Real-world Efficacy of Acupoint Stimulation on Clinical Pregnancies Among Women Following Frozen-Thawed Single Blastocyst Transfer: a retrospective cohort study of 4863 cycles

doi:10.21203/rs.3.rs-4672336/v1

Download PDF

Research Article

Real-world Efficacy of Acupoint Stimulation on Clinical Pregnancies Among Women Following Frozen-Thawed Single Blastocyst Transfer: a retrospective cohort study of 4863 cycles

https://doi.org/10.21203/rs.3.rs-4672336/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Acupoint stimulation is a frequently used complementary treatment among individuals undergoing assisted reproductive technology, however, the evidence for its efficacy remains a subject of considerable debate.

Methods

A single-center retrospective cohort study of 4863 cycles was to determine the efficacy of varying acupoint stimulation protocols on clinical pregnancies among 3,907 women aged < 43 years during single frozen-thawed blastocyst transfer. Based on the timing and dosage of acupoint stimulation they fulfilled, the participations were categorized into four groups: group T0 (n = 1227) without acupoint stimulation, group T1(n = 1607) with one session of acupoint stimulation administered within 2 hours prior to embryo transfer procedure, group T2-3 (n = 1600) with a total of 2–3 sessions of acupoint stimulation conducted during the period from progesterone administration day to 2 hours prior to transfer procedure, and group T ≥ 4 (n = 429) having at least 4 sessions of acupoint stimulation from the 5th day of exogenous estrogen initiated and continuing until 2 hours prior to the embryo transfer procedure. In this study acupoint stimulation was facilitated through the application of medicated patches in conjunction with electromagnetic wave heat therapy (AMP-EWH). The primary outcome was clinical pregnancy, with 8 weeks of pregnancy follow-up.

Results

Clinical pregnancies occurred among 734 of 1227 cycles in group T0(59.8%), 1030 of 1607 cycles in group T1(64.1%), 999 of 1600 cycles in group T2-3(62.4%), and 289 of 429 cycles in group T ≥ 4(67.4%). Upon various statistical analysis models compared to the CPR of group T0, group T1 showed OR 1.199,(95%CI,1.029 to 1.397), p = 0.020, in Model crude; 1.173,(0.999 to 1.376),0.054 ,Model 1;1.184,(1.008 to 1.391),0.040, Model 2;1.170,(0.985 to 1.390),0.074, Model PSM; group T2-3 showed CPR OR 1.116, (95%CI, 0.959 to 1.300), p = 0.157, in Model crude; 1.062, (0.905 to 1.300), 0.460, Model 1; 1.076 (0.916 to 1.265), 0.373, Model 2; and 1.069 (0.898 to 1.274), 0.454, Model PSM; and group T ≥ 4 showed CPR OR 1.387, (95%CI, 1.100 to 1.748), p = 0.006, in Model crude; 1.380, (1.084 to 1.758), 0.009, Model 1; 1.394 (1.093 to 1.778), 0.008, Model 2; and 1.672 (1.219 to 2.295), 0.001, Model PSM.

Conclusions

AMP-EWH administrated at least 4 sessions resulted in a significantly higher clinical pregnancy rate of single frozen-thawed blastocyst transfer among women under 43 years old. These findings support the use of AMP-EWH as an effective adjunct to frozen-thawed blastocyst transfer for enhancing clinical pregnancy rates.

Frozen-thawed single blastocyst transfer

Acupoint stimulation

Acupoint medication patches conjunction with electromagnetic wave heat (AMP-EWH)

Timing

Dosage

Endometrial receptivity

Assisted reproductive technology (ART) have evolved significantly. Among these, elective single embryo transfer, particularly of single frozen-thawed blastocyst transfer (denoted as FET) has emerged as a pivotal strategy in ART(1). Despite its clinical significance, blastocyst transfer success rates remain suboptimal in some cases, and subject to complex interplay of patient-specific factors. Actually, there is no shortage of patients who have encountered implantation failures even repeated implantation failures without identifiable reasons(2). Given the widespread implementation of preimplantation genetic testing (PGT) for embryo screening remains limited, the pursuit of complementary therapies to optimize endometrial receptivity is therefore of paramount importance to embryo implantation and subsequent pregnancy establishment(3). There remains a quest for non-invasive, adjunctive therapies that can further augment the receptive state of the endometrium.

Traditional Chinese Medicine (TCM) offers a rich repository of therapeutic interventions. One of which is acupoint therapy, it involves the stimulation of specific points on the body, addressing a wide array of health conditions. The past few decades have witnessed a surge in research exploring the adjunctive role of acupuncture to ART. Based on meridian theory of TCM, acupuncture can regulate the function of “kidney qi” to improving reproductive function via stimulating acupoints located on specific meridian. This integrative approach has permeated various stages of in vitro fertilization-embryo transfer (IVF-ET)(4), encompassing ovulation induction(5), anesthesia during oocyte retrieval(6), embryo transfer procedures(7), and the management of threatened miscarriag(8, 9). The therapeutic avenues of acupoint stimulation have expanded to include manual acupuncture, moxibustion, the application of acupoint plasters, transcutaneous electrical acupoint stimulation (TEAS), acupoint injection, and so on(10), reflecting a diversification of traditional techniques to meet modern medical demands. Notably, acupuncture has attracted considerable interest due to its potential to enhance ART outcomes. Significantly, the integration of acupuncture into frozen–thawed embryo transfer (FET) protocols has shown promising results, thereby fostering a burgeoning area of research within the field of reproductive medicine(11). To augment the efficacy of acupuncture-like stimulation while minimizing its invasiveness, we focus on exploring the acupoint stimulation by application of acupoint medication patches in conjunction with electromagnetic wave heat therapy (AMP-EWH). Compared with typical manual acupuncture or electro-acupuncture, AMP-EWH carries seldom risk of infective complications or visceral injury and mitigates any potential needle-related anxieties or fear. Noteworthily, the acupoint drug patch provides a sustained therapeutic effect over 4 to 6 hours, optimizing its clinical efficacy. Currently, in the field of reproductive medicine, acupoint medicated patching is predominantly utilized for the treatment of infertility with follicular dysplasia(12), threatened miscarriage(13) and for alleviating hyperemesis gravidarum(14), indicating its high efficacy and safety. To our knowledge, the effects of AMP-EWH on frozen-thawed blastocyst transfer have not previously been reported.

This retrospective cohort study, based on a large sample size of infertile women following single frozen-thawed blastocyst transfers in actual clinical setting, aims to figure out three main issues. Firstly, to ascertain whether adjunctive acupoint stimulation protocols can make single frozen-thawed blastocyst transfers more likely to result in a clinical pregnancy. Secondly, if it does help, to identify which acupoint stimulation protocol offers a more pronounced advantage regarding pregnancy outcomes. And thirdly, to assess the different performance of acupoint stimulation protocol across various patient subgroups, then to determine the group that stands to benefit the most from this acupoints stimulation intervention.

Sample selection and ethical approval

This was a single-center retrospective cohort study, a total of 3907 individual patients who met our inclusion criteria had experienced 4863 single frozen-thawed blastocyst transfer cycles included at Reproductive Medicine Centre of the First Affiliated Hospital of Wenzhou Medical University during the period from January 2022 to December 2023. The study was approved by the Ethics Committee of Reproductive Medicine of the First Affiliated Hospital of Wenzhou Medical University (NO: KY2023-R202).

Study design

Based on whether patients received acupoint stimulation by AMP-EWH prior to embryo transfer, they were categorized into two groups: the non-acupoint stimulation group (group T0, with 1227 cycles) and the acupoint stimulation group (group T ≥ 1, with total 3636 cycles). Furthermore, group T ≥ 1 was classified into three subsets (group T1, groupT2-3, and group T ≥ 4), depending on the timing and dosage of AMP-EWH therapy they fulfilled. Group T1 was consisted of 1607 cycles, group T2-3 of 1600 cycles, and group T ≥ 4 of 429 cycles.

Inclusion criteria and exclusion criteria

The inclusion criteria were as follows: (1) Infertility women with age < 43 years; (2) A single autologous frozen-thawed blastocyst transferred. The exclusion criteria were as follows: (1) Blastocysts from pre-implantation genetic testing (PGT); (2) Sperm from surgical retrieval; (3) Male with chromosomal abnormality (using his own sperm for fertilization); (3) Female with chromosomal abnormality; (4) Recurrent spontaneous abortion; (5) Female with basic diseases, such as diabetes, hypertension, endocrinopathy, and autoimmune diseases; (6) Unmanaged abnormal uterine conditions, included hydrosalpinx, endometrial polyps, uterine adhesions, endometrial fluid collection, submucous leiomyoma, severe adenomyosis and uterine fibroids (≥ 5cm); (7) Incomplete pre-8-week pregnancy follow-up record.

Outcome measurements

The primary outcome measure of this study was clinical pregnancy rate (CPR), while the secondary outcome measures were β-HCG positivity rate, biochemical pregnancy loss rate, and ectopic pregnancy rate. Clinical pregnancy was defined as at least one intrauterine gestational sac detected by transvaginal ultrasound 4 or 5 weeks after embryo transfer. β-HCG positivity was defined as a serum β-hCG level ≥ 10 mIU/mL at 12 days following blastocyst transfer procedure. Biochemical pregnancy loss was defined as the absence of gestational sac both intrauterine and extrauterine until 6 weeks after the blastocyst transfer with prior confirmation of a positive HCG test. Ectopic pregnancy was defined as the gestational sac outside the uterine cavity and without any sac intrauterine.

Endometrial preparation protocols and FET processes

All participants received the same hormonal (estrogen and progesterone) protocol for endometrial preparations before blastocyst transfer. Patients started hormon replacement therapy with exogenous estradiol on day 3–6 of the menstrual cycles. In all cases, patients were given oral administration of 2mg estradiol valerate (Progynova; Bayer-Schering Pharma AG, Germany) twice daily, which was increased after 5 days to 3 mg twice daily. A transvaginal ultrasound monitoring (SonoScape S15; Kaili Bio-medical technology, China) was performed to assess the endometrium thickness and pattern after 10 days of estradiol administration.

If the endometrium thickness was less than 8 mm, estrogen administration would be continuously increased to 4 mg twice daily with/without vaginal administration of micronized estradiol (Femoston; Abbott, the Netherlands) for another 5 days. Once the endometrium thickness was more than 8 mm, or the endometrium thickness was the thickest compared with all previous fresh and FET cycles, besides exogenous estrogen, 40mg intramuscular progesterone would be injected once daily on the same day and 20mg Dydrogesterone twice daily combined in the following morning, transforming the endometrium to secretary phase. Frozen-thawed blastocyst transfer procedure was planned on the 6th day of progesterone administration.

On the morning of the FET day before thawing the blastocyst, endometrium thickness, pattern and its frequency of peristalsis were assessed by a transvaginal ultrasound. If the endometrium condition was favorable, the frozen blastocyst thawing could be performed. Afterward, the embryo transfer procedure was administrated under abdominal ultrasound guidance, if the quality of the thawed embryo remained well. And 90 mg vaginal progesterone (Crinone 8% gel, Serono, Merck, UK) once per day replaced intramuscular progesterone administration.

In 12 days after blastocyst transfer, the level of serum β-hCG was determined to assess the outcomes of FETs, and luteal phase support medications was continued if an HCG test with a positive result (β-hCG ≥ 10mIU/mL). Clinical pregnancy was confirmed by transvaginal ultrasound in 4–5 weeks after FETs.

Acupoint stimulation protocols

We managed the acupoint stimulation through the application of acupoint medication patches in conjunction with electromagnetic wave heat therapy (AMP-EWH). The acupoints selected for stimulation included CV3(Zhongji), CV4(Guanyuan), CV6(Qihai), bilateral Zigong, GV3(Yaoyangguan), GV4(Mingmen), BL23(Shenshu), and Balao. The production of acupoint medication patches: Curcuma zedoaria (Jiang Huang), Evodia rutaecarpa (Wu Zhu Yu), Syzygium aromaticum (Ding Xiang), Euphorbia helioscopia (Tou Gu Cao), Cinnamomum cassia (Rou Gui), Ligusticum chuanxiong (Chuan Xiong), Angelica sinensis (Dang Gui), in a ratio of 3:3:3:3:1:2:2. The herbs were ground into a fine powder and homogeneously mixed. Fresh ginger juice was then incorporated to form a medicated paste. The ointments were prepared in advance by the Chinese Pharmacy Department of the First Affiliated Hospital of Wenzhou Medical University, one day prior to their intended use. An appropriate number of blank acupoint patches (Air-permeable surgical non-woven tape, Zhejiang Aoqi Medical Dressings Co., Ltd., China) were prepared, with a lentil-sized portion of the ointment placed at the center of each patche. These medicated patches were subsequently affixed to the designated acupoints as mentioned above. Following the application of acupoint patches, localized electromagnetic wave heat therapy (conducted by specific electromagnetic wave therapy device, Chongqing Zhongzhi Medical Instruments Co., Ltd., China) was subsequently administered for 15 minutes exposure. Each session of acupoint patch application was maintained for a period ranging from 4 to 6 hours, after which patients were instructed to remove the patches independently upon returning home.

