Depression during pregnancy is a very common problem for women. It affects almost 15% of pregnant women every year. It usually needs more attention as it is very difficult to treat. However, researchers from the University of Pittsburgh conducted a new study to find effective ways to treat this problem. They have revealed a simple survey given during the first trimester through an app can accurately predict which mothers will develop depression after some time. This app is known as MyHealthPregnancy. Â
In this study, researchers involved 944 pregnant women. These women did not have any history of depression. These women were asked some questions about their family background, medical history and their stress levels. Some of them were also asked additional questions about their food insecurity. Researchers then checked them for the intensity of their depression once every trimester. They used 80% of the data for training 6 different machine learning models. On the other hand, they used the remaining 20% to test their model’s accuracy in predicting depression. Â
When researchers observed the collected data carefully, they found that this model was 89% accurate in predicting depression. It only used 14 factors out of 55 possible variables. It used some factors such as anxiety history, their relationship status, stress levels and worries. When they focused on understanding more about a food security, they found that food insecurity plays a very important role in predicting depression. When researchers included this factor, it improved model’s accuracy to 93%. They only needed 9 variables for predicting the risk of depression. Â
One of the reputed doctors said that some factors such as quality of sleep, worries about labor pain and delivery and access to food can increase the risk of depression in the future. Depression during pregnancy can affect a mother’s and baby’s health. So, healthcare professionals should make some effective strategies to reduce this risk. They should use this model as it accurately predicts depression. Â
Reference Link:Â Â
Tamar Krishnamurti et al, Predicting first time depression onset in pregnancy: applying machine learning methods to patient-reported data, Archives of Women’s Mental Health (2024). Â
