Marianne Thoresen

Rodent models of neonatal hypoxic–ischemic (HI) injury require a subset of animals to be immobilized for continuous temperature monitoring during the insult and subsequent treatment. Restrained animals are discarded from the analysis due to the effect of restraint on the brain injury as first demonstrated by Thoresen et al 1996. However, the effects of restraint on responses to hypothermic (HT) post‐insult therapy are not well described. We examine the effects of restraint associated with different probe placements on HI brain injury. We have conducted a meta‐analysis of 23 experiments comparing probe rats (skin n = 42, rectal n = 35) and free‐moving matched non‐probe controls ( n = 80) that underwent HI injury (left common carotid artery ligation and 90 min 8% O 2 ) at postnatal day 7 (P7), followed by 5 h of NT (37°C) or HT (32°C). On P14, brain regions were analyzed for injury (by neuropathology and area loss), microglial reactivity and brain‐derived neurotrophic factor (BDNF). HI injury was mitigated in NT skin and rectal probe rats, with greater neuroprotection among the rectal probe rats. Following HT, the skin probe rats maintained the restraint‐associated neuroprotection, while brain injury was significantly exacerbated among the rectal probe rats. Microglial reactivity strongly correlated with the acquired injury, with no detectable difference between the groups. Likewise, we observed no differences in BDNF signal intensity. Our findings suggest a biphasic neuroprotection from restraint stress, which becomes detrimental in combination with HT and the presumed discomfort from the rectal probe. This finding is useful in highlighting unforeseen effects of common experimental designs or routine clinical management.