Patients in group T1 had one session of acupoint stimulation administered exclusively within 2 hours prior to the embryo transfer procedure. Group T2-3 received 2–3 sessions of acupoint stimulation conducted from the day of progesterone administration to 2 hours before blastocyst transfer procedure, during the secretory phase of the endometrium. Group T ≥ 4 had at least 4 sessions of AMP-EWH from the 5th day of exogenous estrogen initiated and continuing until 2 hours prior to the embryo transfer procedure, from the proliferative to secretory phase of the endometrium. Although acupoint stimulation was ideally scheduled every three days, some patients, due to personal scheduling conflicts, occasionally underwent sessions every four to five days. The timeline for the acupoint stimulation sessions, as implemented for five subsets (group T0, T ≥ 1, T1, T2-3, and T ≥ 4), was depicted in Fig. 2.

Statistical analysis

Data analysis was performed using SPSS 27.0 (IBM Corp., Armonk, NY). Multiple imputation was employed to handle the missing data (b-FSH, b-LH, b-E2, and AMH,) in incomplete cases. A total of 5 imputed datasets were generated and the final statistical estimates for missing data presented in our study were the average of the results from the five imputed datasets. We initially assessed multicollinearity using variance inflation factor (VIF) statistics. A VIF greater than 5 is considered indicative of substantial multicollinearity. In our regression models, we strategically elected to omit variables that demonstrated high intercorrelations.

To account for the non-independence of observations within clusters, we utilized Generalized Linear Mixed Models (GLMM), adjusting those independent variables as fixed effects, female age at embryo transfer, male age at embryo cryopreservation, duration of embryo cryopreservation, BMI, type of infertility (primary/secondary), number of previous failed embryo transfer cycles, peripheral serum levels of b-FSH/ b-LH/ b-E2/ AMH, peripheral serum E2 level and endometrium thickness on the day of progesterone administration, endometrium thickness change between the day of progesterone administration and the day of FET, ovulation induction protocol/number of oocytes retrieved/ mature oocytes / high-quality cleavage stage embryos in the the ovarian stimulation cycle from which the transferred blastocyst originated and embryo quality evaluation after thawed (derived from normal fertilization or not/comprehensive score / stage of embryo development); meanwhile, incorporating random effects for the embryo incubator and the physician performing the embryo transfer procedure. Different covariates were selected to construct four distinct regression models for the statistical comparison of pregnancy outcomes.

The baseline characteristics of the women and the outcome parameters of clinical intervention prior to embryo transfer in each group were summarized using counts (%) for categorical variables, and means (SD) or median (IQR) for numeric variables. When conducting those comparisons, we utilize the standardized mean difference (SMD) as a meaningful measure. The SMD value being less than 0.1 indicates that the difference is trivial or negligible, 0.10–0.34 considered a small effect size, more than 0.35 is a moderate effect size, and above 0.65 is a large effect size. For the pregnancy outcomes analysis, we compared the proportions of embryo transfer cycles with β-HCG positivity, biochemical pregnancy loss, ectopic pregnancy and clinical pregnancy between 2 groups individually ( T ≥ 1vs.T0, T1vs.T0, T2-3vs.T0, T2-3 vs. T1, T ≥ 4vs.T0, T ≥ 4vs.T1 and T ≥ 4vs.T2-3, respectively) for all participants, using risk mean differences (MDs) and odds risks (ORs) associated with 95%CIs. Further, in subgroup analyses, we compared the clinical pregnancy rates as referred to group T0 with odds risks (ORs) associated with 95%CIs. Two-sided p < 0.05 indicated statistical significance.

Patient and cycle characteristics

Within our study cohort, a total of 3907 individual patients who met our inclusion criteria had experienced 4863 single blastocyst transfer cycles at our centre during the period from January 2022 to December 2023. Excluding 4515 cycles based on the exclusion criteria. Among the overall population, 3636 cycles (56.1%) were supplemented with acupoints stimulation, while 1227 cycles (43.9%) proceeded without such intervention. Furthermore, based on the number of acupoint stimulation sessions received prior to frozen-thawed blastocyst transfer, the groups were categorized as follows: T1 with 1607 patients, T2-3 with 1600 patients, and T ≥ 4 with 429 patients. The grouping is presented in Fig. 1.

Baseline characteristics: As Table 1 presented, female age (32.31 to 32.73 years) was quite balanced across all groups (p = 0.317, SMD = 0.082). The majority of patients (62.0-68.1%) were between the ages of 29–37. The mean BMI was slightly higher in group T ≥ 4(22.47 kg/m2) compared to group T0(22.22 kg/m2) (p > 0.05), and approximately 69.3–80.2% of overall women had a BMI less than 24.0 kg/m2. The type (primary or secondary) and the duration of infertility were similar in all groups (p > 0.05).

Significantly fewer previous ET failure attempts experienced were noted in groups that received acupoint stimulated compared to the group T0, with a mean of 0.57 vs.0.74, respectively (p < 0.001, SMD = 0.139). Peripheral serum levels of reproductive hormones, including b-FSH, b-LH, and b-E2 were generally consistent among the groups, with only minor variations observed in b-E2 levels (p = 0.039, SMD = 0.119).

In terms of the prior ovarian stimulation cycles, there were no significant differences in female and male ages, as well as AMH levels, across the study groups (p > 0.05, SMD < 0.100). Ovulation induction protocols varied among the groups. The GnRH-antagonist protocol was the most prevalent, being utilized in 81.4–83.4% of cycles across all groups, with a particularly high prevalence in group T ≥ 4. Meanwhile the GnRH-agonist down-regulation protocol used in 8.9% of cycles in group T0, which was slightly higher compared to other groups (p = 0.003, SMD = 0.148). In contrast to groups that received acupoint stimulation, group T0 exhibited a significantly higher number of oocytes retrieved and a greater formation of high-quality Day 3 embryos (p = 0.019, SMD = 0.120; p = 0.045, SMD = 0.138, respectively). No significant differences were noted among the groups in terms of fertilization method and oocytes maturity rate (p = 0.597, SMD = 0.042; p = 0.218, SMD = 0.061, respectively).

When considering the transferred embryos status, the duration of blastocyst cryopreservation was shortest in group T2-3 and longest in group T0, ranging from 6.92 to 8.62 months, the difference had statistical significance (p = 0.013, SMD = 0.086). The majority of transferred blastocyst (93.0-94.4%) were at the D5 stage of development and the remaining 5.6-7.0% were at the D6 stage. No statistically significant differences were found in the comprehensive assessment of embryo quality among all groups, regardless before frozen or after thawed stage (p = 0.872, SMD = 0.027; p = 0.861, SMD = 0.031, respectively). Prior to cryopreservation, approximately 49.1–51.1% of the transferred blastocysts were rated as top quality, 23.7–25.7% as good, and 24.9–25.2% as fair. After thawed, the distribution of embryo quality remained consistent, the top was 49.0-50.1%, the good was 23.3–25.5% and the fair was 25.5–25.8%. Upon detailed comparison of the distribution of embryo expansion grade (3,4 and 5/6), the inner cell mass grade (A, B and C), and the trophectoderm grade (A, B and C) among the groups, no significant differences were identified (p = 0.384, SMD = 0.047; p = 0.455, SMD = 0.036; p = 0.119, SMD = 0.083, respectively). Detailed characteristics are shown in Table 1.

