Correlation between anti-müllerian hormone in polycystic ovarian syndrome with metformin: a systematic review and meta-analysis

Figures & data

Figure 1. Flowchart of the systematic literature search strategy and results.

Table 1. Characteristics of the studies included in the meta-analysis.

Author	Year	Region	Design	Population	Criteria of PCOS	Time	Sample	Age (y)	BMI	Daily	Method	AMH (Mean ± SD)		Conclusion
								(Mean ± SD)	(Mean ± SD)	dosage		Pre	Post
Nazirudeen, R	2023	India	RCT	PCOS women (BMI > 23)	Rotterdam criteria	6 months	15	24.35 ± 5.10	31.52 ± 4.40	1000mg	ECLIA	5.77 ± 4.38	2.03 ± 0.63	MET reduces AMH levels in obese PCOS.
Daneshjou, D	2022	Iran	RCT	Infertile PCOS women	The NIH criteria.	2 months	15	28.93 ± 2.76	27.01 ± 0.73	1000mg	NA	6.92 ± 1.87	6.69 ± 0.78	MET decrease AMH levels
Rajasekaran, K	2022	India	RCT	Infertile PCOS women	Rotterdam criteria	3 months	50	30.00 ± 4.00	27.00 ± 4.20	1700mg	NA	10.91 ± 4.10	10.22 ± 9.14	MET decrease AMH levels
Hughes, L	2022	USA	RCT	Infertile PCOS women	NA	6 treatment cycles	79	27.80 ± 3.50	33.80 ± 8.40	NA	ELISA	12.70 ± 9.20	12.5 ± 10.0	MET alone did not affect AMH levels.
Usman, Fatimah	2020	Indonesia	Prospective cohort	PCOS with IR women	1.oligo or anovulation; 2.hyperandrogenism, 3.PCOS features on transvaginal ultrasound or laparoscopy	3 months	70	28.61 ± 3.88	IQR:19-33	2550mg	ELISA	7.70 ± 0.60	7.59 ± 0.58	MET reduces AMH levels in obese PCOS.
Chhabra, N	2018	India	Prospective cohort	Infertile PCOS women	Rotterdam criteria	3 months	32	30.03 ± 3.00	NA	1700mg	NA	11.87 ± 5.60	7.40 ± 3.90	MET alone best reduced AMH levels, comparing to other insulin sensitizer groups.
Foroozanfard, F	2017	Iran	Prospective cohort	Infertile PCOS women	Rotterdam criteria	8 weeks	30	25.20 ± 4.20	26.20 ± 3.80	1500mg	ELISA	10.0 ± 3.70	7.80 ± 3.70	1. MET decreased AMH. 2.AMH change was linked to BMI
Wiweko, B	2017	Indonesia	RCT	PCOS women	Rotterdam criteria	6 months	20	28.25 ± 3.99	28.02 ± 6.02	1500mg	Coulter gen II	9.30 ± 5.06	7.47 ± 4.59	MET decreased AMH more than DLBS3233
Madsen, H N	2015	Denmark	RCT	Hyperandrogenemic PCOS women	supranormal levels of testosterone oligomenorrhea or amenorrhea	6 months	41	31.51 ± 4.92	33.70 ± 6.36	1700mg	Coulter gen II	7.45 ± 6.17	6.50 ± 4.01	MET reduces AMH levels in obese PCOS.
Saleh, B O	2015	Iraq	Cross-sectional	PCOS women	Rotterdam criteria	3 months-3 years	20	27.50 ± 1.32	30.85 ± 0.95	1500mg	ELISA	4.49 ± 0.54	3.03 ± 0.53	1. MET resulted in the reduction of AMH levels 2. obese PCOS patients are more likely to respond to MET therapy with maximal doses (2550 mg/day)
Nascimento, A D	2013	Brazil	Prospective cohort	PCOS with IR women	Rotterdam criteria	8 weeks	16	26.30 ± 1.00	29.10 ± 1.70	1500mg	ELISA	6.99 ± 0.85	5.81 ± 0.78	No alteration in AMH levels was observed after treatment
Neagu, M	2012	Romania	Prospective cohort	Infertile PCOS women	AE/PCOS 2006 criteria	2 months	11	NA	NA	2550mg	NA	8.99 ± 0.99	6.28 ± 0.46	Serum AMH level after MET treatment were considerable lower
Tomova, A	2011	USA	RCT	PCOS women	Rotterdam criteria	6 months	13	26.59 ± 4.92	30.73 ± 8.50	2550mg	ELISA	6.40 ± 4.69	5.36 ± 3.46	1. Metformin decreases AMH, but not statistically significant. 2. Obese PCOS patients respond well to high-dose metformin.
Panidis, D	2011	Greece	Prospective cohort	PCOS women (BMI < 25)	The NIH criteria	6 months	15	21.07 ± 3.21	21.52 ± 1.01	1700mg	ELISA	9.24 ± 3.70	7.77 ± 2.82	There was no statistically significant change with MET
Fábregues, F	2011	Spain	Prospective cohort	Infertile PCOS women	Rotterdam criteria	6 months	11	26.90 ± 6.02	28.30 ± 6.46	1700mg	ELISA	6.40 ± 0.58	4.73 ± 0.54	MET decrease AMH levels, but the effect was not statistically significant.
Romualdi, D	2011	Italy	Prospective cohort	Obese PCOS women	Rotterdam criteria	6 months	28	26.43 ± 5.60	33.55 ± 4.62	1700mg	EIA	5.21 ± 1.36	4.33 ± 1.71	AMH values remained unchanged (data not shown)
Falbo, A	2010	Italy	Prospective cohort	Young normal weight PCOS women	The presence of both chronic anovulation and clinical and/or biochemical hyperandrogenism	12 months	20	27.83 ± 3.61	22.84 ± 1.64	1700mg	EIA	anov: 5.23 ± 1.56 ov: 5.75 ± 1.59	anov: 3.23 ± 2.64 ov: 4.09 ± 3.91	MET decrease serum AMH levels.
Carlsen, S M	2009	Norway	Prospective cohort	PCOS women	Rotterdam criteria	6 months	20	30.60 + 5.90	33.40 + 7.50	2550mg	ELISA	15.30 + 11.50	15.20 + 12.00	Protracted MET treatment did not influence ovarian volumes
Bayrak, A	2007	USA	Prospective cohort	PCOS women	Chronic oligoanovulation, hyperandrogenism, or clinical signs of hyperandrogenism	1 weeks	10	29.50 ± 4.80	30.20 ± 4.50	1700mg	ELISA	2.82 ± 1.8	2.92 ± 1.6	One week of metformin therapy did not alter AMH levels.
Piltonen, T	2005	Finland	Prospective cohort	PCOS women	Rotterdam criteria	6 months	26	30.06 ± 5.64	30.10 ± 7.26	1500mg	EIA	12.25 ± 2.10	11.41 ± 2.24	Metformin treatment for 6 months lowered AMH levels

