A comprehensive meta-analysis of exogenous estrogen, progesterone, and testosterone in animal models of ischemic and hemorrhagic stroke.
Exogenous sex hormones have been extensively studied for their influence on stroke risk and outcome. This meta-analysis served to update the pre-clinical acute ischemic stroke (AIS) literature and provide the first synthesis of the intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH) literature on how estrogen, progesterone, and testosterone affect post-stroke outcomes.
This study was pre-registered with PROSPERO (CRD42024544794). Medline, EMBASE, Scopus, and Web of Science were searched; studies using animal models of stroke investigating exogenous estrogen, progesterone, or testosterone, alone or in combination, compared to non-treated controls were included. Assessments of injury volume, edema, and behaviour (neurological deficits, sensorimotor and cognitive outcomes) were analyzed via hierarchical meta-analyses. Risk of bias was assessed via SYRCLE and CAMARADES, and evidence certainty via an adaptation of the GRADE tool.
In total, 211 studies were included. Estrogen and progesterone improved all post-AIS outcomes (SMDs = 0.32-1.30, 95% CIs [0.02, 2.07], very low to moderate certainty of evidence), whereas testosterone had mostly null effects (very low to moderate certainty). Fewer studies investigated hemorrhagic stroke, with null effects of estrogen (very low to low certainty) and conflicting results of progesterone (SMDs = 0.15-1.16 [-2.20, 2.58], very low to moderate certainty) in ICH, as well as benefit of progesterone in SAH (SMD = 2.63 [0.98, 4.30], very low certainty). Uncertainty in our evidence arose from low scientific and translational rigor. Sex and gonadal status were consistent moderators of these effects, and gonadal depletion length (i.e., the 'timing hypothesis') was a significant moderator of estrogen's effect on post-AIS injury volume.
Estrogen and progesterone are promising cerebroprotectants for AIS. Further focussed and rigorous pre-clinical research on remaining research gaps (e.g., dosage parameters) are needed to guide clinical investigations and maximize the likelihood of translational success. The impact of testosterone and sex hormones in hemorrhagic stroke remain inconclusive due to lack of research.
This study was pre-registered with PROSPERO (CRD42024544794). Medline, EMBASE, Scopus, and Web of Science were searched; studies using animal models of stroke investigating exogenous estrogen, progesterone, or testosterone, alone or in combination, compared to non-treated controls were included. Assessments of injury volume, edema, and behaviour (neurological deficits, sensorimotor and cognitive outcomes) were analyzed via hierarchical meta-analyses. Risk of bias was assessed via SYRCLE and CAMARADES, and evidence certainty via an adaptation of the GRADE tool.
In total, 211 studies were included. Estrogen and progesterone improved all post-AIS outcomes (SMDs = 0.32-1.30, 95% CIs [0.02, 2.07], very low to moderate certainty of evidence), whereas testosterone had mostly null effects (very low to moderate certainty). Fewer studies investigated hemorrhagic stroke, with null effects of estrogen (very low to low certainty) and conflicting results of progesterone (SMDs = 0.15-1.16 [-2.20, 2.58], very low to moderate certainty) in ICH, as well as benefit of progesterone in SAH (SMD = 2.63 [0.98, 4.30], very low certainty). Uncertainty in our evidence arose from low scientific and translational rigor. Sex and gonadal status were consistent moderators of these effects, and gonadal depletion length (i.e., the 'timing hypothesis') was a significant moderator of estrogen's effect on post-AIS injury volume.
Estrogen and progesterone are promising cerebroprotectants for AIS. Further focussed and rigorous pre-clinical research on remaining research gaps (e.g., dosage parameters) are needed to guide clinical investigations and maximize the likelihood of translational success. The impact of testosterone and sex hormones in hemorrhagic stroke remain inconclusive due to lack of research.
Authors
Kung Kung, Suerte Suerte, Khiabani Khiabani, Parranto Parranto, Gannon Arnott Gannon Arnott, Kalisvaart Kalisvaart, Nakagawa Nakagawa, Klahr Klahr, Colbourne Colbourne
View on Pubmed