Anxiety, depression and health-related quality of life in patients with thoracic aortic disease: a longitudinal study in a cardiothoracic outpatient clinic

Figures

• 

  Figure 1. STROBE-compliant flowchart of selection and exclusions.

• 

  Figure 2. Age distribution (n = 111) showing range 32–85 years with patient counts per age group.

• 

  Figure 3. Sex distribution (68 male, 43 female).

• 

  Figure 4. Summarizes surgical status at initia/first presentation (pre = preoperative, post = postoperative, excluded: missing core data, excluded: withdrawal of permission.

• 

  Figure 5. Distribution of EQ-5D-5L index values using the German tariff. The distribution is left-skewed with peak values around 0.9 and tail extending to −0.5, indicating most patients reported moderate to good health states (0.55-0.9).

• 

  Figure 6. Histogram of EQ-VAS (0–100 scale) scores. Distribution of subjective health. EQ-5D-5L index (0–1, higher = better HRQoL); EQ-VAS (0–100, higher = better). EQ-5D-5L: left-skewed (peak ∼0.9, tail to −0.5).

• 

  Figure 7. Distribution of PHQ-4 total scores (0–12). The histogram exhibits a right-skewed pattern with dominant peak at 0–1, steadily decreasing frequencies toward higher scores, and relatively few observations in the 6–12 range, confirming most scores fall within normal-to-mild distress while a minority reaches moderate-to-severe levels.

• 

  Figure 8. Boxplots of primary outcomes across 256 visits (EQ-5D-5L index left-skewed median ∼0.7; EQ-VAS symmetric median 70 IQR 60–80; PHQ-4 right-skewed median ∼3 tail to 12). Substantial inter-individual variability evident, particularly in PHQ-4 distress scores.

• 

  Figure 9. (a) individual trajectories of EQ-5D-5L index values across outpatient visits among patients with at least two assessments. Each line represents one patient. Most trajectories show stable or improving HRQoL over successive visits, with a minority exhibiting persistently low or declining scores, illustrating marked inter-individual heterogeneity. (b) Spaghetti plot/line plot with Individual trajectories of EQ-VAS scores (0–100) across outpatient visits among patients with at least two assessments. Each line represents one patient and illustrates substantial inter-individual variability, with most trajectories showing stable or improving self-rated health over time and a minority remaining in lower score ranges. (c) Spaghetti plot/line plot with Individual trajectories of PHQ-4 total scores (0–12) across outpatient visits among patients with at least two assessments. Each line represents one patient, revealing marked inter-individual heterogeneity: most trajectories show early peaks followed by decline over successive visits, while a minority exhibits persistent high distress or fluctuating patterns.

• 

  Figure 10. Forest plot of linear mixed-effects model β-coefficients (95% CI) for EQ-VAS (0–100). Model: EQ-VAS∼Time_since_surgery_days + Sex (0 = male ref, 1 = female) + Op_status (0 = postop ref, 1 = preop)+(1|Patient_ID). Positive β indicates improved subjective health; negative β indicates deterioration. Time since surgery and postoperative status are related to higher EQ-VAS scores, suggesting recovery over time, whereas female sex and preoperative status show trends toward lower perceived health. Clinically, this underscores the importance of postoperative psychosocial follow-up.

• 

  Figure 11. Forest plot of logistic regression odds ratios (95% CI) for PHQ-4 ≥6. Model predictors: age, sex (male vs. female reference), visit number, surgical status (pre- vs. postoperative reference). OR >1 indicates increased odds of clinically relevant distress; OR <1indicates reduced odds. Preoperative status (vs. postoperative) shows OR≈2.0 with CI excluding 1. Male sex shows protective trend (OR <1). Cluster-robust standard errors account for within-patient correlation.

• 

  Figure 12. Forest plot of mixed-effects model β-coefficients (95% CI) for EQ-5D-5L index. Positive β indicates better HRQoL. Time since surgery shows positive effect; preoperative status (vs. postoperative reference) and female sex (vs. male reference) show negative associations.