapy also significantly increased natural pregnancy rates. However, further research is necessary to evaluate its impact on overall pregnancy rates and assisted reproductive outcomes.

Abbreviations

PCOS

Polycystic ovary syndrome

MET

Metformin

COM

Metformin combined with semaglutide therapy

CRP

C-reactive protein

WHR

Waist-to-hip ratio

UA

Uric acid

ALT

Alanine aminotransferase

AST

Aspartate aminotransferase

TG

Triglycerides

TC

Total cholesterol

HDL-C

High-density lipoprotein cholesterol

LDL-C

Low-density lipoprotein cholesterol

WBCs

White blood cells

CRP

C reactive protein

HbA1c

Glycosylated hemoglobin

FBG

Fasting blood glucose

FINS

Fasting insulin

HOMA-IR

Homeostasis model assessment of insulin resistance

SHBG

Sex hormone binding globulin

DHEA-S

Dehydroepiandrosterone sulfate

E2

Estradiol

TEST

Testosterone

PRL

Prolactin

FAI

Free androgen index

LH

Luteinizing hormone

FSH

Follicle stimulating hormone

AMH

Anti-Müllerian hormone

CVAI

Chinese visceral adiposity index

hCG

Human chorionic gonadotrophin

CV

Coefficients of variation

ELISA

Enzyme-linked immunosorbent assay

GLP-1 RAs

Glucagon-like peptide-1 receptor agonists

WC

Waist circumference

HC

Hip circumference

HPO

Hypothalamic–pituitary–ovarian

SD

standard deviation

Author contributions

Haiyan Chen and Xiaohui Lei contributed equally to this work. Both Haiyan Chen and Xiaohui Lei were involved in the study concept and design, interpreting the data, compiling the statistical dataset, performing the analyses, and writing and revising the manuscript. Zhuoran Yang and Yuxin Xu were responsible for the follow-up of sub-centers and the collection of clinical data. Cong Wang, Hu Du, and Dongfang Liu contributed to data interpretation and the critical revision of the manuscript. All authors reviewed and approved the final version of the manuscript, and no other individuals made substantial contributions to this work.

Funding

The work was supported by the: Chongqing medical scientific research project (Joint project of Chongqing Health Commission and Science and Technology Bureau) (2025MSXM106), the Chongqing Municipal Health Commission (no. 2022ZDXM004, China), the First batch of key Disciplines on Public Health in Chongqing, the National Natural Science Foundation of China (Grants 81501199 to C.W.), the Natural Science Foundation Project of Chongqing CSTC (cstc2017jcyjAX0016 to C.W. and CSTB2022NSCQ-MSX1008 to C.W.).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study was approved by the hospital’s ethics committee of the Second Affiliated Hospital of Chongqing Medical University (Approval No. 2024204) and registered on Chinese Clinical Trial Registry (ChiCTR2400090908) in October 15, 2024. All participants received written and oral information and signed informed consent before any examination.

Consent for publication

Not applicable.

Disclaimers

Any opinions or recommendations discussed are solely those of the author(s).

Source(s)of support

Chongqing medical scientific research project (Joint project of Chongqing Health Commission and Science and Technology Bureau) (2025MSXM106), the First batch of key Disciplines on Public Health in Chongqing, the National Natural Science Foundation of China (Grants 81501199 to C.W.), the Natural Science Foundation Project of Chongqing CSTC (cstc2017jcyjAX0016 to C.W. and CSTB2022NSCQ-MSX1008 to C.W.).

Competing interests

The authors declare no competing interests.