Table 1

Demographic characteristics of study subjects.
	T0 (n = 1227)	T ≥ 1 (n = 3636)	T1 (n = 1607)	T2-3 (n = 1600)	T ≥ 4 (n = 429)	P	SMD
During FET cycles
Female age (y)	32.35(4.41)	32.43(4.38)	32.31(4.44)	32.47(4.36)	32.73(4.23)	0.317	0.082
21–28	304(24.8)	818(22.5)	371(23.1)	361(22.6)	86(20.0)
29–37	761(62.0)	2368(65.1)	1036(64.5)	1040(65.0)	292(68.1)
38–42	162(13.2)	450(12.4)	200(12.4)	199(12.4)	51(11.9)
BMI (kg/m²)	22.22(3.22)	22.35(3.30)	22.34(3.31)	22.34(3.31)	22.47(3.22)	0.058	0.078
< 24.0	925(75.4)	2921(80.3)	1174(73.1)	1158(72.4)	297(69.2)
≥ 24.0	302(24.6)	715(19.7)	423(26.9)	442(27.6)	132(30.8)
Type of infertility						0.712	0.043
primary	393(32.0)	1203(33.1)	542(33.7)	516(32.3)	145(33.8)
secondary	834(68.0)	2433(66.9)	1065(66.3)	1084(67.8)	284(66.2)
Duration of infertility(y)	3.31(2.63)	3.35(2.49)	3.41(2.52)	3.29(2.42)	3.35(2.64)	0.537	0.051
< 5.0	941(76.7)	2777(76.4)	1209(75.2)	1236(77.3)	332(77.4)
≥ 5.0	286(23.3)	859(23.6)	398(24.8)	364(22.8)	97(22.6)
No. of previous failed ET cycles
mean (SD)	0.74(1.12)	0.57(0.97)	0.57(0.94)	0.57(1.00)	0.58(1.04)	< 0.001	0.139
0	687(56.0)	2316(63.7)	1011(62.9)	1028(64.3)	277(64.6)
1	331(27.0)	883(24.3)	407(25.3)	378(23.6)	98(22.8)
2	120(9.8)	252(6.9)	111(6.9)	110(6.9)	31(7.2)
≥ 3	89(7.3)	185(5.1)	78(4.9)	84(5.3)	23(5.4)
Serum level of hormones
b-FSH (IU/L)	7.34(2.29)	7.18(2.09)	7.15(1.98)	7.19(2.17)	7.22(2.20)	0.128	0.089
b-LH (IU/L)	5.76(3.45)	5.57(3.20)	5.60(3.31)	5.52(3.08)	5.67(3.26)	0.264	0.047
b-E2 (pmol/L)	169.25(80.58)	161.47(81.77)	162.01(85.12)	161.29(78.93)	159.82(79.44)	0.035	0.119
Pior ovarian stimulation cycles
Female age (y)	31.69(4.38)	31.89(4.38)	31.75(4.44)	31.95(4.38)	32.21(4.15)	0.102	0.097
21–28	262(21.4)	706(19.4)	326(20.3)	311(19.4)	69(16.1)
29–37	839(68.4)	2560(70.4)	1113(69.3)	1128(70.5)	319(74.4)
38–42	126(10.3)	370(10.2)	168(10.5)	161(10.1)	41(9.6)
Male age (y)	33.70(4.80)	33.87(4.66)	33.85(4.75)	33.83(4.60)	34.12(4.54)	0.473	0.080
AMH (ng/ml)	5.40(4.05)	5.31(4.03)	5.35(3.98)	5.25(4.00)	5.42(4.34)	0.758	0.041
<1.10	72(5.9)	218(6.0)	86(5.4)	102(6.4)	30(7.0)
1.10–2.50	211(17.2)	657(18.1)	286(17.8)	290(18.1)	81(18.9)
2.51-6.00	555(45.2)	1613(44.4)	715(44.5)	716(44.8)	182(42.4)
>6.00	389(31.7)	1148(31.6)	520(32.4)	492(30.8)	136(31.7)
Ovulation induction protocol						0.003	0.148
GnRH-antagonist	999(81.4)	3059(84.1)	1359(84.6)	1342(83.9)	358(83.4)
PPOS	103(8.4)	336(9.2)	130(8.1)	162(10.1)	44(10.3)
GnRH-agonist	109(8.9)	201(5.5)	97(6.0)	82(5.1)	22(5.1)
Mild stimulation	16(1.3)	40(1.1)	21(1.3)	14(0.9)	5(1.2)
No. of oocyte retrieved	17.90(9.72)	17.17(8.88)	17.47(8.67)	17.05(8.88)	16.54(9.57)	0.019	0.120
Method of fertilization						0.597	0.042
IVF	817(66.6)	2368(65.1)	1030(64.1)	1059(66.2)	279(65.0)
ICSI	280(22.8)	822(22.6)	368(22.9)	356(22.3)	98(22.8)
IVF/ICSI	130(10.6)	446(12.3)	209(13.0)	185(11.6)	52(12.1)
Oocyte mature rate	0.85(0.18)	0.86(0.17)	0.86(0.16)	0.86(0.17)	0.86(0.18)	0.218	0.061
Normal fertilization rate	0.76(0.16)	0.74(0.16)	0.74(0.16)	0.74(0.16)	0.75(0.17)	0.039	0.125
No. of high-quality cleavage stage (D3) embryo
mean (SD)	6.15(4.37)	5.84(3.88)	5.98(3.78)	5.77(3.94)	5.57(4.04)	0.045	0.138
≤ 4	505(41.2)	1543(42.4)	664(41.3)	671(41.9)	208(48.5)
≥ 5	722(58.8)	2093(57.6)	943(58.7)	929(58.1)	221(51.5)
Embryos Status
Duration of blastocyst cryopreservation (m)
mean (SD)	8.62(16.74)	7.17(13.49)	7.41(13.99)	6.92(13.12)	6.94(12.96)	0.013	0.086
median (IQR)	3.0(1.0–7.0)	2.0(1.0–6.0)	2.0(1.0–6.0)	2.0(1.0–6.0)	2.0(1.0–6.0)
Stage of embryo development						0.491	0.041
D5	1141(93.0)	3419(94.0)	1505(93.7)	1509(94.3)	405(94.4)
D6	86(7.0)	217(6.0)	102(6.3)	91(5.7)	24(5.6)
Method of fertilization for the transferred blastocyst						0.657	0.022
IVF	878(71.6)	2557(70.3)	1118(69.6)	1138(71.1)	301(70.2)
ICSI	349(28.4)	1079(29.7)	489(30.4)	462(28.9)	128(29.8)
Blastocyst derived from normal fertilization						0.056	0.116
yes	1182(96.3)	3554(97.7)	1573(97.9)	1561(97.6)	420(97.9)
no	45(3.7)	82(2.3)	34(2.1)	39(3.4)	9(2.1)
Embryo quality comprehensive assessment before frozen						0.872	0.027
Top	606(49.4)	1823(50.1)	789(49.1)	818(51.1)	216(50.3)
Good	315(25.7)	897(24.7)	413(25.7)	379(23.7)	105(24.5)
Fair	306(24.9)	916(25.2)	405(25.2)	403(25.2)	108(25.2)
Embryo quality comprehensive assessment post thawed						0.861	0.031
Top	601(49.0)	1819(50.0)	789(49.1)	815(50.9)	215(50.1)
Good	313(25.5)	883(24.3)	406(25.3)	373(23.3)	104(24.2)
Fair	313(25.5)	934(25.7)	412(25.6)	412(25.8)	110(25.6)
Embryo expansion grade post thawed						0.384	0.047
3	171(13.9)	544(15.0)	220(13.7)	251(15.7)	73(17.0)
4	1021(83.2)	3002(82.6)	1339(83.3)	1314(82.1)	349(81.4)
5/6	35(2.8)	90(2.5)	48(3.0)	35(2.2)	7(1.6)
Embryo inner cell mass grade post thawed
A	540(44.0)	1666(45.8)	725(45.1)	756(47.3)	185(43.1)	0.455	0.036
B	685(55.8)	1960(53.9)	876(54.5)	841(52.6)	243(56.6)
C	2(0.2)	10(0.3)	6(0.4)	3(0.2)	1(0.2)
Embryo trophectoderm grade post thawed
A	364(29.7)	1218(33.5)	520(32.4)	555(34.7)	143(33.3)	0.119	0.083
B	552(45.0)	1494(41.1)	681(42.4)	636(39.8)	177(41.3)
C	311(25.3)	924(25.4)	406(25.3)	409(25.6)	109(25.4)
1. Grouping: T0, without any acupoint stimulation; T ≥ 1, with acupoint stimulation at least one time; T1, with acupoint stimulation only one time; T2-3, with acupoint stimulation 2–3 times; T ≥ 1, with acupoint stimulation 4 or more time. 2. Values are shown as mean ± SD, median (IQR) or n (%). 3. Abbreviation: FET, frozen-thawed embryo transfer; BMI, body mass index; b-FSH, basal-follicle stimulate hormone; b-LH, basal-luteal hormone; b-E2, basal-estradiol; AMH, anti-müllerian hormone; IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection; SMD, standard mean difference (statistical effect size). 4. Three embryo quality categories for blastocyst: Top- (AB, AB, or BA), Good- (BB), and Fair (C for inner cell mass or trophectoderm); 5. The red number highlights the P which was less than 0.05.

Outcome parameters of clinical intervention prior to embryo transfer

As presented in Table 2, on the day of progesterone administration, a statistically significant difference in the duration of exogenous estrogen use was observed across the groups (p = 0.011, SMD = 0.101), with group T ≥ 4 exhibiting a longer duration, averaging 14.65 days. Peripheral serum levels of estradiol (E2) and progesterone (P) did not significantly vary among the groups (p = 0.374, SMD = 0.110 for E2; p = 0.137, SMD = 0.071 for P), indicating a relative consistency in hormonal profiles. The mean endometrium thickness across groups was not significantly different (p = 0.773), with a negligible SMD of 0.018. Ultrasound assessment revealed no significant variation in the distribution of endometrial patterns (p = 0.918), with the majority of patients (69.1–70.3%) showing a distinct trend towards type B patterns.

On the day of blastocyst transfer, there were no significant differences observed in the endometrial parameters among the groups. Specifically, 57.3 to 60.3% of patients presented with an endometrium thickness ranging from 9.5 to 14mm, while approximately 93.8–95.8% of women had an endometrial pattern classified as type C. Upon stratification by the frequency of endometrial peristaltic waves per minute, categorized as none, 1–2 times, and ≥ 3 times, a notable divergence was observed in the outcomes between group T0 and those received acupoint stimulation (p = 0.008, SMD = 0.152). This suggests a unique response profile in the absence of acupoint therapy. Mild localized rash was reported as a side effect of AMP-EWH, with no significant difference observed across the acupoint stimulation groups (p = 0.387, SMD = 0.075). Moreover, no severe adverse events related to AMP-EWH were reported during the study.

Table 2

Clinical parameters pre-blastocyst transfer
	T0 (n = 1227)	T ≥ 1 (n = 3636)	T1 (n = 1607)	T2-3 (n = 1600)	T ≥ 4 (n = 429)	P	SMD
On the day of progesterone administration
Duration of exogenous estrogen used, d
mean (SD)	14.10(4.85)	14.15(5.09)	13.85(4.90)	14.32(5.17)	14.62(5.46)	0.011	0.101
median (IQR)	13.0(10.0–16.0)	13.0(10.0–16.0)	12.0(10.0–16.0)	13.0(10.0–16.0)	13.0(10.0–17.0)
Peripheral serum levels of E2, pmol/L
mean (SD)	1998.07(3030.09)	1841.54(2724.86)	1848.97(2605.30)	1819.71(2769.77)	1895.12(2987.29)	0.374	0.110
median (IQR)	687.0(427.0-2648.0)	612.0(405.0-2527.4)	648.0(412.0-2594.0)	596.0(393.3-2349.8)	604.0(407.0-2793.5)
Peripheral serum levels of P, nmol/L
mean (SD)	1.33(0.88)	1.29(0.85)	1.31(0.86)	1.27(0.81)	1.34(0.94)	0.137	0.071
median (IQR)	1.13(0.68–1.76)	1.12(0.66–1.68)	1.12(0.66–1.74)	1.11(0.65–1.63)	1.17(0.66–1.66)
Endometrial thickness (EMT1), mm
mean (SD)	9.52(1.72)	9.51(1.64)	9.52(1.69)	9.50(1.59)	9.51(1.65)	0.773	0.018
<7.0	31(2.5)	78(2.1)	34(2.1)	36(2.3)	8(1.9)
7.0-8.4	293(23.9)	870(23.9)	393(24.5)	375(23.4)	102(23.8)
8.5–13.0	852(69.4)	2568(70.6)	1122(69.8)	1144(71.5)	302(70.4)
>13.0	51(4.2)	120(3.3)	58(3.6)	45(2.8)	17(4.0)
Ultrasound assessment of endometrium pattern						0.918	0.043
Type A	364(29.7)	1114(30.6)	496(30.9)	486(30.4)	132(30.8)
Type B	863(70.3)	2522(69.4)	1111(69.1)	1114(69.6)	297(69.2)
On the day of blastocyst transfer
Endometrial thickness (EMT2), mm						0.272	0.060
mean (SD)	10.19(2.03)	10.29(1.97)	10.33(2.00)	10.25(1.92)	10.26(2.02)
< 7.0	38(3.1)	73(2.0)	25(1.6)	37(2.3)	11(2.6)
7.0-9.4	435(35.5)	1270(34.9)	566(35.2)	549(34.3)	155(36.1)
9.5–14.0	719(58.6)	2167(59.6)	956(59.5)	965(60.3)	246(57.3)
> 14.0	35(2.9)	126(3.5)	60(3.7)	49(3.1)	17(4.0)
Ultrasound assessment of endometrium pattern						0.078	0.090
Type B	51(4.2)	198(5.4)	100(6.2)	76(4.8)	22(5.1)
Type C	1176(95.8)	3438(94.6)	1507(93.8)	1524(95.3)	407(94.9)
Endometrial thickness change						0.061	0.100
Compaction	107(8.7)	246(6.8)	104(6.5)	112(7.0)	30(7.0)
No change	539(43.9)	1524(41.9)	674(41.9)	659(41.2)	191(44.5)
Expansion	581(47.4)	1866(51.3)	829(51.6)	829(51.8)	208(48.5)
Frequency of endometrial peristalsis, per minute						0.008	0.152
0	984(80.2)	2750(75.6)	1239(77.1)	1195(74.7)	316(73.7)
1–2	199(16.2)	807(22.2)	326(20.3)	377(23.6)	104(24.2)
≥ 3	44(3.6)	79(2.2)	42(2.6)	28(1.8)	9(2.1)
Side effects of acupoint therapy
Mild rash	NA	56(1.5)	20(1.2)	21(1.3)	9(2.1)	0.387	0.075
Severe reaction	NA	0	0	0	0
1. Endometrial thickness (EMT) was defined as the maximal distance between the endometrial and myometrial junctions at about 1.5 cm from the uterine fundus through the median sagittal section of the uterus. 2. The EMT on the day of progesterone administration and on ET day were recorded as EMT1 and EMT2, respectively. Based on EMT change ratio= the patients were divided into three groups: the endometrial compaction group (the ratio≤-5%), non-change group ( -5% ~ +5%), and expansion group ( ≥ + 5%). 3. Endometrium pattern, Type A (Trilinear or Multilayered Endometrium), characterized by a distinct trilinear appearance or a multilayered endometrial structure, a strong echo in the outer layers, hypoechoic or anechoic areas, which are darker compared to the surrounding tissue in the innermost layer typically, and an unmistakable linear echo visible within the uterine cavity. Type B (Weak Trilinear), the trilinear pattern is less pronounced, with a weak appearance. There is an isolated echo present in the middle layer, and the echoes within the uterine cavity are inconspicuous or not easily discernible. Type C (Homogeneous Pattern), marked by a strong echo throughout, indicating a more homogeneous structure, and absence of an intrauterine midline echo, suggesting a uniform appearance without the distinct layers seen in the other types. 4. NA: not available.

Four Reproductive outcome endpoints in entire population

In our study, the crude model and 3 adjusted models (Model 1, Model 2 and PSM) were engaged in investigation, aiming to mitigate the influence of potential confounders, rigorously evaluate and compare the efficacy of various treatment strategies on reproductive outcomes, thereby enhancing the validity and generalizability of our findings. See Table 3 and Fig. 3,4 for details, with group T0 serving as the control.

