Abstract：
OBJECTIVE: Finding typical patterns - phenotypes - of sleep behaviors characterizing parasomnias in different age and sex groups.
METHODS: We analyzed YouTube videos on sleep-related behaviors likely representing parasomnias. We applied the search terms \"sleepwalking\", \"somnambulism\", \"sleep eating\" \"sleep sex\", \"sleep talking\" and \"aggression in sleep\" in six languages. We classified those persons shown on the videos into estimated biological sex and age (child, adult, elderly) groups. We scored the activity types by a self-made scale and applied binary logistic regression to analyze the association between sleep behaviors versus sex and age groups by the STATA package, providing a 95 % confidence interval and the probability of statistical significance.
RESULTS: 224 videos (102 women; 68 children, 116 adults, and 40 elderly people) were scored. Elderly people had significantly (P < 0.012) less odds of ambulation in sleep likely consistent with somnambulism compared to adults and children. Adult females performed complex manual activities during sleepwalking more often, than males (P < 0.012). Elderly males had 40-fold odds compared to adults and children, to perform aggressive movements and 70-fold odds of complex movements in bed, compared to adults. Elderly people presented emotional behaviors less frequently than adults (P < 0.004), and females showed them twice as often as males. Adults sleep-talked full sentences more often than children and elderly people (P < 0.001).
CONCLUSIONS: Our results support the existence of age- and sex-specific parasomnia phenotypes, denoting possible safety measures. The remarkably low odds of sleepwalking in the elderly highlight the possibility of different pathomechanisms in higher age groups compared to children.
METHODS: We analyzed YouTube videos on sleep-related behaviors likely representing parasomnias. We applied the search terms \"sleepwalking\", \"somnambulism\", \"sleep eating\" \"sleep sex\", \"sleep talking\" and \"aggression in sleep\" in six languages. We classified those persons shown on the videos into estimated biological sex and age (child, adult, elderly) groups. We scored the activity types by a self-made scale and applied binary logistic regression to analyze the association between sleep behaviors versus sex and age groups by the STATA package, providing a 95 % confidence interval and the probability of statistical significance.
RESULTS: 224 videos (102 women; 68 children, 116 adults, and 40 elderly people) were scored. Elderly people had significantly (P < 0.012) less odds of ambulation in sleep likely consistent with somnambulism compared to adults and children. Adult females performed complex manual activities during sleepwalking more often, than males (P < 0.012). Elderly males had 40-fold odds compared to adults and children, to perform aggressive movements and 70-fold odds of complex movements in bed, compared to adults. Elderly people presented emotional behaviors less frequently than adults (P < 0.004), and females showed them twice as often as males. Adults sleep-talked full sentences more often than children and elderly people (P < 0.001).
CONCLUSIONS: Our results support the existence of age- and sex-specific parasomnia phenotypes, denoting possible safety measures. The remarkably low odds of sleepwalking in the elderly highlight the possibility of different pathomechanisms in higher age groups compared to children.
摘要：
目的：在不同年龄和性别群体中寻找表征失眠症的典型睡眠行为模式-表型。
方法：我们分析了YouTube视频中可能代表失眠症的睡眠相关行为。我们应用了搜索词“梦游”，“梦游症”，\"睡眠吃\"\"睡眠性\",六种语言的“睡眠说话”和“睡眠中的侵略”。我们将视频中显示的人分为估计的生物性别和年龄(儿童，成人,老年人)团体。我们通过自制的量表对活动类型进行评分，并应用二元逻辑回归分析STATA软件包的睡眠行为与性别和年龄组之间的关系，提供95%的置信区间和统计显著性的概率。
结果：224个视频（102名妇女；68名儿童，116名成人,和40名老年人)得分。与成人和儿童相比，老年人在睡眠中行走的几率显著降低（P<0.012），可能与梦游症相符。成年女性在梦游过程中更频繁地进行复杂的手动活动，高于男性（P<0.012）。与成人和儿童相比，老年男性有40倍的几率，在床上进行激进的运动和70倍的复杂运动，与成年人相比。老年人表现情绪行为的频率低于成年人(P<0.004),女性展示它们的频率是男性的两倍。成年人睡觉时通满句子的频率高于儿童和老年人（P<0.001）。
结论：我们的结果支持存在年龄和性别特异性的失眠症表型，表示可能的安全措施。老年人梦游的几率非常低，这凸显了与儿童相比，高年龄段人群存在不同病理机制的可能性。
方法：我们分析了YouTube视频中可能代表失眠症的睡眠相关行为。我们应用了搜索词“梦游”，“梦游症”，\"睡眠吃\"\"睡眠性\",六种语言的“睡眠说话”和“睡眠中的侵略”。我们将视频中显示的人分为估计的生物性别和年龄(儿童，成人,老年人)团体。我们通过自制的量表对活动类型进行评分，并应用二元逻辑回归分析STATA软件包的睡眠行为与性别和年龄组之间的关系，提供95%的置信区间和统计显著性的概率。
结果：224个视频（102名妇女；68名儿童，116名成人,和40名老年人)得分。与成人和儿童相比，老年人在睡眠中行走的几率显著降低（P<0.012），可能与梦游症相符。成年女性在梦游过程中更频繁地进行复杂的手动活动，高于男性（P<0.012）。与成人和儿童相比，老年男性有40倍的几率，在床上进行激进的运动和70倍的复杂运动，与成年人相比。老年人表现情绪行为的频率低于成年人(P<0.004),女性展示它们的频率是男性的两倍。成年人睡觉时通满句子的频率高于儿童和老年人（P<0.001）。
结论：我们的结果支持存在年龄和性别特异性的失眠症表型，表示可能的安全措施。老年人梦游的几率非常低，这凸显了与儿童相比，高年龄段人群存在不同病理机制的可能性。
