Fatigue Potentially Reduces the Effect of Transcranial Magnetic Stimulation on Depression Following COVID-19 and Its Vaccination

Abstract

COVID-19's long-term effects, known as Long-COVID, present psychiatric and physical challenges in recovered patients. Similarly, rare long-term post-vaccination side effects, resembling Long-COVID, are emerging (called Post-Vaccine). However, effective treatments for both conditions are scarce. Our clinical experience suggests that transcranial magnetic stimulation (TMS) often aids recovery in Long-COVID and Post-Vaccine patients. However, its effectiveness is reduced in patients with severe fatigue. Therefore, we retrospectively analysed Tokyo TMS Clinic's outpatient records (60 in total; mean age, 38 years) to compare Long-COVID and post-vaccine patients' characteristics and symptoms, assess the impact of TMS on their symptoms, and investigate the role of fatigue in depression recovery with TMS. The primary outcome was the regression coefficient of the initial fatigue score on depression score improvement using TMS. Secondary outcomes included psychiatric/physical scores before and after TMS and their improvement rates. We found no differences in the initial symptoms and background factors between Long-COVID and Post-Vaccine patients. After ten TMS sessions, all psychiatric and physical symptom scores improved significantly. TMS improves depression, insomnia, anxiety, and related neuropsychiatric symptoms, which were the primary complaints in this study. Thus, we conclude that TMS improves depression and anxiety. The effectiveness of TMS in treating depression in Long-COVID and Post-Vaccine patients decreased as fatigue severity increased. In conclusion, TMS relieved depressive symptoms following COVID-19 and vaccination; however, fatigue may hinder its effectiveness.

Keywords: fatigue; long-COVID; transcranial magnetic stimulation.

Figure 2

Correlations between each psychiatric/physical-symptom test score. Upper panels: Scatterplot matrix for each normalised psychiatric/physical test score. Values on the vertical and horizontal axes are expressed as percentages. Each figure shows the regression line and 95% confidence interval. Lower tables: The upper rows show the correlation coefficients between each normalised psychiatric/physical test score, and the lower rows show the p-value of the correlation coefficient; p < 0.05 is considered significant. Where the correlation coefficients were significant, we coloured them in three colours according to their magnitude.

Figure 3

Multivariate analysis of covariance between ΔQIDS (%) and each psychiatric/physical-symptom test score at the first visit. (A–C) MANCOVA results for each of the three patterns. The partial regression coefficients of PS_Initial visit (%) are significantly positive in all models. In Model 3 (C), the VIF is below 2. In each right-hand diagram, the horizontal axis shows the predicted value of ΔQIDS (%) obtained by adding the intercept and all values obtained by multiplying each independent variable by its partial regression coefficient. The vertical axis represents the measured value of the ΔQIDS (%). * Significant (p < 0.05).

Figure 1

List of target data and exclusions.