Tabel 3 Statistical analysis of pregnancy outcomes among the groups, using 4 Models. Group T0 serves as the reference the control point for all pregnancy outcome indicators, with values presented as odds ratios alongside their 95% confidence intervals, unless otherwise indicated.
	T0 (n = 1227)	T ≥ 1 (n = 3636)	P1	T1 (n = 1607)	P2	T2-3 (n = 1600)	P3	T ≥ 4 (n = 429)	P4
HCG positive
No. of event	867	2660		1187		1146		327
Risk estimate, % (95%CI)	70.7 (68.2 to 73.2)	73.2 (71.8 to 74.6)		73.9 (71.7 to 76.1)		71.6 (69.4 to 73.8)		76.2 (72.1 to 80.3)
Mean difference, % (95% CI)	0	2.5 (-0.4 to 5.4)		3.2 (-0.1 to 6.5)		0.9 (-2.3 to 4.3)		5.6 (0.7 to 10.5)
Model Crude	1	1.132 (0.981 to 1.306)	0.090	1.174 (0.994 to 1.385)	0.059	1.048 (0.890 to 1.235)	0.574	1.331 (1.033 to 1.716)	0.027
Model 1	1	1.100 (0.948 to 1.277)	0.209	1.146 (0.964 to 1.361)	0.122	1.006 (0.848 to 1.194)	0.944	1.329 (1.021 to 1.729)	0.034
Model 2	1	1.113 (0.958 to 1.293)	0.163	1.158 (0.974 to 1.377)	0.097	1.020 (0.858 to 1.211)	0.826	1.342 (1.030 to 1.750)	0.030
Model PSM	1	1.102 (0.915 to 1.327)	0.305	1.147 (0.952 to 1.382)	0.149	0.980 (0.814 to 1.181)	0.834	1.561 (1.107 to 2.200)	0.011
Biochemical pregnancy loss
No. of event	131	325		148		140		37
Risk estimate, % (95%CI)	10.7 (8.9 to 12.5)	8.9 (7.9 to 9.9)		9.2 (7.8 to 10.6)		8.8 (7.4 to 10.2)		8.6 (5.9 to 11.3)
Mean difference, % (95% CI)	0	-1.7 (-3.7 to 0.3)		-1.5 (-3.7 to 0.7)		-1.9 (-4.1 to 0.3)		-2.1 (-5.2 to 1.0)
Model Crude	1	0.821 (0.663 to 1.017)	0.071	0.849 (0.662 to 1.087)	0.194	0.802 (0.624 to 1.031)	0.085	0.790 (0.539 to 1.158)	0.227
Model 1	1	0.836 (0.673 to 1.037)	0.103	0.857 (0.668 to 1.099)	0.224	0.824 (0.640 to 1.061)	0.134	0.796 (0.542 to 1.170)	0.246
Model 2	1	0.832 (0.670 to 1.033)	0.096	0.853 (0.665 to 1.096)	0.214	0.821 (0.637 to 1.058)	0.128	0.790 (0.537 to 1.161)	0.230
Model PSM	1	0.773 (0.589 to 1.014)	0.063	0.871 (0.668 to 1.136)	0.310	0.764 (0.577 to 1.012)	0.072	0.708 (0.436 to 1.151)	0.164
Ectopic pregnancy
No. of event	2	17		9		7		1
Risk estimate, % (95%CI)	0.2 (-0.03 to 0.4)	0.5 (0.3 to 0.7)		0.6 (0.2 to 1.0)		0.4 (0.05 to 0.8)		0.2 (-0.2 to 0.6)
Mean difference, % (95% CI)	0	0.3 (-0.1 to 0.7)		0.4 (0.0 to 0.8)		0.3 (-0.1 to 0.7)		0.07 (-0.5 to 0.6)
Model Crude	1	2.877 (0.664 to 12.47)	0.158	3.450 (0.744 to 15.99)	0.114	2.691 (0.558 to 12.98)	0.217	1.431 (0.129 to 15.822)	0.770
Model 1	1	1.055 (0.721 to 1.544)	0.783	1.073 (0.695 to 1.655)	0.751	1.049 (0.678 to 1.623)	0.829	1.011 (0.529 to 1.932)	0.973
Model 2	1	1.056 (0.721 to 1.547)	0.780	1.074 (0.696 to 1.658)	0.746	1.049 (0.678 to 1.624)	0.829	1.013 (0.530 to 1.936)	0.969
Model PSM	1	1.063 (0.669 to 1.691)	0.796	1.078 (0.678 to 1.715)	0.750	1.024 (0.641 to 1.637)	0.919	1.005 (0.453 to 2.232)	0.990
Clinical pregnancy
No. of event	734	2318		1030		999		289
Risk estimate, % (95%CI)	59.8 (57.1 to 62.5)	63.8 (62.2 to 65.4)		64.1 (61.8 to 66.5)		62.4 (60.0 to 64.8)		67.4 (62.9 to 71.9)
Mean difference, % (95% CI)	0	3.9 (0.8 to 7.0)		4.3 (0.8 to 7.8)		2.6 (-0.9 to 6.1)		7.5 (2.2 to 12.8)
Model Crude	1	1.181 (1.035 to 1.349)	0.014	1.199 (1.029 to 1.397)	0.020	1.116 (0.959 to 1.300)	0.157	1.387 (1.100 to 1.748)	0.006
Model 1	1	1.145 (0.996 to 1.315)	0.057	1.173 (0.999 to 1.376)	0.051	1.062 (0.905 to 1.247)	0.460	1.380 (1.084 to 1.758)	0.009
Model 2	1	1.157 (1.006 to 1.331)	0.041	1.184 (1.008 to 1.391)	0.040	1.076 (0.916 to 1.265)	0.373	1.394 (1.093 to 1.778)	0.008
Model PSM	1	1.199 (1.008 to 1.427)	0.041	1.170 (0.985 to 1.390)	0.074	1.069 (0.898 to 1.274)	0.454	1.672 (1.219 to 2.295)	0.001
1. P1 designated the statistical significance of the pregnancy outcomes of group T ≥ 1 vs. group T0; P2, group T1 vs. group T0; P3, group T2-3 vs. group T0; P4, group T ≥ 4 vs. group T0. 2. Mean difference was calculated with reference to the incidence rate of the group T0, representing the comparative risk difference between group T0 and the others. 3. Multifactorial binary logistic regression analysis was performed by general linear mixed model (GLMM). 3 models (Model 1, Model 2 and Model PSM) employed their respective covariate adjustments. 4. Model 1: adjusted for female age at FET, number of previous failed ET cycles, number of high quality D3 embryo, stage of blastocyst D5/D6, embryo quality comprehensive assessment post thawed as fixed effects; embryo incubator and embryo transfer as random effects. 5. Model 2: building upon Model 1, additional covariates as fixed effects: duration of infertility, BMI, type of infertility, male age at embryo cryopreservation, AMH, number of oocytes retrieved in prior cos cycle, duration of transferred blastocyst cryopreservation, the transferred blastocyst derived from normal fertilization (yes or no), endometrial thickness on the day of progesterone administration, while retaining the random effects as Model 1. 6. Model PSM: initially, conducting propensity score matching 1:1 between T ≥ 1 and T0, T1 and T0, T2-3 and T0, then T ≥ 4 and T0, individually, utilizing these covariates for the matching process: female age at FET, number of previous failed ET cycles, number of high quality D3 embryo, stage of blastocyst D5/D6, embryo quality comprehensive assessment post thawed age, duration of infertility, BMI, type of infertility, male age at embryo cryopreservation, AMH, number of oocytes retrieved in prior cos cycle, duration of transferred blastocyst cryopreservation, the transferred blastocyst derived from normal fertilization(yes or no). Subsequently, we performed GLMM on each matched set, incorporating the matched covariates as fixed effects and considering embryo incubator and embryo transfer as random effects. 7. The red number highlights the P which was less than 0.05.

β-HCG positivity: In current study, group T0, T ≥ 1, T1, T2-3 and T ≥ 4 had this event rates of 70.7%, 73.2%, 73.9%, 71.6% and 76.2%, respectively. When contrasted with Group T0, the individual mean difference(95%CI) of T ≥ 1, T1, T2-3 and T ≥ 4 was 2.5 ( -0.4 to 5.4), 3.2 ( -0.1 to 6.5), 0.9 ( -2.3 to 4.3) and 5.6 (0.7 to 10.5), respectively. Group T ≥ 4 consistently displayed an absolute advantage of higher HCG positivity rate across all four models examined (OR, 1.331 [95% CI, 1.033 to 1.716], p = 0.027 in Model crude; 1.329 [1.021 to 1.729], p = 0.034 in Model 1; 1.342 [1.030 to 1.750], p = 0.030 in Model 2; 1.561 [1.107 to 2.200], p = 0.011 in Model PSM). While an increasing trend was observed in the odds ratios (ORs) for the other acupoint stimulation groups (T ≥ 1, T1 and T2-3) compared to the baseline group T0, these did not attain statistical significance(p > 0.05), regardless which model used for analysis.

Biochemical pregnancy loss: The biochemical pregnancy loss rates were assessed comparatively among all groups. The risk estimates for biochemical pregnancy were as follows: group T0 at 10.7% (95% CI: 8.9 to 12.5), group T ≥ 1 8.9% (7.9 to 9.9), group T1 9.2% (7.8 to 10.6), group T2-3 8.8% (7.4 to 10.2), and group T ≥ 4 8.6% (5.9 to 11.3). The mean difference relative to group T0 was − 1.7% (95% CI: -3.7 to 0.3) for T ≥ 1, -1.5% (-3.7 to 0.7) for T1, -1.9% (-4.1 to 0.3) for T2-3, and − 2.1% (-5.2 to 1.0) for T ≥ 4, indicating a general trend of lower biochemical pregnancy loss rates in the acupoint stimulation groups. In the Model crude analyses, with group T0 as control, the odds ratios (ORs) for T ≥ 1, T1, T2-3 and T > 4 were 0.821 (95% CI: 0.663 to 1.017), 0.849 (0.662 to 1.087), 0.802 (0.624 to 1.031) and 0.790 (0.539 to 1.158), respectively. After adjustments in Model 1, Model 2 and Model PSM, similar patterns emerged, with slightly varied ORs but consistently no significant difference.

Ectopic pregnancy

The rates of ectopic pregnancy were assessed across groups T0, T ≥ 1, T1, T2-3, and T ≥ 4, with respective risk estimates of 0.2%, 0.5%, 0.6%, 0.4%, and 0.2%. In the initial Model crude assessment, the odds ratios (ORs) were 2.877(95%CI, 0.664 to 12.471), p = 0.158 for group T ≥ 1; 3.45 (0.744 to 15.995), 0.114 for group T1; 2.691 (0.558 to 12.979), 0.217 for group T2-3 and 1.431 (0.129 to 15.822), 0.770 for group T ≥ 4. However, these associations were not statistically significant. Subsequent adjusted analyses utilizing Model 1, Model 2 and Model PSM all yielded ORs with p-values more than 0.745, suggesting no significant differences in ectopic pregnancy rates.

Clinical pregnancy

Individually, group T0, T ≥ 1, T1, T2-3 and T ≥ 4 exhibited a clinical pregnancy risk estimate (95%CI) of 59.8% (57.1–62.5%), 63.8% (62.2–65.4%), 64.1% (61.8–66.5%), 62.4% (60.0–64.8%) and 67.4% (62.9–71.9%), respectively. Furthermore, in contrasting to group T0, the mean difference (95%CI) of T ≥ 1, T1, T2-3 and T ≥ 4 was 3.9% (0.8–7.0%), 4.3% (0.8–7.8%), 2.6% (-0.9–6.1%) and 7.5% (2.2–12.8%), respectively. Group T ≥ 4 showed the highest clinical pregnancy rate (CPR), indicating a statistically significant increase compared to group T0.

