Aim:
Objective: To investigate the association between anxiety and infertility-related variables and to evaluate the performance of a random forest model in predicting and classifying infertility type using clinical and psychological features.

Material and methods:
Methods: A cross-sectional study was conducted among 268 infertile women systematically recruited from the infertility center at Safeer Al-Imam Al-Hussain Surgical Hospital in Karbala, Iraq. Data were collected using standardized questionnaires covering sociodemographic characteristics, medical, obstetric, and infertility histories. Anxiety severity was assessed using the Generalized Anxiety Disorder-7 scale. Descriptive statistics and linear regression analyses were performed. A random forest model was developed using an 80/20 training-testing split to predict infertility type based on anxiety severity and related clinical variables.

Results:
Results: Anxiety was highly prevalent, affecting 83.9% of participants. Mild anxiety was observed in 35.1%, while moderate anxiety was reported in 31.3%. Linear regression showed that infertility duration and infertility type were significantly associated with anxiety severity. Longer infertility duration was associated with higher anxiety scores. The random forest model demonstrated high classification accuracy and low log loss in predicting infertility type. Secondary infertility was the most common type and was associated with longer infertility duration and higher anxiety severity.

Conclusions:
Conclusions: Anxiety is strongly associated with infertility, particularly infertility duration and type. Combining conventional statistical analysis with machine learning may improve understanding of the clinical and psychological complexity of infertility and support more individualized patient care.