Abstract / Summary
Background: Post-stroke fatigue is common and disabling, yet rarely distinguished from co-occurring depression, limiting understanding of its neural substrates. We examined resting-state fMRI correlates of post-stroke fatigue while accounting for depressive symptoms. Methods: We studied 103 stroke patients and 29 age-matched controls from a longitudinal neuroimaging cohort. Fatigue was assessed via the SF-36 Vitality subscale and depression via the Geriatric Depression Scale - Short (GDS-15). Linear mixed-effects models evaluated associations between vitality and resting-state functional connectivity, adjusting for depression and covariates. Residualized vitality models served as sensitivity analyses to isolate fatigue variance independent of depressive symptoms, with FDR correction applied within feature families. Results: In depression-adjusted models, higher vitality (lower fatigue) was associated with increased frontoparietal-limbic connectivity (β = 0.0013, FDR q = 0.048), though this did not survive depression residualization. In a residualized sensitivity analysis, higher vitality (lower fatigue) was associated with reduced thalamus-insula and thalamus-ventral attention network connectivity (thalamus-insula: β = −0.0071, FDR q = 0.011; thalamus-VAN: β = −0.0043, FDR q = 0.012); these associations were attenuated and did not survive correction in the primary depression-adjusted model (q ≈ 0.075). Conclusions: Fatigue was associated with both frontoparietal-limbic and thalamocortical connectivity, although the thalamic associations were attenuated after adjustment for depressive symptoms in the primary analysis and were most evident in residualized sensitivity analyses. These findings suggest that fatigue and depressive symptoms may involve both shared and potentially distinct neural correlates. Point estimates were smaller in stroke participants than controls, providing no evidence that the observed associations were stroke-specific. Future work should determine whether these relationships are unique to stroke or reflect broader neural correlates of fatigue across clinical populations.