Upon statistical analyses using various models, group T ≥ 1 showed a significant advantage in CPR over group T0 in model crude, model 1 and model PSM, but not in model 2, with OR (95%CI) and p value as follows: 1.181 (1.035 to 1.349), p = 0.014 in model crude; 1.157 (1.006 to 1.331), p = 0.041 in model 1; 1.199 (1.008 to 1.427), p = 0.041 in model 2 and 1.145 (0.996 to 1.315), p = 0.057 in model PSM. Group T1 demonstrated a slightly higher rate conducted by Model 1 and Model PSM (OR, 1.173 [95%CI, 0.999 to 1.376], p = 0.051; 1.170 [0.985 to 1.390], p = 0.074, respectively). Moreover, group T1 exhibited a statistically significant increase in CPR as indicated by Model crude and Model 2 (OR, 1.199 [95%CI, 1.029 to 1.397], p = 0.020; 1.184 [1.008 to 1.391], p = 0.074, respectively). Group T2-3 exhibited a suggestive trend towards an increased CPR; however, this elevation did not reach statistical significance regardless which analytical approach employed. The effect sizes for group T2-3 as compared to group T0 were as follows: OR, 1.116 (95%CI, 0.959 to 1.300), p = 0.157 according to the crude model; 1.062 (0.905 to 1.247), 0.460, Model 1; 1.076 (0.916 to 1.265), 0.373, Model 2; 1.069 (0.898 to 1.274), 0.454, Model PSM.

The incidence of clinical pregnancy, a critical outcome measure in our study, was found to be highest in group T ≥ 4 among all groups. Notably, when compared to the baseline group T0, the odds ratios (ORs) from various models all pointed towards a robust increased likelihood of clinical pregnancy in group T ≥ 4. Model crude demonstrated an OR,1.387(95%CI,1.100 to 1.748) and a p-value of 0.006. Upon adjustment using Model 1, the OR slightly attenuated to 1.380 but remained statistically significant (p = 0.009). Further adjustment in Model 2 yielded an OR (95%) of 1.394 (1.093 to 1.778), reinforcing the robustness of the association with a p-value of 0.008. Additionally, the Model PSM demonstrated a notably impressive OR (95%) of 1.672(1.219 to 2.295) and a p-value of 0.001, further supporting the strong association.

Pairwise comparisons of pregnancy outcomes administrated among the acupoint stimulation groups (group T1, T2-3, and T ≥ 4).

See Table 4 for details. For the HCG positivity rates, no significant differences were observed between group T2-3 versus T1 and group T ≥ 4 versus T1, regardless which Model employed for analyses. However, a significant increase was detected between T ≥ 4 and T2-3, with adjusted ORs 1.317 (95%CI, 1.019 to 1.702), p = 0.035, in Model 1 and 1.311(1.013 to 1.697), p = 0.013, in Model 2. The biochemical pregnancy loss rates showed no significant differences across all pairwise comparisons in all models (p-values ranging from 0.649 to 0.986). Ectopic pregnancy rates were not significantly different between any of the groups in any model, with p-values consistently above the threshold for significance (p = 0.375 to 0.913). In terms of clinical pregnancy incidence, a non-significant trend was observed for T2-3 versus T1 and T ≥ 4 versus T1. Notably, group T ≥ 4 versus T2-3 showed a significant difference with adjusted ORs 1.298 (95%CI, 1.026 to 1.641), p = 0.029, in Model 1 and 1.295(1.023 to 1.640), p = 0.022, in Model 2.

Table 4

Statistical analysis of pregnancy outcomes among the acupoint stimulation groups (group T1, T2-3, and T ≥ 4) using 4 Models. Mean difference and odds ratios (ORs) for pairwise comparisons were presented alongside their respective 95% confidence intervals (95% CIs), served as the metrics for these assessments.
	T2-3 vs. T1		T ≥ 4 vs. T1		T ≥ 4 vs. T2-3
	OR (95%CI)	P	OR (95%CI)	P	OR (95%CI)	P
HCG positivity rate
Mean difference, %(95%CI)	-2.2(-5.0 to 1.3)	0.152	2.4(-2.0 to 7.3)	0.327	4.6(-0.03 to 9.4)	0.056
Model Crude	0.893(0.765 to 1.044)	0.155	1.134(0.885 to 1.454)	0.320	1.270(0.992 to 1.627)	0.058
Model 1	0.882(0.750 to 1.037)	0.129	1.162(0.898 to 1.502)	0.254	1.317(1.019 to 1.702)	0.035
Model 2	0.884(0.751 to 1.042)	0.142	1.160(0.895 to 1.502)	0.262	1.311(1.013 to 1.697)	0.013
Model PSM	0.874(0.740 to 1.032)	0.112	1.168(0.824 to 1.656)	0.381	1.305(0.932 to 1.838)	0.121
Biochemical pregnancy loss rate
Mean difference(95%CI)	-0.5(-2.0 to 3.9)	0.648	-0.6(-4.0 to 2.4)	0.706	-0.1(-3.0 to 3.0)	0.936
Model Crude	0.945(0.742 to 1.205)	0.649	0.930(0.638 to 1.357)	0.708	0.984(0.674 to 1.438)	0.935
Model 1	0.970(0.759 to 1.238)	0.804	0.937(0.641 to 1.368)	0.734	0.966(0.660 to 1.413)	0.858
Model 2	0.975(0.763 to 1.245)	0.840	0.936(0.640 to 1.367)	0.730	0.959(0.655 to 1.405)	0.831
Model PSM	0.998(0.778 to 1.280)	0.986	0.823(0.506 to 1.339)	0.432	1.014(0.615 to 1.672)	0.955
Ectopic pregnancy rate
Mean difference(95%CI)	-0.1(-1.0 to 0.04)	0.611	-0.3(-1.0 to 0.03)	0.378	-0.2(-1.0 to 0.9)	0.582
Model Crude	0.780(0.290 to 2.101)	0.623	0.415(0.052 to 3.283)	0.405	0.532(0.065 to 4.333)	0.555
Model 1	0.978(0.657 to 1.456)	0.912	0.942(0.506 to 1.756)	0.851	0.964(0.517 to 1.798)	0.907
Model 2	0.975(0.655 to 1.451)	0.901	0.942(0.505 to 1.755)	0.850	0.966(0.518 to 1.802)	0.913
Model PSM	0.982(0.655 to 1.473)	0.932	0.937(0.424 to 2.075)	0.872	1.014(0.449 to 2.288)	0.973
Clinical pregnancy rate
Mean difference(95%CI)	-1.7(-5.0 to 2.0)	0.329	3.3(-2.0 to 8.2)	0.211	4.9(-0.01 to 10.3)	0.059
Model Crude	0.931(0.806 to 1.075)	0.330	1.156(0.922 to 1.450)	0.208	1.242(0.991 to 1.556)	0.060
Model 1	0.909(0.782 to 1.057)	0.217	1.180(0.933 to 1.493)	0.167	1.298(1.026 to 1.641)	0.029
Model 2	0.911(0.782 to 1.059)	0.225	1.179(0.931 to 1.494)	0.172	1.295(1.023 to 1.640)	0.022
Model PSM	0.894(0.767 to 1.044)	0.156	1.264(0.918 to 1.742)	0.151	1.255(0.915 to 1.721)	0.159
1. Model 1, Model 2 and Model PSM, employing their individual covariate adjustments as detailed in Table 3, were all performed by general linear mixed model (GLMM) for multifactorial binary logistic regression analysis. 2. The red number highlights the P which was less than 0.05.

Subgroup-Specific Findings of Clinical Pregnancy Rates

We conducted an in-depth analysis of the CPRs, comparing group T1, T2-3, and T ≥ 4 to the reference group T0 individually. The multifactorial adjusted Model 2 as mentioned above was utilized to calculate odds ratios (ORs) along with their corresponding 95% confidence intervals (CIs), as detailed in Table 5 and Fig. 5.

Table 5

Stratifying the population into subgroups, the CPR of each group (T0, T1, T2-3, and T ≥ 4) was evaluated individually. Odds Ratios (OR) for pairwise comparisons, derived from the multifactorial adjusted Model 2 alongside their respective 95% confidence intervals (95% CI), were served as the metrics for this assessment.
	Events of clinical pregnancy / Total No. (%)				Multifactorial Binary Logistic Regression Analysis (by GLMM-Model 2)
	T0	T1	T2-3	T ≥ 4	T1 vs. T0		T2-3 vs. T0		T ≥ 4 vs. T0
	T0	T1	T2-3	T ≥ 4	OR (95%CI)	P	OR (95%CI)	P	OR (95%CI)	P
Overall	734/1227(59.8)	1030/1607(64.1)	999/1600(62.4)	289/429(67.4)	1.184(1.008 to 1.391)	0.040	1.076(0.916 to 1.265)	0.373	1.394 (1.093 to 1.778)	0.008
Subgroups
Female age at blastocyst transfer
21-28y	205/304(67.4)	269/371(72.5)	264/361(73.1)	70/86(81.4)	1.316(0.933 to 1.854)	0.117	1.291(0.913 to 1.826)	0.148	2.134(1.163 to 3.916)	0.014
29-37y	456/761(59.9)	674/1036(65.1)	648/1040(62.3)	204/292(69.9)	1.199(0.981 to 1.465)	0.077	1.053(0.862 to 1.288)	0.612	1.552(1.148 to 2.098)	0.004
38-42y	73/162(45.1)	87/200(43.5)	87/199(43.7)	15/51(29.4)	1.038(0.662 to 1.629)	0.870	0.990(0.628 to 1.560)	0.965	0.543(0.262 to 1.125)	0.100
BMI
< 24.0 kg/m²	554/925(59.9)	729/1174(62.1)	704/1158(60.8)	199/297(67.0)	1.107(0.919 to 1.333)	0.284	1.015(0.841 to 1.224)	0.876	1.418(1.062 to 1.894)	0.018
≥ 24.0 kg/m²	180/302(59.6)	301/433(69.5)	295/442(66.7)	90/132(68.2)	1.504(1.083 to 2.089)	0.015	1.280(0.925 to 1.769)	0.136	1.369(0.868 to 2.159)	0.177
No. of previous failed ET cycles
0–1	646/1018(63.5)	932/1418(65.7)	902/1406(64.2)	264/375(70.4)	1.095(0.919 to 1.305	0.308	1.001(0.839 to 1.195)	0.990	1.404(1.073 to 1.830)	0.013
2	52/120(43.3)	61/111(55.0)	53/110(48.2)	17/31(54.8)	1.887(1.044 to 3.411)	0.036	1.155(0.645 to 2.067)	0.627	1.567(0.638 to 3.845)	0.326
≥ 3	36/89(40.4)	37/78(47.4)	44/84(52.4)	8/23(34.8)	2.121(1.016 to 4.427)	0.045	2.802(1.339 to 5.861)	0.006	1.223(0.407 to 3.678)	0.719
Endometrium thickness on the day of progesterone administration
<7.0 mm	12/31(38.7)	16/34(47.1)	14/36(38.9)	5/8(62.5)	1.660(0.429 to 6.417)	0.458	0.568(0.146 to 2.207)	0.410	2.994(0.331 to 27.08)	0.325
7.0–13.0 mm	686/1145(59.9)	971/1515(64.1)	953/1519(62.7)	272/404(67.3)	1.178(1.005 to 1.398)	0.046	1.082(0.913 to 1.277)	0.356	1.404(1.093 to 1.805)	0.008
>13.0 mm	36/51(70.6)	43/58(74.1)	32/45(71.1)	12/17(70.6)	1.671(0.588 to 4.750)	0.333	0.891(0.314 to 2.525)	0.827	1.347(0.292 to 6.214)	0.701
Anti-Müllerian Hormone
<2.5 ng/ml	146/278(52.5)	216/365(59.2)	209/388(53.9)	69/109(63.3)	1.244(0.889 to 1.741)	0.203	0.948(0.681 to 1.319)	0.751	1.457(0.896 to 2.371)	0.129
≥ 2.5 ng/ml	588/949(62.0)	814/1242(65.5)	790/1212(65.2)	220/320(68.8)	1.153(0.959 to 1.386)	0.129	1.118(0.928 to 1.347)	0.239	1.366(1.029 to 1.812)	0.031
Stage of embryo development
D5	707/1141(62.0)	986/1505(65.5)	971/1509(64.3)	279/405(68.9)	1.150(0.973 to 1.359)	0.101	1.068(0.903 to 1.262)	0.442	1.363(1.060 to 1.752)	0.016
D6	27/86(31.4)	44/102(43.1)	28/91(30.8)	10/24(41.7)	1.911(0.998 to 3.657)	0.051	1.285(0.647 to 2.554)	0.472	2.176(0.815 to 5.811)	0.121
Embryo quality assessment
Top	406/601(67.6)	579/789(73.0)	577/815(70.8)	157/215(73.0)	1.267(0.995 to 1.613)	0.054	1.124(0.885 to 1.427)	0.339	1.366(0.954 to 1.955)	0.088
Good	186/313(59.4)	25./406(61.6)	224/373(60.1)	75/104(72.1)	1.032(0.754 to 1.412)	0.844	0.971(0.705 to 1.337)	0.858	1.681(1.018 to 2.776)	0.042
Fair	142/313(45.4)	204/412(49.5)	198/412(48.1)	57/110(51.8)	1.187(0.872 to 1.615)	0.275	1.114(0.819 to 1.516)	0.490	1.355(0.858 to 2.140)	0.193
1. Model 2 was performed by general linear mixed model (GLMM) for multifactorial binary logistic regression analysis, the fixed effects and random effects as mentioned in Table 3. 2. The red number highlights the P which was less than 0.05.

