{"paper_id":"cdb2449b-0f10-408c-b490-54d47e4420fb","body_text":"The acceptance of assisted reproductive technology (ART) treatments has grown among\ninfertile couples attempting to achieve pregnancy ( Jie  et al ., 2015 ). In the US, the use of\nintracytoplasmic sperm injection (ICSI) in fresh IVF cycles grew from 36.4% in 1996\nto 76.2% in 2012 ( Boulet  et al .,\n2015 ). The International Committee for Monitoring Assisted Reproductive\nTechnologies in Europe reported that ICSI is used in 65% of ART cycles ( Dyer  et al ., 2016 ). ICSI\npatients usually undergo a higher number of IVF cycles, which could mean that ICSI\nis offered to patients with poorer prognoses ( Tannus\n et al ., 2017 ).\nPregnancies arising from ICSI have inherent risks not observed in pregnancies from\nnormal conception. Patients offered ICSI are generally older than individuals\nattempting pregnancy through normal conception, and this difference may increase the\nrisk of genetic disorders in the embryonic karyotype ( Pendina  et al ., 2014 ). In addition, ICSI\nrequires ovarian stimulation, which may increase the risk of multiple pregnancy,\nectopic pregnancy, fetal growth restriction, and miscarriage ( Zhu  et al ., 2016 ;  Jackson  et al ., 2015 ;  Egbe  et al ., 2016 ;  Johnston  et al ., 2015 ;  Perkins  et al ., 2015 ). The incidence of early\npregnancy loss in ICSI patients undergoing fresh embryo transfers amounted to 14.9%\n vs.  26.2% in individuals submitted to frozen embryo transfers\n( Xu  et al ., 2017 ). Early\nmiscarriages in patients given ICSI have decreased the success rate of ICSI. Several\nfactors associated with spontaneous abortion in patients offered ICSI have been\ndiscussed. Most studies showed an association between maternal age and the BMI\naffecting oocyte or endometrial quality ( Hahn\n et al ., 2014 ;  Metwally  et al ., 2010 ;  Hourvitz  et al ., 2006 ). Other predictors have been\nrelated to the quality of the embryo, sperm, and IVF laboratory technology ( Brandes  et al ., 2011 ;  Xu  et al ., 2017 ). A few studies\nhave also shown that fertility treatment itself may be a risk factor for early\nmiscarriage. The risk of miscarriage was reportedly higher following frozen embryo\ntransfers when compared to cases of spontaneous pregnancy ( Xu  et al ., 2017 ).\nAccording to previous studies, and considering the emotional and psychological burden\nimposed by miscarriage on infertile women offered ICSI, awareness of the risk\nfactors associated with ICSI is of great help. Therefore, these findings may be used\nto evaluate assisted reproductive treatments and counsel infertile women suffering\nwith miscarriage. This study aimed to determine the prevalence and the risk factors\nlinked to miscarriage in women treated with ICSI.\n\nThis cross-sectional study included all infertile pregnant women treated with ICSI at\nthe Fatemezahra Infertility and Reproductive Center within a period of five years\n(October 1, 2010 to September 30, 2015). The study protocol received approval from\nthe Ethics Committee of the Bobal University of Medical Science (No. 697). All\nparticipants gave informed consent before joining the study.\nThe following inclusion criteria were adopted: patients on ICSI cycles aged 20-43\nyears receiving fresh embryo transfers. The included patients underwent biochemical\nand hormonal testing, in addition to vaginal ultrasound, hysterosalpingography\n(HSG), and semen analysis. HSG was performed before the patients were included in\nthe study. Patients with structural anomalies underwent hysteroscopy to have the\ndefects removed. Individuals with a history of recurrent miscarriage and uterine\nfibroids, patients with non-removable structural uterine anomalies, and women given\nfrozen embryo transfers were excluded. Pregnant patients treated with the routine\nprotocol in effect at our center (long protocol) were enrolled in the study.\nPatients whose charts missed information on pregnancy outcome were contacted by\nphone and interviewed to capture the missing data point.\nPregnancy was defined as follows: ß-HCG> 25u 16 days after the embryo\ntransfer; patients with low ß-HCG levels on day 16 were tested again two days\nlater. After pregnancy was confirmed, the patients were followed for ß-HCG\nlevels and vaginal ultrasound examination was performed to detect a fetal heartbeat.\nConsidering the patients included in this study, the risk factors collected and\nidentified through questionnaires were maternal age, paternal age, the body mass\nindex (BMI), polycystic ovary syndrome (PCOS), infertility causes, response to\novulation induction treatment including the number of oocytes, type of treatment\ntype, hormone levels including follicle-stimulating hormone (FSH), luteinizing\nhormone (LH), thyroid-stimulating hormone (TSH), prolactin (PRL), on the third day\nof the menstrual cycle.