HADS Depression Clause Samples

HADS Depression. M18-Trajectory model‌ Trajectory clustering is applied to imputed dataset. For depression, the model that best describes the data is a quadratic one with four latent classes (Fig. I1) (lowest BIC). Figure I2 shows the resulting trajectories for the model and the actual measurements of the patients that compose each class. The largest trajectory group accounted for 56% of the patients (labeled ‘Low’) and was composed of patients who generally had relatively low levels of depression throughout the one-year period. The trajectory labeled ‘High’, estimated to account for 27% of the patients, reported initial moderate/high depression score that remained moderate/high over the one year. The group labeled ‘Decreasing’, accounting for 11% of the patients, began with a moderate/high depression score that improved over time. The final group, labeled ‘Increasing’ and accounting for 6% of the population, started with an initial low score but their score varied substantially during the observation period and are characterized by an increasing mean trajectory. Also shown in Figure I2 are 95% confidence intervals around each trajectory. The fact that the confidence intervals are tight around each trajectory indicates the adequacy of the model.