Table 1 Characteristics of the included studies
Study | Country | Design | Sample size | Mean age (years) | Men (%) | Timing of GDF-15 measurement | Methods of GDF-15 measurement | Methods for determining GDF-15 cutoff | Cutoff of GDF-15 | Follow-up duration (months) | Definition of poor functional outcome | No. of patients with poor functional outcome | Variables adjusted |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Worthmann [14] | Germany | PC | 57 | 74 | 47 | At admission | IRMA | ROC curve analysis | 1630 ng/l | 3 | mRS: 2–6 | 19 | Age and sex |
Groschel [15] | Germany | PC | 264 | 70.3 | 55.3 | At admission | ECLIA | Median | 1191 ng/l | 3 | mRS: 2–6 | 113 | Age, sex, NIHSS at admission, TOAST classification, SBP, DM, CAD, HDL-C, and thrombolysis |
Geng [16] | China | RC | 102 | 71.3 | 58.8 | At admission | RIA | Median | 1580 ng/l | 3 | mRS: 2–6 | 29 | None |
Lu [17] | China | RC | 120 | 67.8 | 60.8 | Within 24 h of stroke onset | ELISA | Median | 1594 ng/l | 3 | mRS: 2–6 | 65 | None |
Yin 2019 | China | Post-hoc analysis | 3066 | 62.3 | 64.1 | Within 24 h of admission | ELISA | UL of healthy individual in previous studies | 1200 ng/l | 3 | mRS: 3–6 | 762 | Age, sex, current smoking, alcohol consumption, and NIHSS at admission |
Breniere [18] | France | PC | 173 | 74.6 | 51 | At admission | ELISA | T3 versus T1-2 | 2088 ng/l | 3 | mRS: 3–6 | 92 | None |
Dong [19] | China | PC | 83 | 62.4 | 63.9 | At admission | ELISA | Median | 864 ng/l | 3 | mRS: 3–6 | 42 | Age, sex, comorbidities, and NIHSS at admission |
Jeong [22] | Korea | RC | 62 | 66.2 | 39 | Within 24 h of admission | ELISA | Median | 493 ng/l | 3 | mRS: 3–6 | 20 | Age, sex, comorbidities, eGFR, and timing of recanalization |
Cai [21] | China | RC | 148 | 58.4 | 62.2 | Within 24 h of admission | ELISA | ROC curve analysis | 826.5 ng/l | 3 | mRS: 3–6 | 61 | None |
Li [23] | China | RC | 156 | 58.4 | 55 | Within 24 h of stroke onset | ELISA | ROC curve analysis | 802.1 ng/l | 3 | mRS: 3–6 | 64 | None |