Stratified by female age at embryos transfer

Within the age subgroup of 21–28 years, groups T1 and T2-3 showed non-significant increases in CPR, with ORs of 1.316 (95% CI, 0.933 to 1.854), p = 0.117 and 1.291 (0.913 to 1.826), p = 0.148, respectively. Group T ≥ 4 demonstrated a significantly higher CPR, with an OR 2.134 (95% CI, 1.163 to 3.916), p = 0.014. In the age of 29–37 years, findings was similar with the subgroup of 21–28 years, group T ≥ 4 showed a distinct higher CPR of OR 1.552 (95%CI, 1.148 to 2.098), p = 0.004, and for groups T1 and T2-3, there was an increasing trend, with ORs of 1.199 (95% CI, 0.981 to 1.465), p = 0.077, and 1.053 (0.862 to 1.288), p = 0.612, respectively, although these increases did not reach statistical significance. Regarding the age subgroup of 38–42 years, group T ≥ 4 was found to lower than group T0, with an OR of 0.543 (95% CI, 0.262 to 1.125), p = 0.100. Furthermore, the CPRs for groups T1 and T2-3 were comparable to those of group T0 (p = 0.870, 0.965, respectively) in women aged 38–42 years.

Stratified by BMI Categories

In terms of BMI (kg/m²), individuals under 24.0kg/m² of groups T1 and T2-3 did not showed a significant difference from group T0 when CPR outcomes compared, with ORs of 1.107 (95% CI, 0.919 to 1.333), p = 0.284, and 1.015 (0.841 to 1.224), p = 0.876, respectively. Group T ≥ 4 showed a significant increase,with OR 1.418 (95% CI, 1.062 to 1.894), p = 0.018. In the subgroup of 24.0 kg/m² and above, group T1 demonstrated a significant increase with OR 1.504 (95% CI, 1.083 to 2.089), p = 0.015. And both the T2-3 and T ≥ 4 groups demonstrated an increasing trend, with ORs of 1.280 (95% CI, 0.925 to 1.769), p = 0.136 and 1.369 (0.868 to 2.159), p = 0.177, respectively.

Stratified by number of previous failed ET cycles

Across all subgroups stratified by the number of previous failed ET cycles, the acupoint stimulation groups consistently exhibited an upward trend. Particularly, notable improvements were observed in individuals with 0–1 failed cycles, where group T ≥ 4 demonstrated a pronounced effect, with an OR of 1.404 (95% CI, 1.073 to 1.830), p = 0.013. In the subgroup with 2 failed cycles, group T1 exhibited a significantly superiority with an OR of 1.887 (95% CI, 1.044 to 3.411), p = 0.036. Similarly, for individuals facing 3 or more failed cycles, both groups T1 and T2-3 displayed significant advantages, with ORs of 2.121 (95% CI, 1.016 to 4.427), p = 0.045, and 2.802 (1.339 to 5.861), p = 0.006, respectively.

Stratified by endometrium thickness on the day of progesterone administration

Among patients with an endometrium thickness < 7.0mm, both groups T1 and T ≥ 4 demonstrated ORs of 1.660 (95% CI, 0.429 to 6.417), p = 0.458, and 2.994 (0.331 to 27.08), p = 0.325, respectively. Group T2-3 displayed a decreasing trend with an OR of 0.568 (95% CI, 0.146 to 2.207), p = 0.4102. In patients with an endometrium thickness between 7.0-13.0mm, all three groups receiving acupoint treatment showed varying degrees of increased CPR. Particularly, Groups T1 and T ≥ 4 exhibited outstanding effectiveness in improving CPR outcomes with ORs of 1.184 (95% CI, 1.003 to 1.398) p = 0.046 and 1.404(1.093 to 1.805), p = 0.008, respectively. In the subset of patients with an endometrium thickness > 13.0mm, both group T1 and T ≥ 4 demonstrated a similar trend toward increased CPR, while group T2-3 showed a decreasing trend, and none of statistical significance was attained in either of the comparisons to group T0.

Stratified by anti-Müllerian hormone (AMH) level

Subgroup analysis based on AMH levels revealed no statistically significant differences in clinical pregnancy outcomes across groups, with the exception of the subgroup with AMH levels ≥ 2.5 ng/mL, where group T ≥ 4 exhibited a significant advantage, with OR of 1.366(95%CI,1.029 to 1.812), p = 0.031.

Stratified by stage of embryo development

In both D5 and D6 embryos, all three groups subjected to acupoint stimulation demonstrated a trend toward higher CPR. Notably, for D5 blastocyst transfer cycles, group T ≥ 4 manifested a statistically significant superiority with an OR of 1.363 (95% CI, 1.060 to 1.752), p = 0.016. However, for D6 blastocyst transfer cycles, the comparative advantage of each group (T1, T2-3, or T ≥ 4) over group T0 did not achieve statistical significance, potentially due to the limited sample size.

Stratified by embryo quality assessment

The acupoint treatment groups all enhanced CPR versus the control group T0 across various levels of embryo quality assessment which was stratified as top, good, and fair, with the exception of group T2-3 in the good-quality embryo with an OR of 0.971, (95%CI,0.705 to 1.337), p = 0.858. Conversely, in this good category, group T ≥ 4 exhibited a remarkable advantage over group T0 in improving CPR with an OR of 1.681 (95% CI, 1.018 to 2.776), p = 0.042.

In this retrospective clinical cohort study involving women undergoing single frozen-thawed blastocyst transfer process, administration of AMP-EWH vs without any acupoint stimulation did result in statistically significantly different clinical pregnancy rates validated through multiple mixed effects models. Especially, when acupoint stimulation is administered throughout the full course of endometrium preparation (from the proliferative to secretory phase) for FET, interventions with at least four sessions are more likely to demonstrate promoting favorable clinical pregnancy outcomes. Furthermore, among stratified women, subgroup analyses revealed inconsistent findings across overall patient-reported clinical outcomes, and then it would identify the target population for each acupoint stimulation regimen. These findings do support the use of AMP-EWH to improve clinical pregnancy rates among women aged < 43 years undergoing single frozen-thawed blastocyst transfer, if employing distinct intervention strategies tailored to specific populations could enhance clinical pregnancy outcomes more.

Findings from our study support a recent meta-analysis including 14 RCTs with a total of 1,130 participants(11) that found significant effects of acupuncture adjuvant to FET on the outcomes of clinical pregnancy rate (RR = 1.54,95%CI [1.28, 1.85], I2 = 34%; 14 trials), biochemical pregnancy rate (RR = 1.51,95%CI [1.21, 1.89];5 trials), endometrial thickness (MD = 0.97, 95%CI [0.43, 1.51]; 12 trials), and endometrial pattern (RR = 1.41, 95%CI [1.13, 1.75]; 7 trials). A similar finding was reported from a retrospective research of 923 Chinese women with recurrent implantation failure(15), of definite increase in biochemical pregnancy rates (p = 0.08) and clinical pregnancy rates (p = 0.049) than non-acupuncture, meanwhile, a potentially higher live birth rates (p = 0.160) observed in the acupuncture group after propensity score-matched (PSM) logistic analysis; unfortunately, the developmental stage of the transplanted embryo was not explicitly stated in the text. Another single-blind, prospective, randomized controlled trial among the infertile women undergoing FET (blastocyst) in hCG-induced natural cycles(16) indicated the TEAS (transcutaneous electrical acupuncture point stimulation) group was associated with significant higher risks in embryo implantation, clinical pregnancy and live birth than the mock TEAS group (p-values of 0.024, 0.038 and 0.033, respectively). However, the presence of divergent outcomes across studies regarding acupuncture and FET has led to a complex landscape in the interpretation of acupuncture's efficacy. A double-blind randomized controlled trial of 226 subfertile patients undergoing FET (cleavage)(17) found comparable clinical pregnancy and live birth rates in patients who had a single session (25 minutes) of real or placebo acupuncture, which administrated immediately after embryo transfer procedure. And a meta-analysis of 20 trials with 5130 women that compared acupuncture with sham acupuncture controls or no adjuvant treatment(18) suggested acupuncture may be effective when compared to no adjunctive treatment, with increased clinical pregnancies and live births, but was not an efficacious treatment when compared with sham controls.

As a multitude of clinical studies emerged, a growing body of evidence indicated that the frequency of acupoint treatment sessions increased, their role in facilitating favorable pregnancy outcomes became significantly pronounced. Recently, a systematic review and meta-analysis conducted among 38 RCTs involving 5,991 participants(19) that found timing and dosage of acupuncture were crucial factors affecting pregnancy outcomes in IVF-ET. Acupuncture was associated with improved pregnancy outcomes, i.e. CPR (RR = 1.71, 95% CI:1.36–2.16, p < 0.00001) and LBR (RR = 2.40, 95% CI:1.20–4.79, p = 0.01) in frozen cycles. Improvements in CPR were observed across all dosage groups, but only the high-dosage group showed a significant increase in LBR (RR = 1.75, 95% CI:1.05–2.92, p = 0.03). A meta-analysis of 14 RCTs involving 1564 infertile women(20)found out the dose-related efficacy of acupuncture. For the moderate or high-dosage group, CPR and part of ER (endometrial receptivity) parameters (i.e. endometrial thickness, endometrial pattern, and resistance index) were significantly improved in the acupuncture group (i.e., CPR: OR = 2.00, 95% CI [1.24, 3.22], p = 0.004, I² = 0% in one menstrual cycle; OR = 2.49, 95% CI [1.67, 3.72], p < 0.05, I² = 0% in three menstrual cycles) when the dosage of acupuncture was restricted. However, the low-dosage group did not show significant improvements.

Our current study, through 4863 cycles observational summary, examines the impact of different courses of AMP-EWH treatment at various stages of endometrium preparation for single frozen-thawed blastocyst transfer in women aged < 43 years. When compared to non-acupoint stimulation (group T0), participants who underwent AMP-EWH were categorized into a single group T ≥ 1 exhibited a significantly higher CPR (OR 1.181, [95%CI,1.035 to 1.349], p = 0.014, in Model crude;1.157, [1.006 to 1.331], 0.041, Model 2; 1.199, [1.008 to 1.427], 0.041, Model PSM). Additionally, its efficacy most benefits from group T ≥ 4, administrating AMP-EWH for at least 4 sessions during the entire endometrium preparation process until 2 hours prior to blastocyst transfer procedure. Compared to group T0, the advantages of group T ≥ 4 were highlighted in terms of both HCG positivity rates (HCGPR) and clinical pregnancy rates (CPR) irrespective of which model employed for data analyses, HCGPR was of ORs 1.331,(95%CI, 1.033 to 1.716), p = 0.027,in Model crude; 1.329, (1.021 to 1.729), 0.034, Model 1; 1.342,(1.030 to 1.750), 0.030, Model2; and 1.561(1.107 to 2.200), 0.011, Model PSM; and CPR of ORs 1.387,(95%CI, 1.100 to 1.748), p = 0.006, in Model crude; 1.380, (1.084 to 1.758), 0.009, Model 1; 1.394 (1.093 to 1.778), 0.008, Model 2; and 1.672 (1.219 to 2.295), 0.001, Model PSM. To determine whether acupoint stimulation at varying timing and intensities had differential effects on patients' pregnancy outcomes of frozen-thawed blastocyst transfer, further pairwise comparisons were conducted among the participants who had completed acupoint stimulation across a range of session frequencies. When compared to groups T1 and T2-3, group T ≥ 4 had a higher likelihood of achieving HCG positivity and clinical pregnancy, although some of these comparative results did not reach statistical significance. Interestingly, the pregnancy outcomes did not uniformly improve with a mere increase in the number of acupuncture sessions. For instance, when comparing group T2-3 with group T1, there was a noted trend towards decreased rates of HCG positivity and clinical pregnancy in group T2-3, suggesting that the timing of acupoint stimulation may also exert a crucial influence on the pregnancy outcomes following FET (single blastocyst). This finding is consistent with the results of a systematic review and meta-analysis mentioned above(19), and further supported by a RCT(21) that reported the full-course (high-dose) acupuncture group demonstrated a marked improvement in biochemical pregnancy rate and clinical pregnancy rate in IVF-ET cycles, compared to the non-acupuncture cohort. However, the comparative analysis does not reveal substantial differences in these outcomes between high-dose (receiving acupuncture treatment once a week, a total of 5 times before the FET and on the day of transplantation) and low-dose arms (receiving acupuncture treatment exclusively on the day of embryo transplantation, 1 h before and 30 min after the embryo transfer).