\nThe patients were first categorized as follows based on age and BMI:\n-Age: ≤30 years / 30.1-35 years / 35.1-40 years / >40 years\n-BMI: <18.5 (skinny), 18.5-25 (normal), 25.1-30 (overweight), 30.1-35 (obese),\n>35.1 (very obese)\nThey were subsequently divided into two groups: patients whose pregnancies ended in\nmiscarriage and patients whose pregnancies produced live births. The chi-square\ntest, Fisher's exact test and the t-test were used in statistical analysis. A\n p <0.05 was considered significant. All data were treated as\nconfidential and used solely for the purposes of the study. No interventions were\ncarried out during the study so as to minimize patient risk. Permission to access\npatient data was granted by the managers and staff in charge.\n\nThe study included 860 patients, and 161 of them were pregnant. Of the 161 charts\nexamined, 145 met the inclusion criteria. Patients with inaccurate or missing\ndocuments were also excluded. In the end, 120 cases were included in the study.\nThe pregnancy success rate was 18.72% (161 pregnancies in 860 patients), while the\nmiscarriage rate was 20% (24 miscarriages in 120 cases). Of the 96 live births, 79\n(82.3%) were from singleton pregnancies - 45 males (37.5%) and 34 females (35.4%);\n15 (15.6%) were from twin pregnancies; and two (1.2%) were from triplet\npregnancies.\nThe miscarriage rate was 40% in the galactorrhea group and 18% in the\nnon-galactorrhea group. The difference between the two groups was statistically\nsignificant ( p =0.04). However, prolactin levels did not show a\nsignificant relationship in either of the two groups ( Table 1 ). The calculated effect size was -0.3.\nComparison of hormone levels in women following ICSI\nThe mean age of the pregnant women enrolled in the study was 32.8±5.1, which\nwas within the range of 20 to 43, while the mean age of their partners was\n32.8±6.0. The mean duration of infertility was 5.26±3.8 years, which\nwas within the range of one to 19 years. No association was found between maternal\nage and miscarriage rate in the age groups ( p =0.69). In the other\nage groups - individuals aged >35 and ≤35 years, with respective\nmiscarriage rates of 27.3% and 19.3% - the difference was not significant\n( p =0.69). The calculated effect size showed that the chances of\nsuccess decreased by 8% among individuals aged >35 years (Effect size =- 0.1).\nThe age groups of the pregnant women included in the study following ICSI are shown\nin  Table 2 .\nAge groups of pregnant women following ICSI\nThe mean BMI of the miscarriage group was 26±3.7,  versus \n26.1±3.8 of the non-miscarriage group, suggesting the absence of a\nsignificant correlation between the BMI and pregnancy rate ( p =0.9).\n Table 3  shows the stratification of the\nBMI for pregnant women following ICSI.\nBMI groups of pregnant women following ICSI\nMale factor infertility was diagnosed in 31.7% (38 patients) of the cases, while\nfemale factor was observed in 12.5% (15 patients) of the cases. Fifty-five percent\n(66 patients) had both types of infertility, and 8% (one patient) had unexplained\ninfertility. No association was found between cause of infertility and miscarriage\n( p =0.55) ( Table 4 ).\nCauses of infertility in pregnant women following ICSI\nThe mean duration of infertility in the miscarriage group was 6.6±3.8\n vs.  4.9±3.7 years in the non-miscarriage group,\nindicating the existence of a significant relationship between duration of\ninfertility and miscarriage rate ( p =0.05). Primary infertility was\nobserved in 77.5% (93 patients) of the cases and secondary infertility in 22.5% (27\npatients) of the cases. The miscarriage rates in the above groups were 19.4% and\n22.2%, respectively. Statistically, there was no significant relationship between\nthe type of infertility and miscarriage ( p =0.73). The miscarriage\nrate was 23.8% in patients with PCOS  vs.  17.9% in patients without\nPCOS. The miscarriage rate was higher in the group with PCOS, but the difference was\nnot statistically significant ( p =0.44).\nThe number of ampules used in ovulation induction, the day of oocyte retrieval, the\nnumber of eggs, and endometrial thickness were not significantly correlated with\nmiscarriage ( Table 5 ).\nMean values of related parameters seen in women following ICSI\nAs far as having a history of surgery is concerned, 26.7% (32 patients) of the cases\nhad undergone surgery including Cesarean sections, laparoscopy/laparotomy,\nappendectomy, D & C (dilation and curettage), procedures for the removal of\novarian or breast cysts, and endometriomas. No association was found between having\na history of surgery and miscarriage ( p =0.4).\nSixty-two (84.9%) of the 73 patients who underwent hysterosalpingography had normal\nHSG findings. Eleven (15.1%) had abnormal HSG findings and suffered from conditions\nsuch as endometriosis, tubal obstruction, and hydrosalpinx. However, no significant\ncorrelation was found between the two groups in terms of miscarriage rates\n( p =0.3).