An individual-based model for predicting the prevalence of depression

Major depressive disorder (depression) is a common psychiatric illness. Here, a discrete-time individual-based model is proposed to predict the time evolution of the percentage of people suffering from depression. A normalized index I-i is introduced to reflect the psychological health condition of the ith individual: low values of I-i correspond to mentally healthy; high values, to depressive state. Changes on I-i are driven by rules that depend on the psychiatric histories and socio-demographic features of the individuals, on the risk factors affecting them, and on the recovery rate. Computational simulations were performed by using official data from Brazil and Germany in the latest years. Despite the prevalence in women being higher, the model fits the data only if women are more cognitively resilient to depression compared to men; that is, when exposed to the same risk factors, the value of I-i for women is lower than the value of I-i for men.