Abstract：
OBJECTIVE: The objective of this study is to assess the concordance and its association with sociocultural background of a four-question survey with accelerometry in a multiethnic adolescent population, regarding sleep components. Based on questions from the Pittsburgh Sleep Quality Index and adapted to a school context, the questionnaire focussed on estimating sleep onset time, wake-up time and sleep duration on both weekdays and weekends. This subjective survey was compared with accelerometry data while also considering the influence of sociocultural factors (sex, place of living, ethnic community and socio-economic status).
METHODS: Adolescents aged 10.5-16 years (n = 182) in New Caledonia completed the survey and wore an accelerometer for seven consecutive days. Accelerometry was used to determine sleep onset and wake-up time using validated algorithms. Based on response comparison, Bland-Altman plots provided agreement between subjective answers and objective measures. We categorized participants\' answers to the survey into underestimated, aligned and overestimated categories based on time discrepancies with accelerometry data. Multinomial regressions highlighted the sociocultural factors associated with discrepancies.
RESULTS: Concordance between the accelerometer and self-reported assessments was low particularly during weekends (18%, 26% and 19% aligned for onset sleep time, wake-up time and sleep duration respectively) compared with weekdays (36%, 53% and 31% aligned, respectively). This means that the overall concordance was less than 30%. When considering the sociocultural factors, only place of living was associated with discrepancies in onset sleep time and wake-up time primarily on weekdays. Rural adolescents were more likely to overestimate both onset sleep time (B = -1.97, p < 0.001) and wake-up time (B = -1.69, p = 0.003).
CONCLUSIONS: The study found low concordance between self-assessment and accelerometry outputs for sleep components. This was particularly low for weekend days and for participants living in rural areas. While the adapted four-item questionnaire was useful and easy to complete, caution should be taken when making conclusions about sleep habits based solely on this measurement.
METHODS: Adolescents aged 10.5-16 years (n = 182) in New Caledonia completed the survey and wore an accelerometer for seven consecutive days. Accelerometry was used to determine sleep onset and wake-up time using validated algorithms. Based on response comparison, Bland-Altman plots provided agreement between subjective answers and objective measures. We categorized participants\' answers to the survey into underestimated, aligned and overestimated categories based on time discrepancies with accelerometry data. Multinomial regressions highlighted the sociocultural factors associated with discrepancies.
RESULTS: Concordance between the accelerometer and self-reported assessments was low particularly during weekends (18%, 26% and 19% aligned for onset sleep time, wake-up time and sleep duration respectively) compared with weekdays (36%, 53% and 31% aligned, respectively). This means that the overall concordance was less than 30%. When considering the sociocultural factors, only place of living was associated with discrepancies in onset sleep time and wake-up time primarily on weekdays. Rural adolescents were more likely to overestimate both onset sleep time (B = -1.97, p < 0.001) and wake-up time (B = -1.69, p = 0.003).
CONCLUSIONS: The study found low concordance between self-assessment and accelerometry outputs for sleep components. This was particularly low for weekend days and for participants living in rural areas. While the adapted four-item questionnaire was useful and easy to complete, caution should be taken when making conclusions about sleep habits based solely on this measurement.
摘要：
目的：这项研究的目的是评估多种族青少年人群中使用加速度测量法进行的四个问题调查的一致性及其与社会文化背景的关联。关于睡眠成分。根据匹兹堡睡眠质量指数的问题，并适应学校环境，问卷的重点是估计睡眠开始时间，工作日和周末的起床时间和睡眠时间。这项主观调查与加速度测量数据进行了比较，同时还考虑了社会文化因素的影响(性别，居住的地方,民族社区和社会经济地位)。
方法：新喀里多尼亚10.5-16岁（n=182）的青少年完成了调查，并连续7天佩戴了加速度计。使用经过验证的算法，使用加速法确定睡眠开始和唤醒时间。根据响应比较，Bland-Altman地块在主观答案和客观措施之间达成了共识。我们将参与者对调查的答案归类为被低估，根据与加速数据的时间差异，对齐和高估类别。多项回归强调了与差异相关的社会文化因素。
结果：加速度计和自我报告评估之间的一致性很低，尤其是在周末（18％，26%和19%与开始睡眠时间一致，唤醒时间和睡眠持续时间)与工作日(36%，53%和31%对齐，分别)。这意味着总体一致性小于30%。当考虑到社会文化因素时，只有居住地与起始睡眠时间和唤醒时间的差异有关，主要是在工作日。农村青少年更可能高估睡眠时间（B=-1.97，p<0.001）和唤醒时间（B=-1.69，p=0.003）。
结论：研究发现睡眠成分的自我评估和加速度测量输出之间的一致性较低。对于周末和居住在农村地区的参与者来说，这尤其低。虽然改编的四项问卷很有用，而且很容易完成，在仅根据这种测量得出有关睡眠习惯的结论时，应谨慎行事。
方法：新喀里多尼亚10.5-16岁（n=182）的青少年完成了调查，并连续7天佩戴了加速度计。使用经过验证的算法，使用加速法确定睡眠开始和唤醒时间。根据响应比较，Bland-Altman地块在主观答案和客观措施之间达成了共识。我们将参与者对调查的答案归类为被低估，根据与加速数据的时间差异，对齐和高估类别。多项回归强调了与差异相关的社会文化因素。
结果：加速度计和自我报告评估之间的一致性很低，尤其是在周末（18％，26%和19%与开始睡眠时间一致，唤醒时间和睡眠持续时间)与工作日(36%，53%和31%对齐，分别)。这意味着总体一致性小于30%。当考虑到社会文化因素时，只有居住地与起始睡眠时间和唤醒时间的差异有关，主要是在工作日。农村青少年更可能高估睡眠时间（B=-1.97，p<0.001）和唤醒时间（B=-1.69，p=0.003）。
结论：研究发现睡眠成分的自我评估和加速度测量输出之间的一致性较低。对于周末和居住在农村地区的参与者来说，这尤其低。虽然改编的四项问卷很有用，而且很容易完成，在仅根据这种测量得出有关睡眠习惯的结论时，应谨慎行事。