\n\nAccording to our results, longer duration of infertility and galactorrhea were\nassociated with increased miscarriage rates, although miscarriage patients did not\nhave higher serum prolactin levels. If we consider that the related\n p -value presents only the chance responsible for the observed\ndifference in the women with galactorrhea, the calculated effect size may present an\nassurance that there is an association between galactorrhea and miscarriage rates.\nIn woman with galactorrhea, the probability of having a successful pregnancy\ndecreases by 13.9%. Each 7.2 women with galactorrhea have at least one miscarriage\nwhen compared to women without galactorrhea (number needed to harm).\nGalactorrhea is a relatively common problem. However, it is often missed at\npresentation.  Ugwa  et al .\n(2016)  reported that most of their patients had normal prolactin levels\nregardless of galactorrhea. A possible implication is that routine breast\nexamination might be needed for infertile women throughout infertility treatment to\nallow earlier diagnosis of galactorrhea.\nThe miscarriage rate in our study was 20%. The miscarriage rate reported for patients\nsubmitted to ART was 21% in a study by  Wang\n et al . (2001)  and 18% in a study by  Aflatoonian  et al . (2011) . The\nabove rates are in agreement with our findings. The causes for the high miscarriage\nrates seen in these treatment methods included older maternal age, history of\nmiscarriage, the procedure used for ART, increased genetic problems inherent to\nthese methods, and the causes of infertility, infertility itself, multiple\npregnancies, embryo quality, hyperstimulation, high and low BMI levels, and\nincreased risk of miscarriage.\nAlthough female age and BMI are first line prognostic factors in human reproduction,\nwe did not find significant differences between the age or BMI groups with\nmiscarriage.  Winter  et al. \n(2002)  concluded that there was no significant relationship between age,\nobesity, and the miscarriage rate, while  Dai et al . (2018)  pointed out that women over the\nage of 40 had significantly higher early miscarriages rates (60.6%) than women under\nthe age of 40. Weight gain and obesity had no effect on miscarriage rates in a study\nby  Tian  et al . (2007) .\nHowever,  Moragianni  et al .\n(2012)  concluded that ART in patients with a BMI ≥ 30 yielded\nsignificantly lower implantation and clinical pregnancy rates. In our study, the\nnon-significant association can be attributed to the lower maternal age and BMI.\nMost of the patients in our study were younger than 30 (74.2%), and 50% of the women\nhad a normal BMI within the range of 18.5-25, and 40% were obese (BMI 25.1-30). This\nfinding may have been affected by the limited size of our sample. Our results,\ntherefore, must be interpreted with caution. Calculating the effect size revealed a\nweak relationship between age >35 and miscarriage rate, and that the chances of\nhaving a successful pregnancy dropped by 8% in the group aged >35 years. We also\ncalculated the effect size for the association between BMI >30 and miscarriage,\nand found the two were poorly associated and the chances of having a successful\npregnancy dropped by 72.3% in women with a BMI >30. Further studies with more\nfocus on age subgroups and their associations with miscarriage are needed. A marked\nassociation was found between duration of infertility and rate of miscarriage. This\nfinding may be attributed to possible increases in age or BMI among couples with the\nlonger duration of infertility.\nThe miscarriage rate was higher in the PCOS group, but the difference was not\nstatistically significant. Although  Beydoun\n et al . (2009)  reached the same conclusion,  Kamalanathan  et al . (2013)  and\n Luo  et al.  (2017)  found\nthat the rate of miscarriage in PCOS patients was higher than in non-PCOS patients,\nand that the elevated prevalence of miscarriage was related to a high prevalence of\nobesity in individuals with PCOS. In our study, more than half of the participants\nhad a normal BMI, which to some extent precluded the addition of the effects of a\nhigher BMI and having PCOS.\nThe results showed that there was no significant correlation between rate of\nmiscarriage and cause of infertility.  Bahceci &\nUlug (2004)  found no significant association between cause of infertility\nand rate of early miscarriage. This is not consistent with the results described by\n Hajishafiha  et al .\n(2011) , in a study that found a significant relationship between the\ncauses of infertility and the rate of miscarriage.\n\nIncreasing galactorrhea and the duration of infertility may increase the risk of\nmiscarriage. Therefore, women require early care and counseling before pregnancy\nalong with meticulous care during the pregnancy. In doing so, it might be possible\nto minimize the adverse effects of pregnancy for patients.","source_license":"CC-BY-4.0","license_restricted":false}