Methods and Materials
Study population. Data were analyzed from a single hospital in Shenzhen, a Southern Chinese city, in accordance with the relevant guidelines and regulations as set out by the approving body, the Ethics Committee, Shenzhen Baoan Women’s and Children’s Hospital, Jinan University (IRB:LLSC 2018-08-01). All the study participants donated a 5ml venous blood after gave a written informed consent. All participants’ names and other identifying information were removed in all sections of this manuscript.
From September to November 2018, we recruited at random 72 fertile women undergoing pre-pregnancy physical examinations at our hospital, where the rate of response was >95%. The subjects were between the ages of 20 and 40 years old, and were all local Han Chinese and long-term residents in Shenzhen, China. With the exception of four persons who suffered from uterine fibroids, recurrent miscarriage, oligomenorrhea and cervicitis, 68 healthy women were included in this study. After obtaining the informed consent, information of the subjects such as age, height, weight, AAM, menstrual cycle and period, and reproductive history were collected by trained nurses through a questionnaire survey, and 5ml venous blood was collected for flow cytometry analyses. Indicators including peripheral blood red blood cell (RBC), white blood cell (WBC) and hemoglobin (Hb) were obtained from the results of physical examination.
Follow-up of pregnancy status. The pregnancy status of the subjects were followed-up by interview until December 2019, which was verified by serum human chorionic gonadotropin (HCG) and ultrasound, and the first day of last menstrual period (LMP) was recorded. Among of these subjects, there were 35 pregnancies. Each pregnant woman was scheduled to come back to our hospital for antepartum examinations according to our hospital routine methods. Moreover, each pregnancy early signs of miscarriage was timely detected and treated by our hospital, and 7 ones were maintained to birth (defined as threatened miscarriage); However another 8 ones were aborted (defined as spontaneous miscarriage, with the exception of chromosome abnormality, congenital deformity of the fetuses, and other secondary diseases of the gravidas and fetuses). The information on weight gain at early pregnancy, thyroid stimulating hormone (TSH), free T4 (FT4) and thyroperoxidase antibodies (TPO-ab) at early pregnancy, and blood pressure, fast glucose, triglyceride (TG), cholesterol (CHOL), and HbA1c during mid-trimester pregnancy was collected from medical records of antepartum examinations.
Flow cytometry analyses. Pre-pregnancy immune functions were tested according to our routine methods(21). Briefly, peripheral white cells were isolated from whole blood at pre-pregnancy physical examinations by adding FACS LYSING SOLUTION (BD Biosciences, USA). Then the precipitates were collected by centrifugation. After washed twice by phosphate-buffered saline (PBS, sigma, USA), and cells were labelled according to our published paper(21). Antibodies for APC-H7-conjugated anti-CD3; PE-Cy7-conjugated anti-CD4 and anti-CD28; FITC-conjugated anti-CD45RA; PE-conjugated anti-CD25; BV421-conjugated anti-CD56, anti-CD127; BV510-conjugated anti-CD8; Alexa Fluor 647-conjugated anti-CCR7 and anti-CXCR5; Alexa Fluor 488-conjugated anti-CD183, and BV421-conjugated anti- CD194 were purchased from BD Biosciences.
Samples were run on a BD LSR Fortessa (BD Biosciences) at Shuangzhi Purui Medical Laboratory Co., Ltd. (China), and the data was analyzed using FlowJo 10.1 software (Tree Star Inc, USA).
Data Collection of public GWAS studies. The published GWAS data for AAM and Tfh cells were collected from UK Biobank (http://www.nealelab.is/blog/2017/9/11/details-and-considerations-of-the-uk-biobank-gwas ) and TwinsUK(22), respectively.
Statistical analysis and public GWAS data analysis. The SPSS 25.0 statistical software (SAS Institute Inc., Chicago, IL, USA) was used to process and analyze the data. Data were presented as n , proportion, median and min/max. We first performed association analyses between immune functions and AAM by using Spearman correlation analyses. Manny-Whitney U tests were then used to compare immune cells between healthy pregnancies and threatened or spontaneous miscarriages. Spearman correlation analyses also were used to analyze the associations between Tfh cells and the various biological indexes during pregnancy. Lastly, we used the Kaplan-Meier estimate and Cox regression analyses to study the associations between Tfh cells and miscarriages. The relative risk (RR) and its 95% confidence interval (CI) were calculated. All tests were two sided, and the level of significance was set at 0.05. Statistical figures were produced by using GraphPad Prism8.0 (GraphPad Software, USA).
To test the genetic association of Tfh cells and AAM, MR-Base software (http://www.mrbase.org/ ) were used for two sample MR test according to the method as described(23). Briefly, inverse-variance weighted approach, weighted median method, MR-Egger method, and single mode and weighted mode method were used for MR tests as a web application. Then we performed a Causal direction test to measure MR causal estimates. Lastly, a leave-one-out analysis was also performed to assess whether the MR estimate is driven or biased by a single SNP.