We conducted stratified subgroup analysis utilizing a GLMM-Model 2 for multifactorial binary logistic regression of clinical pregnancy rates (CPR) to identify AMP-EWH effect modifiers and refine efficacy estimates across various subgroups. In subgroup comparison to group T0, group T ≥ 4 showed superior outcomes of CPR in patients ≤ 37 years old with a BMI < 24.0 kg/m², 0–1 previous failed embryo transfer attempts, an endometrium thickness of 7.0–13.0 mm on progesterone administration day, embryo quality graded as good, and D5 blastocyst transferred. Conversely, group T1 had a significantly enhanced CPR in patients with a BMI ≥ 24.0 kg/m² and two prior ET failures. Exclusively, within the cohort of patients with three or more unsuccessful attempts of ET, group T2-3 highlighted a significant advantage. These findings underscored the potential for personalized treatment strategies based on patient-specific characteristics in FET (single blastocyst).

During the FET cycle, the selection of acupoints varies across different study protocols. Among them, certain acupoints are most frequently utilized, such as the following combinations of acupoint selection strategies were employed. BL23(Shenshu bilateral), BL32(Ciliao bilateral), and KI3(Taixi bilateral)(15); CV3(Zhongji), CV4(Guanyuan), SP6(Sanyinjiao) and Zigong bilaterally(16); ST36(Zusanli), SP6 (Sanyinjiao), SP10 (Xuehai) and LI4 (Hegu)(17); Zigong, CV4(Guanyuan), LR3(Taichong), and LI4(Hegu)(21); CV4(Guanyuan), zigong, ST36(zusanli), SP6(Sanyinjiao), BL23(Shenshu), and BL32(Ciliao)(22); CV12(Zhongwan), ST25(Tianshu), CV4(Guanyuan), CV3(Zhongji), Zigong, ST34(Liangqiu), ST36(Zusanli), ST37(Shangjuxu), and ST39(Xiajuxu)(23). Indentified from a systematic review and meta-analysis(20), SP6(Sanyinjiao), Zigong, RE4(Shenmai), ST36(Zusanli), and LR3(Taichong) were the top five acupoints selected in reproductive medicine. In present study, the acupoints selected for stimulation included CV3(Zhongji), CV4(Guanyuan), CV6(Qihai), bilateral Zigong, GV3(Yaoyangguan), GV4(Mingmen), BL23(Shenshu), and Balao. By applying medicinal patches and adjunct electromagnetic wave heat to these acupoints, a comprehensive regulation of the body's ‘qi’ and blood, as well as the balance of yin and yang, can be achieved. This promotes uterine blood circulation and improves the endometrial growth environment, thereby enhancing the endometrial receptivity and creating more favorable conditions for embryo implantation. However, the selection of acupoints for each patient should be tailored according to individual differences and medical conditions, and should be appropriately adjusted by an experienced TCM practitioner.

A critical evaluation of the existing literature revealed a substantial disparity in study outcomes(24). The discrepancies were attributed to differences in study populations, patient demographics, variations in acupuncture stimulation modalities, the development stage of transferred embryos (D3/D5/D6), endometrial preparation protocols, acupoints selection, treatment timing, stimulation dosage, study sample size, and inconsistencies in research methodologies. Given these reasons, the replicability of the findings across studies is questionable, leading to divergent, and even contradictory, outcomes. The identified disparities highlight a pressing necessity for a unified approach to acupuncture research in FET. Standardization of patient selection criteria, treatment protocols, and outcome measures are imperative to ensure the reliability and validity of future studies. Adhering to such standards will facilitate a more coherent understanding of acupuncture's role and potential benefits in reproductive medicine.

There is a robust consensus on the safety profile of acupuncture therapy, while opinions on the correlation between acupuncture and pregnancy outcomes in IVF-ET vary. The safety of acupuncture is corroborated by an overview from systematic reviews particularly without relation to miscarriage risks(25), further substantiated by a study that has scrutinized the effects of acupuncture administered during pregnancy, intrapartum and postpartum(26, 27), and additionally confirmed by a retrospective chart review(28) that performed on the 212 on-going pregnancy women once received real or placebo acupuncture on the day of embryo transfer, no statistically significant difference was found in maternal adverse outcomes (gestational diabetes and hypertensive disorder), the preterm delivery rate, the mode of delivery, and baby Apgar scores and birthweight. And no evidence was found that acupuncture poses any risk to the mother or unborn fetus(29). In our study, the incidence rates of biochemical pregnancy loss, ectopic pregnancy and adverse effects associated with acupoint stimulation were statistically assessed to evaluate the safety profile of AMP-EWH. All the three groups that received acupoint intervention (groups T1, T2-3, and T ≥ 4) demonstrated a trend towards reduced biochemical pregnancy loss rates and slightly increased ectopic pregnancy rates, however, these trends did not achieve statistical significance when compared to the control group T0. Particularly, the occurrence of ectopic pregnancies was minimal, leading to a broad 95% confidence interval for the odds ratios (ORs) in the intergroup comparisons. The side effects of AMP-EWH were characterized by mild localized rashes with an incidence rate of ≤ 2.1%, and there were no reported cases of severe allergic reactions or infections. Patients who experienced mild localized rashes showed marked improvement following symptomatic treatment, without interruption of acupoint stimulation or the subsequent FET process.

Despite the growing interest in the complementary role of acupuncture in IVF-ET, there is a dearth of research elucidating the underlying mechanisms by which acupuncture influences various aspects of the procedure. The scarcity of studies investigating the physiological impact of acupuncture on the different stages of IVF-ET, from ovulation induction to embryo implantation, limits our comprehensive understanding of its therapeutic potential. Research in this area is critical, as it could uncover the biological pathways through which acupuncture may modulate endocrine function(30), target the uterus(31), improve endometrium receptivity(32) to prepare for embryo transfer, improve blood flow to the uterus(33–35), reduce stress and anxiety(36), and potentially create a more favorable uterine environment for embryo implantation. The process of implantation requires a reciprocal interaction between blastocyst and endometrium, culminating in a small window of opportunity during which implantation can occur(37). This interaction involves the embryo, with its inherent molecular program of cell growth and differentiation, and the temporal differentiation of endometrial cells to attain uterine receptivity. Implantation itself is governed by an array of endocrine, paracrine and autocrine modulators, of embryonic and maternal origin(38). Implantation failure is thought to occur as a consequence of impairment of embryo developmental potential and/or impairment of uterine receptivity and the embryo-uterine dialogue.

When to FET, new techniques for monitoring changes in the endometrium at the level of cell growth(39, 40), gene regulation(41) and protein expression(42) may lead to the identification of better markers for implantation. Therefore, that would provide insights into the possible mechanisms by which acupoint stimulation may induce changes in the endometrium. To date, the intricate biological mechanisms that directly govern endometrial proliferation and differentiation, and by extension, implantation, remain largely enigmatic(43).

The impact of acupuncture on endometrial changes and its subsequent effects on embryo implantation are even more obscure. Current investigations into the modulatory effects of acupuncture on endometrial receptivity mainly rely on non-invasive sonographic assessments(23, 44), except a few endometrial tissue samples subjected to miRNA-seq analysis(45), assessment of HOXA10 expression(16), and whole transcriptome sequencing(22) to elucidate the molecular underpinnings of the endometrial response. In this study, group T ≥ 4 demonstrated a significant advantage over group T0 in both HCG positivity rate and clinical pregnancy rate, and with an trend of improvement relative to groups T1 and T2-3. This may be attributed to the continuous and cumulative acupoint stimulations for group T ≥ 4 patients throughout the proliferative and secretory phases of endometrial preparation with HRT, as well as on the day of embryo transfer, thereby enhancing endometrial receptivity during the window of implantation and positively influencing pregnancy outcomes. Regrettably, this study did not include mechanistic research. To further substantiate the effects of acupoint stimulation, future studies could consider conducting pragmatic randomized controlled trials (RCTs) with larger sample sizes. A comparison between a group without acupoint treatment and a group receiving continuous acupoint treatment throughout the FET cycle could be made, with live birth rate as the primary outcome measure. Additionally, ultrasound assessment of endometrial receptivity indices should be measured.

Strengths

As far as we know, it is the first time to evaluate the real-world effect of AMP-EWH on pregnancy outcomes in women following single frozen-thawed blastocyst transfer.

The study benefits from a robust sample size of 4,863 cycles, conducting a real-world clinical practice, offering insights that are directly applicable to patient care outside of controlled trial environments. The high follow-up rate for the source population's CPR ensures the study outcomes. The division of the study population into distinct groups based on different acupoint treatment protocols (T0, T ≥ 1, T1, T2-3, T ≥ 4) allows for a nuanced understanding of treatment effects across various levels of intervention. The use of Generalized Linear Mixed Models (GLMM) helps to control for a range of confounding variables, providing a sophisticated approach to adjusting for potential confounders, which has managed 3 adjusted model, with various covariates serving as fixed effects and embryo culture incubators and embryo transfer physicians accounting for random effects, Notably, incorporating propensity score matching in the model PSM prior to multifactorial adjustment by GLMM, enhances the study's ability to balance the treatment groups for observed covariates, reducing selection bias and improving the validity of the comparisons. Initially, the study's approach to analyzing different acupoint stimulation protocols and identifying the most effective one contributes to the evidence base for personalized treatment strategies. Further subgroup analysis helps to pinpoint the target patient groups that may benefit the most from different acupoint stimulation and offers new insights into opportunities for physicians to intervene.

Limitations

As a retrospective study, our research faces the typical constraints of its design, including possible biases in data collection and the potential for selection bias in participant recruitment. Despite a robust follow-up rate, concerns about data quality and completeness remain, with the possibility of missing data affecting outcomes. Although we employed GLMM and PSM, unmeasured confounding factors could still confound our results. Causality cannot be definitively established due to the observational nature of our study, and caution is warranted in interpreting the temporality and biological plausibility of our findings. The generalizability of our results may be limited by the specific patient population and treatment settings. While the study attempts to control for differences between incubators and physicians as random effects, the extent to which these factors influence outcomes may not be fully captured. Additionally, multiple subgroup analyses are potentially subject to bias owing to the limited sample size of group T ≥ 4, which can affect the robustness and generalizability of the results.

Acupoint stimulation via the application of acupoint medication patches combined with electromagnetic wave heat therapy (AMP-EWH) exerts beneficial influences on clinical pregnancy outcomes following frozen-thawed blastocyst transfer. Notably, the protocol incorporating 4 or more sessions of AMP-EWH may confer a notably enhanced effect in improving clinical pregnancy rates. Current evidence underscores the real-world clinical value of AMP-EWH as an innovative complementary approach for both clinicians and patients engaged in the process of frozen-thawed blastocyst transfer. Nonetheless, considering the current studies' methodological limitations and publication bias, a measured and prudent endorsement of AMP-EWH for broader clinical application is advisable. Further high-quality, large-scale, randomized, multicenter, double-blinded trials should be conducted to substantiate the efficacy of AMP-EWH in frozen-thawed blastocyst transfer, to explore its optimal dosages and timing for treatment, affirm its safety profile, and to elucidate the underlying mechanism of those observed benefits.

