Fritz Ann-Kristina, Amrein Irmgard, Wolfer David P
Institute of Anatomy, University of Zurich, Zurich, Switzerland.
Institute of Human Movement Sciences and Sport, ETH Zurich, Zurich, Switzerland.
Am J Med Genet C Semin Med Genet. 2017 Sep;175(3):380-391. doi: 10.1002/ajmg.c.31565. Epub 2017 Jun 27.
Although most nervous system diseases affect women and men differentially, most behavioral studies using mouse models do not include subjects of both sexes. Many researchers worry that data of female mice may be unreliable due to the estrous cycle. Here, we retrospectively evaluated sex effects on coefficient of variation (CV) in 5,311 mice which had performed the same place navigation protocol in the water-maze and in 4,554 mice tested in the same open field arena. Confidence intervals for Cohen's d as measure of effect size were computed and tested for equivalence with 0.2 as equivalence margin. Despite the large sample size, only few behavioral parameters showed a significant sex effect on CV. Confidence intervals of effect size indicated that CV was either equivalent or showed a small sex difference at most, accounting for less than 2% of total group to group variation of CV. While female mice were potentially slightly more variable in water-maze acquisition and in the open field, males tended to perform less reliably in the water-maze probe trial. In addition to evaluating variability, we also directly compared mean performance of female and male mice and found them to be equivalent in both water-maze place navigation and open field exploration. Our data confirm and extend other large scale studies in demonstrating that including female mice in experiments does not cause a relevant increase of data variability. Our results make a strong case for including mice of both sexes whenever open field or water-maze are used in preclinical research.
尽管大多数神经系统疾病对女性和男性的影响存在差异，但大多数使用小鼠模型的行为学研究并未纳入两性的实验对象。许多研究人员担心，由于发情周期的存在，雌性小鼠的数据可能不可靠。在此，我们回顾性评估了在水迷宫中执行相同位置导航实验的5311只小鼠以及在同一旷场实验区域中测试的4554只小鼠的性别对变异系数（CV）的影响。计算了作为效应大小度量的科恩d值的置信区间，并以0.2作为等效边际检验其等效性。尽管样本量很大，但只有少数行为参数对CV显示出显著的性别效应。效应大小的置信区间表明，CV要么等效，要么至多显示出很小的性别差异，占CV组间总变异的不到2%。虽然雌性小鼠在水迷宫习得和旷场实验中可能略有更大的变异性，但雄性小鼠在水迷宫探针实验中的表现往往不太可靠。除了评估变异性，我们还直接比较了雌性和雄性小鼠的平均表现，发现它们在水迷宫位置导航和旷场探索中表现相当。我们的数据证实并扩展了其他大规模研究，表明在实验中纳入雌性小鼠不会导致数据变异性的显著增加。我们的结果有力地支持了在临床前研究中使用旷场或水迷宫时纳入两性小鼠的做法。
