More Realistic Forecasting of Future Life Events After Psilocybin for Treatment-Resistant Depression

Abstract

Background: Evidence suggests that classical psychedelics can promote enduring changes in personality, attitudes and optimism, as well as improvements in mental health outcomes. Aim: To investigate the effects of a composite intervention, involving psilocybin, on pessimism biases in patients with treatment-resistant depression (TRD). Methods: Patients with TRD (n = 15) and matched, untreated non-depressed controls (n = 15) performed the Prediction Of Future Life Events (POFLE) task. The POFLE task requires participants to predict the likelihood of certain life events occurring within a 30-day period, after which the actual rate of event occurrence is reported; this gives an index of potential pessimism versus optimism bias. Psilocybin was administered in two oral dosing sessions (10 and 25 mg) one week apart. Main outcome measures were collected at baseline and one week after the second dosing session. Results: Patients showed a significant pessimism bias at baseline [t(14) = -3.260, p = 0.006; 95% CI (-0.16, -0.03), g = 1.1] which was related to the severity of their depressive symptoms (r_s = -0.55, p = 0.017). One week after psilocybin treatment, this bias was significantly decreased [t(14) = -2.714, p = 0.017; 95% CI (-0.21, -0.02), g = 0.7] and depressive symptoms were greatly improved [t(14) = 7.900, p < 0.001; 95% CI (16.17, 28.23), g = 1.9]; moreover, the magnitude of change in both variables was significantly correlated (r = -0.57, p = 0.014). Importantly, post treatment, patients became significantly more accurate at predicting the occurrence of future life events [t(14) = 1.857, p = 0.042; 95% CI (-0.01, 0.12), g = 0.6] whereas no such change was observed in the control subjects. Conclusion: These findings suggest that psilocybin with psychological support might correct pessimism biases in TRD, enabling a more positive and accurate outlook.

Keywords: cognitive bias; forecasting; pessimism; psilocybin; treatment-resistant depression.

FIGURE 1

Schematic diagram of the study design.

FIGURE 2

(A) Depressive symptoms. Patients had significantly higher BDI scores than controls at baseline (U = 0.0E0, p < 0.001; g = 5.0). The patients’ BDI scores were significantly reduced 1 week after psilocybin treatment [t(14) = 7.900, p < 0.001; 95% CI (16.17, 28.23), g = 1.9], but remained significantly greater than the controls (U = 34.50, p = 0.001; g = 1.1). (B) Depressive symptoms over time: Patients. There was a significant reduction in the patients’ BDI scores over time [X²(3, N = 15) = 32.35, p < 0.0001]. Baseline BDI scores did not significantly differ from scores at the beginning of dosing day 1 (p = 0.944), confirming that the depressive symptoms had not naturally declined prior to psilocybin treatment. BDI scores were significantly reduced on dosing day 2 when compared with baseline (p < 0.001) and dosing day 1 (p = 0.033) scores. One-week after both dosing sessions, BDI scores were significantly reduced when compared to baseline (p < 0.001), dosing day 1 (p < 0.001) and dosing day 2 (p = 0.009) scores. Data expressed as mean ± SEM (p < 0.05^∗; p < 0.01^∗∗; p < 0.001^∗∗∗). Note: Time (days) between: (i) baseline and dosing day 1 (M = 35.0, SD = 23.3); (ii) dosing day 1 and dosing day 2 (M = 7.9, SD = 2.6) (iii) dosing day 2 and one-week post follow-up (M = 7.8, SD = 2.5; (iv) baseline and one-week post follow-up (M = 50.8; SD = 25.3).

FIGURE 3

(A) Accuracy. Patients were significantly more accurate in their predictions at the follow-up relative to baseline [t(14) = 1.857, p = 0.042; 95% CI (–0.01, 0.12), g = 0.6]. There were no significant changes in accuracy for the control subjects [t(14) = 0.603, p = 0.556; 95% CI (–0.06, 0.09)]. No between-groups differences in accuracy were found at baseline [t(28) = –0.196, p = 0.423] or the follow-up [t(28) = 0.676, p = 0.252; 95% CI (–0.04, 0.08)]. (B) Cognitive bias. The greater the digression from zero on the y-axis, the greater the bias. Patients had a significant pessimism bias at baseline [t(14) = –3.260, p = 0.006; 95% CI (–0.16, –0.03), g = 1.1]. No significant biases were found for the controls [t(14) = 1.823, p = 0.090; 95% CI (–0.09, 0.01)]. Baseline bias scores were significantly lower (i.e., greater pessimism) for the patients than the controls [t(28) = 3.704, p < 0.001; 95% CI (0.06, 0.22), g = 1.3]. Patients’ pessimism bias was significantly reduced [t(14) = –2.714, p = 0.017] and they no longer displayed any biases following psilocybin treatment [t(14) = 0.768, p = 0.455; 95% CI (–0.06, 0.02)]. No significant biases were found for the controls [t(14) = 0.460, p = 0.653; 95% CI (–0.05, 0.03)] and there was no significant between-groups difference [t(28) = –0.265, p = 0.793; 95% CI (–0.07, 0.05)] at the follow-up. Data expressed as mean ± SEM (p < 0.05^∗; p < 0.01^∗∗; p < 0.001^∗∗∗). ^∗ Only the TRD patients received psilocybin treatment.

FIGURE 4

(A) Relationship between depressive symptoms and bias. At baseline, the patients’ bias scores were moderately and significantly related to their BDI scores (r_s = -0.55, p = 0.017). No significant relationships were found between BDI and bias scores for the control subjects (r_s = -0.22, p = 0.219). (B) Relationship between the change in depressive symptoms and bias. Bias and BDI scores changed significantly more for the patients than the control subjects’ bias [t(28) = -3.167, p = 0.004; 95% CI (-0.24, -0.05), g = 1.2] and BDI scores [t(28) = 7.228, p < 0.001; 95% CI (15.09, 27.04), g = 2.6]. The decrease in BDI score was significantly related to the decrease in pessimism for the patients (r = -0.57, p = 0.014). There was no significant relationship found between the change in BDI and bias scores (r_s = -0.007, p = 0.980) for the control subjects.