ART: Assisted Reproductive Technology; FET: Frozen–Thawed Embryo Transfer; IVF-ET: In Vitro Fertilization-Embryo Transfer; ICSI: Intracytoplasmic Sperm Injection; PGT: Preimplantation Genetic Testing; TCM: Traditional Chinese Medicine; TEAS: Transcutaneous Electrical Acupoint Stimulation; AMP-EWH: Acupoint Medication Patches in Conjunction with Electromagnetic Wave Heat Therapy; BMI: Body Mass Index; b-FSH: Basal-Follicle Stimulate Hormone; b-LH: Basal-Luteal Hormone; b-E2: Basal-Estradiol; AMH: Anti-Müllerian Hormone; HCG: Human Chorionic Gonadotropin; CPR: Clinical Pregnancy Rate; VIF: Variance Inflation Factor; GLMM: Generalized Linear Mixed Models; PSM: Propensity Score-Matched; SMD: Standard Mean Difference (statistical effect size); CI: Confidence Interval; OR: Odd ratio.

Ethics approval and consent to participate

The present study was reviewed and preapproved by the Review of Ethics Committee in Clinical Research of the First Affiliated Hospital of Wenzhou Medical University (protocol number, KY2023-R202). An informed consent form was signed by all participants at the beginning of the study.

Consent for publication

Not applicable.

Availability of data and materials

No new data were generated or analysed in support of this research. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors had no conflicts of interest to declare.

Funding

This work was supported by Basic Scientific Research project of Wenzhou science and Technology Bureau (Y20220006 to Yili Teng), and Wenzhou Key Laboratory of Reproduction and Genetics (2022HZSY0051).

Authors’ contributions

Yili Teng and Shiyi Lin contributed to the conception and design of the study. Huan Zhang provided methodological support. Yili Teng, Peipei Pan, Haiyan Yang, Xuefeng Huang, Xia Chen and Shiyi Lin contributed equally to the acquisition of data, drafting of the article and critical review of the final draft. All the authors have approved the final version for publication.

Acknowledgements

Not applicable.

D. Wei et al., Frozen versus fresh single blastocyst transfer in ovulatory women: a multicentre, randomised controlled trial. Lancet 393, 1310-1318 (2019).
J. Kwak-Kim et al., The role of immunologic tests for subfertility in the clinical environment. Fertil Steril 117, 1132-1143 (2022).
F. Parisi et al., The pathophysiological role of estrogens in the initial stages of pregnancy: molecular mechanisms and clinical implications for pregnancy outcome from the periconceptional period to end of the first trimester. Hum Reprod Update 29, 699-720 (2023).
H. Qin, J. Feng, X. Wu, Effects and mechanisms of acupuncture on women related health. Front Med 18, 46-67 (2024).
L. Yu, Q. Xia, Z. Sun, J. Song, Efficacy of Acupoint Application on In Vitro Fertilization Outcome in Patients with Polycystic Ovary Syndrome: A UHPLC-MS-Based Metabolomic Study. Evid Based Complement Alternat Med 2022, 9568417 (2022).
P. Humaidan, E. Stener-Victorin, Pain relief during oocyte retrieval with a new short duration electro-acupuncture technique--an alternative to conventional analgesic methods. Hum Reprod 19, 1367-1372 (2004).
A. S. Dehghani, K. Homayouni, Z. Kanannejad, Z. Kanannejad, The effect of acupuncture on the day of embryo transfer on the in vitro fertilization outcomes: An RCT. Int J Reprod Biomed 18, 209-214 (2020).
J. Feng, Q. Li, X. Zhang, S. Zhao, T. Guo, Acupuncture Treatment for Threatened Abortion. Med Acupunct 35, 43-47 (2023).
J. M. Wu et al., [Adjuvant therapy of acupoint catgut embedding for threatened abortion after in vitro fertilization-embryo transfer: a randomized controlled trial]. Zhongguo Zhen Jiu 39, 689-693 (2019).
M. Xu, M. Zhu, C. Zheng, Effects of acupuncture on pregnancy outcomes in women undergoing in vitro fertilization: an updated systematic review and meta-analysis. Arch Gynecol Obstet 309, 775-788 (2024).
C. Zhu et al., Effects of acupuncture on the pregnancy outcomes of frozen-thawed embryo transfer: A systematic review and meta-analysis. Front Public Health 10, 987276 (2022).
L. Y. Chen Jiaojie, Zou Jianqin, Huang Ye, Luo Juanzhen, Eﬀect analysis of heat-sensitive moxibustion combined with acupoint application in the treatment of infertility with follicular dysplasia under the guidance of infrared thermal tomography technology. Electronic Journal of Practical Gynecological Endocrinology 36, 3 (2022).
T. L. Cheng Xiaoman, Wang Fen, Cheng Zixia, Nie Ting, Curative Effect Observation of Antai Formula Combined with Acupoint Application on Early Threatened Abortion. Shaanxi Journal of Traditional Chinese Medicine 44, 4 (2023).
E. Y. Wan Qingzhi, Wang Bo, A Meta-analysis of treating hyperemesis gravidarum by acupoint application. Clinical Journal of Chinese Medicine 12, 5 (2020).
D. Gan et al., Acupuncture benefits to women with recurrent implantation failure: A propensity score-matched cohort study. Heliyon 9, e18193 (2023).
Z. Shuai, F. Lian, P. Li, W. Yang, Effect of transcutaneous electrical acupuncture point stimulation on endometrial receptivity in women undergoing frozen-thawed embryo transfer: a single-blind prospective randomised controlled trial. Acupunct Med 33, 9-15 (2015).
E. W. So, E. H. Ng, Y. Y. Wong, W. S. Yeung, P. C. Ho, Acupuncture for frozen-thawed embryo transfer cycles: a double-blind randomized controlled trial. Reprod Biomed Online 20, 814-821 (2010).
C. A. Smith et al., Acupuncture performed around the time of embryo transfer: a systematic review and meta-analysis. Reprod Biomed Online 38, 364-379 (2019).
X. Wang et al., The Timing and Dose Effect of Acupuncture on Pregnancy Outcomes for Infertile Women Undergoing In Vitro Fertilization and Embryo Transfer: A Systematic Review and Meta-Analysis. J Integr Complement Med, (2024).
X. Zheng et al., The Dose-Related Efficacy of Acupuncture on Endometrial Receptivity in Infertile Women: A Systematic Review and Meta-Analysis. Front Public Health 10, 858587 (2022).
Y. R. Wang, S. Z. Wang, K. Robert, J. P. Zhao, [Effect of acupuncture intervention at different time points on the pregnancy rate of frozen embryo transfer]. Zhongguo Zhen Jiu 42, 987-990 (2022).
J. Cheng et al., Whole Transcriptome Sequencing Reveals How Acupuncture and Moxibustion Increase Pregnancy Rate in Patients Undergoing In Vitro Fertilization-Embryo Transplantation. Biomed Res Int 2019, 4179617 (2019).
G. Luo et al., [The effect of warm needling pretreatment for endometrial receptivity of frozen embryo transfer]. Zhongguo Zhen Jiu 37, 831-835 (2017).
L. E. Hullender Rubin, B. J. Anderson, L. B. Craig, Acupuncture and in vitro fertilisation research: current and future directions. Acupunct Med 36, 117-122 (2018).
X. Wang et al., An Overview of Systematic Reviews of Acupuncture for Infertile Women Undergoing in vitro Fertilization and Embryo Transfer. Front Public Health 9, 651811 (2021).
C. Savona-Ventura, T. Mahmood, The role of traditional complementary physical interventions in obstetrics - A scientific review commissioned by the European board and college of obstetrics and gynaecology (EBCOG). Eur J Obstet Gynecol Reprod Biol 279, 84-87 (2022).
J. Xu, I. Z. MacKenzie, The current use of acupuncture during pregnancy and childbirth. Curr Opin Obstet Gynecol 24, 65-71 (2012).
M. T. Y. Seto, K. W. Cheung, T. K. Lo, E. H. Y. Ng, Pregnancy outcomes of women randomized to receive real versus placebo acupuncture on the day of fresh or frozen-thawed embryo transfer. Eur J Obstet Gynecol Reprod Biol 218, 119-122 (2017).
C. M. Witt et al., Safety of acupuncture: results of a prospective observational study with 229,230 patients and introduction of a medical information and consent form. Forsch Komplementmed 16, 91-97 (2009).
J. Feng et al., The Efficacy of Complementary and Alternative Medicine in the Treatment of Female Infertility. Evid Based Complement Alternat Med 2021, 6634309 (2021).
W. Djaali, K. Abdurrohim, D. R. Helianthi, Management of Acupuncture as Adjuvant Therapy for In Vitro Fertilization. Med Acupunct 31, 361-365 (2019).
H. X. Dong et al., Effect of Acupuncture on Endometrial Blood Flow in Women Undergoing in vitro Fertilization Embryo Transfer: A Single Blind, Randomized Controlled Trial. Chin J Integr Med 30, 10-17 (2024).
X. M. You et al., [Clinical study on the effect of acupuncture on pregnancy outcome in patients with prethrombotic infertility after repeated implantation failure]. Zhen Ci Yan Jiu 48, 939-945 (2023).
W. W. Su, J. S. Tian, X. A. Gao, [Effect of warming acupuncture on uterine blood perfusion in patients with failed high-quality freeze-thawed embryo transfer]. Zhongguo Zhen Jiu 40, 498-502 (2020).
Y. Y. Lai et al., Immediate Effect of Electro-acupuncture on Endometrial Blood Flow in Patients with Recurrent Implantation Failure: A Randomized Controlled Trial. Chin J Integr Med 30, 291-298 (2024).
P. G. Guven, Y. Cayir, B. Borekci, Effectiveness of acupuncture on pregnancy success rates for women undergoing in vitro fertilization: A randomized controlled trial. Taiwan J Obstet Gynecol 59, 282-286 (2020).
A. Owusu-Akyaw, K. Krishnamoorthy, L. T. Goldsmith, S. S. Morelli, The role of mesenchymal-epithelial transition in endometrial function. Hum Reprod Update 25, 114-133 (2019).
S. Ojosnegros, A. Seriola, A. L. Godeau, A. Veiga, Embryo implantation in the laboratory: an update on current techniques. Hum Reprod Update 27, 501-530 (2021).
T. M. Rawlings et al., Modelling the impact of decidual senescence on embryo implantation in human endometrial assembloids. Elife 10, (2021).
J. J. Kim, Preparing for implantation. Elife 10, (2021).
S. Grewal, J. G. Carver, A. J. Ridley, H. J. Mardon, Implantation of the human embryo requires Rac1-dependent endometrial stromal cell migration. Proc Natl Acad Sci U S A 105, 16189-16194 (2008).
A. Luddi et al., Emerging role of embryo secretome in the paracrine communication at the implantation site: a proof of concept. Fertil Steril 115, 1054-1062 (2021).
B. Oldak et al., Complete human day 14 post-implantation embryo models from naive ES cells. Nature 622, 562-573 (2023).
E. Stener-Victorin, U. Waldenström, S. A. Andersson, M. Wikland, Reduction of blood flow impedance in the uterine arteries of infertile women with electro-acupuncture. Hum Reprod 11, 1314-1317 (1996).
Y. Mu et al., Integrated miRNA-seq analysis reveals the molecular mechanism underlying the effect of acupuncture on endometrial receptivity in patients undergoing fertilization: embryo transplantation. 3 Biotech 10, 6 (2020).

No competing interests reported.

Download PDF

Editor invited by journal
16 Aug, 2024
Editor assigned by journal
17 Jul, 2024
Submission checks completed at journal
17 Jul, 2024
First submitted to journal
02 Jul, 2024

You are reading this latest preprint version

Real-world Efficacy of Acupoint Stimulation on Clinical Pregnancies Among Women Following Frozen-Thawed Single Blastocyst Transfer: a retrospective cohort study of 4863 cycles

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Sample selection and ethical approval

Study design

Inclusion criteria and exclusion criteria

Outcome measurements

Endometrial preparation protocols and FET processes

Acupoint stimulation protocols

Statistical analysis

Results

Patient and cycle characteristics

Outcome parameters of clinical intervention prior to embryo transfer

Four Reproductive outcome endpoints in entire population

Subgroup-Specific Findings of Clinical Pregnancy Rates

Stratified by female age at embryos transfer

Stratified by BMI Categories

Stratified by number of previous failed ET cycles

Stratified by endometrium thickness on the day of progesterone administration

Stratified by anti-Müllerian hormone (AMH) level

Stratified by stage of embryo development

Stratified by embryo quality assessment

Discussion

Strengths

Limitations

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1