Abstract
Background
Fatigue affects about half of stroke survivors, but its natural history and associations are uncertain. We aimed to determine the trajectory of fatigue between 6 and 12 months after stroke, and factors associated with changes in fatigue between 6 and 12 months after stroke, in patients recruited to three large clinical trials.
Methods
We performed three large randomised, placebo-controlled trials of fluoxetine 20mg daily for six months, for stroke recovery. Short Form 36 (SF-36) vitality scores measured post-stroke fatigue at 6 and 12 months. We undertook an individual patient data meta-analysis (IPDM) to explore the proportion of patients with fatigue at 6 months (SF-36 vitality score ≤45) who moved to a ‘no fatigue’ category at 12 months, the proportion with ‘no fatigue’ at 6 months who moved to a fatigue category, and the proportion experiencing a clinically relevant improvement in fatigue (increase in SF36 vitality of 10 points) from 6 to 12 months. We performed regression analyses to identify which baseline factors and whether the mood (self-assessed combined anxiety/depression) and pain components of EuroQol 5D-5L (EQ 5D-5L) at 6 months influenced fatigue trajectory between 6 and 12 months.
Results
A total of 5907 patients were included. Fatigue (SF-36 vitality scores) was recorded at 6 and 12 months in 5025 (85%) patients. Fatigue (SF-36 vitality ≤45) at 6 months was present in 1489 (29.6%, 95%CI: 28.4 to 30.9%) participants, of whom 568 (38.1%, 95%CI: 35.7% to 40.7%) had no fatigue at 12 months. Fatigue at 12 months was present in 1474 (29.3%, 28.0 to 30.6%) participants, of whom 553 (15.6%, 95% CI 14.5% to 16.9%) had no fatigue at 6 months.
Patient consent (rather than proxy consent), being able to lift arms off bed at baseline, and severe/extreme pain at 6 months were independently associated with persistence of fatigue through to 12 months, whereas stroke type (embolic or small vessel disease vs haemorrhagic) at baseline and self-assessed slight anxiety/depression at 6 months were associated with fatigue resolution by 12 months.
Among those without fatigue at 6 months, Asians (compared to whites), and those reporting moderate or severe anxiety/depression problems were less likely to have fatigue at 12 months.
Conclusion
More than a third of patients with fatigue at 6 months experienced a clinically relevant reduction in fatigue between 6 and 12 months, whereas one sixth of patients without fatigue at 6 months developed fatigue at 12 months. Although the individual regression analyses identified statistically significant associations between some independent variables and fatigue at 12 months, there was no consistent pattern across the analyses. Future research should identify factors associated with the development and recovery of fatigue after stroke.