BMI body mass index, NA not available, SD standard deviation, Pre before metformin, Post after metformin, ELISA enzyme linked immunosorbent assay, EIA Enzyme Immunoassay, Coulter gen II Beckman Coulter Gen II ELIS, ECLIA electrochemiluminescence immunoassay.

Table 2. Changes in LH and FSH hormones and number of follicles before and after medication.

Author	Pretreatment				Post-treatment
	LH	FSH	LH/FSH	AFC	LH	FSH	LH/FSH	AFC
Piltonen, T 2005	NA	NA	NA	11.90 ± 0.50	NA	NA	NA	8.3 ± 0.6
Bayrak, A 2007 (IR patients with PCOS)	9.42 ± 4.61	5.56 ± 2.06	1.73 ± 0.73	38.80 ± 19.30	9.10 ± 4.70	5.96 ± 1.59	1.46 ± 0.46	50.75 ± 19.26
Bayrak, A 2007 (NIR patients with PCOS)					9.70 ± 5.06	5.16 ± 2.56	1.47 ± 0.57	29.40 ± 14.70
Falbo, A 2010 (Met-anov group)	11.89 ± 3.87	5.85 ± 1.57	NA	NA	△-0.12 ± 0.31	△0.06 ± 0.03	NA	NA
Falbo, A 2010 (Met-ov group)	12.56 ± 3.45	5.86 ± 1.32	NA	NA	△-0.76 ± 0.25	△-0.03 ± 0.04	NA	NA
Fábregues, F 2011	13.10 ± 1.20	NA	NA	28.06 ± 6.80	12.10 ± 1.30	NA	NA	24.20 ± 30.00
Panidis, D 2011	12.81 ± 7.65	6.56 ± 1.46	NA	NA	13.60 ± 16.69	6.49 ± 3.76	NA	NA
Chhabra, N 2018	NA	NA	NA	13.06 ± 1.8	NA	NA	NA	9.8 ± 1.7
Daneshjou, D 2022	NA	4.99 ± 0.72	NA	NA	NA	5.98 ± 0.96	NA	NA
Rajasekaran, K 2022	7.00 ± 3.43	7.37 ± 1.34	0.94 ± 0.62	24.8 ± 4.1	6.82 ± 4.90	5.40 ± 1.31	1.25 ± 0.96	22.5 ± 4.0
Nazirudeen, R 2023	15.24 ± 11.33	6.83 ± 3.24	2.21 ± 1.17	NA	8.4	NA	0.48	NA

Table 3. Quality Assessment of meta-analysis based on the Newcastle-Ottawa Scale judgment.

