Abstract
BACKGROUND: Assisted reproductive techniques (ART) are an effective solution for infertility treatment. Miscarriage is a common and distressing complication, the incidence of which is much higher in couples undergoing ART than in natural pregnancies. This study aimed to identify and classify risk factors associated with miscarriage in couples undergoing ART. METHODS: In this systematic review study, a search was conducted in the databases including PubMed, Scopus, Web of Science, Ovid, BSCO Host, IEEE, Embase, Proquest, Cochrane Library, between January 2014 and February 2025, based on the PRISMA 2020 guidelines. Data from the articles included in the study were collected using a structured data collection form and then analyzed descriptively. RESULTS: A total of 17 studies were included in the study according to the eligibility criteria. Risk factors were classified into three main categories: baseline characteristics (demographics, lifestyle, and medical history), clinical characteristics (hormonal profiles, uterine abnormalities), and treatment characteristics (fetal quality, stimulation protocols). The study showed that parental age, high or low BMI, previous miscarriages, and unhealthy lifestyle habits (such as smoking, alcohol consumption, stress) were significant risk factors. From a clinical perspective, hormonal imbalance (such as Abnormal Follicle-Stimulating (FSH), Anti-Mullerian Hormone (AMH), and Thyroid Stimulating Hormone (TSH) levels), thin endometrium, and poor ovarian reserve were associated with an increased risk of miscarriage. Treatment characteristics such as embryo transfer type and protocol, ovarian stimulation protocols, embryo grading, and freeze-thaw cycles were effective in predicting miscarriage. CONCLUSION: Miscarriage is influenced by various and diverse factors. Understanding the predictive risk factors enables physicians to provide targeted counseling and preventive interventions. Screening, increased education, lifestyle modification, and ART treatment programs can also increase reproductive success. In the future, designing predictive models using artificial intelligence (machine and deep learning) could help improve decision-making and predict miscarriage. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12884-026-08819-6.