Study	Selection				Comparability	Outcome			Quality score
	Representativeness of the exposed cohort	Selection of the non exposed cohort	Ascertainment of exposure	Demonstration that outcome of interest was not present at start of study	Comparability of cohorts on the basis of the design or analysis	Assessment of outcome	Was follow-up long enough for outcomes to occur	Adequacy of follow up of cohorts
Nazirudeen, R	*	*	*	*	*	*	*	*	********
Daneshjou, D	*	*	*	*	*	*	*	*	********
Rajasekaran, K	*	*	*	*	*	*	*	*	********
Hughes, L	*	*	*	*	*	*			******
Usman, Fatimah	*	*	*	*	*	*			******
Chhabra, N	*	*	*	*	*	*			******
Foroozanfard, F	*	*	*	*	*	*	*		*******
Wiweko, B	*	*	*	*	*	*	*		*******
Madsen, H N	*	*	*	*	*	*			******
Saleh, B O	*	*	*	*	*	*			******
Nascimento, A D	*	*	*	*	*	*			******
Neagu, M	*	*	*	*	*	*			******
Tomova, A	*	*	*	*	*	*	*	*	********
Panidis, D	*	*	*	*	*	*	*		*******
Fábregues, F	*	*	*	*	*	*	*		*******
Romualdi, D	*	*	*	*	*	*	*		*******
Falbo, A	*	*	*	*	*	*	*	*	********
Carlsen, S M	*	*	*	*	*	*			******
Bayrak, A	*	*	*	*	*	*			******
Piltonen, T	*	*	*	*	*	*	*		*******

Figure 2. Funnel Plot of Serum AMH Changes before and after Metformin Treatment.

Figure 3. Forest plot of Serum AMH Changes before and after Metformin Treatment.

Table 4. Sensitivity analysis of serum AMH Changes before and after MET treatment.

Remove literature	Heterogeneity test	MD 95%CI	P-value
Piltonen, T 2005	I²=85%, p < 0.00001	1.36 [0.81, 1.92]	p <0 .00001
Bayrak, A 2007 (IR patients with PCOS)	I²=85%, p <0 .00001	1.38 [0.84, 1.92]	p < 0.00001
Bayrak, A 2007 (NIR patients with PCOS)	I²=85%, p < 0.00001	1.38 [0.85, 1.92]	p <0 .00001
Carlsen, S M 2009	I²=85%, p <0 .00001	1.34 [0.81, 1.87]	p <0 .00001
Falbo, A 2010 (Met-anov group)	I²=85%, p <0 .00001	1.30 [0.76, 1.84]	p <0 .00001
Falbo, A 2010 (Met-ov group)	I²=85%, p <0 .00001	1.32 [0.79, 1.86]	p < 0.00001
Fábregues, F 2011	I²=83%, p <0 .00001	1.30 [0.75, 1.86]	p <0 .00001
Panidis, D 2011	I²=85%, p <0 .00001	1.33 [0.79, 1.87]	p <0 .00001
Romualdi, D 2011	I²=85%, p < 0.00001	1.37 [0.81, 1.93]	p < 0.00001
Tomova, A 2011	I²=85%, p < 0.00001	1.34 [0.80, 1.87]	p <0 .00001
Neagu, M 2012	I²=80%, p < 0.00001	1.20 [0.70, 1.70]	p <0 .00001
Nascimento, A D 2013	I²=85%, p < 0.00001	1.35 [0.78, 1.92]	p < 0.00001
Madsen, H N 2015	I²=85%, p <0 .00001	1.35 [0.81, 1.89]	p < 0.00001
Saleh, B O 2015	I²=83%, p < 0.00001	1.33 [0.74, 1.91]	p <0 .00001
Foroozanfard, F 2017	I²=85%, p <0 .00001	1.29 [0.76, 1.83]	p <0 .00001
Wiweko, B 2017	I²=85%, p < 0.00001	1.32 [0.79, 1.86]	p <0 .00001
Chhabra, N 2018	I²=84%, p < 0.00001	1.23 [0.71, 1.75]	p <0 .00001
Usman, Fatimah 2020	I²=51%, p = 0.004	1.45 [1.06, 1.84]	p <0 .00001
Hughes, L 2022	I²=85%, p <0 .00001	1.36 [0.82, 1.89]	p < 0.00001
Daneshjou, D 2022	I²=85%, p < 0.00001	1.41 [0.86, 1.96]	p <0 .00001
Rajasekaran, K 2022	I²=85%, p < 0.00001	1.35 [0.81, 1.89]	p <0 .00001
Nazirudeen, R 2023	I²=84%, p <0 .00001	1.25 [0.72, 1.78]	p <0.00001

Figure 4. Forest plot for subgroup analysis of Criteria of PCOS.

Figure 5. Forest plot for subgroup analysis of Detection Methods.

Figure 6. Forest plot for subgroup analysis of BMI.

Figure 7. Forest plot for subgroup analysis of Daily dosage.

Figure 8. Forest plot for subgroup analysis of Treatment duration.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.